article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
studies of lateral interactions between adsorbed particles are motivated by the need to understand the origin of the wide variety of ordered overlayers and phase transitions at fractional adsorbate coverage on metal surfaces .
these interactions have contributions ranging from short - range and van der waals to long - range dipole - dipole , and lattice - mediated interactions @xcite .
hard - square short - range interactions and dipole - dipole long - range interactions are the major contributions to the lateral interactions for bromine adsorption on ag(100 ) @xcite . in this paper
, we explore the validity and applicability of such a model for adsorption of chlorine on ag(100 ) .
halide electrosorption on single - crystal metal electrode surfaces is a good model system for studying the properties of the electrode - electrolyte interface in an electrochemical cell . due to its relative simplicity
, it can be used to distinguish between the various contributions of different parameters according to the effect of their inclusion on the overall behavior of the system .
a mean - field approach is not sufficient , even for the description of one of the simplest halide - electrosorption systems br / ag(100 ) .
however , a simple lattice - gas model with constant parameters is sufficient to describe its equilibrium @xcite , and dynamic @xcite properties .
while the electrosorption of br on single - crystal ag(100 ) from aqueous solution has been extensively studied as an example of adlayer formation in an electrochemical system @xcite , less attention has been given to the electrosorption of cl @xcite on ag(100 ) .
a lattice - gas model with constant parameters is not sufficient to describe cl / ag(100 ) , therefore this system can be used to further investigate the nature and characteristics of the lateral interactions between the adsorbed halide atoms . in particular , we here develop a self - consistent picture of variable lattice - gas parameters based on the resident charge on the adatoms being coverage dependent or electrode - potential dependent ( through the coverage ) .
the rest of this paper is organized as follows . in section [ sec : ii ]
we describe an electrostatic model of the adlayer that is used in the simulations , the lateral interaction energies , and the monte carlo methods used .
a brief description of the experimental procedure is given in section [ sec : exp ] .
the results of fitting the simulations to experimental data are detailed in section [ sec : results ] , followed by a brief comparison with br / ag(100 ) in section [ sec : compare ] .
our conclusions are summarized in section [ sec : conclusions ] .
the adsorption of cl ions occurs at the fourfold hollow sites of the ag(100 ) surface @xcite , which form a square lattice as shown in fig .
[ fig : model ] . to approximate the equilibrium behavior of this system
, we use a lattice - gas model , in which the lattice sites correspond to the adsorption sites .
mitchell _ et al . _
@xcite used an off - lattice model for the br / ag(100 ) system to show that the br adsorbates spend most of the time near the four - fold hollow sites of the ag(100 ) surface , thus justifying the lattice - gas treatment of halide adsorption . to describe the energy associated with a configuration of adsorbates on the surface , a grand - canonical effective hamiltonian @xcite is used , @xmath2 where @xmath3 is a sum over all pairs of sites , @xmath4 are the lateral interaction energies between particles on the @xmath5th and @xmath6th lattice sites , measured in mev / pair , @xmath7 is the electrochemical potential , measured in mev / particle , and @xmath8 is the total number of lattice sites .
the local occupation variable @xmath9 is @xmath10 if site @xmath5 is occupied and @xmath11 otherwise .
the long - range interactions , @xmath12 , depend on the distance , @xmath13 , between ions @xmath5 and @xmath6 ( measured in ag(100 ) lattice spacing units , @xmath14 @xcite ) as @xmath15 where the infinite value for @xmath16 indicates nearest - neighbor exclusion , and negative values of @xmath12 denote long - range repulsion .
the coverage isotherms were simulated using a square @xmath17 lattice with periodic boundary conditions to reduce finite - size effects .
the electrochemical potential @xmath18 is related to the bulk ionic concentration @xmath19 and the electrode potential @xmath20 ( measured in mv ) . in the dilute - solution approximation ,
the relationship is @xmath21 where @xmath22 is an arbitrary constant , @xmath23 is a reference concentration ( here taken to be @xmath24 ) , and @xmath25 is the elementary charge unit @xcite .
the reference potential @xmath26 is chosen sufficiently negative such that the coverage vanishes at @xmath26 for all values of @xmath19 used , and @xmath27 has the sign convention that @xmath28 favors adsorption .
the relationship between @xmath29 , @xmath19 , and @xmath20 is discussed further in the appendix .
when cl ions adsorb on the surface , a fraction of their charge is transfered through the external circuit .
this fraction , @xmath30 , is negative and is directly related to the average resident charge per ion , @xmath31 @xcite .
this relationship is an approximation and is more valid as the potential at the adsorbate approaches the value of the potential in the solution . for the current system s ionic strength , this condition is only approximately satisfied and may be considered as a source of error .
we have previously shown @xcite that for cl / ag(100 ) , the electrosorption valency @xmath0 depends on the coverage @xmath32 , which is defined as @xmath33 in order to investigate such a dependence more thoroughly , we here propose a model with a coverage - dependent next - nearest - neighbor lateral interaction energy @xmath1 , as well .
this is motivated by two assumptions : that @xmath0 is coverage dependent and that the major contribution to @xmath1 is due to electrostatic dipole - dipole interactions . a simple electrostatic picture of the adlayer , in which an adsorbate s resident charge and its image charge form a dipole , suggests a relationship between the electrosorption valency and the dipole moment .
if @xmath0 is coverage dependent ( @xmath34 ) , then the resident charge is also coverage dependent ( @xmath35 ) , and hence @xmath1 , which is proportional to @xmath36 , is coverage dependent as well ( @xmath37 ) .
assuming for simplicity that @xmath0 depends linearly on @xmath32 , @xmath38 the resident charge @xmath39 becomes @xmath40 and @xmath1 takes the form : @xmath41 where @xmath42 is a prefactor proportional to the square of a `` dipole distance '' .
this effect is known as _ depolarization _ , and may under extreme conditions lead to the possibilty of a first - order phase transition @xcite k. the model does produce a first - order transition at room temperature for some parameters that do not fit the experimental data . ] . if @xmath0 is coverage independent ( i.e. , @xmath43 ) , then @xmath1 becomes coverage independent , as well .
equilibrium monte carlo ( mc ) simulations were used to measure the equilibrium coverage as a function of @xmath29 , which was then converted to @xmath20 using eq .
( [ eq : mbar ] ) .
at each mc step a lattice site , @xmath5 , was chosen randomly , and a change in its occupation variable @xmath44 was attempted with a metropolis acceptance probability @xcite : @xmath45\ ; , \label{eq : p}\ ] ] where @xmath46 is the energy difference between the initial state and the proposed state .
the value of @xmath1 was updated in the simulations at each mc step , corresponding to the new proposed coverage .
the energy difference @xmath46 was calculated from eq .
( [ eq : h ] ) using two methods , a truncated sum of the contributions of neighboring occupied sites up to five lattice spacings away @xcite , and a mean - field - enhanced truncated sum up to three lattice spacings @xcite . in the latter method , energy contributions from adparticles more than three
lattice spacings away are calculated using a mean - field estimate as detailed in ref .
a detailed description of the experimental procedure is given in ref .
@xcite , and only a brief summary follows here .
the ag(100 ) single - crystal electrodes were chemically etched in cyanide solution , rinsed in milli - q water , and carefully annealed in a hydrogen flame .
they were then quickly transfered into the electrochemical cell after cooling in a stream of argon . using the hanging meniscus technique , a platinum wire counter electrode , and a saturated calomel electrode as reference electrode ,
the electrochemical measurements were carried out at a temperature of ( @xmath47)@xmath48c .
for the chronocoulometric experiments , the potential was set at initial values between @xmath491.375 and @xmath490.300 v vs sce until adsorption equilibrium was established , and then stepped to the final potential of @xmath491.400 v , where the chloride is completely desorbed from the surface .
the equilibrium mc simulations were performed at a temperature of @xmath50c for various values of @xmath51 , and @xmath52 .
the value of @xmath42 was varied from @xmath11 to @xmath53 mev in steps of @xmath54 mev , while @xmath55 and @xmath56 were varied from @xmath11 to @xmath57 in steps of @xmath58 , resulting in @xmath59 different simulations .
the resulting simulated isotherms were fitted to experimental data .
this large number of simulations necessitated the choice of the relatively small system size with @xmath60 .
the system displays a second - order phase transition from a disordered , low - density phase to an ordered @xmath61 phase @xcite at an intermediate value of the electrochemical potential . for the steepest part of the isotherm ( fig .
[ fig : bestfit ] ) , where critical slowing down due to the phase transition is important , the system was allowed to relax for a longer time to reach equilibrium than for the points further away from the phase transition .
we checked for finite - size effects , and no significant difference was observed when comparing the resulting isotherms with isotherms simulated using larger system sizes of @xmath62 , and @xmath63 .
a small value of @xmath64 can be used because @xmath32 is not the order parameter corresponding to the @xmath61 phase , and so its fluctuations , which are proportional to d@xmath65d@xmath29 , only diverge logarithmically with @xmath64 @xcite .
in addition , better statistics were collected close to the phase transition , using longer runs and more sampling than for points far from the phase transition , where the fluctuations are smaller .
using eq .
( [ eq : mbar ] ) , the simulated isotherms were converted from the @xmath29 scale to the @xmath20 scale and then fit to the experimental isotherms .
the fitting was done by varying the value of @xmath66 to minimize @xmath67 for each set of values for @xmath42 , @xmath55 , and @xmath56 . here , @xmath68 is the least - squares sum per degree of freedom , and @xmath69 and @xmath70 are the experimental and simulated coverages , respectively , corresponding to the same value of @xmath20 .
linear interpolation was used between the simulated data points to calculate @xmath70 for all values of @xmath20 . from fits to @xmath71 mm and @xmath72 mm experiments , a grid of of values of @xmath67 , was collected .
three different models were studied .
( i ) a constant @xmath0 ( i.e. , also constant @xmath1 ) model for which we only considered the simulations with @xmath43 .
( ii ) a model in which only the line with @xmath73 was considered .
these values for @xmath55 and @xmath56 were estimated from values of @xmath0 obtained from the experimental data in ref .
@xcite using the method in ref
. @xcite .
( iii ) the third model studied was the coverage dependent model with @xmath74 , where no constraints on the values of any of the parameters were imposed , and the whole parameter space was searched for a global minimum . a three - dimensional plot of the parameter space is shown in fig .
[ fig : param](a ) , where the diameters of the symbols are proportional to @xmath75 for fits to @xmath72 mm ( circles ) and 10 mm ( squares ) of the mean - field - enhanced simulations .
the parameter space for the non - mean - field - enhanced method ( not shown ) is similar .
figure [ fig : param](a ) shows the existence of several local minima and a good overlap between the minima for both concentrations .
the position of the accepted global minimum is indicated by the arrows .
figure [ fig : param](b ) is a projection onto the @xmath55 , @xmath56 plane , which shows that the minima are concentrated within one region close to the @xmath43 plane , suggesting a relatively weak dependence of @xmath0 on the coverage .
moreover , the @xmath67 values in the @xmath43 plane ( model ( i ) ) and for @xmath76 ( model ( ii ) ) are significantly larger than the @xmath67 values for the accepted global minimum ( model ( iii ) ) .
figures [ fig : param](c ) and [ fig : param](d ) show that while the magnitude of @xmath55 decreases monotonically with increasing @xmath42 , there is no general trend for @xmath56 as a function of @xmath42 . due to the existence of several shallow minima of @xmath67 , and due to the limited resolution of the grid in parameter space , we list in table 1 the best - fit parameter values for each model , along with all fits that have @xmath67 within @xmath77 of the best - fit value
. one can see in table 1 that there are two possible sets of parameters that fit to the 10 mm data for model ( iii ) with mean - field - enhanced simulations . to discriminate between these possible fits to the 10 mm data we check if they are also possible fits to the 20 mm experimental data .
the first set ( @xmath78 mev ) is the overall best fit for 10 mm .
however , it fits only slightly better than the second set ( @xmath79 mev ) , and it is not a possible fit to the 20 mm experimental data .
the only parameter set ( @xmath79 mev , @xmath80 , @xmath81 ) that is a possible fit to both the 10 mm and 20 mm experimental data , is the accepted best fit for model ( iii ) . using the same approach for the non - mean - field - enhanced method , models ( i ) and ( ii ) , and concentration , results in a unique set of parameter values for each model , simulation method , and concentration .
these accepted fits are summarized in table 2 .
notice in table 2 that for the accepted global minimum of model ( iii ) , and for both mean - field - enhanced and non - mean - field - enhanced methods , the values of @xmath55 and @xmath56 are reasonably close to the ones calculated from the experimental data @xcite using the method of ref .
also note that the fits for 10 mm are consistently better than the fits to the 20 mm experiments .
this might be due to violation of the dilute - solution limit at the higher concentration .
this may also be the reason that the values of @xmath82 , which are expected to be the same for both concentrations when fit to a single simulation , differ consistently by about @xmath83 mev between the concentrations .
the values of @xmath42 are consistently less negative for the longer - ranged lateral interactions of the mean - field - enhanced than for the non - mean - field method .
this is not surprising .
the quality of the fits is better for model ( iii ) , as can be seen from the plots corresponding to each of the models ( figs .
[ fig : bestfit ] , [ fig : expgamma ] , and [ fig : consgamma ] ) . models ( ii ) and ( i ) fit worse at either the lower - coverage part of the isotherm ( model ( ii ) ) , see fig .
[ fig : expgamma ] , or at the upper part of the isotherm ( model ( i ) ) , see fig .
[ fig : consgamma ] .
in contrast , the best - fit simulations in fig .
[ fig : bestfit ] fit the experiments well over the whole range of coverages .
the plots for the non - mean - field - enhanced fits ( not shown ) are similar .
a plot of @xmath0 vs @xmath20 is shown is fig .
[ fig : gamma_vs_e](a ) for the best - fit values of @xmath55 and @xmath56 corresponding to the 10 mm experimental data for each of the three models considered . also shown
are the values of @xmath0 obtained in ref .
@xcite by the method of ref .
while model ( ii ) is expected to fit the ref .
@xcite values of @xmath0 , model ( iii ) also predicts a mean value of @xmath0 close to , but slightly more negative than , the mean of the values of @xmath0 from ref .
the @xmath1 values corresponding to the best fits of the three models are shown in fig .
[ fig : gamma_vs_e](b ) .
since the fits are sometimes better for the mean - field - enhanced method , and sometimes worse , the mean of the sets of parameters obtained by including and excluding the mean - field enhancement are reported here as our final results : @xmath84 , and @xmath85 mev . in the far - field approximation used here , the lateral interaction energy between two parallel dipoles separated by a distance @xmath86 is given by @xmath87 where @xmath88 is the dipole moment , @xmath89 is an effective `` dipolar distance , '' and the second and third equalities result from eqs .
( [ eq : phi ] ) and ( [ eq : q ] ) , respectively .
consequently , @xmath90 for the @xmath91 mev final result , the dipole moment ranges between @xmath92 e for @xmath93 and @xmath94 e for @xmath95 , and the `` dipolar distance '' @xmath96 .
density functional theory calculations of a model without water @xcite suggest a dipole moment of @xmath97 e , which is only about @xmath98 higher than the fitted value at low coverage . from nave physical intuition
, the dipolar distance might be expected to be of the order of the ionic diameter of cl , about @xmath99 .
the value of @xmath89 obtained here involves several assumptions .
one main assumption used is that the dipole is a classical point dipole , which should be quite reasonable as the distance between neighboring dipoles is about an order of magnitude larger than the obtained value for @xmath89 .
moreover , we have assumed that the interaction is a classical electrostatic dipole interaction , while there could be quantum effects at the shortest length scales . on the other hand , when viewed as a charge distribution , the dipole is expected to have a much smaller `` dipolar distance '' than navely expected @xcite . using the relationship between the dipole moment and the helmholtz capacitance @xmath100 @xcite , @xmath101 and our values for the dipole moment and @xmath0
, we obtain a @xmath100 that is in the range of @xmath102 to @xmath103 @xmath104f/@xmath105 .
this range is comparable to the reported value of @xmath100 for ag(110 ) in chloride ion solution , @xmath106f/@xmath105 @xcite , which would yield @xmath107 .
this value of @xmath89 is of the same order of magnitude as our value . finally , although models ( i ) and ( ii ) are still possible , a self - consistent ( @xmath108 ) entirely electrostatic model not only fits better to the experimental data but also , with no constraints imposed , predicts values of @xmath0 that are reasonably compatible with those obtained from the experimental data using the method of ref . @xcite .
br electrosorption on the same substrate ag(100 ) displays different characteristics .
the electrosorption valency and the next - nearest - neighbor lateral interaction energy are not only more negative , @xmath109 and @xmath110 mev , but also independent of the coverage @xcite . in other words , while mutual depolarization @xcite is present for cl / ag(100 ) , this effect is not significant for br / ag(100 ) .
simulations of br / ag(100 ) using @xmath111 , when fit to three experimental data sets , yielded @xmath112 @xcite
. moreover , the value obtained for @xmath0 @xmath113 , is consistent with previous results from both simulations and experimental analysis @xcite .
since cl is more electronegative than br , it is expected to have a less negative @xmath0 or a more negative resident charge , consistent with our results .
a discrepancy between the values of the electrosorption valency @xmath0 obtained from fitting simulations to experiments , and the value obtained from the analysis @xcite of experimental data was reported in a previous study by abou hamad _
et al . _
this discrepancy suggested that the long - range interactions are dominated by electrostatic dipole / dipole effects for cl electrosorption on the ag(100 ) single crystal surface . in this work , a large set of mc simulations ( @xmath59 simulations ) over a grid in parameter space
was fitted to two sets of chronocoulometry experimental data for different concentrations .
the existence of local minima of @xmath67 in parameter space suggests alternative models . to within the resolution of our grid and the accuracy of the experimental data
, we have shown that while other , simpler , models are still possible , a purely electrostatic model can be used to describe the cl / ag(100 ) system .
it also predicts an electrosorption valency that is compatible with the value obtained through direct experimental data analysis by the method of ref .
additional sets of experimental data , with different concentrations around 10 and 20 mm of the adsorbate ion , along with a finer grid in parameter space would enable a more decisive determination of the most appropriate model and its parameter values .
we thank m.t.m . koper and s. frank for useful discussions and helpful comments .
we also thank m.t.m .
koper for bringing refs .
@xcite to our attention .
this work was supported in part by nsf grant no .
dmr-0240078 , and by florida state university through the school of computational science and the center for materials research and technology , the research centre jlich , and the netherlands organization for scientific research ( nwo ) .
we define the electrosorption valency as @xcite @xmath115 where @xmath116 is the charge on the metal , @xmath32 is the coverage , and @xmath104 is the _ chemical _ potential of the solution ( here taken as the dilute - solution approximation , @xmath117 ) . considering an auxiliary thermodynamic potential @xmath118 related to the surface tension @xmath119 as @xmath120 and using the lippman equation ,
@xmath121 we have @xmath122 consequently , one obtains the maxwell relation , @xcite @xmath123 the relation , @xmath124 , gives the electrosorption valency as the change in @xmath104 necessary to keep @xmath32 constant under a change in @xmath20 . but to keep @xmath32 constant , the _ electrochemical _ potential @xmath29 must remain constant , so that @xmath125 in the three - dimensional parameter space , obtained from fits of mean - field - enhanced simulations to the 10 mm experimental data ( squares ) and the 20 mm experimental data ( circles ) .
the diameters of the symbols are proportional to @xmath75 .
( a ) is a three dimensional view , ( b ) is a projection onto the @xmath55 , @xmath56 plane , and ( c ) and ( d ) are projections onto the @xmath55 , @xmath42 plane and the @xmath56 , @xmath42 , plane respectively .
the parameter space for the non - mean - field - enhanced method ( not shown ) is similar .
[ fig : param],width=576 ] vs @xmath20 for the values of @xmath55 and @xmath56 obtained from the best fits for models ( i ) , ( ii ) , and ( iii ) to the 10 mm experimental data .
the values of @xmath0 obtained in ref .
@xcite by the method of ref .
@xcite are also shown .
( b ) corresponding plots of @xmath1 vs @xmath20 for the three models.[fig : gamma_vs_e],title="fig:",width=384 ] vs @xmath20 for the values of @xmath55 and @xmath56 obtained from the best fits for models ( i ) , ( ii ) , and ( iii ) to the 10 mm experimental data .
the values of @xmath0 obtained in ref .
@xcite by the method of ref .
@xcite are also shown .
( b ) corresponding plots of @xmath1 vs @xmath20 for the three models.[fig : gamma_vs_e],title="fig:",width=384 ] .fitting parameters for cl adsorption on a ag(100 ) single - crystal surface , for models ( i ) , ( ii ) , and ( iii ) , with and without mean - field interactions .
possible fits to within @xmath77 of the best @xmath67 for each model are also shown . @xmath42 and @xmath66 are in units of mev . [
cols="^,^,^,^,^,^,^,^,^,^,^ " , ] | we present monte carlo simulations using an equilibrium lattice - gas model for the electrosorption of cl on ag(100 ) single - crystal surfaces . fitting the simulated isotherms to chronocoulometry experiments
, we extract parameters such as the electrosorption valency @xmath0 and the next - nearest - neighbor lateral interaction energy @xmath1 . both coverage - dependent and coverage independent @xmath0
were previously studied , assuming a constant @xmath1 [ i. abou hamad , th .
wandlowski , g. brown , p.a .
rikvold , j. electroanal .
chem .
554 - 555 ( 2003 ) 211 ] . here
, a self - consistent , entirely electrostatic picture of the lateral interactions with a coverage - dependent @xmath1 is developed , and a relationship between @xmath1 and @xmath0 is investigated for cl on ag(100 ) .
_ * keywords : * _ chlorine electrosorption ; lateral interactions ; electrosorption valency ; chronocoulometry ; continuous phase transition ; lattice - gas model ; monte carlo simulation . |
despite their humble appearances , the dwarf spheroidal ( dsph ) satellites of the milky way provide a source of persistent intrigue .
mysteries concerning their origin , evolution , mass density , and dynamical state make it difficult to know where to place these common galaxies in the context of standard ( e.g. cold dark matter ) models of structure formation .
are they primordial building blocks of bigger galaxies , or debris from galaxy interactions ?
while dsph galaxies have stellar populations similar in number to those of globular clusters ( @xmath2 ) , their stars are spread over a much larger volume ( @xmath3-@xmath4 kpc compared to @xmath5-@xmath6 pc in globular clusters ) resulting in the lowest luminosity ( i.e. , baryonic ) densities known in any type of galaxy . in many cases
it is unclear how these galaxies could have avoided tidal disruption by the milky way over their lifetimes without the addition of considerable unseen mass .
this characteristic of dsph galaxies suggests that the dynamics of these systems are dominated either by significant amounts of unseen matter , or that these galaxies are all far from dynamical equilibrium .
in general , the jeans equations ( equations ( 4 - 21 ) , ( 4 - 24 ) , and ( 4 - 27 ) of binney & tremaine 1987 @xcite , hereafter , bt87 ) provide a robust description of the mass distribution , @xmath0 , of a collisionless gravitational system such as a dsph galaxy in viral equilibrium , equation ( [ eq : em ] ) below .
their general form permits any number of mass components ( stellar , gas , dark ) , as well as anisotropy in the velocity dispersion tensor and a non - spherical gravitational potential .
when applied to spherical stellar systems and assuming at most only radial or tangential velocity anisotropy , these equations can be simplified to estimate the radial mass distribution ( equation 4 - 55 of bt87 ) : @xmath7 where @xmath8 is the spatial density distribution of stars , @xmath9 is the mean squared stellar radial velocity at radius @xmath1 .
the dimensionless isotropy parameter , @xmath10 , compares the system s radial and tangential velocity components : @xmath11 apart from the constraints on the geometry and the functional form of the anisotropy , equation ( [ eq : mrjeans ] ) is model - independent , making it an appealing tool .
it is relevant that equation ( [ eq : mrjeans ] ) and ( [ eq : jns1 ] ) below are applicable to any tracer population that in equilibrium and satisfies the collisionless boltzman equation .
kinematic datasets for individual dsph galaxies have historically been far too small ( typically containing radial velocities for @xmath12 30 stars ; see mateo 1998 ) to allow for a precise determination of @xmath0 using relations similar to equation ( [ eq : mrjeans ] ) .
instead , authors have been forced to adopt additional strong assumptions that reduce the jeans equation to even simpler forms and where the relevant distributions ( @xmath13 and @xmath14 in equation [ eq : mrjeans ] ) are represented by parametric models .
specifically , if one assumes isotropy of the velocity dispersion tensor ( i.e. , @xmath15 ) , spherical symmetry , and that the starlight traces the mass distribution ( effectively a single - component king model ( irwin and hatzidimitriou 1995 ) ) , then one obtains for the m / l ratio ( richstone and tremaine 1986 ) : @xmath16 where @xmath17 is the one - dimensional central velocity dispersion , @xmath18 is the central surface brightness , and @xmath19 is the half - light radius .
the parameter @xmath20 is nearly equal to unity for a wide range of realistic spherical dynamical models so long as the mass distribution is assumed to match that of the visible matter . with this approach
the modern variant of the classical ` king fitting ' procedure ( king 1966 ) the measured central radial velocity dispersion and surface brightness yield estimates of such quantities as the global and central m / l ratios .
in all eight of the mw s measured dsphs , large central velocity dispersions have conspired with their low surface brightnesses to produce large inferred m / l values .
this line of reasoning has led to a general belief that dsph galaxies are almost completely dark - matter dominated , and their halos have assumed the role of the smallest non - baryonic mass concentrations identified so far in the present - day universe .
this analysis fails for galaxies that are far from dynamical equilibrium , for example due to the effects of external tidal forces from the milky way ( fleck and kuhn 2003 ; klessen and kroupa , 1998 ) .
numerical models aimed to investigate this ( oh et al .
1995 ; piatek and pryor 1995 ) generally found that tides have negligible effects on the central dynamics of dsph galaxies until the survival time of the galaxy as a bound system becomes comparable to the orbital time ( about 1 gyr for the closer dsph satellites of the milky way ) .
observations agree with this broad conclusion by finding that remote dsph galaxies are no less likely to contain significant dark matter halos than systems located closer to their parent galaxy ( mateo et al .
1998 ; vogt et al . 1995 ) .
however , so - called resonance models ( fleck and kuhn 2003 ; kuhn 1993 ; kuhn et al .
1996 ) have been proposed that imply the central velocity dispersions can be significantly altered due to the inclusion of stars streaming outward from the barycenter of a galaxy and projected near the galaxy core .
recent versions of these models invariably imply a significant extension of the affected galaxies along the line - of - sight ( more precisely , along the line between the center of the dwarf and the milky way ; kroupa 1997 ; klessen and kroupa 1998 ) and a massive tidal stream along the satellite s orbit .
observations do not reveal strong evidence of significant line - of - sight distortions in dsph galaxies ( hurley - keller et al 1999 ; klessen et al .
2003 ) , other than sagittarius ( e.g. ibata et al .
1995 ) ; thus , for the purposes of this paper , we will assume that dsph galaxies are generally close to a state of dynamical equilibrium .
even with this enabling assumption , the classical analysis of dsph masses as we describe it above is far from ideal for a number of reasons .
first , though recent work ( e.g. irwin and hatzidimitriou 1995 ) has helped to greatly improve estimates of dsph structural parameters ( @xmath21 and @xmath22 in equation [ eq : moverl ] ) , the errors in the velocity dispersions often dominated by poisson uncertainties due to the small number of kinematic tracers
contribute the principle source of uncertainty in m / l estimates .
second , there is little reason to suppose the assumption that mass follows light in dsphs is valid . in all other galaxies
the bulk of the matter resides in a dark halo extending far beyond the luminous matter , a trend that becomes more exaggerated toward smaller scales ( kormendy and freeman 2004 ) . finally ,
velocity anisotropies , if they exist , may mimic the presence of dark matter , and so represent a tricky degeneracy in the model even if the assumption of isotropy is dropped .
modern instrumentation is poised to deliver dramatically larger kinematic datasets to help minimize the first problem .
for example , the michigan / magellan fiber system , now operational at the magellan 6.5-m telescopes , obtains spectra from which high - precision radial velocities can be measured of up to 256 objects simultaneously . as a result , it is now feasible to obtain thousands of individual stellar spectra in many dsph systems , enlarging sample sizes by more than an order of magnitude .
this not only reduces the statistical uncertainty of the dispersion estimation , but can also provide information on the spatial variation of the dispersion across the face of a galaxy .
these rich datasets therefore allow for fundamentally improved results even using fairly conventional analysis techniques .
for example , by parameterizing the velocity anisotropy , wilkinson et al .
( 2002 ) and kleyna et al .
( 2002 ) show that samples of @xmath23 stars can begin to break the degeneracy between anisotropy and mass in spherical systems .
but these large datasets also allow us to aim higher . in this paper
, we introduce and develop the formalism for a qualitatively different sort of analysis designed to make the most efficient use of large kinematic datasets . rather than adopting a model that parameterizes the various distributions used in the jeans equation ( e.g. , @xmath0 , @xmath13 , @xmath24 , or @xmath10 )
, we operate on the star count and radial velocity data directly to estimate the mass distribution non - parametrically .
we estimate the true three - dimensional mass distribution from the projected stellar distribution and the line - of - sight velocity distribution . in this first application of our technique
, we still require the assumptions of viral equilibrium , spherical symmetry , and velocity isotropy ( @xmath15 ) . in section 2 , we introduce notation and definitions of the three - dimensional as well as projected stellar and phase - space densities and review the jeans equation . in sections 3 and 4 ,
we show how to use very general shape constraints on the mass distribution to estimate the detailed form of the velocity dispersion profile and the radial mass distribution .
we illustrate the process on simulated data , demonstrating that , when furnished with large datasets , nonparametric analysis is a powerful and robust tool for estimating mass distributions in spherical or near - spherical systems ( section 5 ) .
we illustrate the application of our approach to an existing but relatively small kinematic dataset for the fornax dsph galaxy at the end of section 5 .
our emphasis in this paper is the application of our technique to dsph kinematics , but the method described in this paper is applicable to any dynamically hot system in which the radial velocity field is sampled discretely at the positions of a tracer population .
thus , our methodology would work for , say , samples of globular clusters or planetary nebulae surrounding large elliptical galaxies , or for individual stars within a globular cluster .
our approach follows earlier non - parametric analysis of globular cluster kinematics by gebhardt and fischer ( 1995 ) and merritt et al .
( 1997 ) , though the details of our method differ significantly .
let @xmath25 and @xmath26 denote the 3-dimensional position and velocity of a star within a galaxy .
we will regard these as jointly distributed random vectors , as in bt87 , p. 194 .
suppose that @xmath27 and @xmath28 have a joint density that is spherically symmetric and isotropic , so that @xmath29 = f_0(r , v)d{{\mbox{\boldmath$x$}}}d{{\mbox{\boldmath$v$}}},\ ] ] where @xmath30 and @xmath31 .
we also suppose that @xmath32 has been centered to have a mean of zero so that @xmath33 suppose finally that the mass density @xmath34 is spherically symmetric and let @xmath35 the total mass within @xmath1 of the center .
the goal of this paper is to come up with a means of estimating @xmath0 non - parametrically , specifically without assuming any special functional form for @xmath36 or @xmath34 . in the presence of spherical symmetry
, the relevant form of the jeans equation is @xmath37,\ ] ] where @xmath38 and @xmath39 in statistical terminology , @xmath40 is the conditional expectation of @xmath41 given @xmath42 ( this is the same as the conditional expectation on @xmath43 , hence the factor of 3 ) , and the marginal density of @xmath44 is @xmath45 . in astronomical terms
, @xmath40 is the one - dimensional radial - velocity dispersion profile squared , and @xmath45 is the true , three - dimensional density profile of the tracer population . using the notation of bt87 , these correspond to @xmath24 and @xmath13 , respectively . to estimate @xmath0 from equation ( [ eq : jns1 ] ) clearly requires estimates of the functions @xmath46 and @xmath47 . for dsph galaxies , the data available for this consist of a large sample of positions and photometry of stars in individual systems ( irwin and hatzidimitriou 1995 ; hereafter ih95 ) , and much smaller samples of stars with positions and velocities ( e.g. walker et al .
2004 ) . of course , it is not possible yet to determine complete three - dimensional positional or velocity information for any of these stars , but only the velocity in the line of sight and the projection of position on the plane orthogonal to the line of sight . with a proper choice of coordinates ,
these observables become @xmath48 , and @xmath49 .
these can be made equivalent to , say , right ascension and declination , and radial velocity , respectively .
the incomplete observation of @xmath32 does not cause a problem here , given the assumed isotropy , since @xmath50 can be replaced by @xmath51 in equation ( [ eq : mu ] ) without changing the value of @xmath40 .
the incomplete observation of position poses a more serious problem that is known as wicksell s ( 1925 ) problem in the statistical literature .
the procedures for estimating @xmath46 and @xmath47 are consequently different ; they will be considered separately in sections 3 and 4 .
let @xmath53 and @xmath54 denote the true three - dimensional distance of a star and its two - dimensional projected distance from the origin , respectively .
we denote the corresponding densities as @xmath55 and @xmath56 .
then @xmath57 , and @xmath58 for an example , consider plummer s distribution , @xmath59_{+}^{7\over 2},\ ] ] where @xmath60 is a constant , and @xmath61 is a parameter that is related to the velocity dispersion through @xmath62}$ ] . here
@xmath1 is measured in units of 100 pc and @xmath63 in km / s .
we also employ here the notation @xmath64_{+}$ ] to denote taking the larger of @xmath65 and @xmath66 . with this definition ,
we take @xmath64_+^\alpha$ ] to be shorthand for @xmath67_+)^\alpha$ ] ; that is , the value of @xmath65 is either positive or zero before raising to the power @xmath68 . for this case ,
@xmath69 $ ] and @xmath70 $ ] .
we use the plummer distribution in our simulations below .
ih95 provide data from which @xmath46 can be estimated .
these data consist of a sample of stars with projected radii @xmath71 and counts @xmath72 of the number of stars for which @xmath73 where @xmath74 divide the stars into bins .
let @xmath75 denote the distribution function of @xmath76 , so that @xmath77 = \int_{0}^{s } g_{\bf s}(t)dt$ ]
. then @xmath78 may be estimated from its empirical version @xmath79 in this context and throughout this paper , the @xmath80 symbol denotes functions estimated directly from empirical data .
since the samples of stars used by ih95 to determine the surface density of most of the milky way dsph galaxies are quite large , there is generally very little statistical uncertainty in our estimate of @xmath81 from equation ( [ eq : estgee ] ) .
if we now model @xmath46 by a step function , say @xmath82 for @xmath83 and @xmath84 , then @xmath85 may be recovered directly . in this case , the integral in equation ( [ eq : fg ] ) is easily computed leading to @xmath86 where @xmath87.\ ] ] then @xmath85 may be recovered from @xmath88.\ ] ] the summation is to be interpreted to be zero when @xmath89 , and the right side of equation ( [ eq : gf ] ) can in practice be estimated by replacing @xmath75 with its empirical version @xmath90 . the results of applying this method to a sample of 150,000 positions drawn from a plummer model with @xmath91 are shown in figures [ fig : counts ] and [ fig : fbar ] . figure [ fig : counts ] shows simulated counts of the number of stars in equally - spaced bins ranging from @xmath92 .
the estimate of @xmath46 from these simulated data is shown in figure [ fig : fbar ] . in both figures
the dotted line denotes the true function .
the estimation of @xmath46 follows wicksell s ( 1925 ) original analysis and can almost certainly be improved .
we have not done that here , because the sample sizes are so large , but related questions are currently under investigation by pal ( 2004 ) .
for estimating @xmath47 and @xmath94 it is convenient to work with squared radius @xmath95 , as in groeneboon and jongbloed ( 1995 ) and hall and smith ( 1992 ) .
the joint density of @xmath96 is then @xmath97 relation ( 7 ) represents a special case on integrating over all three velocity components and using @xmath98 .
now let @xmath99 and @xmath100 so that @xmath101 is the conditional expectation of @xmath102 given @xmath103 , and similarly for @xmath104 .
we also define the important function @xmath105 then , reversing the orders of integration and recognizing a beta integral ( see bt87 , p. 205 ) , we can write @xmath106 and @xmath107 where the prime notation ( @xmath108 ) denotes the derivative . clearly , @xmath109 in equation ( [ eq : mu ] ) . thus , the expression for @xmath0 from the jeans equation ( equation ( [ eq : jns1 ] ) ) can be rewritten as @xmath110 our estimator for @xmath46 was obtained in section 3 , so we focus here on determining @xmath111 . for our problem , we benefit from noting that the function @xmath112 is subject to certain shape restrictions which are useful in the estimation process . from equation ( [ eq : psi ] )
we can see that @xmath112 is a non - negative decreasing function , while from equation ( [ eq : jns2 ] ) it is evident that @xmath112 has a non - negative second derivative ( that is , it is upward convex ) .
in fact , @xmath113 must be an increasing function that vanishes when @xmath114 , physically consistent with its direct proportionality to @xmath0 in equation ( [ eq : jns2 ] ) .
the next step in the estimation of @xmath0 is to estimate @xmath112 .
suppose that there is a sample @xmath115 of projected positions @xmath116 and radial velocities @xmath117 measured with known error , @xmath118 .
thus , @xmath119 , where where @xmath120 are independent of @xmath121 with zero mean values and known variances , @xmath122 . in practice , there may be selection effects : an astronomer may choose to sample some regions of a galaxy more intensely than others , or external factors ( weather , moonlight , telescope / instrument problems ) may cause undersampling of some regions . to address this , we must include a selection function , @xmath123 , into the model . here , @xmath123 is the probability that a star in a galaxy is included in the kinematic sample , given that it is there . thus the selection probability is assumed to depend on projected position , @xmath124 , only .
as before , let @xmath125 ; then the joint density of @xmath126 and @xmath127 in the sample is @xmath128 where @xmath129 d\theta,\ ] ] @xmath130 is as in equation ( [ eq : vy ] ) , and @xmath131 is a normalizing constant , determined by the condition that the integral of @xmath132 be one . integrating over @xmath133 in ( [ eq : slctn1 ] ) , the actual density of @xmath134 is @xmath135 and @xmath131 is determined by the condition that @xmath136 . here
@xmath137 may be determined from the complete photometric data for a given galaxy to give the projected stellar distribution as a function of @xmath134 .
if @xmath46 is a step function , then combining equations ( [ eq : fg ] ) and ( [ eq : gf ] ) with the relation @xmath138 leads to @xmath139\ ] ] in the interval @xmath140 for @xmath141 .
if an astronomer can specify @xmath123 , then @xmath131 and @xmath142 can be computed directly using equation ( 15 ) .
then @xmath143 is an unbiased estimator of @xmath144 for each @xmath145 ; that is , @xmath146 for each @xmath145 .
if an astronomer can not specify @xmath147 , then it must be estimated from the data .
the first step is to estimate @xmath148 , which can in principle be done in many ways .
the simplest is to use a kernel estimator of the form @xmath149 where @xmath150 is a probability density function , called the _ kernel _ , and @xmath61 is a positive factor called the _ bandwidth _ that may be specified by the user or computed from the data .
silverman ( 1986 ) is a recommended source for background information on kernel density estimation .
possible choices of @xmath150 and the computation of @xmath61 are discussed there .
other methods for estimating @xmath148 include local polynomials [ 79 - 82 , loader 1999 ] and log - splines [ 178 - 179 , gu 2002 ] .
whichever method is adopted , once @xmath151 has been estimated , @xmath152 may be subsequently estimated by @xmath153 then @xmath112 can be estimated by replacing @xmath152 with @xmath154 in ( [ eq : psisharp ] ) . in this case @xmath155 is no longer exactly unbiased , but it is at least consistent in the statistical sense that as the number of data points increases , the estimator continuously tends ever closer to the true value .
generally , @xmath156 does not satisfy the shape restrictions when viewed as a function of @xmath145 : @xmath156 is unbounded as @xmath145 approaches any of the data points @xmath126 .
it is decidedly not monotone nor convex ( see figure [ fig : psi ] ) . for these reasons @xmath155
may be called the _ naive _ estimator .
this naive estimator can be improved by imposing physically justified shape restrictions on @xmath0 .
one can imagine a wide range of possible model forms for @xmath0 , but in the spirit of attempting a non - parametric solution , we adopt a simple spline of the form @xmath157^p_{+},\ ] ] where @xmath74 are knots , @xmath158 are constants , and @xmath159^p$ ] .
the simplest case is for @xmath160 , but this form for @xmath0 leads to a divergent estimate of the central velocity dispersion , @xmath161 . for @xmath162 , @xmath163 as @xmath164 ; thus , @xmath165 provides an estimate of the central mass density with the correct asymptotic behavior at the galaxy center .
unfortunately , translating the shape restrictions on @xmath0 into conditions on @xmath158 is much more difficult when @xmath162 than for @xmath166 . for this paper
we have chosen to compromise with @xmath167 because it leads to a good estimate of @xmath161 yet is still amenable to the application of the shape restrictions we described above .
we plan to tackle the case @xmath168 in a future paper
. for our adopted case ( @xmath167 ) the expression ( equation ( [ eq : spln1 ] ) ) for the mass , @xmath0 , is a quadratic spline : @xmath169 ^ 2_{+}.\ ] ] one natural constraint is that @xmath0 remain bounded as @xmath170 . expanding equation ( [ eq : spln2 ] ) to read @xmath171 for @xmath172
, we see that this constraint is equivalent to requiring that @xmath173 , and @xmath174 . from this first equality constraint ( @xmath173 )
, we can trivially say @xmath175 and , therefore , @xmath176
^ 2_{+ } - [ r - r_{m-1}]_{+}^{2}\big\}.\ ] ] we also require that @xmath0 be a non - decreasing function of @xmath1 ( no negative mass ) , or equivalently that the derivative @xmath177 be non - negative .
it is clear from equations ( [ eq : spln2 ] ) and ( [ eq : spln3 ] ) that @xmath177 is a piecewise linear function that is constant on the interval @xmath178 .
thus , the condition that @xmath0 be non - decreasing is equivalent to requiring that @xmath179 for @xmath180 .
that is , @xmath181 for @xmath182 .
for the case @xmath183 , this constraint can be written as @xmath184 .
this is clear from noting that we do not change this summation if we replace @xmath185 with @xmath186 ; thus we can write @xmath187 , where we have used the facts that @xmath188 and that @xmath174 as required by equation ( [ eq : spln2 ] ) .
these are the constraints imposed on @xmath189 that we employ below to estimate @xmath0 .
if we solve for @xmath111 in equation ( [ eq : jns2 ] ) , we can write @xmath190 where @xmath191_{+}^{2}-[r - r_{m-1}]_{+}^{2}\big),\ ] ] for notational convenience , we define @xmath192 , then solve for @xmath193 in terms of the coefficients @xmath194 to get @xmath195 where @xmath196 for @xmath197 . the next step is to impose the shape restrictions implicit in equation ( [ eq : jns2 ] ) .
if @xmath155 were square integrable , this could be directly accomplished by minimizing the integral of @xmath198 ^ 2 $ ] , or equivalently by minimizing @xmath199 with respect to the coefficients , @xmath194 .
the function @xmath155 is not integrable , essentially because @xmath200 , but @xmath201 can still be minimized .
letting @xmath202 be the vector @xmath203^t$ ] , this criterion may be expressed in matrix notation as @xmath204 where the elements of the vector @xmath205 are given by @xmath206 and the elements of @xmath207 are given by @xmath208 for @xmath209 .
thus , the estimation problem for @xmath112 leads to a quadratic programming problem of minimizing @xmath201 in equation ( [ eq : kappa ] ) subject to condition ( [ eq : cnstrnt1 ] ) with equality when @xmath210 .
if we can determine @xmath211 and @xmath212 , we can estimate the coefficients @xmath194 which in turn gives us our estimate for @xmath0 .
observe that the @xmath211 depend explicitly on both the deduced stellar density distribution , @xmath46 , and the velocity dispersion profile derived from a set of radial velocity tracers , @xmath47 .
the @xmath212 depend only on @xmath45 .
let @xmath213 be defined as @xmath214 \nonumber \\ & & \ \ \ \ \ \ \ \ \ \ + { 1\over n}\sum_{i : y_i\le z}{u_i^2-\sigma_i^2\over w(y_i)}\left[-{4\over 3 } y_i^{3\over 2}+2zy_i^{1\over 2 } \right],\end{aligned}\ ] ] where the weighting function , @xmath152 , may be specified or estimated as described earlier in this section .
then @xmath211 can then be computed numerically as @xmath215 this form for the @xmath211 is simpler to evaluate than the double integral form for these terms given earlier . in a similar fashion we can write @xmath212 as @xmath216 - { ( s+t ) \min[s , t]^2\over 2 } + { \min[s , t]^3\over 3}\right ] \gamma_i(s ) \gamma_j(t ) dt ds \nonumber \\\end{aligned}\ ] ] these integrals ( [ eq : zi ] ) and ( [ eq : qij ] ) can be computed explicitly and matlab codes are available at website _
http://www.stat.lsa.umich.edu/~wangxiao _ to do so .
if we define @xmath217 as the solution to the quadratic minimization problem ( see equation ( [ eq : kappa ] ) ) , our estimates of @xmath218 , @xmath219 , @xmath40 , and @xmath0 can be denoted using analogous notation and are given by @xmath220 @xmath221 @xmath222 and , @xmath223 ^ 2_+ -\left[r - r_{m-1}\right]_+^2\right\}.\ ] ] the latter two functions specify the projected radial velocity and mass profiles , respectively .
to illustrate an application of our approach , we have drawn a sample of 1500 @xmath224 pairs from a plummer model with @xmath225 and @xmath226 ( see figure [ fig : su ] ) .
the naive estimator @xmath227 may be computed from these data using equation ( [ eq : psisharp ] ) and it is shown in figure [ fig : psi ] , along with the improved estimator @xmath228 . the need for the improved estimator for @xmath112 is clear since , as expected , @xmath227 is a highly irregular function which reacts strongly to the presence of individual data points .
the estimated velocity dispersion profile , @xmath229 is presented in figure [ fig : mu ] and the estimated mass profile , @xmath230 , is illustrated in figure [ fig : mass ] . in all cases
, the dotted line represents the true functions computed directly from the plummer model from which the @xmath224 pairs were drawn . in these figures @xmath231 , @xmath232 , and @xmath233 .
the last plataux on the right in figure [ fig : mass ] ( @xmath234 ) deserves comments , because it illustrates an important feature of the problem in general and our method in particular : estimates become unreliable for large values of @xmath1 , because the data become sparse , and shape restricted methods are especially prone to problems such problems
woodroofe and sun ( 1993 ) .
deciding exactly where reliability ends is difficult , but the confidence bands of figures [ fig : masscb ] and [ fig : diffsample ] provide some indication . in the example of figure 6 , they correctly predict that reliability deteriorates abruptly for @xmath234 and , interestingly , make the same prediction for a range of sample sizes .
there are several sources of error in our derived velocity dispersion and mass profile estimators which may be divided into the broad classes of modeling error , systematic error , and statistical error .
while quite general , our assumptions also represent some over - simplification and these too introduce uncertainty to our results .
the assumption of spherical symmetry , for example , ignores the known ellipticities seen in most dsph galaxies ( ih95 , mateo 1998 ) . but even if we assume our working assumptions were all correct , systematic error will develop from our approximations of distributions as step functions , piecewise linear functions , and quadratic splines . on general grounds , this error will be small provided that the underlying true functions are smooth , so we feel justified in ignoring this source of systematic error .
we have also ignored the statistical error in the estimation of @xmath46 in section [ sect : h ] , because the sample sizes in ih95 are so large .
that leaves the statistical error in the kinematic data as the principle source of uncertainty in our estimation of @xmath112 and the estimators of @xmath47 and @xmath94 . to quantify the magnitude of this error ,
let @xmath235 denote the error committed when estimating @xmath0 , and let @xmath236 denote the @xmath237 percentile of its distribution distribution , so that @xmath238 = \alpha$ ] .
then @xmath239 = 1 - 2\alpha,\ ] ] and @xmath240 $ ] is a @xmath241 percent confidence interval for @xmath0 . within the context of plummer s model
, the @xmath236 can be obtained from a monte carlo simulation .
if we generate @xmath242 samples from plummer s model , we obtain @xmath242 values for the errors @xmath243 , which we denote as @xmath244
. then @xmath236 may be estimated by the value of @xmath145 for which the @xmath245 of the @xmath246 are less than or equal to @xmath145 .
we applied this procedure to a grid of @xmath1 values with @xmath247 and @xmath248 and then connected the resulting bounds to in a continuous piecewise linear fashion to obtain the 95% and 68% confidence bounds shown in figure [ fig : masscb ] .
the 95% and 68% confidence bounds for @xmath47 shown in figure [ fig : mucb ] were obtained in a similar manner .
the procedure just outlined requires knowing the true @xmath0 and the distributions of @xmath134 and @xmath49 from the plummer model .
but the exercise is still useful since it shows the intrinsic accuracy of the estimation procedure .
for example , when the sample size is @xmath249 , the estimator of @xmath0 has very little accuracy for @xmath250 .
somewhat surprisingly , the estimator for @xmath0 retains reasonable accuracy for @xmath251 despite the lack of much data beyond @xmath252 ( see figure [ fig : counts ] ) .
we have also explored how our estimates fare with significantly smaller samples of tracers with known positions and velocities ( i.e. , fewer @xmath224 pairs ) .
figure [ fig : diffsample ] shows @xmath253 percent confidence bounds for the estimates of @xmath0 and @xmath40 based on sample sizes @xmath254 , @xmath255 , @xmath256 and @xmath257 stars . as one might expect , the confidence bounds expand as the sample size decreases , but , interestingly , the radial location of where the estimator performs poorly only very slowly migrates to smaller values of @xmath1 as the sample shrinks .
the requirement that we know @xmath0 and the distribution of @xmath134 and @xmath49 ahead of time in order to estimate the confidence bounds can be avoided by using a bootstrap procedure . in its simplest form , the bootstrap can be implemented as follows : given a sample of @xmath258 values ( corresponding to the observable projected position and radial velocity for each star in a kinematic sample ) , first compute @xmath230 .
then generate @xmath259 samples with replacement objects , one selects a new sample of @xmath186 values from the original values @xmath260 .
the new sample differs from the original in that each value @xmath260 is returned to the sample before choosing the next value @xmath261 . in this way
, some values from the original sample will invariably be chosen more than once to form the bootstrap sample . ] from the given sample and compute estimators @xmath262 for each of these samples .
let @xmath263 and estimate the percentiles of @xmath264 as described above . letting @xmath265 denote the estimated @xmath237 percentile , @xmath266 $ ] is the bootstrap confidence interval for @xmath0 . for more details ,
see efrom and tibshirani ( 1993 ) who provide a good introduction to bootstrap methods .
this approach offers a practical way of assessing the statistical errors in @xmath229 and @xmath230 of an actual kinematic sample .
there is also the possibility of obtaining large - sample approximations to the error distribution . from ( [ eq : zi ] )
, it is clear that the @xmath211 are approximately normally distributed in large samples .
this suggests that the distribution of @xmath267 should be approximately the same as the distribution of a shape - restricted estimator in a normal model .
unfortunately , calculating the latter is complicated .
some qualitative features are notable , however .
monotone regression estimators are generally quite reliable away from the end points , but subject to the _ spiking problem _ near the end points . for non - decreasing regression ,
the estimator has a positive bias near large values of the independent variable ( in this case , @xmath1 ) , and a negative bias at small values of the independent variable ( i.e. for @xmath268 ) .
the positive bias at large @xmath1 is very clearly seen in figure [ fig : masscb ] and is exacerbated by the sparsity of data .
the negative bias near @xmath114 is not evident because it is ameliorated by two effects .
first , the mass distribution is known to be always positive ( @xmath269 ) and goes to zero at @xmath114 ( @xmath270 ) .
this eliminates bias at @xmath114 .
second , our assumption ( [ eq : spln3 ] ) requires @xmath230 to be of order @xmath271 for small @xmath1 , while @xmath0 is on physical grounds known to approach zero as @xmath272 for small @xmath1 . as an application to real observational data , we have use the radial velocity data from walker et al ( 2005 ) for the fornax dsph galaxy .
this dataset consists of velocities for 181 stars observed one at a time over the past 12 years ( mateo et al .
1991 has a description of some of the earliest data used here ) .
data from newer multi - object instruments are not included here ( see walker et al 2005b for a first analysis of these newer results ) .
full details about the observations can be found in the papers referenced above ; for the present application we simply adopt these results along with their quoted errors to see what sort of estimator for @xmath0 we can extract from these data . from the simulations we already know that the dataset is relatively small for this method , but these simulations also imply that we have no reason to expect a strong bias in our mass estimator due to the sample size , only a potentially large uncertainty ( at least a factor of 2 ) in the final mass estimate .
the histogram of the star counts from ( irwin and hatzidimitriou 1995 ) is shown in the first panel of figure [ fig : fornax1 ] .
the remaining two panels in figure [ fig : fornax1 ] show the derived estimates for the spatial density of stars in fornax @xmath273 and the true projected radial distribution of stars in fornax @xmath274 .
the bottom panels of figure [ fig : fornax21 ] give the estimated projected radial distribution of stars from ( walker et al , 2004 ) and the estimated selection function @xmath275 derived from the data . figure [ fig : fornax22 ] show the mass profile , @xmath230 and the squared velocity dispersion profile , @xmath229 .
previous estimates of the fornax mass have been based primarily on the classical , or king ( 1966 ) analysis outlined in the introduction of this paper , and give @xmath276 range from @xmath4 to @xmath277 m@xmath278 ( mateo et al .
1991 ; walker et al .
these models resemble truncated isothermal spheres and implicitly assume spherical symmetry , velocity isotropy , that mass follows light , and a specific parametric form for the joint density function @xmath36 .
our mass distribution estimator provides an independent measurement employing neither of the latter two assumptions .
if one believes the second plateau seen at @xmath279 kpc in figure [ fig : fornax22 ] ( first panel ) is a real feature , this would indicate a fornax mass of @xmath280 .
however , we note that the simulations using plummer models ( figure [ fig : diffsample ] ) show similar plateaux at large radii related not to the underlying mass , but rather to the scarcity of remaining data points . the first plateau ( @xmath281 kpc ) , by contrast , covers a region for which the plummer simulations indicate even a relatively small dataset can yield a reasonable estimate of the mass .
we therefore consider the two plateaux in the @xmath0 curve of figure [ fig : fornax22 ] to bracket the region of plausible fornax mass as measured from an n @xmath12 200 sample by the estimation technique .
the resulting interval of @xmath282m@xmath278 @xmath283m@xmath284 @xmath285 @xmath286 m@xmath278 is of considerably larger mass than the results of the classical analysis .
the estimation approach then offers an attractive alternative to variants of the classical analysis that require parametric dynamical models to interpret kinematic data for dsphs .
our simulations show that this new non - parametric analysis will provide significantly more accurate results as kinematic samples grow larger .
finally , we can make some comments regarding the effect of velocity anisotropy in our results .
if we consider an extreme case where the outermost bins of the fornax velocity dispersion profile are dominated by stars in tangential orbits ( that is , @xmath287 ; see equations ( [ eq : mrjeans ] ) and ( [ eq : jbeta ] ) ) , then the outer bins give us the mass directly ( from equation [ eq : mrjeans ] ) as @xmath288 , or @xmath289 m@xmath278 , which is equivalent to the lower limit provided by the mass estimation technique .
thus , even if we consider an extreme breakdown of the isotropy assumption , the data alone support a larger mass than the simple king analysis . since it is more likely that the velocity distribution has intermediate anisotropy ( kleyna et al .
2002 ) , we conclude that the the mass of fornax to the outermost measured data point is between @xmath290 and @xmath286 m@xmath291 for an observed luminosity of @xmath292 l@xmath278 for fornax , this gives a global m / l ratio ranging from 7 to 22 .
thus , even the most luminous , most baryonic - dominated dsph satellite of the milky way is dominated by dark matter .
we have presented a new method for estimating the distribution of mass in a spherical galaxy . along the way , we have made some choices , generally preferring the simplest of several alternatives for example , the use of quadratic splines in section 4 and simple histograms in section 3 . in spite of this , the algorithm is a bit complicated .
is it really better than simpler methods for example , using kernel smoothing to estimate the distribution of projected radii and line of sight velocities and then using inversion to estimate @xmath0 ?
our method differs from the simpler approach through its essential use of shape restrictions . from a purely statistical viewpoint ,
the monotonicity of @xmath113 in equation ( [ eq : jns2 ] ) provides a lot of useful information [ robertson , et.al .
( 1988 ) ] .
early in our investigations , we tried the kernel smoothing outlined above and found that the resulting estimators did not satisfy the shape restriction , leading to negative estimates of the mass density .
imposing the shape restrictions directly guarantees non - negative estimates , at least .
it also reduces the importance of tuning parameters , like the bandwidth in kernel estimation or the number of bins in our work . imposing the shape restrictions also complicates the algorithm and leaves a quadratic programming problem at the end .
we think that the game is worth the candle .
there is some similarity between our method and that of merritt and saha ( 1993 ) in that both use basis functions , splines in our case , and both impose shape restrictions , the condition that @xmath293 in theirs .
our method makes more esssential use of shape restrictions and treats the inversion problem in greater detail . | thanks to instrumental advances , new , very large kinematic datasets for nearby dwarf spheroidal ( dsph ) galaxies are on the horizon .
a key aim of these datasets is to help determine the distribution of dark matter in these galaxies .
past analyses have generally relied on specific dynamical models or highly restrictive dynamical assumptions .
we describe a new , non - parametric analysis of the kinematics of nearby dsph galaxies designed to take full advantage of the future large datasets .
the method takes as input the projected positions and radial velocities of stars known to be members of the galaxies , but does not use any parametric dynamical model , nor the assumption that the mass distribution follows that of the visible matter .
the problem of estimating the radial mass distribution , @xmath0 ( the mass interior to true radius @xmath1 ) , is converted into a problem of estimating a regression function non - parametrically . from the jeans equation
we show that the unknown regression function is subject to fundamental shape restrictions which we exploit in our analysis using statistical techniques borrowed from isotonic estimation and spline smoothing .
simulations indicate that @xmath0 can be estimated to within a factor of two or better with samples as small as 1000 stars over almost the entire radial range sampled by the kinematic data .
the technique is applied to a sample of 181 stars in the fornax dsph galaxy .
we show that the galaxy contains a significant , extended dark halo some ten times more massive than its baryonic component . though applied here to dsph kinematics , this approach can be used in the analysis of any kinematically hot stellar system in which the radial velocity field is discretely sampled . |
recently , topological properties of time - reversal - invariant band insulators in two and three dimensions have been extensively studied@xcite . a class of insulators preserving the time reversal symmetry is called topological insulators characterized by non - trivial topological [email protected] topological insulators
have been intensively studied because of the existence and potential applications of robust surface metallic states . both in two and three dimensions , the topological phases are typically realized in the systems with strong spin - orbit interaction@xcite .
all the known topological insulators contain heavy or rare metal elements , such as bismuth or iridium , which poses constraints on the search for topological materials .
irrespective of constitutents , ubiquitous mutual coulomb repulsions among electrons have been proposed to generate effective spin - orbit couplings @xcite .
it has been proposed that an extended hubbard model on the honeycomb lattice can generate an effective spin - orbit interaction from a spontaneous symmetry breaking at the hartree - fock mean - field level leading to a topologically non - trivial phase@xcite . since the honeycomb - lattice system , which is dirac semimetals in the non - interacting limit , becomes a topologically nontrivial insulator driven by the coulomb interaction , this phase
is often called a topological mott insulator ( tmi ) .
this phenomenon is quite unusual not only because an emergent spin - orbit interaction appears from the electronic mutual coulomb interaction , but also it shows an unconventional quantum criticality that depends on the electron band dispersion near the fermi point@xcite .
however , this proposed topological phase by utilizing the ubiquitous coulomb repulsions has not been achieved in real materials even though the tmi is proposed not only in various solids @xcite but also in cold atoms loaded in optical lattices @xcite . even in simple theoretical models such as extended hubbard models ,
it is not clear whether the tmis become stable against competitions with other orders and quantum fluctuations .
reliable examination of stable topological mott orders in the extended hubbard model is hampered by competing symmetry breakings such as cdws .
couplings driving the topological mott transitions are also relevant to formations of a cdw , which has not been satisfactorily discussed in the previous studies . since the emergence of the tmi in the honeycomb lattice requires the coulomb repulsion between the next nearest neighbor sites , the long - period cdw instability must be considered on equal footing , which is not captured in the small - unit - cell mean - field ansatz employed in the previous studies .
examining charge fluctuations with finite momentum over entire brillouin zones is an alternative way to clarify the competitions among tmis and cdws , as studied by employing functional renormalization group methods @xcite .
however , first order thermal or quantum phase transitions not characterized by diverging order - parameter fluctuations are hardly captured by such theoretical methods .
the most plausible symmetry breking competing with tmis indeed occurs as a first order quantum phase transition as discussed later .
the quantum many - body fluctuations beyond the mean - field approximation severely affects the stability of the tmi .
the stability of the tmi and estimation of the critical value of interaction on the honeycomb lattice has mainly been considered by mean - field calculations which can not treat the correlation effect satisfactorily .
however , there exists a reliable limit where the tmi becomes stable : for infinitesimally small relevant coulomb repulsions , the quadratic band crossing with vanishing fermi velocities cause the leading instability toward the tmi , as extensively examined by using perturbative renormalization group methods@xcite .
however , examining the instabilities toward the tmi in dirac semimetals requires elaborate theoretical treatments . in this study , for clarification of the competitions among tmis and other symmetry breakings
, we first examine the long - period cdw at the level of mean - field approximation that turns out to be much more stable compared to that of short period . indeed , this cdw severly competes the tmi on the honeycomb lattice .
the tmi on the honeycomb lattice studied in the literatures is consequently taken over by the cdw . we , however , found a prescription to stabilize the tmis on the honeycomb lattice : by reducing the fermi velocity of the dirac cones , the tmi tends to be stabilized .
we examine the realization of the tmis in the extended hubbard model on the honeycomb lattice by controlling the fermi velocity and employing a variational monte carlo method@xcite with many variational parameters@xcite , multi - variable variational monte carlo ( mvmc)@xcite , together with the mean - field approximation .
finally , we found that , by suppressing the fermi velocity to a tenth of that of the original honeycomb lattice , the tmi emerges in an extended parameter region as a spontaneous symmetry breaking even when we take many - body and quantum fluctuations into account .
this paper is organized as follows . in section
[ sec : model and method ] , we introduce an extended hubbard model and explain the order parameter of tmi .
we also introduce the mvmc method . in section [ sec : stability ] , we first show how the long - range cdw becomes stable over the tmi phase in standard honeycomb lattice models .
then we pursue the stabilization of tmi by modulating fermi velocity at the dirac cone at the mean - field level .
finally we study by the mvmc method the effect of on - site coulomb interaction which was expected to unchange the stability of the tmi phase at the level of mean - field approximation .
section [ sec : dis ] is devoted to proposal for realization of tmis in real materials such as twisted bilayer graphene .
in this section , we study ground states of an extended hubbard model on the honeycomb lattice at half filling defined by @xmath0 where the single particle parts of @xmath1 are defined as @xmath2 and @xmath3 is the spin - orbit interaction . here @xmath4 is a creation ( annihilation ) operator for a @xmath5- spin electron , @xmath6 is an electron density operator , @xmath7 represents the hopping of electrons between site @xmath8 and @xmath9 , and @xmath10 are on - site ( off - site ) coulomb repulsion .
bracket @xmath11 denotes the next - neighbor pair , @xmath12 is the strength of the spin - orbit interaction and @xmath13 is the @xmath14=@xmath15-spin operator . in eq.([hso ] ) , the @xmath16-th site is in the middle of the next nearest neighboring pair @xmath8 and @xmath9 as shown in fig . [
fig : honeycomb ] , and @xmath17 is the vector from the site @xmath8 to @xmath9 .
we start with the hopping matrix @xmath7 in eq.([eq : h0 ] ) for the bond connecting a pair of the nearest - neighbor sites @xmath18 , @xmath19 as the simplest extended hubbard model on the honeycomb lattice .
later , we will examine the effect of third neighbor hoppings @xmath20 .
we take @xmath21 as the unit of energy and set @xmath22 throughout the rest of this paper . for the non - interacting limit , @xmath23 ,
the system becomes a topological insulator when @xmath12 is nonzero , which is identical to the topological phase of the kane - mele model @xcite .
for the off - site coulomb repulsion , we mainly consider the second neighbor interaction ( @xmath24 ) , which is necessary for the emergence of the correlation - induced topological insulator @xcite .
the second neighbor coulomb repulsions @xmath25 effectively generate the spin - orbit interactions , which are identical to @xmath12 , and induce topological insulator phases even for @xmath26 .
we note that the coulomb repulsion of on - site or the nearest neighbor site do not affect the stability of the tmi phase at the level of the mean - field approximation .
indeed , our mvmc results show that this is essentially true beyond the level of mean - field approximation which will be discussed in the later section . therefore , for the moment , we focus only on the effect of @xmath27 for the consideration of interaction effects . .
next nearest neighbor interaction @xmath25 and the loop current @xmath28 flowing between the next nearest neighbor are shown by red and blue arrows .
, width=245 ] here the tmi is the broken symmetry phase characterized by the order parameter @xmath28 defined by @xmath29 where the self - consistent mean fields for the second neighbor bonds are given by @xmath30 here , @xmath31 denotes the expectation value for next nearest neighbor bonds and the order parameter @xmath28 is physically interpreted as spin dependent loop currents flowing within a hexagons constituting the honeycomb lattice . at the level of the mean - field approximation ,
this quantum phase transition is understood by decoupling two - body electron correlation term of the next nearest neighbor bond , @xmath32 , as @xmath33 we also note that this phase transition to tmi , which is proposed not only on the honeycomb lattice but also in several other lattice models , belongs to an unconventional universality class , which depends on the dimension of the system and the dispersion of the electron band@xcite . in this section
, we pursue the topological mott phase transition by employing the mean - field analysis and the variational monte carlo method . for the latter method
, we use a trial wave function of the gutzwiller - jastrow form , @xmath34 with a one body part , @xmath35 where @xmath36 is the variational parameters and @xmath37 is the number of the electrons in the system .
though this form of the wave function restricts itself to the hilbert subspace with the zero total @xmath38-component of @xmath14=@xmath15 , @xmath39 , it can describe topological phases on the honeycomb lattice as long as we use complex variables for @xmath36 . here ,
@xmath40 and @xmath41 are the gutzwiller and jastrow factors defined as @xmath42 and @xmath43 respectively , with which the effects of electron correlations are taken into account beyond the level of mean - field approximation .
the expectation value , @xmath44 , is minimized with respect to variational parameters , @xmath36 , @xmath45 , and @xmath46 , by using the monte - carlo sampling and using the stochastic reconfiguration method by calculating gradient of the energy and the overlap matrix in the parameter space@xcite .
we optimize the parameters by typically 2000 stochastic reconfiguration steps . in the present implementation of the variational monte carlo method with complex variables ,
the feasible system size for the calculation is about up to 300 , from which we speculate properties in the thermodynamic limit . for this purpose
, we perform the size extrapolation using the standard formula @xmath47 where @xmath48 is the linear dimension of the system size .
we note that the order of taking the limit in the right hand side of eq.([zeta_limit ] ) is also important for the validity of the extrapolation , which is known as the textbook prescription for the defining spontaneous symmetry breakings .
another practical way to determine the spontaneous symmetry breakings is in principle the finite size scaling of the correlation for the order parameter .
however , the latter finite size scaling is not practically easy .
the correlation of @xmath28 becomes too small because @xmath28 itself is about the order of @xmath49 and the correlation becomes the order of @xmath50 which becomes comparable to the statistical error of the monte carlo sampling . for the size extrapolation , we fit the data of the finite size calculations by a polynomial of @xmath51 , that is , we assume the size dependence by @xmath52 the above assumption for the finite size scaling is based on a practical observation and an analogy to the finite size scaling in the spin wave theory @xcite . as a practical observation , @xmath53 for the limit @xmath54 and @xmath55
is scaled by @xmath51 .
we examine the ground states of the extended hubbard model on the honeycomb lattice by tuning the on - site coulomb repulsion @xmath56 and the second neighbor coulomb repulsion @xmath25 . even for the parameter sets favorable to the tmis that have been studied in the pioneering works on the tmi @xcite
, we show that the tmi is not stabilized when we take into account other competing orders overlooked in the literature .
however , here , we reveal that , by tuning the fermi velocities of the dirac cones , the tmi is indeed stabilized . in this section
, we consider long - period cdw states and show that 6-sublattice order is stable when @xmath25 becomes dominat .
this state is schematically shown in fig.[fig : cdw6sub](a ) , where the electron density per site is disproportionated into four inequivalent values .
when we pick up very rich sites or very poor sites , they constitute triangular lattices .
the moderately rich or poor sites constitute the honeycomb lattices .
sites labelled by 1 ( red ) , 2 ( light red ) , 3 ( light blue ) and 4 ( blue ) correspond to the sites where the electron densities are very rich , rich , poor and very poor , respectively .
( b ) the growth of order parameter of the cdw state .
( c ) resulting phase diagram of honeycomb lattice for @xmath56 and @xmath25 at the level of mean - field approximation .
we do not find the region of stable tmi phase .
, width=302 ] figure [ fig : cdw6sub](b ) shows the growth of the order parameters of the 6-sublattice cdw state for several different parameters of @xmath56 . in this calculation , we have used the mean - field approximation , where we defined the mean field by @xmath57 where @xmath58 is defined through @xmath59 here , the condition of half filling is satisfied by the following identity , @xmath60 as can be seen from fig.[fig : cdw6sub](b ) , the order parameter grows above @xmath61 even if we set @xmath62 .
this is a quite serious problem for the stabilization of the tmi since the critical value of @xmath25 for the tmi at the mean - field approximation is also about @xmath63 .
in addition , the energy gain due to the formation of the topological mott order is much smaller than that for the cdw state .
when we consider larger values of @xmath56 , then the system becomes an antiferromagnetic state .
furthermore , the mean - field approximation often overestimates the ordered phase , because it does not take into account fluctuation effects seriously . actually , by using the mvmc method , we could not find the region where the topological insulator becomes the ground state in the parameter space of @xmath56 and @xmath27 . the resulting phase diagram by the mean - field approximation is shown in fig .
[ fig : cdw6sub ] ( c ) . there
the antiferromagnetic phase is denoted as nel , where the spin on different sites of the bipartite aligns in the anti - parallel direction .
as examined above , we found that the cdw state dominates and could not find parameter regions where the tmi is stable .
however , we find that the tmi becomes stable by modulating the fermi velocity at the dirac cones in the electronic band dispersion actually it is confirmed that , when the fermi velocity is 0 and then the dirac cones change to the quadratic band crossing points , the phase transition from a zero - gap semiconductors to a tmi occurs with infinitesimal coulomb repulsions @xcite .
for the honeycomb lattice , it is possible to change dirac semimetals to quadratic band crossings by introducing the third neighbor hopping @xmath20 ( schematically shown in fig.[fig : t3 ] ( a ) ) .
then the part of the hamiltonian @xmath64 is replaced with @xmath65 where the third neighbor hoppings are given as @xmath66 here the tnn stands for the third neighbor .
we find that the fermi velocity linearly decreases by introducing @xmath20 and becomes @xmath67 at @xmath68 as @xmath69 , which is also shown in fig.[fig : t3 ] ( b ) .
though tuning nominal value of @xmath70 in the graphene is difficult , tuning of the fermi velocity at the dirac cone has been proposed in bilayer graphene by changing the relative orientation angle between two layers@xcite , which is effectively equivalent to the tuning of @xmath70 . and
fermi velocity @xmath71 at dirac point . here
, @xmath71 is shown by the ratio to the fermi velocity at @xmath72 . , width=321 ] figure [ fig : mfphase ] shows the phase diagram calculated by the mean - field approximation for @xmath73 , @xmath74 , and @xmath75 .
compared to the nel ordered phase and the tmi , the cdw phase is not largely affected by the fermi velocity .
therefore , the region of tmi recovers .
( a ) 0.3 , ( b ) 0.4 and ( c ) 0.45 .
sm denotes the semimetal .
, width=321 ] next , we examine the stability by using the mvmc method . in our mvmc method , we impose the translational symmetry on the variational parameters to reduce the computational cost , as @xmath76 here , @xmath77 is taken as any bravais vector of the 6 sublattice unit cell illustrated in fig.[fig : cdw6sub](a ) , in order to examine possible spontaneous symmetry breakings including the cdw shown in fig.[fig : cdw6sub](a ) and the tmi on an equal footing .
therefore the number of the sites for each calculation is taken as @xmath78 .
we have calculated for @xmath79 , where about 2500 variational parameters are used for the calculation of the largest size . to perform the extrapolation to the small external filed limit , @xmath80 , we have also calculated for several different strengths of the spin - orbit interaction , @xmath81 , and @xmath82 .
figure [ fig : vmcdata](a ) shows the numerical results for the order parameter of the tmi at @xmath83 and @xmath84 .
the sudden drops in @xmath28 around @xmath85 signals the emergence of the 6-sublattice cdw state . indeed ,
when @xmath27 further increases , @xmath28 vanishes .
this is physically quite natural because @xmath28 is interpreted as the loop current and hard to be stabilized inside the cdw phase where electrons are locked at specific sites . for @xmath86 at @xmath84 .
( b ) size extrapolation at @xmath87 for several different values of @xmath27 . here
, four different sizes ( @xmath79 ) are calculated .
lines are results of quadratic fittings .
( c)extrapolation of @xmath12 is shown for different values of @xmath27 .
lines are results of quadratic fitting . ,
width=321 ] same calculations are carried out for @xmath88 , which is employed in the size extrapolation of @xmath28 .
figure [ fig : vmcdata](b ) shows the size extrapolation at @xmath87 as a typical example .
since the cdw becomes dominant at @xmath85 , the results for @xmath89 are shown in fig.[fig : vmcdata](b ) .
when @xmath25 is small , @xmath28 empirically follows the size dependence @xmath90 with a constant @xmath91 as we see in fig.5(b ) .
this extrapolation is performed for four different values of @xmath12 , and then we get the values for @xmath92 .
these data are shown in fig .
[ fig : vmcdata](c ) , where the second extrapolation to @xmath54 is performed .
final results for @xmath93 are shown in fig .
[ fig : vmcresult](a ) , where the results of @xmath94 are also shown .
the errorbars are defined from the errors of the extrapolation , where the largest error of the first extrapolation is added to the errors of the second extrapolation .
we note that the error bars arising from the statistical errors of the monte carlo sampling are much smaller .
relatively large errorbars for small @xmath27 at @xmath94 is possiblly because of the existence of the critical point at a finite value of @xmath27 , which is about @xmath95 .
when @xmath27 exceeds this critical value and @xmath28 in the thermodynamic limit remains nonzero , the error becomes smaller as can be seen from fig .
[ fig : vmcresult](a ) . for @xmath83 , the order grows from small value of @xmath96 .
however , at @xmath94 , non - zero @xmath28 can not be detected for small values of @xmath27 .
though the estimate of the universality class from these data is difficult , theoretically it is expected to belong to that of the gross - neveu model@xcite , and our result does not contradict this criticality . for @xmath97 , we do not find the value of @xmath27 where @xmath28 in the thermodynamic limit remains nonzero , and therefore phase transitions is not expected .
this is shown in fig .
[ fig : vmcresult](b ) , where the size extrapolation at @xmath87 is shown .
there , @xmath28 becomes @xmath67 at @xmath98 for all @xmath27 , which is completely different from the behavior at @xmath94 and @xmath83 . the resulting phase diagram is shown in fig .
[ fig : ueffect](a ) .
although we expect that the mvmc results show larger critical values @xmath99 for the transition at @xmath100 , the estimated results indicate @xmath101 slightly smaller than the mean - field results shown in fig.[fig : mfphase ] .
the reason that @xmath101 becomes smaller at @xmath100 in the mvmc results is probably an artifact arising from a peculiar size dependence near the essential singularity at @xmath102 , as we see even in the mean - field calculation shown in fig .
[ fig : ueffect](b ) where , the size dependence in the mean - field calculation are shown .
the possible errors in in the estimate of @xmath28 is as large as @xmath103 and the resultant errors in the estimate of @xmath101 is around 0.1 .
therefore , the stability of the tmi phase over the cdw and nel phases in the region @xmath104 for @xmath100 is robust .
here we note that the boundary between the cdw and tmi phases around @xmath105 does not change when we take into account the quantum fluctuations ( by calculating with the vmc method ) , because it is a strong first - order transition .
are shown .
( b)size extrapolation at @xmath106 and @xmath87 for several different values of @xmath27 . here , four different sizes ( @xmath79 ) are calculated .
lines are results of quadratic fittings . , width=321 ] now we discuss the effect of the on - site coulomb repulsion . though the onsite coulomb repulsion does not affect the value of @xmath28 and therefore stability of the tmi in the mean - field approximation , our mvmc result shows that increasing @xmath56 decreases the value of @xmath28 if @xmath27 is fixed as shown in fig .
[ fig : ueffect](c ) .
while the increasing @xmath56 quantitatively decreases the value of @xmath28 , its effect does not destroy the stability of the tmi phase at @xmath107 at least if @xmath108 .
it also suppresses the emergence of the cdw .
therefore it may help to enlarge the region of the tmi phase .
the same effect is expected when we consider the coulomb repulsion for the nearest neighbor sites @xmath109 .
that is , it decreases the value of @xmath28 beyond the level of the mean - field approximation but does not essentially affect the phase transition .
however , we also note @xmath109 may cause another type of cdw , and stabilization of tmi should be examined against this cdw phase when @xmath109 is large . and @xmath70 at @xmath110 obtained by mvmc calculations .
( b)size dependence of mean - field calculation for several different values of @xmath27 at @xmath83 .
( c ) @xmath27 dependence of order parameter of tmi for @xmath111 and @xmath112 at @xmath113 .
decrease in @xmath28 is observed by introducing @xmath56 .
, width=321 ] as a qualitative difference from the mean - field approximation , we found that effect of @xmath56 decreases the value of the order parameter of the tmi . in the case of the border between the semimetal and the nel ordered phases ,
it is expected that @xmath114 , namely the critical values for @xmath56 becomes larger by treating fluctuation effects carefully .
furthermore , we expect that @xmath114 becomes a function of @xmath27 , which may enhance the fluctuation of the nel order and suppress the phase transition . in our calculation
, we did not find nel ordered states or mixed states of tmi and nel ordered states . on the other hand ,
the cdw phase is expected to be much more stable against fluctuations and the mean - field solution gives a reasonably good description because of a large scale of the energy gain for the cdw phase in comparison to other phases as adequately shown even by the mean - field approximation . at the boundary of the tmi to nel
ordered phases , the universality class may change as suggested in the kane - mele - hubbard model@xcite .
here , we discuss the realization of the tmi in the real solids . a primary candidate of tmis is graphene . as a well - known fact ,
graphene is nothing but a two - dimensional honeycomb network of carbon atoms and hosts dirac electrons .
however , it is also a well - known fact that , in free - standing graphene and graphene on substrates , significant single - particle excitation gaps have not been observed yet@xcite .
below , we examine possible routes toward the realization of the tmi in graphene - related systems .
first of all , as already studied above , the suppression of the fermi velocity of the dirac electrons is crucial for the stabilization of the tmis .
as extensively studied in the literature @xcite , twisted bilayer graphene ( tblg ) offers dirac electrons with tunable fermi velocities . by choosing stacking procedures , the quadratic band crossing , in other words , the zero fermi velocity limit ,
is also achieved , which has been already observed in experiments@xcite .
next , we need to clarify competitions with any other possible symmetry breakings in the tblg with the fermi velocity smaller than that of graphene .
the suppressed fermi velocities may possibly cause instabilities towards not only the tmis but also other competing orders as discussed in this paper .
for clarification of the competition we need an _ ab initio _ estimation of effective coulomb repulsions which directly correspond to the coulomb repulsions in the extended hubbard model@xcite .
the _ ab initio
_ study on the effective coulomb repulsions employs a many - body perturbation scheme called constrained random phase approximation ( crpa)@xcite . the crpa estimation gives the following values for the coulomb repulsions : the on - site and off - site coulomb repulsions are given as @xmath115 , @xmath116 , and @xmath117 with @xmath118 ev for free - standing graphene .
if we neglect longer - ranged coulomb repulsions , we expect the nel or cdw orders by employing these crpa estimates of @xmath119 , and @xmath25 in graphene and tblg .
therefore , the free - standing graphene and tblg do not offer a suitable platform for the tmis .
however , by choosing dielectric substrates , the strength of the coulomb repulsions , @xmath120 , and @xmath121 , is suppressed due to enhancement of dielectric constant as @xmath122 , and @xmath123 , where @xmath124 is defined by dielectric constants of each materials as @xmath125 .
( here , we ignored the possible reduction of the effective dielectric constants at small distances . ) then , we may approach the parameter region , where tmis become stable , as shown in fig.[fig : ueffect](b ) .
if we neglect @xmath126 , dielectric substrates with @xmath127 are enough to stabilize the tmis .
even when the nearest - neighbor coulomb repulsion @xmath126 is taken into account , tmi is expected to be stable as long as @xmath126 is not strong . in the above discussion , we neglected further neighbor coulomb repulsions , namely , the third neighbor coulomb repulsions @xmath128 , the fourth neighbor ones @xmath129 , and so on . to justify the above discussion , we need to screen the further neighbor coulomb repulsions by utilizing dielectric responses of the substrates and/or adatoms . here
we note that the screening from atomic orbitals on the same and neighboring sites effectively reduces the coulomb repulsions as extensively studied by using crpa@xcite .
the relative strength of the on - site and second neighbor coulomb repulsions , @xmath130 , may also be controllable by utilizing the adatoms , which are expected to efficiently screen the on - site coulomb repulsions if the adatoms is located just on top of the carbon atoms .
if we combine the control of the @xmath130 with the suppression of the further - neighbor coulomb repulsions , the above discussion may be relevant . by utilizing the adatoms ,
the nearest - neighbor coulomb repulsions @xmath126 are also expected to be well - screened by adatoms on nearest - neighbor carbon - carbon bonds .
the suppression of @xmath126 is helpfull for suppressing the cdws competing with the tmi . finally , we estimate the single - particle excitations gap @xmath131 induced by the tmi .
the excitation gap is crucial for actual applications of the tmi as a spintronics platform .
if we set @xmath132 and expect the tblg with @xmath133 ev , we obtain the excitation gap up to 0.1 ev , where we use the mean - field estimation of the gap @xmath134@xcite with the value of the mvmc result for @xmath28 and assumed this formula is valid in the presence of electron correlation .
the estimated gap scale is substantially larger than the room temperatures .
our theoretical results support that topological insulators with such a large excitation gap @xmath135 ev are possibly obtained by using abundant carbon atoms .
in this paper , we have studied the realization of the tmi phase for the electronic systems on a honeycomb lattice by using the mean - field calculation and mvmc method .
we found that the cdw of the 6 sublattice unit cell is much more stable than the previously estimated cdw with smaller unit cells for the simplest case where the electronic transfer is limited to the nearest neighbor pair .
for the stabilization of the tmi we need to suppress the fermi velocity at the dirac point than the standard dirac dispersion for the case with only the nearest neighbor transfer . in the case of the honeycomb lattice , this is realized by introducing the third neighbor hopping @xmath20 and we have given quantitative criteria for the emergence of the tmi .
related real material is a bilayer graphene where the fermi velocity is tuned by changing the rotation angle between two parallel layers@xcite .
actually , the quadratic band crossing is realized when the rotation angle is @xmath67 ( known as the ab stacking bilayer graphene@xcite ) , which is mimicked by @xmath136 . since smaller values of fermi velocity stabilizes the tmi at smaller values of @xmath27 ,
its effective control may offer a breakthrough in the realization of two dimensional tmis .
we need further analyses for experimental methods of controlling the stability of the tmis and _ ab initio _ quantitative estimates of the stability for bilayer graphenes , which are intriguing future subjects of our study .
the authors thank financial support by grant - in - aid for scientific research ( no .
22340090 ) , from mext , japan .
the authors thank t. misawa and d. tahara for fruitful discussions . a part of this research
was supported by the strategic programs for innovative research ( spire ) , mext ( grant number hp130007 and hp140215 ) , and the computational materials science initiative ( cmsi ) , japan . | realization and design of topological insulators emerging from electron correlations , called topological mott insulators ( tmis ) , is pursued by using mean - field approximations as well as multi - variable variational monte carlo ( mvmc ) methods for dirac electrons on honeycomb lattices .
the topological insulator phases predicted in the previous studies by the mean - field approximation for an extended hubbard model on the honeycomb lattice turn out to disappear , when we consider the possibility of a long - period charge - density - wave ( cdw ) order taking over the tmi phase .
nevertheless , we further show that the tmi phase is still stabilized when we are able to tune the fermi velocity of the dirac point of the electron band . beyond the limitation of the mean - field calculation
, we apply the newly developed mvmc to make accurate predictions after including the many - body and quantum fluctuations . by taking the extrapolation to the thermodynamic and weak external field limit
, we present realistic criteria for the emergence of the topological insulator caused by the electron correlations . by suppressing the fermi velocity to a tenth of that of the original honeycomb lattice ,
the topological insulator emerges in an extended region as a spontaneous symmetry breaking surviving competitions with other orders .
we discuss experimental ways to realize it in a bilayer graphenesystem . |
closely spaced double layer electron systems possess an additional , layer degree of freedom , which in certain conditions stabilizes ground states with no counterpart in the single layer case .
notable examples include fractional quantum hall states ( qhs ) at even denominator fillings , such as @xmath0 @xcite and @xmath1 @xcite , or a peculiar qhs at total filling factor @xmath2 ( layer filling factor 1/2 ) @xcite .
the @xmath2 qhs in interacting double layers displays striking transport properties such as enhanced inter - layer tunneling @xcite and counterflow superfluidity @xcite , and has been likened to a bcs exciton condensate @xcite .
dipolar superfluidity has been posited to also occur at zero magnetic field @xcite in spatially separated , closely spaced two - dimensional electron and hole systems , thanks to the pairing of carriers in opposite layers .
although remarkable progress has been made in the realization of high mobility electron - hole bilayers @xcite , an unambiguous signature of electron - hole pairing remains to be experimentally observed .
the common thread in these phenomena is the inter - layer coulomb interaction being comparable in strength to the intra - layer interaction , leading to many - particle ground states involving the carriers of _ both _ layers .
the emergence of graphene @xcite as an electronic material has opened fascinating avenues in the study of the electron physics in reduced dimensions .
thanks to its atomically thin vertical dimension , graphene allows separate two - dimensional electron systems to be brought in close proximity , at separations otherwise not accessible in other heterostructures , and tantalizing theoretical predictions are based on this property @xcite . in light of these observations , it is of interest to explore electron physics in closely spaced graphene double layers . here
we discuss the fabrication , and key electron transport properties in this system , namely individual layer resistivity and coulomb drag .
we introduce a model to describe the layer density dependence on gate and inter - layer bias , and show that calculations agree well with experimental results in zero and high magnetic fields .
coulomb drag measurements reveal two distinct regimes : ( i ) diffusive drag at elevated temperatures , and ( ii ) mesoscopic fluctuations - dominated drag at low temperatures .
while we focus here on graphene double layers separated by a thin metal - oxide dielectric , a system with which the authors are most familiar with @xcite , we also note recent progress in graphene double layers separated by hexagonal boron nitride @xcite .
the fabrication of independently contacted graphene double layers starts with the mechanical exfoliation from natural graphite of the bottom graphene layer onto a 280 nm thick sio@xmath3 dielectric , thermally grown on a highly doped si substrate .
electron beam ( e - beam ) lithography , metal ( ni or cr - au ) contact deposition followed by lift - off , and o@xmath3 plasma etching are used to define a hall bar device .
the al@xmath3o@xmath4 inter - layer dielectric is then deposited by atomic layer deposition ( ald ) , and using an 2 nm thick evaporated al film to nucleate the ald growth . the total inter - layer dielectric thickness for the samples used our study ranges from 4 nm to 9 nm . to fabricate the graphene top layer , a second monolayer graphene is mechanically exfoliated on a sio@xmath3/si substrate .
after spin - coating poly(metyl metacrylate ) ( pmma ) on the top layer and curing , the underlying sio@xmath3 substrate is etched with naoh , and the top layer along with the alignment markers is detached with the pmma membrane .
the pmma membrane is then aligned with the bottom layer device , and a hall bar is subsequently defined on the top layer , completing the graphene double layer .
we focus here on data collected from two samples , labeled 1 and 2 , both with a @xmath5 nm thick al@xmath3o@xmath4 inter - layer dielectric , and with an inter - layer resistance larger than 1 g@xmath6 .
the layer mobilities are @xmath710,000 @xmath8/v@xmath9s for both samples .
the layer resistivtities are measured using small signal , low frequency lock - in techniques as function of back - gate bias ( v@xmath10 ) , and inter - layer bias ( v@xmath11 ) applied on the top layer .
the bottom layer is maintained at the ground ( 0 v ) potential during measurements .
the data discussed here are collected using a pumped @xmath12he refrigerator with a base temperature @xmath13 k. o@xmath4 dielectric .
the two layer densities can be independently controlled using the back - gate ( @xmath14 ) bias along with the inter - layer ( @xmath15 ) bias applied on the top layer .
lower right : optical micrograph of a graphene double - layer device .
the red ( blue ) contour marks the bottom ( top ) layer .
the scale bar is 5 @xmath16 m .
( b),(c ) band diagram of a graphene double layer under an applied back - gate [ panel ( b ) ] or inter - layer [ panel ( c ) ] bias . adapted from ref .
[ 19 ] . ] to understand the layer resistivity dependence on gate and inter - layer bias , it is instructive to examine a band structure model which relates the applied @xmath14 and @xmath15 biases to the top ( @xmath17 ) and bottom ( @xmath18 ) layer densities [ figs .
1(b , c ) ] .
the applied @xmath14 can be written as the sum of the electrostatic potential drop across the bottom sio@xmath3 dielectric and the fermi energy of the bottom layer : @xmath19 @xmath20 represents the fermi energy of graphene relative to the charge neutrality ( dirac ) point at a carrier density @xmath21 ; @xmath21 and @xmath20 are positive ( negative ) for electrons ( holes ) .
@xmath22 is the sio@xmath3 dielectric capacitance per unit area .
similarly , an applied @xmath15 can be written as the sum of the electrostatic potential drop across the al@xmath3o@xmath4 dielectric , and the fermi energy of the two layers : @xmath23\ ] ] a positive @xmath15 applied on the top layer induces electrons ( holes ) in the bottom ( top ) layer , which explains the negative sign for the right hand side terms in eq .
while the above equations implicitly assume that the two graphene layers are charge neutral at @xmath24 v , a finite doping of the two layers can be included in the above model using additive constants to the left hand side in eqs .
alternatively , @xmath14 and @xmath15 can be referenced with respect to the bias values which render both layers charge neutral , a convention which we adopt in discussing our experimental results @xcite . and @xmath15 .
( b ) longitudinal resistivity measured in the bottom ( left panel ) and top ( right panel ) layer in sample 1 at @xmath13 k. adapted from ref . [ 19 ] . ]
figure 2(a ) data show contour plots of @xmath25 and @xmath26 calculated according to eqs .
( 1,2 ) , with the following dependence of the fermi energy on density @xmath27 , consistent with masseless particles with a fermi velocity @xmath28 .
the bottom sio@xmath3 dielectric capacitance is @xmath29 nf@xmath9@xmath30 , a value determined experimentally by probing the capacitance of metal pads located in proximity of the graphene double layer and confirmed by hall measurements .
the inter - layer capacitance value used here is @xmath31
nf@xmath9@xmath30 @xcite , and the fermi velocity @xmath32 cm / s .
figure 2(a ) data show that the charge neutrality point of the bottom layer follows an almost linear dependence on @xmath14 and @xmath15 , while the density of the top layer is controlled primarily by the inter - layer bias , and to a much lesser extent by the back - gate bias . to better understand these observations , let us first neglect @xmath20 in eqs .
( 1,2 ) : @xmath26 is controlled by @xmath15 only , and @xmath18 depends linearly on @xmath14 and @xmath15 , with @xmath22 and @xmath33 respectively as proportionality coefficients .
although this behavior resembles fig .
2(a ) data , noticeable departures can be observed as a result of the reduced density of states in graphene , and consequently non - negligible @xmath20 .
first , @xmath26 does depend on @xmath14 , implying an incomplete screening of the back - gate induced electric field by the bottom layer , or equivalently a finite density of the states in the bottom layer , hence finite @xmath34 .
second , the @xmath35 line shows a noticeable non - linearity as a function of @xmath14 and @xmath15 near zero gate bias , a consequence of the non - negligible fermi energy of the top layer .
figure 2(b ) data show contour plots of the longitudinal resistivities @xmath36 ( left panel ) and @xmath37 ( right panel ) measured in the bottom and top layers , respectively , at a temperature @xmath13 k. a comparison with fig .
2(a ) data reveals a very good agreement between the experimental data and calculations , validating the model of eqs .
interestingly , this model can be used to make a direct measurement of the fermi energy in graphene , by employing one of the two layers as a resistively detected kelvin probe @xcite . indeed , setting @xmath38 in eq .
( 2 ) yields @xmath39 .
the latter relation implies that the inter - layer bias required to bring the top layer to the charge neutrality point is equal to the fermi energy of the bottom layer expressed in units of ev .
consequently , tracking the top layer charge neutrality point in fig .
2(b ) ( right panel ) yields the bottom layer fermi energy as a function of @xmath14 .
we next address the layer density dependence on @xmath14 and @xmath15 in a perpendicular magnetic field ( @xmath40 ) . in the presence of a @xmath40-field
the electrons occupy landau levels ( ll ) with energies @xmath41 , where @xmath42 is an integer denoting the ll index . including the spin and valley degrees of freedom each ll has a degeneracy of @xmath43 .
quantum hall states ( qhss ) develop at filling factors @xmath44 .
neglecting ll disorder - induced broadening , the fermi energy dependence on density therefore writes @xmath45 , where @xmath46 $ ] and @xmath47 is the nearest integer function .
figure 3(a ) data show contour plots of @xmath25 ( left panel ) and @xmath26 ( right panel ) calculated as a function of @xmath14 and @xmath15 at @xmath48 t using eqs .
( 1,2 ) , along with the same capacitance values used in fig .
2(a ) : @xmath29 nf@xmath9@xmath30 , @xmath31 nf@xmath9@xmath30 . in addition
, a finite broadening is assumed for each ll using a lorentzian density of states distribution with a width @xmath49 mev for the @xmath50 ll and @xmath51 mev for @xmath52 @xcite .
figure 3(b ) data show the measured @xmath36 ( left panel ) and @xmath37 ( right panel ) as a function of @xmath14 and @xmath15 in sample 1 , at @xmath13 k. the data display alternating regions of vanishing resistivity corresponding to quantum hall states , and high resistivity regions corresponding to half filled lls . a comparison with fig .
3(b ) data reveals a very good agreement of the charge neutrality point dependence on @xmath14 and @xmath15 for both layers , effectively validating eqs .
( 1,2 ) model in high magnetic fields .
similar to fig .
2(b ) data , the @xmath15 value when the top layer is charge neutral is equal to the fermi energy of the bottom layer .
consequently , the top layer charge neutrality line of fig .
2(b ) traces the bottom layer fermi energy as a function of @xmath14 , and shows the staircase dependence corresponding to discrete lls .
an important attribute of a double layer system is the ability to probe the coulomb drag between the two layers . a charge current ( @xmath53 ) flown in one ( drive ) layer results in a net momentum transfer to the opposite ( drag ) layer , thanks to the coulomb interaction between electrons in the two layers .
if no current is allowed to flow in the drag layer , a voltage @xmath54 builds up in order to counter the momentum transfer .
the drag resistivity is defined as @xmath55 , where @xmath56 and @xmath57 are the length and width of the sample area over which the drag voltage is measured . in effect , the drag resistivity is proportional to the scattering rate between electrons located in opposite layers .
the dependence of @xmath58 on @xmath25 , @xmath26 , @xmath59 , and the spacing @xmath60 between the two layers can provide insight into the interaction between the two layers , as well as the ground state in individual layers .
coulomb drag between closely spaced carrier systems has been probed in a variety of gaas / algaas heterostructures which , depending on the type of sample used , allowed experimental access to electron - electron drag @xcite , hole - hole drag @xcite , and electron - hole drag @xcite .
assuming that the ground state in each layer is a fermi liquid , and that the layer spacing is sufficiently large such that the fermi wave - vector @xmath61 , the drag resistivity depends on the layer densities , temperature , and spacing as @xmath62 . in fig .
4 we show contour plots of @xmath58 as a function of @xmath25 , and @xmath26 , measured in sample 1 at different temperatures .
the data were measured by flowing current in the bottom layer , and measuring the drag voltage in the top layer as a function of @xmath14 and @xmath15 , which were subsequently converted into layer densities using eqs .
we note that because the drag voltage is orders of magnitude smaller than the longitudinal voltage drop in the drive layer , very small inter - layer leakage currents can introduce artifacts in the experimental data ; care was taken here to ensure the measured drag voltage is not affected by inter - layer leakage .
figure 4 data explore @xmath58 in the four quadrants of the @xmath63 plane , and reveals several noteworthy findings .
first , @xmath58 is negative ( positive ) for same ( opposite ) carrier polarity .
second , @xmath58 increases with increasing @xmath59 .
third , @xmath58 vanishes at large @xmath64 values , and in the vicinity of the @xmath65 points .
lastly , the @xmath58 vs. @xmath63 data in different quadrants are very similar , except for the sign change depending on the carrier type .
as we explain below all these observations are consistent with coulomb drag of masseless fermions where the ground state in each layer is a fermi liquid .
several theoretical studies examined the coulomb drag in graphene to date @xcite . in the weak - coupling limit , defined as @xmath66 ,
the drag resistance scales as @xcite @xmath67 in the strong - coupling limit , defined as @xmath68 , @xmath58 is expected to have a weaker dependence on layer density , and to be independent of the layer spacing as @xcite @xmath69 a close examination of fig .
4 data reveals a @xmath59-dependence which is stronger than @xmath70 , but weaker than @xmath71 , in the @xmath59-range probed here . a best fit to @xmath58 vs. @xmath64 along the diagonals @xmath72 , and @xmath73 yields @xmath74 , with an exponent @xmath75 little dependent on temperature .
the @xmath76-value is lower than @xmath77 , expected in the weak - coupling limit , but larger than 1 expected in the strong - coupling limit .
this can be readily understood since @xmath78 varies between 0.4 and 1 in the density range where @xmath58 is not vanishingly small . at low densities , where @xmath79 disorder - induced puddles in the two layers lead to a decrease of @xmath58 towards zero .
the experimental @xmath58 dependence on @xmath25 , @xmath26 , and @xmath59 is in good agreement with the expected theoretical dependence of diffusive coulomb drag in the fermi liquid regime , although a quantitative matching between experiment and theory remains to be established .
as the temperature is reduced below 50 k , the drag resistivity reveals a dramatic departure from the diffusive coulomb drag of fig .
4 . in fig .
5(a ) we show a contour plot of @xmath58 measured as a function of @xmath14 and @xmath15 , at @xmath80 k in sample 2 , and using the top ( bottom ) layer as the drag ( drive ) layer .
unlike the diffusive coulomb drag of fig .
4 , the data of fig .
5(a ) show a pattern of large @xmath58 fluctuations centered around a zero average .
the mesoscopic fluctuations increase in amplitude with reducing @xmath59 , and completely obscure the diffusive drag at low temperatures @xcite .
the @xmath58 mesoscopic fluctuations emerge as a result of phase coherent transport at low temperature , and represent the counterpart of universal conductance fluctuations in coulomb drag @xcite .
similar mesoscopic fluctuations have been observed in coulomb drag probed between two - dimensional electron systems in gaas / algaas heterostructures , albeit at lower temperatures @xcite . a comparison of fig .
5(a ) data with the carrier density contour plot of fig .
2(a ) calculations reveal an interesting observation . the locus of constant @xmath58 in fig .
5(a ) corresponds to the lines of constant @xmath17 in the @xmath81 plane of fig .
alternatively stated , the @xmath58 mesoscopic fluctuations track the drag layer constant density lines in the @xmath81 plane .
to better illustrate this observation in fig .
5(b ) ( left panel ) we show a contour plot of @xmath58 vs. @xmath82 ; these data represent fig .
5(a ) @xmath58 data with @xmath81 axes converted into @xmath82 using eqs . ( 1,2 ) . the right panel of fig .
5(b ) ( right panel ) shows the drag conductivity @xmath83 vs. @xmath82 .
we note that the @xmath84 fluctuation amplitude has the same order of magnitude as @xmath85 .
figure 5(b ) data manifestly show that the coulomb drag mesoscopic fluctuations depend mainly on the drag layer density ( @xmath26 ) , and are largely insensitive to the drive layer density ( @xmath25 ) .
this observation is manifestly at variance with the onsager reciprocity relation , according to which interchanging the drag and drive layers should not affect the measured @xmath58 .
examination of fig .
5(b ) data reveals a regular , almost periodic @xmath58 vs. @xmath26 pattern . to better illustrate this pattern , in fig .
6 we show @xmath58 vs. @xmath26 measured at a fixed drive layer density , @xmath86 @xmath30 , and at @xmath80 k. the data reveals an almost periodic dependence of @xmath87 vs. @xmath26 reminiscent of fabry - perot interference @xcite .
a simple analysis can relate the top layer fermi energy ( @xmath88 ) change corresponding to @xmath58 maxima to a cavity length ( @xmath56 ) , via @xmath89 . a typical value for @xmath88 of 10 mev deduced from fig . 6 data , corresponds to @xmath90 @xmath16 m , a value comparable to the phase coherence length in graphene on sio@xmath3 @xcite at @xmath80 k. vs. @xmath26 measured at a fixed bottom layer density @xmath86 @xmath30 in sample 2 , at @xmath80 k and using the top ( bottom ) layer as the drag ( drive ) layer .
the top axis represents the fermi energy of the drag layer .
the data show an almost periodic fluctuation pattern . ]
in summary , we present a magnetotransport and coulomb drag study in graphene double layers , consisting of two independently contacted graphene monolayers separated by a thin al@xmath3o@xmath4 dielectric .
the coulomb drag probed in this system reveals two regimes : ( i ) diffusive drag at elevated ( @xmath91 k ) temperatures , and ( ii ) mesoscopic fluctuations dominated drag at low temperatures .
the temperature dependence of the diffusive drag is consistent with the fermi liquid theory , while the density dependence suggests the layers are close to the strong - coupling regime @xcite .
the coulomb drag mesoscopic fluctuations observed at low temperature depend mainly on the drag layer density , and are largely insensitive to the drive layer density , an observation which is at variance with the onsager reciprocity relation .
kun yang , k. moon , l. zheng , a.h .
macdonald , s.m .
girvin , d. yoshioka , and s.c .
zhang , phys .
lett . * 72 * , 732 ( 1994 ) ; k. moon , h. mori , kun yang , s.m .
girvin , a.h .
macdonald , l. zheng , d. yoshioka , and s.c .
zhang , phys .
b * 51 * , 5138 ( 1995 ) .
the @xmath96 ratio is determined from the slope of the bottom layer charge neutrality point traced as a function of @xmath14 and @xmath15 .
@xmath33 is determined using the @xmath96 ratio and the measured @xmath22 value .
we note that @xmath33 consists of the al@xmath3o@xmath4 dielectric capacitance plus series contributions associated with the graphene - dielectric interfaces ; b. fallahazad _ et al . _ ,
. lett . * 100 * , 093112 ( 2012 ) . | we review the fabrication and key transport properties of graphene double layers , consisting of two graphene monolayers placed in close proximity , independently contacted , and separated by an ultra - thin dielectric .
we outline a simple band structure model relating the layer densities to the applied gate and inter - layer biases , and show that calculations and experimental results are in excellent agreement both at zero and in high magnetic fields .
coulomb drag measurements , which probe the electron - electron scattering between the two layers reveal two distinct regime : ( i ) diffusive drag at elevated temperatures , and ( ii ) mesoscopic fluctuation - dominated drag at low temperatures .
we discuss the coulomb drag results within the framework of existing theories .
= 1 graphene , double layer , coulomb drag , quantum hall 73.22.pr , 73.43.-f , 73.22.gk , 71.35.-y |
terrestrial planets and cores of giant planets are generally considered to have formed through accretion of many small bodies called planetesimals . to simulate accretion processes of planets ,
two methods , which are complementary to each other , have been applied .
the first one is @xmath0-body simulations in which orbits of all bodies are numerically integrated and gravitational accelerations due to other bodies are calculated in every time step ( e.g. , kokubo and ida , 1996 ; chambers and wetherill , 1998 ; richardson et al . , 2000 ;
morishima et al . , 2010 ) .
@xmath0-body simulations are accurate and can automatically handle any complicated phenomena , such as resonant interactions and spatially non - uniform distributions of planetesimals .
gravity calculations are accelerated by such as tree - methods ( richardson et al . , 2000 ) or special hardwares ( kokubo and ida , 1996 ; grimm and stadel , 2014 ) , and a large time step can be used with sophisticated integrators , such as mixed variable symplectic ( mvs ) or wisdom - holman integrators ( duncan et al . , 1998 ; chambers , 1999 ) .
even with these novel techniques , @xmath0-body simulations are computationally intense and the number of particles or the number of time steps in a simulation is severely limited .
the second approach is statistical calculations in which planetesimals are placed in two dimensional ( distance and mass ) eulerian grids , and the time evolutions of the number and the mean velocity of an ensemble of planetesimals in each grid are calculated using the phase - averaged collision and stirring rates ( e.g. , greenberg et al . , 1978 ; wetherill and stewart , 1989 , 1993 ; inaba et al . , 2001 ; morbidelli et al . , 2009 ; kobayashi et al . ,
while this approach does not directly follow orbits of individual particles , it can handle many particles , even numerous collisional fragments .
largest bodies , called planetary embryos , are handled in a different manner than small bodies , taking into account their orbital isolation . the mutual orbital separation between neighboring embryos is usually assumed to be 10 mutual hill radii .
the last assumption is not always guaranteed , particularly in the late stage of planetary accretion ( e.g. , chambers and wetherill , 1998 ) . to handle orbital evolution of embryos more accurately , eulerian hybrid codes - body calculations . ]
have been developed ( spaute et al . , 1991 ; weidenschilling et al .
, 1997 ; bromley and kenyon , 2006 ; glaschke et al . , 2014 ) , in which small planetesimals are still handled by the eulerian approach whereas orbital evolutions of largest embryos are individually followed by such as @xmath0-body integrations .
gravitational and collisional interactions between embryos and small planetesimals are handled using analytic prescriptions .
glaschke et al .
( 2014 ) took into account radial diffusion of planetesimals due to gravitational scattering of embryos and their code can approximately handle gap opening around embryos orbits .
a lagrangian hybrid method has also been introduced by levison et al .
( 2012 ) ( ldt12 hereafter ) . in their lipad code , a large number of planetesimals
are represented by a small number of lagrangian tracers .
this type of approach is called a super - particle approximation and is also employed in modeling of debris disks ( kral et al . , 2013 ; nesvold et al . , 2013 ) and planetesimal formation ( johansen et al
. , 2007 ; rein et al . , 2010 ) .
orbits of individual tracers are directly followed by numerical integrations , and interactions between planetesimals ( stirring and collisions ) in tracers are handled by a statistical routine .
embryos are represented by single particles and the accelerations of any bodies due to gravity of embryos are handled in the @xmath0-body routine .
lagrangian hybrid methods have several advantages than eulerian hybrid methods .
for example , lagrangian methods can accurately handle spatial inhomogeneity in a planetesimal disk ( e.g. , gap opening by embryos ) , planetesimal - driven migration , resonant interactions between embryos and small planetesimals , and eccentric ringlets .
a drawback of lagrangian methods would be computational cost as orbits of all tracers need to be directly integrated .
therefore , it is desirable that lagrangian hybrid methods can handle various problems accurately even with a moderately small number of tracers . in this paper , we develop a new lagrangian hybrid code for planet formation . while we follow the basic concept of particle classes introduced by ldt12 , recipes used in our code are different from those used in the lipad code in many places .
the largest difference appears in the methods to handle viscous stirring and dynamical friction .
the lipad code solves pair - wise interactions while our method gives the accelerations of tracers using the phase - averaged stirring and dynamical friction rates .
while the lipad code conducts a series of three body integrations in the stirring routine ( in the shear dominant regime ) and in the routine of sub - embryo migration during a simulation , our code avoids them by using unique routines .
the complete list of comparison with the lipad code turns out to be rather long and is given in appenedix g. in section 2 , we explain our method . in section 3 ,
we show various tests of the new hybrid code and compare them with analytic estimates and pure @xmath0-body simulations .
the computational speed and limitations of our code are discussed in section 4 .
the summary is given in section 5 .
for the sake of clarity , specific derivations are deferred to appendices .
. schematic illustration of our hybrid code . accelerations due to gravity of embryos are handled by the @xmath0-body routine .
tracer - tracer interactions and back reaction of tracers on sub - embryos including collision are handled by the statistical routine .
the particle classes in our hybrid code are the same as those introduced by ldt12 .
the code has three classes of particles : tracers , sub - embryos , and full - embryos ( fig . 1 ) . in the present paper
, we do not consider fragmentation of planetesimals and dust - tracers are not introduced .
a tracer represents a large number of equal - mass planetesimals on roughly the same orbits .
the mass of a planetesimal and the number of planetesimals in the tracer @xmath1 ( @xmath1 for the index of the tracer ) are defined to be @xmath2 and @xmath3 .
therefore , the tracer mass is given by @xmath4 . through collisional growth ,
@xmath2 increases and @xmath3 decreases while @xmath4 remains close to its original value .
we allow mass exchanges between tracers through collisions so the tracer mass , @xmath4 , is not necessarily fixed but has the upper and lower limits , @xmath5 and @xmath6 ( @xmath7 ) .
we employ @xmath8 and @xmath9 in this paper , where @xmath10 is the minimum mass of a sub - embryo .
the mass @xmath10 is usually the initial mass of a tracer , which is the same for all tracers at the beginning of a simulation . if @xmath11 and @xmath12 , the tracer is promoted to a sub - embryo . if @xmath13 , the sub - embryo is further promoted to a full - embryo , where we employ the numerical factor @xmath14 of 100 , as recommended by ldt12 .
the number @xmath3 is an integer in our model .
orbits of any types of particles are directly integrated .
we use a mixed variable symplectic integrator known as symba ( duncan et al . , 1998 ) , which can handle close encounters between particles .
this integrator is also used by ldt12 .
the collisional and gravitational interactions between full - embryos with tracers are directly handled in every time step of orbital integrations , as is the case of pure @xmath0-body simulations .
on the other hand , tracer - tracer interactions are handled in a statistical routine which is described in subsequent sections in great detail . while the time step for the orbital integration @xmath15 is @xmath16 of the orbital period , the time step @xmath17 for the statistical routine can be taken to be much larger as long as @xmath18 is sufficiently smaller than the stirring and collisional timescales . in this paper , we employ @xmath19 for all test simulations .
we confirmed that almost the same outcome is obtained even with a smaller @xmath18 for all the test simulations . in interactions between a tracer and a full - embryo in the @xmath0-body routine ,
the tracer is assumed to be a single particle with the mass equivalent to the total mass of planetesimals in the tracer .
this is not a problem as long as the embryo is sufficiently massive compared with tracers .
however , if the embryo mass is similar to tracer masses as is the case immediately after its promotion , and if embryo - tracer interactions are handled by the direct @xmath0-body routine , the embryo suffers artificially strong kicks from tracers . to avoid this issue , sub - embryos are introduced .
accelerations of sub - embryos due to gravitational interactions with tracers are handled by the statistical routine whereas accelerations of tracers due to gravitational interactions with sub - embryos are handled by the direct @xmath0-body routine ( fig . 1 ) .
collisions of planetesimals in tracers with sub - embryos are also handled in the statistical routine to avoid artificially large mass jumps in sub - embryos , contrary to ldt12 ( see appendix g for more discussion ) .
fig . 2 .
schematic illustration of neighboring tracer search .
the filled circle is the target tracer and open circles are interloping tracers ( interlopers ) .
the surface number density of planetesimals in an interloper is simply given by the number of planetesimal in the interloper divided by the area of the region between two arcs .
our statistical routine uses the surface number density of nearby tracers .
we first describe how we determine the surface number density for tracer - tracer interactions .
the surface number density of tracers for interactions between a sub - embryo and tracers is derived in a similar way , but slightly modified so that close and distant encounters are treated differently ( section 2.5 ) .
consider a target tracer @xmath1 , surrounded by nearby interloping tracers .
the cylindrical co - ordinate system @xmath20 is introduced and we consider a curved region , called the region @xmath1 , located at @xmath21 and centered at the position of the tracer @xmath1 ( fig . 2 ) . the radial half width and the angular half width in the @xmath22 direction of the region @xmath1 are given by @xmath23 and @xmath24 , respectively .
the area of this region is given by @xmath25 for neighboring tracer search , tracers are sorted into 2d @xmath26 grids and only nearby cells of the target tracer are checked whether other tracers are in the designated region . if the interloper @xmath27 is in the region @xmath1 , the surface number density of planetesimals in the interloper @xmath27 around the target @xmath1 is simply expressed as @xmath28 the density @xmath29 is an instantaneous density . at other times , the interloper @xmath27 may not be in the region @xmath1 but other interlopers having planetesimal masses similar to @xmath30 may be in the region @xmath1 .
after averaging over time , the encounter rate with planetesimals in a certain mass range is expected to converge regardless of the size of @xmath31 .
a small @xmath32 is favorable to capture a small radial structure accurately .
however , it is not reasonable to use @xmath32 smaller than a typical radial length of gravitational influence of a planetesimal , which is usually about @xmath33 . here
, @xmath34 is the hill radius given by @xmath35 where @xmath36 is the semi - major axis of the tracer @xmath1 and @xmath37 is the mass of the central star .
since the mass of a possible largest planetesimal is @xmath10 , we employ @xmath38 for tracer - tracer interactions . for interactions between a sub - embryo and tracers ,
a larger @xmath23 needs to be used and we employ 10 hill radii of a possible largest sub - embryo with the mass of @xmath39 . an advantage of the choice of @xmath23 given by eq .
( [ eq : drt0 ] ) is that we can omit the procedure for orbital isolation of planetesimals .
orbital isolation of the largest bodies occurs if the sum of orbital spaces of these bodies is less than the width of the region of interest ( wetherill and stewart , 1993 ) . if a planetesimal is assumed to occupy an orbital space of 10 hill radii ( kokubo and ida , 1998 ) and if the tracer masses are @xmath10 , the sum of the orbital spaces of planetesimals in a pair of tracers are always larger than @xmath40 .
since our model allows a tracer mass lower than @xmath10 , isolation can potentially occur if the tracer masses and the numbers of planetesimals in the tracers are both small .
such cases are rather rare and we can ignore them .
therefore , orbital isolation can occur only for embryos and that is handled by the @xmath0-body routine .
an additional discussion is given in appendix g. a small @xmath24 is favorable to handle non - axisymmetric structures , in which the surface density is not azimuthally uniform .
such structures usually appear if external massive bodies exist ( kortenkamp and wetherill , 2000 ; nagasawa et al .
, 2005 ; levison and morbidelli , 2007 ; queck et al . , 2007 ) .
a small @xmath24 can also be chosen , if we want to suppress the number of interlopers to save computational costs ( section 4.1 ) . in this paper , we usually adopt @xmath41 , but dependency on @xmath24 is also examined . even if the tracer @xmath27 is in the region @xmath1 ,
the tracer @xmath1 is occasionally outside the region @xmath27 used in neighboring search around the tracer @xmath27 . to make sure that the tracer @xmath1 and the tracer @xmath27 are mutually counted ( or excluded ) as interlopers
, the number density @xmath29 is evaluated using eq .
( [ eq : nj ] ) only if @xmath42 or @xmath43 for @xmath44 .
otherwise and only if the tracer @xmath1 is in the region @xmath27 , @xmath29 is given by @xmath45 , where @xmath46 is the area of the region @xmath27 .
although we only considered encounters with interlopers , planetesimals in the target @xmath1 may encounter with other planetesimals in the same target if @xmath47 .
appendix a describes how we correct this effect .
this effect is found to be too small to be identified .
let the semimajor axis , the orbital eccentricity , the orbital inclination , the longitudes of pericenter and ascending node , and the mean longitude of the target be @xmath36 , @xmath48 , @xmath49 , @xmath50 , @xmath51 , and @xmath52 , and those for the interloper be @xmath53 , @xmath54 , @xmath55 , @xmath56 , @xmath57 , and @xmath58 .
the difference between the semimajor axes is defined as @xmath59 .
when the target encounters with the interloper , they may collide with each other at a certain probability .
their orbital elements are also modified by mutual gravitational scattering . to handle these processes in the statistical routine in a simple manner , we make following assumptions . 1
. we use the collision probability and the mean change rates of orbital elements derived by previous studies which usually solved hill s equations .
the equations of motion are reduced to hill s equations if hill s approximation is applied ( hill , 1878 ; nakazawa et al . , 1989 ) .
the conditions to apply hill s approximation are 1 .
@xmath48 , @xmath49 , @xmath54 , @xmath55 @xmath60 , 2 .
@xmath2 , @xmath30 @xmath61 , and 3 .
2 . the semimajor axis differences @xmath63 and the phases of the interlopers , @xmath56 , @xmath57 , @xmath58 are uniformly distributed .
namely , if a certain interloper in the region @xmath1 is found , we assume that other interlopers with the same @xmath54 , @xmath55 , but different @xmath63 , @xmath56 , @xmath57 , and @xmath58 are distributed uniformly over the region @xmath1 .
3 . the longitudes , @xmath50 and @xmath51 , of the target change randomly and large enough on the evolution timescales of @xmath48 and @xmath49 .
this approximation is unnecessary if approximation 2 is valid and either @xmath64 or @xmath65 , because the collision and stirring rates do not depend on @xmath50 and @xmath51 in such a situation .
assumption 1 ( hill s approximation ) seems to work well in most of cases of planetary accretion .
greenzweig and lissauer ( 1990 ) showed that the collision probability of a planet with planetesimals on circular and co - planar orbits agrees with that based on hill s approximation within @xmath66 as long as @xmath67 .
our test simulations for late stage planetary accretion ( section 3.5 ) indicate that hill s approximation also works well even at moderately high orbital eccentricities ( @xmath68 ) .
assumption 2 is acceptable if many planetesimals mutually interact via gravitational scattering and collisions without orbital isolation .
these interactions lead to randomization of spatial positions and phases .
we extensively adopt these uniformities even if the target is a sub - embryo isolated from other embryos , but with some approximated cares ( section 2.6 ) .
assumption 3 is likely to be reasonable since the timescale of phase change is comparable to the stirring timescale in general ( e.g , tanaka and ida , 1996 ) . with assumptions 2 and 3
, we can assume that the time averaged stirring and collision rates of an individual target are equivalent to the stirring and collision rates averaged over phase space .
while this equivalence resembles the so - called ergodic hypothesis , the difference is that we assume the equivalence on a short timescale .
although assumption 3 may not be strictly correct , it greatly simplifies the problem and it is worthwhile to examine its validity or usefulness .
fig . 3 .
definitions of the eccentricity vectors .
the inclination vectors are defined in a similar manner .
integrations over distributions of orbital elements of interlopers can be transformed to integrations over relative orbital elements of interlopers with respect to the target ( nakazawa et al . , 1989 ) .
the eccentricity and inclination vectors of the target @xmath1 are given as @xmath69 the same characters but with the subscript @xmath27 represent the eccentricity and inclination vectors for the interloper @xmath27 .
the relative eccentricity and inclination vectors are given as ( fig .
3 ) @xmath70 where @xmath71 and @xmath72 are the longitudes for the relative vectors .
nakazawa et al .
( 1989 ) adopted assumptions 1 - 3 and derived the number of planetesimals @xmath73 colliding with the target @xmath1 during a time interval @xmath17 as [ their eq .
( 49 ) ] @xmath74 where @xmath75 and @xmath76 are the mean squares of @xmath54 and @xmath55 of interlopers , @xmath29 is the surface number density of interlopers , @xmath77 is the mean semi - major axis , @xmath78 is the orbital frequency at @xmath79 , @xmath80 is the distribution function of @xmath81 and @xmath82 normalized as @xmath83 , and @xmath84 is the non - dimensional collision rate averaged over @xmath63 , @xmath71 , and @xmath72 ( appendix b.1 ) . in the above , the reduced mutual hill radius , @xmath85 , is given as @xmath86 the expression of @xmath87 for the case of the rayleigh distributions of @xmath54 and @xmath55 is given by eq .
( 35 ) of nakazawa et al .
( 1989 ) . in our method , on the other hand , @xmath87 for a certain interloper is given by the dirac delta function .
thus , the expected number of planetesimals @xmath73 in the interloper @xmath27 colliding with the target @xmath1 during @xmath17 is given as @xmath88 where @xmath29 is given by eq .
( [ eq : nj ] ) .
the stirring rates can be derived in a similar manner using the non - dimensional stirring rates ( section 2.5 ) . in our method , the averaging over distributions of eccentricities and inclinations of interlopers will be made by averaging over multiple interlopers .
although it is not necessary to assume the distribution functions of eccentricities and inclinations of interlopers in calculations of the collision and stirring rates , our stirring routine naturally reproduces rayleigh distributions ( section 3.2 ) . if the system is perturbed by a massive external body and the forced eccentricity is much larger than free eccentricities , the phases , @xmath56 and @xmath57 , may not be uniformly distributed .
even in such a case , our approach may still be applied by replacing the eccentricity and inclination vectors by the free eccentricity and inclination vectors whose directions are uniformly distributed , provided that the forced eccentricity vectors are roughly the same between the target and interlopers .
thus , while uniformities of @xmath71 and @xmath72 are assumed , we do not necessarily assume uniformities of the relative longitudes @xmath89 and @xmath90 . the averaging over @xmath91 and @xmath92 is done by averaging over interlopers [ @xmath93 . while this relaxing of the assumption does not alter eq .
( [ eq : dndt ] ) , it is important for dynamical friction ( section 2.5 )
. the forced eccentricity vectors are not necessarily the same between the target and interlopers , for example , if the mass - dependent dissipative force works .
in such a situation , our approach is not valid , although it is still worthwhile to examine how inaccurate it is .
the expected change rate of the mass @xmath2 of a planetesimal in the tracer @xmath1 due to collisions with planetesimals in the interloper @xmath27 is simply given by multiplying the planetesimal mass @xmath30 to eq .
( [ eq : dndt ] ) ( ida and nakazawa , 1989 ; greenzweig and lissauer , 1990 ; inaba et al . ,
2001 ) @xmath94 to avoid double counting , we only consider a case for @xmath42 or @xmath43 for @xmath44 .
using the mass change rate , the probability that the mass of a planetesimal in the target @xmath1 increases by @xmath95 due to a collision with a planetesimal in the interloper @xmath27 during the time step @xmath17 is given as @xmath96 .
if this probability is larger than a random number which takes between 0 and 1 , we assume that all planetesimals in the target @xmath1 collide with planetesimals in the interloper .
the same procedure is applied to all other interlopers in the region @xmath1 to check whether collisions occur between planetesimals in the target and those in interlopers .
if a collision occurs and mutually colliding bodies merge , the changes in the masses and the numbers of planetesimals in the tracers are given as follows : @xmath97 where the characters with the additional subscript @xmath98 mean those after the merging .
note that the total mass is conserved in the collisions : @xmath99 . before determining whether a collision occurs or not , we need to choose an appropriate @xmath95 .
the mass @xmath95 is usually the mass of the interloping planetesimal @xmath30 , but can take different values for convenience as follows : @xmath100 & \mbox{case ( a ) ; $ k_{jb } > 0 $ } , \\ k_j m_{j}/k_i & \mbox{case ( b ) ; $ k_{jb } = 0 $ } , \end{array}\right .
\label{eq : delm}\ ] ] where int ( ) means the decimals are dropped off for the number in the parentheses .
the form in the upper row of eq .
( [ eq : delm ] ) means that if @xmath30 is smaller than the threshold mass , we do not consider individual collisions but the accumulated mass change of the target planetesimal .
we adopt the threshold mass of @xmath101 .
we first estimate the post - collision mass of the interloper @xmath102 using eq .
( [ eq : inte ] ) and @xmath95 for case ( a ) of eq .
( [ eq : delm ] ) for any interlopers .
if the resulting mass , @xmath102 , is found to be smaller than the lower limit of the tracer mass , @xmath6 , we use @xmath95 of case ( b ) given by the lower row in eq .
( [ eq : delm ] ) instead of that of case ( a ) .
if a collision occurs for case ( b ) , all the planetesimals in the interloper @xmath27 merge with those in the target @xmath1 and the interloper is deleted .
therefore , the mass of any tracer is kept to be lager than @xmath6 .
while we adopt @xmath103 , this value can be arbitrarily chosen : a small lower limit may be able to give accurate mass evolution while increase of the number of tracers can be suppressed with a large lower limit .
the procedure to change masses and orbital elements of planetesimals due to collisional merging is schematically summarized in fig . 4 . for case ( a ) in eq .
( [ eq : delm ] ) , in which @xmath104 and @xmath105 , only some of planetesimals in the interloper @xmath27 collide with planetesimals in the target @xmath1 . in this case
, we split the interloper @xmath27 into two tracers : the interloper @xmath27 and the impactor .
the number of planetesimals in the interloper after the split is given by @xmath106 in eq .
( [ eq : inte ] ) and the position and velocity of the post - split interloper are unchanged during the collision between the target and the impactor . for case
( b ) , all the planetesimal in the interloper are involved in the collision and we call the interloper the impactor . in both cases , the planetesimal mass and the number of planetesimals in the impactor
are given by @xmath95 and @xmath3 , respectively , and the collision process is almost the same except some small differences .
schematic illustration of the procedure to handle collisional merging between planetesimals in two tracers ( the target and the interloper ) .
the figure shows how the masses of planetesimal ( @xmath2 and @xmath30 ) , the numbers of planetesimals in the tracers ( @xmath3 and @xmath107 ) , the semimajor axes ( @xmath36 and @xmath53 ) , and the orbital eccentricities ( @xmath48 and @xmath54 ) change during the procedure .
when the target and the impactor are judged to collide , they do not physically contact each other .
even their stream lines do not usually cross each other . since we assume uniformity in semimajor axis of an interloper , it is possible to make two stream lines cross each other by changing the semi - major axis of the impactor without changing other orbital elements .
the crossing point of the stream lines is the location where the impact takes place .
this impact point can be found as follows .
first , @xmath22 and @xmath108 of the target and the impactor are aligned by changing the true anomaly of the target @xmath109 by @xmath110 and that for the impactor @xmath111 by @xmath112 .
the angular changes , @xmath110 and @xmath112 , are derived by solving the following set of equations : @xmath113 where @xmath114 is the longitude of the impacting position measured from the reference direction @xmath115 .
( [ eq : xx0 ] ) and ( [ eq : xx ] ) give two solutions for @xmath110 mutually shifted by @xmath116 , and we choose one with a smaller absolute value .
then , the true anomalies are changed by moving the colliding pair along their keplerian orbits .
second , the semimajor axis of the impactor is scaled by @xmath117 .
this is done by scaling the impactor position by @xmath117 and the velocity by @xmath118 .
the positions of the target and the impactor become almost identical through these procedures without changing their eccentricity and inclination vectors .
then , the merging between the target and the impactor takes place . the position and velocity of the target after the merging with the impactor
are given by the mass weighted mean values ( i.e. , the mass center for the position ) .
after that , the semi - major axis of the target is adjusted to @xmath119 by scaling the position and the velocity so that the @xmath120-component of the total orbital angular momentum of the target and the interloper is conserved during the merging : @xmath121 where @xmath122 and @xmath123 are the orbital eccentricity and inclination of the target after the merging . if the target tracer mass is equal to or exceeds the upper limit , @xmath124 , we split it into two tracers .
if @xmath125 and @xmath126 , we also split the tracer to avoid simultaneous emerging of two sub - embryos .
the numbers of planetesimals in the split tracers are given by @xmath127 if @xmath128 is even , and @xmath129 if @xmath128 is odd .
we give weak velocity changes , symmetrically in momenta , to the two new tracers so that their relative eccentricity and inclination are about @xmath85 .
this will avoid an encounter velocity artificially too low between them in the next encounter .
the semi - major axes of the two tracers are adjusted so that the orbital separation between two tracers is randomly chosen between 0 to @xmath23 but the @xmath120-component of the total orbital angular momentum is conserved . for case
( b ) , we set the semi - major axis of one of the new tracers be the one for the deleted interloper instead of choosing the orbital separation randomly . finally ,
the true anomalies of the two tracers ( or one if no splitting is made ) are changed by @xmath130 and @xmath131 . through the processes described above , tracer masses remain between @xmath6 and @xmath5 .
the number of tracers remains unchanged or increases for case ( a ) of eq .
( [ eq : delm ] ) whereas it remains unchanged or decreases for case ( b ) . as a result , the total number of tracers remains similar to the initial value even after many collisions between planetesimals .
our several experiments show that increase of the tracer number from the initial value is suppressed less than 30@xmath132 in most cases .
our current code also has an option of hit - and - run collisions and the procedure is described in detail in appendix e. in this case , neither the number of planetesimals nor that of tracers changes .
the orbital elements of planetesimals evolve due to gravitational interactions between them .
the interactions are separated into viscous stirring ( vs ) and dynamical friction ( df ) .
viscous stirring increases eccentricities and inclinations in expense of the tidal potential ( i.e. , radial diffusion ) and dynamical friction leads to energy equipartition between particles ( ida , 1990 , stewart and ida , 2000 ; ohtsuki et al . , 2002 ) .
dynamical friction can be separated into stirring and damping parts ( dfs and dfd ) and sometimes the latter part only is called dynamical friction ( binney and tremaine , 1987 ; rafikov , 2003 ; ldt12 ) .
we handle the stirring and damping parts of dynamical friction separately as shown below .
the evolution of eccentricity @xmath48 and inclination @xmath49 of the target @xmath1 due to viscous stirring and dynamical friction caused by interactions with the interloper @xmath27 is given as ( ida , 1990 ; tanaka and ida , 1996 ) @xmath133 where @xmath134 , @xmath135 , @xmath136 , @xmath137 , and @xmath138 are the non - dimensional rates of viscous stirring and dynamical friction .
the expressions of these rates are given in appendix b.2 .
since we do not assume uniformities of @xmath89 and @xmath90 to handle eccentric and/or inclined ringlets , as discussed in section 2.3 , the terms with these relative angles remain in the expressions for dynamical friction .
the way to split dynamical friction into the two terms is explained in detail in appendix c. the evolutions of @xmath48 due to viscous stirring and dynamical friction are then given as @xmath139 similar expressions can be obtained for @xmath49 .
mutual encounters between planetesimals causes random walk in their semimajor axes .
the radial diffusion coefficient @xmath140 for the target @xmath1 is approximately given by ( appendix d ) @xmath141 .
\label{eq : di0}\ ] ] ignoring the effect of curvature , the probability function of the change in @xmath36 after the time step @xmath17 is given by @xmath142 where the function is normalized as @xmath143 .
we choose @xmath144 at random with a weight given by @xmath145 . if @xmath146 , we set @xmath147 .
this is rare but can occur if @xmath148 .
the stirring and damping rates of orbital elements need to be converted into the acceleration of the tracer . since the time step for the statistical routine @xmath17 is usually much smaller than the stirring time , @xmath149 , the absolute values of the changes in the orbital elements during @xmath17 are much smaller than the magnitudes of the orbital elements themselves .
consider a situation that encounters between planetesimals occur randomly in directions of velocities .
a certain acceleration occurs to a planetesimal in an encounter , and in another encounter an acceleration with the same magnitude to the first encounter but in the opposite direction can occur . in first order ,
the changes of orbital elements cancel out in these two encounters . however , in second order , the orbital eccentricity and inclination increase a little on average , because the orbital elements are slightly changed after the first encounter .
this is the way how viscous stirring works , and to handle it , we introduce second order gauss planetary equations .
the stirring part of dynamical friction is also handled by these equations .
on the other hand , the damping part of dynamical friction is handled by the standard gauss planetary equations of first order ( e.g. , murray and dermott , 1999 ) .
if we use the standard gauss planetary equations for the stirring part of dynamical friction , the standard deviations of orbital eccentricity and inclination become unphysically small for equal - mass planetesimals or mass - dependent segregation of these orbital elements occurs for a case of a mass distribution .
we assume that @xmath150 and @xmath151 in the following formulation .
note that these assumptions are already used in hill s approximation . in this subsection
only , we drop the subscript @xmath1 for the tracer @xmath1 to avoid long , confusing subscripts .
the expected changes of orbital elements during the time step @xmath17 are @xmath152 , @xmath153 , @xmath154 , @xmath155 , and @xmath156 , where @xmath157 , @xmath158 , @xmath159 , and similar expressions are obtained for the inclination .
the change of the velocity vector is split into three components as @xmath160 where @xmath161 and @xmath162 are the components of the velocity change in the radial and tangential directions in the orbital plane , @xmath163 is the velocity change normal to the orbital plane , @xmath164 , @xmath165 , and @xmath166 are unit vectors .
the velocity vector of the tracer itself is given as @xmath167 , where @xmath168 and @xmath169 are the radial and tangential components .
all the components of the velocity change are further split into the stirring and damping terms with subscripts s and d , respectively : @xmath170 in our stirring routine , all the components of the stirring part are chosen using random numbers , but their averaged functional forms follow gaussian distributions with the zero mean values : @xmath171 , where the angle brackets in this subsection mean averaging over multiple velocity changes . the acceleration due to an individual encounter may not necessarily follow gaussian . on the other hand , since we handle the time averaged accelerations of tracers due to multiple encounters that occur in a random fashion , their distributions are limited to gaussian .
first we consider the velocity changes in the orbital plane .
the energy conservation gives the following equation , if we take up to the second order terms @xmath172 where @xmath173 .
if we ignore the second order terms and the first term in the r.h.s . of eq .
( [ eq : enec ] ) as @xmath174 , this gives @xmath175 as @xmath176 note again that @xmath152 is given by eq .
( [ eq : pda ] ) and thus @xmath175 is gaussian .
the angular momentum conservation gives @xmath177 where @xmath178 is the change of @xmath179 and @xmath180 is the distance from the central star .
inserting eq .
( [ eq : angc ] ) into eq .
( [ eq : enec ] ) , we have @xmath181 averaging eq .
( [ eq : angc2 ] ) over time , the first order terms vanish and we have only the second order terms . using @xmath182 , we have @xmath183 where we used @xmath184 and @xmath185 [ @xmath186 ; eq .
( [ eq : di0 ] ) or eq .
( [ eq : db2 ] ) ] , and ignored the term proportional to @xmath187 as it is small enough as compared with other terms for @xmath188 . inserting @xmath189 into eq .
( [ eq : de2 ] ) , we have @xmath190 we give @xmath191 assuming it is gaussian .
if the r.h.s side of eq . ( [ eq : vrs ] ) is negative , we set @xmath192 , add the r.h.s divided by @xmath193 to @xmath194 , and re - calculate @xmath154 , which will be handled as shown below .
this occurs in the shear dominant regime [ @xmath195 . to give the velocity changes @xmath196 and @xmath197 for the damping parts of dynamical friction , we use the standard first order gauss planetary equation for @xmath179 ( murray and dermott , 1999 ) that can also be derived from eq .
( [ eq : angc2 ] ) : @xmath198 , \label{eq : dedf}\ ] ] where @xmath87 and @xmath199 are the true and eccentric anomalies .
we employ the following forms @xmath200 setting @xmath201 in eq .
( [ eq : dedf ] ) , we have @xmath202 where we assumed @xmath203 .
next , we consider the velocity change perpendicular to the orbital plane .
the specific angular momentum vector @xmath204 after the velocity change @xmath163 in the inertial cartesian coordinate system are given as @xmath205 where @xmath206 , @xmath207 , @xmath208 , and @xmath209 are the rotation matrices given by eqs .
( 2.119 ) and ( 2.120 ) of murray and dermott ( 1999 ) .
using @xmath210 , the @xmath120-component of eq .
( [ eq : ang4 ] ) is given as @xmath211 where @xmath212 is the change of @xmath1 and @xmath213 is the argument of pericenter .
this equation leads to the standard gauss planetary equation for @xmath1 ( murray and dermott , 1999 ) , if we ignore the second order terms .
averaging eq .
( [ eq : ang5 ] ) over time , the first term in the r.h.s . vanishes . setting the mean change of @xmath214 as @xmath215 and using @xmath216 , we have @xmath217 we give @xmath218 , assuming it is gaussian . for the damping part of dynamical friction
, we adopt the form @xmath219 ignoring the second order terms in eq .
( [ eq : ang5 ] ) , and setting @xmath220 , we have @xmath221 where we took averaging over the orbital period and used @xmath222 .
if the planetesimal mass @xmath2 is equal to or exceeds @xmath10 , the tracer is promoted to a sub - embryo .
the acceleration of a sub - embryo due to gravitational interactions with a single planetesimal in a tracer is taken into account in the @xmath0-body routine to hold the correct mutual hill radius .
on the other hand , the acceleration of the sub - embryo due to interactions with other @xmath223 planetesimals in the tracer is handled by the statistical routine .
collisions of planetesimals with sub - embryos are also handled in the statistical routine .
a large embryo tends to form a gap around in its feeding zone ( tanaka and ida , 1997 ) , although other nearby embryos , if they exit , tend to fill the gap .
the surface density in the feeding zone usually decreases through this process .
we take into account a possible difference of surface densities inside and outside of the feeding zone whereas spatial variation of the surface density inside of the feeding zone is not considered .
this assumption allows us to still use the stirring and collision rates averaged over phases and semi - major axis for interactions between a sub - embryo and tracers .
a collision between the sub - embryo @xmath1 and the tracer @xmath27 occurs only if @xmath224 , where @xmath225 is the jacobi integral of hill s equations given by @xmath226 where @xmath59 and @xmath227 is the half width of the feeding zone given by @xmath228 even if @xmath224 , collisions do not occur with trojans or if @xmath229 is too small ( see fig . 2 of ohtsuki and ida , 1998 ) .
ida and nakazawa ( 1989 ) showed that collisions occur only if @xmath230^{1/2}a_ih_{ij},\ ] ] where @xmath231 is the minimum orbital separation for occurrence of collision .
we do not consider interactions with trojans in the statistical routine .
planetesimals in the feeding zone ( @xmath224 and @xmath232 ) can potentially enter the hill sphere of the embryo and orbits of planetesimals are strongly deflected by close encounters with the embryo . for tracer - tracer interactions ,
the number density is calculated by eqs .
( [ eq : ss ] ) and ( [ eq : nj ] ) .
if a sub - embryo is the target , however , we search for planetesimals in the feeding zone .
their surface number density is given by @xmath233 we miss some of tracers in the feeding zone if they are radially outside of the region @xmath1 of neighboring search .
the factor @xmath234 in eq .
( [ eq : nj2 ] ) is introduced to correct this effect , where @xmath235 is the orbital period of the tracer @xmath27 , and @xmath236 is the period in which the tracer @xmath27 is radially in the region @xmath1 during @xmath235 provided that the tracer @xmath27 is azimuthally in the range of the region @xmath1 .
the expression of @xmath236 is given in appendix f. in the stirring routine , we use @xmath223 in eq .
( [ eq : nj2 ] ) instead of @xmath107 as the back reaction from a single planetesimal is taken into account in the @xmath0-body routine .
using eq .
( [ eq : nj2 ] ) , collisional growth and velocity changes of sub - embryos are handled in the statistical routine in the same manner as tracer - tracer interactions .
if there is no spatial density variation between the inside and outside of the feeding zone , the same collision and stirring rates are derived whether the target is a tracer or a sub - embryo .
even if the mutual distance between the centers of a tracer and a sub - embryo is less than the sum of the radii of the constituent planetesimal and the sub - embryo , orbital integrations of both bodies are continued in the @xmath0-body routine using softened forces that avoid singularity .
we employ the k1 kernel of dehnen ( 2001 ) for the softening .
gravitational interactions occur even if @xmath237 due to distant encounters ( weidenschilling 1989 , hasegawa and nakazawa , 1990 ) .
we evaluate the change rates of the orbital elements of the sub - embryo due to individual distant encounters without assuming uniformity in @xmath63 .
let the changes of the squared relative orbital eccentricity and inclination due to a single distant encounter be @xmath238 and @xmath239 . hasegawa and nakazawa ( 1990 ) derived the changes @xmath240 and @xmath241 averaged over @xmath71 and @xmath72 [ their eqs .
( 38 ) and ( 39 ) ] .
the averaged changes of @xmath242 and @xmath243 of the sub - embryo @xmath1 during a single encounter with a planetesimal in the interloper @xmath27 are given by @xmath244 and @xmath245 ( ida , 1990 ) .
using the keplerian shear velocity given by ( bromley and kenyon , 2011 ) @xmath246 the frequency of encounters is given as @xmath247 , where we used eq .
( [ eq : nj ] ) and @xmath23 in this case is the radial extent of planetesimals in the tracer @xmath27 and is assumed to be infinitesimally small .
the change rates of @xmath242 and @xmath243 due to distant encounters with the interloper @xmath27 are then given as @xmath248 where @xmath249 and @xmath250 .
the rates given by eqs .
( [ eq : dedis ] ) and ( [ eq : didis ] ) are added to eqs .
( [ eq : devs ] ) and ( [ eq : divs ] ) if @xmath237 .
dynamical friction due to distant encounters is negligible ( ida , 1990 ) .
an embryo gravitationally scatters surrounding planetesimals .
back reaction on the embryo causes its radial migration .
individual gravitational encounters usually cause a random walk of the embryo .
however , if the accumulated torque is not cancelled out , the embryo migrates over a long distance .
this is called planetesimal - driven migration ( ida et al . , 2000 ;
kirsh et al . ,
planetesimal - driven migration of full - embryos is automatically taken into account in the @xmath0-body routine .
we developed a unique method to handle planetesimal - driven migration of sub - embryos explained in the following . in our code , the angular momentum change @xmath251 of the sub - embryo @xmath1 due to gravitational interactions with tracers during @xmath17 is stored in the @xmath0-body routine . if @xmath251 is released instantaneously by giving the tangential acceleration , the sub - embryo may migrate unrealistically too fast due to strong kicks from tracers that have masses similar to the sub - embryo .
the direction of migration may change rapidly in such a situation .
this problem can be avoided by limiting the angular momentum released during @xmath17 to the theoretical prediction , @xmath252 , derived from the statistical routine . if @xmath253 , @xmath251 is fully released .
if @xmath254 , on the other hand , the remaining portion @xmath255 is added to @xmath251 stored during the next time step interval .
what is done by the scheme is basically time averaging of the torque on the sub - embryo over a timescale much longer than @xmath17 so that strong spikes of the torque due to close encounters are smoothed out .
more explanations are given in section 3.3 , together with test results .
the predicted angular momentum change , @xmath252 , is determined from the surface density and velocities of planetesimals in neighboring tracers .
bromley and kenyon ( 2011 , 2013 ) and ormel et al . ( 2012 ) showed three different modes of planetesimal - driven migration , referred to as type i iii migration analogous to gas - driven migration .
type i and iii migration occurs when an embryo interacts with planetesimals in close encounters .
type i migration is considered to occur when the velocity dispersion of planetesimals is large , and its migration rate is very slow as is determined by the difference between torques from outer and inner disks . on the other hand , type iii migration is fast self - sustained migration that occurs when the velocity dispersion of planetesimals is relatively small , and the embryo migrates due to the torque from the one side ( inner or outer side ) of the disk , leaving a relatively enhanced disk in the other side ( ida et al . , 2000 ;
kirsh et al .
, 2009 ; levison et al . , 2010 ; capobianco et al . , 2011 ) .
we ignore type i migration in the evaluation below , as the migration rate is well represented by the formula for type iii migration even at large velocity dispersions ( kirsh et al .
, 2009 ; bromley and kenyon , 2011 ) .
type ii migration occurs when an embryo opens up a gap in the planetesimal disk so the torque is primarily from planetesimals in distant encounters .
this torque is particularly important if the asymmetries of the surface density and the velocity dispersion between the inner and outer disks are large .
such a situation occurs if an embryo in a gaseous disk shepherds outer small planetesimals that rapidly spiral toward the central star without the embryo .
the radial migration rate of type iii planetesimal - driven migration due to an interaction with the tracer @xmath27 is ( ida et al . , 2000 )
@xmath256 where the sign is positive if @xmath257 and vice versa , and @xmath258 is the non - dimensional migration rate given in appendix b.3 . if @xmath259 stored in the @xmath0-body routine is positive / negative , we count only tracers with semimajor axes larger / smaller than @xmath36 because planetesimals in close encounters cause attractive forces . in the evaluation of @xmath29 in eq .
( [ eq : dat3j ] ) , we use @xmath260 instead of @xmath31 used in eq .
( [ eq : nj ] ) .
the type iii migration rate is then given as @xmath261 where @xmath262 is the factor that represents slowing down for massive embryos .
the slowing down occurs when the embryo can not migrate over the feeding zone during the synodic period for that width , as cancellation of angular momentum exchange occurs in the second encounter ( ida et al . , 2000 , kirsh et al . , 2009 ) .
bromley and kenyon ( 2011 ) showed that the slowing down factor is given by a migration length during the synodic period for the half - width of the feeding zone @xmath263 relative to @xmath263 , if this ratio is less than unity : @xmath264 .
\label{eq : slow}\ ] ] we employ @xmath265 after bromley and kenyon ( 2011 ) .
this length is somewhat shorter than the actual half - length of the feeding zone , @xmath227 [ eq . ( [ eq : de0 ] ) ] , and the reason for this value being applied is because the strong attractive force comes from a relatively narrow region . if the embryo mass is small enough , @xmath266 ; in this regime , the migration rate is independent of the embryo mass . one may think that the migration rate can be simply estimated by summing up the torques from both sides in eqs .
( [ eq : dat3j ] ) and ( [ eq : dat3 ] ) rather than summing up for one side and that @xmath259 calculated in the @xmath0-body routine is unnecessary .
however , we find that such an approach can work only if the embryo is massive enough so that @xmath267 . in this case
, the all tracers in the feeding zone strongly interact with the embryo and this produces large velocity asymmetry between the unperturbed side and the other side that the embryo migrated through . on the other hand ,
if the embryo is small , it migrates over the feeding zone through interactions with a small fraction of tracers in the feeding zone , and only small velocity asymmetry is produced .
in such a situation , the evaluated torques from both sides roughly cancel out each other and the fast migration can not be reproduced .
if we evaluate the torques from tracers in an azimuthally limited region , where velocity asymmetry exists , summation of both sides may work .
however , an appropriate choice of the azimuthal width is unclear to us and the number of tracers in a narrow region is too few to give a reliable migration rate .
this is why we evaluate the migration rate due to the one - sided torque as an upper limit , while the actual angular momentum change is evaluated in the @xmath0-body routine . for type ii migration
, we evaluate the torques from tracers in both sides of the disk as done for the stirring rate due to distant encounters . the change in the semimajor axis difference , @xmath268 , during a single encounter for @xmath269
is given as ( bromley and kenyon , 2011 ) @xmath270 where @xmath271 and @xmath272 .
this formula is expected to be independent of @xmath81 and @xmath82 after averaged over @xmath71 and @xmath72 ( hasegawa and nakazawa , 1990 ) . in a similar manner of eq .
( [ eq : dedis ] ) , the change rate of the semimajor axis of the embryo is given as @xmath273 the type ii migration rate is then given as @xmath274 if the surface density is symmetric for the inner and outer disk , the first term in the parentheses in the r.h.s .
( [ eq : dat20 ] ) vanishes after summation . in this case
, the embryo slowly migrate inward .
if density asymmetry exists , the first term remains and relatively fast migration takes place in either inward or outward direction .
the migration rate of the embryo suggested from the statistical routine is given as @xmath275 the predicted angular momentum change during @xmath17 is given as @xmath276 where @xmath277 is the angular momentum of the sub - embryo . to produce smooth type iii migration , at least a few , preferably @xmath278
, tracers need to exist constantly in the feeding zone in the region @xmath1 . if we choose a very small @xmath24 for the region @xmath1 , that makes sub - embryo migration noisy .
we find that a large @xmath259 sometimes remains unreleased even when an embryo reaches a disk edge .
this usually causes the embryo migrate artificially further away from the disk edge . to avoid this problem ,
we remove @xmath259 quickly on the timescale of @xmath279 orbital periods at the disk edge , although this violates conservation of the total angular momentum of the system .
we regard the embryo being at the disk edge if the total mass of tracers in the feeding zone ahead of the embryo in the migrating direction is less than the embryo mass . for a small embryo ,
sometimes no tracer exists in the feeding zone due to statistical fluctuations even if the embryo is not at the disk edge . to avoid this issue , we also count tracers within @xmath280 ahead of the embryo even if they are not in the feeding zone .
the global gravitational forces of tracers are added to the equations of motion for tracers and sub - embryos . to calculate the forces , we assume that the disk is axisymmetric and the vertical distribution is given by gaussian for tracers . for simplicity , at each radial location
, we consider only one scale height of tracers with various planetesimal masses whereas in reality different mass groups have different scale heights .
the scale height of planetesimals in a radial grid at @xmath281 is given by @xmath282 where the summation is done for each radial bin , and the total mass of the tracers @xmath283 in the grid is @xmath284 the surface density of tracers are given by @xmath285 , where @xmath286 is the radial grid width .
the spatial density of tracers is given as @xmath287 the @xmath281 and @xmath120 components of the tracers global gravitational force are given by ( nagasawa et al .
, 2000 ) @xmath288 dr'dz ' \nonumber \\ f_{{\rm tr } , z}(r , z ) & = & -4 g \int^{\infty}_{-\infty } \int_{r ' } \frac{\rho_{\rm tr}(r',z')r'(z - z')}{\sqrt{(r+r')^2 + ( z - z')^2 } } \nonumber \\ & & \times \left[\frac{e(\xi)}{(r - r')^2 + ( z - z')^2}\right]dr'dz',\end{aligned}\ ] ] where @xmath289 is the gravitational constant , @xmath290 and @xmath291 are the first and second kind of elliptic integrals , and @xmath292 is given by @xmath293 these components of the force are tabulated in ( @xmath294 ) grids and interpolated for use in simulations . for the accelerations on sub - embryos
, @xmath295 is used instead of @xmath3 in eqs .
( [ eq : htr ] ) and ( [ eq : mtr ] ) .
it is not necessary to calculate the global gravitational force components frequently and we usually update them every @xmath296 times the shortest orbital period of particles .
the global gravitational force modifies the precession rates of the longitudes of pericenter and ascending node ; it usually accelerates the absolute precession rates .
the precession rates do not usually matter to dynamics of tracers in terms of such as encounter velocities .
these rates are , however , likely to be important for stability of the system , when multiple sub - embryos appear in the system .
the embryos mutually interact through secular resonances which cause oscillations in eccentricities and inclinations of embryos .
the global gravitational force reduces these oscillation periods . as a result ,
the maximum eccentricities and inclinations of sub - embryos during the oscillations are suppressed to relatively small values .
we identify this effect in simulations in section 3.5 ( results without the global gravitational force are not shown there , though ) .
although its effect is much smaller than dynamical friction , it does not seem to be negligible .
[ cols="^,^,^,^,^,^,^ " , ] table 1 .
parameters used for hybrid simulations .
theme is a very brief expression of what experiments are performed .
@xmath297 is the initial total mass of planetesimals .
region represents the initial inner and outer edges of the planetesimal disk .
@xmath298 is the power - law exponent for the surface density of planetesimals ( @xmath299 ) .
@xmath0 is the initial number of planetesimals .
@xmath300 is the initial total number of tracers .
@xmath301 is the initial number of embryos ( @xmath302 means that no embryo is introduced in the initial state but embryos appear as a result of collisional growth ) .
@xmath303 is the initial root - mean - square eccentricity of tracers .
@xmath15 is the time step for orbital integrations .
@xmath24 is the angular half - width of the region used in neighboring search ( fig . 2 ) .
collision represents collisional outcome .
gas disk represents whether damping effects due to a gas disk are taken into account . in all the simulations , the physical density of any body is 2.0 g @xmath304 .
the initial distributions of @xmath179 and @xmath1 are given by the rayleigh distributions with @xmath303 = 2@xmath305 .
we show six different types of tests of the hybrid code : collisional damping and stirring ( section 3.1 ) , gravitational scattering and radial diffusion ( section 3.2 ) , planetesimal - driven migration of sub - embryos ( section 3.3 ) , runaway to oligarchic growth ( section 3.4 ) , accretion of terrestrial planets from narrow annuli ( section 3.5 ) , and accretion of cores of jovian planets ( section 3.6 ) .
the parameters used in the test simulations are summarized in table 1 .
time evolution of @xmath303 and @xmath305 of planetesimals due to mutual collisions only for ( a ) @xmath306 = 0 and ( b ) @xmath306 = 0.8 .
all effects related to mutual gravity are ignored . the total number of planetesimals is @xmath307 represented by 2000 ( solid curves ) and 500 ( dotted curves ) tracers .
planetesimals are placed between 0.95 au and 1.05 au and the total mass is 0.1 @xmath308 . in the hybrid simulations
, we only use tracers between 0.96 and 1.04 au in calculations of @xmath303 and @xmath305 to avoid the effect due to radial expansion .
evolution curves calculated using the analytic formula ( ohtsuki , 1999 ) are also shown by dashed curves for comparison . in the panel ( b ) , @xmath306 = 0.72 is used for the analytic curves . the first test we present is velocity evolution of tracers due to mutual collisions only to check whether the collision probability and the velocity changes at collisions are calculated accurately .
all mutual - gravity related effects are ignored in the test : gravitational scattering , gravitational focusing in the collisional probability [ the second term in the bracket of the r.h.s of eq .
( [ eq : pcol ] ) ] , and enhancement of the impact velocity ( appendix e.2 ) .
collisions are assumed to result in inelastic re - bounds ( hit - and - run collisions ) , and post collision velocities are characterized by the normal and tangential restitution coefficients , @xmath306 and @xmath309 .
appendix e.2 explains how to handle hit - and - run collisions between planetesimals in two tracers , with assumptions of @xmath310 and @xmath311 [ see eq .
( [ eq : vnd ] ) for more detail ] .
these coefficients , however , can be modified to any values . in the tests , we assume no friction ( i.e. , @xmath311 ) . collisional velocity evolution is a well - studied problem for planetary rings and several authors have developed analytical formulae ( goldreich and tremaine , 1978 ; hmeen - antilla , 1978 ; ohtsuki , 1992 , 1999 ) .
we compare our simulation results with analytic estimates of ohtsuki ( 1999 ) because of the easiness of use .
figure 5 shows the time evolution of @xmath303 and @xmath305 of @xmath307 planetesimals for ( a ) @xmath306 = 0 and ( b ) @xmath306 = 0.8 .
the radius of all planetesimals is 0.4 km .
the planetesimal disk extends from 0.95 to 1.05 au and its total mass is 0.1 @xmath308 .
these parameters give the disk normal optical depth of @xmath312 .
two simulations are performed to see resolution effects for both fig .
5a and 5b : one with @xmath313 and @xmath314 and another one with @xmath315 and @xmath316 , where @xmath300 is the total number of tracers . the total angular momentum , @xmath317 , is well conserved for all test simulations shown in fig . 5 ; @xmath318 . for @xmath310 ( fig .
5a ) , we find excellent agreements between our results and the analytic estimates of ohtsuki ( 1999 ) , in particular for the case of @xmath319 .
however , the decreases in @xmath303 and @xmath305 stall at certain values in hybrid simulations @xmath320 for @xmath319 ( 500 ) whereas in the analytic curves , @xmath179 and @xmath1 decrease down to @xmath321 , where @xmath322 is the particle radius . in the hybrid simulations , eccentric ringlets form , in which the longitudes of pericenter of all particles are nearly aligned . this issue is probably related to the initial condition . in creation of the initial positions and velocities , the longitudes of pericenter
are randomly chosen , and the sum of the eccentricity vectors of all tracers is not exactly zero .
the eccentricities of the ringlets in the end state are , in fact , close to the expected values of the residual eccentricity , @xmath323 . since the residual eccentricity decreases with increasing the number of tracers , the eccentricity of the ringlet in the end state decreases .
even if the residual eccentricity is fixed to be zero , the same problem is likely to occur unless all particles effectively interact with each other .
the ringlets are not only eccentric but also inclined because a similar discussion is applied to inclination .
the evolution curves of @xmath179 and @xmath1 also weakly depends on @xmath24 ; a small @xmath24 gives a better agreement with the analytic curves .
it is found that relative velocities are slightly underestimated if @xmath24 is large and radial excursions of particles are larger than @xmath23 .
this leads a slightly lower damping rate .
collisions cause not only damping but also stirring as the collisional viscosity converts the tidal potential to the kinetic energy .
if @xmath324 , where @xmath325 is the critical restitution coefficient , the stirring rate exceeds the damping rate and @xmath303 and @xmath305 increase with collisions ( goldreich and tremaine , 1978 ; ohtsuki , 1999 ) .
we find that our method is slightly more dissipative than the analytic estimate and gives @xmath326 for reasons that we were not able to figure out .
figure 5b shows the same as fig .
5a but for the case of @xmath327 .
it is found that our result for @xmath327 and @xmath313 ( 500 ) well agrees with the analytic estimate with @xmath328 ( 0.70 ) .
overall , our method can reproduce analytic results using a @xmath329 higher @xmath306 . in simulations of planetary accretion , we use @xmath310 for hit - and - run collisions and
the offset in @xmath306 is not a problem .
note that orbital alignments do not occur for the case of fig . 5b , because collisional stirring randomizes orbital phases .
fig . 6 . fig .
time evolution of @xmath303 and @xmath305 of planetesimals due to gravitational interactions : ( a ) @xmath330 , @xmath331 , @xmath332 , ( b ) @xmath333 , @xmath331 , @xmath332 , and ( c ) @xmath334 , @xmath335 , @xmath336 ( see table 1 for the characters ) .
the panel ( c ) is the case of the bimodal size distribution with @xmath337 , @xmath338 , @xmath339 , and @xmath340 , where the characters with the subscripts 1 and 2 are those for small and large planetesimals , and the mass ratio of a large planetesimal to a small one is 5 . in all cases ,
planetesimals are placed between 0.95 au and 1.05 au , and @xmath303 and @xmath305 are calculated using tracers between 0.96 au and 1.04 au .
evolution curves calculated using the analytic formula ( ohtsuki et al . , 2002 ) and from the pure @xmath0-body simulations ( only for the panels ( a ) and ( c ) ) are also shown for comparison .
fig . 7 .
snapshots on the @xmath341 plane for the simulations shown in fig .
6a : ( a ) the hybrid simulation and ( b ) the pure @xmath0-body simulation .
time evolutions of the standard deviation of @xmath342 for the simulations shown in fig . 7 .
results for six individual hybrid simulations are shown by colored dotted curves , whereas their mean value is shown by a black solid curve .
the result for the hybrid simulation shown in fig .
7a is a black dotted curve . for comparison , the result from the pure n - body simulation ( fig .
7b ) is shown by a black dashed curve .
the next test we present is velocity evolution of tracers due to mutual gravity only in the absence of any collisional effects .
figure 6 shows evolution of @xmath303 and @xmath305 for a planetesimal disk extending from 0.95 and 1.05 au .
the panels ( a ) and ( b ) are for mono - sized planetesimals represented by 200 tracers and the numbers of planetesimals @xmath0 are @xmath343 and @xmath344 , respectively .
the panel ( c ) is for a bimodal size distribution case ; 1500 small planetesimals and 200 large planetesimals are placed in the disk .
the mass ratio of a large planetesimal to a small one is 5 .
the number of tracers for small planetesimals is 150 and that for large planetesimals is 100 .
the results are compared with analytic estimates of ohtsuki et al .
for the panels ( a ) and ( c ) , we also performed pure @xmath0-body simulations using the parallel - tree code pkdgrav2 ( morishima et al . , 2010 ) with the same initial conditions .
the agreements between the analytic estimates and pure @xmath0-body simulation results are excellent , so results from both methods are trustable as benchmarks .
overall , we find good agreements between the results from the hybrid code and the analytic estimates , although our model gives slightly lower @xmath303 and @xmath305 than the analytic estimates , except fig .
6b shows a good agreement .
we also perform the same hybrid simulations with different values of @xmath300 and @xmath24 , and the absolute levels of @xmath303 and @xmath305 at a certain time are found to be almost independent of these parameters , unless @xmath300 and @xmath24 are too small .
the underestimates of @xmath303 and @xmath305 of our calculations is probably caused by the underestimates of @xmath136 and @xmath137 that we employ .
the analytic calculations of tanaka and ida ( 1996 ) give slightly lower @xmath136 and @xmath137 for @xmath345 and @xmath346 than those from direct three - body orbital integrations ( ida , 1990 , rafikov and slepian , 2010 ) .
the influence of this difference is less important for large @xmath347 and @xmath348 and that is why the hybrid simulation in fig .
6b shows an excellent agreement with the analytic result , contrary to fig .
figure 7 shows snapshots on the @xmath341 plane for the simulation of fig .
snapshots of the corresponding pure @xmath0-body simulation are also shown for comparison .
the hybrid code can reproduce not only @xmath303 and @xmath305 but also the distributions of @xmath179 and @xmath1 .
ida and makino ( 1992 ) carried out @xmath0-body simulations and found that the distributions of @xmath179 and @xmath1 can be well approximated by a rayleigh distribution .
we perform the kolmogorov - smirnov ( ks ) test to check whether the distributions of @xmath179 and @xmath1 in the hybrid simulations resemble a rayleigh distribution .
the test gives the significance level , @xmath349 ( @xmath350 ) ( press et al . ,
a large @xmath349 means that the likelihood that two functions are the same is high .
the time - averaged ( @xmath351 yr ) values of @xmath349 for the distributions of @xmath179 and @xmath1 are 0.44 and 0.57 , respectively .
we also perform the same ks test using the outputs of the pure @xmath0-body simulation , and the values of @xmath349 for @xmath179 and @xmath1 are 0.62 and 0.19 .
thus , the distributions of @xmath179 and @xmath1 from our hybrid simulations resemble a rayleigh distribution at a level similar to those from pure @xmath0-body simulations .
figure 7 also shows that a planetesimal disk radially expands with time due to viscous diffusion at a rate similar to that for the pure @xmath0-body simulation . to compare more quantitatively
, we plot the time evolution of the standard deviation of @xmath342 ( the normalized orbital energy ) of the disk in fig .
8 . since the radial diffusion of a low resolution hybrid simulation
is found to be noisy , we take average over several simulations .
the off - sets between simulations at @xmath352 are caused by random numbers that are used for choosing initial semimajor axes of tracers .
we find a good agreement in the increasing rates of the standard deviations of @xmath342 between the hybrid simulations and the pure @xmath0-body simulation .
this comparison indicates that the diffusion coefficient given by eq .
( [ eq : di0 ] ) is a good approximation .
contrary to collisional velocity evolution shown in section 3.1 , conservation of the angular momentum is violated by gravitational scattering as we do not explicitly solve pair - wise gravitational interactions in the statistical routine .
provided that the radial migration distance of a tracer due to random walk is about @xmath353 , the error in the total angular momentum is estimated as @xmath354 , which we found consistent with the actual errors in the test simulations .
fig . 9 .
self - sustained ( type iii ) migration of a sub - embryo .
we employ @xmath355 ( where @xmath2 is the embryo mass ) , @xmath356 , and @xmath357 .
the embryo is displayed as a red filled circle with a horizontal bar with a half length of 10 hill radii .
fig . 10 .
time evolution of the semi - major axis of a sub - embryo due to self - sustained migration : ( a ) @xmath358 and @xmath359 , ( b ) @xmath360 and @xmath359 , and ( c ) @xmath358 and @xmath361 . for all cases , @xmath356 and @xmath362 .
analytic migration rates ( bromley and kenyon , 2011 ) are shown by dashed lines . in section 2.6.2
, we described how we handle migration of sub - embryos . to test the routine of sub - embryo migration
, we perform a suite of hybrid simulations starting with initial conditions similar to those in kirsh et al .
( 2009 ) and bromley and kenyon ( 2011 ) .
the initial planetesimal disk extends from 14.5 au to 35.5 au and its total mass is 230 @xmath308 represented by 230 tracers ( thus @xmath363 ) . a sub - embryo with a mass of @xmath2 is initially placed at 25 au and growth of the embryo is turned off . in this subsection , the index @xmath1 is given to a sub - embryo .
each tracer has 10000 planetesimals and tracer - tracer interactions are quite small compared with gravitational effects of the embryo .
an example of migration of an earth - mass embryo ( @xmath364 ) is shown in fig .
9 using four snapshots on the @xmath365-@xmath179 plane . for this example , the migration is in the mass independent regime [ @xmath266 ; eq .
( [ eq : slow ] ) ] .
the embryo first has some strong encounters with tracers and a stirred region is produced .
the embryo tends to migrate away from the stirred region due to strong torques exerted by tracers in the unperturbed region .
this process trigers fast self - sustained migration . during the migration
, the embryo continuously encounters with unperturbed tracers on the leading side while leaving the stirred tracers in the trailing side .
the fast migration continues until the embryo reaches the disk edge .
embryo migrations for three cases with different embryo masses and initial velocity dispersions are shown in fig .
10 . in each case ,
six simulations are performed , using different random number seeds to generate initial orbital elements of tracers .
the migration rates suggested by bromley and kenyon ( 2011 ) are shown by dotted lines .
they confirmed good agreements between their formula and @xmath0-body simulations .
overall , our method can reproduce migration rates suggested by bromley and kenyon ( 2011 ) .
however , our method gives almost the same probabilities of inward and outward migrations if @xmath366 , whereas @xmath0-body simulations show that inward migration is predominant ( kirsh et al .
, 2009 ; bromley and kenyon , 2011 ) .
if the embryo is small , the direction of migration is determined by first few strong encounters , and once migration starts it is self - sustained regardless of its direction .
in realistic planetary accretion , even if inward migration is predominant , the embryo encounters with inner massive embryos and then turns its migration direction outward ( minton and levison , 2014 ) . therefore , we believe that the artificially high occurrence of outward migration is not a problem . for a large sub - embryo ,
inward migration is clearly predominant even in our hybrid simulations ( fig .
10c ) . to understand these trends and why our method works , we consider conditions necessary to produce smooth self - sustained migration
. if planetesimals are small enough compared with the embryo , smooth self - sustained migration of the embryo takes place ( ordered migration ) .
if planetesimals are large , on the other hand , the embryo suffers strong kicks and the direction of its migration may change rapidly in a random fashion ( stochastic migration ) . whether migration is ordered or stochastic is determined by the size distribution of planetesimals .
this argument is similar to that discussed for planetary spins ( dones and tremaine , 1993 ) .
the relative contributions of the ordered and stochastic components are estimated as follows .
consider that an embryo migrates over a distance @xmath367 through @xmath368 encounters with surrounding planetesimals : @xmath369 where again @xmath370 and @xmath371 is the change of @xmath98 @xmath372 $ ] during a single encounter . in the above ,
the angle brackets mean the averaging over encountered planetesimals .
the expected value of @xmath373 is given as @xmath374 , \label{eq : stoch } \end{aligned}\ ] ] where we assume that @xmath375 and that there is no correlation between @xmath376 and @xmath371 to transform to the second row .
if the first term in the r.h.s of eq .
( [ eq : stoch ] ) is larger than the second term , the migration is stochastic rather than ordered .
this means that the embryo is likely to migrate over @xmath367 via random walk rather than self - sustained migration .
if we assume that the ordered torque exerted on the embryo is the one - sided ( type iii ) torque , @xmath377 is determined by the non - dimensional migration rate @xmath258 ( appendices b.3 ) . the value of @xmath378 is estimated using the non - dimensional viscous stirring rates , @xmath136 and @xmath137 ( appendix d ) , and we consider contribution of planetesimals in one side assuming that large velocity asymmetry exists .
using these non - dimensional rates , we have @xmath379 .
this ratio is @xmath380 for the shear dominant regime [ @xmath195 and @xmath381 for the dispersion dominant regime ( we assume @xmath382 ) .
therefore , even with planetesimals as massive as the embryo ( @xmath383 ) , the migration is likely to be self - sustained migration rather than random walk if the embryo migrates longer than @xmath384 .
this length is comparable to the feeding zone width . in the above discussion
, the planetesimal mass @xmath30 can be replaced by the tracer mass @xmath385 to discuss migration calculated in our @xmath0-body routine .
therefore , our @xmath0-body routine can reproduce self - sustained migration of the embryo even with tracers as massive as the embryo .
this is true only if the ordered torque exerted on the embryo is as strong as the one - sided torque .
this is not always the case .
the torque becomes nearly one - sided if a strong velocity asymmetry of planetesimals exists around the embryo .
if the embryo jumps over its feeding zone due to a strong kick from a tracer that is as massive as the embryo , the embryo can not see velocity asymmetry at that radial location .
thus , the condition to apply the one - sided torque is that radial migration distance of the embryo due to a single encounter is much smaller than its feeding zone .
this condition generally turns out to be @xmath386 .
it is likely that the condition @xmath387 or less recommended by various authors ( kirsh et al .
2009 ; ldt12 ; minton and levison , 2014 ) comes from this criterion . in our method , alternatively , instantaneous large changes of @xmath36 of the sub - embryo are suppressed by limiting the migration distance of the sub - embryo to the theoretical prediction that is derived assuming that planetesimals are much less massive than the embryo .
therefore , our method can reproduce a long distance self - sustained migration .
the condition to reproduce predominant inward migration is more severe because we need to consider migration of an embryo due to torques from both sides of the disk . in an unperturbed state ,
the total torque from both sides is order of @xmath85 of the one - sided torque in the shear dominant regime ( bromley and kenyon , 2011 ) and is order of @xmath81 of the one - sided torque in the dispersion dominant regime ( ormel et al . , 2012 ) .
thus , ( @xmath388 ) is multiplied to the second term in the r.h.s .
( [ eq : stoch ] ) .
then , the condition that the migration is ordered when the embryo migrates over the feeding zone is roughly given by @xmath389 .
the initial condition of fig .
10c barely satisfies this criterion provided that @xmath30 is the tracer mass instead of the planetesimal mass .
our method indeed reproduces predominant inward migration .
once the embryo migrates over the feeding zone , velocity asymmetry is obvious and self - sustained migration immerses .
snapshots on the @xmath341 plane for the hybrid simulations ; ( a ) @xmath300 = 400 and ( b ) @xmath300 = 4000 .
tracers are displayed as colored circles and their colors represent the mass of a planetesimal as shown by the color bar .
sub - embryos are displayed as red filled circles with a horizontal bar with a half length of 10 hill radii [ for the panel ( b ) , only sub - embryos more massive than @xmath390 have a horizontal bar for consistency with the panel ( a ) ] .
a radius of a circle is proportional to @xmath391 .
snapshots of the cumulative number and the horizontal velocity ( @xmath392 , where @xmath393 is the keplerian velocity at 1au ) for ( a ) @xmath300 = 400 and ( b ) @xmath300 = 4000 .
the diagonal dashed line is the number of planetesimals represented by a single tracer with a mass of @xmath10 for each run .
the vertical dotted line represents @xmath10 . for comparison ,
the results from ( c ) wetherill and stewart ( 1993 ) and ( d ) inaba et al .
( 2001 ) are shown ; both from fig .
10 of inaba et al .
( 2001 ) .
fig . 12 . continued .
most of the authors who developed statistical planetary accretion codes ( weidenschilling et al . , 1997 ;
kenyon and luu , 1998 ; inaba et al .
, 2001 ; morbidelli et al . , 2009 ; ormel et al . ,
2010a ; ldt12 ; amaro - seoane et al . , 2014 ) tested their codes using fig .
12 of wetherill and stewart ( 1993 ) as a benchmark .
we also perform simulations starting with the same condition of wetherill and stewart ( 1993 ) .
inaba et al .
( 2001 ) performed the same statistical simulations using their own coagulation code and the code of wetherill and stewart ( 1993 ) after modifying settings of these two codes as close as possible .
both results are shown in fig .
10 of inaba et al .
( 2001 ) and we compare them with our results . as the initial condition , @xmath394 planetesimals are placed between 0.915 au and 1.085 au .
individual planetesimals have an equal - mass of @xmath395 g. the nebular gas is included and the gas / solid ratio is 44.4 .
see table 1 of wetherill and stewart ( 1993 ) for more details .
we take into account aerodynamic gas drag ( adachi et el . , 1976 ) instead of 2.0 used in morishima et al .
( 2010 ) . ] and tidal damping ( papaloizou and larwood , 2000 ) , but not type i gas - driven migration .
the gas disk model is described in morishima et al .
( 2010 ) in detail . in our simulations , the planetesimal disk radially expands with time , while the solid surface density remains constant in the previous statistical simulations .
however , the effect of radial expansion is small on the timescale of the simulation ( @xmath396 yr ) .
figure 11 shows snapshots on the @xmath341 plane for the two hybrid simulations with different initial numbers of tracers , @xmath397 and 4000 . in the early stage of evolution ( @xmath398 = 10,000 yr ) ,
large embryos grow very quickly and @xmath179 is enhanced locally where large embryos exist .
this stage is called the runaway growth stage .
once large embryos dominantly control the velocity of the entire planetesimal disk ( @xmath399 yr ) , growth of embryos slows down ( ida and makino , 1993 ) .
this stage is called the oligarchic growth stage . even though large embryos tend to open up gaps around their orbits , the gaps are not cleared out because the embryos migrate quickly due to occasional short - distance planetesimal - driven migrations and embryo - embryo interactions and because other embryos tend to fill gaps . during the oligarchic growth stage ,
the orbital separation between neighboring embryos is typically 10 hill radii as found by kokubo and ida ( 1998 , 2000 ) .
tanaka and ida ( 1997 ) showed that gaps open up only if the orbital separation of neighboring embryos is larger than 20 hill radii .
thus , formation of clear gaps is unlikely to occur in this stage .
figure 12 shows the mass and velocity distributions at different times for two hybrid simulations shown in fig .
the mass spectra of the hybrid simulations are smooth at the transition mass @xmath10 at which a tracer is promoted to a sub - embryo ( see also appendix g ) . for comparison , the results from wetherill and stewart ( 1993 ) and inaba et al .
( 2001 ) are also shown .
the hybrid simulations agree well with these statistical simulations , although some small differences are seen . in the hybrid simulations ,
the number of planetesimals represented by tracers has a lower limit shown by a dashed line in fig .
thus , the tail of the mass spectrum in the massive side during the runaway growth is not correctly described .
this causes a systematic delay in growth of large bodies , as already discussed by ldt12 .
the delay is particularly clear in the low - resolution hybrid simulation at 10,000 yr , although the delay is small .
this resolution effect should be seen even in the high resolution hybrid simulation while we see a good agreement with the statistical simulations at 10,000 yr .
this is probably because our stirring routine slightly underestimates eccentricities and inclinations ( section 3.2 ) and this effect cancels out with the resolution effect to some degree .
once the growth mode reaches oligarchic growth ( @xmath400 25,000 yr ) , the masses of the largest bodies in the low resolution simulation almost catch up with those in the high resolution simulation and those in the statistical simulations .
13 . snapshots on the @xmath341 plane for the simulations starting from equal - mass 2000 planetesimals in a narrow annulus between 0.7 au and 1.0 au : ( a ) the hybrid simulation starting with 200 tracers and ( b ) the pure @xmath0-body simulation . in the panel ( a )
, tracers are displayed as colored circles , and the colors represent the mass of planetesimal .
sub - embryos are displayed as red filled circles .
sub - embryos more massive than @xmath390 @xmath401 have a horizontal blue bar with a half length of 10 hill radii .
the same color code is used in the pure @xmath0-body simulation .
a radius of a circle is proportional to @xmath391 ( for the pure @xmath0-body simulation , @xmath11 for all particles ) .
fig . 14 . fig . 14 .
time evolutions of various quantities for three hybrid simulations ( orange , brown , and red solid curves ) and three pure @xmath0-body simulations ( blue , purple , and black dashed curves ) : ( a ) the mass - weighted root - mean - square eccentricity and inclination , @xmath402 and @xmath403 , for embryos and planetesimals , ( b ) the total number of all bodies @xmath0 , the total number of embryos @xmath301 , the total number of tracers @xmath300 , ( c ) the total mass of embryos @xmath404 , the effective mass of the system @xmath405 , and the mass of the largest body @xmath406 .
the last section focused only on the early stages of planetary accretion .
further , the damping effects due to the gas disk make it difficult to judge whether the code works appropriately . here
we test our code for the entire stages of accretion without the gas disk .
we perform hybrid simulations of accretion of terrestrial planets starting from a narrow annulus , similarly to those adopted in morishima et al .
( 2008 ) and hansen ( 2009 ) , until completion of accretion ( @xmath407 myr ) . as the initial condition , 2000 equal - mass planetesimals ( @xmath408 )
are placed between 0.7 au to 1.0 au , and the total disk mass is @xmath409 .
the initial number of tracers is 200 . for comparison
, we also perform pure @xmath0-body simulations with the same initial condition using pkdgrav2 .
a hybrid simulation takes about three days using a single core whereas a pure @xmath0-body simulation takes a few weeks using eight cores . as found in morishima et al .
( 2008 ) , the evolution is almost deterministic due to efficient dynamical friction if the initial number of planetesimals is large enough ( @xmath410 ) .
if the initial @xmath0 is small while keeping the same total mass , the evolution becomes rather stochastic and the final systems have relatively large diversity ( hansen , 2009 ) .
an example of the hybrid simulations is shown in fig .
13 together with an example of the pure @xmath0-body simulations . through the runaway growth stage , tens of embryos form ( @xmath411 myr )
. with increasing masses of embryos , mutual orbital crossings and impacts occur , and a few distinctively large embryos emerge ( @xmath412= 1 myr ) . in the late stage , the largest embryos sweep up small remnant planetesimals ( @xmath413 myr ) . at the end of accretion ,
the largest three embryos have orbital eccentricities and inclinations as small as those for the present - day earth and venus due to dynamical friction of remnant bodies ( @xmath414 myr ) . in the hybrid simulation ,
all the largest embryos are sub - embryos until the end of the simulation .
we perform three hybrid simulations and three pure @xmath0-body simulations in total , and all simulations show similar results , except one hybrid simulation produces only two distinctively large embryos . to see the accretion history more in detail , we plot various quantities as a function of time in fig . 14 . overall agreements between the hybrid simulations and the pure @xmath0-body simulations are excellent . good agreements between the hybrid and pure @xmath0-body simulations suggest that our statistical routine works well even in the late stages of planetary accretion .
figure 14a shows the evolution of the mass - weighted root - mean - square eccentricity and inclination , @xmath415 and @xmath416 , for embryos and planetesimals . with
increasing time , @xmath417 and @xmath418 for both embryos and planetesimals increase primarily due to viscous stirring of largest bodies .
after @xmath419 10 myr , @xmath417 and @xmath418 of embryos starts to decrease because small embryos with large @xmath179 and @xmath1 collide with large embryos with low @xmath179 and @xmath1 .
figure 14b shows the evolution of the total number of all bodies @xmath0 and the total number of embryos @xmath301 .
the number @xmath301 has a peak around 1 myr both in the hybrid and pure @xmath0-body simulations .
the number of bodies which orbits are numerically integrated , @xmath420 , increases by @xmath421 at maximum and the maximum value is taken around 1 myr ( fig .
figure 14c shows the total mass of embryos , @xmath404 , the effective mass of the system @xmath405 , which is sensitive to masses of largest embryos , and the mass of the largest body , @xmath406 .
it is found that the values of @xmath405 and @xmath406 in the hybrid simulations are slightly lower than those in the pure @xmath0-body simulations at @xmath422 myr . during the giant impact stage roughly between @xmath398 = 0.2 myr and a few myr ,
the contribution of embryo - embryo impacts to mass gain of embryos is significant , comparably to embryo - planetesimal impacts .
the embryo - embryo collision rate sensitively depends on @xmath179 and @xmath1 of embryos .
the hybrid simulations give slightly higher @xmath179 and @xmath1 ( fig .
14a ) and thus lower growth rates of embryos than those in the pure @xmath0-body simulations during the giant impact stage .
this is likely to be a resolution effect . in the low resolution hybrid simulations
, the number of nearby tracers around embryos fluctuates and that makes dynamical friction somewhat noisy .
we also perform the same simulations but starting with a larger number of tracers ( @xmath300 = 400 ) and find better agreements in @xmath405 with the pure @xmath0-body simulations .
fig . 15 .
snapshots on the @xmath341 plane for the hybrid simulation in the jovian planet region .
the simulation starts with equal - mass @xmath423 planetesimals .
tracers are displayed as colored circles and colors represent the mass of a planetesimal .
sub - embryos ( @xmath424 ) are displayed as red filled circles and full - embryos ( @xmath425 ) are displayed as blue filled circles .
a radius of a circle is proportional to @xmath391 , but enhanced by a factor of 6 for embryos for good visibility .
fig . 16 .
time evolutions of ( a ) semimajor axes and ( b ) masses of embryos that are more massive than 0.4@xmath308 at 2 myr .
the same color is used for the same embryo in both panels . the embryos represented by blue , red , and purple curves
are called embryo 1 , 2 , and 3 , respectively , in the main text .
the final test we present is accretion of cores of jovian planets starting with a wide planetesimal disk .
levison et al . ( 2010 ) and capobianco et al . ( 2011 ) showed importance of self - sustained outward migration of embryos , as they can grow rapidly via collisions with little - perturbed outer bodies .
however , these authors placed large embryos in the disk of small planetesimals in the initial conditions , and it is not clarified if outward migration plays a significant role in more realistic accretion scenarios .
we set the initial condition similar to that employed by levison et al .
( 2010 ) , but without any embryos .
we place @xmath423 equal - mass planetesimals between 4 au and 16 au represented by 10000 tracers .
the initial planetesimal mass @xmath426 is @xmath427 g and the radius is 2.4 km .
the total mass of planetesimals is 200 @xmath308 and this gives the solid surface density about seven times the minimum - mass solar nebula ( hayashi , 1981 ) .
the simulation is performed up to 2 myr and it takes roughly five days to complete using a single core .
the nebula gas effects are included in the same manner as the simulations in section 3.4 , and the gas - to - solid ratio is 56.7 , which is the same as the minimum - mass solar nebula beyond the snow line ( hayashi , 1981 ) . we do not consider enhanced cross sections of embryos due to their atmospheres ( inaba and ikoma , 2003 ) . the nebular gas is assumed to dissipate exponentially on the timescale of 2 myr .
capobianco et al .
( 2011 ) showed that outward migration of an embryo is usually triggered in the nebular gas , if the planetesimal radius is in a certain range .
the planetesimal radius we employ is in this range in the region inside of @xmath41910 au [ see their eq . ( 21 ) ] .
note that , however , planetesimal radii increase with time in our simulation unlike those in levison et al .
( 2010 ) and capobianco et al .
( 2011 ) .
the snapshots on the @xmath341 plane are shown in fig .
15 and the time evolutions of the masses and the semimajor axes of six largest embryos at the end of the simulation are shown in fig .
several embryos appear near the inner edge of the disk after 0.1 myr .
after some embryo - embryo interactions , one of the embryos ( embryo 1 ) starts self - sustained outward migration . since embryo 1 encounters with unperturbed planetesimals on its way , it grows very rapidly from 0.1 @xmath308 to 3 @xmath308 during only 50,000 yr .
after embryo 1 reaches the outer edge , it turns its migration direction inward .
however , it encounters with another smaller embryo ( embryo 2 ) and inward migration is halted .
there , embryo 1 opens up a gap and starts slow outward migration due to distant encounters with inner planetesimals .
embryo 1 is promoted to a full - embryo at @xmath428 0.2 myr during its inward migration .
embryo 2 and another embryo ( embryo 3 ) also migrate outward during the outward migration of embryo 1 , and they also grow more rapidly than inner embryos , but not as rapidly as embryo 1 .
they eventually change their migration directions inward and come back to near the inner edge .
embryo 2 occasionally attempts to migrate outward even after that , although its outward migration is not sustained .
two outward migrations at @xmath428 0.55 and 0.65 myr are halted by encounters with another embryo at @xmath419 7 au , which merges with embryo 2 at @xmath428 0.65 myr .
embryo 2 is promoted to a full - embryo at this time .
we do not clearly understand what prevents embryo 2 from subsequent outward migrations .
the radial profile of the surface density of planetesimals is no longer smooth due to occasional gap formation by embryos , and the outward migrations of embryo 2 seem to be halted at local low density regions .
another possibility is that migrations of full - embryos are halted by stochastic effects since they are not massive enough as compared with tracers .
this possibility may be tested using different transition masses from a sub - embryo to a full embryo .
the masses of embryo 2 and embryo 3 respectively reach 3.6 @xmath308 and 2.5 @xmath308 at 2 myr .
their growth is slow at this time as they open up a gap in the planetesimal disk .
obviously , we need to perform more simulations with different parameter sets and conduct more careful analysis .
nevertheless , the example we showed here is a good demonstration that our code can produce rapid growth of embryos during their outward migrations .
ngo ( 2012 ) performed simulations of accretion of cores of jovian planets using the lipad code .
the initial conditions of his simulations are similar to ours . in his simulations without fragmentation ( his figs .
5.14 and 5.15 ) , sub - embryos are almost radially stationary even though the routine of sub - embryo migration is included .
once sub - embryos are promoted to full - embryos , they suddenly start rapid radial migrations .
it is unclear to us why such large differences between sub - embryos and full - embryos are seen in his simulations . in spite of the large difference of embryo migration between
his and our simulations , the final masses of the largest embryos are close to each other .
the cpu time @xmath429 per timestep of orbital integration as a function of the number of embryos @xmath301 and the number of tracers @xmath300 is roughly given as @xmath430 where @xmath431 , @xmath432 , and @xmath433 are the cpu times for a single orbital integration ( a kepler drift ) , for a single direct gravity calculation of a pair in the @xmath0-body routine , and for a single set of calculations for a pair in the statistical routine , and @xmath434 is the typical number of interlopers in the neighboring search region ( fig . 2 ) .
for a standard latest computer , @xmath435 s. from some of simulations , we find that @xmath436 and @xmath437 . as seen in eq .
( [ eq : tcpu ] ) , the load on the statistical routine is controllable by modifying @xmath434 and @xmath438 . since the radial width of the neighboring search region
is given by eq .
( [ eq : drt0 ] ) , @xmath439 for a fixed surface density .
if we choose an azimuthal width as @xmath440 , @xmath434 becomes independent of @xmath300 and the cpu time for the statistical routine can be linearly scaled with @xmath300 .
practically , one may choose a @xmath24 so that the load on the statistical routine does not exceed the load on the @xmath0-body routine .
the actual computational load on the statistical routine in some of our simulations is as follows .
recall that in the present paper , we call the statistical routine every 30 steps of orbital integrations ( @xmath441 ) . if the time step of the statistical routine is halved , its load is doubled . in the hybrid simulations shown in fig . 5 ( @xmath331 and @xmath442 ) , the load on the statistical routine is only 10 % relative to the total load .
the rest of 90 % is on the @xmath0-body routine and more than half of it is occupied by the kepler equation solver .
there is no direct gravity calculation in these simulations , as no embryo is included . in the hybrid simulation shown in fig .
15 ( @xmath443 and @xmath444 ) , the load on the statistical routine is about 60 % before embryos appear .
if the number of embryos is larger than about 10 , the load on the direct gravity calculation exceeds that for the kepler equation solver , and the relative load on the statistical routine decreases . at 1.6 myr of fig .
15 , at which 21 embryos exist , the load on the statistical routine is about 45 % , that on the direct gravity calculation is 35 % , and that on the orbital integration is 20 % .
overall , the load of the statistical routine is comparable to or less than that on the @xmath0-body routine
. we also comment on parallelization .
the current version of our code is serial .
however , the direct @xmath0-body routine is already parallelized and we test how much simulations are accelerated by parallelization in case without the statistical routine . in the tests , we fix the number of embryos to be 10 and vary the number of tracers .
we find that simulations are accelerated only if the number of tracers is more than @xmath396 .
in contrast , the pure @xmath0-body simulations using the tree code , pkdgrav2 , are accelerated by parallelization if @xmath445 ( morishima et al . , 2010 ) .
the difference comes from the cpu times per step . to be benefitted by parallelization
, the cpu time needs to be somewhat long to ignore latency .
the computational cost per step for hybrid simulations is as small as those for @xmath0-body simulations without particle - particle interactions .
that ironically makes parallelization inefficient .
we consider only random walk of tracers in tracer - tracer interactions .
thus , self - sustained migration occurs only to embryos in our code although small bodies ( @xmath446 ) may migrate in reality . to examine
if migrations of small bodies play an important role for overall accretion history of embryos , high resolution simulations ( with smaller @xmath10 ) are necessary .
minton and levison ( 2014 ) showed that self - sustained migration occurs only if the embryo is much more massive than surrounding planetesimals and sufficiently separated from other embryos .
such conditions are usually fulfilled near the transition from the runaway growth stage to the oligarchic growth stage .
the mass of the largest embryo at the transition is roughly given by @xmath447 ( ormel et al .
, 2010b ; morishima et al . ,
2013 ) , where @xmath448 is the isolation mass and @xmath426 is the initial mass of planetesimals .
therefore , if @xmath426 is very small ( e.g. , weidenschilling , 2011 ) , a very large number of tracers is necessary to represent an embryo at the transition by a single particle in our code .
however , minton and levison ( 2014 ) also pointed out that the growth time scale needs to be longer than the migration time for an embryo to avoid encounters with other embryos during migration .
if @xmath426 is very small , this criterion is not generally fulfilled for the largest embryo at the transition due to a very short growth time scale .
thus , masses of candidate embryos for self - sustained migration are probably not too small to be practically resolved by our code even if @xmath426 is very small . the argument here , of course , depends on disk parameters and radial location and needs more careful estimates .
our code can probably handle eccentric / inclined ringlets that are usually produced by external massive bodies ( see section 2.3 ) .
however , we have not tested such simulations in this paper , and they remain for future work . a potential issue of our code in handling eccentric / inclined ringlets is accuracy of the global gravity forces ( section 2.7 ) . only for this calculation
, we assume that a ring is axisymmetric and symmetric with respect to the invariant plane .
these assumptions may not be appropriate if the radial / vertical thickness or the ringlet is much smaller than the radial / vertical excursion of individual particles .
this is the case of some of narrow ringlets seen around solar system giant planets .
if the global self - gravity plays an important role in structure evolution of narrow ringlets , we need a more advanced method to calculate it .
the current code includes an option of hit - and - run bouncing ( appendix e ) but not fragmentation .
we plan to implement collisional fragmentation to our code in subsequent work .
since gas drag damps @xmath179 and @xmath1 of small fragments effectively , growth of embryos is accelerated by fragmentation of planetesimals , particularly in the oligarchic growth stage . on the other hand , since the total solid mass available for planetary accretion decreases due to rapid inward migration of fragments in the gaseous disk , the final masses of planets also decrease , unless planetary atmospheres effectively capture fragments .
these conclusions were derived by simulations using statistical accretion codes ( e.g. , inaba et al . , 2003 ; chambers , 2008 ; kobayashi et al . , 2010 , 2011 ; amaro - seoane et al . ,
2014 ) . in these studies ,
the torques on embryos exerted by small fragments were not taken into account .
levison et al . ( 2010 ) showed that small fragments shepherded by embryos push bash the embryos toward the central star , leading to rapid inward migration of the embryos . on the other hand , kobayashi et al .
( 2010 ) pointed out that shepherded fragments are quickly ground down due to mutual collisions and very small ground fragments collide with or migrate past the embryos before pushing back the embryos .
however , this has not been confirmed by direct lagrangian type simulations and is of interest in future work .
we developed a new particle - based hybrid code for planetary accretion .
the code uses an @xmath0-body routine for interactions with embryos while it can handle a large number of planetesimals using a super - particle approximation , in which a large number of small planetesimals are represented by a small number of tracers .
tracer - tracer interactions are handled by a statistical routine .
if the embryo mass is similar to tracer masses and if embryo - tracer interactions are handled by the direct @xmath0-body routine , the embryo suffers artificially strong kicks from tracers . to avoid this issue ,
sub - embryos are introduced .
accelerations of sub - embryos due to gravitational interactions with tracers are handled by the statistical routine whereas accelerations of tracers due to gravitational interactions with sub - embryos are handled by the direct @xmath0-body routine .
our statistical routine first calculates the surface number densities and the orbital elements of interloping planetesimals around each target tracer ( section 2.2 ) . using the phase - averaged collision probability , whether a collision between the interloper and the target occurs is determined ( section 2.4 ) .
if a collision occurs , the velocity changes are accurately calculated by matching the positions of the interloper and the target without changing their eccentricity and inclination vectors . using the phase - averaged stirring and dynamical friction rates ,
the change rates of the orbital elements of the tracers are calculated ( section 2.5 ) .
these rates are converted into the accelerations of the tracers using both first and second order gauss planetary equations .
planetesimal - driven migration of sub - embryos is basically handled by the @xmath0-body routine but the migration rate is limited to the theoretical prediction derived by the statistical routine ( section 2.6 ) .
this unique routine can reproduce smooth , long - distance migrations of sub - embryos .
we performed various tests using the new hybrid code : velocity evolution due to collisions , and that due to gravitational stirring and dynamical friction , self - sustained migration of sub - embryos , formation of terrestrial planets both in the presence and absence of the gaseous disk , and formation of cores of jovian planets .
all the test simulations showed good agreements with analytic predictions and/or pure @xmath0-body simulations for all cases , except that the last test did not have a robust benchmark .
the computational load on the portion of the statistical routine is comparable to or less than that for the @xmath0-body routine .
the current code includes an option of hit - and - run bouncing but not fragmentation that remains for future work .
we are grateful to the reviewers for their many valuable comments , which greatly improved the manuscript .
we also thank satoshi inaba for kindly giving us his simulation data .
this research was carried out in part at the jet propulsion laboratory , california institute of technology , under contract with nasa .
government sponsorship acknowledged .
simulations were performed using jpl supercomputers , aurora and halo .
in section 2.2 , we only consider encounters with interlopers .
however , planetesimals in the target @xmath1 may encounter with other planetesimals in the same target if @xmath47 . to take into account this effect
, we modify apparent planetesimal numbers in interlopers , if masses of interloping planetesimals are similar to those in the target tracer . for this purpose only , we assume axisymmetric distribution of planetesimals and introduce the mass bin for planetesimals .
we also use the radial bin used in the neighboring search .
we first calculate the total mass of tracers , @xmath449 , in the radial and mass bins where the target locates .
the summation is done for tracers in any azimuthal locations , excluding the target . since the actual total mass seen from an individual planetesimal in the target tracer
is @xmath450 , we use a corrected @xmath451 for the interloper , @xmath452 if it is in the same radial and mass bins with the target . if there is no other tracer in the same radial and mass bins , the correction is not applied .
we find that this self - encounter effect is very small at all as we do not identify any difference in simulations with and without the correction .
the effect can also be checked by simulations with a same total number of planetesimals but using different numbers of tracers .
although a simulation with a small number of tracers has larger statistical fluctuations in various physical quantities , such as the mean eccentricity , any differences in time - averaged quantities are not well identified .
thus , we can omit this correction , and mass bins are unnecessary in our code .
with hill s approximation , the equations of motion for the relative motion between the target and the interloper are reduced to hill s equations ( hill 1878 ; nakazawa et al . , 1989 ) .
the time and the distance of hill s equations can be normalized by the inverse of orbital frequency , @xmath453 , and @xmath454 , respectively , where @xmath455 is the mean semimajor axis and @xmath85 is the reduced mutual hill radius [ eq .
( [ eq : rhill ] ) ] .
the solution of the relative motion in the absence of mutual gravity is expressed by non - dimensional relative orbital elements , @xmath98 , @xmath456 , @xmath457 , @xmath71 and @xmath72 .
the elements @xmath98 , @xmath456 and @xmath457 are given as @xmath458 where @xmath459 is the difference of the semimajor axes , the relative eccentricity @xmath81 and the relative inclination @xmath82 are defined in eq .
( [ eq : relei ] ) together with the phases @xmath71 and @xmath72 .
the non - dimensional collision rate @xmath460 is defined as @xmath461 where @xmath462 is unity for collisional orbits and zero otherwise . in the dispersion dominant regime [ @xmath463 , @xmath464
is given as ( greenzweig and lissauer , 1990 ; inaba et al .
, 2001 ) @xmath465 where @xmath466 with the physical radii @xmath467 and @xmath468 , @xmath469 and @xmath470 are the complete elliptic integrals of the first and second kinds , respectively , and @xmath471 . in the shear dominant regime [ @xmath472 ,
if the inclination is small enough ( @xmath473 ; see goldreich et al .
( 2004 ) for this criterion ) all pairs of particles entering the mutual hill sphere collide , and @xmath464 becomes independent of @xmath456 and @xmath457 as ( ida and nakazawa , 1989 ) @xmath474 in the shear dominant regime with moderate inclinations ( @xmath475 ) , @xmath464 depends only on @xmath476 as ( ida and nakazawa , 1989 ) @xmath477 following inaba et al .
( 2001 ) , we connect the collision rates in different regimes as @xmath478.\ ] ] the collision rate given by this expression well agrees with that derived by numerical integrations ( ida and nakazawa , 1989 ; greenzweig and lissauer , 1990 ) . the non - dimensional stirring and dynamical friction rates are defined as @xmath479 where @xmath480 , @xmath481 , and @xmath482 are the changes of @xmath483 , @xmath484 , and @xmath485 [ eq . ( [ eq : relei ] ) ] during a single encounter .
the non - dimensional dynamical friction rate for inclination is obtained by replacing @xmath486 by @xmath487 in eq .
( [ eq : pdfd ] ) and is known to be the same with @xmath138 ( tanaka and ida , 1996 ) .
the rates for the shear dominant regime are ( ida 1990 , ohtsuki et al . , 2002 )
@xmath488 for @xmath489 , we adopt a functional form that is similar to the one with the rayleigh distribution ( ohtsuki et al . ,
2002 ) but the factors are modified .
the rates for the dispersion dominant regime are given as ( tanaka and ida , 1996 ) @xmath490 \nonumber \\ & & \times \ln{(\lambda^2 + 1)},\\ q_{\rm vs , high } & = & \frac{36}{\pi \tilde{i}_{\rm r } \sqrt{\tilde{e}_{\rm r}^2 + \tilde{i}_{\rm r}^2 } } \left[k(\zeta)- \frac{12\tilde{i}_{\rm r}^2}{\tilde{e}_{\rm r}^2 + 4\tilde{i}_{\rm r}^2}e(\zeta)\right ] \nonumber \\ & & \times \ln{(\lambda^2 + 1)},\\ p_{\rm df , high } & = & \frac{288}{\pi \tilde{i}_{\rm r } ( \tilde{e}_{\rm r}^2 + 4\tilde{i}_{\rm r}^2 ) \sqrt{\tilde{e}_{\rm r}^2 + \tilde{i}_{\rm r}^2 } } e(\zeta ) \ln{(\lambda^2 + 1)},\end{aligned}\ ] ] where @xmath491 in @xmath492 and @xmath493 , we omit the term @xmath494 as it is negligible compared with @xmath495 in the dispersion dominant regime .
we connect the stirring and dynamical friction rate at the shear and dispersion dominant regimes in a similar manner of ohtsuki et al .
( 2002 ) : @xmath496 the coefficients are @xmath497 the approximated formulae well agree with the rates derived by three - body orbital integrations [ fig . 8 of ida ( 1990 ) and fig . 1 of ohtsuki et al .
( 2002 ) ] , except for the region with @xmath498 and @xmath499 ( see section 3.2 ) .
the non - dimensional migration rate @xmath258 for one - sided torque is defined as @xmath500 where @xmath371 is the change of @xmath98 during a single encounter and the negative sign represents the back reaction on the embryo . at @xmath501 , @xmath502 [ ida et al . , 2000
; their eq .
it is expected that @xmath503 in the shear dominant regime is about the same value as the migration rate is almost constant in this regime ( kirsh et al . , 2009 ) :
@xmath504 for the dispersion dominant regime , ida et al .
( 2000 ) derived the migration rate averaged over @xmath71 and @xmath72 as [ their eq .
( 18 ) ] @xmath505 inserting this equation into eq .
( [ eq : rmig ] ) , we have @xmath506 where the integration range of @xmath98 is limited between @xmath507 and @xmath456 , and @xmath508 is approximately treated as a constant in the integration .
analogous to the stirring rates , we connect the migration rates at the shear and dispersion dominant regimes but the value is assumed not to exceed @xmath509 @xmath510.\ ] ] while this expression must be correct in the low and high velocity limits , its accuracy at the intermediate regime needs to be confirmed by comparing with the rate derived from three - body orbital integrations .
when the non - dimensional collisional and stirring rates are derived , uniformities of @xmath71 and @xmath72 are assumed .
however , we do not assume uniformities of the relative longitudes , @xmath89 and @xmath511 .
alignments of the longitudes occur , for example , if the system is perturbed by a giant planet and the forced eccentricity is much larger than the free eccentricities .
ida ( 1990 ) showed that the eccentricity evolution of a particle @xmath1 due to dynamical friction of particles @xmath27 is @xmath512p_{\rm df } \label{eq : dfj0 } \\ & = & \frac{c_1}{\nu_{j}h_{ij}^2}\left[\nu_{j}{e}_{j}^2 - \nu_{i}e^2_{i } + ( \nu_{i}-\nu_{j})e_{i}e_{j}\cos{\varpi_{ij } } \right]p_{\rm df } , \label{eq : dfj1 } \end{aligned}\ ] ] where the characters appeared in these equations are defined around eqs .
( [ eq : dedfs ] ) and ( [ eq : dedfd ] ) .
if @xmath91 s of interlopers are randomly distributed and the distribution of @xmath513 is the rayleigh distribution , the term with @xmath514 becomes small enough after averaging over interlopers ( ida , 1990 ) .
the equilibrium state for this case leads to the well - known energy equipartitioning ( @xmath515 ) with dynamical friction only . with viscous stirring ,
however , the energy equipartitioning is achieved only if the mass distribution is very steep ( rafikov , 2003 ) . if the distribution of @xmath91 is not random , as may be the case of a forced system , the term with @xmath514 in eq .
( [ eq : dfj1 ] ) can not be ignored . in a forced system ,
the eccentricity vectors may be split as @xmath516 and @xmath517 , where @xmath518 is the forced eccentricity vector , @xmath519 and @xmath520 are the free eccentricity vectors , respectively ( murray and dermott , 1999 ) . inserting these forms into eq .
( [ eq : dfj0 ] ) and setting @xmath521 due to the uniformity of @xmath71 , we can find an equation similar to eq .
( [ eq : dfj1 ] ) but @xmath48 and @xmath54 replaced by the free eccentricities @xmath522 and @xmath523 and @xmath91 replaced by the relative longitude of the free eccentricities . since there is no correlation between the directions of @xmath520 and @xmath519 , the third term in the square bracket of the r.h.s of eq .
( [ eq : dfj1 ] ) can be neglected .
thus , dynamical friction tends to lead the energy equipartitioning of the free eccentricities , if @xmath518 is shared for all particles . a similar conclusion can be derived for inclinations .
now we consider splitting eq .
( [ eq : dfj1 ] ) into two terms : a stirring term that is preferably larger than the viscous stirring term [ as required in eq .
( [ eq : vrs ] ) ] and a damping term that decreases the free eccentricity .
we find the forms that can fulfill such conditions as @xmath524 the same equations are given in eqs .
( [ eq : dedfs ] ) and ( [ eq : dedfd ] ) . in a forced system , the terms in the parenthesis of the r.h.s . of eq .
( [ eq : dedfdap ] ) are reduced to @xmath525 using @xmath521 and @xmath526 due to the uniformity of @xmath71 . the l.h.s . of eq .
( [ eq : dedfdap ] ) is also reduced to @xmath527 because the orientation of @xmath519 changes randomly due to gravitational interactions with other tracers [ @xmath528 . thus , the damping term [ eq .
( [ eq : dedfdap ] ) ] decreases the free eccentricity , as required . usually in astrophysics
, only this damping term is called dynamical friction ( binney and tremaine , 1987 ) . a similar splitting is possible for inclinations , and the split terms
are given by eqs .
( [ eq : didfs ] ) and ( [ eq : didfd ] ) .
the classical theory of random walk shows that the diffusion coefficient @xmath140 for the one dimensional problem is @xmath529 , where @xmath144 is the change in the semimajor axis for the tracer @xmath1 during the time step @xmath17 and the angle brackets mean averaging over multiple changes .
the change of @xmath36 due to a single encounter with other planetesimal is defined to be @xmath530 , where @xmath371 is again the change of @xmath98 during an encounter .
thus , the diffusion coefficient @xmath531 for the tracer @xmath1 due to encounters with planetesimals in the tracer @xmath27 is given as @xmath532 where @xmath533 is the change of @xmath98 during @xmath17 and the subscript for the angle brackets means averaging over planetesimals in the tracer @xmath27 .
petit and hnon ( 1987 ) performed three - body orbital integrations and showed that @xmath534 for @xmath535 .
although it is unclear if this factor is applicable for @xmath536 , we use the relationship for any cases .
this gives @xmath537/2 , using @xmath538 .
the averaged change @xmath539 is given by @xmath540 where we used @xmath541 that is derived from the jacobi integral of hill s equations [ eq .
( [ eq : jaco ] ) ] , and @xmath136 and @xmath137 are defined in eqs .
( [ eq : pvs3 ] ) and ( [ eq : qvs3 ] ) . inserting eq .
( [ eq : db2 ] ) into eq .
( [ eq : did ] ) , the diffusion coefficient @xmath531 is given as @xmath542 finally , the diffusion coefficient due to encounters with all interlopers is given by @xmath543 .
ohtsuki and tanaka ( 2003 ) derived the viscosity for the equal - mass planetesimals disk and their viscosity is @xmath544 .
the diffusion coefficients derived by ormel et al .
( 2012 ) and glaschke et al .
( 2014 ) are similar to ours .
in section 2.4 , we show how the total number of tracers changes due to merging between planetesimals . on the other hand , in hit - and - run collisions
, there should be no change in the number . before describing how to handle hit - and - run collisions between planetesimals in two tracers ,
we first show how to handle a hit - and - run collision between embryos or between a full embryo and a tracer .
consider an impact between the target with mass @xmath2 and the impactor with mass @xmath30 ( @xmath545 ) .
let the impact velocity and the impact angle be @xmath546 and @xmath547 ( @xmath548 for a head - on collision ) .
if @xmath549 , where @xmath550 is the critical velocity , the collision results in escape of the projectile after bouncing on the target . based on thousands of sph simulations , genda et al . ( 2012 ) derived the formula of @xmath550 as @xmath551 where @xmath552 is the mutual escape velocity , @xmath553 , and @xmath554 .
the coefficients are @xmath555 , @xmath556 , @xmath557 , @xmath558 , and @xmath559 . if the impact is judged as a hit - and - run collision ( @xmath549 ) , the post - impact velocities of the projectile and the target are determined as follows .
we do not consider any mass exchanges between the target and the impactor for simplicity following kokubo and genda ( 2010 ) .
the impact velocity vector is decomposed to the normal and tangential components ( @xmath560 and @xmath561 ) relative to the vector pointing toward the center of the target from the center of the impactor .
the tangential component of the post - impact relative velocity @xmath562 is assumed to be the same as @xmath561 .
the normal component of the post - impact relative velocity is given as @xmath563 the velocity change described here is similar to but slightly different from that of kokubo and genda ( 2010 ) , who always set @xmath564 and adjust @xmath565 .
the normal and restitution coefficients , @xmath306 and @xmath309 , are defined as @xmath566 and @xmath567 . while the above case basically assumes @xmath310 and @xmath311 , these coefficients
can be changed to any values in simulations ( in section 3.1 , we use @xmath310 and 0.8 ) .
now we consider hit - and - run collisions between planetesimals in two tracers .
we will handle them as consistent as possible with those between embryos .
if a collision is judged to occur in the statistical routine , the target and the impactor are moved to the impact position without changing their eccentricity and inclination vectors , as described in section 2.4.2 .
the relative velocity , @xmath568 , at infinity ( without acceleration due to mutual gravity ) is calculated at this impact position . if @xmath569 , only some fraction of planetesimals in the tracer @xmath27 are involved in the collision ( fig .
the impact velocity is given as @xmath570 .
we ignore deflection of the relative orbit during acceleration due to mutual gravity so that the direction of the impact velocity vector is identical to that of the relative velocity vector @xmath571 .
the problem is that we can not directly use the relative position vector of the tracers because their mutual separation is not exactly the sum of the radii of a colliding pair after matching their positions . instead of using the actual relative position vector
, we make a fictitious relative position vector @xmath572 . the fictitious impact angle @xmath547 is given by @xmath573 .
assuming there is no correlation between @xmath571 and @xmath572 , the probability that the impact angle is between @xmath547 and @xmath574 is given as @xmath575 ( e.g. , shoemaker and wolfe , 1982 ) .
we choose an arbitrary @xmath547 between 0 and 90 degrees following this distribution and also choose the tangential direction of @xmath572 relative to @xmath571 at random .
then , we judge whether the collision results in merging or bouncing using eq .
( [ eq : gen ] ) in which @xmath30 is replaced by @xmath95
. if it is merging , we follow the process described in section 2.4.2 , using @xmath571 as the impact velocity vector .
if it turns out to be a hit - and - run collision , we calculate the post impact velocities using @xmath576 and @xmath577 from eq .
( [ eq : vnd ] ) .
then their velocities are modified so that their relative velocity at infinity is given as @xmath578 , again ignoring deflection of the orbits due to mutual gravity . if the interloper is split into two tracers [ case ( a ) in eq .
( [ eq : delm ] ) ] before the collision , we merge them by summing their momenta after matching their positions .
the merged tracer has the same @xmath30 and @xmath107 as those of the original interloper before splitting .
the semimajor axis of the interloper is adjusted so that the @xmath120-component of the total angular momentum of is conserved .
finally , the mean longitudes of the target and interloper are placed to the original values .
here we explain how to derive the correction factor @xmath234 appeared in eq .
( [ eq : nj2 ] ) .
let us define the inner and outer radii of the region @xmath1 for neighboring search around the embryo be @xmath579 and @xmath580 ( fig . 2 ) .
the interloper @xmath27 has the semimajor axis @xmath53 , the eccentricity @xmath54 , and the inclination @xmath55 .
the maximum and minimum distances of the interloper @xmath27 , @xmath581 and @xmath582 , from the central star projected on the invariant plane ( @xmath21 ) are given by @xmath583 if @xmath584 , we define the eccentric anomaly @xmath585 ( @xmath586 ) at @xmath587 as @xmath588 and the mean anomaly is also defined as @xmath589 . if @xmath590 , we set @xmath591 . in a similar manner ,
if @xmath592 , we define the eccentric anomaly @xmath593 ( @xmath586 ) at @xmath594 as @xmath595 and the mean anomaly @xmath596 . if @xmath597 , we set @xmath598
. then , the period @xmath236 in which the interloper @xmath27 is radially in the region @xmath1 during the orbital period @xmath235 is given as @xmath599
here we summarize the differences between the lipad code ( ldt12 ) and our new code . in the lipad code , to derive the spatial density of planetesimals ,
tracers are placed in mass and radial grids and the vertical distribution is assumed to be gaussian .
our code has neither these fixed grids nor the assumption for the vertical distribution ( except for the global gravity calculation described in section 2.7 ) .
thus , our code has lagrangian natures more than the lipad code , although these differences are probably not fundamental for the test simulations shown in this paper .
our scheme is more beneficial for systems with non - axisymmetric and/or inclined structures .
the advantage of the assumption of the vertical distribution in the lipad code is that changes in collision and stirring rates at different @xmath120 are explicitly taken into account .
changes in collision rates with z are taken into account in our model too , although implicitly , as we move both the target and the interloper to the collisional point ( section 2.4 ) ; collisions occur more often near the mid - plane than at high @xmath600 . in the stirring routine ( section 2.5 )
, our method ignores this effect and this probably introduces some inaccuracies in orbital evolutions of individual particles .
nevertheless , the velocity evolution of the system as a whole is reproduced reasonably well by our method as shown in the test simulations ( see more discussion for viscous stirring and dynamical friction given below ) .
the number of planetesimals in a tracer , @xmath3 , is an integer in our model ( section 2.1 ) , contrary to ldt12 .
if a tracer with a non - integer number of planetesimals is promoted to a single sub - embryo , the extra mass , @xmath601 ( if @xmath602 ) , needs to be incorporated into the sub - embryo or transferred to other tracers which may have very different planetesimal masses . both are not accurate although it is not clear to us how it is handled in the lipad code . in our code , orbital isolation between planetesimals
does not occur as we choose a narrow radial half width @xmath23 of the neighboring search region given by eq .
( [ eq : drt0 ] ) .
if we adopt a larger @xmath23 , orbital isolation can potentially occur . in this case
we may need to consider how isolated bodies interact through distant perturbations .
in general , however , the radial resolution is sacrificed with a large @xmath23 . in our current scheme ,
interactions between runaway bodies are approximately represented by occasional close encounters rather than distant perturbations .
the lipad code explicitly handles orbital isolation and distant perturbations ( e.g. , their fig .
this means that they use the radial grid size larger than ours .
even with a large @xmath23 , the exitisting lagrangian codes can not handle orbital isolation of planetesimals that are much less massive than @xmath10 , because a small number of runaway bodies can not be represented by tracers with nearly fixed masses .
collisions between tracers and sub - embryos are handled in the @xmath0-body routine in the lipad code .
thus , after the promotion the first impact of a tracer doubles the mass of a sub - embryo .
this makes artificial kinks in mass spectra at the transition mass , @xmath10 .
the influence of this effect on evolution of the overall system is small as discussed by ldt12 , particularly if the masses of the largest embryos are much larger than @xmath10 . in our method ,
collisions of planetesimals with sub - embryos are handled by the statistical routine ( secs .
2.4 and 2.6 ) .
the mass spectra are smooth at @xmath10 ( see fig .
12 ) , because only some of planetesimals in a tracer collide with a sub - embryo .
note that we assume uniformities in phase space ( section 2.3 ) that seem to be valid at least in test simulations shown in this paper , while no assumption for phase - space is necessary for the @xmath0-body routine .
the impact velocity between tracers is calculated more accurately in our code than the lipad code .
as described in section 2.4 , we estimate the impact velocity between tracers by matching their positions without changing their eccentricity and inclination vectors . in the lipad code , only the interloper s longitude is aligned to that of the target by rotating the interloper s coordinates , not letting the interloper move along its keplerian orbit .
this changes the longitude of pericenter of the interloper by the rotation angle .
ldt12 managed to minimize this issue by choosing the azimuthally closest interloper for the collision with the target , as originally done in levison and morbidelli ( 2007 ) . on the other hand , our method allows any interlopers to collide with the target .
the timescale of collisional damping shown in fig .
2 of ldt12 is clearly too short , probably because they described the initial condition of a wrong simulation .
the largest difference between our code and the lipad code probably appears in the methods to handle viscous stirring and dynamical friction . in the lipad code ,
the probability that an interloper enters within a mutual hill radius of the target is calculated .
if an encounter occurs , the acceleration of a smaller body of the pair is given by solving a pair - wise gravitational interaction .
the acceleration of a lager body of the pair is not calculated in this routine , but the damping forces are given by the chandrasekhar s formula of dynamical friction .
instead of taking into account stirring on the larger body , they set lower limits of @xmath179 and @xmath1 of the larger body based on energy equipartitioning .
some inaccuracies seem to be introduced in their approach .
first , energy equipartitioning is not usually achieved except with a very steep mass spectrum ( rafikov , 2003 ) .
second , the chandrasekhar s formula of dynamical friction is not applicable to the shear - dominant regime ( e.g. , ida , 1990 ; ohtsuki et al . , 2002 ) .
in our method , we do not solve individual pair - wise gravitational interactions .
instead , we calculate the time averaged changes of the orbital elements due to multiple encounters with tracers , using the phase - averaged stirring rates ( section 2.5.1 )
. these changes of orbital elements are then converted into the accelerations of tracers ( section 2.5.2 ) .
this approach allows us to handle viscous stirring , the stirring and damping parts of dynamical friction due to interactions between all pairs in the same manners regardless of their mass ratios .
our approach is based on uniformities of the phases ( section 2.3 ) , and methods which explicitly solves pair - wise gravitational interactions ( i.e. , the acceleration of a smaller body in the lipad code ) are probably more accurate than ours .
we would like to point out , however , that the lipad code also partly assumes uniformities of the phases because it uses the phase - averaged collision / encounter probability of greenzweig and lissauer ( 1990 ) ; note that @xmath464 we use and @xmath603 in the lipad code are the same quantity in different formats .
ldt12 included planetesimal - driven migration of sub - embryos by adding torques on them calculated by a series of three - body integrations ( sun , the target , and the interloper ) during lipad simulations .
each interloper is randomly chosen from one in seven hill radii of the target .
it has the same semimajor axis , orbital eccentricity , and inclination as one of the tracers but its phases are modified . as far as we understand
, their point is that the torque on the sub - embryo is smoothed by phase averaging
. this may be true in the dispersion dominant regime ( @xmath604 ) . in the shear dominant regime
, however , the phase dependence of the torque is weak and the averaged toque turns out to be as strong as the torque from a single interloping tracer at all [ see fig . 8 of ohtsuki and tanaka ( 2003 ) , for example ] .
thus , sub - embryos should occasionally suffer strong kicks from tracers . although it is unclear to us whether the point described above is indeed an issue for the lipad code , sub - embryos in lipad simulations ( ngo , 2012 ) look radially much more stationary than those in our simulations ( section 3.6 ) . in our method ,
the angular momentum change of a sub - embryo is stored in the @xmath0-body routine but its release rate is limited to the theoretical prediction derived by the statistical routine ( section 2.6.2 ) .
theoretical justifications of our method are discussed in section 3.3 .
the lipad code performs a series of three - body integrations during a simulation for two cases .
the first one is for the routine of sub - embryo migration .
the another one is for the stirring routine in the shear dominant regime .
our code does not need to perform three body integrations .
migration of sub - embryos are handled as explained above . stirring in the shear dominant regime
is handled using the approximated formulae of the phase - averaged stirring rates ( section 2.5 and appendix b ) .
adachi , i. , hayashi , c. , nakazawa , k. 1976 .
the gas drag effect on the elliptical motion of a solid body in the primordial solar nebula .
56 , 17561771 .
kortenkamp , s.j . , wetherill g.w , 2000 . terrestrial planet and asteroid formation in the presence of giant planets .
i. relative velocities of planetesimals subject to jupiter and saturn perturbations .
icarus 143 , 6073 .
morishima , r. , schmidt , m.w . ,
stadel , j. , moore , b. 2008 . formation and accretion history of terrestrial planets from runaway growth through to late time : implications for orbital eccentricity .
j. 685 , 12471261 .
nagasawa , m. , lin , d.n.c . ,
thommes , e.w . , 2005 .
dynamical shake - up of planetary systems .
i. embryo trapping and induced collisions by the sweeping secular resonance and embryo - disk tidal interaction .
j. 635 , 578598 . | we introduce a new particle - based hybrid code for planetary accretion .
the code uses an @xmath0-body routine for interactions with planetary embryos while it can handle a large number of planetesimals using a super - particle approximation , in which a large number of small planetesimals are represented by a small number of tracers .
tracer - tracer interactions are handled by a statistical routine which uses the phase - averaged stirring and collision rates .
we compare hybrid simulations with analytic predictions and pure @xmath0-body simulations for various problems in detail and find good agreements for all cases .
the computational load on the portion of the statistical routine is comparable to or less than that for the @xmath0-body routine .
the present code includes an option of hit - and - run bouncing but not fragmentation , which remains for future work .
accretion ; planetary formation ; planetary rings ; planets , migration ; origin , solar system |
graphene has garnered interest from broad spectrum of communities , ranging from those aiming at atomic scale circuit devices to those searching for new topological phases . both communities sought after ways to gap the massless dirac spectrum .
the realization of a gate - induced band - gap in the bernal stacked bi - layer graphene @xcite following the prediction in ref .
@xcite brought the holy grail of graphene based transistor one step closer to reality .
however , the sub - gap conductance measured by @xcite with weak temperature dependence well below the optically measured gap as large as 250 mev@xcite introduced a new puzzle and obstacle : the gapped bilayer is not as insulating as it should be .
dominant transport along physical edge of the samples proposed earlier by @xcite have been ruled out by corbino geometry measurements@xcite , which observed two - dimensional variable range hopping type temperature dependence , independent of geometry . in this paper
we predict existence of topological gapless channel of transport along recently imaged ab - ba tilt boundary network @xcite which solves the puzzle .
the predicted topological edge state holds the promise of the first realization of topological surface(edge ) state hosted by structural topological defect .
though there has been much theoretical interest in topological gapless modes hosted by structural topological defects@xcite no such topological gapless mode has been observed so far .
the lattice dislocations in three dimensional crystals previously discussed occur deep in the sample that is not directly accessible .
however , the tilt boundary of interest have recently been observed@xcite .
the tilt boundary is a structural topological line defect along which each neighboring layer is displaced by one inter - atomic spacing
. such defect can occur due to the third dimension added by the stacking of the graphene layers ; it forms a boundary between two inequivalent stacking structures frequently referred to as ab and ba .
here we show that the tilt boundaries host gapless modes of topological origin and form the first example of a naked structural defect hosting topological electronic states .
( @xmath2 ) sublattice sites .
@xmath3 represent hopping matrices for a tight - binding model .
, scaledwidth=50.0% ] topological aspects of gapped multi - layer graphene have been previously discussed@xcite and it was pointed out that they should exhibit quantum valley hall effect with corresponding edge states . however , to this date there has been no experimental detection of proposed edge state @xcite .
moreover , little is known about how the topological aspects of gapped multi - layer graphene relates to topological insulators @xcite .
the idea of classifying different topological insulator ( superconductor ) candidates based on symmetries @xcite have played a key role in the field of topological insulators . in particular the observation that additional symmetries such as the crystalline symmetries can enlarge the possibilities of topological phases@xcite led to the discovery of three - dimensional topological crystalline insulators@xcite .
on one hand we propose feasible experiments to detect topological edge states at naturally occurring tilt boundaries . at the same time , we make first concrete application of the spt approach @xcite for two dimensional ( 2d ) system and study a large class of gapped graphene systems placing the quantum valley hall insulator in the larger context and predicting conditions for topological superconductors .
the rest of the paper is organized as follows . in section
[ sec : micro ] we show that a ab - ba tilt boundary in gated bilayer graphene supports gapless edge states through explicit microscopic calculations .
specifically we consider an abrupt boundary in tight - binding model and then investigate the effect of strain using ab - initio calculation . in section [ sec : topo ] we show that these edge - states are protected by no valley mixing , electron number conservation , and time reversal ( @xmath1 ) symmetries within the framework of spt .
hence we identify chirally stacked gated @xmath4-layer graphene layers as _ time - reversal symmetric _
@xmath0-type spt . in section [ sec : expt ]
we discuss experimental implications .
finally in section [ sec : summary ] we summarize the results and comment on practical implications .
+ c + fig .
[ fig : bilayer ] and fig .
[ fig : domainwall](a ) show tilt boundaries of interest in gapped bernal stacked bi - layer graphene . in the case sketched , strain is concentrated at the tilt boundary with the top layer stretched by one inter - atomic spacing with respect to the bottom layer . for a general orientation , tilt boundaries can involve both strain and shear . as the tilt boundaries in layered graphene form a type of topological line defects in structure , they can be characterized using the tangent vector @xmath5 and the burger s vector @xmath6 .
the tangent vector @xmath5 points along the tilt boundary which can point along any direction with respect to the burger s vector @xmath6 .
when the tilt boundary only involves strain as in the case depicted in fig .
[ fig : bilayer ] and fig .
[ fig : domainwall](a ) , the @xmath6 is perpendicular to @xmath5 . in the opposite extreme limit of @xmath7 ,
shear is concentrated at the boundary .
independent of the angle between @xmath6 and @xmath5 , the burger s vector magnitude is the inter - atomic spacing i.e. @xmath8 for a bilayer system , as it is shown explicitly for the strain tilt boundary in fig .
[ fig : domainwall](a ) .
since @xmath9 is a fraction of the bravis lattice primitive vector magnitude @xmath10 , the bilayer domain boundary is a partial dislocation from quasi two - dimensional view . in a general mult - layer a vertical array of these partial dislocations form a tilt - boundary . in typical samples , the domain wall separating the ab and ba stacked domains has substantial width spanning 5 - 20 nm and the angle between @xmath6 and @xmath5 ranges between @xmath11 and @xmath12 @xcite .
[ fig : bilayer ] allow us to makes two important microscopic observations about tilt - boundaries .
( 1 ) as the boundary requires a shift of one - layer with respect to the other by one inter - atomic spacing along the bond direction , there are three natural directions for the tilt - boundary to run for each fixed angle between @xmath6 and @xmath5 .
( 2 ) the boundary is arm - chair for @xmath13 ( pure shear ) whereas it is zigzag for @xmath14 ( pure strain case shown in fig .
[ fig : bilayer ] ) . based on observation ( 1 )
we expect a given type of tilt boundary to possibly form a triangular network seen in experiments@xcite .
the observation ( 2 ) combined with earlier microscopic studies of boundary condition effects on edge states in ref .
@xcite implies that electronic spectrum at tilt - boundaries with @xmath13 will be gapped though the gap magnitude will be small when the tilt boundary is spread over finite width .
in this section , we consider the electronic structure of tilt boundaries with @xmath14 and return to more general case in the section [ sec : expt ] . as it is shown schematically in fig .
[ fig : bilayer ] , the bulk of each domain is gapped in the presence of inter - layer hopping and the external electric field . the latter is important for breaking the inversion symmetry between the layers and gapping otherwise touching bands @xcite .
below we present two separate microscopic calculations of the ab - ba tilt boundary electronic structure for @xmath14 , which shows gapless edge states .
we consider a tight - binding hamiltonian with nearest - neighbor intra- and inter - layer hopping . for the ab - stacking region ( see fig .
[ fig : domainwall](a ) ) , @xmath15 where @xmath16 is a layer index , @xmath17 and @xmath18 are intra - layer and interlayer nearest neighbor hopping respectively , and @xmath19 is the chemical potential difference between two layers due to the gate voltage .
( m , n ) labels the position of the two site unit - cell with @xmath20 and @xmath21 annihilating electrons at the two sites of layer @xmath22 . for the ba stacked region ,
the only change is in the inter - layer term with @xmath23 replacing the inter - layer term in @xmath24 .
as we address the effect of strain through ab - initio simulation , we focus here on a sharp tilt boundary as shown in fig .
[ fig : domainwall ] .
we plot the energy spectrum in fig .
[ fig : domainwall](b ) with the model parameters set to be @xmath25ev , @xmath26ev , and @xmath27ev .
the size of the system was 200 unit - cell in each direction under periodic boundary condition with each domain spanning 100 unit - cell width separated by two sharp tilt boundaries . from the spectra it is clear that @xmath28 and @xmath29 valleys each have two edge states per spin .
further investigation of the wave function shows right and left moving edge states associated with the given valley are spatially separated between the two edges : the edges offer valley filtering ( see fig.[fig : domainwall](c ) ) . .
the fermi level is indicated with the blue dashed line and the bands with marked linear dispersion relation intersecting near the fermi level are outlined by the purple ( light ) curves .
( b ) top and side views of the charge distribution for a state near the @xmath30 point of ( a ) .
the yellow rectangles indicate the ab - ba domain wall structure .
, scaledwidth=48.0% ] the electronic structure in realistic tilt boundary will be affected by both the span over which the lattice structure transition from ab to ba stacking , and the strain concentrated at the tilt boundary . to address these issues we further carried out first - principles calculation using density functional theory ( dft ) within the local density approximation ( lda ) @xcite .
we constructed a periodic supercell with two tilt boundaries characterized by burgers vectors identical to the ones considered within tight binding calculation , but the domain wall is set to have finite width over which the one unit - cell mismatch is spread .
we have tried several configurations with domains and tilt boundaries of different widths to find qualitatively similar results . in the rest of this paper we focus on a representative example with 3.1 nm wide domain wall between 1.5 nm wide domains .
the choice of narrow width for the domains was due to the limitation in the simulation capability .
to this 2-d system , we then applied a slightly exaggerative perpendicular electric field of 5 v / nm . by relaxing the domain boundaries until the forces reaches below 0.5 ev / nm
, we took the effect of both strain and the width into account .
we have tried several configurations with domains and tilt boundaries of different width but otherwise similar setting to find qualitatively similar results .
the dft - lda simulation results are presented in fig .
[ fig : ldaband ] , which confirms the existence of the gapless edge states predicted by the tight - binding model . the electronic structure along the extended direction of edge
is shown in fig . [
fig : ldaband](a ) , focusing on the region near @xmath28 point . in this figure two gapless 1d dirac dispersion
is clearly resolved from the gapped bulk specta , with two distinct dirac points @xmath30 and @xmath31 in the vicinity of the @xmath28 point .
the strain concentrated at the tilt boundary causes energy splitting of @xmath30 and @xmath31 states ; the energy of 1d dirac points is increased by compressive strain and decreased by tensile strain .
this energy splitting will become negligibly small in realistic tilt boundaries with much wider span , as the strain will become smaller .
[ fig : ldaband](a ) shows that the existence of gapless edge states found in our tight - binding calculation are robust against long range perturbations such as tilt boundary width or subtle bond - length variation inside the domain wall , as well as the strain at the tilt boundary .
we now turn to the spatial distribution of the gapless edge states . for illustration ,
[ fig : ldaband ] ( b ) shows the wavefunction amplitude of the state slightly above the @xmath30 point .
as expected from the tight - binding results , the charge distribution is prominent inside the domain wall but rapidly decays away from the tilt boundary .
the dirac points of edge states and the decaying feature of edge states always exist , making us believe it is a universal feature .
meanwhile , the charge for the states near @xmath30 is highly localized on the layer subject to compressive strain , regardless of the direction of the applied electric field .
the situation for the states in vicinity of @xmath31 is similar except that the charge prefer to highly localized on the layer under tensile strain . hence forming a layer selective contact to an isolated edge state
could be a mechanism for valley filtering@xcite . finally , we comment on the so far ignored effect of interaction .
if the edge boundaries have substantial width , forward scattering part of the coulomb interaction will be the dominant correlation effect to the edge states and lead to luttinger liquid behavior @xcite .
in order to address the robustness of the edge states , we investigate topological aspects of the low energy effective theory in the continuum limit . we first show that the ab - ba tilt boundary can be mapped to a gate - polarity boundary of uniform bi - layer . based on this mapping and results of refs .
@xcite on the gate - polarity boundary , we discuss the valley chern number of the tilt boundary edge states .
we than apply the notion of spt @xcite and identify chirally - stacked multi - layer graphene as a realization of @xmath32-type spt , protected by time - reversal ( @xmath1 ) , absence of valley mixing , charge conservation symmetries .
this identification enables us to address effects of symmetry changes : topological quantum phase transitions .
there are recent studies of such perturbations for specific cases such as rashba spin - orbit coupling @xcite and magnetic ordering @xcite . through our first application of spt classification scheme by @xcite to a concrete physical system of multi - layer graphene
, we obtain an exhaustive systematic study of topological quantum phase transition possibilities .
the low energy effective hamiltonian near the @xmath28 valley for uniformly ab or ba stacked bi - layer is @xmath33 where @xmath34 or @xmath35 sign should be used for ab or ba stacking respectively . in eq . , @xmath36 , @xmath37 s are pauli matrices acting on the sub - lattice indices , and @xmath38 s are pauli matrices acting on the layer indices .
the effective hamiltonian near @xmath29 is @xmath39 .
now it is straight forward to show that ba stacking is equivalent to ab stacking subject to the opposite gate polarity . at zero field
the hamiltonian of the ab stacked bilayer can be transformed to that of the ba stacking by interchanging the two layers via the following unitary transformation : @xmath40 , with @xmath41 . for a gated bi - layer however , the gate polarity has to flip since @xmath42 hence at the level of low energy effective theory , the tilt boundary between ab and ba stacking under uniform external field is equivalent to the gate polarity domain wall of structurally uniform bi - layer proposed by @xcite . the above equivalence combined with earlier results on valley chern number of gated chirally stacked multi - layer offers the topological origin of the helical edge states observed in the microscopic calculation of section [ sec : micro ] .
first for bi - layer and then for general @xmath4-layers , it was shown that low energy effective theory of chiral stacked @xmath4-layer under uniform vertical electric field @xmath19 has finite chern number per spin for each valley of equal magnitude and opposite sign @xcite : @xmath43 the chern numbers in eq .
can be obtained by integrating the berry curvature over momenta @xmath44 continuing the linearized dispersion to infinity .
this combined with the equivalence relation of eq .
means the valley chern numbers change sign at the tilt boundary .
such sign change leads to @xmath45 across the tilt boundary and @xmath4 branches of valley helical edge states @xcite , as long as the two valleys @xmath28 and @xmath29 remain distinct .
hence the two valley helical edge states per spin observed in section [ sec : micro ] originate from the valley chern number change across the tilt boundary , as in the gate polarity boundary edge states @xcite . hence , our prediction is the tilt boundaries will be the first experimentally observed crystalline topological defects to host topological gapless mode due to change in the chern number .
we now apply the procedure for identifying the class of spt based on symmetries of free fermion hamiltonian developed by @xcite , which predicts possible number of protected edge ( surface ) states .
this procedure allows us to consider additional symmetries in the multi - layer graphene in addition to the @xmath46 , @xmath1 , and @xmath47 taken into account in the pioneering work by @xcite , and by @xcite .
the procedure consists of three steps : ( 1 ) find a gapless dirac hamiltonian ( by keeping the kinetic term only ) with the same symmetries .
then we find all the symmetry preserving mass terms that can gap out the gapless part and are amenable to classification using clifford algebra .
this is based on the assumption that the spt order is robust as long as the energy gap stays finite and the symmetries remain the same and hence any gapped hamiltonian can be adiabatically transformed into a gapped dirac hamiltonian .
( 2 ) express the hamiltonian and the conserved quantities associated with symmetries in the majorana basis .
this leads to the clifford algebra ( i.e. real representation of the dirac algebra ) associated with the gapless part of the hamiltonian .
( 3 ) find the space of mass matrices that anti - commutes with all the generators of this clifford algebra .
the resulting space may have disconnected pieces , the number of which gives the classification of the spt .
two mass matrices are topologically distinct if and only if they belong to two different pieces . applying this procedure to chiral multi - layer graphene will enable us to study phase transitions into different spts upon symmetry changes . for chiral multi - layer graphene system , we assume no valley mixing , electron number conservation @xmath48 , and time reversal ( @xmath1 ) symmetries .
the relevant gapless dirac hamiltonian is : @xmath49 where @xmath50 , and @xmath51 , and @xmath52 , in which @xmath53 is given by the number of layers .
this hamiltonian can be written in the majorana fermion basis using the following decomposition @xmath54 where @xmath55 denotes the majorana fermion satisfying @xmath56 where @xmath57 ( @xmath58 ) denotes the a , or b sublattice ( @xmath28 or @xmath29 valley ) indices , and @xmath59 denotes the flavor of the majorana fermions ( + or - ) . in the majorana fermion basis the hamiltonian is represented as follows : @xmath60 where @xmath61 is a real anti - symmetric matrix ( differential operator ) , and @xmath62 is an eight component vector whose components are @xmath63 .
now we express the conserved quantities associated with the symmetries of the hamiltonian in the majorana fermion basis .
first , no valley mixing combined with total electron number conservation symmetry leads to separate conservation of the electron number at each valley @xmath64 and @xmath65 .
hence the total electron number @xmath66 , and the valley polarization @xmath67 are conserved . in the majorana fermion basis ,
@xmath68 where @xmath69 ( @xmath70 ) pauli matrices act on the majorana flavors ( valley indices ) .
so defined @xmath71 and @xmath72 satisfy @xmath73 . under time
reversal symmetry @xmath1 , @xmath28 and @xmath29 valley indices are exchanged i.e. , @xmath74 . hence @xmath1 acts like @xmath75 in the valley basis with the matrix part of the time reversal operator satisfying @xmath76 .
on the other hand , no valley mixing implies the hamiltonian is invariant under @xmath77 transformation which acts like @xmath78 in the valley basis .
hence , in the presence of no valley mixing symmetry , we can define a new time reversal operator @xmath79 , which acts like @xmath80 . in terms of majorana fermions
@xmath81 in order to find the relevant clifford algebra , we need to form anti - commuting generators in terms of @xmath61 in eq . combined with symmetries , @xmath82 , @xmath83 , and @xmath84
however , the symmetries require @xmath85=\left[\mathcal{a},\hat{q}_{v}\right]=0 $ ] and @xmath86=0 . moreover , symmetry operators satisfy @xmath87 , @xmath88 , and @xmath89 relations . using these relations
, we find the full set of generators of the relevant clifford algebra as @xmath90 , @xmath91 , and @xmath92 as @xmath93 for @xmath94 .
the resulting full set of anti - commutation relations is @xmath95 and it defines a clifford algebra cliff(3,2 ) .
now , we will find the space of mass matrices , @xmath96 , that can gap out dirac hamiltonian associated with this clifford algebra in order to obtain spt classification .
the mass term with matrix representation @xmath97 should satisfy the following algebra : @xmath98 where we have normalized @xmath99 . solving the above equation yields the allowed space for the mass matrix , @xmath100 .
it has been shown @xcite that @xmath96 which solves eq .
for the case of cliff@xmath101 is @xmath102 the spt classification is then given by the number of disconnected pieces in the space of mass matrix @xmath100 i.e. its zeroth homotopy group : @xmath103 .
using eq . , it can be verified that @xmath104 @xcite .
consequently , each class of the time reversal invariant multilayer graphene in the absence of intervalley scattering is indexed by a @xmath32-valued number : in this case the valley chern number .
[ tab1 ] [ cols="^,^,^,^,^",options="header " , ] now we are in a good position to consider symmetry changes . important to note here that spontaneously ordered phases
can be considered alongside systems under external field , as once a system is deep inside the ordered phase it can be treated within mean - field theory .
we first consider the symmetry reduction possibilities while maintaining spin degeneracy ( see table [ table : spinless ] ) .
if we only break the time reversal symmetry , the system is characterized by two independent topological indices @xmath105 , hence the classification is given by @xmath106 .
we refer to these spt phases as intra - valley quantum ( anomalous ) hall ( qah ) states .
such phases may be realized by placing trigonally - strained graphene @xcite under an external magnetic field , as the sign of the pseudo - magnetic field caused by strain is opposite for the two valleys .
further reducing the symmetry by introducing inter - valley scattering leads to inter - valley qah state indexed by a single integer @xmath32 : the total chern number . breaking
electron number conservation turns the above insulators into superconductors . following the procedure above
, we obtain the same classification for the topological superconductors in 2d , resulting in topological valley superconductor ( tvsc ) and intra- or inter - valley topological superconductors .
table [ table : spinless ] summarizes all symmetry reduction possibilities and their classifications starting form gated multi - layer graphene .
now we consider extending our classification to take the electron spin into account as a dynamical degree of freedom .
this will allow us to consider interaction effects at the level of spin ordering . with both spin and valley degrees of freedom ,
there are three symmetry operators related to time reversal : @xmath107 exchanges two valleys , @xmath108 which acts on the spin indices flips spin , and @xmath109 does both .
if both @xmath110 and @xmath111 are broken , the classification reduces to that of spinless electrons ( see upper part of table [ table : spinful ] ) .
however , taking any of these symmetry operators into account leads to new classes ( see lower part of table [ table : spinful ] ) .
when all @xmath112 , @xmath110 ( and as a result @xmath111 ) and no valley - mixing symmetries are imposed the state corresponds to the so - called `` layer antiferromagnetic '' ( laf ) phase predicted in refs .
@xcite and possibly occurring as a ground state in bi - layer graphene at neutrality point @xcite .
in this phase the product of spin and valley of edge quasi - particles is locked to their momentum .
therefore , one may index this state by its _ spin - valley chern number _
@xmath113 @xcite .
in this phase , for each valley , the edge states associated with two spins are counter - propagating . as long as
these two counter - propagating modes do not couple , there can be any number of them per each valley , leading to @xmath114 spin - valley hall conductivity @xcite .
however there is a form of symmetry allowed coupling between a pair of sets of counter propagating edges ; this coupling can gap out the edge modes @xcite .
therefore , the number of symmetry protected edge modes for each valley is @xmath113 mod @xmath115 .
hence , we obtain a @xmath116 classification for this phase labeled by @xmath117 as supposed to @xmath32 classification which would be implied by refs .
@xcite .
another interesting possibility is breaking @xmath112 , and @xmath110 , while respecting their product @xmath111 and no - valley mixing symmetry .
this leads to the quantum spin hall ( qsh ) phase , in which @xmath118 , and @xmath119 due to @xmath111 symmetry , while there is no constraint on the @xmath120 .
hence , unlike usual 2d qsh states @xcite , we obtain @xmath32 .
this is because multiple type of time reversal operators can be defined @xcite in the presence of no valley mixing symmetry .
this and other spt possibilities are summarized in table [ table : spinful ] . with the classification at hand , we return to what it implies for the fate of the gapless edge states at the tilt boundary of chirally stacked and gated n - layer graphene .
in general , a topological phase transition between two phases within the same class requires the bulk gap to close and reopen ( as in inter - plateaux transitions in integer quantum hall effect ) .
gapless edge states are guranteed at a physical boundary between such two phases . the edge states at the ab - ba tilt boundary we established in the section ii are examples of such edge states .
hence the edge states will remain gapless as long as the time ( valley ) reversal and charge conservation symmetry are maintained .
on the other hand , symmetry change can either yield a trivial phase which does not support an edge state or a different type of spt with different type of edge states .
according to tables i and ii , ruining the no valley mixing symmetry is the only way to render the system trivial and gap the edge states .
however this requires large momentum transfer which generally requires fine tuning unless the unit - cell becomes enlarged either for the entire system @xcite or for the edge through arm - chair edge , or short - range disorder such as a vacancy breaks @xmath121-@xmath122 sub - lattice symmetry @xcite .
as both rarely occur , we anticipate gapless edge states at most tilt boundaries .
in particular , the natural zigzag boundary formation for the tilt boundaries make such edge states more robust than the edge states in gate polarity boundaries @xcite . among symmetry change possibilities leading to another spt , transition from qvh with spin degeneracy to laf phase where spin is
a dynamic degree of freedom is of particular interest as laf is suspected to be the ground state of bi - layer graphene near neutrality point @xcite . upon this phase transition
the nature of edge states change from spin degenerate valley helical states ( qvh ) to spin - valley hall edge states ( laf ) .
in such transitions , the bulk gap has to close and reopen ; this is indeed seen in the experiment of @xcite .
in this section , we discuss the experimental implications of our findings . specifically we propose the transport through the network of tilt boundaries as a solution to the long standing mystery of sub - gap transport@xcite .
further we propose feasible experiments to test the proposal . in order to discuss the topological transport through the network of tilt boundaries observed in refs .
@xcite we should first discuss the effect of the arbitrary angle between the burger s vector @xmath6 and the tangent vector @xmath5 .
microscopic study of tilt boundaries at various angles between @xmath6 and @xmath5 will be presented in the future@xcite .
however , it has been known that a single domain bernal stacked gapped bilayer ribbon should support gapless edge state for zig - zag edges , but not for arm - chair edges @xcite .
this is because the arm - chair edge enlarges the unit - cell along the direction parallel to the boundary and makes the projection of k and k@xmath123 valley identical .
however , @xcite showed that the polarity boundary edge states only develop barely visible gap which is orders of magnitude less then the bulk gap even for a sharp boundary , and the gap decreases quickly when the polarity boundary becomes smooth .
these arguments apply to our tilt boundary and the edge state will develop a small gap when @xmath13 .
however , given the large width of the observed tilt boundaries we expect that all straight tilt boundaries will have nearly gapless edge states except those with small angles between @xmath6 and @xmath5 .
when the tilt boundary meanders and changes directions , likely there will be portions with small gap segmenting the gapless regions and the transport will occur through hopping between the gapless regions . the observed 2d network of such tilt boundaries would yield 2d variable range hopping temperature dependence @xmath124@xcite at low temperatures governed by the 2d connectivity and the small characteristic gaps of gapped regions .
this explains the observed temperature @xmath1 dependence of resistance at low temperatures @xcite .
we propose following experiments to test our proposal .
( 1 ) four terminal transport measurements with two of the contacts , say contacts 1 and 3 , at two ends of a tilt boundary .
this would yield highly anisotropic transport proving dominant transport along the tilt boundary i.e. , @xmath125 .
( 2 ) scanning tunneling spectroscopy measurements of local density of states .
this should measure a gapless spectrum at the tilt boundary but exhibit a gapped spectrum with the gap magnitude of the optical gap away from the tilt boundary .
( 3 ) thermoelectric imaging .
the mid - gap density of state at tilt boundaries would appear in scanning thermopower images .
unpublished thermopower imaging data by @xcite indeed show a network with local density of state near fermi energy , that is reminiscent of the tilt boundary network .
( 4 ) edge current imaging using scanning squid which can detect magnetic field generated by edge currents .
we showed that spin - degenerate tilt boundaries of gated multi - layer graphene support topological gapless edge states protected by three symmetries : time ( valley ) reversal , no - valley - mixing , and electron number conservation .
we demonstrated the existence of gapless edge states through a tight - binding model calculation and a first principal calculation , where the latter took strain effects into account .
we then addressed the symmetry protection of the edge states and consequences of symmetry changes within the framework of spt @xcite .
the framework of spt allowed us to place the 2d topological phase supporting the edge states , namely qvh , among various topological insulator / superconductor phases alongside previously postulated qah , laf and qsh .
while previous literature postulated qvh , qah , laf and qsh to be all supporting number of edge states growing with the number of layers @xmath4 ( i.e. @xmath0-type in the language of classification ) , we found that the symmetry of laf only protects odd number of edge modes for each valley .
hence laf is a @xmath116-topological insulator much like quantum spin hall insulator@xcite .
transition between these different spt s require closing and re - opening of the bulks gap as already been observed in ref .
@xcite .
we predict the naturally occurring tilt boundary @xcite to be the first topological structural defect hosting topologically protected gapless mode of transport , most importantly , our findings on tilt boundaries combined with the recent observations @xcite solve the long standing mystery of sub - gap transport @xcite .
our explanation can be tested through proposed transport , scanning tunneling spectroscopy and thermopower imaging experiments , and scanning squid experiments .
experimental confirmation of the tilt - boundary transport origin of the sub - gap transport will open doors to control the sub - gap transport and enable device application of gated multi - layer graphene systems .
: we thank paul mceuen for numerous discussions and sharing his unpublished data .
we thank joe stroscio for sharing his unpublished data on thermopower imaging .
we thank e.j .
mele for useful discussions and sharing the preprint @xcite and j. sethna for useful discussions regarding structural topological defect aspect of the tilt boundary .
e .- a.k . and d.n .
were supported by nsf award eec-0646547 through cornell center for nanoscience .
e .- a.k . and a.v .
were supported in part by nsf career grant dmr-0955822 .
was also supported in part by nsf grant dmr-1120296 through cornell center for materials research .
y.l . and l.y . were supported by nsf grant no .
the computational resources have been provided by lonestar of teragrid at the texas advanced computing center ( tacc ) .
45ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1038/nmat2082 [ * * , ( ) ] link:\doibase 10.1103/physrevb.74.161403 [ * * , ( ) ] http://dx.doi.org/10.1038/nature08105 [ * * , ( ) ] http://dx.doi.org/10.1038/nphys1822 [ * * , ( ) ] link:\doibase
10.1021/nl102459 t [ * * , ( ) ] , @noop `` , '' @noop ( ) , @noop * * , ( ) http://dx.doi.org/10.1038/nphys1220 [ * * , ( ) ] link:\doibase 10.1103/physrevx.2.031013 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.115120 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.99.236809 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.100.036804 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.95.226801 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.96.106802 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.195125 [ * * , ( ) ] link:\doibase 10.1103/physrevb.85.085103 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.106.106802 [ * * , ( ) ] http://dx.doi.org/10.1038/nmat3449 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.075418 [ * * , ( ) ] link:\doibase 10.1103/physrev.136.b864 [ * * , ( ) ] link:\doibase 10.1103/physrev.140.a1133 [ * * , ( ) ] http://dx.doi.org/10.1038/nphys547 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.104.216406 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.107.256801 [ * * , ( ) ] link:\doibase 10.1103/physrevb.83.155447 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.195438 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.106.156801 [ * * , ( ) ] in link:\doibase 10.1063/1.3149495 [ _ _ ] , , vol . , ( ) pp . , link:\doibase 10.1126/science.1191700 [ * * , ( ) ] http://dx.doi.org/10.1038/nphys1420 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.205106 [ * * , ( ) ] http://dx.doi.org/10.1038/nnano.2011.251 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.95.146802 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.205116 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.98.186809 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.98.266402 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.89.266603 [ * * , ( ) ] @noop `` , '' link:\doibase 10.1103/physrevlett.100.026802 [ * * , ( ) ] @noop * * , ( ) @noop * * @noop `` , '' @noop `` , '' | despite much interest in engineering new topological surface(edge ) states using structural defects , such topological surface states have not been observed yet .
we show that recently imaged tilt boundaries in gated multi - layer graphene should support topologically protected gapless edge states .
we approach the problem from two perspectives : the microscopic perspective of a tight - binding model and an ab - initio calculation on a bilayer , and the symmetry protected topological ( spt ) states perspective for a general multi - layer .
hence we establish the tilt boundary edge states as the first concrete example of edge states of symmetry enriched @xmath0-type spt , protected by no valley mixing , electron number conservation , and time reversal @xmath1 symmetries .
further we discuss possible phase transitions between distinct spt s upon symmetry changes .
combined with recently imaged tilt boundary network , our findings offer a natural explanation for the long standing puzzle of sub - gap conductance in gated bilayer graphene , which can be tested through future transport experiments on tilt boundaries . in particular , the tilt boundaries offer an opportunity for in - situ imaging of topological edge transport |
in the last years the physics of quasi - one - dimensional compounds has been the object of intense theoretical and experimental investigations . in this work
we study the ground state ( gs ) and magnetic excitations of two bipartite chains motivated by low - dimensional inorganic @xcite and organic @xcite ferrimagnets : the @xmath1 chain in fig . [ fsmhei](a ) and the @xmath2 chain in fig .
[ fsmhei](b ) , respectively .
we shall use the simplest approach for interacting electron systems on a lattice with @xmath3 sites , namely the one - band hubbard model ( hm ) : @xmath4 where @xmath5 ( @xmath6 ) is the creation ( annihilation ) operator for electrons with spin @xmath7 at site @xmath8 ; @xmath9 is the hopping parameter and @xmath0 is the on site coulomb repulsion . for the extremely localized regime ( @xmath10 ) and
@xmath11 , where @xmath12 is the number of particles , the hm can be mapped onto the antiferromagnetic heisenberg model ( ahm ) @xcite : @xmath13 where @xmath14 . we should mention that macdo et al .
@xcite , using a variety of numerical methods , have derived the ferrimagnetic nature of @xmath1 chains modeled by the hm in the strong and weak coupling regimes .
here we focus on the half - filled band case ( @xmath15 ) and extend these previous studies to much larger systems using lanczos exact diagonalization technique .
it is also worth mentioning that the two lattices present three electronic bands in the tight - binding limit ( @xmath16 ) : one flat band at @xmath17 for the two chains ; and two dispersive ones , @xmath18 and @xmath19 ^ 2}$ ] , with @xmath20 and @xmath21 for the @xmath1 and @xmath2 chains , respectively .
the flat band is closely associated with ferrimagnetic properties of both chains at half - filling @xcite .
a particular feature of the @xmath1 chain is a local invariance under the exchange of the @xmath22 sites in any cell @xmath23 @xcite .
the eigenvalues of the exchange operator being @xmath24 .
this symmetry leads to a conserved local spatial parity that in the heisenberg limit asserts that the @xmath22 sites form either a singlet ( @xmath25 ) or a triplet ( @xmath26 ) bond state .
therefore for the symmetry sector in which all @xmath26 the spectrum is identical to that of the alternating spin-@xmath27/spin-@xmath28 ahm chain @xcite .
a theorem due to lieb and mattis @xcite asserts that the gs of the ahm on a bipartite lattice has total spin @xmath29 , where @xmath30 and @xmath31 are the number of lattice sites at the sub - lattices @xmath28 and @xmath32 , respectively .
so , if @xmath33 the system display unsaturated ferromagnetic ( f ) gs . the coexistence of both f and antiferromagnetic ( af ) long - range order implies in ferrimagnetism , as rigorously proved by tian @xcite .
another crucial step was provided by lieb @xcite , who proved that the gs of the hm for @xmath15 has @xmath29 for any @xmath34 . in this case ,
tian and collaborators @xcite further established the ferrimagnetic ( ferri ) long - range order of the gs if @xmath33 .
the unit cell of the two chains of interest has @xmath32 sites in one sublattice and @xmath28 in the other ( see fig .
[ fsmhei ] ) , so that @xmath35 , where @xmath36 is the number of unit cells . however , although ferrimagnetism is expected , the specific magnetic structure of a unit cell may strongly depend on the coulomb coupling , as well as on quantum fluctuations , and is not known _ a priori _ from the more general results of the theorems stated above .
we can probe the magnetic order through the magnetic structure factor ( msf ) : @xmath37 which is related to the zero - field static magnetic susceptibility by @xmath38 , where @xmath39 is the boltzmann constant and @xmath40 is the temperature .
the condition for a f ( af ) ordered state is that @xmath41\sim n$ ] , so that in a long - range ferri gs the two conditions must be fulfilled .
this is indeed the case for the two chains , as shown in figs .
[ fsmhei ] and [ fsmhub ] , both in the strong and weak coupling limits . for the @xmath42 ( a ) and @xmath2 ( b ) chains in the heisenberg limit ( @xmath10 ) .
the size ranges from @xmath43 to @xmath44 .
the insets display the size dependence of the ferromagnetic [ @xmath45 and antiferromagnetic [ @xmath46 peaks .
dashed lines are guides for the eye.,title="fig:",width=249 ] for the @xmath42 ( a ) and @xmath2 ( b ) chains in the heisenberg limit ( @xmath10 ) .
the size ranges from @xmath43 to @xmath44 .
the insets display the size dependence of the ferromagnetic [ @xmath45 and antiferromagnetic [ @xmath46 peaks .
dashed lines are guides for the eye.,title="fig:",width=241 ] for the @xmath42 ( a ) and @xmath2 ( b ) chains using the hm for @xmath47 .
the size ranges from @xmath48 to @xmath49 for the @xmath1 chain and to @xmath50 for the @xmath2 .
the inset presents the size dependence of the ferromagnetic [ @xmath45 and antiferromagnetic [ @xmath46 peaks .
dashed lines are guides for the eye.,title="fig:",width=245 ] for the @xmath42 ( a ) and @xmath2 ( b ) chains using the hm for @xmath47 .
the size ranges from @xmath48 to @xmath49 for the @xmath1 chain and to @xmath50 for the @xmath2 .
the inset presents the size dependence of the ferromagnetic [ @xmath45 and antiferromagnetic [ @xmath46 peaks .
dashed lines are guides for the eye.,title="fig:",width=245 ] due to the critical nature of both chains at low temperatures , the correlation length @xmath51 and @xmath52 satisfy power law behaviour : @xmath53 and @xmath54 as @xmath55 . since @xmath56 at @xmath57 , using scaling arguments and the results of fig .
[ fsmhei ] , we have @xmath58 , i. e. , @xmath59 , in agreement with the values @xmath60 and @xmath61 derived using renormalization group techniques @xcite . in fig .
[ smfu ] we present the local magnetization ( @xmath62 ) and the average local value of double occupancy ( @xmath63 ) as function of @xmath0 . since the system has a spontaneously broken @xmath64 symmetry , we can choose the spin sector @xmath65 to study these quantities , where @xmath66 is the total spin @xmath67 component .
in the heisenberg limit ( @xmath10 , @xmath63=0 ) , the bulk magnetization estimated from chains up to @xmath44 are : @xmath68 and @xmath69 , for the @xmath42 chain , in agreement with density matrix renormalization group ( dmrg ) calculations @xcite and with values for @xmath70 and @xmath71 in the spin-@xmath27/spin-@xmath28 chain @xcite . for the @xmath2 chain , @xmath72 , @xmath73 and @xmath74 .
notice that a ferri structure sustain quantum fluctuations , which strongly reduce the local magnetization from the classical value , although in any case the unit cell magnetization remains @xmath75 .
in the non - interacting limit ( @xmath16 ) we see in figs .
[ smfu](a ) and [ smfu](b ) that @xmath76 and @xmath77 for the @xmath1 ( @xmath2 ) chain .
these results being related to the band structure : filling the bands up to @xmath15 and choosing the spin sector @xmath65 , there exist @xmath78 unpaired electrons in the flat band which occupy @xmath79 and @xmath80 sites only .
it is important to note that in this limit the global gs does not have an unique value of @xmath81 , but rather @xmath82 .
as the coulomb coupling is turned on , the gs , with a unique @xmath35 , evolves adiabatically from the weak to the strong coupling regime ( fig .
[ smfu ] ) . and average site double occupancy @xmath63 as function of @xmath0 for @xmath50 in the @xmath42 and @xmath2 chains in the spin sector @xmath65 .
dashed lines are guides for the eye.,title="fig:",width=249 ] and average site double occupancy @xmath63 as function of @xmath0 for @xmath50 in the @xmath42 and @xmath2 chains in the spin sector @xmath65 . dashed lines are guides for the eye.,title="fig:",width=245 ]
systems with a ferri gs naturally have f ( lowering the gs spin ) and af ( rising the gs spin ) magnons as their elementary magnetic excitations .
the two chains have three spin wave branches : an af mode , defined as @xmath83 ; and two f ones , derived from @xmath84 , where @xmath85 is the gs energy and @xmath86 are lowest excitation energies in the sector @xmath87 , with @xmath88 and @xmath89 .
the acoustical and optical f modes are identified from symmetry considerations .
these modes are depicted in fig .
[ enq ] for the two chains using the ahm : the af mode has a gap , denoted by @xmath90 ; the f1 mode is gapless , i. e. , the goldstone mode , consistent with the symmetry broken phase of the chains ; the gap of the f2 mode is denoted by @xmath91 . for the @xmath42 chain the gapped f2 branch is flat , fig .
[ enq](a ) , and the other modes are dispersive .
since these dispersive modes preserve the local triplet bond , they are identical to those found in the spin-@xmath27/spin-@xmath28 @xcite chain .
surprisingly , linear spin wave theory ( lswt ) @xcite predicts that @xmath92 , very close to our estimated value : @xmath93 .
moreover , a good agreement is found for the gapless f1 branch in the long wave - length limit
. however , both lswt and mean field ( mf ) theory @xcite predicts @xmath94 , deviating from our estimated exact diagonalization value : @xmath95 , which is in excelent agreement with numerical and analytical calculations for the spin-@xmath27/spin-@xmath28 chain @xcite . on the other hand ,
the interacting spin wave theory @xcite derives a better result for @xmath90 , but it implies in a higher shift for @xmath91 ( flat mode ) not observed in our data of fig .
[ enq](a ) .
this flat mode indicates the localized nature of the excitation , which is associated with the formation of a singlet state ( even parity symmetry ) between the @xmath22 sites in one cell , while the other cells have odd parity symmetry ( @xmath22 sites in a triplet state ) .
thus , since the symmetries of the f modes ( dispersive and flat ) are distinct , a possible level crossing is not avoided , as in fact observed in the data of fig .
[ enq](a ) : for @xmath96 the localized excitation is below the dispersive one . ( a ) and @xmath2 ( b ) chains for @xmath44 ( circles ) , 8 ( triangles down ) , 6 ( triangles up ) .
solid lines are the lswt results from ref .
@xcite ( @xmath1 ) and ref .
@xcite ( @xmath2 ) ; dashed lines are guides for the eye.,title="fig:",width=245 ] ( a ) and @xmath2 ( b ) chains for @xmath44 ( circles ) , 8 ( triangles down ) , 6 ( triangles up ) .
solid lines are the lswt results from ref .
@xcite ( @xmath1 ) and ref .
@xcite ( @xmath2 ) ; dashed lines are guides for the eye.,title="fig:",width=245 ] [ enq ] for the @xmath2 chain we can see in fig .
[ enq](b ) that the three branches are dispersive . like in the @xmath42 chain ,
the af mode is gapped with @xmath97 .
the lswt @xcite conducts to a qualitative but not quantitative correct af branch .
however , a good agreement is found for the gapless f1 mode in the long - wavelength limit .
moreover , the gap for the f2 mode is estimated to be @xmath98 .
we can clearly observe in fig .
[ enq](b ) the level repulsion between the f branches , which in this case have the same symmetry .
the af gaps found above are responsible for plateaus in the magnetization per spin [ @xmath99 as function of the applied magnetic field @xmath100 .
in fact , it has been shown @xcite that if @xmath101 a plateau may appear in the magnetization curve of the ahm . in the last equation ,
@xmath102 is the site total spin , @xmath103 is the magnetization and @xmath104 is the number of sites in one unit period of the gs for a given value of @xmath100 .
the hamiltonians that we are considering here have @xmath105 and three sites per unit cell , so , unless the system spontaneously breaks the translation invariance ( e.g. , by growing the unit cell ) , we expect a plateau at @xmath106 .
this is indeed observed in fig .
[ maghei ] and its width matches exactly @xmath90 , which measures the region of stability of the ferri phase . for higher fields
the magnetization increases in the expected way @xcite , as shown in the inset ( 2 ) of figs .
[ maghei ] ( a ) and [ maghei ] ( b ) , before the magnetization saturates at @xmath107 for @xmath108 and @xmath109 , respectively .
this field dependent behaviour contrasts with the linear one predicted by mf theory @xcite . for the @xmath42 ( a ) and @xmath2 ( b ) chains using the ahm .
the inset ( 1 ) presents a finite size scaling study for @xmath110 ; the inset ( 2 ) is a curve traced from the midpoints of the steps using the finite size results , except at plateau regions.,title="fig:",width=245 ] for the @xmath42 ( a ) and @xmath2 ( b ) chains using the ahm .
the inset ( 1 ) presents a finite size scaling study for @xmath110 ; the inset ( 2 ) is a curve traced from the midpoints of the steps using the finite size results , except at plateau regions.,title="fig:",width=245 ] now we turn to study the evolution of the af and the f2 spin wave gaps as function of @xmath0 using the hm : @xmath111 ( plateau width ) and @xmath112 , respectively . in order to reduce finite size effects in calculating @xmath113
we have used the boundary condition that minimizes the gap between the gs and excited levels .
our estimated curves are depicted in fig .
[ gapsfu ] using the scaling _ ansatz _
@xcite for extrapolation : @xmath114 , with @xmath115 for the @xmath1 chain ; @xmath116 for the @xmath2 chain .
note that @xmath117 and @xmath118 for the @xmath1 and @xmath2 chains , respectively .
this can be understood by noticing that @xmath119 is given by the gap between the two dispersive tight - binding electronic bands .
moreover , @xmath120 , @xmath121 , for the @xmath1 chain , whereas for the @xmath2 chain @xmath122 increases linearly before reaching the tight - binding limit @xmath118 . for both chains ,
we observe a crossover to the heisenberg limit : @xmath123 , in agreement with the results of fig .
further , we notice that @xmath124 nullifies linearly with @xmath0 , for both chains , following the behaviour of the gap between the flat electronic bands @xcite . ) and f2 ( @xmath91 ) spin wave gaps in units of @xmath9 for the @xmath1 ( a ) and @xmath2 ( b ) chains as function of @xmath0 using the hm .
solid ( dashed ) line is the af ( f2 ) gap for the ahm with @xmath125 ; dotted lines are guides for the eye.,title="fig:",width=257 ] ) and f2 ( @xmath91 ) spin wave gaps in units of @xmath9 for the @xmath1 ( a ) and @xmath2 ( b ) chains as function of @xmath0 using the hm .
solid ( dashed ) line is the af ( f2 ) gap for the ahm with @xmath125 ; dotted lines are guides for the eye.,title="fig:",width=245 ]
in this work we reported a detailed exact diagonalization study of the half - filled hubbard model on bipartite quasi - one - dimensional lattices .
these chains have experimental motivation in inorganic and organic ferrimagnetic polymers .
we have focused on the description of their ferrimagnetic ground state properties and magnetic excitations . in particular , we have revealed the non - trivial physical behaviour exhibited by these chains as function of the coulomb coupling strength .
we acknowledge several useful discussion with a. l. malvezzi .
this work was supported by cnpq , finep , facepe and capes ( brazilian agencies ) .
j. silvestre and r. hoffman , inorg .
( 1985 ) 4108 ; m. drillon _
et al _ , j. magn .
128 ( 1993 ) 83 .
y. v. korshak , t. v. medvedera , a. a. ovchinnikov and v. n. spector , nature 326 ( 1987 ) 370 ; h. nishide , adv .
( 1995 ) 937 .
e. p. raposo and m. d. coutinho - filho , mod .
b 9 ( 1995 ) 817 .
a. m. s. macdo , m. c. dos santos , m. d. coutinho - filho , and c. a. macdo , phys .
74 ( 1995 ) 1851 .
h. tasaki , prog . of theor .
99 ( 1998 ) 489 , and references therein .
f. c. alcaraz and a. l. malvezzi , j. phys .
a : math . gen . 30
( 1997 ) 767 .
e. h. lieb and d. c. mattis , j. math .
phys . 3 ( 1962 ) 749 .
tian , j. phys . a : math .
( 1994 ) 2305 .
e. h. lieb , phys .
62 ( 1989 ) 1201 .
shen , z .- m . qiu and g .- s .
tian , phys .
72 ( 1994 ) 1280 ; g .- s .
tian and t .- h .
lin , phys .
b 53 ( 1996 ) 8196 .
e. p. raposo and m. d. coutinho - filho , phys .
b 59 ( 1999 ) 14384 ; phys
78 ( 1997 ) 4853 .
g. sierra , m. a. martn - delgado , s. r. white , d. j. scalapino and j. dukelsky , phys .
b 59 ( 1999 ) 7973 .
s. yamamoto , s. brehmer and h .- j .
mikeska , phys .
b 57 ( 1998 ) 13610 ; n. b. ivanov , phys .
b 62 ( 2000 ) 3271 ; s. yamamoto , phys .
b 69 ( 2004 ) 064426 . c. vitoriano , f. b. de brito , e. p. raposo and m. d. coutinho - filho , mol . cryst .
cryst . 374
( 2002 ) 185 . c. vitoriano , m. d. coutinho - filho and e. p. raposo , j. phys .
35 ( 2002 ) 9049 .
t. nakanishi and s. yamamoto , phys .
b 65 ( 2002 )
214418 . f. b. de brito , m. sc .
dissertation , universidade federal de pernambuco , 2001 . for the @xmath2 an analytical solution
is found only for @xmath126 , in which case the gapped f mode is flat .
m. oshikawa , m. yamanaka , and i. affleck , phys .
78 ( 1997 ) 1984 .
l. capriotti , f. becca , s. sorella and a. parola , phys .
rev b 67 ( 2003 ) 172404 . | we present an exact diagonalization study of the half - filled hubbard model on bipartite quasi - one - dimensional lattices .
in particular , we emphasize the dependence of the ferrimagnetic ground state properties , and its associated magnetic excitations , on the coulomb repulsion @xmath0 . |
our task is to compute the maximal quantum violation @xmath80 of a two - party bell inequality defined by the vector of coefficients @xmath81 , given a fixed set of measurements operators @xmath82 for alice .
let s write @xmath83 for the assemblage created on alice s side by bob s measurements on the state @xmath0 . with this
, we have the conditional probabilities @xmath84 and we have to maximize @xmath85 for fixed @xmath86 and @xmath1 .
the following sdp program is a relaxation of the above problem : @xmath87 it is well known that one can always find a quantum state @xmath0 and quantum measurements @xmath88 for bob which attain the maximum @xmath89 .
hence , @xmath90 , and the above sdp provides the exact quantum bound of @xmath80 for a fixed set of alice s measurements @xmath1 on the bell inequality defined by coefficients @xmath81 . | we investigate the relation between the incompatibility of quantum measurements and quantum nonlocality .
we show that a set of measurements is not jointly measurable ( _ i.e._incompatible ) if and only if it can be used for demonstrating einstein - podolsky - rosen steering , a form of quantum nonlocality .
moreover , we discuss the connection between bell nonlocality and joint measurability , and give evidence that both notions are inequivalent .
specifically , we exhibit a set of incompatible quantum measurements and show that it does not violate a large class of bell inequalities .
this suggest the existence of incompatible quantum measurements which are bell local , similarly to certain entangled states which admit a local hidden variable model . the correlations resulting from local measurements on an entangled quantum state can not be explained by a local theory .
this aspect of entanglement , termed quantum nonlocality , is captured by two inequivalent notions , namely bell nonlocality @xcite and epr steering @xcite .
the strongest form of this phenomenon is bell nonlocality , witnessed via the violation of bell inequalities .
steering represents a strictly weaker form of quantum nonlocality @xcite , witnessed via violation of steering inequalities @xcite .
both aspects have been extensively investigated in recent years , as they play a central role in the foundations of quantum theory and in quantum information processing .
interestingly quantum nonlocality is based on two central features of quantum theory , namely entanglement and incompatible measurements . specifically , performing ( i ) arbitrary local measurements on a separable state , or ( ii ) compatible measurements on an ( arbitrary ) quantum state
can never lead to any form of quantum nonlocality .
hence the observation of quantum nonlocality implies the presence of both entanglement and incompatible measurements .
it is interesting to explore the converse problem .
two types of questions can be asked here ( see fig .
[ fig1 ] ) : ( a ) do all entangled states lead to quantum nonlocality ?
( b ) do all sets of incompatible measurements lead to quantum nonlocality ?
an intense research effort has been devoted to question ( a ) .
first , it was shown that all pure entangled states violate a bell inequality @xcite , hence also demonstrating epr steering .
for mixed states , the situation is much more complicated .
there exist entangled states which are local , in the sense that no form of quantum nonlocality can be demonstrated with such states when using non - sequential measurements @xcite .
these issues become even more subtle when more sophisticated measurement scenarios are considered @xcite .
question ( b ) has received much less attention so far . in the case of projective measurements
, it was shown that incompatible measurements can always lead to bell nonlocality @xcite .
note that in this case , compatibility is uniquely captured by the notion of commutativity @xcite .
however , for general measurements , _ i.e._positive - operator - valued - measures ( povms ) , no general result is known . in this case
, there are several inequivalent notions of compatibility . here
we focus on the notion of joint measurability , see e.g.@xcite , as this represents a natural choice in the context of quantum nonlocality .
several works discussed question ( b ) for povms @xcite .
the strongest result is due to wolf et al .
@xcite , who showed that any set of two incompatible povms with binary outcomes can always lead to violation of the clauser - horne - shimony - holt bell inequality .
however , this result may not be extended to the general case ( of an arbitrary number of povms with arbitrarily many outcomes ) , since pairwise joint measurability does not imply full joint measurability in general @xcite . here
we explore the relation between compatibility of general quantum measurements and quantum nonlocality .
we start by demonstrating a direct link between joint measurability and epr steering .
specifically , we show that for any set of povms that is incompatible ( _ i.e._not jointly measurable ) , one can find an entangled state , such that the resulting statistics violates a steering inequality .
hence the use of incompatible is a necessary and sufficient ingredient for demonstrating epr steering .
this raises the question of how joint measurability relates to bell nonlocality . specifically , the question is whether , for any set of incompatible povms ( for alice ) , one can find an entangled state and a set of local measurements ( for bob ) , such that the resulting statistics violates a bell inequality .
here we give evidence that the answer is negative .
in particular , we exhibit sets of incompatible measurements which can provably not violate a large class of bell inequalities ( including all full correlation bell inequalities , also known as xor games , see @xcite ) .
we therefore conjecture that non joint measurability and bell nonlocality are inequivalent . hence , similarly to local entangled states
, there may exist incompatible quantum measurements which are bell local . _ steering vs joint measurability .
_ we start by defining the relevant scenario and notations .
we consider two separated observers , alice and bob , performing local measurements on a shared quantum state @xmath0 .
alice s measurements are represented by operators @xmath1 such that @xmath2 , where @xmath3 denotes the choice of measurement and @xmath4 its outcome . upon performing measurement @xmath3 , and obtaining outcome @xmath4 , the ( unnormalized ) state held by bob
is given by @xmath5 the set of unnormalised states @xmath6 , referred to as an _ assemblage _ , completely characterizes the experiment , since @xmath7 is the probability of alice getting the output @xmath4 ( for measurement @xmath3 ) and given that information and bob s state is described by @xmath8 .
importantly , one has that @xmath9 for all measurements @xmath10 and @xmath11 , ensuring that alice can not signal to bob . in a steering test @xcite ,
alice want to convince bob that the state @xmath0 is entangled , and that she can steer his state .
bob does not trust alice , and thus wants to verify alice s claim . asking alice to perform a given measurement @xmath3 , and to announce the outcome @xmath4
, bob can determine the assemblage @xmath12 via local quantum tomography . to ensure that steering did indeed occur ,
bob should verify that the assemblage does not admit a decomposition of the form @xmath13 where @xmath14 . clearly ,
if a decomposition of the above form exists , then alice could have cheated by sending the ( unentangled ) state @xmath15 to bob and announce outcome @xmath4 to bob according to the distribution @xmath16 .
note that here @xmath17 represents a local variable of alice , representing her choice of strategy .
assemblages of the form are termed unsteerable and form a convex set @xcite .
hence any steerable assemblage can be detected via a set of linear witnesses called steering inequalities @xcite . by observing violation of a steering inequality
, bob will therefore be convinced that alice can steer his state . for a demonstration of steering , it is necessary for the state @xmath0 to be entangled .
however , not all entangled states can be used to demonstrate steering @xcite ; at least not when non - sequential measurements are performed on a single copy of @xmath0 .
moreover , steering also requires that the measurements performed by alice are incompatible . to capture the compatibility of a set of quantum measurements we use here the notion of joint measurability , see e.g. @xcite
a set of @xmath18 povms @xmath1 is called jointly measurable if there exists a measurement @xmath19 with outcome @xmath20 $ ] where @xmath21 gives the outcome of measurement @xmath3 , that is @xmath22 where @xmath23 stands for the elements of @xmath24 except for @xmath25 .
hence , all povm elements @xmath1 are recovered as marginals of the _ mother observable _
@xmath19 .
importantly , the joint measurability of a set of povms does not imply that they commute @xcite .
hence joint measurability is a strictly weaker notion of compatibility for povms . moreover , joint measurability is not transitive .
for instance ,
pairwise joint measurability does not imply full joint measurability in general @xcite ( see below ) .
our main result is to establish a direct link between joint measurability and steering .
specifically , we show that a set of povms can be used to demonstrate steering if and only if it is not jointly measurable .
more formally we prove the following result .
the assemblage @xmath26 , with @xmath27 , is unsteerable for any state @xmath0 acting in @xmath28 if and only if the set of povms @xmath29 acting on @xmath30 is jointly measurable .
the if part is straightforward .
our goal is to show that @xmath26 admits a decomposition of the form when @xmath29 is jointly measurable , for any state @xmath0 .
consider @xmath19 , the mother observable for @xmath29 , and define alice s local variable to be @xmath31 , distributed according to @xmath32 , where @xmath33 .
next alice sends the local state @xmath34 . when asked by bob to perform measurement @xmath3 , alice announces an outcome @xmath4 according to @xmath35 .
we now move to the only if part .
consider an arbitrary pure state @xmath36 with schmidt number @xmath37 .
notice that we can always write @xmath38 , where @xmath39 is an ( unormalized ) maximally entangled state in @xmath40 , and @xmath41 is diagonal matrix that contains only strictly positive numbers .
the assemblage resulting from a set of povms @xmath29 on @xmath0 is given by @xmath42 where @xmath43 is the transpose of @xmath44 .
our goal is now to show that if @xmath45 is unsteerable then @xmath29 is jointly measurable .
as @xmath45 is unsteerable , we have that @xmath46 which allows us to define the positive definite operator @xmath47 form which we can recover the assemblage @xmath48 as marginals , _
i.e._@xmath49 .
since the diagonal matrix @xmath41 is invertible , we can define @xmath50 .
it is straightforward to check that @xmath19 is a mother observable for @xmath51 : ( i ) it is positive , ( ii ) sums to identity , and ( iii ) has povm elements @xmath44 as marginals .
hence @xmath51 is jointly measurable , which concludes the proof .
note finally an interesting point that follows from the above . considering a set of incompatible measurements acting on @xmath30 , any pure entangled state of the schmidt number @xmath37
can be used to demonstrate epr steering .
_ bell nonlocality vs joint measurability .
_ it is natural to ask whether the above connection , between joint measurability and steering , can be extended to bell nonlocality .
recall that in a bell test , both observers alice and bob are on the same footing , and test the strength of the shared correlations . specifically
, alice chooses a measurement @xmath3 ( bob chooses @xmath52 ) and gets outcome @xmath4 ( bob gets @xmath53 ) .
the correlation is thus described by a joint probability distribution @xmath54 .
the latter can be reproduced by a pre - determined classical strategy if it admits a decomposition of the form @xmath55 where @xmath17 represents the shared local ( hidden ) variable , and @xmath56 .
any distribution that does not admit a decomposition of the above form is said to be bell nonlocal .
the set of local distributions , _
i.e._of the form is convex , and can thus be characterized by a set of linear inequalities called bell inequalities @xcite .
hence violation of a bell inequality implies bell nonlocality . in quantum theory , bell
nonlocal distributions can be obtained by performing suitably chosen local measurements , @xmath1 and @xmath57 , on an entangled state @xmath0 . in this case
, the resulting distribution @xmath58 does not admit a decomposition of the form .
bell nonlocality is however not a generic feature of entangled quantum states .
that is , there exist mixed entangled states which are local , in the sense that the statistics resulting from arbitrary non - sequential local measurements can be reproduced by a local model @xcite .
given the above , we investigate now how joint measurability relates to bell nonlocality .
first the above theorem implies that , if the set of povms @xmath59 used by alice is jointly measurable , then the statistics @xmath54 can always be reproduced by a local model , for any state @xmath0 and measurements of bob @xmath60 .
the converse problem is much more interesting .
the question is whether for any set of povms @xmath59 that is not jointly measurable , there exists a state @xmath0 and a set of measurements @xmath60 such that the resulting statistics @xmath54 violates a bell inequality .
this was shown to hold true for the case of sets of two povms with binary outcomes @xcite . in this case ,
joint measurability is equivalent to violation of the chsh bell inequality .
here we give evidence that this connection does not hold in general .
specifically , we exhibit a set of povms which is not jointly measurable but nevertheless can not violate a large class of bell inequalities .
consider the set of three dichotomic povms ( acting on @xmath61 ) given by the following positive operators @xmath62 for @xmath63 , where @xmath64 are the pauli matrices , and @xmath65 .
indeed , one has that @xmath66 .
this set of povms should be understood as noisy pauli measurements .
the set is jointly measurable if and only if @xmath67 , although any pair of povms is jointly measurable for @xmath68 @xcite ( see also @xcite ) .
hence in the range @xmath69 , the set @xmath70 forms a _ hollow triangle _ : it is pairwise jointly measurable but not fully jointly measurable .
we now investigate whether the above hollow triangle can lead to bell inequality violation .
the most general class of bell inequalities to be considered here are of the form : @xmath71 where @xmath72 all ( tight ) bell inequalities of the above form for @xmath73 are known ( see appendix ) . using a numerical method based on semi - definite - programming ( sdp ) @xcite ( see appendix )
we could find the smallest value of the parameter @xmath74 for which a given inequality can be violated using the set of povms .
the results are summarized in table i. notably , we could not find a violation in the range @xmath69 where the set @xmath70 is a hollow triangle .
in fact , no violation was found for @xmath75 , whereas pairwise joint measurability is achieved for @xmath76 , thus leaving a large gap .
note also that pairwise joint measurability implies violation of the chsh inequality here , since we have povms with binary outcomes @xcite .
we thus conjecture that there is a threshold value @xmath77 , such that all hollow triangles with @xmath78 do not violate any bell inequality .
[ table ] .
bell inequality violation with incompatible povms .
specifically , we consider the sets given in equations and .
for each set , we determine the smallest value of the parameter @xmath74 , such that the set becomes jointly measurable ( jm ) , and achieve bell inequality violation .
we consider tight bell inequalities with up to @xmath79 measurements for bob ( see appendix ) .
note that pairwise joint measurability is equivalent to violation of the chsh bell inequality . [ cols="^,^,^",options="header " , ] |
classical be stars are non - supergiant , b - type stars that are surrounded by hot gaseous disks .
this circumstellar gas is responsible for many observational characteristics such as hydrogen balmer emission lines , ir flux excess , and short- and long - term flux variability @xcite .
optical and infrared interferometry has become an important tool in characterizing be stars and their disks @xcite .
the first interferometric survey of be stars was made by @xcite to resolve the h@xmath2 emission in seven be stars .
their survey showed that the emitting regions are flattened , which is strong observational evidence of a disk - like geometry . @xcite combined optical interferometry and spectropolarimetry to derive the disk position angle on the sky , and they found good agreement between these techniques .
@xcite used the navy precision optical interferometer ( npoi ) to observe the h@xmath2 emission from the disks of seven be stars .
their observations showed that a direct correlation exists between the disk sizes and the net h@xmath2 luminosities .
infrared observations have begun to reveal the spatial properties of the continuum and line emission of be star disks .
@xcite made the first chara array long - baseline interferometric observations in the @xmath0-band of four bright be stars , @xmath3 cas , @xmath4 per , @xmath5 tau , and @xmath6 dra , and they were able to resolve the disks and to constrain their geometrical and physical properties .
@xcite studied the geometry and kinematics of the be star @xmath6 cma in the br@xmath3 emission line and in the nearby continuum using the vlti / amber instrument .
@xcite observed the be binary system @xmath7 sco using spectrally - resolved interferometry with the vlti / amber and chara / vega instruments .
their observations show that the disk varies in size from 4.8 mas in h@xmath2 , to 2.9 mas in br@xmath3 , and to 2.4 mas in the @xmath0-band continuum .
@xcite completed a survey of eight be stars with vlti / amber and measured the disk extensions in the br@xmath3 line and the nearby continuum .
their study suggests that the disk kinematics are dominated by keplerian rotation and that the central stars have a mean ratio of angular rotational to critical velocity of @xmath8 .
in addition , @xcite used the vlti / midi instrument to determine the @xmath9-band ( 10 @xmath10 m ) disk angular size for seven be stars .
interferometry offers us the means to explore be star disks in large numbers and to begin to understand their properties as a whole .
here we present results from such a survey that we conducted in the @xmath0-band continuum using the chara array long - baseline interferometer . in section 2
, we list our sample stars , present our observational data sets , and describe the data reduction process . in section 3 , we describe a method that we implemented to correct the interferometric measurements for the flux of stellar companions .
we discuss in section 4 the spectral energy distributions and their use in estimating the stellar angular diameter and infrared excesses of be stars .
in section 5 , we present fits of the interferometric visibilities using simple geometrical models , and in section 6 , we discuss the results with a particular comparison of the @xmath0-band and h@xmath2 disk sizes .
finally , we summarize our results and draw our conclusions in section 7 .
we selected 24 be stars as targets for this interferometric survey .
the main selection criteria were that the stars are nearby and bright , well within the limiting magnitude of the chara classic tip - tilt servo system ( @xmath11 11 ) and the near - ir fringe detector ( @xmath12 8.5 ) .
the selected be stars had to have declinations north of about @xmath1315@xmath14 to be accessible with the interferometer at low air - mass values .
furthermore , most of the targets have recently shown hydrogen emission and a near - ir flux excess .
we relied particularly on spectrophotometric and h@xmath2 observations conducted by @xcite , @xcite , @xcite , and @xcite .
the targets and their adopted stellar parameters are presented in table [ param ] .
columns 1 and 2 list the star names , columns 3 to 5 list the spectral classification from the compilation by @xcite and the stellar effective temperature @xmath15 and gravity @xmath16 from @xcite ( see their table 9 `` apparent parameters '' ) .
the stars hd 166014 and hd 202904 are not listed by @xcite , so we used the parameters for these two from @xcite .
columns 6 and 7 list predictions for the position angle @xmath17 of the projected major axis of the disk that should be @xmath18 different from the intrinsic polarization angle @xcite and for @xmath19 , the ratio of the minor to major axis sizes according to the estimated stellar inclination from @xcite . measuring the instrumental transfer function of the chara array interferometer is performed by observing calibrator stars with known angular sizes before and after each target observation .
the calibrator stars are selected to be close to the targets in the sky , unresolved with the interferometer s largest baseline , and without known spectroscopic or visual binary companions .
we collected photometric data on each calibrator star in order to construct their spectral energy distribution ( sed ) and to determine their angular diameter .
the collected @xmath20 photometry ( available from @xcite ) is transformed into calibrated flux measurements using procedures described by @xcite and @xcite .
the stellar effective temperature @xmath15 and the gravity @xmath16 ( generally from the compilation of @xcite ) are used to produce a model flux distribution that is based on kurucz stellar atmosphere models .
note that we generally used johnson @xmath21 magnitudes compiled by @xcite and @xmath22 magnitudes from @xcite , who list tycho @xmath23 and @xmath24 magnitudes that are slightly different from johnson @xmath23 , @xmath24 magnitudes .
the photographic @xmath25 and @xmath26 magnitudes were collected from @xcite , which are only slightly different from johnson @xmath25 , @xmath26 magnitudes .
the near - ir @xmath27 photometry was collected from the 2mass survey @xcite .
the use of non - standard photometry introduces errors in the best - fit , limb - darkened angular diameter of the calibrators that are comparable to or smaller than the estimated uncertainties given in table 2 .
we also collected measurements of @xmath28 and applied a reddening correction to the model sed before fitting the sed of the calibrators .
we then computed a limb - darkened angular diameter @xmath29 by direct comparison of the observed and model flux distributions .
based upon the limb - darkening coefficients given by @xcite , we transformed the limb - darkened angular diameter to an equivalent uniform - disk angular diameter @xmath30 assuming a projected baseline of 300 m. columns 1 and 2 of table [ calib ] list the calibrator star and its corresponding target , respectively , columns 3 and 4 list the calibrator effective temperature @xmath15 and reference source , columns 5 and 6 give the surface gravity @xmath16 and reference source , column 7 gives the spectral classification , and columns 8 and 9 list the adopted interstellar reddening @xmath28 and the reference source , respectively .
column 10 of table [ calib ] lists the best - fit limb - darkened angular diameter @xmath29 derived from fitting the calibrator sed , and column 11 lists the corresponding uniform - disk angular diameter @xmath30 .
the observations were conducted between 2007 october and 2010 november using the chara classic beam combiner operating in the @xmath0-band ( effective wavelength = 2.1329 @xmath10 m ; @xcite ) .
we need a minimum of two interferometer baselines at substantially different projected angles on the sky to map the circumstellar disks around our targets , and we generally used the south - west baseline of length @xmath31 278 m oriented at 39@xmath14 west of north ( s1/w1 ) and the south - east baseline of length @xmath31 330 m oriented at 22@xmath14 east of north ( s1/e1 ) .
each target and its calibrator were observed throughout a given night in series of 200 scans recorded with a near - ir detector on a single pixel at a frequency that ranged between 500 - 750 hz , depending on the seeing conditions of each particular night of observation .
the interferometric raw visibilities were then estimated by performing an integration of the fringe power spectrum .
we used the chara data reduction software ( reduceir ; @xcite ) to extract and calibrate the target and the calibrator interferometric visibilities .
then the raw visibilities were calibrated by comparing them to the time - interpolated calibrator visibilities and rescaling them according to the predicted calibrator visibility for the given projected baseline and stellar angular diameter .
the resulting calibrated visibilities are listed in table [ calibvis ] ( given in full in the electronic version ) .
column 1 of table [ calibvis ] lists target hd number , column 2 lists the heliocentric julian date of mid - observation , column 3 lists the telescope pair used in each observation , columns 4 and 5 list the @xmath32 and @xmath33 frequencies , respectively , column 6 lists the projected baseline , column 7 lists the effective baseline ( see 5.2 ) , columns 8 and 9 list the calibrated visibility and its corresponding uncertainty , respectively , and lastly , columns 10 and 11 list the visibility measurements and uncertainties corrected for the flux of stellar companions for those cases with known binary parameters .
we will discuss this correction in detail in the next section .
the internal uncertainties from fitting individual fringes are generally smaller than @xmath34 .
the scatter in the data depends mostly on the target magnitude and seeing conditions at the time of the observations , which usually varies with a fried parameter in the range @xmath35 cm .
the visibility uncertainties for the brightest targets range between 2@xmath36 and 5@xmath36 , while those for the faintest ones reach up to 8@xmath36 .
overall we obtained a relatively good set of observations at different hour angles for each star in our sample , with the exceptions of hd 58715 and hd 148184 where the position angle coverage was limited to only one projected baseline .
figure set 1 shows the distribution of the observations in the @xmath37 plane for our sample .
our measurements of the sizes and orientations of be star disks are based upon the observed decline in visibility caused by the extended angular distribution of disk flux .
however , a drop in visibility can also occur if a stellar companion is within the effective field of view of the interferometer , and if the binary component is ignored , then the disk dimensions will be overestimated ( in the extreme case apparently implying the presence of a large disk where none is present ) .
binary systems are relatively common among b - stars in general and be stars in particular @xcite , so it is important to investigate what role they play in the interpretation of our interferometric measurements . fortunately , the signatures of binary companions in interferometric observations are well understood @xcite , and given the projected separation and magnitude difference we can determine how the companion will affect our measurement . according to the van cittert - zernicke theorem ,
the complex visibility is related to the fourier transform of the angular spatial distribution in the sky , and the measured fringe amplitude is proportional to the real part of the visibility @xcite . in the ideal case of a binary star consisting of two point sources ,
the visibility varies according to a cosinusoidal term with a frequency that depends on the projected separation and the baseline and wavelength of observation .
the interferometric fringe observed will display an amplitude that depends on the real part of the visibility according to the projected separation and binary flux ratio . for observations like ours that record the flux over a wide filter band , the fringe pattern
is only seen over a range in optical path delay that is related to the coherence length .
binaries with separations smaller than the coherence length will display the full amplitude variation expected from the visibility dependence on binary separation and flux ratio @xcite , while those with separations larger than the coherence length will appear as separated fringe packets @xcite .
thus , there are several separation ranges that are key to this discussion .
first , we may ignore those binary companions that are outside of the field of view of the interferometer ( @xmath38 arcsec ; 3.6 ) and separated by more than atmospheric seeing disk .
second , there are binary companions that are effectively within the field of view , but whose projected separations are large enough that the fringe packets of the components do not overlap ( @xmath39 mas for an observation with a 300 m baseline ) .
this is by far the most common case for our observations , and indeed the fringe packet of the companion is usually located far beyond the recorded scan . in this situation
the flux of the companion will act to dilute the measured visibility of the fringe but will not change its morphology ( 3.3 ) .
third , there are binaries that have such small projected separations ( @xmath40 mas for an observation with a 300 m baseline ) that their fringe packets overlap and create an oscillatory pattern in the observed visibility due to the interference between fringe packets ( 3.3 ) .
this probably occurred in only a few cases among our observations ( 3.7 ) . ideally , we should model the fringe visibility in terms of the binary projected separation and flux ratio together with a parametrization of the be disk properties ( 5.1 ) . however , because this is a survey program , our observational results are generally too sparse in coverage of baseline and position angle range to attempt such a general solution .
for example , most of the known companions have angular separations that are relatively large , and the fringe packet of a companion would have only been recorded in short baseline observations .
however , such short baseline data would not resolve the be disks , which is our primary scientific goal .
there are a few cases where the projected separations are small and the binary creates oscillations in visibility with projected baseline , but again our baseline coverage is generally too sparse to measure these together with the disk properties .
consequently , we made the decision to rely solely on published data on the binary companions of our targets ( 3.2 ) in order to perform the corrections to the measured visibility where it was necessary to do so ( i.e. , the known companion was sufficiently bright and close enough to influence our visibility measurements ) .
the visibility correction procedure we adopted is outlined in the following subsections .
the method is built upon the scheme described by @xcite and considers how a binary companion influences the appearance of the combined fringe packets .
we use estimates of the flux ratio ( 3.5 , 3.6 ) and the orbital projected separation ( 3.2 ) for each observation to build a numerical relation between the be star visibility and net observed visibility .
then we interpolate within these relations to determine the be star visibility alone ( 3.3 ) .
we also extend this approach to correct the visibilities for two multiple star systems ( 3.4 ) .
many be stars in our sample are known binaries or multiple systems .
we checked for evidence of the presence of companions through a literature search with frequent consultation of the washington double star catalog ( @xcite ) , the fourth catalog of interferometric measurements of binary stars ( @xcite ) , and the third photometric magnitude difference catalog .
we only considered those companions close enough to influence the interferometry results ( i.e. , those with separations less than a few arcsec ) .
we show the binary search results in table [ binary ] , which includes visual binaries ( typical periods @xmath41 y ) and spectroscopic binaries ( typical periods @xmath42 y )
. columns of table [ binary ] list the star name , number of components , reference code for speckle interferometric observations , and then in each row , the component designation , orbital period , angular semimajor axis , the estimated @xmath0-band magnitude difference between the components @xmath43 , a y or n for whether or not a correction for the flux of companions was applied to the data , and a reference code for investigations on each system .
entries appended with a semi - colon in table [ binary ] indicate parameter values with large uncertainties . these include the single - lined spectroscopic binaries , where we simply assumed a primary mass from the spectral classification and @xmath44 for the companion to derive the semimajor axis @xmath45 , which was transformed to an angular semimajor axis using the distance from _ hipparcos _ @xcite .
we find that ten of the 24 be stars in our sample have no known companion , and five others have companions that are too faint ( all single - lined spectroscopic binaries ) to influence the interferometric measurements .
thus , no corrected visibilities are listed for these 15 targets in table 3 .
however , the companions were bright enough ( @xmath46 ) and close enough ( separation @xmath47 ) for the remaining nine targets that we had to implement the corrections method outlined below .
we need to determine the companion s separation and position angle at the time of each observation in order to find the angular separation projected along the baseline we used .
this was done by calculating the binary relative separation for the time of observation from the astrometric orbital parameters using the method outlined by @xcite .
we show in table [ orbital ] the adopted orbital parameters for those binaries where we made visibility corrections .
note that no entry is present for hd 166014 , where only one published measurement exists ( see appendix ) , and we simply assumed that the projected separation was larger than the recorded scan length .
there are four cases where we present new orbital elements that are based upon published astrometric measurements , and we caution that these are preliminary and used only to estimate the separations and position angles at the times of the chara array observations .
details about these preliminary fits are given in the appendix .
the changes in visibility caused by a binary companion can be described equivalently in terms of the real part of the complex visibility or the amplitude of interfering fringe packets @xcite . here
we develop a fringe packet approach to the problem that simulates the binary changes and that is directly applicable to the fringe amplitudes that we measure .
we begin by considering how the fringe patterns of binaries overlap in order to assess the changes in visibility caused by a binary consisting of two unresolved stars , and then we extend the analysis to the situation where one star ( be plus disk ) is partially resolved . the fringe packet for star @xmath48 observed in an interferometric scan of changing optical path length
has the form @xmath49 where @xmath50 , @xmath45 is the scan position relative to the center of the fringe , @xmath51 is the coherence length given by @xmath52 ( equal to 13 @xmath10 m for the chara classic @xmath53 filter ) , and @xmath4 is a phase shift introduced by atmospheric fluctuations @xcite . if two stars are present in the field of view ( 3.6 ) with a projected separation along the scan vector of @xmath54 , then their fringe patterns may overlap and change the composite appearance of the fringe according to @xcite @xmath55 where @xmath56 and @xmath57 are the monochromatic fluxes of the stars
. then the fringe visibility is calculated by @xmath58 in this discussion we will ignore the very small decline in fringe amplitude related to the angular diameters of the stars themselves , because their angular diameters are all very small ( see table 6 below ) .
we show a series of such combined fringe patterns in the panels of figure [ app1 ] for an assumed flux ratio of @xmath59 . in the top panel ,
the projected separation is zero , and the two patterns add to make the fringe pattern of a single unresolved star . the associated visibility that we would measure equals one in this case .
however , in the second panel from the top , we show how a projected separation of 1 @xmath10 m results in a much lower visibility because the peaks associated with star 1 are largely eliminated by the troughs associated with star 2 . in the third panel from the top ,
the separation is just large enough ( comparable to the coherence length ) that the fringe pattern of the companion emerges from the blend , and the lower panel shows a separated fringe packet in which both fringe patterns are clearly visible .
note that the relationship between the projected angular separation ( mas = milli - arcsecond ) and the separation of fringe centers ( @xmath10 m ) is given by @xmath60 where @xmath23 is the projected baseline in meters .
for instance , using a 330 m baseline , the longest scan length is 150 @xmath10 m , which corresponds to a separation of @xmath31 94 mas .
we show in figure [ app2 ] the net visibility that would be measured as a function of projected separation @xmath54 .
this shows that in general the observed visibility will be less than that of a single star .
as expected , for very close separations the visibility varies cosinusoidally with @xmath54 with a frequency of @xmath61 . on the other hand , for projected separations larger than the coherence length @xmath51
, the visibility approaches the value @xmath62 equal to the semiamplitude of the flux diluted fringe pattern of the target .
now suppose that star 1 is a be star with a disk that is partially resolved , so that if it were observed alone , it would show a visibility @xmath63 .
consequently , its fringe pattern would have an semiamplitude given by @xmath64 .
we show a selection of model binary fringe patterns in figure [ app3 ] again for @xmath59 and a specific separation of @xmath65 @xmath10 m .
the panels show from top to bottom the progressive appearance of the combined fringe patterns as @xmath66 drops from 1 to 0.25 .
now the visibility drops in tandem until @xmath67 where the maximum and minimum are set by the fringe pattern of the companion .
we show the relationship between the be star visibility @xmath66 and the net observed visibility @xmath68 in figure [ app4 ] ( solid line for @xmath65 @xmath10 m ) . at this separation
, there is some slight destructive interference between the fringe patterns that decreases the maximum amplitude for star 2 , and as @xmath66 declines to zero , the net visibility attains the amplitude of star 2 alone @xmath69 .
figure [ app4 ] also shows the @xmath70 relationship for two other separations .
the dotted line shows the case of zero separation for maximum constructive interference , and here the visibility declines linearly to @xmath69 as @xmath66 tends to zero .
finally , the dashed line shows the case for a very large separation in which the fringe pattern of star 2 falls beyond the recorded portion of the scan . here
the visibility starts at its diluted value of @xmath62 at @xmath71 and declines to near zero at @xmath72 .
thus , to correct the observed visibilities for the presence of a companion , we need a diagram like figure [ app4 ] for each observation of a target .
we calculated the projected separation of the stars at the time of the observation from the dot product of the relative position vector ( from the angular orbital elements given in table 5 ) and the @xmath37 spatial frequencies for the baseline used .
then we created an associated @xmath70 diagram for each observation based upon the projected separation and effective flux ratio .
the corrected visibility @xmath66 was then found by interpolating in the relationship at the observed @xmath68 value . in some rare circumstances
, we encountered a double - valued @xmath73 relation , so no correction was attempted because of this ambiguity .
there are six systems in our sample with two or more close companions .
both companions of hd 5394 ( @xmath3 cas ) are faint , so no correction was made . in the cases of hd 23862 ( pleione ) and
hd 200120 ( 59 cyg ) , the inner companion is very faint , so corrections were made for only the outer , brighter companion .
the spectroscopic pair that comprises the b component of hd 4180 ( @xmath74 cas ) has such a small angular semimajor axis that it was treated as a single object , and thus this system was also corrected as a binary star ( 3.3 ) .
this left two systems , hd 198183 ( the triple @xmath75 cyg ) and hd 217675 ( the quadruple @xmath74 and ) , that required corrections for the flux of additional components .
the very close b pair of @xmath74 and was regarded as a single object ( see appendix ) , so both @xmath75 cyg and @xmath74 and were treated as triple star systems .
we made visibility corrections for these two systems in much the same way as for the binaries , except in this case the fringe normalizations were assigned by @xmath76 again , we formed model visibilities from the coaddition of the fringe patterns , determined the @xmath70 relationships for the time and baseline configuration of each observation , and then used the inverted relation @xmath73 to determine the corrected visibility .
unfortunately , there are significant uncertainties surrounding both the magnitude differences and orbital elements for the companions of hd 198183 and hd 217675 , and these introduce corresponding uncertainties in the amounts of visibility correction .
our results on these two systems must therefore be regarded as representative visibility solutions rather than definitive ones .
however , the corrected interferometry visibilities are all close to one for these two be stars , which suggests that their disks are only marginally resolved if at all . on the other hand ,
the much lower uncorrected visibilities of these two show that the influence of the companions is clearly present .
both targets will be important subjects for future , multiple baseline observations with the chara array to determine their orbital properties .
our visibility correction procedure requires a knowledge of both the projected separation of the stars ( 3.2 ) and their monochromatic flux ratio .
unfortunately , the magnitude differences between the be star primary and the companion are generally available only in the @xmath24-band , and we need to estimate the magnitude differences in the @xmath0-band . we must consider the color difference between the components and how much brighter the be star plus disk appears in the @xmath0-band compared to the @xmath24-band . the predicted magnitude difference is given by @xmath77 where @xmath78 are the monochromatic @xmath0-band fluxes for the be companion , the be star , and the be disk , respectively .
we can estimate the first term from the color differences of the be star and companion using @xmath79 where we will assume that the disk contribution is negligible in the @xmath24-band so that @xmath80 . in the absence of other information , we estimated the color differences @xmath81 by assuming that both the be star and companion are main sequence objects , and we used the relationship between @xmath81 and magnitude difference from a primary star of effective temperature @xmath15(be ) @xcite for main sequence stars from @xcite to find @xmath81 for both stars .
we determined the infrared flux excess term @xmath82 from our observed estimate of @xmath83 @xcite , which are related by @xmath84^k } \over { [ 1 + f_d / f_1 + f_2/f_1]^v}}\ ] ] where the superscripts indicate the filter band .
if we assume that the disk contributes no flux in the @xmath24-band , then we can rearrange this equation to find the @xmath0-band flux excess relative to that of the be star alone .
then we can combine the results from the two equations above to predict the observed the @xmath0-band magnitude difference @xmath85 our estimates for @xmath86 are listed in tables 4 and 5 .
note that there are several instances where we give a range for the magnitude difference ; these are be stars that are single - lined spectroscopic binaries with a companion of an unknown type .
we consider two hypothetical cases .
first , we assume that the companion is a main sequence star of one solar mass , and we use the @xcite main sequence relation to obtain the magnitude and color differences of the companion .
the second case is to assume that the companion is a hot subdwarf ( similar to the case of @xmath4 per ; see @xcite ) with a typical effective temperature of 30 kk and a stellar radius of @xmath87 .
we then estimate @xmath86 by adopting the main sequence radius for the be star according to its effective temperature and by using the planck function for both stars in order to estimate the monochromatic @xmath0-band flux ratio .
table [ binary ] lists those cases with a hyphen in the @xmath43 column giving the magnitude difference range between that for a hot subdwarf ( smaller ) and for a solar - type companion ( larger ) .
however , we made no visibility corrections in most of these cases because the nature of the companion is so uncertain ( with the exception of @xmath4 per where the companion s spectrum was detected and characterized by gies et al .
1998 ) .
the chara classic observations were recorded on a single pixel of the near infrared observer camera ( niro ) , and the physical size of the pixel corresponds to a square of dimensions @xmath38 arcsec on the sky .
the flux of companions with separations small compared to 0.8 arcsec will be more or less completely included in the observations , but the flux of companions with larger separations from the central be star may be only partially recorded according to the separation and seeing conditions at the time of observation .
therefore , we need to calculate the effective flux ratio of companion to target based upon the relative amounts of flux as recorded by this one pixel . seeing is usually computed in real time according to the chara tip - tilt measurements of the @xmath24-band flux of the targets . by assuming that the @xmath0-band seeing varies with wavelength by @xmath88 @xcite , the @xmath0-band seeing disk is about @xmath310.76 times that in @xmath24 .
the effective flux ratio is given by @xmath89 where @xmath90 and @xmath91 are the net intensity contributions of the primary and secondary component , respectively , recorded by the pixel .
we assume a gaussian profile for the point spread function as projected on the detector , @xmath92\ ] ] where @xmath93 are the coordinates of the central position of the star on the detector chip , and @xmath94 is related to the seeing ( @xmath95 ) .
the intensity distributions of the primary and the secondary components integrated over one pixel on the detector are given by @xmath96 @xmath97 where @xmath98 is the separation of the binary , and @xmath99 is the actual ratio of primary to secondary flux , which is derived from the magnitude difference of the two components in the @xmath0-band , @xmath100 we used the visibility correction scheme described in 3.3 and 3.4 with the predicted projected angular separation ( from the orbital elements in table 5 and the observed @xmath37 spatial frequencies in table 3 ) and the effective flux ratio ( 3.5 and 3.6 ) to derive estimates of the be star visibility alone . the corrected visibilities and their corresponding uncertainties are listed in the last two columns of table [ calibvis ] for the nine cases with significant companions .
these uncertainties do not include the contributions to the error budget from uncertainties in projected separation and flux ratio . in most cases
the projected separations are much larger than the coherence length ( typically 9 to 27 mas for 300 m to 100 m baselines ) , which corresponds to the separated fringe packet case ( see lower panel of fig . 2 ) .
thus , the correction depends mainly on the flux dilution term , @xmath101 , and typical uncertainties of 0.2 mag for @xmath86 only amount to correction uncertainties of @xmath102 in visibility , i.e. , generally smaller than the measurement errors . on the other hand , binaries with projected separations less than the coherence length ( mainly observations of @xmath74 cas and @xmath4 per ; see table 5 ) have corrections that reflect the fringe amplitude oscillation caused by beating between the fringe patterns of the components ( fig .
these corrections may amount to @xmath103 in visibility , @xmath104 for the relatively faint companions considered here .
figure 3 shows that the full range of the correction varies over a projected separation difference of @xmath105 , which corresponds to a projected angular separation difference of approximately 0.7 to 2.2 mas for 300 m and 100 m baselines , respectively .
the predictive accuracy of the orbital separation probably has comparable uncertainties for the cases of @xmath74 cas and @xmath4 per , so it is possible that the uncertainties in the corrections may be similar to the corrections themselves in these close separation instances .
nevertheless , we found that implementing the corrections even in these close cases tended to reduce the scatter in the fit of the be disk visibilities ( 5.1 ) , so we will adopt the corrected visibilities in the subsequent analysis of all nine targets with significant companions .
the spectral energy distributions ( seds ) of our sample stars can help us to estimate the photospheric angular diameter of the be star and the ir flux excess ( from a comparison of the observed and extrapolated stellar ir fluxes ) .
the observed ir flux excess can thus be directly compared with the disk flux fraction derived from fits of the visibility measurements .
one difficulty with this approach is that the ir flux excess is usually derived assuming that the disk contributes no flux in the optical range , so the stellar sed can be normalized to the optical flux .
however , this may lead to an underestimate of the ir excess if the disk flux emission is significant in the optical , which may be the case for stars with dense and large circumstellar disks .
the disk flux fraction declines at lower wavelength because of the drop in free - free opacity , and the disk contribution is negligible in the ultraviolet ( uv ) part of the spectrum .
therefore , we decided to make fits of the uv flux to set the photospheric flux normalization , which will provide a more reliable estimate of the ir flux excess .
we collected uv spectra for all our targets from the archive of the _ international ultraviolet explorer ( iue ) satellite _ , maintained at the nasa mikulski archive for space telescopes at stsci . in most cases , we used the available low dispersion , swp ( 1150 - 1900 ) and lwp / lwr ( 1800 - 3300 ) spectra , and in cases where there were fewer than two each of these , we used high - dispersion spectra for these two spectral ranges . in the case of hd 217891 ,
all but one of the spectra were made with the small aperture , so we formed an average spectrum and rescaled the flux to that measured by the td-1 satellite ( @xcite ) . in the case of hd 203467
, there were no long - wavelength spectra available in the _ iue _ archive , so we used a combination of _ iue _ swp spectra ( 1150 1900 ) , skylab objective - prism spectrophotometry ( 1900 2300 ; @xcite ) , and oao-2 spectral scans ( 2300 3200 ; @xcite ) .
fluxes from each spectrum were averaged into 10 bins from 1155 to 3195 , and then the fluxes from all the available spectra for a given target were averaged at each point in this wavelength grid .
we created model spectra to compare with the _ iue _ observations by interpolating in the grid of model lte spectra obtained from r. kurucz .
these models were calculated for solar abundances and a microturbulent velocity of 2 km s@xmath106 .
the interpolation was made in effective temperature and gravity using estimates for these parameters ( see table 1 ) from the be star compilation of @xcite .
for those sample stars in binaries , we formed a composite model spectrum by adding a model spectrum for each companion that was scaled according to the @xmath0-band magnitude difference listed in table [ orbital ] .
we then made a non - linear , least - squares fit of the observed uv spectrum with a model spectrum transformed according to the extinction curve of @xcite and normalized by the stellar , limb - darkened angular diameter @xmath29 , and we assumed a standard ratio of total to selective extinction of @xmath107 for the interstellar extinction curve .
finally , we considered an extension of the fitted photospheric sed into the @xmath0-band , and we determined an ir flux excess from @xmath108 where the monochromatic @xmath0-band fluxes are @xmath109 for the be star component , @xmath110 for the sum of the photospheric fluxes of the be and all companions ( if any ) , and @xmath111 is the observed flux from 2mass @xcite .
our results are listed in table [ sedinfo ] that gives the hd number of the star , the derived reddening @xmath28 and its uncertainty , the limb darkened stellar angular diameter @xmath112 and its uncertainty , and the infrared excess from the disk @xmath113 and its uncertainty .
the final columns list the photospheric fraction of flux @xmath114 and its uncertainty that are related to the ir flux excess by @xmath115 our derived interstellar reddening values are generally smaller than those derived from the optical colors because the disk flux contribution increases with increasing wavelength through the optical band which mimics interstellar reddening @xcite .
furthermore , our derived angular diameters may be smaller in some cases from previous estimates because of the neglect of the flux of the companions in earlier work .
note that we have neglected the effects of rotation ( oblateness and gravity darkening ) on the sed , and these may influence the flux normalization @xcite .
here we show how we can use the interferometric visibility measurements to estimate the stellar and disk flux contributions and to determine the spatial properties of the disk emission component .
we used a two - component geometrical disk model to fit the chara classic observations and to measure the characteristic sizes of the circumstellar disks .
the model consists of a small uniform disk representing the central be star and an elliptical gaussian component representing the circumstellar disk @xcite . because the fourier transform function is additive , the total visibility of the system is the sum of the visibility function of the central star and the disk , @xmath116 where @xmath117 , @xmath118 , and @xmath119 are the total , stellar , and disk visibilities , respectively , and @xmath114 is the ratio of the photospheric flux contribution to the total flux of the system . the visibility for a uniform disk star of angular diameter @xmath120 is @xmath121 where @xmath122 is the first - order bessel function of the first kind and @xmath123 is the spatial frequency of the interferometric observation , @xmath124 .
the central be star is usually mostly unresolved even at the longest baseline of the interferometer , which brings its visibility close to unity , @xmath125 .
the disk visibility is given by a gaussian elliptical distribution @xmath126\ ] ] where @xmath127 is the full - width at half - maximum ( fhwm ) of the spatial gaussian distribution along the major axis and @xmath128 is given by @xmath129 where @xmath19 is the axial ratio and @xmath17 is the position angle of the disk major axis .
thus , the gaussian elliptical model has four free parameters : the photospheric contribution @xmath114 , the axial ratio @xmath19 , the position angle @xmath17 , and the disk angular size @xmath127 .
the brightness distribution of the disk projected onto the sky is a function of the inclination and position angle of the major axis .
interferometric observations with a given baseline @xmath130 will sample the disk elliptical distribution according to the angle between the projection of the baseline on the sky and the position angle of the disk major axis .
it is helpful to consider rescaling the baseline to account for the changes in the disk size with direction . for a flat disk inclined by an angle @xmath48 and oriented with the major axis at a position angle @xmath17 ( measured from north to east )
, we can rescale using the effective baseline @xmath131 @xcite @xmath132 where @xmath133 is the baseline position angle at the time of the observations .
this new quantity , the effective baseline , takes into consideration the decrease in the interferometric resolution due to the inclination of the disk in the sky , and thus for the purposes of analysis , it transforms the projected brightness distribution of the disk into a nearly circularly symmetric brightness distribution .
thus , the disk part of the visibility can be considered as a function of @xmath131 alone , and below we will use this parameter to present the interferometric results . however
, if there is also a stellar flux contribution , than its projection and visibility will be a function of the projected baseline @xmath130 only ( if the star appears spherical in the sky ) .
consequently , models with both stellar and disk contributions should be presented for both the major and minor axis directions in order to show the range in visibility with baseline direction in a single plot . along the minor axis , for example , @xmath134 , so that the relation becomes @xmath135 , and consequently for edge - on systems with @xmath136 a given @xmath131 may correspond to a @xmath130 far larger than available with the array . such observations ( if possible )
would start to resolve the stellar disk and lead to low net visibility ( see the dotted line in fig .
8.3 ) .
the simple geometrical representation of the be star system may in some cases present a solution family or degeneracy that exists between two fundamental parameters of the gaussian elliptical model : the gaussian full width at half maximum @xmath127 and the stellar photospheric contribution @xmath114 . in order to explain the ambiguity in the model
, we consider the case of a locus of ( @xmath114 , @xmath127 ) that produces the same visibility measurement at a particular baseline .
we illustrate this relation in figure [ test1 ] , which shows an example of a series of ( @xmath114 , @xmath127 ) gaussian elliptical visibility curves that all produce a visibility point @xmath24 = 0.8 at a projected baseline of 200 m ( for @xmath75 = 2.1329 @xmath10 m and a be star angular diameter @xmath137 = 0.3 mas ) .
the plot shows models for ( @xmath114 , @xmath127 ) = ( 0.156 , 0.6 mas ) ( solid line ) , ( @xmath114 , @xmath127 ) = ( 0.719 , 1.2 mas ) ( dotted line ) , and ( @xmath114 , @xmath127 ) = ( 0.816 , 2.4 mas ) ( dashed line ) . figure [ test2 ] shows the relationship between the gaussian elliptical full width at half maximum and the stellar photospheric contribution for the family of curves that go through the observed point @xmath24 = 0.8 at a 200 m baseline .
larger circumstellar disks are associated with larger stellar flux contributions , and vice - versa , which demonstrates that a single interferometric measurement does not discriminate between a bright small disk and a large faint one .
additional measurements at different baselines are necessary to resolve this ambiguity . note that if the interferometric observations are all located on one projected baseline in the ( @xmath138 ) plane , then the disk properties are defined in only one dimension .
thus , in such circumstances it is not possible to estimate the axial ratio @xmath19 or the position angle of the disk major axis @xmath17 .
only a lower limit for @xmath127 can be set in such cases .
the circumstellar disk is modeled with a gaussian elliptical flux distribution centered on the be star .
this disk model has four independent parameters ( @xmath19 , @xmath17 , @xmath114 , @xmath127 ) , and one assumed parameter , the stellar diameter @xmath120 ( table 6 ) .
the fitting procedure consists of solving for the model parameters using the idl non - linear least squares curve fitting routine mpfit ( @xcite ) , which provides a robust way to perform multi - parameter surface fitting .
model parameters can be fixed or free depending on the ( @xmath1 ) distribution of the observations .
for the cases with limited coverage in the ( @xmath1 ) plane , setting some model parameters to fixed values was necessary in order to fit successfully the data .
we adopted additional constraints on some model parameters that are well defined from studies such as inclination estimates from @xcite and intrinsic polarization angles from @xcite and @xcite . also , we occasionally used values of the ir flux excess derived from the seds of be stars to estimate the stellar photospheric contribution @xmath114 when needed .
our fitting results are summarized in table [ visfit ] , where the cases with a fixed parameter are identified by a zero value assigned to its corresponding uncertainty .
column 1 of table [ visfit ] lists the hd number of the star , columns 2 and 3 list the axial ratio @xmath19 and its uncertainty , columns 4 and 5 list the disk position angle along the major axis @xmath17 and its uncertainty , columns 6 and 7 list the photospheric contribution @xmath114 and its uncertainty , columns 8 and 9 list the angular fwhm of the disk major axis @xmath127 and its uncertainty , column 10 lists the reduced @xmath139 of the fit , columns 11 and 12 list the corrected photospheric contribution @xmath114(corr ) and its uncertainty ( see 5.5 ) , respectively , columns 13 and 14 list the disk - to - star radius ratio @xmath140 and its uncertainty , and finally , column 15 indicates the cases of fully resolved disks ( y ) , marginally resolved disks ( m ) , and unresolved disks ( n ) in our be star sample .
plots of the best - fit solutions showing the visibility curves of the system disk - plus - star as a function of the effective baseline in meters along with our interferometric data are presented in figure set 8 .
the solid lines in figure set 8 represent the best - fit visibility model of the disk along the major axis , the dotted lines represent the best - fit visibility model of the disk along the minor axis , and the star signs represent the interferometric data .
we find that the circumstellar disks of the four be stars , hd 23630 , hd 138749 , hd 198183 , and hd 217675 were unresolved at the time of our chara array observations , while the circumstellar disks of hd 23862 , hd 142926 , hd 164284 , hd 166014 , hd 200120 , and hd 212076 were only marginally resolved .
those are the cases where we had to fix @xmath19 , @xmath17 , and/or @xmath114 in order to make gaussian elliptical model fits to the data . on the other hand
, we successfully resolved the circumstellar disks around the other 14 be stars and were able to perform four - parameter gaussian elliptical fits for most of them as listed in table [ visfit ] .
modeling a circumstellar disk with an elliptical gaussian intensity distribution is convenient but not completely realistic .
the flux distribution in the model assumes that light components from the circumstellar disk and the central star are summed and that no mutual obscuration occurs .
it is important to note that in the case of small disks most of the model disk flux is spatially coincident with the photosphere of the star , so the assignment of the flux components becomes biased .
to illustrate this effect , we show in figure [ gaucorr ] the model components for a case with a faint disk .
the dotted line in figure [ gaucorr ] shows the assumed form of the intensity of the uniform disk of the star ( angular diameter of 0.68 mas ) , the dashed line shows the gaussian distribution of the circumstellar disk along the projected major axis ( fwhm = 0.55 mas ) , and the solid line shows the sum of the two intensity components . in the case where the fwhm of the circumstellar emission is similar to or smaller than the stellar diameter , most of the disk flux occurs over the stellar photosphere where the sum produces a distribution similar to that of a limb - darkened star .
the interpretation of the results obtained from the gaussian elliptical fits must be regarded with caution in situations where the derived disk radius is smaller than the star s radius and a significant fraction of the disk flux is spatially coincident with that of the star .
consequently , we decided to correct the gaussian elliptical fitting results in two ways .
first , the disk radius was set based upon the relative intensity decline from the stellar radius , and we adopted the disk radius to be that distance where the gaussian light distribution along the major axis has declined to half its value at the stellar equator . the resulting ratio of disk radius to star radius is then given by @xmath141 where @xmath127 is the gaussian full - width at half maximum along the major axis derived from the fits , and @xmath120 is the angular diameter of the central star .
secondly , the model intensity over the photosphere of the star from both the stellar and disk components was assigned to the flux from the star in an optically thin approximation .
the fraction of the disk flux that falls on top of the star is @xmath142 , where @xmath143 is found by integrating over the stellar disk the gaussian spatial distribution given by @xcite @xmath144\ ] ] where @xmath19 is the axial ratio and @xmath145 are the sky coordinates in the direction of the minor and major axes .
consequently , this fraction of the model disk flux should be reassigned to the star and removed from the disk contribution
. then the revised ratio of total disk to stellar flux is @xmath146 which is lower than the simple estimate of @xmath147 .
we list in columns 11 to 14 of table [ visfit ] the revised values of the stellar flux contribution @xmath114 and the be disk radius @xmath148 ( along with the corresponding uncertainties ) obtained by applying this correction using the stellar angular diameters from table 6 .
it is important to validate our results against other published data wherever possible .
however , since this is the first large scale study of the @xmath0-band emission of northern be stars , it is difficult to make direct comparisons .
we list in table 8 the available measurements of gaussian elliptical model parameters of be star disks for the stars in our sample ( excluding the work of @xcite that is included in our analysis ) .
there is only one other @xmath0-band measurement available by @xcite , but even in this case a direct comparison of @xmath127 is difficult because of the different assumptions made about the remaining parameters .
measurements of the disk emission in other bands will likely yield different diameter estimates ( 6.2 ) .
nevertheless , we can compare our results on the geometry of the disks with previous work , and we can consider the parameter estimates resulting from other kinds of observations .
we begin by comparing the stellar and disk flux contributions that we derived from the gaussian elliptical fits ( table 7 ) with those estimated from the ir - excess in the spectral energy distributions ( table 6 ) . among the 14 stars in our sample with reliable detections of disk emission , we fit for the stellar flux contribution in six cases .
the corrected parameter @xmath149 is plotted together with the sed estimate of this ratio in figure 10 .
the uncertainties are significant , but these two estimates of stellar flux in the @xmath0-band appear to be consistent ( perhaps surprisingly so , because of the known temporal flux variations and large time span between the 2mass photometry and our interferometric observations ) .
next we consider the disk axial ratio @xmath19 that is related approximately to the disk normal inclination @xmath48 by @xcite @xmath150 where the second term accounts for the increase in the minor axis size caused by the increase in disk thickness with radius .
we expect that the disk inclination will be the same as the stellar spin inclination because the disk is probably fed by equatorial mass loss .
@xcite estimated the stellar inclinations for most of our targets by comparing the projected rotational velocity @xmath151 with the critical rotational velocity @xmath152 , for which the equatorial centripetal and gravitational accelerations balance , and by assuming that be stars as a class share a common ratio of angular rotational to critical velocity .
the axial ratios derived from their estimates of spin inclination are given in the final column of table 1 .
there are seven targets with fitted estimates of @xmath19 among our 14 reliable detections , and we plot these values together with @xmath153 from the inclinations from @xcite in figure 11 ( indicated by square symbols ) .
figure 11 includes other @xmath19 estimates from previous interferometry ( table 8) . with two exceptions ( @xmath154 cyg = hd 202904 above and @xmath3 cas = hd 5394 below ) ,
the estimates from interferometry and rotational line broadening are in broad agreement .
three of our targets have prior interferometric estimates of @xmath19 , and our results agree within the uncertainties .
note that the increase in disk thickness with radius implies that @xmath19 will appear larger in bands where the disk emission extends to larger radius ( h@xmath2 ) , and in two cases ( @xmath155 per = hd 22192 and @xmath5 tau = hd 37202 ) @xmath19(h@xmath2 ) is larger than @xmath19(@xmath0-band ) , while they are the same in the third case ( @xmath3 cas = hd 5394 ) .
the disk normal is expected to have the same position angle in the sky as that for the intrinsic polarization ( from scattering in the inner disk ; @xcite ) .
we show in figure 12 a comparison of the position angles derived from interferometry with those from polarimetric studies @xcite .
these generally agree within the uncertainties , but there are some exceptions ( @xmath154 cyg = hd 202904 and several cases with large axial ratio @xmath19 where it is difficult to determine position angle ) .
our results are fully consistent for four targets with previous interferometric estimates of @xmath17 .
all these comparisons lend support to our strategy of fixing the @xmath114 , @xmath19 , and @xmath17 parameters in those cases where there were ambiguities with the full four - parameter , gaussian elliptical fit .
we clearly detected the visibility decline due to the disk in 14 stars in our sample .
the ratio of disk to stellar radius ( table 7 , column 13 ) varies from 1.5 to 10 with a mean value of 4.4 among this subsample , but our detections are probably biased towards those cases with larger and brighter disks .
we compare these @xmath0-band diameters with those measured for h@xmath2 ( table 8) in figure 13 , and this demonstrates that in most cases the h@xmath2 disk diameters are much larger .
@xcite attributed this difference to the larger opacity of h@xmath2 compared to the free - free and bound - free opacities that dominate the disk emission in the near - infrared .
there are three targets that fall below the trend : @xmath154 cyg = hd 202904 ( but with an uncertainty that may be consistent with the trend ) , @xmath4 per = hd 10516 , and @xmath6 dra = hd 109387 .
it is curious that the latter two are both short - period binaries , and it may be that the usually larger h@xmath2 emitting regions are truncated by tidal forces , so that their @xmath0-band and h@xmath2 disk sizes are comparable . after setting aside these three discrepant cases ,
the unweighted slope of the relation starting from an origin at @xmath156 is @xmath157 ( s.d . of the mean ) , and this slope is indicated by a dotted line in figure 13 .
such a correlation is helpful in predicting the size of the circumstellar disk in the @xmath0-band from h@xmath2 observations , and vice - versa , especially since be star disks are generally highly variable and simultaneous , multiwavelength observations are usually difficult to obtain .
next , we consider the connection between disk size and brightness .
if the disk gas near the central be star has a temperature that is some fraction of the stellar temperature , then we might expect to find a correlation between the surface intensity @xmath26 of the inner disk and star in the limit of high disk optical depth , @xmath158 for some constant @xmath159 .
we can explore this relationship by considering the gaussian elliptical model prediction for the disk brightness at the stellar equator @xcite , @xmath160 the mean stellar intensity in the model is @xmath161 then we can equate these with proportionality constant @xmath159 to obtain @xmath162 this relation predicts that the disk to star flux ratio ( left hand side ) is approximately proportional to the ratio of disk to star projected area ( right hand side ) .
the gaussian elliptical parameters from table 7 and the stellar angular diameters from table 6 were used to calculate the terms on both sides of the equation , and they are plotted together in figure [ fxratio ] ( where we assumed a minimum uncertainty of @xmath163 for those cases where the formal uncertainties may be underestimated ) .
we see that the be stars with the largest disk to stellar flux ratio are often those with a large ratio of circumstellar to stellar angular diameter . however , a counter example is the case of @xmath6 dra = hd 109387 that is plotted with a large but faint disk ( lower right location in fig .
[ fxratio ] ) .
the fit of the gaussian elliptical model parameters in this case is constrained by a set of 100 m baseline measurements from @xcite ( see fig .
8.10 ) , and if a fit was made from the new measurements alone , then the disk size would be much smaller ( as also suggested by the smaller h@xmath2 disk size shown in fig .
13 ) . given these uncertainties ,
our measurement for @xmath6 dra may be set aside from consideration here .
an unweighted , nonparametric , correlation test for the remaining 13 be stars yields a kendall s statistic of @xmath164 , which has null rejection probability of @xmath165 , i.e. , a value which is not small enough to reject with confidence the null hypothesis of no correlation .
we suspect that the poor correlation results from a break down among the faint disk stars of our simplifying assumption that the disks are optically thick , which may only be applicable to the brightest disk cases .
we show for completeness in figure [ fxratio ] a linear fit of the sample of 13 stars that has a constant of proportionality of @xmath166 ( s.d . of the mean ) .
we caution again that such a relation may only hold for dense , optically thick disks and that our results may be biased against detection of fainter disks in general .
we are planning to investigate further the question of disk surface intensity in another paper that will apply physical models and radiative transfer calculations to fit the observed visibilities .
we found that the criterion for a confident detection of the circumstellar disk is usually a decline in visibility below @xmath167 at the longest baselines available .
we can use this criterion to estimate the limitations on disk sizes that we can detect for be stars at different distances .
the visibility measured along baselines aligned with the projected major axis of the disk is a function of the photospheric flux fraction @xmath114 , the ratio of disk to stellar radius @xmath140 , and the stellar angular diameter @xmath120 ( eq .
17 , 18 , 19 ) .
thus , given @xmath140 and @xmath120 we can find @xmath127 , and the remaining parameter to estimate is @xmath114 . in practice
this could be estimated from a simultaneous analysis of the sed of the be star .
however , for our purpose here , we adopted the relationship between the disk to star flux ratio and projected surface areas ( eq .
29 and fig .
14 ) to estimate @xmath114 from @xmath19 and @xmath140 .
this relationship is poorly defined for fainter disks ( @xmath168 ) , where the disks become optically thin and the ratio of areas argument no longer applies .
the stellar angular diameter is found from the assumed stellar radius and distance , and we made estimates for two cases , b0 v and b8 v types for the be star , and for three distances corresponding to visual magnitudes 3 , 5 , and 7 .
we used stellar radii and absolute magnitudes for these classifications from the compilation of @xcite , and we neglected any extinction in the calculation of distance from the magnitude difference .
figure 15 shows the resulting predicted visibilities for a @xmath0-band measurement with a projected baseline of 300 m as a function of @xmath140 for these different cases .
each plot shows how the visibility at this baseline declines as the disk size increases , and we can use these to estimate the smallest disk detectable . for example , we see in figure 15a for a b0 v star of apparent magnitude 5 that the curve dips below @xmath167 at @xmath169 from which we would infer that only disks larger than this would be detected with the chara array .
as expected , we can detect smaller disks in nearer ( brighter ) be stars .
figure 15b shows the case for a later b8 v type that is somewhat more favorable for detection of smaller disks .
note that at small disk radii we simply assume that all the flux is stellar , so the limiting visibility near @xmath170 corresponds to the stellar visibility .
be stars are generally fast rotators , but it is difficult to determine the actual equatorial rotational velocity @xmath171 from the observed projected rotational velocity @xmath151 ( measured from the rotational doppler broadening of the lines ; @xcite ) because of the unknown stellar inclination @xmath48 .
if we assume that the stellar spin and disk rotational axes are coaligned , then we can use the disk axial ratio to set the stellar inclination ( eq .
we collected the seven @xmath19 values derived from our fits ( table 7 ) together with estimates from previous work ( table 8) to find mean @xmath172 factors for 12 be stars in our program .
we adopted the projected photospheric rotational velocities @xmath173 and the critical rotational velocities @xmath152 from @xcite , who derived these values after making corrections for the effects of gravity darkening .
then we divided @xmath173 by @xmath172 to find the equatorial rotational velocity @xmath171 .
our results appear in table 9 that lists the hd number of the star , @xmath173 and @xmath152 @xcite , @xmath171 , and two ratios , @xmath174 and @xmath175 .
the ratio of angular rotational velocities is given for the roche approximation using expressions from @xcite , @xmath176.\ ] ] we find that most of the stars in this subsample are rotating very quickly , and two targets may have attained the critical rate ( @xmath3 cas = hd 5394 and 48 lib = hd 142983 ) . however , there are two targets with much more moderate rotational velocities , @xmath177 psc = hd 217891 and @xmath154 cyg = hd 202904 .
the former has a rather large axial ratio @xmath178 , and it is possible that the uncertainties would allow a small inclination and hence large rotational velocity .
however , the case of @xmath154 cyg is more difficult to understand .
@xcite argue on the basis of the narrow lines in the spectrum that the axial ratio should be close to @xmath179 if it is actually a rapid rotator .
this is much larger than the value we derive from the gaussian elliptical fits , @xmath180 .
additional interferometric observations are needed to confirm or remove this discrepancy .
we find a mean angular velocity ratio of @xmath181 ( s.d . ) for the sample of 12 be stars or @xmath182 ( s.d . ) if we remove @xmath177 psc and @xmath154 cyg from the sample .
these results support earlier be star angular velocity ratio estimates of 0.88 @xcite , 0.7
1.0 @xcite , 0.8
1.0 @xcite , and @xmath183 @xcite .
the consistency of these results suggests that rapid rotation in be stars plays a key role in the be phenomenon .
our chara array interferometric survey of northern sky be stars has led to the resolution of extended @xmath0-band continuum emission from the disks of 14 stars among of our sample of 24 stars .
we interpreted the visibility measurements with a simple geometrical model that assumes a gaussian elliptical brightness distribution for the circumstellar disks .
the model fits yield estimates of the disk angular diameter , axial ratio , position angle , and the photospheric and disk flux contributions in the @xmath0-band .
we demonstrated that the results are consistent with earlier interferometric studies of the disk h@xmath2 emission and with other estimates for the disk parameters from spectroscopic and polarimetric studies .
we determined estimates for the stellar angular diameters and infrared flux excesses from fits of the ultraviolet and near - infrared spectral energy distributions .
we find that the mean ratio of the @xmath0-band disk to stellar diameter is 4.4 with a range from 1.5 to 10 for the detected cases .
the ratio is similar among both early- and late - type be stars , which suggests that be star disks tend to scale in size and flux with the properties of the central star ( although we caution that fainter , smaller disks may have escaped detection ) .
the diameters of the @xmath0-band emitting region are much smaller than those for the h@xmath2 emission in most cases , and the difference is probably due to the higher opacity of h@xmath2 that extends the emission to the lower density gas found at larger radii
. the apparent axial ratio of the disk emission is related to the disk inclination , and we can reliably assign the derived inclination to the stellar spin axis also .
we used the interferometrically observed axial ratios for a sub - sample of 12 be stars to convert the projected rotational velocity from spectroscopy to the equatorial rotational velocity , and we found that these stars generally rotate close to their critical velocities .
we were surprised to find that so many of our targets ( 14 of 24 ) are members of binary and multiple star systems , and we developed methods to account for the influence of the flux of the companions on the interferometric visibilities .
some be stars may have been spun up through mass transfer in a close binary system @xcite , and we have identified the spectral features of the hot , stripped - down , mass donor star in several cases @xcite .
@xcite has found that a large fraction of close , spectroscopic binaries among solar - type stars have distant third companions , and it is possible that the tertiary received the excess angular momentum from the natal cloud that needed to be removed from the central region in order to form a close binary during star formation .
thus , the presence of a wide companion around a be star may indicate that the central object was originally a close binary that subsequently experienced interaction , mass transfer , and spin - up of the gainer .
the derivation of orbits for these wide companions will be especially useful to measure masses of be stars and to see , for example , if they are overluminous for their mass because of large scale internal mixing caused by mass transfer and fast rotation @xcite . our survey provides a first epoch of diameter measurements that will serve in future studies of the growth and dissipation of disks that are frequently observed in be stars @xcite .
the gaussian elliptical models provide a useful but first - order description of the disks , and it will be important to explore the development of disk asymmetries related to disk instabilities ( such as one - armed spiral features ; @xcite ) .
such studies will require closure phase measurements in addition to interferometric visibility , and instruments at the chara array like the michigan infrared combiner @xcite and vega @xcite that use multiple telescopes and baselines will eventually allow us to reconstruct images of be star disks .
such images will be vitally important to test physical models of the disks and to determine the nature of the companion stars .
we are grateful to vincent coud du foresto , william hartkopf , carol jones , robert kurucz , brian mason , rafael millan - gabet , john monnier , david obrien , deepak raghavan , lewis roberts , philippe stee , and christopher tycner for important contributions to this work .
this material is based upon work supported by the national science foundation under grants ast-0606861 and ast-1009080 ( gies ) and ast-0606958 and ast-0908253 ( mcalister ) .
str acknowledges partial support from nasa grant nnh09ak731 .
yamina touhami also gratefully acknowledges the support of a georgia space grant consortium fellowship and a nasa / nexsci visiting fellowship to the california institute of technology .
institutional support has been provided from the gsu college of arts and sciences and from the research program enhancement fund of the board of regents of the university system of georgia , administered through the gsu office of the vice president for research and economic development .
the _ iue _
spectra presented in this paper were obtained from the mikulski archive for space telescopes ( mast ) at stsci .
stsci is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas5 - 26555 .
support for mast for non - hst data is provided by the nasa office of space science via grant nnx09af08 g and by other grants and contracts .
_ hd 4180 . _ @xcite present a single - lined spectroscopic orbit for hd 4180 plus interferometric observations from npoi of the resolved system a , b .
the components of this binary system are too close for speckle resolution , and the object appeared single in observations by @xcite .
@xcite observed spectral features corresponding to two similar late - b or early - a stars that showed doppler shifts on a timescale of approximately 4 days , and these probably form in the close ( ba , bb ) system that was suspected by @xcite .
we have estimated the @xmath0-band magnitude difference using the magnitude difference from npoi @xmath184 mag @xcite , the estimated spectral types from @xcite , the near - ir color calibration from @xcite , and the flux excess @xmath185 mag from @xcite . _
_ @xmath3 cas is a single - lined spectroscopic binary with a faint ( undetected ) companion @xcite .
although this binary could be resolved in our chara visibility observations , the expected large magnitude difference makes detection very difficult .
the distant and faint b companion @xcite will have no influence on our measurements . @xcite and @xcite discuss chara array @xmath186-band and @xmath25-band measurements of the disk size and orientation . _
_ @xmath4 per is a double - lined spectroscopic system with a hot subdwarf companion @xcite .
we used the fuv flux ratio , temperatures , and gravities from @xcite and derived magnitude differences by scaling model spectral energy distributions from @xcite . _
_ alcyone is a bright pleiades member with seven visual components listed in the wds , but all these components have separations greater than 79 arcsec .
the star appears single in speckle @xcite and ao observations @xcite . _
_ pleione is a single - lined spectroscopic binary with a low mass companion , possibly a hot subdwarf or a m - dwarf @xcite .
the next companion is chara 125 that has a separation of @xmath187 @xcite , but it is not always detected in speckle measurements , indicating a large magnitude difference of @xmath188 mag . @xcite
present a radial velocity study that suggests that the orbital period is @xmath189 y , and the eccentricity is found to be large .
we assumed that the orbital period is the same as that associated with the spectroscopic shell episodes @xmath190 y @xcite and that speckle companion chara 125 is the companion in this orbit @xcite .
we then made a preliminary orbital fit of the speckle data @xcite to arrive at the elements presented in table 5 , although we caution that this is probably only one of a family of possible solutions .
there are five other , fainter but wider ( @xmath191 ) , visual companions in the wds , which should not affect our measurements . _
hd 37202 . _
@xcite present an analysis of the single - lined spectroscopic orbit for @xmath5 tau .
no other components are found in speckle @xcite or interferometric observations @xcite .
interferometric observations of disk asymmetries are described by @xcite , @xcite , and @xcite . _
hd 58715 . _
@xcite suggest that @xmath177 cmi is a single - lined , spectroscopic binary with a period of 218 d , but this result has not yet been confirmed by other investigators .
interferometric studies by @xcite and @xcite show no evidence of a close companion .
furthermore , no companion is found in speckle data @xcite and ao imaging ( @xcite ) .
eight faint and distant companions are listed in the wds .
_ hd 109387 .
_ @xcite show that @xmath6 dra is a single - lined , spectroscopic binary with a faint companion .
@xcite found that the addition of a hot companion improved the fit of the @xmath0-band interferometry , but @xcite point out that density exponent derived by @xcite is significantly lower than that determined from h@xmath2 interferometry .
this discrepancy casts some doubt about the detection of the companion .
there are no published speckle measurements of this star . _
hd 138749 .
_ @xmath192 crb has a companion whose separation has increased from @xmath193 in 1976 to @xmath194 in 2010 according to the wds , but the position angle varied by only @xmath195 over the same interval , and this suggests a large orbital eccentricity and/or an inclination @xmath196 .
the @xmath22-band magnitudes were measured by @xcite , and these suggest that the system consists of a b6 vnne primary and a a2 v secondary . assigning masses for these classifications and assuming that the semimajor axis is close to the smallest observed ( @xmath197 ) , we have calculated a preliminary period that is given in table [ orbital ] .
there are 28 measurements ( from 1976 to 2010 ) of separation and position angle in the fourth catalog of interferometric measurements of binary stars @xcite that we used to derive the preliminary orbital elements that are given in table 5 .
there are no obvious radial velocity variations indicative of a spectroscopic binary @xcite . _
hd 142926 . _ @xcite present a single - lined spectroscopic orbit for 4 her .
they argue that the companion star must be a small and faint object since they see no evidence of its spectral features .
no other companions are observed with speckle interferometry @xcite nor are any companions listed in the wds . _
hd 142983 . _
the spectrum of 48 lib is dominated by shell features that vary on a timescale of a decade , and it is very difficult to study the photospheric spectrum of the star ( b3 : iv : e shell ) to search for radial velocity variations @xcite .
unfortunately , there are no published speckle observations , and there are no companions indicated in the wds .
we assume it is a single object .
recent @xmath186-band interferometric observations are discussed by @xcite . _
hd 148184 .
_ @xcite presents a preliminary single - lined orbit for @xmath198 oph with a period of 34.121 d. there are no available speckle observations , and no companion is indicated in the wds .
@xcite obtained h@xmath2 interferometric observations , and they make no mention of evidence of a companion . _ hd 164284 .
_ the visual companion of 66 oph was first discovered by @xcite and confirmed by @xcite .
we assumed that the current separation corresponds to the angular semimajor axis , and we estimated the orbital period by assigning masses assuming main sequence stars , the temperature from @xcite , and the measured @xmath199 mag @xcite .
there are only four measurements in the fourth catalog of interferometric measurements of binary stars @xcite , but we used these to arrive at the preliminary orbital elements presented in table 5 .
@xcite discuss spectroscopy of the star and pulsational behavior , but no mention is made of a spectroscopic binary companion .
_ hd 166014 .
_ @xcite reported a marginal detection of a close companion to @xmath74 her at a separation of 60 mas , but this was not confirmed in later speckle observations by @xcite .
however , the interferometric measurements for this star showed lower visibilities at short baselines , which could not be fitted by the standard gaussian elliptical model .
consequently , we assumed that this decreased visibility is due to the presence of the companion found by @xcite .
we applied the binary flux dilution correction to fit the data by assuming a companion that is 2.5 mag fainter than the be star in the @xmath0-band .
there is no known spectroscopic companion , and @xcite found no evidence of radial velocity variability . _
hd 198183 . _
the @xmath75 cyg system consists of at least four stars .
component c is distant and faint , so we ignored its flux .
the ab system has a long period ( @xmath200 y ; @xcite ) , and we used the orbital elements from @xcite to estimate the position of a , b .
@xcite and @xcite determined an astrometric orbit for the close pair mca 63 aa , ab that apparently consists of similar magnitude stars .
however , this close pair was not detected in recent speckle observations by @xcite , presumably because their separation was too small at that time .
@xcite notes the presence of some - short term line profile variability that might be explained as a composite spectrum consisting of the be star ( aa ) plus a single - lined spectroscopic binary , so it is possible that b or ab has a companion .
the magnitude difference of a , b was determined in the tycho system by @xcite , and we converted this to a johnson @xmath201 mag using the formulae from @xcite .
the magnitude and color differences of a , b suggest that component b is a late b - type star , and we used @xmath202 , the effective temperature of aa from @xcite , and the main sequence relations from @xcite to estimate the @xmath0-band magnitude difference .
we encountered two problems related to the visibility correction for the inner companion ab .
first , the orbit from @xcite predicted that the aa and ab fringe patterns would cross each other for the baseline orientations of observations from jd 2,455,365 , and consequently , large visibility variations were predicted when none were observed .
we suspect that the orbit needs revision , and for the purposes of the visibility correction , we altered the epoch of periastron from by 1982.668 to by 1981.526 so that the predicted , projected separations at the time of our observations were always larger than half the fringe scan length .
the second problem concerned the flux contribution of ab .
w. hartkopf ( usno ) kindly retrieved his speckle data on the close aa , ab pair and derived an approximate magnitude difference of @xmath203 mag .
because we have no color information for the close pair , we simply assumed that @xmath0-band magnitude difference was the same .
we found that with ab this bright , the dilution correction was too large and led to corrected visibilities larger than one . by increasing this estimate to @xmath204 mag ,
the mean of the corrected visibilities was approximately one .
this indicates that the be star disk was probably unresolved , unless the flux contributions of the companions are actually significantly lower than these estimates .
_ hd 200120 .
_ 59 cyg has a nearby b companion @xcite plus three other very distant and faint components .
b. mason ( usno ) kindly provided us with a preliminary orbit for a , b ( table 5 ) that we used to estimate the position and separation at the times of our chara array observations .
the be star is also a spectroscopic binary with a hot subdwarf companion @xcite , and consequently we assume that the smaller @xmath43 ( brighter ) estimate is more reliable in table 4 . _ hd 202904 . _
@xmath154 cyg has four faint and distant companions listed in the wds , but there is no close companion detected in speckle interferometric observations @xcite .
@xcite discuss spectroscopic radial velocities that may be consistent with binary motion for a period of 11.4 y , but further measurements are required to verify their suggestion .
_ hd 203467 .
_ there are no companions of 6 cep listed in the wds , and , unfortunately , there are no published speckle observations of this star .
spectroscopic observations are discussed by @xcite who show that the profiles vary with a 1.621 d cycle , a period that is probably related to pulsation or rotation . _
hd 209409 .
_ there are no companions of @xmath74 aqr listed in the wds , and no companions were found by @xcite using adaptive optics observations with vlt / naco .
@xcite discuss spectroscopy of this be - shell star and note no evidence of a binary companion .
@xcite obtained preliminary @xmath0-band interferometry with vlti / amber , but they did not resolve the disk in the continuum . _ hd 217675 .
_ @xcite present a re - analysis of all the existing plus new astrometric and radial velocity measurements for @xmath74 and .
they show that the system has a @xmath205 hierarchy and the pairs share a wide orbit with a period of 117 y. the a component is probably a spectroscopic binary consisting of the be star and late - b star companion in a 5.7 y orbit , while the b component consists of a pair of similar late - b stars in a 33 d spectroscopic orbit .
we treated the b component as a single object because the binary separation is so small , and then we corrected the visibilities using the orbits from @xcite and the magnitude differences of @xmath206 @xcite and @xmath207 @xcite .
the flux ratios in the @xmath0-band were estimated using the @xmath208 color calibration from @xcite and spectral types from @xcite . _
hd 217891 .
_ no companions of @xmath177 psc are listed in the wds and none were found in adaptive optics observations by @xcite .
@xcite discuss radial velocity measurements that appear to be relatively constant .
abt , h. a. 1987 , in physics of be stars , proc .
a. slettebak & t. p. snow ( cambridge , uk : cambridge univ . press ) , 470 allende prieto , c. , & lambert , d. l. 1999 , , 352 , 555 ammons , s. m. , robinson , s. e. , strader , j. , et al .
2006 , , 638 , 1004 armstrong , j. t. , mozurkewich , d. , vivekanand , m. , et al .
1992 , , 104 , 241 baize , p. 1993
, , 99 , 205 baize , p. 1983
, , 51 , 479 balachandran , s. 1990 , , 354 , 310 balega , i. i. , & balega , y. y. 1988 , sov .
, 14 , 393 balega , i. i. , balega , y. y. , vasyuk , v. a. , & mcmanus , j. j. 1992 , in iau colloq . 135 : complementary approaches to double and multiple star research ( asp conf .
32 ) , ed .
h. a. mcalister & w. i. hartkopf ( san francisco : asp ) , 469 boden , a. f. 2000 , in principles of long baseline stellar interferometry ( jpl publ .
00 - 009 ) , ed .
p. r. lawson ( pasadena : nasa jpl ) , 9 boesgaard , a. m. , & friel , e. d. 1990 , , 351 , 467 boesgaard , a. m. , & tripicco , m. j. 1986 , , 303 , 724 carciofi , a. c. , okazaki , a. t. , le bouquin , j .- b .
, et al . 2009 , , 504 , 915 claret , a. 2000 , , 363 , 1081 cohen , m. , wheaton , w. a. , & megeath , s. t. 2003 , , 126 , 1090 colina , l. , bohlin , r. , & castelli , f. 1996 , hst instrument science report
cal / scs-008 ( baltimore : stsci ) cranmer , s. r. 2005 , , 634 , 585 cutri , r. m. , skrutskie , m. f. , van dyk , s. , et al .
2003 , 2mass all - sky catalog of point sources ( pasadena : univ .
massachusetts and ipac / california inst . tech . ) da silva , r. , milone , a. c. , & reddy , b. e. 2011 , , 526 , a71 dachs , j. , hanuschik , r. , kaiser , d. , ballereau , d. , & bouchet , p. 1986
, , 63 , 87 delaa , o. , stee , ph . , meilland , a. , et al .
2011 , , 529 , a87 dougherty , s. m. , & taylor , a. r. 1992 , , 359 , 808 dougherty , s. m. , waters , l. b. f. m. , burki , g. , et al .
1994 , , 290 , 609 dyck , h. m. , benson , j. a. , & schloerb , f. p. 1995
, , 110 , 1433 edvardsson , b. , andersen , j. , gustafsson , b. , et al .
1993 , , 275 , 101 ekstrm , s. , meynet , g. , chiappini , c. , hirschi , r. , & maeder , a. , , 489 , 685 fabricius , c. , & makarov , v. v. 2000 , , 356 , 141 fitzpatrick , e. l. 1999 , , 111 , 63 floquet , m. , neiner , c. , janot - pacheco , e. , et al .
2002 , , 394 , 137 frmat , y. , zorec , j. , hubert , a .- m . , & floquet , m. 2005 , , 440 , 305 fuhrmann , k. 1998 , , 338 , 161 gerbaldi , m. , faraggiana , r. , & caffau , e. 2007 , , 472 , 241 gies , d. r. 2001 , in the influence of binaries on stellar population studies ( assl vol .
264 ) , ed .
d. vanbeveren ( dordrecht : kluwer ) , 95 gies , d. r. , bagnuolo , w. g. , jr . , baines , e. k. , et al .
2007 , , 654 , 527 gies , d. r .. bagnuolo , w. g. , jr . ,
ferrara , e. c. , et al .
1998 , , 493 , 440 gies , d. r. , mckibben , w. p. , kelton , p. w. , opal , c. b. , & sawyer , s. 1990 , , 100 , 1601 gray , d. f. 2005 , the observation and analysis of stellar photospheres , 3rd ed .
( cambridge uk : cambridge univ . press ) gray , r. o. , napier , m. g. , & winkler , l. i. 2001 , , 121 , 2148 grundstrom , e. d. 2007 , ph.d .
dissertation , georgia state univ .
grundstrom , e. d. , & gies , d. r. 2006 , , 651 , l53 harmanec , p. 1987 , in physics of be stars , iau coll .
a. slettebak & t. p. snow ( cambridge , uk : cambridge univ . press ) , 339 harmanec , p. , habuda , p. , tefl , s. , et al .
2000 , , 364 , l85 hartkopf , w. i. , mason , b. d. & worley , c. e. 2001 , , 122 , 3472 henize , k. g. , wray , j. d. , parsons , s. b. , & benedict , g. f. 1979 , catalog of far - ultraviolet objective - prism spectrophotometry : skylab experiment s-019 , nasa ref .
1031 ( washington , dc : nasa ) herbison - evans , d. , hanbury brown , r. , davis , j. , & allen , l. r. 1971 , , 151 , 161 horch , e. p. , meyer , r. d. , & van altena , w. f. 2004 , , 127 , 1727 huang , w. , & gies , d. r. 2008 , , 683 , 1045 huang , w. , gies , d. r. , & mcswain , m. v. 2010 , , 722 , 605 janson , m. , bonavita , m. , klahr , h. , et al .
2011 , , 736 , 89 jarad , m. m. , hilditch , r. w. , & skillen , i. 1989 , mnras , 238 , 1085 jones , c. e. , tycner , c. , sigut , t. a. a. , benson , j. a. , & hutter , d. j. 2008 , , 687 , 598 karata , y. , & schuster , w. j. 2006 , , 371 , 1793 koubsk , p. , harmanec , p. , kubt , j. , et al .
1997 , , 328 , 551 koubsk , p. , hummel , c. a. , harmanec , p. , et al .
2010 , , 517 , a24 koubsk , p. , yang , s. , floquet , m. , et al .
2003 , in magnetic fields in o , b and a stars : origin and connection to pulsation , rotation and mass loss ( asp conf .
305 ) , ed .
l. a. balona , h. f. henrichs , & r. medupe ( san francisco : asp ) , 295 kraus , s. , monnier , j. d. , che , x. , et al .
2012 , , 744 , 19 lafrasse , s. , mella , g. , bonneau , d. , et al .
2010 , in optical and infrared interferometry ii , proceedings of the spie , vol . 7734 , ed .
w. c. danchi , f. delplancke , & j. k. rajagopal ( bellingham , wa : spie ) , 4e lanz , t. , & hubeny , i. 2003 , , 146 , 417 lejeune , t. , & schaerer , d. 2001 , , 366 , 538 luthardt , r. , & menchenkova , e. v. 1994 , , 284 , 118 maintz , m. , rivinius , th . , stahl , o. , tefl , s. , & appenzeller , i. 2005 , publ .
, 93 , 21 mamajek , e. e. , meyer , m. r. , & liebert , j. 2002 , , 124 , 1670 markwardt , c. b. 2009 , in astronomical data analysis software and systems xviii ( asp conf
411 ) , ed .
d. a. bohlender , d. durand , & p. dowler ( san francisco : asp ) , 251 masana e. , jordi c. , & ribas i. 2006 , , 450 , 735 mason , b. d. , hartkopf , w. i. , mcalister , h. a. , & sowell , j. r. 1993 , , 106 , 637 mason , b. d. , ten brummelaar , t. , gies , d. r. , hartkopf , w. i. , & thaller , m. l. 1997 , , 114 , 2112 mason , b. d. , wycoff , g. l. , hartkopf , w. i. , douglass , g. g. , & worley , c. e. 2001 , , 122 , 3466 mason , b. d. , hartkopf , w. i. , gies , d. r. , henry , t. j. , & helsel , j. w. 2009 , , 137 , 3358 mcdavid , d. 1999 , , 111 , 494 meade , m. r. , & code , a. d. 1980 , , 42 , 283 meilland , a. , delaa , o. , stee , ph . , et al .
2011 , , 532 , a80 meilland , a. , millour , f. , kanaan , s. , et al .
2012 , , 538 , a110 meilland , a. , millour , f. , stee , ph . , et al .
2007 , , 464 , 73 meilland , a. , stee , ph .
, chesneau , o. , & jones , c. 2009 , , 505 , 687 miroshnichenko , a. s. , bjorkman , k. s. , & krugov , v. d. 2002 , , 114 , 1226 monet , d. g. , levine , s. e. , canzian , b. , et al .
2003 , aj , 125 , 984 monnier , j. d. , anderson , m. , baron , f. , et al .
2010 , in optical and infrared interferometry ii , proceedings of the spie , vol . 7734 , ed .
w. c. danchi , f. delplancke , & j. k. rajagopal ( bellingham , wa : spie ) , 0 g mourard , d. , brio , ph . , perraut , k. , et al .
2011 , , 531 , a110 neiner , c. , floquet , m. , hubert , a. m. , et al .
2005 , , 437 , 257 nemravov , j. , harmanec , p. , koubsk , p. , et al .
2012 , , 537 , a59 nemravov , j. , harmanec , p. , kubt , j. , et al .
2010 , , 516 , a80 obrien , d. p. , mcalister , h. a. , raghavan , d. , et al .
2011 , , 728 , 111 oudmaijer , r. d. , & parr , a. m. 2010 , , 405 , 2439 peters , g. j. , gies , d. r. , grundstrom , e. d. , & mcswain , m. v. 2008 , , 686 , 1280 philip , a. d. , & egret , d. 1980 , , 40 , 199 pols , o. r. , cot , j. , waters , l. b. f. m. , & heise , j. 1991 , , 241 , 419 porter , j. m. , & rivinius , th .
2003 , , 115 , 1153 pott , j .- u . ,
woillez , j. , ragland , s. , et al .
2010 , , 721 , 802 prugniel , ph . , &
soubiran , c. 2001 , , 369 , 1048 quirrenbach , a. , bjorkman , k. s. , bjorkman , j. e. , et al .
1997 , , 479 , 477 raghavan , d. , farrington , c. d. , ten brummelaar , t. a. , et al .
2012 , , 745 , 24 raghavan , d. , mcalister , h. a. , torres , g. , et al .
2009 , , 690 , 394 rivinius , th . ,
baade , d. , & tefl , s. 2003 , , 411 , 229 rivinius , th . ,
tefl , s. , & baade , d. 2006 , , 459 , 137 roberts , l. c. , jr . , turner , n. h. , & ten brummelaar , t. a. 2007 , , 133 , 545 rudjak , d. , boi , h. , harmanec , p. , et al .
2009 , , 506 , 1319 saad , s. m. , kubt , j. , hadrava , p. , et al .
2005 , , 296 , 173 schaefer , g. h. , gies , d. r. , monnier , j. d. , et al .
2010 , , 140 , 1838 schrder , c. , reiners , a. , & schmitt , j. h. m. m. 2009 , , 493 , 1099 sharov , a. s. , & lyutyi , v. m. 1988 , sov .
astron . , 32 , 303 skrutskie , m. f. , cutri , r. m. , stiening , r. , et al .
2006 , , 131 , 1163 smith , m. a. , lopes de oliveira , r. , motch , c. , et al .
2012 , , 540 , a53 soubiran , c. , le campion , j .- f .
, cayrel de strobel , g. , & caillo , a. 2010 , , 515 , a111 starikova , g. a. 1982 , sov .
lett . , 8 , 166 stee , ph .
2011 , in active ob stars : structure , evolution , mass loss , and critical limits , iau symp .
272 , ed . c. neiner
, g. wade , g. meynet , & g. peters ( cambridge , uk : cambridge univ . press ) , 313 stee , ph . ,
delaa , o. , monnier , j. d. , et al .
2012 , a&a , 545 , a59 tefl , s. , le bouquin , j .- b . ,
carciofi , a. c. , et al .
2012 , , 540 , a76 tefl , s. , rivinius , th . ,
carciofi , a. c. , et al .
2009 , a&a , 504 , 929 tanaka , k. , sadakane , k. , narusawa , s .- y .
, et al . 2007 , , 59 , l35 tannirkulam , a. , monnier , j. d. , millan - gabet , r. , et al . 2008 , , 677 , l51 ten brummelaar , t. a. , mcalister , h. a. , ridgway , s. t. , et al . 2005 , , 628 , 453 thompson , g. i. , nandy , k. , jamar , c. , et al .
1978 , catalogue of stellar ultraviolet fluxes ( td1 ) : a compilation of absolute stellar fluxes measured by the sky survey telescope ( s2/68 ) aboard the esro satellite td-1 ( london : science research council ) tokovinin , a. a. 1985 , , 61 , 483 tokovinin , a. , mason , b. d. , & hartkopf , w. i. 2010 , , 139 , 743 tokovinin , a. , thomas , s. , sterzik , m. , & udry , s. 2006 , , 450 , 681 touhami , y. 2012 , ph.d .
dissertation , georgia state univ .
touhami , y. , gies , d. r. , & schaefer , g. h. 2011 , , 729 , 17 touhami , y. , richardson , n. d. , gies , d. r. , et al .
2010 , , 122 , 379 townsend , r. h. d. , owocki , s. p. , & howarth , i. d. 2004 , , 350 , 189 tycner , c. , gilbreath , g. c. , zavala , r. t. , et al .
2006 , , 131 , 2710 tycner , c. , hajian , a. r. , armstrong , j. t. , et al .
2004 , , 127 , 1194 tycner , c. , jones , c. e. , sigut , t. a. a. , et al .
2008 , , 689 , 461 tycner , c. , lester , j. b. , hajian , a. r. , et al .
2005 , , 624 , 359 valenti , j. a. , & fischer , d. a. 2005 , , 159 , 141 van leeuwen , f. 2007 , , 474 , 653 wegner , w. 1994 , , 270 , 229 wright , c. o. , egan , m. p. , kraemer , k. e. , & price , s. d. 2003 , , 125 , 359 young , a. t. 1974 , , 189 , 587 yudin , r. v. 2001 , , 368 , 912 zhuchkov , r. y. , malogolovets , e. v. , kiyaeva , o. v. , et el .
2010 , astron .
, 54 , 1134 lcccccc hd 004180 & @xmath74 cas & b2 ve & 14438 & 3.284 & 164 & 0.582 + hd 005394 & @xmath3 cas & b0 i ve & 26431 & 3.800 & 20 & 0.764 + hd 010516 & @xmath4 per & b0.5 i ve & 25556 & 3.899 & 117 & 0.322 + hd 022192 & @xmath155 per & b4.5
ve & 15767 & 3.465 & 125 & 0.280 + hd 023630 & @xmath209 tau & b5 iiie & 12258 & 3.047 & 124 & 0.727 + hd 023862 & @xmath210 tau & b8 vpe & 12106 & 3.937 & 159 & 0.438 + hd 025940 & 48 per & b4 ve & 16158 & 3.572 & 55 & 0.798 + hd 037202 & @xmath5 tau & b1 i ve & 19310 & 3.732 & 122 & 0.071 + hd 058715 & @xmath177 cmi & b8 ve & 11772 & 3.811 & 140 & 0.779 + hd 109387 & @xmath6 dra & b6 iiipe & 13982 & 3.479 & 102 & 0.660 + hd 138749 & @xmath192 crb & b6vnne & 14457 & 3.745 & 177 & 0.200 + hd 142926 & 4 her & b9 pe & 12076 & 3.917 & 70 & 0.300 + hd 142983 & 48 lib & b3 i ve & 15000 & 3.500 & 50 & 0.405 + hd 148184 & @xmath198 oph & b1.5 vpe & 28783 & 3.913 & 20 & 0.947 + hd 164284 & 66 oph & b2 iv & 21609 & 3.943 & 18 & 0.685 + hd 166014 & @xmath74 her & b9.5 iii & 9800 & 3.500 & 89 & 0.868 + hd 198183 & @xmath75 cyg & b5 ve & 13925 & 3.167 & 30 & 0.826 + hd 200120 & 59 cyg & b1.5 ve & 21750 & 3.784 & 95 & 0.310 + hd 202904 & @xmath154 cyg & b2.5 vne & 19100 & 3.900 & 27 & 0.887 + hd 203467 & 6 cep & b2.5 ve & 17087 & 3.377 & 76 & 0.799 + hd 209409 & @xmath74 aqr & b7 i ve & 12942 & 3.701 & 96 & 0.364 + hd 212076 & 31 peg & b1.5 vne & 12942 & 3.701 & 148 & 0.955 + hd 217675 & @xmath74 and & b6 iiipe & 14052 & 3.229 & 25 & 0.022 + hd 217891 & @xmath177 psc & b5 ve & 14359 &
3.672 & 38 & 0.943 + lcccccccccc hd 004222 & hd 004180 & 8970 & 1 & 4.20 & 2 & a2 v & 0.009 & 3 & @xmath211 & @xmath212 + hd 006210 & hd 005394 & 6065 & 4 & 3.86 & 4 & f6 v & 0.009 & 5 & @xmath213 & @xmath214 + hd 011151 & hd 010516 & 6405 & 6 & 4.00 & 6 & f5 v & 0.003 & 7 & @xmath215 & @xmath216 + hd 020675 & hd 022192 & 6577 & 6 & 4.28 & 6 & f6 v & 0.006 & 7 & @xmath217 & @xmath218 + hd 024167 & hd 023630 & 8200 & 1 & 4.01 & 8 & a5 v & 0.016 & 8 & @xmath219 & @xmath220 + hd 024357 & hd 023862 & 6890 & 1 & 4.30 & 2 & f4 v & 0.008 & 8 & @xmath221 & @xmath222 + hd 025948 & hd 025940 & 6440 & 1 & 4.07 & 8 & f5 v & 0.008 & 7 & @xmath221 & @xmath223 + hd 037147 & hd 037202 & 7200 & 1 & 4.13 & 8 & f0 v & 0.003 & 8 & @xmath224 & @xmath225 + hd 057006 & hd 058715 & 6166 & 9 & 3.77 & 9 & f8 v & 0.001 & 7 & @xmath226 & @xmath227 + hd 111456 & hd 109387 & 6313 & 10 & 4.70 & 10 & f6 v & 0.000 & 7 & @xmath228 & @xmath229 + hd 142640 & hd 142983 & 6481 & 4 & 4.09 & 4 & f6 v & 0.005 & 7 & @xmath230 & @xmath231 + hd 144585 & hd 148184 & 5831 & 11 & 4.03 & 11 & g5 v & 0.005 & 7 & @xmath232 & @xmath233 + hd 159139 & hd 166014 & 9550 & 2 & 4.17 & 2 & a1 v & 0.023 & 8 & @xmath234 & @xmath235 + hd 161941 & hd 164284 & 10512 & 12 & 3.67 & 12 & b9.5 v & 0.180 & 3 & @xmath236 & @xmath237 + hd 166233 & hd 164284 & 6661 & 12 & 3.57 & 12 & f2 v & 0.003 & 7 & @xmath238 & @xmath239 + hd 168914 & hd 166014 & 7600 & 2 & 4.20 & 2 & a7 v & 0.000 & 13 & @xmath240 & @xmath241 + hd 192455 & hd 203467 & 6251 & 4 & 4.05 & 4 & f5 v & 0.003 & 7 & @xmath242 & @xmath243 + hd 196629 & hd 198183 & 6996 & 14 & 4.25 & 15 & f0 v & 0.007 & 8 & @xmath244 & @xmath245 + hd 203454 & hd 202904 & 6146 & 16 & 4.50 & 16 & f8 v & 0.001 & 8 & @xmath246 & @xmath247 + hd 211575 & hd 209409 & 6300 & 17 & 4.00 & 17 & f3 v & 0.006 & 7 & @xmath248 & @xmath249 + hd 213617 & hd 212076 & 7259 & 18 & 4.40 & 18 & f1 v & 0.017 & 8 & @xmath250 & @xmath251 + hd 217877 & hd 217891 & 5953 & 6 & 4.29 & 6 & f8 v & 0.003 & 7 & @xmath252 & @xmath253 + hd 217926 & hd 217891 & 6528 & 14 & 3.63 & 8 & f2 v & 0.005 & 7 & @xmath254 & @xmath255 + hd 218470 & hd 217675 & 6407 & 19 & 4.07 & 19 & f5 v & 0.004 & 7 & @xmath256 & @xmath257 + lcccccccccc hd 004180 & 54756.622&s1/w1&@xmath1375.187 & @xmath13627.643 & 278.114 & 162.970 & 0.705 & 0.042 & 0.689 & 0.046 + hd 004180 & 54756.629&s1/w1&@xmath1344.906 & @xmath13630.398 & 278.055 & 164.379 & 0.693 & 0.042 & 0.838 & 0.045 + hd 004180 & 54756.637&s1/w1&@xmath1315.062 & @xmath13631.747 & 278.024 & 166.438 & 0.821 & 0.047 & 0.826 & 0.051 + hd 004180 & 54756.649&s1/w1&19.818 & @xmath13631.622 & 278.027 & 169.637 & 0.736 & 0.044 & 0.865 & 0.048 + hd 004180 & 54756.661&s1/w1&52.167 & @xmath13629.866 & 278.067 & 173.303 & 0.767 & 0.048 & 0.793 & 0.053 + hd 004180 & 54757.610&s1/w1&@xmath1391.896 & @xmath13625.513 & 278.157 & 162.450 & 0.812 & 0.045 & 0.863 & 0.049 + hd 004180 & 54757.626&s1/w1&@xmath1353.847 & @xmath13629.731 & 278.070 & 163.890 & 0.569 & 0.032 & 0.670 & 0.035 + hd 004180 & 54757.633&s1/w1&@xmath1321.944 & @xmath13631.555 & 278.029 & 165.906 & 0.724 & 0.041 & 0.762 & 0.045 + hd 004180 & 54757.645&s1/w1&11.396 & @xmath13631.820 & 278.023 & 168.790 & 0.609 & 0.032 & 0.680 & 0.035 + hd 004180 & 54757.657&s1/w1&44.336 & @xmath13630.437 & 278.054 & 172.357 & 0.790 & 0.035 & 0.863 & 0.038 + hd 004180 & 54757.665&s1/w1&77.830 & @xmath13627.335 & 278.120 & 176.648 & 0.668 & 0.029 & 0.717 & 0.031 + hd 004180 & 54759.958&s1/e1&230.947 & @xmath13702.578 & 325.380 & 229.379 & 0.609 & 0.024 & 0.670 & 0.026 + hd 004180 & 54759.965&s1/e1&251.849 & @xmath13693.810 & 324.740 & 232.929 & 0.577 & 0.022 & 0.627 & 0.024 + hd 004180 & 54759.973&s1/e1&273.565 & @xmath13683.594 & 323.945 & 236.659 & 0.596 & 0.024 & 0.651 & 0.026 + hd 004180 & 54759.985&s1/e1&297.619 & @xmath13670.821 & 322.880 & 240.810 & 0.552 & 0.024 & 0.597 & 0.027 + hd 004180 & 54759.993&s1/e1&324.424 & @xmath13654.529 & 321.402 & 245.410 & 0.498 & 0.035 & 0.547 & 0.038 + hd 004180 & 54767.629&e1/e2&122.971&0.450&54.103&53.424 & 0.948 & 0.035 & 1.002 & 0.038 + hd 004180 & 54767.637&e1/e2&124.960&3.844&55.004&54.111 & 0.861 & 0.030 & 0.899 & 0.033 + hd 004180 & 54767.641&e1/e2&126.845&7.417&55.902&54.758 & 0.837 & 0.030 & 0.865 & 0.033 + hd 004180 & 54767.649&e1/e2&128.894&11.841&56.948&55.457 & 0.876 & 0.028 & 0.903 & 0.030 + hd 004180 & 54767.657&e1/e2&130.759&16.587&57.990&56.086 & 0.934 & 0.032 & 0.961 & 0.035 + hd 004180 & 54767.665&e1/e2&131.999&20.340&58.760&56.500 & 0.910 & 0.035 & 0.933 & 0.038 + hd 004180 & 54767.668&e1/e2&133.025&24.046&59.475&56.838 & 0.875 & 0.035 & 0.893 & 0.039 + hd 004180 & 54767.692&e1/e2&135.194&38.016&61.787&57.521 & 0.946 & 0.045 & 0.974 & 0.049 + hd 004180 & 54767.700&e1/e2&135.336&42.509&62.411&57.550 & 0.990 & 0.055 & 1.025 & 0.060 + hd 004180 & 54767.708&e1/e2&135.106&48.520&63.159&57.450 & 0.924 & 0.050 & 0.960 & 0.055 + hd 004180 & 55130.751&w1/w2 & @xmath13234.308&64.271 & 106.895 & 106.693 & 0.935 & 0.055 & 0.936 & 0.060 + hd 004180 & 55130.766&w1/w2 & @xmath13240.004&47.195 & 107.615 & 107.615 & 0.952 & 0.051 & 0.988 & 0.055 + hd 004180 & 55130.778&w1/w2 & @xmath13243.174&31.895 & 107.904 & 107.769 & 0.871 & 0.046 & 0.952 & 0.050 + hd 004180 & 55130.794&w1/w2 & @xmath13244.593&16.546 & 107.858 & 107.308 & 0.872 & 0.043 & 1.007 & 0.047 + hd 004180 & 55130.805&w1/w2 & @xmath13244.315&1.671 & 107.492 & 106.283 & 0.772 & 0.042 & 0.930 & 0.046 + hd 005394 & 54754.703&s1/w1&139.084 & @xmath13607.765 & 274.306 & 231.158 & 0.303 & 0.019 & & + hd 005394 & 54754.711&s1/w1&165.175 & @xmath13600.436 & 273.983 & 227.589 & 0.324 & 0.021 & & + hd 005394 & 54754.722&s1/w1&194.453 & @xmath13590.507 & 273.525 & 223.502 & 0.350 & 0.022 & & + hd 005394 & 54754.730&s1/w1&219.570 & @xmath13580.451 & 273.038 & 219.937 & 0.360 & 0.022 & & + hd 005394 & 54754.746&s1/w1&259.566 & @xmath13561.204 & 272.040 & 214.171 & 0.394 & 0.022 & & + hd 005394 & 54754.754&s1/w1&280.443 & @xmath13549.397 & 271.385 & 211.130 & 0.408 & 0.023 & & + hd 005394 & 54758.765&s1/e1 & @xmath13354.366 & @xmath13576.322 & 297.658 & 296.711 & 0.135 & 0.013 & & + hd 005394 & 54758.773&s1/e1 & @xmath13335.716 & @xmath13592.279 & 299.530 & 297.903 & 0.145 & 0.011 & & + hd 005394 & 54758.781&s1/e1 & @xmath13314.532 & @xmath13608.507 & 301.370 & 298.796 & 0.124 & 0.011 & & + hd 005394 & 54758.789&s1/e1 & @xmath13294.521 & @xmath13622.229 & 302.876 & 299.257 & 0.132 & 0.010 & & + hd 005394 & 54758.804&s1/e1 & @xmath13255.926 & @xmath13644.924 & 305.268 & 299.276 & 0.147 & 0.011 & & + hd 005394 & 54758.968&s1/e1&235.831 & @xmath13655.016 & 306.292 & 248.023 & 0.186 & 0.015 & & + hd 005394 & 54758.976&s1/e1&256.817 & @xmath13644.451 & 305.219 & 244.459 & 0.181 & 0.011 & & + hd 005394 & 54760.816&s1/e1 & @xmath13210.001 & @xmath13666.447 & 307.424 & 298.080 & 0.163 & 0.010 & & + hd 005394 & 54760.828&s1/e1 & @xmath13177.144 & @xmath13678.694 & 308.604 & 296.556 & 0.151 & 0.008 & & + hd 005394 & 54760.840&s1/e1 & @xmath13137.760 & @xmath13690.286 & 309.689 & 294.109 & 0.161 & 0.010 & & + hd 005394 & 54766.656&e1/e2&127.804&@xmath137.489&56.326&48.422 & 0.969 & 0.058 & & + hd 005394 & 54766.668&e1/e2&130.796&1.022&57.547&49.382 & 0.882 & 0.062 & & + hd 005394 & 54766.672&e1/e2&131.950&5.101&58.097&50.392 & 0.840 & 0.062 & & + hd 005394 & 54766.680&e1/e2&132.995&9.479&58.661&53.292 & 0.877 & 0.061 & & + hd 005394 & 54766.699&e1/e2&134.968&22.648&60.211&54.232 & 0.848 & 0.050 & & + hd 005394 & 54766.703&e1/e2&135.250&27.140&60.691&55.091 & 0.850 & 0.054 & & + hd 005394 & 55116.718&s1/e1 & @xmath13459.951 & @xmath13432.035 & 277.633 & 276.146 & 0.219 & 0.036 & & + hd 005394 & 55116.726&s1/e1 & @xmath13453.050 & @xmath13446.835 & 279.962 & 278.908 & 0.103 & 0.032 & & + hd 005394 & 55153.031&e1/w1 & @xmath13372.570&592.870 & 308.070 & 233.694 & 0.393 & 0.027 & & + hd 005394 & 55153.039&e1/w1 & @xmath13393.023&580.994 & 308.609 & 232.103 & 0.364 & 0.027 & & + hd 005394 & 55154.023&e1/w1 & @xmath13354.456&602.648 & 307.604 & 235.204 & 0.314 & 0.018 & & + hd 010516 & 54759.000&s1/e1&222.132 & @xmath13699.878 & 323.058 & 288.495 & 0.753 & 0.038 & 0.787 & 0.040 + hd 010516 & 54759.008&s1/e1&244.525 & @xmath13690.563 & 322.307 & 292.424 & 0.718 & 0.036 & 0.728 & 0.038 + hd 010516 & 54759.012&s1/e1&262.220 & @xmath13682.355 & 321.615 & 295.266 & 0.778 & 0.039 & 0.778 & 0.041 + hd 010516 & 54759.020&s1/e1&282.941 & @xmath13671.709 & 320.675 & 298.271 & 0.759 & 0.033 & 0.749 & 0.035 + hd 010516 & 54759.032&s1/e1&302.160 & @xmath13660.737 & 319.655 & 300.718 & 0.773 & 0.041 & 0.763 & 0.043 + hd 010516 & 54759.040&s1/e1&327.099 & @xmath13644.721 & 318.072 & 303.338 & 0.774 & 0.039 & 0.772 & 0.042 + hd 010516 & 54759.907&s1/e1&@xmath1348.629 & @xmath13738.941 & 325.810 & 218.245 & 0.661 & 0.023 & 0.645 & 0.025 + hd 010516 & 54759.915&s1/e1&@xmath1327.475 & @xmath13740.209 & 325.889 & 225.058 & 0.682 & 0.025 & 0.678 & 0.026 + hd 010516 & 54759.918&s1/e1&@xmath136.288 & @xmath13740.771 & 325.924 & 231.666 & 0.698 & 0.026 & 0.715 & 0.027 + hd 010516 & 54759.926&s1/e1&13.171 & @xmath13740.665 & 325.917 & 237.546 & 0.674 & 0.027 & 0.711 & 0.029 + hd 010516 & 54759.930&s1/e1&31.856 & @xmath13740.004 & 325.876 & 248.579 & 0.659 & 0.030 & 1.000 & 0.100 + hd 010516 & 54759.938&s1/e1&51.441 & @xmath13738.719 & 325.796 & 253.709 & 0.618 & 0.030 & 0.692 & 0.032 + hd 010516 & 54759.946&s1/e1&70.139 & @xmath13736.921 & 325.683 & 214.956 & 0.623 & 0.032 & 0.710 & 0.034 + hd 010516 & 54760.899&s1/e1 & @xmath1358.610 & @xmath13738.096 & 325.757 & 222.431 & 0.748 & 0.038 & 0.745 & 0.040 + hd 010516 & 54760.907&s1/e1&@xmath1335.711 & @xmath13739.799 & 325.864 & 229.572 & 0.819 & 0.042 & 0.841 & 0.044 + hd 010516 & 54760.915&s1/e1 & @xmath1313.076 & @xmath13740.667 & 325.917 & 236.415 & 0.743 & 0.043 & 0.788 & 0.045 + hd 010516 & 54760.922&s1/e1&9.379 & @xmath13740.733 & 325.921 & 243.200 & 0.740 & 0.047 & 0.811 & 0.049 + hd 010516 & 54760.930&s1/e1&32.475 & @xmath13739.973 & 325.874 & 248.976 & 0.733 & 0.044 & 0.823 & 0.047 + hd 010516 & 54760.938&s1/e1&52.866 & @xmath13738.602 & 325.789 & 256.563 & 0.767 & 0.046 & 0.867 & 0.049 + hd 010516 & 54760.946&s1/e1&80.824 & @xmath13735.640 & 325.602 & 260.335 & 0.708 & 0.042 & 0.801 & 0.045 + hd 010516 & 54760.950&s1/e1&95.289 & @xmath13733.607 & 325.472 & 263.227 & 0.698 & 0.040 & 0.781 & 0.043 + hd 010516 & 54769.723&e1/e2&135.096&34.364&61.330&31.328 & 0.912 & 0.032 & 1.021 & 0.034 + hd 010516 & 54769.755&e1/e2&134.101&54.703&63.720&24.895 & 0.874 & 0.037 & 0.981 & 0.039 + hd 010516 & 54769.762&e1/e2&133.191&59.162&64.120&23.294 & 0.915 & 0.043 & 1.024 & 0.045 + hd 010516 & 54769.766&e1/e2&132.223&62.922&64.424&21.893 & 0.882 & 0.052 & 0.988 & 0.055 + hd 010516 & 54769.774&e1/e2&131.148&66.438&64.682&20.541 & 0.888 & 0.050 & 0.994 & 0.053 + hd 010516 & 54769.782&e1/e2&129.555&70.870&64.970&18.783 & 0.900 & 0.049 & 1.005 & 0.052 + hd 010516 & 55116.910&s1/e1 & @xmath13107.331 & @xmath13731.650 & 325.344 & 198.200 & 0.809 & 0.052 & 0.907 & 0.055 + hd 010516 & 55116.922&s1/e1&@xmath1370.218 & @xmath13736.912 & 325.683 & 211.070 & 0.774 & 0.041 & 0.869 & 0.043 + hd 010516 & 55128.868&e1/e2&101.151&110.159&65.798&7.388 & 0.992 & 0.035 & 1.041 & 0.037 + hd 010516 & 55128.883&e1/e2&90.130&118.704&65.574&11.468 & 0.995 & 0.036 & 1.044 & 0.038 + hd 010516 & 55128.895&e1/e2&84.477&122.399&65.432&13.929 & 0.977 & 0.041 & 1.025 & 0.043 + hd 010516 & 55131.856&e1/e2&102.124&109.303&65.813&7.150 & 0.954 & 0.034 & 1.010 & 0.036 + hd 010516 & 55131.864&e1/e2&97.846&112.933&65.741&8.407 & 0.934 & 0.038 & 0.988 & 0.040 + hd 010516 & 55131.872&e1/e2&94.108&115.844&65.665&9.815 & 0.997 & 0.028 & 1.053 & 0.030 + hd 010516 & 55131.876&e1/e2&89.598&119.070&65.561&11.696 & 0.930 & 0.032 & 0.980 & 0.034 + hd 010516 & 55152.739&e1/w1&688.207&157.539 & 310.617 & 164.924 & 0.610 & 0.014 & 0.649 & 0.015 + hd 010516 & 55152.743&e1/w1&685.155&179.254 & 311.589 & 157.372 & 0.611 & 0.015 & 0.663 & 0.016 + hd 010516 & 55152.751&e1/w1&681.154&199.973 & 312.330 & 149.851 & 0.610 & 0.019 & 0.674 & 0.020 + hd 010516 & 55152.758&e1/w1&676.420&219.393 & 312.862 & 142.520 & 0.635 & 0.024 & 0.710 & 0.025 + hd 010516 & 55152.762&e1/w1&670.547&239.335 & 313.245 & 134.708 & 0.643 & 0.023 & 0.722 & 0.024 + hd 010516 & 55153.754&e1/w1&677.482&215.359 & 312.765 & 144.065 & 0.564 & 0.019 & 0.633 & 0.021 + hd 010516 & 55153.762&e1/w1&668.702&244.959 & 313.323 & 132.452 & 0.588 & 0.022 & 0.664 & 0.023 + hd 010516 & 55153.766&e1/w1&661.393&265.109 & 313.495 & 124.181 & 0.639 & 0.022 & 0.714 & 0.023 + hd 010516 & 55153.774&e1/w1&653.107&284.940 & 313.500 & 115.751 & 0.660 & 0.026 & 0.724 & 0.028 + hd 010516 & 55153.782&e1/w1&643.799&304.566 & 313.344 & 107.126 & 0.644 & 0.024 & 0.687 & 0.025 + hd 022192 & 54754.914&s1/w1&396.051 & @xmath13484.932 & 275.466 & 274.850 & 0.570 & 0.020 & & + hd 022192 & 54754.941&s1/w1&440.224 & @xmath13432.862 & 271.628 & 271.425 & 0.594 & 0.019 & & + hd 022192 & 54754.973&s1/w1&475.710 & @xmath13367.638 & 264.512 & 261.392 & 0.612 & 0.019 & & + hd 022192 & 54754.980&s1/w1&483.126 & @xmath13346.949 & 261.689 & 257.091 & 0.640 & 0.017 & & + hd 022192 & 54754.988&s1/w1&487.960 & @xmath13329.929 & 259.152 & 253.183 & 0.664 & 0.017 & & + hd 022192 & 54754.996&s1/w1&491.828 & @xmath13312.155 & 256.290 & 248.755 & 0.616 & 0.016 & & + hd 022192 & 54755.004&s1/w1&494.654 & @xmath13293.205 & 252.989 & 243.655 & 0.652 & 0.015 & & + hd 022192 & 54755.012&s1/w1&496.121 & @xmath13275.144 & 249.595 & 238.433 & 0.695 & 0.018 & & + hd 022192 & 54755.019&s1/w1&496.414 & @xmath13257.767 & 246.093 & 233.080 & 0.687 & 0.017 & & + hd 022192 & 54755.027&s1/w1&495.522 & @xmath13239.275 & 242.098 & 227.029 & 0.654 & 0.014 & & + hd 022192 & 54761.008&s1/e1&31.538 & @xmath13744.865 & 328.007 & 254.144 & 0.626 & 0.031 & & + hd 022192 & 54761.016&s1/e1&60.588 & @xmath13742.831 & 327.904 & 261.564 & 0.645 & 0.033 & & + hd 022192 & 54761.028&s1/e1&90.950 & @xmath13739.309 & 327.721 & 268.924 & 0.612 & 0.031 & & + hd 022192 & 54761.035&s1/e1&115.975 & @xmath13735.302 & 327.505 & 274.676 & 0.620 & 0.033 & & + hd 022192 & 54761.043&s1/e1&136.228 & @xmath13731.306 & 327.283 & 279.113 & 0.642 & 0.035 & & + hd 022192 & 54761.051&s1/e1&165.376 & @xmath13724.320 & 326.875 & 285.138 & 0.527 & 0.030 & & + hd 022192 & 54761.059&s1/e1&192.011 & @xmath13716.593 & 326.397 & 290.248 & 0.550 & 0.032 & & + hd 022192 & 54767.844&e1/e2&133.292&60.179&64.343&25.844 & 0.951 & 0.035 & & + hd 022192 & 54767.852&e1/e2&132.335&63.837&64.643&24.805 & 0.992 & 0.050 & & + hd 022192 & 54767.856&e1/e2&131.287&67.204&64.889&23.841 & 0.958 & 0.051 & & + hd 022192 & 54767.864&e1/e2&129.725&71.458&65.160&22.626 & 0.915 & 0.055 & & + hd 022192 & 54767.891&e1/e2&120.558&88.550&65.811&18.207 & 0.933 & 0.058 & & + hd 022192 & 54767.899&e1/e2&118.209&91.831&65.857&17.569 & 0.989 & 0.059 & & + hd 022192 & 55131.930&e1/e2&105.236&106.142&65.760&16.662 & 0.994 & 0.031 & & + hd 022192 & 55131.934&e1/e2&101.202&109.690&65.662&17.098 & 0.954 & 0.035 & & + hd 022192 & 55131.950&e1/e2&93.269&115.809 & 65.421&18.612 & 0.990 & 0.037 & & + hd 023630 & 54756.825&s1/w1&167.341 & @xmath13549.381 & 252.672 & 252.672 & 0.964 & 0.069 & & + hd 023630 & 54756.868&s1/w1&282.276 & @xmath13525.282 & 262.360 & 262.360 & 0.921 & 0.058 & & + hd 023630 & 54757.778&s1/w1&35.734 & @xmath13560.721 & 247.197 & 247.197 & 0.982 & 0.043 & & + hd 023630 & 54757.798&s1/w1&95.433 & @xmath13557.466 & 248.832 & 248.832 & 0.960 & 0.046 & & + hd 023630 & 54757.825&s1/w1&176.327 & @xmath13548.027 & 253.285 & 253.285 & 0.864 & 0.046 & & + hd 023630 & 54792.767&e1/e2&135.180&61.139&65.274&65.274 & 0.898 & 0.033 & & + hd 023630 & 54792.779&e1/e2&134.040&66.114&65.756&65.756 & 0.938 & 0.042 & & + hd 023630 & 54792.795&e1/e2&131.545&71.475&65.867&65.867 & 0.880 & 0.064 & & + hd 023630 & 54792.822&e1/e2&124.778&79.887&65.185&65.185 & 0.857 & 0.035 & & + hd 023630 & 54792.838&e1/e2&118.104&85.475&64.142&64.142 & 0.814 & 0.034 & & + hd 023630 & 54793.806&e1/e2&128.331&76.037&65.628&65.628 & 0.995 & 0.043 & & + hd 023630 & 54793.818&e1/e2&124.264&80.381&65.113&65.113 & 0.867 & 0.047 & & + hd 023630 & 54793.834&e1/e2&119.532&84.406&64.379&64.379 & 0.931 & 0.046 & & + hd 023630 & 54793.846&e1/e2&114.828&87.738&63.580&63.580 & 0.918 & 0.052 & & + hd 023630 & 54793.853&e1/e2&110.004&90.682&62.722&62.722 & 0.907 & 0.049 & & + hd 023630 & 54793.873&e1/e2&98.911&96.211&60.708&60.708 & 0.833 & 0.053 & & + hd 023630 & 54793.885&e1/e2&92.072&98.996&59.480&59.480 & 0.915 & 0.062 & & + hd 023630 & 55153.806&e1/w1&690.865&158.091 & 311.812 & 311.812 & 0.896 & 0.061 & & + hd 023630 & 55153.814&e1/w1&690.464&169.092 & 312.756 & 312.756 & 0.936 & 0.066 & & + hd 023862 & 54757.860&s1/w1&271.038 & @xmath13528.484 & 261.309 & 261.988 & 0.752 & 0.035 & & + hd 023862 & 54757.876&s1/w1&310.060 & @xmath13516.946 & 265.211 & 262.910 & 0.759 & 0.027 & 0.873 & 0.031 + hd 023862 & 54757.888&s1/w1&341.209 & @xmath13505.860 & 268.456 & 262.534 & 0.790 & 0.024 & 0.902 & 0.028 + hd 023862 & 54757.919&s1/w1&400.768 & @xmath13478.212 & 274.511 & 260.014 & 0.872 & 0.034 & 0.994 & 0.039 + hd 023862 & 54757.942&s1/w1&436.869 & @xmath13454.827 & 277.464 & 248.860 & 0.818 & 0.035 & 0.930 & 0.040 + hd 023862 & 54757.985&s1/w1&484.700 & @xmath13402.299 & 277.134 & 243.847 & 0.836 & 0.043 & 0.954 & 0.049 + hd 023862 & 54792.830&e1/e2&122.541&81.960&64.861&31.198 & 0.944 & 0.044 & 1.079 & 0.050 + hd 023862 & 54793.826&e1/e2&123.236&81.343&64.966&31.107 & 0.939 & 0.046 & 1.074 & 0.052 + hd 023862 & 54793.849&e1/e2&113.734&88.465&63.394&32.322 & 0.974 & 0.056 & 1.104 & 0.064 + hd 025940 & 54756.711&s1/w1 & @xmath13218.655 & @xmath13593.974 & 278.471 & 241.216 & 0.761 & 0.042 & & + hd 025940 & 54756.726&s1/w1 & @xmath13182.276 & @xmath13605.864 & 278.360 & 241.079 & 0.751 & 0.044 & & + hd 025940 & 54756.746&s1/w1 & @xmath13120.276 & @xmath13620.573 & 278.110 & 241.444 & 0.779 & 0.045 & & + hd 025940 & 54756.754&s1/w1&@xmath1389.8550 & @xmath13625.449 & 278.000 & 241.913 & 0.811 & 0.049 & & + hd 025940 & 54756.765&s1/w1&@xmath1355.292 & @xmath13629.230 & 277.905 & 242.681 & 0.845 & 0.050 & & + hd 025940 & 54756.777&s1/w1&@xmath1323.801 & @xmath13631.092 & 277.855 & 243.593 & 0.876 & 0.051 & & + hd 025940 & 54756.789&s1/w1&8.924 & @xmath13631.455 & 277.845 & 244.748 & 0.792 & 0.046 & & + hd 025940 & 54756.797&s1/w1&9.953 & @xmath13630.324 & 277.876 & 246.029 & 0.748 & 0.040 & & + hd 025940 & 54757.703&s1/w1 & @xmath13231.784 & @xmath13589.029 & 278.493 & 241.323 & 0.841 & 0.050 & & + hd 025940 & 54757.715&s1/w1 & @xmath13201.126 & @xmath13600.025 & 278.425 & 241.119 & 0.783 & 0.048 & & + hd 025940 & 54757.734&s1/w1 & @xmath13142.573 & @xmath13616.044 & 278.200 & 241.223 & 0.715 & 0.042 & & + hd 025940 & 54757.703&s1/w1 & @xmath13231.784 & @xmath13589.029 & 278.493 & 241.323 & 0.843 & 0.049 & & + hd 025940 & 54757.715&s1/w1 & @xmath13201.126 & @xmath13600.025 & 278.425 & 241.119 & 0.800 & 0.049 & & + hd 025940 & 54757.734&s1/w1 & @xmath13142.573 & @xmath13616.044 & 278.200 & 241.223 & 0.730 & 0.042 & & + hd 025940 & 54757.746&s1/w1 & @xmath13108.068 & @xmath13622.708 & 278.064 & 241.609 & 0.648 & 0.034 & & + hd 025940 & 54760.969&s1/e1 & @xmath13157.314 & @xmath13727.311 & 327.390 & 284.001 & 0.785 & 0.044 & & + hd 025940 & 54760.977&s1/e1 & @xmath13130.427 & @xmath13733.384 & 327.725 & 284.706 & 0.792 & 0.043 & & + hd 025940 & 54760.984&s1/e1 & @xmath13101.866 & @xmath13738.510 & 327.993 & 285.508 & 0.807 & 0.043 & & + hd 025940 & 54794.775&e1/e2&135.242&46.831&62.968&60.035 & 0.944 & 0.044 & & + hd 025940 & 54794.779&e1/e2&134.880&51.319&63.493&60.291 & 0.911 & 0.048 & & + hd 025940 & 54794.786&e1/e2&134.312&55.403&63.922&60.470 & 0.949 & 0.055 & & + hd 025940 & 54794.794&e1/e2&133.376&60.087&64.361&60.614 & 0.892 & 0.052 & & + hd 025940 & 54794.802&e1/e2&132.198&64.546&64.725&60.691 & 0.907 & 0.055 & & + hd 025940 & 54794.806&e1/e2&130.896&68.544&65.008&60.711 & 0.913 & 0.054 & & + hd 025940 & 54794.814&e1/e2&129.250&72.798&65.264&60.684 & 0.892 & 0.054 & & + hd 025940 & 54794.821&e1/e2&127.503&76.677&65.459&60.616 & 0.863 & 0.054 & & + hd 025940 & 54794.829&e1/e2&125.009&81.470&65.648&60.474 & 0.947 & 0.062 & & + hd 025940 & 55152.954&e1/w1&474.998&488.627 & 299.815 & 267.938 & 0.808 & 0.025 & & + hd 025940 & 55153.954&e1/w1&468.329&493.237 & 299.244 & 267.042 & 0.746 & 0.017 & & + hd 037202 & 54756.956&s1/w1&315.314 & @xmath13504.611 & 261.789 & 242.145 & 0.400 & 0.024 & & + hd 037202 & 54756.964&s1/w1&340.858 & @xmath13496.489 & 264.961 & 249.143 & 0.354 & 0.024 & & + hd 037202 & 54756.995&s1/w1&400.023 & @xmath13472.348 & 272.327 & 264.031 & 0.402 & 0.023 & & + hd 037202 & 54757.007&s1/w1&420.278 & @xmath13461.613 & 274.658 & 268.510 & 0.371 & 0.021 & & + hd 037202 & 54757.015&s1/w1&434.880 & @xmath13452.671 & 276.174 & 271.435 & 0.358 & 0.022 & & + hd 037202 & 54757.022&s1/w1&446.674 & @xmath13444.449 & 277.230 & 273.545 & 0.380 & 0.021 & & + hd 037202 & 54757.030&s1/w1&457.800 & @xmath13435.568 & 278.014 & 275.251 & 0.390 & 0.022 & & + hd 037202 & 54757.042&s1/w1&467.205 & @xmath13426.849 & 278.424 & 276.385 & 0.354 & 0.019 & & + hd 037202 & 54758.011&s1/w1&433.810 & @xmath13453.368 & 276.069 & 271.232 & 0.461 & 0.040 & & + hd 037202 & 54758.019&s1/w1&446.273 & @xmath13444.747 & 277.198 & 273.478 & 0.477 & 0.045 & & + hd 037202 & 54758.026&s1/w1&455.748 & @xmath13437.307 & 277.889 & 274.962 & 0.435 & 0.040 & & + hd 037202 & 54758.038&s1/w1&466.863 & @xmath13427.191 & 278.415 & 276.351 & 0.465 & 0.045 & & + hd 037202 & 54758.042&s1/w1&473.745 & @xmath13419.831 & 278.498 & 276.932 & 0.483 & 0.043 & & + hd 037202 & 54758.054&s1/w1&481.449 & @xmath13409.981 & 278.215 & 277.172 & 0.430 & 0.039 & & + hd 037202 & 54758.062&s1/w1&486.635 & @xmath13401.723 & 277.629 & 276.920 & 0.407 & 0.040 & & + hd 037202 & 54758.065&s1/w1&490.242 & @xmath13394.520 & 276.857 & 276.373 & 0.430 & 0.040 & & + hd 037202 & 54759.831&s1/e1 & @xmath13492.805 & @xmath13525.895 & 317.086&63.133 & 0.879 & 0.053 & & + hd 037202 & 54759.839&s1/e1 & @xmath13492.697 & @xmath13534.146 & 319.712&62.032 & 0.775 & 0.047 & & + hd 037202 & 54759.847&s1/e1 & @xmath13491.534 & @xmath13542.350 & 322.031&60.727 & 0.735 & 0.047 & & + hd 037202 & 54759.847&s1/e1 & @xmath13490.830 & @xmath13545.410 & 322.822&60.193 & 0.864 & 0.050 & & + hd 037202 & 54759.855&s1/e1 & @xmath13488.463 & @xmath13552.947 & 324.604&58.782 & 0.857 & 0.041 & & + hd 037202 & 54759.862&s1/e1 & @xmath13485.327 & @xmath13560.197 & 326.096&57.325 & 0.822 & 0.030 & & + hd 037202 & 54759.890&s1/e1 & @xmath13459.301 & @xmath13593.680 & 330.240&50.643 & 0.910 & 0.034 & & + hd 037202 & 54760.866&s1/e1 & @xmath13481.087 & @xmath13567.847 & 327.439&55.721 & 0.759 & 0.052 & & + hd 037202 & 54760.870&s1/e1 & @xmath13477.619 & @xmath13573.084 & 328.221&54.610 & 0.846 & 0.061 & & + hd 037202 & 54760.878&s1/e1 & @xmath13469.439 & @xmath13583.359 & 329.438&52.497 & 0.869 & 0.063 & & + hd 037202 & 54794.060&e1/e2&18.406&108.120&48.253&21.920 & 0.928 & 0.058 & & + hd 037202 & 54794.068&e1/e2&12.478&108.366&47.992&23.965 & 0.882 & 0.063 & & + hd 037202 & 54794.853&e1/e2&133.543&67.679&65.868&31.835 & 0.915 & 0.028 & & + hd 037202 & 54794.861&e1/e2&132.529&69.650&65.870&31.021 & 0.874 & 0.029 & & + hd 037202 & 54794.869&e1/e2&131.329&71.550&65.799&30.160 & 0.956 & 0.031 & & + hd 037202 & 54794.923&e1/e2&111.541&87.405&62.346&19.810 & 0.929 & 0.043 & & + hd 037202 & 54794.931&e1/e2&108.236&89.064&61.669&18.368 & 0.891 & 0.043 & & + hd 037202 & 54794.939&e1/e2&104.510&90.774&60.903&16.808 & 0.934 & 0.042 & & + hd 058715 & 54793.003&e1/e2 & 111.322&74.499&58.933&46.343 & 0.932 & 0.057 & & + hd 058715 & 54793.015&e1/e2 & 105.857&75.560&57.220&44.910 & 0.903 & 0.052 & & + hd 058715 & 54793.046&e1/e2&82.961&76.871&49.760&38.982 & 0.993 & 0.044 & & + hd 058715 & 54793.054&e1/e2&77.393&77.381&48.150&37.775 & 0.882 & 0.044 & & + hd 058715 & 54794.007&e1/e2 & 107.874&75.186&57.851&45.434 & 0.928 & 0.048 & & + hd 058715 & 54794.011&e1/e2 & 106.582&75.428&57.447&45.098 & 0.936 & 0.049 & & + hd 058715 & 54794.039&e1/e2&90.678&77.888&52.592&41.179 & 0.989 & 0.063 & & + hd 058715 & 54795.031&e1/e2&93.593&77.497&53.461&41.863 & 0.968 & 0.062 & & + hd 058715 & 54795.043&e1/e2&87.086&78.339&51.536&40.361 & 0.944 & 0.060 & & + hd 058715 & 54795.050&e1/e2&82.285&78.894&50.154&39.312 & 0.904 & 0.053 & & + hd 058715 & 54795.058&e1/e2&75.292&79.617&48.211&37.883 & 0.883 & 0.050 & & + hd 058715 & 54795.066&e1/e2&69.955&80.106&46.791&36.881 & 0.901 & 0.047 & & + hd 058715 & 55152.980&e1/w1 & 686.529 & 183.952 & 312.702 & 258.769 & 0.607 & 0.039 & & + hd 058715 & 55152.983&e1/w1 & 682.963 & 187.829 & 311.635 & 257.515 & 0.635 & 0.042 & & + hd 058715 & 55152.991&e1/w1 & 676.859 & 192.760 & 309.634 & 255.360 & 0.715 & 0.026 & & + hd 058715 & 55152.999&e1/w1 & 670.171 & 197.000 & 307.326 & 252.998 & 0.735 & 0.026 & & + hd 058715 & 55153.976&e1/w1 & 687.046 & 183.269 & 312.844 & 258.950 & 0.587 & 0.026 & & + hd 058715 & 55153.984&e1/w1 & 683.049 & 187.747 & 311.662 & 257.545 & 0.668 & 0.026 & & + hd 058715 & 55153.987&e1/w1 & 677.491 & 192.309 & 309.847 & 255.583 & 0.678 & 0.028 & & + hd 058715 & 55153.995&e1/w1 & 671.522 & 196.209 & 307.798 & 253.474 & 0.718 & 0.030 & & + hd 058715 & 55154.003&e1/w1 & 663.061 & 200.777 & 304.803 & 250.490 & 0.717 & 0.027 & & + hd 109387 & 55009.693&s1/e2&96.843 & @xmath13523.464 & 234.213 & 234.213 & 0.925 & 0.058 & & + hd 109387 & 55009.709&s1/e2&135.285 & @xmath13512.037 & 233.008 & 233.008 & 0.946 & 0.053 & & + hd 109387 & 55009.717&s1/e2&152.810 & @xmath13505.398 & 232.298 & 232.298 & 0.875 & 0.051 & & + hd 109387 & 55009.740&s1/e2&202.086 & @xmath13481.294 & 229.660 & 229.660 & 0.798 & 0.050 & & + hd 109387 & 55044.685&s1/e1&288.533 & @xmath13571.023 & 281.480 & 281.480 & 0.744 & 0.046 & & + hd 109387 & 55340.800&s2/w1&454.859&@xmath1339.202 & 200.854 & 200.854 & 0.872 & 0.055 & & + hd 109387 & 55340.831&s2/w1&432.543&37.965 & 191.026 & 191.026 & 0.915 & 0.060 & & + hd 109387 & 55340.843&s2/w1&417.555&71.308 & 186.360 & 186.360 & 0.907 & 0.045 & & + hd 138749 & 54976.937&e1/e2&44.376&121.248&56.805&52.440 & 0.937 & 0.036 & 1.052 & 0.041 + hd 138749 & 54976.957&e1/e2&30.815&123.292&55.913&53.560 & 0.932 & 0.043 & 1.061 & 0.049 + hd 138749 & 55340.890&s2/w1&460.737 & @xmath13296.485 & 241.040 & 146.197 & 0.945 & 0.048 & 1.038 & 0.052 + hd 138749 & 55368.678&s1/e1 & @xmath13399.306 & @xmath13634.859 & 329.970 & 272.403 & 0.864 & 0.044 & 0.904 & 0.046 + hd 138749 & 55368.694&s1/e1 & @xmath13373.597 & @xmath13651.798 & 330.534 & 280.073 & 0.825 & 0.041 & 0.937 & 0.046 + hd 138749 & 55368.706&s1/e1 & @xmath13344.618 & @xmath13667.870 & 330.650 & 287.443 & 0.899 & 0.052 & 0.956 & 0.056 + hd 138749 & 55381.760&s1/e1 & @xmath13102.050 & @xmath13735.429 & 326.662 & 321.006 & 0.864 & 0.056 & 0.938 & 0.060 + hd 138749 & 55381.787&s1/e1&@xmath1318.564 & @xmath13740.805 & 326.029 & 325.092 & 0.932 & 0.056 & 1.000 & 0.060 + hd 138749 & 55382.740&s1/e1 & @xmath13147.357 & @xmath13729.256 & 327.330 & 317.432 & 0.866 & 0.045 & 0.915 & 0.047 + hd 138749 & 55382.763&s1/e1&@xmath1386.244 & @xmath13737.030 & 326.478 & 322.023 & 0.828 & 0.040 & 0.830 & 0.041 + hd 142926 & 54666.774&s1/e1&30.879 & @xmath13750.619 & 330.524 & 107.198 & 0.897 & 0.096 & & + hd 142926 & 54666.782&s1/e1&53.255 & @xmath13749.322 & 330.506 & 112.612 & 0.857 & 0.106 & & + hd 142926 & 54666.833&s1/e1&209.170 & @xmath13719.762 & 329.770 & 159.389 & 0.837 & 0.140 & & + hd 142926 & 55009.865&s1/e2&160.860 & @xmath13607.210 & 276.365 & 128.805 & 0.933 & 0.053 & & + hd 142926 & 55009.884&s1/e2&200.328 & @xmath13592.994 & 275.381 & 141.721 & 0.963 & 0.056 & & + hd 142926 & 55009.888&s1/e2&214.384 & @xmath13586.912 & 274.907 & 146.379 & 0.896 & 0.051 & & + hd 142926 & 55009.896&s1/e2&227.778 & @xmath13580.553 & 274.378 & 150.831 & 0.899 & 0.050 & & + hd 142926 & 55009.920&s1/e2&267.352 & @xmath13557.979 & 272.215 & 163.967 & 0.860 & 0.071 & & + hd 142926 & 55367.771&s1/e1 & @xmath13226.953 & @xmath13713.827 & 329.549 & 102.526 & 0.913 & 0.051 & & + hd 142926 & 55367.783&s1/e1 & @xmath13195.798 & @xmath13723.839 & 329.908&96.554 & 0.958 & 0.056 & & + hd 142926 & 55367.791&s1/e1 & @xmath13166.932 & @xmath13731.570 & 330.137&92.492 & 0.964 & 0.059 & & + hd 142926 & 55367.799&s1/e1 & @xmath13141.225 & @xmath13737.295 & 330.280&90.200 & 0.855 & 0.048 & & + hd 142926 & 55367.818&s1/e1&@xmath1389.774 & @xmath13745.697 & 330.448&89.627 & 0.979 & 0.054 & & + hd 142926 & 55367.826&s1/e1&@xmath1362.982 & @xmath13748.541 & 330.494&91.432 & 0.967 & 0.060 & & + hd 142926 & 55367.846&s1/e1&@xmath137.623 & @xmath13751.234 & 330.533&99.191 & 0.979 & 0.069 & & + hd 142926 & 55368.838&s1/e1&@xmath1322.053 & @xmath13750.940 & 330.529&96.697 & 0.943 & 0.032 & & + hd 142926 & 55368.846&s1/e1&12.878 & @xmath13751.160 & 330.532 & 103.228 & 0.944 & 0.031 & & + hd 142926 & 55368.857&s1/e1&46.772 & @xmath13749.770 & 330.512 & 110.993 & 0.898 & 0.030 & & + hd 142926 & 55368.865&s1/e1&69.455 & @xmath13747.947 & 330.485 & 116.817 & 0.889 & 0.029 & & + hd 142926 & 55368.873&s1/e1&92.188 & @xmath13745.391 & 330.443 & 123.064 & 0.921 & 0.030 & & + hd 142926 & 55381.829&s1/e1&65.757 & @xmath13748.294 & 330.490 & 115.838 & 0.902 & 0.132 & & + hd 142926 & 55382.779&s1/e1&@xmath1374.464 & @xmath13747.447 & 330.477&90.489 & 0.948 & 0.045 & & + hd 142926 & 55382.790&s1/e1&@xmath1342.888 & @xmath13750.010 & 330.516&93.662 & 0.982 & 0.046 & & + hd 142926 & 55382.802&s1/e1&@xmath1315.243 & @xmath13751.115 & 330.531&97.836 & 0.924 & 0.049 & & + hd 142926 & 55382.806&s1/e1&7.932 & @xmath13751.231 & 330.532 & 102.205 & 0.962 & 0.049 & & + hd 142926 & 55382.814&s1/e1&30.267 & @xmath13750.645 & 330.525 & 107.057 & 0.884 & 0.045 & & + hd 142983 & 55011.700&s1/w2&112.136 & @xmath13295.380 & 139.006 & 70.132 & 0.988 & 0.074 & & + hd 142983 & 55011.708&s1/w2&123.875 & @xmath13296.626 & 141.428 & 69.030 & 0.938 & 0.069 & & + hd 142983 & 55011.716&s1/w2&135.067 & @xmath13297.955 & 143.929 & 68.210 & 0.930 & 0.072 & & + hd 142983 & 55013.692&s1/w2&107.516 & @xmath13294.930 & 138.111 & 70.632 & 0.950 & 0.051 & & + hd 142983 & 55013.700&s1/w2&118.650 & @xmath13296.054 & 140.324 & 69.490 & 0.967 & 0.050 & & + hd 142983 & 55046.674&s1/w1&463.050 & @xmath13336.457 & 251.826 & 116.205 & 0.875 & 0.029 & & + hd 142983 & 55046.682&s1/w1&471.151 & @xmath13342.022 & 256.149 & 118.248 & 0.916 & 0.035 & & + hd 142983 & 55046.689&s1/w1&478.538 & @xmath13348.018 & 260.328 & 120.081 & 0.888 & 0.042 & & + hd 142983 & 55046.701&s1/w1&488.031 & @xmath13358.161 & 266.333 & 122.365 & 0.882 & 0.046 & & + hd 142983 & 55046.713&s1/w1&492.957 & @xmath13366.121 & 270.157 & 123.481 & 0.850 & 0.043 & & + hd 142983 & 55046.721&s1/w1&495.510 & @xmath13373.100 & 272.895 & 124.000 & 0.850 & 0.044 & & + hd 142983 & 55046.729&s1/w1&496.428 & @xmath13379.637 & 274.956 & 124.116 & 0.775 & 0.046 & & + hd 148184 & 54627.787&s2/w2&126.666 & @xmath13224.698 & 113.484 & 110.056 & 0.823 & 0.040 & & + hd 148184 & 54627.799&s2/w2&139.956 & @xmath13227.260 & 117.426 & 113.641 & 0.893 & 0.047 & & + hd 148184 & 54627.807&s2/w2&152.452 & @xmath13230.007 & 121.405 & 117.284 & 0.811 & 0.046 & & + hd 148184 & 54628.787&s2/w2&128.750 & @xmath13225.077 & 114.082 & 110.598 & 0.872 & 0.063 & & + hd 148184 & 54628.799&s2/w2&145.465 & @xmath13228.429 & 119.148 & 115.215 & 0.811 & 0.055 & & + hd 148184 & 54628.810&s2/w2&160.231 & @xmath13231.902 & 124.014 & 119.682 & 0.914 & 0.057 & & + hd 148184 & 54628.822&s2/w2&174.203 & @xmath13235.711 & 128.952 & 124.243 & 0.867 & 0.055 & & + hd 148184 & 54628.834&s2/w2&187.099 & @xmath13239.768 & 133.806 & 128.749 & 0.919 & 0.057 & & + hd 148184 & 55013.786&s1/w2&221.156 & @xmath13292.421 & 161.305 & 155.327 & 0.696 & 0.044 & & + hd 148184 & 55013.806&s1/w2&241.808 & @xmath13300.476 & 169.689 & 163.154 & 0.753 & 0.046 & & + hd 164284 & 54627.825&s2/w2&89.106 & @xmath13328.711 & 149.840 & 137.424 & 0.716 & 0.040 & 0.826 & 0.046 + hd 164284 & 54627.833&s2/w2&102.037 & @xmath13328.316 & 151.262 & 137.230 & 0.709 & 0.044 & 0.807 & 0.050 + hd 164284 & 54627.845&s2/w2&114.585 & @xmath13327.870 & 152.806 & 137.136 & 0.740 & 0.040 & 0.842 & 0.046 + hd 164284 & 54627.853&s2/w2&126.037 & @xmath13327.406 & 154.351 & 137.129 & 0.826 & 0.044 & 0.949 & 0.050 + hd 164284 & 54627.861&s2/w2&137.571 & @xmath13326.880 & 156.033 & 137.199 & 0.840 & 0.044 & 0.955 & 0.050 + hd 164284 & 54627.868&s2/w2&149.041 & @xmath13326.291 & 157.823 & 137.341 & 0.844 & 0.042 & 0.968 & 0.048 + hd 164284 & 54627.876&s2/w2&159.587 & @xmath13325.685 & 159.567 & 137.534 & 0.858 & 0.046 & 0.978 & 0.052 + hd 164284 & 54627.888&s2/w2&171.767 & @xmath13324.899 & 161.691 & 137.827 & 0.872 & 0.048 & 0.997 & 0.055 + hd 164284 & 54627.896&s2/w2&181.990 & @xmath13324.157 & 163.556 & 138.128 & 0.852 & 0.044 & 0.975 & 0.050 + hd 164284 & 54627.907&s2/w2&194.160 & @xmath13323.155 & 165.865 & 138.544 & 0.920 & 0.049 & 1.051 & 0.056 + hd 164284 & 54627.915&s2/w2&203.831 & @xmath13322.248 & 167.759 & 138.912 & 0.806 & 0.044 & 0.925 & 0.050 + hd 164284 & 54627.923&s2/w2&212.586 & @xmath13321.319 & 169.508 & 139.267 & 0.792 & 0.042 & 0.900 & 0.048 + hd 164284 & 54627.935&s2/w2&220.166 & @xmath13320.410 & 171.041 & 139.582 & 0.891 & 0.054 & 1.022 & 0.062 + hd 166014 & 54973.941&e1/e2&117.932&88.141&64.776&63.092 & 0.865 & 0.026 & 0.943 & 0.029 + hd 166014 & 54973.956&e1/e2&108.693&94.991&63.510&60.733 & 0.868 & 0.030 & 0.947 & 0.032 + hd 166014 & 54973.968&e1/e2&104.262&97.711&62.867&59.522 & 0.899 & 0.025 & 0.980 & 0.027 + hd 166014 & 54973.976&e1/e2&99.882&100.132&62.225&58.290 & 0.894 & 0.022 & 0.974 & 0.024 + hd 166014 & 54973.987&e1/e2&92.572&103.694&61.156&56.174 & 0.892 & 0.029 & 0.973 & 0.031 + hd 166014 & 54973.995&e1/e2&88.031&105.652&60.504&54.832 & 0.901 & 0.026 & 0.982 & 0.029 + hd 166014 & 54974.003&e1/e2&82.702&107.736&59.755&53.238 & 0.911 & 0.030 & 0.993 & 0.033 + hd 166014 & 55010.703&s1/e2 & @xmath13352.074 & @xmath13516.042 & 274.847 & 238.510 & 0.681 & 0.045&0.742 & 0.049 + hd 166014 & 55010.711&s1/e2 & @xmath13345.619 & @xmath13522.711 & 275.699 & 237.513 & 0.699 & 0.044 & 0.762 & 0.048 + hd 166014 & 55010.718&s1/e2 & @xmath13338.381 & @xmath13529.463 & 276.454 & 236.262 & 0.780 & 0.051 & 0.850 & 0.056 + hd 166014 & 55365.777&s1/w1&233.844 & @xmath13555.002 & 264.970&115.293 & 0.887 & 0.027 & 0.967 & 0.029 + hd 166014 & 55365.800&s1/w1&292.489 & @xmath13537.299 & 269.148&120.120 & 0.796 & 0.031 & 0.867 & 0.034 + hd 166014 & 55366.781&s1/w1&247.708 & @xmath13551.302 & 265.911&115.971 & 0.793 & 0.041 & 0.864 & 0.044 + hd 166014 & 55366.789&s1/w1&269.495 & @xmath13544.901 & 267.454&117.620 & 0.836 & 0.049 & 0.911 & 0.053 + hd 166014 & 55366.797&s1/w1&293.036 & @xmath13537.107 & 269.189&120.189 & 0.941 & 0.061 & 1.026 & 0.066 + hd 166014 & 55366.804&s1/w1&314.703 & @xmath13529.032 & 270.823&123.257 & 0.940 & 0.056 & 1.025 & 0.061 + hd 166014 & 55366.816&s1/w1&337.270 & @xmath13519.570 & 272.530&127.139 & 0.902 & 0.058 & 0.983 & 0.064 + hd 166014 & 55368.742&s1/e1 & @xmath13462.476 & @xmath13579.387 & 326.159 & 292.999 & 0.853 & 0.041 & 0.930 & 0.045 + hd 166014 & 55368.754&s1/e1 & @xmath13452.749 & @xmath13591.114 & 327.587 & 291.717 & 0.764 & 0.032 & 0.832 & 0.035 + hd 166014 & 55368.761&s1/e1 & @xmath13441.205 & @xmath13603.092 & 328.762 & 289.841 & 0.731 & 0.029 & 0.797 & 0.031 + hd 166014 & 55368.773&s1/e1 & @xmath13421.018 & @xmath13620.687 & 329.975 & 285.966 & 0.772 & 0.033 & 0.841 & 0.036 + hd 166014 & 55368.781&s1/e1 & @xmath13408.099 & @xmath13630.362 & 330.383 & 283.215 & 0.700 & 0.033 & 0.763 & 0.036 + hd 166014 & 55368.793&s1/e1 & @xmath13389.298 & @xmath13642.828 & 330.641 & 278.950 & 0.832 & 0.046 & 0.907 & 0.050 + hd 166014 & 55368.797&s1/e1 & @xmath13377.650 & @xmath13649.772 & 330.653 & 276.188 & 0.903 & 0.049 & 0.984 & 0.053 + hd 166014 & 55368.804&s1/e1 & @xmath13364.233 & @xmath13657.154 & 330.564 & 272.917 & 0.758 & 0.038 & 0.827 & 0.042 + hd 166014 & 55368.820&s1/e1 & @xmath13334.985 & @xmath13671.341 & 330.094 & 265.532 & 0.863 & 0.042 & 0.940 & 0.046 + hd 166014 & 55381.695&s1/e1 & @xmath13475.834 & @xmath13559.216 & 323.048 & 293.996 & 0.771 & 0.049 & 0.841 & 0.053 + hd 166014 & 55381.703&s1/e1 & @xmath13468.553 & @xmath13570.969 & 324.962 & 293.596 & 0.781 & 0.052 & 0.852 & 0.057 + hd 166014 & 55381.711&s1/e1 & @xmath13459.579 & @xmath13583.074 & 326.638 & 292.653 & 0.771 & 0.056 & 0.840 & 0.061 + hd 166014 & 55381.718&s1/e1 & @xmath13450.679 & @xmath13593.395 & 327.833 & 291.405 & 0.752 & 0.051 & 0.819 & 0.056 + hd 166014 & 55381.726&s1/e1 & @xmath13441.645 & @xmath13602.666 & 328.725 & 289.918 & 0.725 & 0.048 & 0.790 & 0.053 + hd 166014 & 55382.687&s1/e1 & @xmath13479.906 & @xmath13551.443 & 321.625 & 293.991 & 0.704 & 0.044 & 0.768 & 0.048 + hd 166014 & 55382.691&s1/e1 & @xmath13475.065 & @xmath13560.569 & 323.283 & 293.975 & 0.793 & 0.048 & 0.864 & 0.052 + hd 166014 & 55382.699&s1/e1 & @xmath13469.400 & @xmath13569.709 & 324.771 & 293.663 & 0.879 & 0.050 & 0.958 & 0.055 + hd 166014 & 55382.707&s1/e1 & @xmath13462.058 & @xmath13579.931 & 326.231 & 292.951 & 0.808 & 0.051 & 0.880 & 0.056 + hd 166014 & 55382.711&s1/e1 & @xmath13454.688 & @xmath13588.915 & 327.340 & 291.998 & 0.824 & 0.049 & 0.898 & 0.054 + hd 166014 & 55383.711&s1/w1&184.316 & @xmath13566.091 & 261.928 & 115.219 & 0.818 & 0.042 & 0.892 & 0.045 + hd 166014 & 55383.718&s1/w1&204.183 & @xmath13562.025 & 263.083 & 114.813 & 0.887 & 0.046 & 0.967 & 0.050 + hd 166014 & 55405.655&w1/s1 & @xmath13202.470&562.417 & 262.976 & 114.824 & 0.846 & 0.050 & 0.922 & 0.054 + hd 166014 & 55405.667&w1/s1 & @xmath13233.675&555.070 & 264.956 & 115.286 & 0.966 & 0.051 & 1.053 & 0.055 + hd 166014 & 55405.694&w1/s1 & @xmath13304.280&533.059 & 270.032 & 121.696 & 0.850 & 0.046 & 0.927 & 0.050 + hd 166014 & 55405.702&w1/s1 & @xmath13324.862&524.943 & 271.591 & 124.917 & 0.954 & 0.049 & 1.040 & 0.053 + hd 166014 & 55405.710&w1/s1 & @xmath13343.167&516.929 & 272.970 & 128.262 & 0.876 & 0.047 & 0.955 & 0.051 + hd 166014 & 55405.718&w1/s1 & @xmath13361.824&507.871 & 274.339 & 132.118 & 0.826 & 0.046 & 0.900 & 0.050 + hd 166014 & 55405.655&w1/s1 & @xmath13202.470&562.417 & 262.976 & 114.824 & 0.846 & 0.050 & 0.922 & 0.054 + hd 166014 & 55405.667&w1/s1 & @xmath13233.675&555.070 & 264.956 & 115.286 & 0.966 & 0.051 & 1.053 & 0.055 + hd 166014 & 55405.694&w1/s1 & @xmath13304.280&533.059 & 270.032 & 121.696 & 0.850 & 0.046 & 0.927 & 0.050 + hd 166014 & 55405.702&w1/s1 & @xmath13324.862&524.943 & 271.591 & 124.917 & 0.954 & 0.049 & 1.040 & 0.053 + hd 166014 & 55405.710&w1/s1 & @xmath13343.167&516.929 & 272.970 & 128.262 & 0.876 & 0.047 & 0.955 & 0.051 + hd 166014 & 55405.718&w1/s1 & @xmath13361.824&507.871 & 274.339 & 132.118 & 0.826 & 0.046 & 0.900 & 0.050 + hd 198183 & 55010.885&s1/e2 & @xmath13248.823 & @xmath13579.262 & 277.372 & 276.761 & 0.635 & 0.072 & 1.083 & 0.124 + hd 198183 & 55010.897&s1/e2 & @xmath13223.674 & @xmath13590.906 & 277.979 & 276.831 & 0.668 & 0.073 & 1.091 & 0.121 + hd 198183 & 55365.845&s1/w1&110.766 & @xmath13603.531 & 269.966 & 251.306 & 0.664 & 0.027 & 1.029 & 0.041 + hd 198183 & 55365.857&s1/w1&151.449 & @xmath13596.899 & 270.935 & 249.108 & 0.704 & 0.026 & 1.069 & 0.040 + hd 198183 & 55365.865&s1/w1&172.252 & @xmath13592.632 & 271.526 & 248.034 & 0.632 & 0.026 & 1.047 & 0.042 + hd 198183 & 55365.873&s1/w1&191.251 & @xmath13588.185 & 272.116 & 247.082 & 0.514 & 0.020 & 0.832 & 0.031 + hd 198183 & 55365.877&s1/w1&212.566 & @xmath13582.540 & 272.826 & 246.042 & 0.572 & 0.022 & 0.983 & 0.034 + hd 198183 & 55365.912&s1/w1&297.314 & @xmath13552.170 & 275.912 & 242.113 & 0.732 & 0.025 & 1.030 & 0.043 + hd 198183 & 55367.967&s1/e1 & @xmath13237.700 & @xmath13712.916 & 330.632 & 328.516 & 0.521 & 0.025 & 0.835 & 0.040 + hd 198183 & 55367.978&s1/e1 & @xmath13207.543 & @xmath13722.002 & 330.518 & 327.446 & 0.641 & 0.025 & 1.048 & 0.041 + hd 198183 & 55367.990&s1/e1 & @xmath13176.748 & @xmath13729.759 & 330.350 & 326.147 & 0.597 & 0.025 & 0.982 & 0.042 + hd 198183 & 55368.002&s1/e1 & @xmath13147.037 & @xmath13735.907 & 330.172 & 324.735 & 0.610 & 0.028 & 1.045 & 0.047 + hd 198183 & 55381.956&s1/e1 & @xmath13162.820 & @xmath13732.799 & 330.267 & 325.503 & 0.587 & 0.046 & 0.991 & 0.078 + hd 198183 & 55381.983&s1/e1 & @xmath1387.6450 & @xmath13744.582 & 329.851 & 321.557 & 0.543 & 0.040 & 0.881 & 0.066 + hd 198183 & 55381.995&s1/e1 & @xmath1349.5940 & @xmath13747.766 & 329.712 & 319.335 & 0.598 & 0.042 & 1.023 & 0.070 + hd 198183 & 55382.842&s1/e1 & @xmath13429.136 & @xmath13600.319 & 324.662 & 324.178 & 0.590 & 0.034 & 1.004 & 0.059 + hd 198183 & 55382.846&s1/e1 & @xmath13418.669 & @xmath13610.820 & 325.806 & 325.498 & 0.526 & 0.034 & 1.000 & 0.100 + hd 198183 & 55382.854&s1/e1 & @xmath13406.037 & @xmath13622.303 & 326.916 & 326.762 & 0.611 & 0.035 & 1.000 & 0.100 + hd 198183 & 55382.862&s1/e1 & @xmath13393.004 & @xmath13633.039 & 327.821 & 327.768 & 0.631 & 0.033 & 0.998 & 0.051 + hd 198183 & 55382.874&s1/e1 & @xmath13365.927 & @xmath13652.508 & 329.141 & 329.133 & 0.600 & 0.032 & 1.000 & 0.100 + hd 198183 & 55383.753&s1/w1 & @xmath1323.4060 & @xmath13610.649 & 268.860 & 259.400 & 0.599 & 0.021 & 0.987 & 0.033 + hd 198183 & 55383.760&s1/w1&6.489 & @xmath13610.952 & 268.812 & 257.503 & 0.542 & 0.019 & 0.816 & 0.028 + hd 198183 & 55383.772&s1/w1&37.555 & @xmath13610.131 & 268.943 & 255.581 & 0.674 & 0.020 & 1.024 & 0.031 + hd 198183 & 55383.780&s1/w1&64.623 & @xmath13608.465 & 269.208 & 253.956 & 0.621 & 0.023 & 0.886 & 0.035 + hd 200120 & 55009.970&s1/e2&@xmath1388.246 & @xmath13616.197 & 273.870&86.097 & 0.933 & 0.042 & 1.015 & 0.045 + hd 200120 & 55009.982&s1/e2&@xmath1357.199 & @xmath13620.607 & 274.201&85.011 & 0.864 & 0.044 & 0.941 & 0.048 + hd 200120 & 55009.993&s1/e2&@xmath1329.247 & @xmath13622.938 & 274.372&85.704 & 0.880 & 0.050 & 0.959 & 0.054 + hd 200120 & 55043.741&s1/e1 & @xmath13471.532 & @xmath13489.003 & 298.873 & 201.254 & 0.766 & 0.040 & 0.832 & 0.043 + hd 200120 & 55043.748&s1/e1 & @xmath13464.247 & @xmath13505.268 & 301.887 & 198.479 & 0.771 & 0.045 & 0.837 & 0.049 + hd 200120 & 55043.756&s1/e1 & @xmath13455.980 & @xmath13521.176 & 304.670 & 195.354 & 0.773 & 0.047 & 0.842 & 0.051 + hd 200120 & 55043.764&s1/e1 & @xmath13447.224 & @xmath13535.966 & 307.115 & 192.074 & 0.805 & 0.048 & 0.871 & 0.052 + hd 200120 & 55043.772&s1/e1 & @xmath13437.289 & @xmath13550.880 & 309.445 & 188.387 & 0.682 & 0.043 & 0.740 & 0.046 + hd 200120 & 55043.780&s1/e1 & @xmath13426.982 & @xmath13564.750 & 311.492 & 184.603 & 0.751 & 0.041 & 0.816 & 0.045 + hd 200120 & 55069.831&e1/e2&100.630&110.012&65.596&43.030 & 0.989 & 0.043 & 1.076 & 0.047 + hd 200120 & 55069.843&e1/e2&94.069&115.029&65.377&40.448 & 0.993 & 0.051 & 1.081 & 0.055 + hd 200120 & 55069.851&e1/e2&88.505&118.781&65.171&38.310 & 0.997 & 0.056 & 1.085 & 0.061 + hd 200120 & 55069.858&e1/e2&82.816&122.213&64.952&36.175 & 0.953 & 0.053 & 1.038 & 0.057 + hd 200120 & 55366.923&s1/w1&317.374 & @xmath13546.760 & 278.143 & 174.911 & 0.968 & 0.038 & 1.050 & 0.042 + hd 200120 & 55366.934&s1/w1&342.486 & @xmath13530.269 & 277.729 & 183.497 & 0.938 & 0.036 & 1.020 & 0.039 + hd 200120 & 55382.892&s1/e1 & @xmath13347.147 & @xmath13641.213 & 320.801 & 156.839 & 0.745 & 0.038 & 0.810 & 0.041 + hd 200120 & 55382.900&s1/e1 & @xmath13324.485 & @xmath13656.784 & 322.303 & 149.544 & 0.731 & 0.035 & 0.795 & 0.038 + hd 200120 & 55382.912&s1/e1 & @xmath13302.793 & @xmath13669.990 & 323.476 & 142.857 & 0.714 & 0.035 & 0.775 & 0.038 + hd 200120 & 55382.919&s1/e1 & @xmath13277.582 & @xmath13683.543 & 324.585 & 135.499 & 0.764 & 0.037 & 0.831 & 0.040 + hd 200120 & 55382.931&s1/e1 & @xmath13245.074 & @xmath13698.552 & 325.702 & 126.769 & 0.842 & 0.041 & 0.917 & 0.045 + hd 202904 & 54688.926&s1/e1&@xmath1318.049 & @xmath13747.138 & 328.809 & 127.427 & 0.872 & 0.045 & & + hd 202904 & 54688.933&s1/e1&8.602 & @xmath13747.284 & 328.799 & 135.631 & 0.766 & 0.039 & & + hd 202904 & 54688.945&s1/e1&43.737 & @xmath13746.215 & 328.870 & 146.820 & 0.768 & 0.042 & & + hd 202904 & 54688.957&s1/e1&78.904 & @xmath13743.691 & 329.033 & 158.310 & 0.759 & 0.046 & & + hd 202904 & 54688.969&s1/e1&109.304 & @xmath13740.306 & 329.238 & 168.389 & 0.807 & 0.049 & & + hd 202904 & 54688.976&s1/e1&138.676 & @xmath13735.936 & 329.483 & 178.197 & 0.644 & 0.040 & & + hd 202904 & 54688.988&s1/e1&176.052 & @xmath13728.726 & 329.836 & 190.705 & 0.607 & 0.038 & & + hd 202904 & 54689.000&s1/e1&203.983 & @xmath13722.046 & 330.107 & 200.028 & 0.674 & 0.042 & & + hd 202904 & 54689.012&s1/e1&236.219 & @xmath13712.834 & 330.392 & 210.711 & 0.726 & 0.048 & & + hd 202904 & 55044.761&s1/e1 & @xmath13462.100 & @xmath13563.445 & 320.602 & 136.641 & 0.779 & 0.040 & & + hd 202904 & 55044.769&s1/e1 & @xmath13453.250 & @xmath13576.138 & 322.517 & 132.771 & 0.818 & 0.049 & & + hd 202904 & 55382.959&s1/e1 & @xmath13212.399 & @xmath13719.802 & 330.186&86.195 & 0.909 & 0.038 & & + hd 202904 & 55382.967&s1/e1 & @xmath13187.013 & @xmath13726.241 & 329.943&88.507 & 0.821 & 0.034 & & + hd 202904 & 55382.979&s1/e1 & @xmath13154.719 & @xmath13733.074 & 329.631&93.215 & 0.917 & 0.044 & & + hd 202904 & 55382.987&s1/e1 & @xmath13128.678 & @xmath13737.548 & 329.396&98.226 & 0.861 & 0.039 & & + hd 203467 & 55044.891&s1/e1 & @xmath13185.852 & @xmath13654.232 & 299.227 & 255.314 & 0.879 & 0.048 & & + hd 203467 & 55044.903&s1/e1 & @xmath13155.004 & @xmath13664.531 & 300.217 & 253.607 & 0.839 & 0.048 & & + hd 203467 & 55044.919&s1/e1 & @xmath13109.489 & @xmath13676.022 & 301.301 & 251.097 & 0.885 & 0.055 & & + hd 203467 & 55044.926&s1/e1&@xmath1385.855 & @xmath13680.349 & 301.703 & 249.823 & 0.827 & 0.050 & & + hd 203467 & 55044.934&s1/e1&@xmath1363.105 & @xmath13683.499 & 301.993 & 248.628 & 0.832 & 0.042 & & + hd 203467 & 55044.942&s1/e1&@xmath1341.188 & @xmath13685.611 & 302.187 & 247.512 & 0.855 & 0.043 & & + hd 203467 & 55044.954&s1/e1&4.236 & @xmath13687.155 & 302.329 & 245.339 & 0.857 & 0.045 & & + hd 203467 & 55043.887&s1/e1 & @xmath13210.332 & @xmath13644.501 & 298.275 & 256.648 & 0.887 & 0.068 & & + hd 203467 & 55043.895&s1/e1 & @xmath13189.240 & @xmath13652.971 & 299.105 & 255.501 & 0.832 & 0.067 & & + hd 203467 & 55043.903&s1/e1 & @xmath13165.099 & @xmath13661.393 & 299.918 & 254.167 & 0.711 & 0.052 & & + hd 203467 & 55043.911&s1/e1 & @xmath13143.541 & @xmath13667.829 & 300.531 & 252.971 & 0.784 & 0.057 & & + hd 203467 & 55043.919&s1/e1 & @xmath13120.150 & @xmath13673.710 & 301.084 & 251.680 & 0.829 & 0.058 & & + hd 203467 & 55043.926&s1/e1&@xmath1395.132 & @xmath13678.781 & 301.557 & 250.319 & 0.736 & 0.053 & & + hd 203467 & 55043.934&s1/e1&@xmath1374.639 & @xmath13682.025 & 301.858 & 249.229 & 0.746 & 0.050 & & + hd 203467 & 55043.938&s1/e1&@xmath1352.206 & @xmath13684.661 & 302.100 & 248.068 & 0.831 & 0.055 & & + hd 203467 & 55043.946&s1/e1&@xmath1329.587 & @xmath13686.367 & 302.257 & 246.938 & 0.693 & 0.045 & & + hd 203467 & 55043.973&s1/e1&47.696 & @xmath13685.077 & 302.138 & 243.468 & 0.731 & 0.042 & & + hd 203467 & 55043.981&s1/e1&74.719 & @xmath13682.014 & 301.857 & 242.422 & 0.767 & 0.047 & & + hd 203467 & 55043.989&s1/e1&97.478 & @xmath13678.358 & 301.518 & 241.614 & 0.905 & 0.054 & & + hd 203467 & 55043.997&s1/e1&120.190 & @xmath13673.701 & 301.084 & 240.878 & 0.864 & 0.060 & & + hd 203467 & 55340.920&s2/w1&93.560 & @xmath13548.997 & 245.010 & 196.078 & 0.855 & 0.042 & & + hd 203467 & 55340.932&s2/w1&132.342 & @xmath13540.130 & 244.655 & 195.480 & 0.895 & 0.044 & & + hd 203467 & 55340.952&s2/w1&180.064 & @xmath13524.571 & 243.999 & 195.365 & 0.934 & 0.046 & & + hd 203467 & 55340.963&s2/w1&215.169 & @xmath13509.503 & 243.321 & 195.725 & 0.903 & 0.045 & & + hd 203467 & 55340.975&s2/w1&243.015 & @xmath13495.085 & 242.633 & 196.279 & 0.889 & 0.049 & & + hd 209409 & 54687.767&s1/e1 & @xmath13428.715 & @xmath13604.476 & 326.044 & 126.315 & 0.811 & 0.048 & & + hd 209409 & 54688.779&s1/e1 & @xmath13451.948 & @xmath13559.568 & 316.460 & 136.914 & 0.759 & 0.043 & & + hd 209409 & 54688.795&s1/e1 & @xmath13433.490 & @xmath13558.143 & 310.926 & 130.173 & 0.763 & 0.041 & & + hd 209409 & 54688.806&s1/e1 & @xmath13413.988 & @xmath13556.905 & 305.301 & 123.146 & 0.734 & 0.036 & & + hd 209409 & 54688.818&s1/e1 & @xmath13392.844 & @xmath13555.770 & 299.436 & 115.670 & 0.816 & 0.041 & & + hd 209409 & 54688.838&s1/e1 & @xmath13352.592 & @xmath13554.012 & 288.922 & 101.995 & 0.826 & 0.049 & & + hd 209409 & 55011.921&s1/w2&55.456 & @xmath13363.928 & 161.963 & 80.066 & 0.936 & 0.062 & & + hd 209409 & 55011.929&s1/w2&68.713 & @xmath13364.034 & 162.990 & 84.897 & 0.943 & 0.067 & & + hd 209409 & 55011.937&s1/w2&81.916 & @xmath13364.163 & 164.222 & 89.801 & 0.855 & 0.057 & & + hd 209409 & 55011.949&s1/w2&107.182 & @xmath13364.481 & 167.148 & 99.393 & 0.854 & 0.057 & & + hd 209409 & 55011.956&s1/w2&119.283 & @xmath13364.667 & 168.805 & 104.067 & 0.958 & 0.057 & & + hd 209409 & 55011.964&s1/w2&130.852 & @xmath13364.868 & 170.540 & 108.576 & 0.794 & 0.050 & & + hd 209409 & 55012.914&s1/w2&49.312 & @xmath13363.887 & 161.560 & 77.863 & 0.840 & 0.040 & & + hd 209409 & 55013.890&s1/w2&7.902 & @xmath13363.736 & 160.068 & 63.799 & 0.870 & 0.044 & & + hd 209409 & 55013.925&s1/w2&70.078 & @xmath13364.046 & 163.107 & 85.400 & 0.804 & 0.036 & & + hd 209409 & 55013.929&s1/w2&80.944 & @xmath13364.153 & 164.124 & 89.437 & 0.925 & 0.034 & & + hd 209409 & 55013.937&s1/w2&92.515 & @xmath13364.285 & 165.360 & 93.794 & 0.954 & 0.043 & & + hd 209409 & 55413.904&s1/e2 & @xmath13107.555 & @xmath13489.067 & 220.303 & 58.038 & 0.869 & 0.030 & & + hd 209409 & 55413.916&s1/e2&@xmath1384.546 & @xmath13488.843 & 218.256 & 61.262 & 0.869 & 0.030 & & + hd 209409 & 55413.923&s1/e2&@xmath1365.526 & @xmath13488.701 & 216.925 & 64.905 & 0.963 & 0.033 & & + hd 209409 & 55413.931&s1/e2&@xmath1341.632 & @xmath13488.575 & 215.724 & 70.513 & 0.963 & 0.036 & & + hd 209409 & 55413.955&s1/e2&9.457 & @xmath13488.494 & 214.950 & 85.310 & 0.936 & 0.043 & & + hd 209409 & 55414.838&s1/e2 & @xmath13242.611 & @xmath13491.727 & 241.230 & 69.816 & 0.945 & 0.031 & & + hd 209409 & 55414.853&s1/e2 & @xmath13217.468 & @xmath13491.022 & 236.260 & 64.101 & 0.955 & 0.029 & & + hd 209409 & 55414.861&s1/e2 & @xmath13199.716 & @xmath13490.591 & 233.031 & 60.842 & 0.950 & 0.028 & & + hd 209409 & 55414.869&s1/e2 & @xmath13178.886 & @xmath13490.148 & 229.549 & 58.007 & 0.940 & 0.028 & & + hd 209409 & 55414.880&s1/e2 & @xmath13157.221 & @xmath13489.753 & 226.293 & 56.358 & 0.931 & 0.026 & & + hd 212076 & 55045.880&s1/w1&359.202 & @xmath13456.312 & 255.499 & 255.368 & 0.967 & 0.043 & & + hd 212076 & 55045.891&s1/w1&382.341 & @xmath13450.810 & 260.068 & 259.829 & 0.935 & 0.039 & & + hd 212076 & 55367.880&s1/e1 & @xmath13490.121 & @xmath13565.572 & 329.265 & 315.754 & 0.780 & 0.048 & & + hd 212076 & 55367.892&s1/e1 & @xmath13484.724 & @xmath13573.431 & 330.348 & 316.900 & 0.873 & 0.050 & & + hd 212076 & 55367.904&s1/e1 & @xmath13477.105 & @xmath13580.702 & 330.659 & 317.324 & 0.845 & 0.047 & & + hd 212076 & 55367.916&s1/e1 & @xmath13466.939 & @xmath13587.908 & 330.315 & 317.148 & 0.921 & 0.048 & & + hd 212076 & 55367.927&s1/e1 & @xmath13454.366 & @xmath13594.928 & 329.352 & 316.409 & 0.902 & 0.052 & & + hd 212076 & 55367.939&s1/e1 & @xmath13436.168 & @xmath13603.074 & 327.452 & 314.849 & 0.991 & 0.056 & & + hd 212076 & 55368.900&s1/e1 & @xmath13479.483 & @xmath13578.680 & 330.639 & 317.267 & 0.874 & 0.046 & & + hd 212076 & 55368.920&s1/e1 & @xmath13461.805 & @xmath13590.962 & 329.972 & 316.896 & 0.900 & 0.048 & & + hd 212076 & 55368.931&s1/e1 & @xmath13443.510 & @xmath13600.003 & 328.268 & 315.526 & 0.911 & 0.049 & & + hd 217675 & 55044.819&s1/e1 & @xmath13475.953 & @xmath13505.723 & 305.541 & 290.158 & 0.743 & 0.043 & 0.950 & 0.055 + hd 217675 & 55044.826&s1/e1 & @xmath13468.254 & @xmath13523.082 & 308.877 & 295.647 & 0.721 & 0.039 & 0.925 & 0.050 + hd 217675 & 55043.811&s1/e1 & @xmath13483.744 & @xmath13483.104 & 300.788 & 282.589 & 0.852 & 0.043 & 1.000 & 0.100 + hd 217675 & 55043.826&s1/e1 & @xmath13472.681 & @xmath13513.528 & 307.073 & 292.661 & 0.817 & 0.042 & 1.045 & 0.054 + hd 217675 & 55043.850&s1/e1 & @xmath13445.976 & @xmath13560.784 & 315.234 & 306.536 & 0.854 & 0.046 & 1.094 & 0.059 + hd 217675 & 55043.858&s1/e1 & @xmath13434.472 & @xmath13576.193 & 317.495 & 310.540 & 0.892 & 0.049 & 1.000 & 0.100 + hd 217675 & 55043.865&s1/e1 & @xmath13421.916 & @xmath13591.043 & 319.495 & 314.126 & 0.866 & 0.056 & 1.112 & 0.071 + hd 217675 & 55043.873&s1/e1 & @xmath13408.693 & @xmath13604.988 & 321.215 & 317.227 & 0.805 & 0.056 & 1.033 & 0.072 + hd 217675 & 55045.811&s1/w1&96.826 & @xmath13618.224 & 275.311 & 228.533 & 0.728 & 0.045 & 0.943 & 0.058 + hd 217675 & 55045.815&s1/w1&112.630 & @xmath13615.927 & 275.479 & 224.681 & 0.702 & 0.046 & 0.888 & 0.059 + hd 217675 & 55045.826&s1/w1&151.470 & @xmath13608.704 & 275.974 & 214.586 & 0.669 & 0.039 & 0.864 & 0.050 + hd 217675 & 55045.850&s1/w1&212.086 & @xmath13592.603 & 276.918 & 196.908 & 0.724 & 0.036 & 0.935 & 0.047 + hd 217675 & 55045.858&s1/w1&238.145 & @xmath13583.671 & 277.346 & 188.508 & 0.657 & 0.034 & 0.867 & 0.043 + hd 217675 & 55046.803&s1/w1&84.6040 & @xmath13619.753 & 275.198 & 231.414 & 0.761 & 0.034 & 0.973 & 0.044 + hd 217675 & 55046.815&s1/w1&118.710 & @xmath13614.946 & 275.549 & 223.161 & 0.780 & 0.038 & 0.984 & 0.049 + hd 217675 & 55046.827&s1/w1&151.743 & @xmath13608.645 & 275.978 & 214.512 & 0.882 & 0.040 & 1.000 & 0.100 + hd 217675 & 55046.834&s1/w1&182.281 & @xmath13601.295 & 276.436 & 205.909 & 0.825 & 0.039 & 1.052 & 0.050 + hd 217675 & 55046.858&s1/w1&241.580 & @xmath13582.395 & 277.402 & 187.361 & 0.804 & 0.039 & 1.058 & 0.050 + hd 217675 & 55046.866&s1/w1&269.911 & @xmath13570.920 & 277.840 & 177.526 & 0.644 & 0.033 & 0.839 & 0.042 + hd 217675 & 55366.956&s1/w1&183.999 & @xmath13600.828 & 276.460 & 205.404 & 0.716 & 0.029 & 0.920 & 0.038 + hd 217675 & 55366.964&s1/w1&209.194 & @xmath13593.508 & 276.867 & 197.806 & 0.749 & 0.031 & 0.956 & 0.040 + hd 217675 & 55366.980&s1/w1&242.014 & @xmath13582.223 & 277.405 & 187.212 & 0.770 & 0.032 & 0.979 & 0.041 + hd 217675 & 55383.926&s1/w1&223.567 & @xmath13588.828 & 277.107 & 193.271 & 0.825 & 0.039 & 1.069 & 0.050 + hd 217675 & 55383.934&s1/w1&248.237 & @xmath13579.852 & 277.509 & 185.111 & 0.824 & 0.037 & 1.059 & 0.047 + hd 217675 & 55383.950&s1/w1&288.692 & @xmath13562.311 & 278.096 & 170.607 & 0.904 & 0.050 & 1.000 & 0.100 + hd 217675 & 55383.965&s1/w1&329.888 & @xmath13540.167 & 278.468 & 154.134 & 0.907 & 0.052 & 1.000 & 0.100 + hd 217675 & 55383.973&s1/w1&348.011 & @xmath13528.757 & 278.499 & 146.223 & 0.811 & 0.050 & 1.043 & 0.064 + hd 217675 & 55383.985&s1/w1&367.622 & @xmath13515.015 & 278.391 & 137.107 & 0.700 & 0.044 & 0.892 & 0.056 + hd 217675 & 55383.993&s1/w1&384.824 & @xmath13501.550 & 278.132 & 128.551 & 0.786 & 0.048 & 1.008 & 0.062 + hd 217891 & 54687.821&s1/e1 & @xmath13456.640 & @xmath13578.466 & 324.245 & 322.770 & 0.741 & 0.049 & & + hd 217891 & 54689.911&s1/e1 & @xmath13265.340 & @xmath13593.779 & 286.138 & 285.201 & 0.816 & 0.055 & & + hd 217891 & 54689.923&s1/e1 & @xmath13237.812 & @xmath13594.869 & 281.859 & 280.178 & 0.741 & 0.051 & & + hd 217891 & 54689.935&s1/e1 & @xmath13208.736 & @xmath13595.854 & 277.774 & 275.059 & 0.734 & 0.043 & & + hd 217891 & 54689.993&s1/e1 & @xmath1327.3410 & @xmath13598.893 & 263.765 & 248.600 & 0.902 & 0.058 & & + hd 217891 & 54690.009&s1/e1&14.6570 & @xmath13598.929 & 263.586 & 244.092 & 0.891 & 0.057 & & + hd 217891 & 54690.907&s1/w1&406.283 & @xmath13410.876 & 254.223 & 184.300 & 0.882 & 0.038 & & + hd 217891 & 54690.919&s1/w1&429.165 & @xmath13408.495 & 260.676 & 187.690 & 0.948 & 0.039 & & + hd 217891 & 54691.927&s1/w1&441.134 & @xmath13407.041 & 264.081 & 189.529 & 0.822 & 0.079 & & + hd 217891 & 54691.943&s1/w1&460.388 & @xmath13404.244 & 269.554 & 192.545 & 0.883 & 0.087 & & + hd 217891 & 55011.982&s1/w2&90.2910 & @xmath13393.355 & 177.562 & 154.737 & 0.783 & 0.044 & & + hd 217891 & 55011.986&s1/w2&102.915 & @xmath13393.067 & 178.764 & 154.193 & 0.814 & 0.050 & & + hd 217891 & 55012.955&s1/w2&44.4740 & @xmath13394.064 & 174.474 & 157.605 & 0.932 & 0.038 & & + hd 217891 & 55012.963&s1/w2&60.8400 & @xmath13393.869 & 175.343 & 156.422 & 0.972 & 0.037 & & + hd 217891 & 55012.979&s1/w2&88.4290 & @xmath13393.394 & 177.398 & 154.826 & 0.952 & 0.036 & & + hd 217891 & 55012.986&s1/w2&102.024 & @xmath13393.089 & 178.675 & 154.228 & 0.908 & 0.037 & & + hd 217891 & 55012.994&s1/w2&115.925 & @xmath13392.726 & 180.155 & 153.746 & 0.919 & 0.034 & & + hd 217891 & 55013.959&s1/w2&61.197 & @xmath13393.864 & 175.365 & 156.398 & 0.898 & 0.044 & & + hd 217891 & 55013.971&s1/w2&83.840 & @xmath13393.486 & 177.005 & 155.057 & 0.971 & 0.041 & & + hd 217891 & 55013.979&s1/w2&95.628 & @xmath13393.238 & 178.053 & 154.494 & 0.941 & 0.046 & & + hd 217891 & 55013.986&s1/w2&106.524 & @xmath13392.977 & 179.135 & 154.058 & 0.864 & 0.038 & & + hd 217891 & 55013.994&s1/w2&119.643 & @xmath13392.619 & 180.580 & 153.640 & 0.894 & 0.042 & & + hd 217891 & 55013.998&s1/w2&132.252 & @xmath13392.228 & 182.112 & 153.349 & 0.821 & 0.038 & & + hd 217891 & 55045.942&s1/w1&416.604 & @xmath13409.857 & 257.121 & 185.806 & 0.909 & 0.036 & & + hd 217891 & 55045.950&s1/w1&433.474 & @xmath13407.991 & 261.901 & 188.348 & 0.935 & 0.036 & & + hd 217891 & 55045.958&s1/w1&446.526 & @xmath13406.324 & 265.617 & 190.369 & 0.877 & 0.034 & & + hd 217891 & 55045.969&s1/w1&457.573 & @xmath13404.699 & 268.757 & 192.102 & 0.919 & 0.035 & & + hd 217891 & 55045.981&s1/w1&472.818 & @xmath13401.951 & 273.033 & 194.488 & 0.744 & 0.033 & & + hd 217891 & 55045.993&s1/w1&480.466 & @xmath13400.193 & 275.110 & 195.646 & 0.889 & 0.047 & & + hd 217891 & 55046.012&s1/w1&492.688 & @xmath13395.930 & 278.084 & 197.241 & 0.860 & 0.044 & & + hd 217891 & 55046.020&s1/w1&495.282 & @xmath13394.131 & 278.481 & 197.390 & 0.808 & 0.042 & & + hd 217891 & 55414.997&s1/e2&22.898 & @xmath13527.803 & 232.421 & 214.139 & 0.851 & 0.043 & & + hd 217891 & 55415.005&s1/e2&44.433 & @xmath13527.677 & 232.969 & 212.205 & 0.947 & 0.051 & & + hd 217891 & 55415.013&s1/e2&62.762 & @xmath13527.505 & 233.709 & 210.730 & 0.880 & 0.052 & & + hd 217891 & 55152.608&e1/w1&690.717&173.262 & 313.305 & 270.167 & 0.819 & 0.033 & & + hd 217891 & 55152.616&e1/w1&690.649&175.406 & 313.507 & 270.612 & 0.783 & 0.036 & & + hd 217891 & 55152.623&e1/w1&689.367&177.274 & 313.164 & 270.591 & 0.722 & 0.045 & & + hd 217891 & 55153.592&e1/w1&687.498&169.694 & 311.552 & 268.309 & 0.815 & 0.036 & & + hd 217891 & 55153.600&e1/w1&689.590&171.435 & 312.629 & 269.391 & 0.791 & 0.047 & & + hd 217891 & 55153.592&e1/w1&687.498&169.694 & 311.552 & 268.309 & 0.815 & 0.036 & & + hd 217891 & 55153.600&e1/w1&689.590&171.435 & 312.629 & 269.391 & 0.791 & 0.047 & & + lcccccccc hd 004180 & 3 & 1 & a , b & 2.824 & 17 & 2.6 & y&2 + & & & ba , bb & 0.01 : & 0.4 : & 0 : & y&3 + hd 005394 & 3 & 1 & a , b & 1800 : & 2070 : & 6.5 : & n&4 + & & & aa , ab & 0.557 & 8.2 : & 3.5 5.5 & n & 5,6 + hd 010516 & 2 & 1 & a , b & 0.347 & 4.9 & 3.1 & y&7 + hd 022192 & 1 & 1 & & & & & n&8 + hd 023630 & 1 & 1 & & & & & n&4 + hd 023862 & 3 & 9 & aa , ab & 35 : & 150 : & 2.1 : & y & 10 + & & & aa1,aa2 & 0.597 & 8.8 & 1.0 5.0 & y & 11 + hd 025940 & 1 & & & & & & n&8 + hd 037202 & 2 & 1 & a , b & 0.364 & 7.7 & 2.7 4.5 & n & 12 + hd 058715 & 1 & 1 & & & & & n&1 + hd 109387 & 2 & & a , b & 0.169 & 3.3 & 1.4 3.1 & n & 13 + hd 138749 & 2 & 14 & a , b & 161 : & 582 : & 1.8 & y & 15 + hd 142926 & 2 & 1 & a , b & 0.126 & 2.7 & 1.0 2.5 & n & 16 + hd 142983 & 1 & & & & & & n & 17 + hd 148184 & 2 & & a , b & 0.0934 & 3.1 & 3.8 5.9 & n & 18 + hd 164284 & 2 & 14 & a , b & 43 : & 103 : & 2.1 & y & 19 + hd 166014 & 2 & 14 & & & & 2.5 & y&3 + hd 198183 & 3 & 14 & a , b & 461 & 770 & 1.2 & y & 20 + & & & aa , ab & 11.63 & 49 & 0.7 : & y & 21 + hd 200120 & 3 & 14 & aa , ab & 161 : & 208 : & 2.6 & y & 22 + & & & aa1,aa2 & 0.077 & 1.1 & 2.8 4.7 & y & 23 + hd 202904 & 1 & 1 & & & & & n & 24 + hd 203467 & 1 & & & & & & n & 25 + hd 209409 & 1 & 1 & & & & & n & 26 + hd 212076 & 1 & 1 & & & & & n & 27 + hd 217675 & 4 & 14 & a , b & 117.4 & 295 & 2.2 & y & 28 + & & & aa , ab & 5.6 & 61 & 2.1 & y & 28 + & & & ba , bb & 0.090 & 1.9 & 0.0 & y & 28 + hd 217891 & 1 & 1 & & & & & n & 29 + lcccccccccc hd 004180 & 1031.550 & 0.00 & 2452792.200 & 180.0 & 17.00 & 115.0 & 267 & 2.6 & @xcite + hd 010516 & 126.673 & 0.00 & 2450091.770 & 180.0 & 4.90 & 72.5 & 117 & 3.1 & @xcite + hd 023862 & 12615.000 & 0.90 & 2453629.000 & 115.0 & 150.47 & 138.0 & 357 & 2.1 & this work + hd 138749 & 73048.440 & 0.50 & 2433000.000 & 30.0 & 581.61 & 98.0&21 & 1.8 & this work + hd 164284 & 9861.539 & 0.40 & 2451150.000 & 165.0 & 103.23 & 52.0 & 180 & 2.1 & this work + hd 198183 & 4272.603 & 0.52 & 2444797.000 & 253.5 & 48.70 & 135.1 & 119 & 0.7 & @xcite , this work + hd 198183 & 168559.275 & 0.35 & 2378057.800 & 322.0 & 770.00 & 146.3 & 144 & 1.2 & @xcite + hd 200120 & 58973.320 & 0.26 & 2460416.414 & 265.5 & 207.60 & 145.8 & 205 & 2.6 & b. mason , priv . comm . + hd 217675 & 2059.966 & 0.22 & 2452859.405 & 55.0 & 61.00 & 152.0 & 318 & 2.1 & @xcite + hd 217675 & 42879.434 & 0.37 & 2455050.859 & 144.2 & 295.00 & 109.6&7 & 2.2 & @xcite + lcccccccc hd 004180 & 0.118 & 0.008 & 0.333 & 0.009 & 0.070 & 0.016 & 0.938 & 0.014 + hd 005394 & 0.096 & 0.008 & 0.446 & 0.012 & 1.442 & 0.230 & 0.265 & 0.057 + hd 010516 & 0.162 & 0.010 & 0.235 & 0.008 & 0.936 & 0.283 & 0.422 & 0.111 + hd 022192 & 0.099 & 0.008 & 0.307 & 0.008 & 0.533 & 0.264 & 0.612 & 0.150 + hd 023630 & 0.014 & 0.008 & 0.638 & 0.019 & 0.662 & 0.234 & 0.544 & 0.118 + hd 023862 & 0.017 & 0.008 & 0.229 & 0.006 & 0.547 & 0.023 & 0.604 & 0.013 + hd 025940 & 0.104 & 0.008 & 0.329 & 0.009 & 0.671 & 0.312 & 0.539 & 0.157 + hd 037202 & 0.044 & 0.009 & 0.445 & 0.015 & 0.785 & 0.280 & 0.485 & 0.127 + hd 058715 & 0.001 & 0.008 & 0.664 & 0.020 & 0.175 & 0.264 & 0.851 & 0.209 + hd 109387 & 0.022 & 0.008 & 0.385 & 0.011 & 0.435 & 0.036 & 0.670 & 0.022 + hd 138749 & 0.000 & 0.008 & 0.296 & 0.008 & 0.198 & 0.019 & 0.833 & 0.014 + hd 142926 & 0.012 & 0.008 & 0.183 & 0.005 & 0.202 & 0.015 & 0.830 & 0.011 + hd 142983 & 0.000 & 0.007 & 0.172 & 0.004 & 1.329 & 0.020 & 0.294 & 0.005 + hd 148184 & 0.354 & 0.010 & 0.201 & 0.007 & 2.008 & 0.280 & 0.157 & 0.041 + hd 164284 & 0.089 & 0.008 & 0.179 & 0.005 & 0.328 & 0.027 & 0.739 & 0.018 + hd 166014 & 0.000 & 0.007 & 0.521 & 0.013 & 0.066 & 0.354 & 0.941 & 0.312 + hd 198183 & 0.000 & 0.007 & 0.199 & 0.005 & 0.314 & 0.029 & 0.749 & 0.020 + hd 200120 & 0.000 & 0.007 & 0.149 & 0.004 & 1.446 & 0.348 & 0.264 & 0.086 + hd 202904 & 0.113 & 0.008 & 0.295 & 0.009 & 0.053 & 0.404 & 0.952 & 0.363 + hd 203467 & 0.156 & 0.010 & 0.197 & 0.007 & 0.948 & 0.031 & 0.418 & 0.012 + hd 209409 & 0.015 & 0.008 & 0.268 & 0.007 & 0.472 & 0.016 & 0.647 & 0.010 + hd 212076 & 0.059 & 0.008 & 0.189 & 0.005 & 0.782 & 0.017 & 0.487 & 0.008 + hd 217675 & 0.046 & 0.008 & 0.396 & 0.010 & 0.003 & 0.458 & 0.997 & 0.443 + hd 217891 & 0.001 & 0.008 & 0.259 & 0.008 & 0.330 & 0.020 & 0.738 & 0.014 + lccccccccccccccc hd 004180 & 0.583 & 0.101 & 101.4 & 13.8 & 0.500 & 0.000 & 1.027 & 0.173 & 3.03 & 0.558 & 0.019 & 3.243 & 0.493 & y + hd 005394 & 0.722 & 0.038 & 38.2&5.0 & 0.082 & 0.036 & 1.236 & 0.063 & 15.63 & 0.190 & 0.032 & 2.946 & 0.134 & y + hd 010516 & 0.100 & 0.000 & 135.5 & 3.5 & 0.682 & 0.017 & 2.441 & 0.198 & 8.18 & 0.700 & 0.016 & 10.437 & 0.838 & y + hd 022192 & 0.251 & 0.562 & 136.8 & 4.5 & 0.518 & 0.000 & 1.030 & 0.264 & 1.50 & 0.612 & 0.090 & 3.502 & 0.898 & y + hd 023630 & 1.000 & 0.000 & 00.0&0.0 & 0.448 & 0.297 & 0.091 & 0.000 & 5.85 & 1.000 & 0.000 & 1.010 & 0.000 &
n + hd 023862 & 0.438 & 0.000 & 159.0 & 0.0 & 0.500 & 0.000 & 0.364 & 0.220 & 3.33 & 0.715 & 0.057 & 1.879 & 2.315 & m + hd 025940 & 0.866 & 0.000 & 108.0 & 0.0 & 0.412 & 0.000 & 0.597 & 0.245 & 2.55 & 0.539 & 0.009 & 2.072 & 0.676 & y + hd 037202 & 0.148 & 0.027 & 125.4 & 1.3 & 0.422 & 0.016 & 1.790 & 0.073 & 4.30 & 0.534 & 0.013 & 4.146 & 0.159 & y + hd 058715 & 0.783 & 0.000 & 139.5 & 0.0 & 0.717 & 0.000 & 0.777 & 0.181 & 2.77 & 0.851 & 0.132 & 1.539 & 0.194 & y + hd 109387 & 1.000 & 0.000 & 0.0&0.0 & 0.803 & 0.025 & 3.214 & 0.749 & 6.36 & 0.805 & 0.025 & 8.407 & 1.932 & y + hd 138749 & 0.200 & 0.000 & 177.0 & 0.0 & 0.500 & 0.000 & 0.261 & 0.177 & 2.06 & 0.907 & 0.056 & 1.332 & 0.470 & n + hd 142926 & 0.270 & 0.082 & 98.4&5.7 & 0.822 & 0.000 & 1.329 & 0.714 & 0.51 & 0.830 & 0.005 & 7.330 & 3.824 & m + hd 142983 & 0.405 & 0.000 & 50.0&0.0 & 0.242 & 0.000 & 0.836 & 0.164 & 0.75 & 0.294 & 0.005 & 4.963 & 0.936 & y + hd 148184 & 0.947 & 0.000 & 20.0&0.0 & 0.123 & 0.000 & 0.858 & 0.142 & 1.59 & 0.157 & 0.004 & 4.385 & 0.690 & y + hd 164284 & 0.685 & 0.000 & 18.0&0.0 & 0.728 & 0.000 & 0.892 & 0.187 & 2.92 & 0.739 & 0.005 & 5.082 & 1.025 & m + hd 166014 & 0.435 & 0.279 & 67.8 & 32.5 & 0.250 & 0.000 & 0.337 & 0.105 & 1.88 & 0.941 & 0.069 & 1.191 & 0.108 & m + hd 198183 & 0.826 & 0.000 & 30.0&0.0 & 0.133 & 0.000 & 0.163 & 0.198 & 4.91 & 0.749 & 0.177 & 1.292 & 0.779 & n + hd 200120 & 0.310 & 0.000 & 95.0&0.0 & 0.143 & 0.000 & 0.554 & 0.190 & 4.67 & 0.264 & 0.023 & 3.852 & 1.219 & m + hd 202904 & 0.258 & 0.129 & 108.8 & 2.9 & 0.541 & 0.370 & 1.211 & 0.786 & 1.75 & 0.604 & 0.319 & 4.225 & 2.544 & y + hd 203467 & 0.799 & 0.000 & 76.0&0.0 & 0.343 & 0.000 & 0.528 & 0.087 & 1.62 & 0.418 & 0.006 & 2.861 & 0.415 & y + hd 209409 & 0.249 & 0.059 & 107.5 & 2.2 & 0.617 & 0.000 & 1.525 & 0.642 & 1.80 & 0.647 & 0.006 & 5.776 & 2.374 & y + hd 212076 & 0.955 & 0.000 & 148.0 & 0.0 & 0.308 & 0.000 & 0.295 & 0.044 & 1.61 & 0.487 & 0.032 & 1.852 & 0.202 & m + hd 217675 & 0.022 & 0.000 & 25.0&0.0 & 0.340 & 0.195 & 0.056 & 0.000 & 2.23 & 1.000 & 0.000 & 1.010 & 0.000 & n + hd 217891 & 0.702 & 0.150 & 30.6 & 17.1 & 0.692 & 0.309 & 0.804 & 0.593 & 1.81 & 0.722 & 0.279 & 3.261 & 2.098 & y + lcccccccc hd 004180 & 1.00 & & & & 1.90 & 0.10 & h@xmath2 & 1 + hd 005394 & 0.70 & 0.02 & 19 & 2 & 3.47 & 0.02 & h@xmath2 & 2 + hd 005394 & 0.58 & 0.03 & 31 & 1 & 3.59 & 0.04 & h@xmath2 & 3 + hd 005394 & 0.74 & 0.05 & 19 & 5 & 0.76 & 0.05 & @xmath25 & 4 + hd 005394 & 0.75 & 0.05 & 12 & 9 & 0.82 & 0.08 & @xmath186 & 4 + hd 010516 & 0.46 & 0.04 & 118 & 5 & 2.67 & 0.20 & h@xmath2 & 2 + hd 010516 & 0.27 & 0.01 & 119 & 1 & 2.89 & 0.09 & h@xmath2 & 3 + hd 022192 & 0.47 & 0.11 & 147 & 11 & 3.26 & 0.23 & h@xmath2 &
2 + hd 022192 & 0.34 & 0.10 & 96 & 2 & 4.00 & 0.20 & h@xmath2 & 5 + hd 022192 & @xmath258 & & 158 & 10 & 111.00 & 16.00 & 15 ghz & 6 + hd 023630 & 0.95 & 0.22 & 19 & & 2.65 & 0.14 & h@xmath2 & 2 + hd 023630 & 0.75 & 0.05 & 45 & 9 & 2.08 & 0.18 & h@xmath2 & 7 + hd 025940 & 0.89 & 0.13 & 68 & & 2.77 & 0.56 & h@xmath2 & 2 + hd 025940 & 0.77 & 0.10 & 115 & 33 & 2.10 & 0.20 & h@xmath2 & 5 + hd 037202 & 0.28 & 0.02 & 122 & 4 & 4.53 & 0.52 & h@xmath2 & 2 + hd 037202 & 0.31 & 0.07 & 118 & 4 & 3.14 & 0.21 & h@xmath2 & 8 + hd 037202 & 0.24 & 0.14 & 123 & 6 & 1.57 & 0.28 & @xmath186 & 9 + hd 037202 & & & 126 & 2 & & & br@xmath3 & 10 + hd 058715 & & & & & 2.65 & 0.10 & h@xmath2 & 2 + hd 058715 & 0.69 & 0.15 & 40 & 30 & 2.13 & 0.50 & h@xmath2 & 7 + hd 058715 & 0.76 & 0.10 & 140 & 6 & 0.33 & 0.18 & @xmath186 & 10 + hd 109387 & 1.00 & & & & 2.00 & 0.30 & h@xmath2 & 11 + hd 142983 & 0.60 & 0.11 & 50 & 9 & 1.72 & 0.20 & @xmath186 & 12 + hd 142983 & & & 50 & & 1.65 & 0.05 & @xmath0 & 13 + hd 148184 & 1.00 & & & & 3.46 & 0.07 & h@xmath2 & 14 + hd 202904 & 1.00 & & & & 1.00 & 0.20 & h@xmath2 & 11 + hd 217891 & 1.00 & & & & 2.40 & 0.20 & h@xmath2 & 11 + lccccc hd 004180 & 208@xmath25913 & 332@xmath25921 & 249@xmath25926 & [email protected] & [email protected] + hd 005394 & 441@xmath25927 & 577@xmath25936 & 599@xmath25937 & [email protected] & @xmath260 + hd 010516 & 462@xmath25933 & 590@xmath25942 & 487@xmath259 6 & [email protected] & [email protected] + hd 022192 & 295@xmath25915 & 397@xmath25920 & 310@xmath25970 & [email protected] & [email protected] + hd 023630 & 149@xmath259 8 & 274@xmath25915 & 216@xmath25920 & [email protected] & [email protected] + hd 025940 & 220@xmath25913 & 386@xmath25921 & 323@xmath25982 & [email protected] & [email protected] + hd 037202 & 326@xmath259 7 & 466@xmath25913 & 333@xmath259 6 & [email protected] & [email protected] + hd 058715 & 231@xmath25914 & 380@xmath25924 & 327@xmath25991 & [email protected] & [email protected] + hd 142983 & 407@xmath25922 & 501@xmath25928 & 484@xmath25956 & [email protected] & [email protected] + hd 202904 & 167@xmath25920 & 468@xmath25930 & 171@xmath259 7 & [email protected] & [email protected] + hd 209409 & 282@xmath25920 & 391@xmath25927 & 288@xmath259 4 & [email protected] & [email protected] + hd 217891 & 100@xmath259 6 & 367@xmath25924 & 135@xmath25944 & [email protected] & [email protected] + | we report on a high angular resolution survey of circumstellar disks around 24 northern sky be stars .
the @xmath0-band continuum survey was made using the chara array long baseline interferometer ( baselines of 30 331 m ) .
the interferometric visibilities were corrected for the flux contribution of stellar companions in those cases where the be star is a member of a known binary or multiple system . for those targets with good ( @xmath1 ) coverage
, we used a four - parameter gaussian elliptical disk model to fit the visibilities and to determine the axial ratio , position angle , @xmath0-band photospheric flux contribution , and angular diameter of the disk s major axis .
for the other targets with relatively limited ( @xmath1 ) coverage , we constrained the axial ratio , inclination angle , and/or disk position angle where necessary in order to resolve the degeneracy between possible model solutions .
we also made fits of the ultraviolet and infrared spectral energy distributions to estimate the stellar angular diameter and infrared flux excess of each target .
the mean ratio of the disk diameter ( measured in @xmath0-band emission ) to stellar diameter ( from sed modeling ) is 4.4 among the 14 cases where we reliably resolved the disk emission , a value which is generally lower than the disk size ratio measured in the higher opacity h@xmath2 emission line .
we estimated the equatorial rotational velocity from the projected rotational velocity and disk inclination for 12 stars , and most of these stars rotate close to or at the critical rotational velocity . |
let us provide the rigorous analysis of the experimental data regarding fermi gas vs wigner solid transition . according to ref.@xcite
, the melting diagram of 2d wigner solid obeys the condition @xmath1 , where @xmath2 is the coefficient assumed to be a constant at the phase transition , @xmath3 is the coulomb energy associated to neighboring pair of electrons , @xmath4 is the 2d density . within conventional fermi gas model , @xmath5 is the average kinetic energy of single electron , where @xmath6 is the fermi integral of the order of @xmath7 , @xmath8 the dimensionless temperature .
note that the average kinetic energy @xmath9 coincides with the thermal energy @xmath10 for classical boltzmann carriers @xmath11 .
in contrast , @xmath12 for degenerate electrons @xmath13 . in general , the solidification of strongly degenerated electrons is believed to occur at certain value of the coulomb to fermi energy ratio @xmath14 .
we therefore conclude that @xmath15 . in refs.@xcite , this ratio has been erroneously defined as @xmath16 , thus provides wrong estimate for wigner crystal solidification . for low - disorder 2d system wigner
solid was claimed@xcite to exist when @xmath17 .
the diagram of fermi gas to wigner solid transition @xcite according to eq.([wigner_solid ] ) at @xmath18@xcite .
the color rectangular figures correspond to density and temperature range of apparent metal - to insulator transition in si - mosfet@xcite ; p - gaas@xcite ; n - gaas@xcite and n - sige@xcite 2d systems , modified wit respect to dimensional density @xmath19 and temperature @xmath20 depicted in table [ tab : table1 ] . ] following ref.@xcite , the phase transition can be parameterized as it follows : @xmath21 here , the dimensional temperature @xmath22 and 2d density @xmath23 contain the valley splitting factor @xmath24 .
then @xmath25 and @xmath26 is the effective borh radius and rydberg energy respectively , @xmath27 is the effective mass .
note , for certain value of @xmath28 the correct values @xmath29 are lower by a factor of @xmath0 with respect to those predicted in refs.@xcite . for actual 2d systems
the values @xmath30 are generalized in table [ tab : table1 ] . in fig.[fig1
] we plot the melting curve@xcite specified by eq.([wigner_solid ] ) and , moreover , the observed range of 2d densities and temperatures attributed to apparent metal - insulator transition@xcite .
evidently , wigner solidification regime remains unaffected .
hence , we suggest the typical 2d systems can be described within routine fermi gas model . in conclusion , we demonstrate that wigner solidification has been never achieved in experiments dealt with apparent metal to insulator transition .
the observed anomalies of 2d transport behavior is explained within conventional fermi gas formalism invoking the important correction to measured resistivity caused by peltier and seebeck effects combined .
we represent the experimental evidence confirming the solidity and universality of the above model . 100 s.v .
kravchenko et al , phys.rev.b .
* 51 * , 7038 , 1995 v.m .
pudalov et al , phys.rev.b.,*60 * , r2154 , 1999 y. hanein et al , phys.rev.lett . , * 80 * , 1288 , 1998 jian huang et al , phys.rev.b . , * 74 * , 201302(r ) , 2006 x.p.a .
gao et al , phys.rev.lett .
, * 93 * , 256402 , 2004 jian huang et al , phys.rev.b . , * 83 * , 081310(r ) , 2011 m. p.
lilly et al , phys.rev.lett . , * 90 * , 056806 , 2003 k.lai et al , phys.rev.b . , * 72 * , 081313(r ) , 2005 k.lai et al , phys.rev.b . , * 75 * , 033314 , 2007 p.t .
coleridge et al , phys.rev.b . , * 56 * , 12764(r ) , 1997 v. senz et al , ann .
phys.(paris ) , * 8 * , 237 , 1999 p. m. platzman and h. fukuyama , phys.rev.b . ,
* 10 * , 3150 , 1974 b.tanatar and d.m.ceperley , phys.rev.b . , * 39 * , 5005 , 1989 t. ando et al , rev.mod.phys . , * 54 * , 437 , 1982 m.v.cheremisin , physica e , * 27 * , 151 , 2005 c.g.m.kirby , m.j.laubitz , metrologia , * 9 * , 103 , 1973 m. m. radonjic et al , phys.rev.b . ,
* 85 * 085133 , 2012 x.p.a .
gao et al , phys.rev.lett .
, * 94 * , 086402 , 2005 x.p.a .
gao et al , arxiv : cond - mat/0308003 , 2003 n.f .
mott , j. non - cryst .
solids , * 1 * , 1 , 1968 b.i . shklovskii and a.l .
efros , electronic properties of doped semiconductors , springer - verlag , berlin , 1984 . | the melting condition for two - dimensional wigner solid ( p.m. platzman , h.fukuyama , 1974 ) is shown to contain an error of a factor of @xmath0 .
the analysis of experimental data for apparent 2d metal - to - insulator transition shows that the wigner solidification ( b.tanatar , d.m.ceperley , 1989 ) has been never achieved . within routine fermi gas model both the metallic and insulating behavior of different 2d system for actual range of carrier densities and temperatures
is explained .
recently , much interest has been focused on the anomalous transport behavior of a wide variety of low density two - dimensional ( 2d ) systems .
it has been found that below some critical density , cooling causes an increase in resistivity , whereas in the opposite , high - density case , the resistivity decreases .
the apparent metal to insulator transition was observed in n - si mosfet @xcite , p - gaas@xcite , n - gaas@xcite , n - sige@xcite and p - sige @xcite 2d systems . |
supernova explosions of massive stars are powered by the gravitational binding energy that is released when the initial stellar core collapses to a compact remnant and its radius shrinks from typically a few thousand kilometers to little more than ten kilometers . for solar - metallicity progenitors with main - sequence masses of less than
the compact leftover will be neutron star . in case of more massive stars
a black hole will be formed , most likely by the fallback on a timescale of seconds to hours of matter that does not become unbound in the stellar explosion .
but also the direct collapse of the stellar core to a black hole on a multiple of the dynamical timescale is possible ( @xcite ) .
since the collapse proceeds essentially adiabatically the total energy of the stellar core is conserved during the implosion .
the gravitational energy is temporarily stored as internal energy , mainly of degenerate electrons and electron neutrinos .
if rotation plays a significant role in the progenitor core , a major fraction of the potential energy may also be converted to rotational energy of the nascent neutron star ( or black hole ) .
the disruption of the massive star in a supernova explosion now means that some fraction of the energy in these reservoirs has to be transferred from the compact central object to the outer stellar layers .
what are the physical mechanisms to mediate this energy transfer to the ejecta ? and on what timescale
do they work ?
proposals in the literature include the hydrodynamical bounce - shock , neutrinos , or magnetic fields .
the former would initiate the explosion on a dynamical timescale , whereas the latter two can establish the energy transfer only on the secular timescales of neutrino diffusion or magnetic field amplification , respectively .
unfortunately , observations have so far not been able to yield direct insight into the processes in the stellar center at the onset of the explosion .
the hope is that a future galactic supernova will change this situation by allowing the measurements of a large number of neutrinos and possibly of a gravitational wave signal in great detail .
the few neutrino events discovered in connection with supernova 1987a were a clear signal of stellar core collapse and neutron star formation , but they were not sufficient to reveal the secrets of the explosion .
up to now we have to exploit the less direct information that is provided by the supernova light , by the characteristic properties of supernovae and diffuse and compact supernova remnants , and by the nucleosynthesis of heavy elements which takes place in the immediate vicinity of the newly formed neutron star .
section [ sec : observations ] will discuss the constraints for the explosion mechanism that are associated with such observations . in sect .
[ sec : theory ] we shall briefly review the different suggestions that have been brought forward to explain the explosions of massive stars and will critically evaluate our knowledge on grounds of theoretical considerations . in sect .
[ sec : neutrinoexplosions ] we shall summarize the status of detailed hydrodynamical supernova simulations and their implications for our understanding of the delayed explosion mechanism by neutrino - energy deposition behind the supernova shock . in sect .
[ sec : results ] we shall present the first results of a new generation of multi - dimensional supernova models which employ a boltzmann solver for the neutrino transport and a state - of - the - art description of neutrino - matter interactions .
section [ sec : conclusions ] will conclude with an appraisal of the results and an outlook on open ends .
progress in our understanding of the processes that lead to the explosion of massive stars is mainly based on elaborate numerical modeling , supplemented by theoretical analysis and constrained by a growing data base of observed properties of supernovae .
the latter may carry imprints from the physical conditions very close to the center of the explosion .
observable features at very large radii , however , can be linked to the actual energy source of the explosion only indirectly through a variety of intermediate steps and processes .
any interpretation with respect to the mechansim that initiates the explosion therefore requires caution . a viable model for the explosion mechanism of massive stars should ultimately be able to explain the observed explosion energies , nucleosynthetic yields ( in particular of radioactive isotopes like @xmath2ni , which are created near the mass cut ) , and the masses of the compact remnants ( neutron stars or black holes ) and their connection with the progenitor mass .
ni mass as functions of the main sequence mass of the progenitor stars for several supernovae and hypernovae [ from @xcite].,scaledwidth=100.0% ] ni mass as functions of the main sequence mass of the progenitor stars for several supernovae and hypernovae [ from @xcite].,scaledwidth=100.0% ] recent evaluations of photometric and spectroscopic data for samples of well - observed type - ii plateau supernovae reveal a wide continuum of kinetic energies and ejected nickel masses .
faint , low - energy cases seem to be nickel - poor whereas bright , high - energy explosions tend to be nickel - rich and associated with more massive progenitors ( @xcite ) .
this direct correlation between stellar and explosion properties , however , is not apparent in an independent analysis by @xcite who speculates that more than one stellar parameter ( rotation or magnetic fields besides the progenitor and core mass ) might determine the explosion physics .
a large range of nickel masses and explosion energies was also found for type ib / c supernovae ( @xcite ) . interpreting results obtained by the fitting of optical lightcurves and spectra
, @xcite proposed that explosions of stars with main sequence masses above 2025@xmath0m@xmath1 split up to a branch of extraordinarily bright and energetic events ( `` hypernovae '' ) at the one extreme and a branch of faint , low - energy or even `` failed '' supernovae at the other .
stars with such large masses might collapse to black holes rather than neutron stars .
the power of the explosion could depend on the amount of angular momentum in the collapsing core , which in turn can be sensitive to a number of effects such as stellar winds and mass loss , metallicity , magnetic fields , binarity or spiraling - in of a companion star in a binary system .
another constraint for supernova theory comes from measured neutron star and black hole masses . a direct mass determination may not be possible for compact remnants inside supernova explosions , but the combined information about progenitor ( ejecta ) mass , kinetic energy , and nickel mass of a supernova ( as deduced from spectroscopic and photometric data ) may be used to determine the mass cut in a progenitor model by reproducing the known parameters with numerical simulations .
masses of binary pulsars ( @xcite ) and x - ray binaries ( @xcite ) also add to our knowledge about compact supernova remnants . in this case , however , binary effects make it even more difficult to establish a link to progenitor properties and therefore conclusions on the supernova mechanism should be drawn with great caution .
complete and incomplete silicon and oxygen burning behind the supernova shock depend on the strength of the shock and the structure of the progenitor .
the abundances of corresponding nucleosynthetic products , if observed in an individual supernova or supernova remnant , can therefore yield information about the exploding star and the explosion energy .
their spatial distribution can also reveal anisotropies and mixing phenomena during the explosion as in case of , e.g. , supernova 1987a or the vela supernova remnant . another serious constraint for the explosion mechanism and the conditions near
the mass cut comes from a comparison of integral yields of supernova nucleosynthesis with galactic abundances .
the ejected mass of closed - neutron shell isotopes with @xmath3 of sr , y , and zr , for example , is limited to less than about @xmath4m@xmath1 per event [ @xcite ] .
these stable nuclei are easily created through the capturing of @xmath5-particles in the neutron - rich matter around the mass cut and thus experience conditions in the close vicinity of the neutron star .
the crucial parameters are entropy and neutron excess . both are determined by the exposure to high neutrino fluxes , because neutrinos exchange energy and lepton number with the stellar medium .
anisotropic processes and large - scale mixing between the deep interior and the hydrogen layer had to be invoked in case of supernova 1987a to explain the shape of the lightcurve , the unexpectedly early appearance of x - ray and @xmath6-ray emission , and doppler features of spectral lines ( for a review , see @xcite ) .
fifteen years after the explosion , the expanding debris exhibits an axially symmetric deformation .
supernova 1987a therefore seems to possess an intrinsic , global asymmetry .
the same conclusion was drawn for other core - collapse supernovae ( type - ii as well as ib / c ) based on the fact that their light is linearly polarized at a level around 1% with a tendency to increase at later phases when greater depths are observed ( @xcite ) .
this has been interpreted as evidence that the inner portions of the explosion , and hence the mechanism itself , are strongly non - spherical ( @xcite ) , possibly associated with a `` jet - induced '' explosion ( @xcite ) .
this is a very interesting and potentially relevant conjecture .
it does , however , not necessarily constrain the nature of the physical process that mediates the energy transfer from the collapsed core of the star to the ejecta and thereby creates the asphericity .
, @xmath7 mode in the convective pattern .
the snapshots ( from top to bottom ) show the entropy distribution ( values between about 4 and 23@xmath8 per nucleon ) at post - bounce times @xmath9ms , 415@xmath0ms , 675@xmath0ms , and 1000@xmath0ms .
note that the radial scales of the figures differ .
the neutron star is at the origin of the axially symmetric ( 2d ) grid and plays the role of an isotropic neutrino `` light bulb '' ( plewa et al .
, in preparation).,scaledwidth=82.0% ] , @xmath7 mode in the convective pattern .
the snapshots ( from top to bottom ) show the entropy distribution ( values between about 4 and 23@xmath8 per nucleon ) at post - bounce times @xmath9ms , 415@xmath0ms , 675@xmath0ms , and 1000@xmath0ms .
note that the radial scales of the figures differ .
the neutron star is at the origin of the axially symmetric ( 2d ) grid and plays the role of an isotropic neutrino `` light bulb '' ( plewa et al .
, in preparation).,scaledwidth=82.0% ] , @xmath7 mode in the convective pattern .
the snapshots ( from top to bottom ) show the entropy distribution ( values between about 4 and 23@xmath8 per nucleon ) at post - bounce times @xmath9ms , 415@xmath0ms , 675@xmath0ms , and 1000@xmath0ms .
note that the radial scales of the figures differ .
the neutron star is at the origin of the axially symmetric ( 2d ) grid and plays the role of an isotropic neutrino `` light bulb '' ( plewa et al .
, in preparation).,scaledwidth=82.0% ] , @xmath7 mode in the convective pattern .
the snapshots ( from top to bottom ) show the entropy distribution ( values between about 4 and 23@xmath8 per nucleon ) at post - bounce times @xmath9ms , 415@xmath0ms , 675@xmath0ms , and 1000@xmath0ms .
note that the radial scales of the figures differ .
the neutron star is at the origin of the axially symmetric ( 2d ) grid and plays the role of an isotropic neutrino `` light bulb '' ( plewa et al .
, in preparation).,scaledwidth=82.0% ] the high space velocities of young pulsars with a mean around 400@xmath0km / s ( @xcite ) have also been used to argue for an intrinsically asymmetric explosion mechanism .
it is very likely that the neutron stars obtained their recoil at the onset of the explosion or shortly afterwards ( for a review , see @xcite ) .
but it is not clear whether the mechanism and the anisotropies of the explosion are directly connected with the pulsar kicks .
also a possible `` jet '' and perhaps a `` counter jet '' in the gaseous filaments of the cas a supernova remnant may neither be linked to the process that started the supernova explosion nor to the runaway velocity of the compact remnant .
the latter could have been accelerated ( maybe by anisotropic neutrino emission ) _
after _ the explosion was launched . the jet - like features
if at all caused by processes at the supernova center and not by local inhomogeneities at the forward shock may have been produced even later , for example by bipolar outflow from a neutron star or a black hole that has accreted from a disk of fallback matter .
interestingly , inspecting the images given by @xcite , the dislocation of the compact source relative to the geometrical center of the remnant appears not to coincide with the direction of the possible `` jets '' ( compare also @xcite ) .
rotation plus magnetic fields were proposed as the `` most obvious '' way to break the spherical symmetry and to explain the global asphericity of core - collapse supernovae ( @xcite ) .
it was argued that current numerical calculations may be missing a major ingredient necessary to yield explosions .
a proper treatment of rotation _ and _ magnetic fields may be necessary to fully understand when and how stellar collapse leads to an explosion .
of course , this might be true .
but a confirmation or rejection will require computer models with ultimately the full physics .
it must be stressed , however , that current observations do not necessitate such conclusions and hydrodynamical simulations suggest other possible explanations .
strong convection in the neutrino - heating region behind the supernova shock can account for huge anisotopies of the inner supernova ejecta , even without invoking rotation .
if the explosion occurs quickly , much power remains on smaller scales until the expansion sets in and the convective pattern gets frozen in . if , in contrast , the shock radius grows only very slowly and the explosion is delayed for several 100@xmath0ms after bounce , the convective flow can merge to increasingly larger structures . in two - dimensional ( 2d ) hydrodynamic calculations including cooling and heating by neutrinos between the neutron star and the shock ( with parameter choices for a central , _ isotropic _ neutrino `` light bulb '' which enabled explosions ) ,
we ( plewa et al . , in preparation ) found situations where the convective pattern revealed a contribution of the @xmath7 , @xmath10 mode that was growing with time and was even dominant at about one second after bounce ( fig .
[ fig : monomode ] ) . @xcite already speculated about such a possibility .
certainly three - dimensional ( 3d ) calculations of the full sphere ( and without the coordinate singularity on the axis of the spherical grid ) are indispensable to convincingly demonstrate the existence of this phenomenon .
the primary energy source for powering supernovae of massive stars is the gravitational binding energy of the newly formed proto - neutron star or proto - black hole ( energy from nuclear reactions contributes at a minor level ) . to initiate and drive the explosion
, energy has to be transferred to the stellar matter above the mass cut to be finally converted to kinetic energy of the ejecta .
this could be achieved by hydrodynamical shocks , neutrinos , or magnetic fields as mediators .
collimated outflows or jets of relativistic plasma might also play a role for very special circumstances .
depending on the mediator , the conditions for efficient energy transfer , the corresponding timescale , and the tapped energy reservoir are different .
a lot of work has been spent in the past 40 years to identify viable supernova mechanisms and to study the involved physics .
the space here is not sufficient to review all important papers .
we shall therefore only focus on some key contributions without attempting completeness .
when the homologously collapsing inner part of the stellar core is abruptly stopped at core bounce by the stiffening of the nuclear equation of state ( eos ) , a hydrodynamical shock is formed at the boundary to the supersonically infalling outer part ( @xcite ) . in this case
the initial energy of the shock is equal to the kinetic energy of the inner core `` at the last moment of good homology '' ( @xcite ) : @xmath11 when @xmath12 is the mass of the homologously collapsing part of the stellar core and @xmath13 its maximum infall velocity .
there is general agreement now , supported by a large number of numerical calculations ( e.g. , @xcite ) , that this energy is not sufficient for the shock to reach the surface of the iron core : the energy losses from behind the shock by photodisintegration of nuclei to @xmath5-particles and free nucleons ( @xmath14erg per 0.1@xmath0m@xmath1 ) as well as neutrino escape after shock breakout through the neutrinosphere ( some @xmath15erg ) are so severe that the shock stalls within milliseconds . only in case of extremely small iron cores ( @xmath16m@xmath1;@xcite ) or an extraordinarily soft nuclear eos ( @xcite )
is the prompt hydrodynamical shock able to disrupt the star .
nevertheless , the hope that the prompt hydrodynamical bounce - shock mechanism might work for some stars in a mass window around 10@xmath0m@xmath1 has not been given up completely ( @xcite ) .
soon after the discovery of pulsars the role of rotation and magnetic fields in the supernova explosion was scrutinized ( @xcite ) . because of flux conservation a seed field in the stellar core can grow significantly during collapse , although for realistic initial fields not to a strength that magnetic pressure could be dynamically important . but winding of field lines in the case of differential rotation , which is natural after the collapse of a spinning core , can further amplify the toroidal field component .
if the magnetic pressure becomes comparable to the thermal pressure , magnetohydrodynamical forces can drive an explosion ( @xcite ) , and might accelerate axial jets ( @xcite ) . also
buoyancy instabilities of highly magnetized matter could produce mass motions ( @xcite ) . in this case
rotational energy of the spinning proto - neutron star is converted to magnetic energy , a part of which might then end up as kinetic energy of the ejecta .
hence , @xmath17 when @xmath18 is the spin period of the forming neutron star with mass @xmath19 . the field , however , grows linearly with time and therefore the timescale to reach sufficiently strong fields is long compared to the accretion timescale of the collapsed core , unless unrealistically large initial values are assumed ( @xcite ) .
recently it was suggested that the magnetorotational instability ( mri ) , which was discussed in the context of accretion disks by @xcite , might also be in action in the supernova core ( @xcite ) .
because of an exponential growth , this would reduce the field amplification timescale drastically and would allow for much smaller initial seed fields , too .
it remains to be shown , however , whether the field can grow to a dynamically effective strength in a self - consistent model for reasonable assumptions about the rotation of stellar cores and despite of diverse effects that can lead to field saturation or hinder field growth .
moreover , it is unclear whether axial jets can be initiated by the mri at the conditions in collapsing stellar cores .
implications for supernova nucleosynthesis are unexplored , but might be problematic if mhd explosions are frequent ( @xcite ) .
the amplification and transport of magnetic fields in convective regions inside the nascent neutron star ( @xcite ) and in the neutrino - heating layer behind the supernova shock ( @xcite ) also deserve further attention with respect to the generation of pulsar fields . while neutrinos drain energy from the shock - heated matter immediately after bounce , the situation changes somewhat later when the density and temperature behind the shock have dropped .
high - energy neutrinos , emitted abundantly from the hot accretion layer that covers the nascent neutron star , are absorbed with a small , but finite probability ( typically @xmath20% ) in the postshock gas .
if this energy deposition is large enough , the stalled supernova shock can be revived . since this can happen on a timescale of hundred or more milliseconds after bounce , the explosion is called `` delayed '' in contrast to the prompt bounce - shock mechanism ( @xcite ) .
the idea that the energy transfer to the ejecta is mediated by neutrinos goes back to ( @xcite ) . during core collapse and subsequent contraction ,
the gravitational binding energy of the forming neutron star is first stored as internal energy ( degeneracy energy of leptons and thermal energy of nucleons ) and only slowly carried away by neutrinos on the diffusion timescale of several seconds .
the kinetic energy required for the supernova explosion is of order @xmath15erg and hence tiny compared to the huge reservoir of binding energy , @xmath21 where the second expression results from the first by using @xmath22 .
but just a smaller fraction ( about 10% ) of this energy is actually radiated in neutrinos during the first hundreds of milliseconds until the explosion starts .
neutrino heating behind the shock is essentially _
unavoidable_. the temperature in this region drops roughly like @xmath23 ( close to isentropic conditions ) , which means that the energy loss rate of the stellar plasma by neutrinos produced in @xmath24 captures on protons and neutrons falls like @xmath25 .
in contrast , the energy deposition by @xmath26-absorption on neutrons and @xmath27-absorption on protons decreases much more slowly .
the heating rate varies roughly proportional to the neutrino energy density ( which scales with the summed luminosity of @xmath26 plus @xmath27 , @xmath28 , divided by the square of the inverse distance @xmath29 from the neutrinosphere ) and the mean squared energy @xmath30 ( because of the dependence of the absorption cross section on the neutrino energy @xmath31 ) , i.e. , @xmath32 .
this means that there must be a radial position , the so - called `` gain radius '' ( @xcite ) , exterior of which neutrino heating dominates neutrino cooling .
the heating rate for matter that is disintegrated to free neutrons and protons can be estimated as @xmath33 per nucleon for a typical luminosity of @xmath26 plus @xmath27 of @xmath34erg / s , 15@xmath0mev for the rms neutrino energy , and @xmath29 measured in units of @xmath35 cm .
the explosion energy should scale as @xmath36 where @xmath37 is the neutrino - heated mass , @xmath38 the nucleon mass , and @xmath39 the timescale of neutrino - energy deposition .
the latter can be assumed to be roughly the time until nucleons have absorbed an energy comparable to their binding energy in the gravitational potential of the neutron star , i.e. , @xmath40 for @xmath41m@xmath1 .
these arguments suggest that the gradual ( in contrast to impulsive ) energy deposition by neutrinos might not be responsible for a dominant fraction of the net explosion energy of a supernova .
instead , neutrino heating helps to compensate the gravitational binding and thus triggers the outward expansion of the supernova ejecta .
the energy per baryon increases further by ongoing energy transfer from neutrinos to the accelerating matter , and finally reaches its limiting value due to the energy release in recombinations of free nucleons to nuclei .
this recombination energy may account for a major part of the supernova energy .
numerical simulations and analytic considerations ( @xcite ) suggest a mass @xmath37 between a few @xmath42m@xmath1 and about @xmath43m@xmath1 . with a typical energy deposited by neutrino heating being around 5@xmath0mev per nucleon ( @xcite )
, the explosion energy could range between some @xmath44erg and several @xmath15erg , when about 8@xmath0mev are set free for each nucleon that recombines .
these numbers naturally match the energies of observed normal supernovae ( cf .
[ sec : observations ] ) .
stars with main sequence masses beyond 20@xmath45m@xmath1 seem to be associated with much more powerful explosions with energies up to several @xmath46erg ( sect . [
sec : observations ] ) .
such energies are probably out of reach for the neutrino - heating mechanism as described above .
the idea has been coined ( @xcite ) that the core of such stars collapses to a black hole , which then continues to accrete the infalling matter of the progenitor star ( `` collapsar '' or `` failed supernova '' ; @xcite ) .
provided that the stellar core has retained enough angular momentum during its pre - collapse evolution ( or was spun up by a merger with a binary companion ) , the accretion will proceed at very high rates through a disk or torus on the timescale of viscous angular momentum transport , which is much longer than the dynamical timescale .
therefore the efficiency of energy release can in principle be much higher than in case of spherical accretion . in terms of the rest - mass energy of the accreted matter
one can get @xmath47 with maximum values between @xmath48 for a non - rotating black hole and @xmath49 for an extreme kerr hole that accretes from a corotating thin disk ( @xcite ) .
since the mass accretion rates are hypercritical and the densities in the disk correspondingly high , photons can not escape and the energy is set free in form of neutrinos rather than electromagnetic radiation ( @xcite ) .
the efficiency may be reduced compared to the quoted maximum values if the density and temperature are too low for rapid neutrino production or the disk is very dense and thus becomes nontransparent to neutrinos . in these cases
the energy is advected into the black hole by the accretion flow instead of being carried away by neutrinos ( @xcite ) .
alternatively , magnetohydrodynamical processes may drive outflows from the disk ( e.g. , @xcite ) .
it is also possible that the rotational energy of the spinning black hole is extracted via the blandford and znajek mechanism ( @xcite ) through magnetic fields that are anchored in the black hole and couple it to the surrounding disk . by the annihilation of neutrinos and antineutrinos to @xmath24-pairs ( @xcite ) or the mentioned mhd processes ,
polar jets could be launched to propagate through the star along the rotation axis ( @xcite ) . in case
the jets remain collimated and highly relativistic until they break out of the stellar surface ( which requires that the star does not possess an extended hydrogen envelope ; @xcite ) gamma - ray bursts ( grbs ) may be triggered .
nonrelativistic jets that expand laterally and sweep up the surrounding progenitor gas , or energetic winds from the accretion torus might be the driving force behind enormously powerful stellar explosions ( @xcite ) which are observed as so - called `` hypernovae '' and grb - supernovae ( @xcite ) .
for an overview and more information , see @xcite .
the different mechanisms for massive star explosions as described above require different properties of the progenitor stars and different physical conditions in the collapsed stellar cores .
the latter are incompletely known , but some of the requirements are more likely fulfilled than others , some combinations of necessary conditions may be more common and more typical , while others might be realized only in rare cases and for very special , exceptional circumstances .
it is possible , if not probable , that the diversity of the observed explosions of massive stars and the large number of associated phenomena mean that more than one mechanism is at work .
it may also be , however , that the majority of ordinary supernovae can be ascribed to the same central process which might then depend extremely sensitively on yet to be determined parameters .
the neutrino - driven mechanism ( @xcite ) involves a minimum of controversial assumptions and uncertain degrees of freedom in the physics of collapsing stars .
it relies on the importance of neutrinos and their energetic dominance in the supernova core . after the detection of neutrinos in connection with supernova 1987a and the overall confirmation of theoretical expectations for the neutrino emission ,
this is not a speculation any more but an established fact . of course
, this does not mean that such a minimal input is sufficient to understand the cause of supernova explosions and to explain all observable properties of supernovae .
but at least it can be taken as a good reason to investigate how far one can advance with a minimum of imponderabilities .
detailed numerical simulations are needed to answer the crucial question in this context : is neutrino - energy deposition efficient enough to drive the explosion ?
spherically symmetric simulations with the current input physics ( neutrino interactions and the equation of state of dense matter ) do not yield explosions by the neutrino - heating mechanism .
there is no controversy about that .
all computations are in agreement , independent of newtonian or relativistic gravity and independent of the neutrino transport being treated in an approximate way by flux - limited diffusion methods ( e.g. , @xcite ) or very elaborately by solving the frequency- and angle - dependent boltzmann transport equation ( @xcite ) .
whether neutrinos succeed in reviving the stalled shock depends on the efficiency of the energy transfer to the postshock layer , which in turn increases with the neutrino luminosity and the hardness of the neutrino spectrum .
wilson and collaborators ( @xcite ) have obtained explosions in one - dimensional ( 1d ) simulations for more than ten years now . in these models
it is , however , _ assumed _ that neutron - finger convection in the hot neutron star boosts the neutrino luminosities .
moreover , a special equation of state with a high abundance of pions in the nuclear matter was used ( @xcite ) , which again leads to higher neutrino fluxes from the neutron star and thus to enhanced energy - deposition behind the shock .
both assumptions are not generally accepted .
two - dimensional ( @xcite ) and 3d simulations ( @xcite ) as well as theoretical considerations ( @xcite ) have shown that the neutrino - heating layer is unstable to convective overturn .
the associated effects support shock revival and can lead to explosions even in cases where spherical models fail . in the multi - dimensional situation accretion and expansion
can occur simultaneously .
downflows of cooler , low - entropy matter that has fallen through the shock , coexist with rising bubbles of high - entropy , neutrino - heated gas ( fig .
[ fig : monomode ] ) . on the one hand ,
the downflows carry cool material close to the gain radius where it absorbs energy readily from the intense neutrino fluxes . on the other hand ,
heated matter rises in bubbles and can expand and cool quickly , which reduces the energy loss by the reemission of neutrinos .
this also increases the postshock pressure and hence pushes the shock farther out .
the gain layer grows and thus more gas can accumulate in the neutrino - heating region .
the gas also stays longer in the gain layer , in contrast to one - dimensional models where the matter behind the accretion shock has negative velocity and is quickly advected down to the cooling layer .
when the gas arrives there , neutrino emission sets in and extracts again the energy which had been absorbed from neutrino heating shortly before . due to the combination of all these effects postshock convection enhances the efficiency of the neutrino - heating mechanism .
therefore the multi - dimensional situation is _ generically different _ from the spherically symmetric case .
nevertheless , the existence of convective overturn in the neutrino - heating layer does not guarantee explosions ( @xcite ) . for insufficient neutrino heating the threshold to an explosion will not be overcome .
since neutrinos play a crucial role , an accurate description of the neutrino physics transport and neutrino - matter interactions is indispensable to obtain conclusive results about the viability of the neutrino - driven mechanism .
all previously published multi - dimensional explosion models , however , have employed some crude approximations or simplifications in the treatment of neutrinos .
in order to take a next step of improvement in supernova modelling , we have coupled a new boltzmann solver for the neutrino transport to the prometheus hydrodynamics code .
the combined program is called mudbath ( * mu*lti - imensional * b*oltzm*a*nn * t*ransport and * h*ydrodynamics ) and allows for spherically symmetric as well as multi - dimensional simulations ( @xcite ) .
below we present some results of 1d and our first 2d supernova simulations with this new code . for the integration of the equations of hydrodynamics we employ the newtonian finite - volume code prometheus ( @xcite ) , which was supplemented by additional problem specific features ( @xcite ) .
prometheus is a direct eulerian , time - explicit implementation of the piecewise parabolic method ( ppm ) of @xcite . as a second - order godunov scheme employing a riemann solver it is particularly well suited for following discontinuities in the fluid flow like shocks or boundaries between layers of different chemical composition .
a notable advantage in the present context is its capability of tackling multi - dimensional problems with high computational efficiency and numerical accuracy .
our code makes use of the `` consistent multifluid advection ( cma ) '' method ( @xcite ) for ensuring accurate advection of different chemical components in the fluid , and switches from the original ppm method to the more diffusive hlle solver of @xcite in the vicinity of strong shocks to avoid spurious oscillations ( the so - called `` odd - even decoupling '' phenomenon ) when such shocks are aligned with one of the coordinate lines in multidimensional simulations ( @xcite ) . the boltzmann solver scheme is described in much detail elsewhere .
the integro - differential character of the boltzmann equation is tamed by applying a variable eddington factor closure to the neutrino energy and momentum equations ( and the simultaneously integrated first and second order moment equations for neutrino number ) . for this purpose
the variable eddington factor is determined from the solution of the boltzmann equation , and the system of boltzmann equation and its moment equations is iterated until convergence is achieved . employing this scheme in multi - dimensional simulations in spherical coordinates
, we solve the radius- ( and energy- ) dependent moment equations on the different angular bins of the numerical grid but calculate the variable eddington factor only once on an angularly averaged stellar background .
this approximation is good only for situations without significant global deformations .
since the iteration of the boltzmann equation has to be done only once per time step , appreciable amounts of computer time can be saved ( @xcite ) .
we point out here that it turned out to be necessary to go an important step beyond this simple `` ray - by - ray '' approach .
physical constraints , namely the conservation of lepton number and entropy within adiabatically moving fluid elements , and numerical requirements , i.e. , the stability of regions which should not develop convection according to a mechanical stability analysis , make it necessary to take into account the coupling of neighbouring rays at least by lateral advection terms and neutrino pressure gradients ( buras et al . , in preparation ) .
general relativistic effects are treated only approximately in our code ( @xcite ) .
the current version contains a modification of the gravitational potential by including correction terms due to pressure and energy of the stellar medium and neutrinos , which are deduced from a comparison of the newtonian and relativistic equations of motion .
the neutrino transport contains gravitational redshift and time dilation , but ignores the distinction between coordinate radius and proper radius .
this simplification is necessary for coupling the transport code to our basically newtonian hydrodynamics .
although a fully relativistic treatment would be preferable , tests showed that these approximations seem to work satisfactorily well ( liebendrfer et al .
, in preparation ) , at least as long as there are only moderate ( @xmath501020% ) deviations of the metric coefficients from unity and the infall velocities do not reach more than 1020% of the speed of light in decisive phases of the evolution . as for the neutrino - matter interactions
, we discriminate between two different sets of input physics . on the one hand
we have calculated models with conventional ( `` standard '' ) neutrino opacities , i.e. , a description of the neutrino interactions which follows closely the one used by bruenn and mezzacappa and collaborators ( @xcite ) .
it assumes nucleons to be uncorrelated , infinitely massive scattering targets for neutrinos . in these reference
runs we have usually also added neutrino pair creation and annihilation by nucleon - nucleon bremsstrahlung ( @xcite ) .
details of our implementation of these neutrino processes can be found elsewhere ( @xcite ) .
a second set of models was computed with an improved description of neutrino - matter interactions . besides including nucleon thermal motions and
recoil , which means a detailed treatment of the reaction kinematics and allows for an accurate evaluation of nucleon phase - space blocking effects , we take into account nucleon - nucleon correlations ( following @xcite ) , the reduction of the nucleon effective mass , and the possible quenching of the axial - vector coupling in nuclear matter ( @xcite ) .
in addition , we have implemented weak - magnetism corrections as described by @xcite .
the sample of neutrino processes was enlarged by also including scatterings of muon and tau neutrinos and antineutrinos off electron neutrinos and antineutrinos and pair annihilation reactions between neutrinos of different flavors [ i.e. , @xmath51 ; @xcite ] .
our current supernova models are calculated with the nuclear equation of state of @xcite using an incompressibility modulus of @xmath52mev ( other values of the incompressibility of bulk nuclear matter in the lattimer & swesty eos cause only minor differences , see @xcite ) . for the density regime below @xmath53g/@xmath54
we switch to an equation of state that considers electrons , positrons and photons , and nucleons and nuclei with an approximative treatment for composition changes due to nuclear burning and shifts of nuclear statistical equilibrium ( @xcite ) .
.input physics for our set of computed models .
see sect .
[ sec : code ] for details .
[ cols= " < , < , < , < , < , < , < " , ] @xmath55 angular wedge of the spherical coordinate grid around the equatorial plane .
+ @xmath56 calculation without neutrino - pair creation by nucleon - nucleon bremsstrahlung .
+ @xmath57 calculation without the neutrino - antineutrino processes of @xcite .
[ table:1 ] -axis is assumed , and periodic conditions are used at the lateral grid boundaries.,scaledwidth=100.0% ] we have performed a number of core - collapse simulations in spherical symmetry and then followed the post - bounce evolution in one and two dimensions .
all described calculations were started from a 15@xmath0m@xmath1 progenitor star , model s15s7b2 , provided to us by s. woosley . adopting the naming
, we label our models by s15n for newtonian runs and s15 g for runs with approximate treatment of general relativity , followed by letters `` so '' when `` standard neutrino opacities '' were used and by `` io '' in case of our state - of - the - art improvement of the description of neutrino - matter interactions .
the model names have a suffix that discriminates between 1d ( `` _ 1d '' ) and 2d simulations ( `` _ 2d '' ) . for the 2d models we used a spherical coordinate grid with 20 or 32 equidistant zones within an angular wedge from @xmath58 to @xmath59 or from @xmath60 to @xmath61 , respectively , around the equatorial plane and assumed periodic conditions at the boundaries ( fig .
[ fig : grid ] ) .
we have varied yet another aspect in our simulations .
some of the models were computed with a version of the transport code where the velocity dependent ( doppler shift and aberration ) terms in the neutrino momentum equation ( and the corresponding terms in the boltzmann equation for the antisymmetric average of the specific intensity ; see @xcite ) were omitted .
these terms are formally of order @xmath62 and should be small for low velocities .
this simplification of the neutrino transport , our so - called `` case a '' , was used in the models which have names ending with the letter `` a '' .
models with neutrino transport including all velocity dependent terms also in the neutrino momentum equation ( `` case b '' ) can be identified by the ending `` b '' of their names .
table [ table:1 ] provides an overview of the computed models .
the shock trajectories and corresponding total neutrino luminosities of all models are displayed in fig .
[ fig : shockradii ] .
no explosions , neither with the standard nor with the improved description of the neutrino opacities , were obtained with the most complete implementation of the transport equations , case b ( see lines labeled with m1 , h1 , h2 , s1 , b1 , corresponding to models s15nso_1d.b , s15gio_1d.b , s15gio_2d.b , s15gso_1d.b and s15gso_1d.b@xmath63 , respectively ) .
the shock trajectories of this sample of models form a cluster that is clearly separated from the models computed with the transport version of case a , which generally show a larger shock radius and therefore more optimistic conditions for explosions .
the simplification of the neutrino transport in case a has a remarkable consequence which is obvious from a comparison of the 2d runs of models s15gio_2d.a and s15gio_2d.b , which both were performed with our approximation of relativistic effects and the state - of - the - art improvement of neutrino - matter interactions ( cf .
[ sec : code ] ) .
while model s15gio_2d.b fails to explode , the stalled shock in model s15gio_2d.a is successfully revived by neutrino heating because very strong convection can develop in the gain region .
the time evolution of both models is displayed by the trajectories of mass shells in fig .
[ fig : massshells ] . the reason for this dramatic difference can be understood in the following way .
some of the velocity dependent terms ( those in which derivatives with respect to the neutrino energy do not show up ) in the neutrino momentum equation have a simple formal interpretation : in regions with mass infall ( _ negative _ velocity ) they effectively act like a _ reduction _ of the neutrino - medium interaction on the right hand side of this equation .
the corresponding changes can be 10% or more for neutrino energies in the peak of the spectrum , depending also on time , radius , and the size of the postshock velocities . as a consequence of the omission of these terms in model s15gio_2d.a , neutrinos
are impeded in their streaming .
therefore the comoving - frame neutrino ( energy ) density is increased ( see fig . [
fig : compare ] ) .
this is associated with slightly lower neutrino losses in the cooling layer around the neutrinosphere and a significantly enhanced neutrino heating between gain radius and shock .
conversely , the inclusion of the velocity - dependent terms in the neutrino momentum equation has negative consequences for the shock propagation in model s15gio_2d.b .
although the differences at early post - bounce times are moderate ( 1030% , depending on the quantity ) the accumulating effects during the first 80@xmath0ms after bounce clearly damp the shock expansion and finally lead to a dramatic shock recession after the initial phase of expansion .
before this happens postshock convection has not become strong enough to change the evolution . with the onset of contraction ,
the postshock velocities decrease ( become more negative ) quickly , neutrino - heated matter is rapidly advected down through the gain radius and loses its energy by the reemission of neutrinos .
the larger negative values of the postshock velocity enhance the influence of the velocity - dependent terms and reduce the heating efficiency even more . because of this disastrous feedback , the gain region shrinks to a very narrow layer , a fact which suppresses the convective activity lateron .
this is demonstrated by figs .
[ fig : snapshots1 ] and [ fig : snapshots2 ] , which show that the convection is weak in model s15gio_2d.b but very strong in model s15gio_2d.a . due to a combination of unfavorable effects and a continuously amplifying negative trend ,
model s15gio_2d.b remains below the explosion threshold while model s15gio_2d.a is just above that critical limit . in the vicinity of the threshold the long - time evolution of the collapsing stellar core
therefore depends very sensitively on `` smaller details '' of the neutrino transport .
the two simulations also demonstrate that a sufficiently large shock radius for a sufficiently long time is crucial for the growth of convective overturn in the neutrino - heating layer .
only when convection behind the shock becomes strong it can be decisive for getting an explosion .
this latter fact is known from previous multi - dimensional simulations ( @xcite ) and is again confirmed by a comparison of model s15gio_2d.a with the corresponding one - dimensional simulation , model s15gio_1d.a .
it is interesting that the newtonian 2d run , model s15nso_2d.a ( with standard opacities ) does marginally fail , whereas model s15gio_2d.a with relativistic corrections ( and state - of - the - art neutrino reactions ) succeeds .
the influence of the neutrino opacities can be directly seen by comparing models s15gso_1d.b and s15gio_1d.b .
it is not dramatic , but sufficient to justify the inclusion of the improvements described in sect .
[ sec : code ] .
neutrino - pair creation by bremsstrahlung makes a minor difference during the considered phases of the evolution ( model s15gso_1d.b@xmath63 vs. model s15gso_1d.b ) .
we also found that ledoux convection sets in _ below _
the neutrinosphere already shortly after bounce ( fig .
[ fig : nsconvection ] ) .
it is persistent until the end of our simulations and slowly digs farther into the star as already found in the simulations of @xcite .
but the convective activity is so deep inside the neutron star that its effects on the @xmath26 and @xmath27 luminosities and on the supernova dynamics are insignificant . for model s15gio_2d.a at 436@xmath0ms after bounce .
the total mass of @xmath2ni is about 0.1@xmath0m@xmath1.,scaledwidth=100.0% ] for model s15gio_2d.a at 436@xmath0ms after bounce .
the total mass of @xmath2ni is about 0.1@xmath0m@xmath1.,scaledwidth=100.0% ]
we believe that our 2d models with a boltzmann solver for the neutrino transport and a state - of - the - art description of neutrino - matter interactions have considerably reduced the uncertainties associated with the treatment of the neutrino physics in previous multi - dimensional simulations . with the most complete implementation of the transport physics
we could not obtain explosions .
this result suggests that the neutrino - driven mechanism fails with the employed input physics , at least in case of the considered 15@xmath0m@xmath1 star .
we do not think that the approximate treatment of general relativistic effects is likely to jeopardize this conclusion . a comparison with fully relativistic one - dimensional calculations ( liebendrfer et al .
, in preparation ) is very encouraging . because of the remarkable similarity of the shock trajectories of different progenitors in spherical symmetry ( @xcite ) , it is probable that our negative conclusion is also valid for other pre - collapse configurations with a similar structure .
significant star - to - star variations of the progenitor properties with a non - monotonic dependence on the stellar mass ( @xcite ) , however , suggest that multi - dimensional core - collapse simulations of a larger sample of stars are needed before one can make final , more generally valid statements
. another concern may be the omission of lateral neutrino fluxes in our multi - dimensional transport approximation .
we describe the neutrino losses from the hot spots that appear at locations where the low - entropy downflows plunge into the neutrinospheric layer , by just a radial flux . in reality
the neutrinos produced in this semi - transparent matter would be radiated away more isotropically , a fact which could lead to a larger reabsorption probability in the surrounding medium .
the corresponding increase of the heating efficiency might not be very large , but could still be relevant because of the close proximity of our models to an explosion .
the supernova problem is highly nonlinear and surprises may lurk behind every corner
. it would therefore be premature to conclude that the neutrino - driven mechanism fails and that not even postshock convection can alter this unquestioned outcome of all current spherical models .
besides studying other progenitors with multi - dimensional simulations , one should also investigate the effects of rotation during the neutrino - heating phase and long - time post - bounce evolution of a supernova .
previous simulations found weaker explosions in case of rotating models ( @xcite ) .
but these calculations were performed with a much simplified treatment of the neutrinos and showed powerful , nearly prompt explosions already in the absence of rotation .
it can not be excluded that even a moderate amount of rotation might have significant consequences when added to nonrotating models which explode only with a long time delay after core bounce or even fail to do so .
naturally , three - dimensional simulations are ultimately desirable and could reveal generic differences compared to 2d , in particular in combination with rotation .
the recent results of , which suggest a large degree of similarity between 2d and 3d results , are groundbreaking but again must be interpreted with caution .
the serious approximations of the neutrino physics and the rapid development of the explosion in these models do not allow too far - reaching conclusions .
another major uncertainty of current supernova models is associated with the incompletely known physics in the nuclear and supranuclear medium . not only does the high - density eos affect the core bounce , shock formation , and prompt shock propagation
, it also determines the internal properties of the hot neutron star and thus its neutrino emission , both indirectly and directly .
hadronic degrees of freedom in addition to nucleons in the supranuclear matter , e.g. , hyperons , pions or kaons , can reduce the stiffness of the eos at very high densities . the possible effects on the supernova explosion
are essentially unexplored .
a softening of the eos can , for example , lead to a faster contraction and heating of the nascent neutron star .
this in turn also determines the density and temperature profile in the vicinity of the neutrinosphere where most of the neutrino flux is produced during the crucial phase of delayed shock revival .
our successfully exploding 2d model , model s15gio_2d.a , at least demonstrates that simulations which include the effects of postshock convection are rather close to an explosion .
therefore modest changes of the neutrino emission and transport ( in our case the omission of the velocity - dependent terms in the neutrino - momentum equation ) seem to be already sufficient to push them beyond the critical threshold .
the properties of the explosion in this case are very encouraging and may support one s belief in the basic viability of the delayed explosion mechanism . at 440@xmath0ms after bounce the shock has arrived at a radius of more than 3000@xmath0 km and is expanding with about 10000@xmath0km / s ( fig . [ fig : shockradii ] ) .
the explosion of this model does not become very energetic ( fig .
[ fig : explosionenergy ] ) .
the energy is just @xmath64erg at that time and grows only very slowly .
this may not be a serious problem if one recalls the large spread of energies of observed supernovae ( supernova 1999br , for example , is estimated to have an ejecta mass of @xmath65m@xmath1 and an explosion energy of about @xmath66erg ; @xcite ) .
moreover , the explosion energy is a very steep function of the physical conditions in models which are above the explosion threshold ( @xcite ) .
a significant increase of the energy may therefore not require large changes .
since the explosion starts rather late ( at @xmath67ms post bounce ) , the proto - neutron star has accreted enough matter to have attained an initial baryonic mass of 1.4@xmath0m@xmath1 .
therefore our simulation does not exhibit the problem of previous successful multi - dimensional calculations which produced neutron stars with masses on the lower side of plausible values ( @xmath68m@xmath1 ) .
also another problem of published explosion models ( e.g. , @xcite ) has disappeared : the ejecta mass with @xmath69 is less than @xmath70m@xmath1 in our calculation ( fig .
[ fig : yemass ] ) , thus fulfilling a constraint pointed out by @xcite for supernovae if they should not overproduce the @xmath3 ( closed neutron shell ) nuclei , in particular @xmath71sr , @xmath72y and @xmath73zr , relative to the galactic abundances .
of course , final statements about explosion energy , ejecta composition , and the neutron star mass ( which may change due to later fallback , especially when the explosion energy remains low ) require to follow the evolution for a longer time .
there is still a long way to go until we will have arrived at a `` standard model '' for massive star explosions .
current numerical simulations are not advanced enough to convincingly and self - consistently demonstrate the viability of any of the explosion scenarios described in sect .
[ sec : theory ] . despite of respectable progress in treating the neutrino transport , general relativity , and multi - dimensional aspects in the most detailed numerical calculations ,
the models are far from being complete . neglecting potentially important physics like , e.g. , magnetic fields
, one can not be sure that one is searching for the actual cause of massive star explosions at the right place .
conversely , the existing simulations also do not bring us in a position to reject the idea that neutrinos deliver the required energy in the ejecta .
with the lack of direct observational information and being confronted with the fact that our knowledge about the physics in collapsing stellar cores is incomplete , there is no alternative to systematically explore the space of possibilities by hydrodynamical simulations .
since `` the output of complex numerical models is as inscrutable as nature herself '' , it seems advisable to first develop reliable and accurate supernova models with a minimum of questionable assumptions and then to move on step by step to expanding them towards larger complexity and more completeness . for the time
being , one s favorate choice of the supernova mechanism is more a matter of taste and belief than based on undisputable facts and logical reasoning .
the `` ifs '' and `` buts '' in current models are still too many .
one should be cautious not to put too much money on any single possibility right now .
we are indebted to k. takahashi for providing us routines to calculate the improved neutrino - nucleon interactions , and to c. horowitz for correction formulae for the weak magnetism .
we also thank m. liebendrfer for making output data of his simulations available to us for comparisons .
the institute for nuclear theory at the university of washington is acknowledged for its hospitality and the department of energy for support during a visit of the summer program on neutron stars , during which most of the work leading to fig .
[ fig : monomode ] was done .
htj , rb and mr are grateful for support by the sonderforschungsbereich 375 on `` astroparticle physics '' of the deutsche forschungsgemeinschaft .
tp was supported in part by the us department of energy under grant no .
b341495 to the center of astrophysical thermonuclear flashes at the university of chicago , and in part by the grant 2.p03d.014.19 from the polish committee for scientific research .
he performed his simulations on the cray sv1 - 1a at the interdisciplinary centre for computational modelling in warsaw .
the 2d simulations with boltzmann neutrino transport were only possible because a node of the new ibm `` regatta '' supercomputer was dedicated to this project by the rechenzentrum garching .
computations were also done on the nec sx-5/3c of the rechenzentrum garching , and on the cray t90 and cray sv1ex of the john von neumann institute for computing ( nic ) in jlich .
heger , a. , woosley , s.e . ,
fryer , c.l . , & langer , n. 2002 , in : _ from twilight to highlight the physics of supernovae _ ,
w. hillebrandt & b. leibundgut ( springer series `` eso astrophysics symposia '' , berlin : springer ) ( astro - ph/0211062 ) liebendrfer , m. , messer , o.e.b . ,
mezzacappa , a. , hix , w.r . , thielemann , f .- k . ,
& langanke , k. 2002 , in _ proc .
11th workshop on nuclear astrophysics _
w. hillebrandt & e. mller ( report mpa / p13 , garching : max - planck - institut fr astrophysik ) 126 ( astro - ph/0203260 ) nomoto , k. , shigeyama , t. , kumagai , s. , yamaoka , h. , suzuki , t. 1994 , in _
supernovae , les houches session liv _ , ed .
bludman , r. mochkovitch , & j. zinn - justin ( amsterdam : elsevier / north - holland ) 489 nomoto , k. , maeda , k. , umeda , h. , ohkubo , t. , deng , j. , & mazzali , p. 2002 , in _ a massive star odyssey , from main sequence to supernova , proc .
iau symposium 212 _ , ed .
van der hucht , a. herrero , & c. esteban ( san francisco : asp ) in press ( astro - ph/0209064 ) woosley , s.e . , zhang , w. , & heger , a. 2002 , in : _ from twilight to highlight the physics of supernovae _ , ed
. w. hillebrandt & b. leibundgut ( springer series `` eso astrophysics symposia '' , berlin : springer ) ( astro - ph/0211063 ) | one of the central problems in supernova theory is the question how massive stars explode . understanding the physical processes that drive the explosion is crucial for linking the stellar progenitors to the final remnants and for predicting observable properties like explosion energies , neutron star and black hole masses , nucleosynthetic yields , explosion anisotropies , and pulsar kicks . in this article
we review different suggestions for the explosion mechanism and discuss the constraints that can or can not be deduced from observations .
the prompt hydrodynamical bounce - shock mechanism has turned out not to work for typical stellar iron cores and empirical values of the compressibility of bulk nuclear matter .
magnetohydrodynamical models on the other hand contain a number of imponderabilities and are still far behind the level of refinement that has been achieved in nonmagnetic simulations . in view of these facts the neutrino - driven mechanism must still be considered as the standard paradigm to explain the explosion of ordinary supernovae , although its viability has yet to be demonstrated convincingly . since spherically symmetric models do not yield explosions , the hope rests on the helpful effects of convection inside the nascent neutron star , which could boost the neutrino luminosity , and convective overturn in the neutrino - heated region behind the stalled shock , which increases the efficiency of neutrino - energy transfer in this layer . here
we present the first two - dimensional simulations of these processes which have been performed with a boltzmann solver for the neutrino transport and a state - of - the - art description of neutrino - matter interactions .
although our most complete models fail to explode , convection brings them encouragingly close to a success .
an explosion could be obtained by just a minor modification of the neutrino transport , in which case the exploding model fulfills important requirements from observations .
we discuss necessary improvements on the route to finally successful models .
supernovae , neutrinos , radiation - hydrodynamics |
i am very glad to take part in this book devoted to celebration of the 60 birth day of remarkable scientist and my dear friend ioseph l. buchbinder .
four - dimensional @xmath0 supersymmetric gauge theories are formulated in terms of @xmath0 vector multiplet coupled to a massless hypermultiplets in certain representations @xmath1 of the gauge group @xmath2 .
all such models possess only one - loop divergences @xcite and can be made finite at certain restrictions on representations and field contents . in the model with @xmath3 hypermultiplets in representations
@xmath4 of the gauge group @xmath5 the finiteness condition has simple and universal form @xmath6 where @xmath7 is the quadratic casimir operator for the adjoint representation and @xmath8 is the quadratic casimir operator for the representation @xmath9 . a simplest solution to eq.([fin ] ) is @xmath10 sym theory where @xmath11 and all fields are taken in the adjoint representation .
it is evident that there are other solutions , e.g. for the case of @xmath12 group and hypermultiplets in the fundamental representation one gets @xmath13 , @xmath14 and @xmath15 .
a number of @xmath0 superconformal models has been constructed in the context of ads / cft correspondence ( see e.g. @xcite , the examples of such models and description of structure of vacuum states were discussed in details e.g. in ref .
@xcite ) . in this paper
we study the structure of the low - energy one - loop effective action for the @xmath0 superconformal theories .
the effective action of the @xmath10 sym theory and @xmath0 superconformal models in the sector of @xmath0 vector multiplet has been studied by various methods . however
a problem of hypermultiplet dependence of the effective action in the above theories was open for a long time .
the low - energy effective action containing both @xmath0 vector multiplet and hypermultiplet background fields in @xmath16 sym theory was first constructed in ref . @xcite and studied in more details in @xcite . in this paper
we will consider the hypermultiplet dependence of the effective action for @xmath0 superconformal models .
such models are finite theories as well as the @xmath10 sym theory and one can expect that hypermultiplet dependence of the effective action in @xmath17 superconformal models is analogous to one in @xmath10 sym theory .
however this is not so evident .
the @xmath10 sym theory is a special case of the @xmath0 superconformal models , however it possesses extra @xmath0 supersymmetry in comparison with generic @xmath0 models . as it was noted in @xcite
just this extra @xmath0 supersymmetry is the key point for finding an explicit hypermultiplet dependence of the effective action in @xmath10 sym theory .
therefore a derivation of the effective action for @xmath0 superconformal models in the hypermultiplet sector is an independent problem . in this paper
we derive the complete @xmath0 supersymmetric one - loop effective action depending both on the background vector multiplet and hypermultiplet fields in a mixed phase where both vector multiplet and hypermultiplet have non - vanishing expectation values .
the @xmath0 supersymmetric models under consideration are formulated in harmonic superspace @xcite .
we develop a systematic method of constructing the lower- and higher - derivative terms in the one - loop effective action given in terms of a heat kernel for certain differential operators on the harmonic superspace and calculate the heat kernel depending on @xmath0 vector multiplet and hypermultiplet background superfields .
we study a component form of a leading quantum corrections for on - shell and beyond on - shell background hypermultiplets and find that they contain , among the others , the terms corresponding to the chern - simons - type actions .
the necessity of such manifest scale invariant @xmath18-odd terms in effective action of @xmath10 sym theory , involving both scalars and vectors , has been pointed out in @xcite .
proposal for the higher - derivative terms in the effective action of the @xmath0 models in the harmonic superspace has been given in @xcite .
we show how the terms in the effective action assumed in p.c .
argyres at al . can be actually computed in supersymmetric quantum field theory .
@xmath0 harmonic superspace has been introduced in @xcite extending the standard @xmath0 superspace with coordinates @xmath19 ( @xmath20 ) by the harmonics @xmath21 parameterizing the two - dimensional sphere @xmath22 : @xmath23 the main advantage of harmonic superspace is that the @xmath0 vector multiplet and hypermultiplet can be described by unconstrained superfields over the analytic subspace with the coordinates @xmath24 where the so - called analytic basis is defined by @xmath25 the @xmath0 vector multiplet is described by a real analytic superfield @xmath26 taking values in the lie algebra of the gauge group .
a hypermultiplet , transforming in the representation @xmath27 of the gauge group , is described by an analytic superfield @xmath28 and its conjugate @xmath29 .
the classical action of @xmath0 sym theory coupled to hypermultiplets consist of two parts : the pure @xmath0 sym action and the @xmath30-hypermultiplet action in the fundamental or adjoint representation of the gauge group .
written in the harmonic superspace its action reads @xmath31 where we used the doublet notation @xmath32 . by construction ,
the action ( [ class ] ) is manifestly @xmath0 supersymmetric . here
@xmath33 denotes the analytic subspace integration measure and @xmath34 is the analyticity - preserving covariant harmonic derivative .
it can be shown that @xmath35 is the single unconstrained analytic , @xmath36 , prepotential of the pure @xmath0 sym theory , and all other geometrical object are determined in terms of it .
so , the covariantly chiral superfield strength @xmath37 @xmath38 is expressed through the ( nonanalytic ) real superfield @xmath39 satisfying the equation @xmath40=0.\ ] ] this equation has a solution in form of the power series in @xmath35 @xcite .
for further use we will write down also the superalgebra of gauge covariant derivatives with the notation @xmath41 : @xmath42
@xmath43 @xmath44=\varepsilon_{\alpha\beta}\bar{\cal d}^{\pm}_{\dot\beta}\bar{\cal w}~ , \quad [ \bar{\cal d}^{\pm}_{\dot\alpha } , { \cal d}_{\beta\dot\beta}]=\varepsilon_{\dot\alpha\dot\beta}{\cal d}^{\pm}_{\beta}{\cal w}~,\ ] ] @xmath45=\frac{1}{2i}\{\varepsilon_{\alpha\beta}\bar{\cal d}^+_{\dot\alpha}\bar{\cal d}^-_{\dot\beta}\bar{\cal w } + \varepsilon_{\dot\alpha\dot\beta}{\cal d}^-_{\alpha}{\cal d}^+_{\beta}{\cal w}\}=\frac{1}{2i}\{\varepsilon_{\alpha\beta}\bar{f}_{\dot\alpha\dot\beta } + \varepsilon_{\dot\alpha\dot\beta}f_{\alpha\beta}\}~.\ ] ] the operators @xmath46 and @xmath47 strictly anticommute @xmath48 a full set of gauge covariant derivatives includes also the harmonic derivatives @xmath49 , which form the algebra @xmath50 and satisfy the obviously commutation relations with @xmath51 and @xmath52 .
the action ( [ class ] ) possesses the superconformal symmetry @xmath53 which is manifest in the harmonic superspace approach .
the low energy effective action at a generic vacuum of @xmath17 gauge theory includes only massless @xmath54 vector multiplets and massless neutral hypermultiplets , since charged vectors and charged hypermultiplets get masses by the higgs mechanism .
the moduli space of vacua for the theory under consideration is specified by the following conditions @xcite : @xmath55=0 , \quad \phi f_i=0 , \quad \bar{f}^i \bar\phi = 0 \quad \bar{f}^{(i } t_i f^{j)}=0}~.\ ] ] here the @xmath56 are the scalar components of @xmath17 vector multiplet and complex scalars @xmath57 are the scalar components of the hypermultiplet .
the structure of a vacuum state is characterized by solutions to eqs .
( [ vacua ] ) .
these solutions can be classified according to the phases or branches of the gauge theory under consideration . in the pure coulomb phase @xmath58 , @xmath59 and unbroken gauge group
is @xmath60 . in the pure higgs phase @xmath61 and
the gauge symmetry is completely broken ; there are no massless gauge bosons . in the mixed phases ,
i.e. on the direct product of the coulomb and higgs branches ( some number of @xmath56 is not equal to zero and some number of @xmath57 is not equal to zero ) the gauge group is broken down to @xmath62 where @xmath63 is some abelian subgroup . further we impose the special restrictions on the background @xmath0 vector multiplet and hypermultiplet .
they are chosen to be aligned along a fixed direction in the moduli space vacua ; in particular , their scalar fields should solve eqs .
( [ vacua ] ) : @xmath64 here @xmath65 is a fixed generator in the cartan subalgebra corresponding to abelian subgroup @xmath63 , and @xmath66 is a fixed vector in the @xmath1-representation space of the gauge group , where the hypermultiplet takes values , chosen so that @xmath67 and @xmath68 eq.([vac ] ) defines a single @xmath54 vector multiplet and a single hypermultiplet which is neutral with respect to the @xmath69 gauge subgroup generated by @xmath70 . at the tree level and energies below the symmetry breaking scale , we have free field massless dynamics of the @xmath0 vector multiplet and the hypermultiplet aligned in a particular direction in the moduli space of vacua . thus the low energy propagating fields are massless neutral hypermultiplets and @xmath69 vector which form the on shell superfields possessing the properties @xmath71 @xmath72 the equations ( [ onsh ] ) eliminate the auxiliary fields and put the physical fields on shell . at the quantum level , however , exchanges of virtual massive particles produce the corrections to the action of the massless fields .
we quantize the @xmath0 supergauge theory in the framework of the @xmath0 supersymmetric background field method @xcite by splitting the fields @xmath73 into the sum of the background fields @xmath73 , parameterized according to ( [ vac ] ) , and the quantum fields @xmath74 and expanding the lagrangian in a power series in quantum fields .
such a procedure allows us to find the effective action for arbitrary @xmath0 supersymmetric gauge model in a form preserving the manifest @xmath0 supersymmetry and classical gauge invariance in quantum theory . in the background - quantum splitting
, the classical action of the pure @xmath0 sym theory can be shown to be given by @xmath75=s_{sym}[v^{++}]+\frac{1}{4}\int d\zeta^{(-4)}duv^{++}(d^+)^2{\cal w}_\lambda\ ] ] @xmath76 @xmath77 and @xmath78 denote the @xmath79- and @xmath80-frame forms of @xmath81 and @xmath82 respectively .
the hypermultiplet action becomes @xmath83 + \int d\zeta^{(-4)}du q^+_a{\cal d}^{++}q^{+a}+\frac{1}{2}\int d\zeta^{(-4)}duq^+_a iv^{++}q^{+a}\ ] ] @xmath84 the terms linear in @xmath82 and @xmath85 in ( [ splitt ] ) , ( [ hypeq ] ) determines the equation of motion and this term should be dropped when considering the effective action . to construct the effective action
, we will follow the faddeev - popov ansatz .
we write the final result for the effective action @xmath86 $ ] @xmath87}=e^{is_{cl}[v^{++},\ ; q^+]}\mbox{det}^{1/2}{\stackrel{\frown}{\box}}_{(4,0 ) } \int { \cal d}v^{++}{\cal d}q^+ { \cal d}{\bf b}{\cal d}{\bf c}{\cal d}\varphi e^{is_{q}},\ ] ] where @xmath88 and action @xmath89 is as follows @xmath90=s_{2}[v^{++ } , q^+ , { \bf b } , { \bf c } , \varphi , v^{++ } , q^+]+s_{int},\ ] ] @xmath91 @xmath92@xmath93 this equations completely determine the structure of the perturbation expansion for calculating the effective action @xmath86 $ ] of the @xmath0 sym theory with hypermultiplets in a manifestly supersymmetric and gauge invariant form .
the action @xmath94 defines the propagators depending on background fields . in the framework of the background field formalism in @xmath0 harmonic superspace there appear three types of covariant matter and gauge field propagators .
associated with @xmath95 is a green s function @xmath96 which satisfies the equation @xmath97 , is @xmath98 the @xmath99 hypermultiplet propagator associated with the action ( [ q2 ] ) has the form @xmath100 it is not hard to see that this manifestly analytic expression is the solution of the equation @xmath101 for the hypermultiplet of the second type described by a chargeless real analytic superfield @xmath102 the equation for green function is + @xmath103 .
the suitable expression for @xmath104 is @xmath105 the operator @xmath106 transforms each covariantly analytic superfield into a covariantly analytic and , using algebra ( [ alg ] ) , can be rewritten as second - order dalemberian - like differential operator on the space of such superfields .
the coefficients of this operator depend on background superfields @xmath107 .
consider the loop expansion of the effective action within the background field formulation .
a formal expression of the one - loop effective action @xmath108 $ ] for the theory under consideration is written in terms of a path integral as follows ( [ path3 ] ) , where the full quadratic action is defined in eq . ( [ q2 ] ) . here
@xmath82 is a quantum vector superfield taking values in the lie algebra of the gauge group and @xmath109 , @xmath110 are two real analytic faddeev - popov fermionic ghosts and @xmath111 is the bosonic nielsen - kallosh ghost , all in the adjoint representation of the gauge group . in the vector sector of the @xmath0 sym theory where the matter hypermultiplet are integrated out , the one - loop effective action @xmath112 $ ] reads @xmath113=\frac{i}{2}\mbox{tr}_{(2,2)}\ln\stackrel{\frown}{\box } -\frac{i}{2}\mbox{tr}_{(4,0)}\ln\stackrel{\frown}{\box}-\frac{i}{2}\mbox{tr}_{ad}\ln({\cal d}^{++})^2 + i\mbox{tr}_{r_q}\ln{\cal d}^{++ } + \frac{i}{2}\mbox{tr}_{r_\omega}\ln({\cal d}^{++})^2.\ ] ] currently , the holomorphic and non - holomorphic parts of the low - energy effective action @xmath114 sym theory on the coulomb branch , including heisenberg - euler type action in the presence of a covariantly constant vector multiplet , are completely known .
the general structure of the low - energy effective action in @xmath114 superconformal theories is @xcite : @xmath115 where @xmath116 and @xmath66 are holomorphic and real analytic function of the ( anti)chiral superconformal invariants .
the @xmath117-term is known to generate four - derivative quantum corrections at the component level which include an famous @xmath118 term .
the hypermultiplet dependent part of the effective action in @xmath10 sym theory in leading order is also known @xcite . for further analysis of the effective action it is convenient to diagonalize the action of quantum fields @xmath119 using a special shift of hypermultiplet variables in the path integral @xmath120@xmath121 where @xmath122 are the new independent variables in the path integral .
it is evident that the jacobian of the replacement ( [ replac ] ) is equal to unity . here
@xmath123 is the background - dependent propagator ( [ prohyp ] ) for the superfields @xmath124 . in terms of the new set of quantum fields we obtain for the following hypermultiplet dependent part of the quadratic action @xmath125 then the vector multiplet dependent part of the quadratic action
gets the following non - local extension @xmath126 expression ( [ green2 ] ) , written as an analytical nonlocal superfunctional , will be a starting point for our calculations of the one - loop effective action in the hypermultiplet sector .
our aim in the current and later sections is to find the leading low - energy contribution to the effective action for the slowly varying hypermultiplet when all derivatives of the background hypermultiplet can be neglected .
we will show that for such a case the non - local interaction is localized . using the relation @xmath127 one can rewrite expression for @xmath128 ( [ green2 ] ) in the form @xmath129@xmath130 then we use the explicit form of the green function ( [ prohyp ] ) and the relation allowing us to express the @xmath131 as a polynomial in powers of @xmath132 @xcite : @xmath133@xmath134 where the operator @xmath135 is @xmath136 the non - local term in ( [ nonlocal ] ) takes the form @xmath137@xmath138
the large braces here contain three terms .
it is easy to see that two first terms include the derivatives which will lead to derivatives of the hypermultiplet in the effective action .
since we keep only contributions without derivatives , the above terms can be neglected . as a result ,
is it sufficient to consider only the third term in the braces .
now we apply the relation @xmath139 allowing to integrate over @xmath140 , and obtain @xmath141 then one uses the on - shell harmonic dependence of hypermultiplet @xmath142 and take the coincident limit @xmath143 ( conditioned by @xmath144 ) .
after that we get @xmath145 . as a result , the term under consideration has the form @xmath146 where the expression @xmath147 is treated further as the slowly varying superfield and all its derivatives are neglected .
namely such an expression was obtained in @xcite by summation of harmonic supergraphs .
thus , the second term in ( [ nonlocal ] ) becomes local in the leading low - energy approximation . as a result ,
the operator in action @xmath128 determining the effective background covariant propagator of the quantum vector multiplet superfield @xmath148 takes the form @xmath149 where @xmath150{\cal d}^-_\alpha + \frac{i}{2}t_{(i}[\bar{\cal d}^+_{\dot\alpha}\bar{\cal w},t_{j)}]\bar{\cal d}^{-\dot\alpha } + t_{(i}[{\cal w},[\bar{\cal w } , t_{j)}]].\ ] ] here @xmath151 is the covariant dalemberian .
thus , using the @xmath0 harmonic superspace formulation of the @xmath0 sym theory with hypermultiplets and techniques of the non - local shift we obtained that the whole dependence on the background hypermultiplet is concentrated in the quantum vector multiplet sector with the modified quadratic action .
therefore the one - loop effective action is given by the expression @xmath152 = \gamma^{(1)}_v[v^{++ } , q^+ ] + \widetilde{\gamma}^{(1)}[v^{++}]~,\ ] ] where the first term in ( [ gamma ] ) is originated from quantum vector multiplet @xmath148 @xmath153=\frac{i}{2}\mbox{tr}\ln ( \stackrel{\frown}{\box}_{ij } + q^{+a } \{t_i , t_j\}q^-_{a})~.\ ] ] second term in ( [ gamma ] ) is the contribution of ghosts and quantum hypermultiplet @xmath154 and does not depend on the background hypermultiplet . as a result , the background hypermultiplet dependence of one - loop effective action is included into the operator ( [ box ] ) , acting on @xmath148 and containing the mass matrix of the vector multiplet @xmath155[\bar{\cal w},t_j]+(i\leftrightarrow j)\right ) + q^{+a}\{t_i , t_j\}q^{-}_{a}~,\ ] ] if @xmath85 is in the fundamental representation , and @xmath156[\bar{\cal w},t_j ] + [ q^{+a},t_i][t_j , q^{-}_{a}]\right ) + ( i\leftrightarrow j)~,\ ] ] if @xmath85 in an arbitrary matrix representation . in the above discussion ,
the gauge group structure of the superfields @xmath157 has been completely arbitrary .
henceforth , the background superfields will be chosen to be aligned along a fixed direction in the moduli space of vacua in such a way that their scalar fields should solve egs .
( [ vacua ] ) .
then the hypermultiplet dependent effective action in the case under consideration takes the universal form @xmath158=\]]@xmath159 as the examples we list the values of @xmath160 and @xmath161 for models considered in @xcite .
@xmath162 @xmath10 sym theory with gauge groups @xmath12 , @xmath163 and @xmath164 . here the hypermultiplet sector is composed of a single hypermultiplet in the adjoint representation of the gauge group .
the background was chosen such that the gauge groups are broken down as follows @xmath165 , @xmath166 , @xmath167 .
all background fields aligned along element @xmath168 of the cartan subalgebra ( with @xmath169 ) .
the mass matrix becomes @xmath170 and traces in eq.([gamma ] ) produce the coefficient @xmath161 which is equal to the number of roots with @xmath171 , i.e. to the number of broken generators @xmath172 the form of the mass matrix shows that in this case @xmath173 .
@xmath174 the model introduced in @xcite .
the gauge group is @xmath175 .
the model contains four hypermultiplets @xmath176 in the fundamental and one hypermultiplet @xmath177 in the antisymmetric traceless representation @xmath178 .
the background fields @xmath37 , @xmath176 , @xmath177 are chosen to solve eqs .
( [ vacua ] ) with the unbroken maximal gauge subgroup @xmath179 : @xmath180 @xmath181 the mass matrix @xmath182 has been calculated in @xcite and it has @xmath183 eigenvectors with the eigenvalue @xmath184
@xmath185 the @xmath0 superconformal model which is the simplest quiver gauge theory @xcite .
gauge group is @xmath186 .
the model contains two hypermultiplets @xmath85 , @xmath187 in the bifundamental representations @xmath188 and @xmath189 of the gauge group . in @xcite
a solutions of ( [ vacua ] ) with non - vanishing hypermultiplet components that specifies the flat directions in massless @xmath0 sym theories has been constructed .
the moduli space of vacua for this model includes the following field configuration @xmath190 @xmath191 which preserves an unbroken gauge group @xmath192 together with the diagonal @xmath54 subgroup in @xmath193 associated with the chosen @xmath37 . in such a background
the mass matrix has eigenvalue @xmath194 and the corresponding @xmath183 .
as the result , the hypermultiplet dependent effective action is given by the expression ( [ gamma2 ] ) .
in the next section we will consider the evaluation of this expression .
the expression ( [ gamma2 ] ) is a basis for an analysis of the hypermultiplet dependence of the effective action . in the framework of the fock - schwinger proper - time representation ,
the effective action ( [ gamma2 ] ) is written as follows @xmath195=\frac{i}{2 } n(\upsilon ) \int d \zeta^{(-4 ) } du \int^\infty_0 \frac{ds}{s}e^{-s(\box + \frac{i}{2}\alpha({h})({\cal d}^+{\cal w}{\cal d}^- + \bar{\cal d}^+\bar{\cal w}\bar{\cal d}^- ) + { \cal m}^2_v)}\times\ ] ] @xmath196 where @xmath197 . here
@xmath198 is a superfield heat kernel , the operation @xmath199 means the functional trace in the analytic subspace of the harmonic superspace @xmath200 , where @xmath201 denotes the trace over the discrete indices .
representation of the effective action ( [ proper ] ) allows us to develop a straightforward evaluation of the effective action in a form of covariant spinor derivatives expansion in the superfield abelian strengths @xmath202 .
the leading low - energy terms in this expansion correspond to the constant space - time background @xmath203 , @xmath204 and on - shell background hypermultiplet .
however , it does not mean that we miss all space - time derivatives in the component effective lagrangian .
grassmann measure in the integral over harmonic superspace @xmath205 generates four space - time derivatives in component expansion of the superfield lagrangian .
therefore the above assumption is sufficient to obtain a component effective lagrangian including four space - time derivatives of the scalar components of the hypermultiplet .
calculation of the effective action ( [ proper ] ) is based on evaluating the superfield heat kernel @xmath198 and lead to a final result for the hypermultiplet dependent low - energy one - loop effective action of the heisenberg - euler type .
we remind that the whole background hypermultiplet is concentrated in @xmath206 .
the explicit form of it is : @xmath207=\frac{1}{(4\pi)^2}n(\upsilon)\int d\zeta^{(-4)}du\int_0^\infty \frac{ds}{s^3}e^{-s(\alpha^2(h){\cal w}\bar{\cal w}+r(\upsilon)q^{+a}q^-_a ) } \times\ ] ] @xmath208 here @xmath209 is given by @xmath210 .
it can be expressed in terms of the two invariants of the abelian vector field @xmath211 and @xmath212 as @xmath213.it is easily to see that the integrand in ( [ manifest ] ) can be expanded in power series in the quantities @xmath214 , @xmath215 .
after change of proper time @xmath216 to @xmath217 we get the expansion in power of @xmath218 and their conjugate .
since the integrand of ( [ manifest ] ) is already @xmath219 , we can change in each term of expansion the quantities @xmath220 , @xmath221 by superconformal invariants @xmath222 and @xmath223 @xcite expressing these quantities from @xmath224 and its conjugate . after that
, one can show that each term of the expansion can be rewritten as an integral over the full @xmath17 superspace .
it is interesting and instructive to evaluate the leading part of the effective action ( [ manifest ] ) that exactly coincides , up to group factor @xmath66 with the earlier results @xcite , @xcite , @xcite : @xmath225 here @xmath226 is the euler s dilogarithm function .
next - to - leading corrections to ( [ lead ] ) can also be calculated .
the remarkable feature of the low - energy effective action ( [ lead ] ) is the appearance of the factor @xmath227 in argument @xmath228 .
this factor is conditioned by the vacuum structure of the model under consideration and depends on the specific features of the symmetry breaking .
now we discuss some terms in the component lagrangian corresponding to the effective action ( [ lead ] ) .
component structure of the effective action ( [ lead ] ) has been studied @xcite in the context of @xmath10 sym theory in bosonic sector for completely constant background fields @xmath229 .
however , it was pointed out above that the superfield effective action ( [ lead ] ) allows us to find the terms in the effective action up to fourth order in space - time derivatives of component fields .
now our aim is to find such terms in the hypermultiplet scalar component sector . to do
that we omit all components of the background superfields besides the scalars @xmath230 in the @xmath0 vector multiplet and scalars @xmath231 in the hypermultiplet and integrate over @xmath232 . to get the leading space - time derivatives of the hypermultiplet scalar components we should put exactly two spinor derivatives on each hypermultiplet superfield .
it yields , after some transformations , to the following term with four space - time derivatives on @xmath233 in component expansion of effective action : @xmath234 @xmath235 @xmath236@xmath237 the straightforward calculation of the components in this expression shows that among the many terms with four derivatives there is an interesting term of the special type . as the first term in expansion over variable @xmath238 we have @xmath239@xmath240 the expression ( [ chern ] ) has a form of the chern - simons - like action for the multicomponent complex scalar filed .
the terms of such form in the effective action were discussed in refs .
@xcite , @xcite in context of @xmath241 sym models and in refs .
@xcite for @xmath242 superconformal models respectively . here
the expression ( [ chern ] ) is obtained as a result of straightforward calculation in the supersymmetric quantum field theory .
in the above consideration a crucial point was the condition that the hypermultiplet @xmath85 satisfies the one - shell conditions ( [ onsh ] ) and the constraint @xmath243 .
here we relax the on - shell conditions and study some of possible subleading contributions with the minimal number of space - time derivatives in the component effective action .
[ ht ] we consider a supergraph given in fig.[f1 ] with two external hypermultiplet legs and with all propagators depending on the background @xmath0 vector multiplet .
here the wavy line stands for the @xmath0 gauge superfield propagator and the solid external and internal lines stand for the background hypermultiplet superfields and quantum hypermultiplet propagator respectively . for simplicity we suppose that the background field is abelian and omit all group factors . the corresponding contribution to effective action
looks like @xmath244 @xmath245 as usually , we extract the factor @xmath246 from the vector multiplet propagator for reconstructing the full @xmath0 measure .
then we shrink a loop into a point by transferring the @xmath95 and @xmath247 from first @xmath248-function to another one and kill one integration . at this procedure
the operator @xmath95 does not act on @xmath85 because we are interesting in the minimal number of space - time derivatives in the component form of the effective action . as a result ,
one obtains @xmath249 @xmath250 further we use twice the relation ( [ polin ] ) allowing us to express the @xmath251 as a polynomial in powers of @xmath132 .
then after multiplying the @xmath252 with the distribution @xmath253 we obtain a polynomial in @xmath132 containing the powers of this quantity from 5-th to 1-st .
the first order is just a contribution of the type which we considered in the previous section , because one derivation @xmath254 is used for transformation @xmath132 into @xmath255 in the coincident limit .
another @xmath256 transforms @xmath85 into @xmath257 .
all that has been already done in section 4 . here
we consider the new contribution to the effective action containing term @xmath258 in the above polynomial : @xmath259@xmath260 the ellipsis means the terms with the powers of @xmath132 other then 2 .
one can show that in the coincident limit they disappear . now transferring @xmath254 on @xmath258 we obtain the expression : @xmath261 where @xmath135 is defined in ( [ delt ] ) .
let us consider each of the two underlined contributions separately .
we use the representation @xmath262 where @xmath263 means the coincident limit @xmath264 .
then we can apply a derivative expansion of the heat kernel .
the goal is to collect the maximum possible number of factors of @xmath265 acting on @xmath266 and having the minimum order in @xmath216 in the integral over @xmath216 .
higher orders in @xmath216 generate the higher spinor derivatives in the effective action .
we take terms @xmath267 from @xmath135 and expand the exponential so as to find @xmath268 . the eq .
( [ repr ] ) allows us to write the leading contribution to @xmath269 as follows @xmath270@xmath271 after trivial integration over p and s this contribution has the form @xmath272 now we fulfil the same manipulations with the second underlined contribution @xmath273 keeping the same order in @xmath216 and @xmath274 as in the expression ( [ gamma(1 ) ] ) .
after that we see that the leading term of the form ( [ gamma(1 ) ] ) is absent in @xmath273 .
then it is not difficult to show that the contribution ( [ gamma(1 ) ] ) is rewritten as follows [ we use @xmath275 @xmath276 the non - zero result arises when all @xmath277 - factors act only on the spinor delta - function .
thus , the contribution under consideration is written as an integral over the measure @xmath278 which looks like `` @xmath279 - part '' of the full @xmath0 harmonic superspace measure @xmath280 .
therefore , the hypermultiplet dependent effective action contains the term @xmath281 presence of such a term in the effective action for @xmath0 supersymmetric models in subleading order was proposed in @xcite .
here we have shown how this term can be derived in the supersymmetric quantum field theory .
it is interesting and instructive to find a component form of such a non - standard superfield action ( [ 3/4 ] ) .
here we consider only a purely bosonic sector of ( [ 3/4 ] ) .
after integration over anticommuting variables , which can be equivalently replaced by supercovariant derivatives evaluated at @xmath282 , we obtain a chern - simons - like contribution to the effective action containing three space - time derivatives @xmath283 this expression is the simplest contribution to the hypermultiplet dependent effective action beyond the on - shell conditions ( [ onsh ] ) for the background hypermultiplet .
of course , there exist other , more complicated contributions including the hypermultiplet derivatives , they also can be calculated by the same method which led to ( [ 3/4 ] ) . here
we only demonstrated a procedure which allows us to derive the contributions to the effective action in the form of integral over @xmath279 - part of the full @xmath0 harmonic superspace .
we have studied the one - loop low - energy effective action in @xmath17 superconformal models .
the models are formulated in harmonic superspace and their field content correspond to the finiteness condition ( [ fin ] ) .
effective action depends on the background abelian @xmath0 vector multiplet superfield and background hypermultiplet superfields satisfying the special restrictions ( [ vacua ] ) , ( [ vac ] ) which define the vacuum structure of the models .
the effective action is calculated on the base of the @xmath0 background field method for the background hypermultiplet on - shell ( [ onsh ] ) and beyond the on - shell conditions . for an on - shell
hypermultiplet we found the universal expression for the effective active action .
for hypermultiplet beyond on - shell , we calculated the special manifestly @xmath0 supersymmetric subleading contribution which is written as an integral over @xmath279 of the full @xmath0 harmonic superspace .
we believe that such contributions deserves a special study .
n.g.p is grateful to i.l .
buchbinder for collaboration and s. kuzenko and i. mcarthur for helpful discussions and correspondence .
the work was supported in part by rfbr grants , project no 06 - 02 - 16346 , no 08 - 02 - 00334-a , grant for lrss , project no 2553.2008.2 and intas grant , project no 05 - 7928 .
a. galperin , e. ivanov , s. kalitzin , v. ogievetsky and e. sokatchev , class . and
quant . grav .
* 1 * ( 1984 ) 469 . ; a. galperin , e. ivanov , v. ogievetsky , e. sokatchev , class .
grav . 2 ( 1985 ) 601
grav . 2 ( 1985 ) 617 . | i review the approach @xcite to the one - loop low - energy effective action in the hypermultiplet sector for @xmath0 superconformal models .
any such a model contains an @xmath0 vector multiplet and some number of hypermultiplets .
we found a general expression for the low - energy effective action in the form of a proper - time integral .
the leading space - time dependent contributions to the effective action are derived and their bosonic component structure is analyzed .
the component action contains terms with three and four space - time derivatives of component fields and has the chern - simons - like form . |
one way to analyze the microscopic structure of atomic nuclei is to study the distribution of quarks and gluons , as well as their correlations , in nuclei .
these nuclear parton distribution functions ( npdfs ) can be accessed using various deep inelastic scattering ( dis ) processes : inclusive scattering of leptons , high - mass dimuon production using proton beams , exclusive electroproduction of vector mesons .
none of the above processes determines npdfs comprehensively , only taken together do these experiments provide stringent constraints on npdfs .
the discussion of the present paper is centered around the nuclear effects of nuclear shadowing and antishadowing ( enhancement ) , which affect npdfs at small values of bjorken variable @xmath5 , @xmath12 .
nuclear shadowing of npdfs is developing into an increasingly important subject because it is involved in the interpretation of the rhic data on jet production , evaluation of hard phenomena in proton - nucleus and nucleus - nucleus collisions at the lhc , estimates of the black limit scattering regime in dis , etc .
the major obstacle that hinders our deeper knowledge of npdfs at small @xmath5 is that , up to the present day , all experiments aiming to study npdfs are performed with fixed ( stationary ) nuclear targets . in these data , the values of @xmath5 and @xmath13 are strongly correlated and one measures npdfs essentially along a curve in the @xmath14 plane rather than exploring the entire plane .
moreover , for @xmath15 gev@xmath2 , the data cover the region @xmath16 , where the effect of nuclear shadowing is just setting in . as a result ,
when one attempts to globally fit the available data by modeling npdfs at some initial scale @xmath17 and then performing qcd evolution , various groups @xcite produce significantly different results .
an alternative to the fitting to the data is to combine the gribov theory @xcite , which relates the nuclear shadowing correction to the total hadron - deuteron cross section to the cross section of diffraction off a free nucleon , with the collins factorization theorem @xcite for hard diffraction in dis .
the resulting leading twist theory of nuclear shadowing was developed in @xcite and later elaborated on in @xcite .
the gribov theory has been applied to the description of nuclear shadowing for many years .
first it was done in the region of small @xmath13 , where generalized vector dominance model gives a good description of diffraction , see review in @xcite , and later in deep inelastic region , where large diffractive masses @xmath18 dominate @xcite .
a number of successful model calculations were performed @xcite before the experimental data from hera became available . a calculation constrained to reproduce
the hera data using the gribov theory was presented in @xcite .
it focuses on the calculation of nuclear shadowing for @xmath19 at intermediate @xmath13 where leading and higher twist effects are equally important .
a fair agreement of the data with the gribov theory has been found .
however , this approach does not involve the use of the collins factorization theorem and , hence , does not address npdfs ( see a detailed comparison in sect .
[ sec : results ] ) .
the present work extends the calculation of npdfs of @xcite with an emphasis on the theoretical ambiguity and accuracy of the predictions and makes a comparison to fixed - target nuclear dis data .
in particular , we demonstrate that the theory of leading twist nuclear shadowing and qcd analysis of hard diffraction at hera enable one to predict in a model - independent way the next - to - leading order npdfs for @xmath20 with 30% accuracy , fig .
[ fig : input ] . for larger @xmath5 , @xmath21
, there appears an additional effect of nuclear antishadowing that requires modeling and whose uncertainty is larger . in addition , the hera diffractive data for @xmath22 contains a sub - leading reggeon contribution , which adds additional ambiguity to our predictions , especially for @xmath23 .
the interactions with @xmath24 nucleons ( which is a model - dependent element of the gribov approach ) give negligible contribution in the nmc fixed - target nuclear dis kinematics , see fig .
[ fig : rescattering ] .
the @xmath25-dependence of the nmc data for @xmath26 is reasonably well reproduced , see fig . [
fig : adep ] . the failure to describe the absolute value of the @xmath27 ratios of the available fixed - target data for @xmath28 and
@xmath29 gev@xmath2 likely indicates the presence of significant higher twist effects in the data .
indeed , when the leading twist shadowing is complemented by higher twist effects , which are modeled by @xmath30 , @xmath31 and @xmath32 meson contributions in the spirit of vector meson dominance , the agreement with the data becomes fairly good , see figs .
[ fig : nmcc12 ] , [ fig : nmcca40 ] and [ fig : nmcpb ] .
all this signals that any leading twist qcd analysis of the available data is unreliable for @xmath28 .
the next - to - leading order ( nlo ) effects in the @xmath27 ratios are found to be quite sizable .
this means that it is not self - consistent to use the leading order parameterizations of npdfs in the nlo qcd calculations , see fig .
[ fig : f2eskola ] . in short , the main goals of the paper are to give a concise summary of the leading twist theory of nuclear shadowing , to assess the theoretical uncertainties of the resulting predictions and to make a comparison to nuclear dis data .
we attempted to give a self - contained presentation and , hence , this paper can be rightfully considered as a _ guide _ to leading twist nuclear shadowing .
in this section , we review the leading twist approach to nuclear shadowing developed in @xcite and further elaborated on in @xcite . .
, title="fig:",width=453,height=453 ] -3 cm the approach is based on the 1969 work by v. gribov @xcite , where the following theorem was proven .
let us consider hadron - deuteron scattering at high energies within the approximation that the radius of the deuteron is much larger than the range of the strong interaction .
then the shadowing correction to the total cross section is expressed in terms of the differential diffractive hadron - nucleon cross section .
this is demonstrated in fig .
[ fig : gribov ] : the forward hadron - deuteron rescattering amplitude giving rise to the nuclear shadowing correction contains the hadron - nucleon diffractive amplitude ( denoted by the shaded blob ) squared .
the relationship between nuclear shadowing and diffraction was used in the analysis of parton densities of deuterium and other nuclei by frankfurt and strikman in @xcite . for deuterium and other sufficiently light ( low nuclear density ) nuclei , nuclear shadowing and diffraction on the nucleon
are related in a model - independent way using the gribov theorem ) using the abramovskii , gribov and kancheli ( agk ) cutting rules @xcite , see @xcite .
hence , in the small-@xmath5 limit ( @xmath33 ) , the relation between shadowing and diffraction is essentially a consequence of unitarity . ] .
the generalization to heavy nuclei involves certain modeling of multiple rescattering contributions , which , however , is under control @xcite .
below we shall recapitulate the derivation of the leading twist nuclear shadowing for npdfs , which can be carried out in three steps . .
the shadowing correction arising from the coherent interaction with any two nucleons of the nuclear target with the atomic mass number @xmath25 , @xmath34 ( the superscript @xmath35 serves as a reminder that only the interaction with two nucleons is accounted for ) , is expressed in terms of the proton diffractive structure function @xmath36 ( the superscript @xmath37 indicates the dependence on four kinematic variables ) as a result of the generalization of the gribov result for deuterium ( see also ref .
this does not require decomposition over twists and is therefore valid even for the case of real photon interactions .
the shadowing correction @xmath34 reads . ]
@xmath38 \ , , \label{eq : step1 } \end{aligned}\ ] ] with @xmath39 the ratio of the real to imaginary parts of the diffractive scattering amplitude ; @xmath40 , @xmath41 and @xmath42 the longitudinal ( in the direction of the incoming virtual photon ) and transverse coordinates of the nucleons involved ( defined with respect to the nuclear center ) ; @xmath43 , @xmath44 and @xmath45 the usual kinematic variables used in diffraction .
throughout this work , we use @xmath46 .
equation ( [ eq : step1 ] ) uses the fact that the @xmath45-dependence of the elementary diffractive amplitude is much weaker than that of the nuclear wave function , and , hence , @xmath47 can be approximately evaluated at @xmath48 .
all information about the nucleus is encoded in the nucleon distributions @xmath49 , see appendix [ sec : appendix ] for details .
finally , @xmath50 is a cut - off parameter ( @xmath51 for quarks and @xmath52 for gluons ) , which will be discussed later in the text . the origin of all factors in eq .
( [ eq : step1 ] ) can be readily seen by considering the corresponding forward double rescattering feynman diagram ( see fig . [
fig : gribov2 ] ) , which accounts for the diffractive production of intermediate hadronic states by the incoming virtual photon : the combinatoric factor @xmath53 is the number of the pairs of nucleons involved in the rescattering process . the factor @xmath54 provides the correct translation of the differential diffractive to the total rescattering cross section ( see the definition later ) , as required by the glauber theory @xcite .
the factor @xmath55 is a correction for the real part of the diffractive scattering amplitude @xmath56 .
since the shadowing correction is proportional to @xmath57 , while the total diffractive cross section is proportional to @xmath58 , the factor @xmath55 emerges naturally , when one expresses nuclear shadowing in terms of the total diffractive cross section ( diffractive structure function ) .
the integration over the positions of the nucleons is the same as in the glauber theory .
similarly , because the recoil of the nucleons is neglected ( the transverse radius of the elementary strong amplitude is much smaller than the scale of the variation of the nuclear density ) , both involved nucleons have the same transverse coordinate @xmath42 .
the integration over @xmath44 represents the sum over the masses of the diffractively produced intermediate states . in order to contribute to nuclear shadowing ( not to break up the nucleus in its transition from the @xmath59-state to the @xmath60-state ) ,
the virtual photon should interact with the nucleons diffractively . the product of the two diffractive amplitudes ( depicted as shaded blobs in fig .
[ fig : gribov2 ] ) gives the diffractive structure function of the proton @xmath36 .
also note that we do not distinguish between diffraction on the proton and neutron in the present work , as the corresponding diffractive amplitudes are equal at small @xmath5 .
the effect of the nucleus is given by the nucleon densities @xmath49 .
for the sufficiently heavy nuclei that we consider , nucleon - nucleon correlations can be neglected and the nuclear wave function squared can be approximated well by the product of individual @xmath49 for each nucleon ( the so - called independent particle approximation ) .
the factor @xmath61 is a consequence of the propagation of the diffractively produced intermediate state between the two nucleons involved .
-nucleus rescattering amplitude that gives the principal contribution to nuclear shadowing.,title="fig:",width=453,height=453 ] -3 cm . the qcd factorization theorems for inclusive @xcite and hard diffractive dis @xcite can be used to relate the structure functions in eq .
( [ eq : step1 ] ) to the corresponding inclusive and diffractive parton distribution functions .
since the coefficient functions ( hard scattering parts ) are the same for both inclusive and diffractive structure functions , the relation between the shadowing correction to npdfs and the proton diffractive parton distribution functions ( pdfs ) is given by an equation similar to eq .
( [ eq : step1 ] ) . the shadowing correction to the npdf of flavor @xmath62 , @xmath63 , @xmath64 ,
is related to the proton ( nucleon ) diffractive pdf @xmath65 of the same flavor @xmath66 \ , .
\label{eq : step2 } \end{aligned}\ ] ] equation ( [ eq : step2 ] ) is very essential in several ways .
firstly , it enables one to evaluate nuclear shadowing for each parton flavor @xmath62 separately . secondly , since the diffractive pdfs obey leading twist qcd evolution
, so does the shadowing correction @xmath64 .
this explains why the considered theory can be legitimately called the leading twist approach . since eq .
( [ eq : step2 ] ) is based on the qcd factorization theorem , it is valid to all orders in @xmath67
. hence , if @xmath65 is known with the next - to - leading order ( nlo ) accuracy , as is the case for the used h1 parameterization for @xmath65 , we can readily make predictions for nlo npdfs . .
equation ( [ eq : step2 ] ) is derived in the approximation of the low nuclear thickness , and it takes into account only the interaction with two nucleons of the target .
the effect of the rescattering on three and more nucleons can be taken into account by introducing the attenuation factor @xmath68 ( see for example @xcite ) , @xmath69 where the meaning of @xmath70 should become clear after the following discussion .
let us consider sufficiently small values of bjorken variable @xmath5 such that the factor @xmath61 in eq .
( [ eq : step2 ] ) can be neglected .
then , introducing @xmath70 as @xmath71 eq .
( [ eq : step2 ] ) can be written in the form equivalent to the usual glauber approximation @xmath72 where @xmath73 is the proton inclusive pdf .
therefore , it is clear that thus introduced @xmath70 has the meaning of the rescattering cross section , which determines the amount of nuclear shadowing in the approximation of eq .
( [ eq : step2appr ] ) . hence
, it is natural to assume that the same cross section describes rescattering with the interaction with three and more nucleons , as postulated by the definition of the attenuation factor @xmath68 by eq .
( [ eq : step3 ] ) . in the language of feynman diagrams ,
the assumed form of the attenuation factor implies that the diffractively produced intermediate state rescatters without a significant change of mass with the same cross section on all remaining nucleons of the target , as depicted in fig .
[ fig : gribov3 ] for the case of the triple scattering .
the approximation of elastic rescattering is not important at small enough @xmath5 where longitudinal distances are much larger than the nuclear size , see discussion below .
corrections to the elastic rescattering approximation can be estimated by taking into account the effects of fluctuations of the strength of the rescattering interaction .
modeling of these effects was performed in @xcite with the conclusion that for a wide range of cross section fluctuations , the reduction of nuclear shadowing ( for fixed @xmath74 ) remains a rather small correction for all nuclei .
-nucleus triple scattering amplitude.,width=453,height=453 ] after introducing the attenuation factor into eq .
( [ eq : step2 ] ) , the complete expression for the shadowing correction , @xmath75 , becomes @xmath76 \ , .
\label{eq : master } \end{aligned}\ ] ] this is our master equation ( see also eq .
( [ eq : master2 ] ) ) .
it contains several sources of model - dependence and theoretical ambiguity .
first , the attenuation factor @xmath68 assumes that multiple rescatterings can be described by a single rescattering cross section ) at any @xmath13 .
however , the quasieikonal approximation employed in eq .
( [ eq : master ] ) is best justified at low @xmath77 , where fluctuations in the strength of the interaction are smaller ( see the discussion in ref .
the qcd evolution equations automatically account for the proper increase of the fluctuations of the effective cross section around its average value @xmath70 with an increase of @xmath13 .
this important effect is omitted if one attempts to apply eq .
( [ eq : master ] ) at @xmath78 with a @xmath13-dependent @xmath79 . ] @xmath80 , i.e. cross section fluctuations are neglected in the interaction with three and more nucleons .
note that in the phenomenologically important kinematic region of fixed - target experiments , @xmath23 and @xmath81 gev@xmath2 , the uncertainty associated with the attenuation factor @xmath68 is negligible since the rescattering contribution to shadowing is small , see fig .
[ fig : rescattering ] .
second , the necessity to introduce the parameter @xmath50 is a consequence of the fact that eq .
( [ eq : master ] ) applies only to the region of nuclear shadowing : the transition to the region of the enhancement of npdfs should be modeled separately .
this is the role of the parameter @xmath50 .
third , there are experimental uncertainties in the determination of the diffractive pdfs @xmath65 which we use as an input in eq .
( [ eq : master ] ) .
equation ( [ eq : master ] ) defines the input npdfs for the dglap evolution equations . as a starting evolution scale @xmath17 ,
we take @xmath82 gev@xmath2 : this is the lowest value of @xmath13 of the h1 diffractive fit @xcite .
nuclear pdfs at @xmath83 are obtained using the nlo qcd evolution equations .
therefore , we predict that nuclear shadowing is a leading twist phenomenon . in the small-@xmath5 limit , which for practical purposes means @xmath84 , the factor @xmath61 in eq .
( [ eq : master ] ) can be safely omitted , which results in a significant simplification of the master formula ( after integration by parts two times ) @xmath85 where @xmath86 ; @xmath87 . in the heavy nucleus limit ( @xmath88 ) and at fixed @xmath74 , @xmath89 where @xmath90 is the nuclear size .
as can be seen , in the theoretical limit of infinitely heavy nucleus , nuclear shadowing equals the ratios of the nuclear to nucleon sizes , i.e. it is a purely geometrical effect . at the same time , in the @xmath91 limit , when the leading twist approximation is violated and the radius of the strong interaction becomes larger than @xmath90 , the @xmath92 ratio should approach unity @xcite .
the master equation ( [ eq : master ] ) uses as input the information on hard diffraction in dis on the proton , which was measured at hera by zeus @xcite and h1 @xcite collaborations .
we use the h1 parameterization of @xmath93 @xcite ( note the superscript @xmath94 indicating that the @xmath45-dependence of diffraction is not measured ) , which is based on the qcd analysis of the 1994 h1 data @xcite ( we use fit b , see appendix a of ref .
the choice of the h1 parameterization is motivated by the following observations : it is available in an easily accessible and usable form , see @xcite and also appendix a of ref .
@xcite .
the diffractive jet production in dis at hera data @xcite is best described by the h1 parameterization .
the fit of alvero , collins , terron and whitmore @xcite somewhat overestimates the data .
another parameterization available in the literature , that of hautmann , kunszt and soper @xcite , is not based on the detailed fit to the available diffractive data .
the 1994 h1 fit is in a fair agreement with the most recent 1997 h1 data @xcite .
however , the 1997 h1 data indicates that the gluon distribution of the 1994 fit is too large by about 25% .
hence , in our analysis we multiplied the gluon diffractive distribution of @xcite by 0.75 . since the diffractive pdf @xmath65 enters eq .
( [ eq : master ] ) at @xmath95 , one has to assume a certain @xmath45-dependence in order to be able to use the h1 results for the @xmath45-integrated @xmath93 . the common choice is to assume that @xmath96 so that after the integration over @xmath45 , one obtains @xmath97 where @xmath98 is the slope of the @xmath45-dependence of @xmath65 .
a priory there is no reason why the slope @xmath98 should be equal for all parton flavors @xmath62 and , hence , we introduce its explicit flavor dependence . in our analysis
we use the following values for @xmath98 . for all quark flavors ,
we use @xmath99 gev@xmath100 , which is determined by the measurement of the @xmath45-dependence of the diffractive structure function @xmath36 , as measured by the zeus collaboration @xcite .
of course , the diffractive slope should increase with decreasing @xmath5 ( diffractive cone shrinkage ) .
however , since the experimental error of the value of @xmath101 is large and no measurements of the @xmath5-dependence of @xmath101 are available , any theoretically expected logarithmic increase of @xmath101 will be within the quoted experimental errors . hence , it is sufficient to use the @xmath5-independent @xmath101 .
the slope of the gluon pdf , @xmath102 , could be different from @xmath101 .
if the gluon - induced diffraction is dominated by small - size ( compared to typical soft physics sizes ) partonic configurations in the projectile , @xmath102 could as low as the slope of @xmath103 diffractive production measured at hera which is substantially lower than @xmath101 . to reflect the uncertainties in the value of @xmath102 we examined two scenarios : @xmath104 gev@xmath100 and @xmath105 gev@xmath100 .
the first one corresponds to the lower end of the values of the @xmath103 photoproduction slope reported at hera @xcite , while the second one is close to @xmath101 and to the @xmath106 slope reported in @xcite .
the analysis of recent zeus data on the slope of @xmath36 @xcite reports the values similar to those reported in @xcite .
the analysis also provides information on the @xmath44-dependence of the slope though not on the @xmath43-dependence of the slope ( average @xmath43 for the data sample is growing with a decrease of @xmath44 ) .
overall , the data appear to be consistent with the regge factorization and for the most of the @xmath44 range , the gluon diffractive pdf appears to give a significant , if not the dominant , contribution .
this suggests that our model with a higher value of @xmath102 is closer to the data , though in view of the lack of the data on the slope of the gluon induced diffraction at large @xmath43 , we feel necessary to keep the lower @xmath102 model as well .
the analysis of the 1994 h1 data @xcite showed that at large @xmath44 , the successful fit to the data requires both the pomeron and the reggeon contributions . in our numerical analysis
we include only the dominant pomeron part , because the subleading reggeon contribution begins to play a role only for @xmath23 , where the theoretical ambiguities are large anyway .
therefore , in our analysis we use the reggeon contribution only to estimate its contribution to the overall uncertainty of our predictions , see appendix [ sec : reggeon ] for details . using eq .
( [ eq:43 ] ) , the rescattering cross section @xmath70 becomes @xmath107 where @xmath39 is the ratio of the real to imaginary parts of the diffractive amplitude @xmath56 .
this ratio can be related to the intercept of the effective pomeron trajectory , @xmath108 , using the gribov - migdal result @xcite @xmath109 where the h1 value for @xmath108 was employed .
the results of the evaluation of @xmath70 are presented in fig .
[ fig : sigma ] .
the left panel presents @xmath74 for anti @xmath110-quarks and the right panel is for the gluons , both cases for @xmath111 gev@xmath2 .
the error bands around the central curves represent the uncertainty in the determination of @xmath74 .
this uncertainty comes from the uncertainties in @xmath101 , @xmath112 ( taken to be 25% ) and the choice of @xmath50 added in quadrature .
two solid curves for the gluon case correspond to the two scenarios for the slope @xmath102 discussed above . for the anti @xmath110-quark and gluon channels at @xmath82 gev@xmath2 .
the error bands represent the uncertainty in the predictions discussed in the text.,width=491,height=491 ] now we would like to examine which values of the diffractive masses or @xmath43 contribute to @xmath74 . at very high energies ( small @xmath5 ) , one enters the regime analogous to the triple pomeron limit of hadronic physics , which corresponds to @xmath113 . in this case , one may need to resum logs of energy in the diffractive block ( logs of @xmath43 ) .
however , deviations from dglap are expected only for @xmath114 , which is beyond the @xmath5 range that we consider @xcite . at extremely small @xmath5 ,
the contribution of small @xmath43 may become dominant .
this contribution was evaluated within color condensate model in @xcite neglecting the large @xmath43 contribution .
to analyze at what @xmath5 small @xmath43 become dominant , it is convenient to introduce the ratio @xmath115 defined as follows @xmath116 the ratio @xmath115 for @xmath110-quark and gluon channels at @xmath82 gev@xmath2 is presented in fig .
[ fig : r ] . in the figure
, the solid curves correspond to @xmath117 ; the dashed curves correspond to @xmath118 ; the dotted curves correspond to @xmath119 ; the dot - dashed curves correspond to @xmath120 . at @xmath82 gev@xmath2 .
the solid curves correspond to @xmath117 ; the dashed curves correspond to @xmath118 ; the dotted curves correspond to @xmath119 ; the dot - dashed curves correspond to @xmath120.,width=491,height=491 ] from fig .
[ fig : r ] one can see how much different @xmath121-regions contribute to nuclear shadowing .
for instance , taking @xmath122 , which roughly corresponds to the smallest @xmath5 which could be reached in the rhic kinematics , one sees that large diffractive masses that correspond to @xmath123 ( dashed curve ) contribute 20% to nuclear shadowing in the quark channel and 30% to nuclear shadowing in the gluon channel .
therefore , fig .
[ fig : r ] indicates that if the color glass condensate model is implemented in a way consistent with the hera diffractive data , it predicts a very small fraction of total shadowing for the rhic kinematic range . for completeness , we rewrite our master equation , eq .
( [ eq : master ] ) , in the form which explicitly includes the diffractive slope @xmath98 @xmath124 \ , .
\label{eq : master2 } \end{aligned}\ ] ] we would like to point out that while the leading twist theory of nuclear shadowing is applicable to the partons of all flavors , see eq .
( [ eq : master2 ] ) , using the low-@xmath5 hera diffractive data , which is heavily dominated by the pomeron contribution , we can not make any quantitative predictions for nuclear shadowing of the valence quarks in nuclei .
nuclear shadowing for the valence quarks is driven by the @xmath45-channel exchanges with non - vacuum quantum numbers ( reggeon contribution ) , whose contribution is largely lost in the kinematic region of the hera data . in practical terms , this means that eq .
( [ eq : master2 ] ) should be applied to evaluate nuclear shadowing for the antiquarks and gluons only . as mentioned above , eq .
( [ eq : master2 ] ) can not describe nuclear modifications of pdfs at @xmath125 for the quarks and @xmath126 for the gluons , where nuclear antishadowing and the emc effects dominate . for a comprehensive picture of nuclear modification for all values of @xmath5
, we refer the reader to the review in @xcite . however , since we use eq .
( [ eq : master2 ] ) to evaluate npdfs at some input scale for qcd evolution , in order to provide sensible results after the evolution , we should have a reasonable estimate of npdfs for all @xmath5 .
we adopted the picture of nuclear modification of pdfs developed in @xcite , which suggests that antiquarks in nuclei are not enhanced and the gluons are antishadowed and which uses the constraints based on the baryon and energy - momentum conservation sum rules . in our case , like in @xcite , we model the enhancement of the gluon npdf in the interval @xmath127 with a simple function @xmath128 and choose the free coefficient @xmath129 by requiring the conservation of the momentum sum rule for npdfs .
for instance for @xmath130ca , this requirement gives @xmath131 and about 2 - 3% enhancement of the fraction of the total momentum of the nucleus carried by the gluons , in accord with the analysis of @xcite .
the master equation ( [ eq : master2 ] ) allows one to determine nlo npdfs at the input scale @xmath82 gev@xmath2 . as an example of such a calculation
, we present ratios of the nuclear ( ca-40 ) to free proton pdfs and the ratio of the nuclear to the free nucleon structure function , @xmath132 , at @xmath111 gev@xmath2 by solid curves in fig .
[ fig : input ] .
the shaded error bands around the solid curves indicate the uncertainty of the predictions . for comparison ,
lo predictions of eskola _ et al . _
@xcite ( based on the lo fit to the dis and drell - yan nuclear data ) for the corresponding ratios are given by the dashed curves .
note that for the gluon ratio , we give two predictions corresponding to two versions of the diffractive slope @xmath102 discussed earlier .
also , since we do not predict nuclear shadowing for the valence quarks , this information should be taken from elsewhere . in our analysis
, we use the parameterization by eskola _
@xcite ( see the upper left panel ) . for the parameterization of the proton pdfs
, we used the nlo fit cteq5 m @xcite .
in @xmath130ca at @xmath133 gev . the leading twist theory results ( solid curves and the corresponded shaded error bands ) are compared to the lo predictions by eskola _
@xcite ( dashed curves).,width=491,height=491 ] figure [ fig : evo ] presents the @xmath13-evolution of the ratios in fig .
[ fig : input ] .
the solid curves correspond to @xmath111 gev@xmath2 ; the dashed curves correspond to @xmath134 gev@xmath2 ; the dot - dashed curves correspond to @xmath135 gev@xmath2 .
the leading twist character of the predicted nuclear shadowing is apparent from this figure : the shadowing correction decreases slowly with increasing @xmath13 and there is still rather significant nuclear shadowing at @xmath135 gev@xmath2 .
the solid curves correspond to @xmath111 gev@xmath2 ; the dashed curves correspond to @xmath134 gev@xmath2 ; the dot - dashed curves correspond to @xmath135 gev@xmath2.,width=491,height=453 ] one can see from fig . [
fig : input ] that our predictions at the lowest values of bjorken @xmath5 significantly differ from those by eskola _
_ however , one should keep in mind that we make our predictions to the nlo accuracy , while the fitting to the nuclear dis data in @xcite is done to the lo accuracy .
one should also note that the parameterization of eskola _ et al . _
@xcite assumes that at small @xmath5 , the ratios @xmath27 and @xmath136 become equal and stay constant ( saturate ) .
we point out that : figure [ fig : input ] presents our predictions for the shapes of the npdfs for @xmath130ca , which are to be used as an input for qcd evolution at the scale @xmath137 gev .
this choice of @xmath138 is motivated by the fact that the 1994 h1 diffractive data has @xmath139 gev@xmath2 and the qcd fit to the data of @xcite starts at @xmath140 gev@xmath2 .
results of such evolution are presented in fig .
[ fig : evo ] .
leading twist theory predicts much more significant nuclear shadowing for quarks and gluons than the fits to the fixed - target data of eskola _
et al._. @xcite .
the latter _ assumed _ that shadowing saturates for small @xmath5 , e.g. for @xmath141 for @xmath130ca , and that higher twist effects are negligible .
nuclear shadowing for the gluons is larger than for the quarks due to the dominance of gluons in diffractive pdfs ( this was further confirmed by the recent zeus data @xcite . within our model of antishadowing for the gluons by the simple function @xmath128 , significant variations of the parameter @xmath129 still lead to the conservation ( with accuracy better than 1% ) of the parton momentum sum rule .
hence , the amount of antishadowing for the gluons is not sensitive to the low-@xmath5 behavior of the gluons .
should we compare our predictions to those by hirai , kumano and miyama or to the updated fit by hirai , kumano and nagai @xcite , the disagreement in the shadowing predictions , especially for the gluons , would be much larger . for the comparison of the parameterizations of @xcite and the first of @xcite , one can consult @xcite .
comparing to the parameterization suggested in the work of li and wang @xcite , we again find a strong disagreement in the quark channel and , a surprisingly good agreement for the gluons .
however , the parameterization of @xcite are not based on the detailed comparison to all available fixed - target data , but rather on the need to fit the rhic data within the hijing model . note also that this parameterization does not include the enhancement either in the quark or gluon channels and , hence , violates the exact qcd momentum sum rule for the parton densities .
the only nlo qcd fit to nuclear dis data by de florian and sassot @xcite produces very small nuclear shadowing for npdfs , which is inconsistent with our predictions as well as with the predictions of @xcite . our predictions for npdfs and the structure function @xmath142 for the nuclei of @xmath143c , @xmath130ca , @xmath144pd , @xmath145au and @xmath146pb and for the kinematic range @xmath147 and @xmath148 gev@xmath2 have been tabulated .
they are available in the form of a simple fortran program from v. guzey upon request , [email protected] , or from v. guzey s web page , http://www.tp2.rub.de/@xmath149vadimg/index.html .
we also study the importance of the effect of multiple rescatterings , which is described by the attenuation factor @xmath150 in eq .
( [ eq : master2 ] ) .
figure [ fig : rescattering ] compares the result of the full calculation of the @xmath151 ratio for @xmath130ca at @xmath133 gev ( solid curve ) with the calculation , when the rescattering effect was ignored ( dashed curve ) , i.e. @xmath150 was set to one in eq .
( [ eq : master2 ] ) .
as seen from fig .
[ fig : rescattering ] , the rescattering effect becomes unimportant for @xmath152 , i.e. in the kinematics of the fixed - target nuclear dis experiments .
ratio for @xmath130ca at @xmath133 gev .
the solid curve is the result of the full calculation ; the dashed curve is obtained by neglecting multiple rescatterings.,width=453,height=453 ] nuclear shadowing corrections to npdfs become significantly larger , when one considers the interactions with the target nucleus at small impact parameters . indeed , since the density of nucleons in the center of the nucleus is larger than the average nucleon density , choosing small impact parameters corresponds to the increase of the number of scattering centers . introducing the impact parameter dependent npdfs , @xmath153 , as was done in @xcite @xmath154 the nuclear shadowing correction to the impact parameter
dependent npdfs can be readily found from eq .
( [ eq : master2 ] ) by simply removing the integration over the impact parameter @xmath155 @xmath156 \ , .
\label{eq : master3 } \end{aligned}\ ] ] the results of the evaluation of the nuclear shadowing correction using eq .
( [ eq : master3 ] ) at the zero impact parameter for anti @xmath110-quarks and gluons in @xmath145au are presented in fig .
[ fig : central ] in terms of the ratios @xmath157 and @xmath158 .
the solid curves correspond to @xmath111 gev@xmath2 ; the dashed curves correspond to @xmath134 gev@xmath2 ; the dot - dashed curves correspond to @xmath135 gev@xmath2 .
note that the factor @xmath159 provides the correct normalization of the impulse approximation term , see @xcite for details . and
@xmath158 for @xmath145au at @xmath111 gev@xmath2 ( solid ) , @xmath134 gev@xmath2 ( dashed ) and @xmath135 gev@xmath2 ( dot - dashed).,width=491,height=491 ] the impact parameter - dependent npdfs have been tabulated and are available upon request from v. guzey or from the following web cite , http://www.tp2.rub.de/@xmath149vadimg/index.html .
they were already used by r. vogt in the analysis of the @xmath103 production at rhic and were found to be in a reasonable agreement with the data @xcite .
our predictions for the @xmath160 ratio , where @xmath132 , can be compared to the nmc data @xcite . however
, since the low-@xmath5 data points correspond to low @xmath13 , we can not make a direct comparison with those points .
therefore , we simply evaluate @xmath160 at @xmath161 gev@xmath2 for the data points with @xmath162 gev@xmath2 .
figures [ fig : nmcc12 ] and [ fig : nmcca40 ] compare predictions of our leading twist model ( upper set of solid curves with the associated error bands denoted by dashed curves ) to the nmc data on @xmath143c and @xmath130ca @xcite ; fig .
[ fig : nmcpb ] makes a comparison to the nmc @xmath163 ratio @xcite .
the lower set of the solid curves is obtained by adding the vmd contribution using eq .
( [ eq : vmd ] ) .
, width=491,height=491 ] one can see from figs .
[ fig : nmcc12 ] , [ fig : nmcca40 ] and [ fig : nmcpb ] that the agreement between the data points and our calculations at low @xmath5 is poor . regardless the fact that our model predicts a significant nuclear shadowing effect for low-@xmath5 , @xmath84 , see fig .
[ fig : input ] , nuclear shadowing rather rapidly decreases when @xmath5 approaches the values of @xmath5 probed in fixed - target nuclear dis experiments .
of course , one might argue that we are comparing our predictions at @xmath111 gev@xmath2 to the data with much lower @xmath13 values .
for instance , in fig .
[ fig : nmcca40 ] for the first five data points , the average values of @xmath13 are @xmath164 gev@xmath2 .
we have explicitly checked that the backward qcd evolution of our predictions down to @xmath165 gev@xmath2 changes the predictions only a little .
therefore , since our approach to nuclear shadowing includes the entire leading twist contribution to the nuclear shadowing correction ( one should keep in mind a significant uncertainty due to the unaccounted reggeon contribution , see appendix [ sec : reggeon ] ) , the disagreement with the nmc low-@xmath5 data compels us to conclude that _ the low-@xmath5 nmc data @xcite could contain significant higher twist effects , which contribute approximately 50% to the nuclear shadowing correction to @xmath142 . _ @xcite .
the lower set of the solid curves is obtained by adding the vmd contribution using eq .
( [ eq : vmd ] ) . ,
width=491,height=491 ] @xcite .
the lower set of the solid curves is obtained by adding the vmd contribution using eq .
( [ eq : vmd ] ) . , width=491,height=491 ] at the same time
, the @xmath25-dependence is reproduced reasonably well ( see below ) indicating that the inadequate modeling of the diffraction at low @xmath166 and @xmath5 is to blame .
indeed , it is very natural to have rather significant higher twist effects at small @xmath13 since , for this kinematics , the contribution of small diffractive masses @xmath167 becomes important .
production of small diffractive masses @xmath167 is dominated by the production of vector mesons , which is definitely a higher twist phenomenon . in this
kinematics , the leading twist h1 parameterization of diffraction @xcite underestimates by approximately factor two the diffractive cross section as illustrated by the following estimate . using the definition of the diffractive differential cross section in terms of the diffractive structure function @xmath168 @xcite and the relation between the differential cross section on the lepton level to the total cross section on the virtual photon level ( the hund convention for the virtual photon flux ) , @xmath169 , the latter can be written as @xmath170 then , if we restrict the integration in eq .
( [ eq : rho1 ] ) by low diffractive masses @xmath167 , @xmath171 gev , the resulting @xmath169 can be compared to the cross section of electroproduction of vector mesons ( dominated by the @xmath30 meson ) .
for instance , a comparison to the hermes data on exclusive leptoproduction of @xmath172 mesons from hydrogen @xcite at low @xmath13 and @xmath173 , @xmath174 gev@xmath2 and @xmath175 gev , demonstrates that the calculation using eq .
( [ eq : rho1 ] ) ( with the restriction @xmath171 gev ) gives only 40% of the experimental value @xmath176 @xmath177b .
this observation means that , in the considered case , there is no duality between the continuum and resonance contributions to low - mass inclusive diffraction .
in order to quantitatively study our conclusion about the significant role of the higher twist contribution to nuclear shadowing , we explicitly add the contribution of @xmath30 , @xmath31 and @xmath32 vector mesons to our leading twist predictions in the spirit of the vector meson dominance ( vmd ) model .
however , since about 50% of the vector meson contribution is already contained in the parameterization of inclusive diffraction , we weigh the contribution of the vector mesons to nuclear shadowing by the factor @xmath178 ( this enables us not to double - count the vector meson contribution ) . therefore , the vmd contribution to the shadowing correction to the structure function @xmath142 reads @xcite @xmath179 where @xmath180 .
the vmd parameters @xmath181 and @xmath182 assume their usual values @xcite . for the nucleon structure function @xmath183 for low @xmath13 and @xmath5 , we used the nmc parameterization @xcite for @xmath184 and the allm fit @xcite for @xmath185 .
in addition , since the values of @xmath13 , where expression ( [ eq : vmd ] ) is used can be as large as 10 gev@xmath2 , we take into account the effect of the vector meson size decrease by introducing the explicit @xmath13 dependence of @xmath181 : @xmath181 decreases by the factor of 4 when @xmath13 increases from @xmath186 gev@xmath2 to 10 gev@xmath2 @xcite .
the inclusion of the vmd contribution to the shadowing correction dramatically improves the agreement between our calculations and the nmc data for the considered cases of @xmath143c and @xmath130ca , see the lower set of the solid curves in figs .
[ fig : nmcc12 ] and [ fig : nmcca40 ] .
this supports our conclusion about the 50% contribution of higher twist effects to nuclear shadowing in the fixed - target nuclear dis kinematics . as seen from fig .
[ fig : nmcpb ] , the inclusion of the vmd contribution does not significantly improve the agreement between our calculations and the data .
since the data points have rather large @xmath13 , @xmath187 gev@xmath2 , it is natural that the contribution of the @xmath30 , @xmath32 and @xmath31 mesons is rather small . at the same time , the amount of shadowing is sensitive to the diffractive masses up to @xmath188 gev , where an enhancement is also possible as compared to the leading twist fit that we employ .
one has to emphasize that one is dealing here with @xmath22 , where the pomeron component of the diffractive pdfs is known to underestimate the data by a large factor , for the recent discussion see @xcite .
this contribution is usually referred to as the reggeon contribution to diffractive pdfs .
however , uncertainties in the theoretical treatment of this contribution are very large due to possible interference between the pomeron and reggeon contribution , unreliable parameter @xmath39 for the reggeon contribution , difficulty to reliably extract the reggeon contribution from the diffractive data .
hence we were forced to neglect this contribution in our numerical analysis .
therefore , the theoretical uncertainty due to the reggeon contribution for @xmath23 is almost out of control and , hence , our calculation for @xmath23 become much less reliable .
we refer the reader to appendix [ sec : reggeon ] where we examine the reggeon contribution and its influence of our predictions of nuclear shadowing and on the comparison of our results to the nmc nuclear dis data .
the domain , where the presented leading twist model is best justified , is @xmath189 and @xmath1 gev@xmath2 . it is worth emphasizing here that our conclusion about the importance of the higher twist ( ht ) effects relies on the existing parameterizations of the leading twist ( lt ) diffractive pdfs which exclude from the fits the region of small diffractive masses .
hence the large @xmath121 region is , to the large extent , an extrapolation of lower @xmath43 data or effectively due to backward evolution of large @xmath43 and high @xmath13 data , which is rather not stable .
so one can not a priory exclude that there exists a lt parameterization which satisfies a local duality requirement in the large @xmath43 , low @xmath13 region .
an investigation of such a possibility would require studying diffraction at lower energies than presently accessible at hera .
one can also study the @xmath25-dependence of nuclear shadowing .
figure [ fig : adep ] presents a comparison of the nmc data on @xmath190 @xcite to our model calculations .
all data points correspond to @xmath191 ; the @xmath192 data point has @xmath193 gev@xmath2 and all other data points have @xmath194 gev@xmath2 .
the solid curve is our main prediction , which is obtained as a sum of the leading twist shadowing and the vmd contribution evaluated using eq .
( [ eq : vmd ] ) .
the dashed curve is obtained by increasing the nuclear shadowing correction of the solid curve by the factor two . as seen from fig .
[ fig : adep ] , our calculation reproduces fairly well the @xmath25-dependence of nuclear shadowing .
however , as was discussed previously in connection with fig .
[ fig : nmcpb ] , even with the higher twist vmd contribution included , our calculation systematically underestimates the absolute value of the shadowing effect . for the heavier nuclei
, we should have had a 2 - 3 times larger nuclear shadowing , as indicated by the dashed curve .
the reason for the discrepancy between our calculations and the data is a significant reggeon contribution to hard diffraction at @xmath191 , which we neglect in our calculation . because of this , the theoretical uncertainty of our results becomes very large for @xmath23 .
-dependence of nuclear shadowing .
the nmc data on @xmath190 @xcite are compared to our lt+vdm predictions ( solid curve ) .
the dashed curve is obtained by scaling up the shadowing correction of the solid curve by the factor two .
, width=491,height=491 ] the leading twist model of nuclear shadowing works best in the kinematics where the vmd and reggeon contributions to hard diffraction are small corrections and where the data on hard diffraction at hera were taken , i.e. for @xmath139 gev@xmath2 and @xmath195 .
an example of the proper application of the leading twist model of nuclear shadowing is presented in fig .
[ fig : adep_lowx ] , which depicts the @xmath25-dependence of @xmath27 at @xmath196 gev@xmath2 for @xmath122 ( solid curve ) , @xmath197 ( dashed curve ) and @xmath198 ( dot - dashed curve ) .
-dependence of nuclear shadowing at @xmath196 gev@xmath2 .
the solid curve corresponds to @xmath122 ; the dashed curve corresponds to @xmath197 ; the dot - dashed curve corresponds to @xmath198 .
, width=491,height=491 ] our conclusion about the importance of the higher twist effects at small @xmath5 and small @xmath13 in the fixed - target data is in a broad agreement with phenomenological approaches to nuclear shadowing , which include both the scaling ( leading twist ) and lowest mass ( @xmath30 , @xmath32 and @xmath31 ) vector meson ( higher twist ) contributions .
( the only possible way to avoid such a conclusion would be the local duality scenario which we discussed earlier in this section . ) in @xcite , the scaling contribution arises as the effect of the diffractive scattering , quite similarly in the spirit to the present work .
however , it is difficult to assess the comparability of the pre - hera parameterization of diffraction used in @xcite with the modern hera data on hard diffraction .
more importantly , the effect of nuclear shadowing was discussed for the nuclear structure function @xmath9 and not for npdfs ( this comment also applies to all other work mentioned below ) . in other approaches , the scaling contribution results from the @xmath199 continuum of the virtual photon wave function @xcite , or from the contributions of higher mass vector mesons @xcite , or from the aligned @xmath200 jets @xcite , or from the asymmetric @xmath201 fluctuations of the virtual photon @xcite .
note also that in the case of the real photon interaction with nuclei ( the accurate data on the real photon diffraction for the relevant energies is available , see @xcite ) , the shadowing data agree well with the gribov s theory , see the discussion in ref .
@xcite .
we would like to point out that a fairly good description of the nmc data was achieved in @xcite , which uses the approach to nuclear shadowing based on its relation to diffraction on the nucleon . as an input for the their calculation
, the authors used the phenomenological parameterizations of the inclusive and diffractive structure functions of the nucleon , which fit well the inclusive and diffractive data .
however , in contrast to our strictly leading twist analysis , the phenomenological parameterizations of @xcite have the @xmath13-dependence of the form @xmath202 , where @xmath129 and @xmath155 are numerical parameters .
hence , the analysis of @xcite effectively includes higher twist contributions , which indirectly confirms our conclusion that a good description of the nmc data @xcite is impossible to achieve without the inclusion of the higher twist effects ( contribution of vector mesons ) .
in addition , we would like to point to other important difference between the present analysis and the analysis of @xcite . in order to evaluate nuclear shadowing as a function of @xmath13 ,
the authors of @xcite apply an equation similar in the spirit to our eq .
( [ eq : master2 ] ) at all @xmath13 . as we explain in the end of sect .
[ sec : theory ] , the application of eq .
( [ eq : master2 ] ) at large @xmath13 violates qcd evolution because one then ignores the proper increase of the fluctuations of @xmath70 as a result of the qcd evolution .
also , neglecting proper qcd evolution , one neglects the contribution of larger @xmath5 effects antishadowing and emc effects to the small-@xmath5 region .
the second major difference is that the use of the qcd factorization theorem for hard diffraction allowed us to make predictions for npdfs .
since this factorization theorem is not used in @xcite , only nuclear structure function @xmath142 is considered .
finally , the well - understood and important effect of the decrease of the coherence length with the increase of @xmath5 , i.e. the factor @xmath61 , was ignored in @xcite .
next we discuss the importance of the next - to - leading order ( nlo ) effects in the nuclear structure function @xmath9 . using the lo parameterization for the ratios of the nuclear to proton pdfs of eskola _ et al .
_ @xcite at the initial scale @xmath203 gev , we perform qcd evolution to @xmath134 gev@xmath2 both with nlo and lo accuracy .
the resulting @xmath204 ratios after the nlo evolution ( dashed curve ) and lo evolution ( dot - dashed curve ) are presented in fig .
[ fig : f2eskola ] .
gev@xmath2 to lo ( dot - dashed curve ) and to nlo ( dashed curve ) accuracy .
our leading twist nuclear shadowing predictions are given by the solid curves and the associated error bands.,width=453,height=453 ] for the proton pdfs , we use cteq5 parameterizations @xcite : cteq5 m for the nlo calculations and cteq5l for the lo calculations . for comparison
, we also present @xmath204 ( solid curves and the associated error bands ) calculated using our leading twist model .
the two solid curves correspond to the two scenarios of nuclear shadowing for gluons ( two models for @xmath102 ) .
a significant difference between the solid and dashed curves in fig .
[ fig : f2eskola ] demonstrates that the effects associated with the nlo qcd evolution and nlo expression for the structure function @xmath9 are important both in the very low-@xmath5 region and in the @xmath5-region of the fixed - target data , @xmath205 .
this gives us another indication that the lo fits to the fixed - target data of @xcite must have significant intrinsic uncertainties , especially at low-@xmath5 , where there is no data and the fits are extrapolations .
in addition , fig .
[ fig : f2eskola ] demonstrates that it is not self - consistent to use the lo fits for npdfs in the nlo calculations of various hard processes with nuclei , which require nlo npdfs as an input .
the determination of the small-@xmath5 behavior of the valence quark distributions relies heavily on the use of the @xmath207 and @xmath208 data for the determination of the @xmath209 difference .
the main problem is that at @xmath210 , the difference is comparable or even smaller than the nuclear shadowing correction for the deuteron , which we denote as @xmath211 , @xmath212
. then the @xmath209 difference reads @xmath213 in view of the discussed uncertainties in the model of diffraction ( higher twist effects , subleading contributions ) in the kinematics , where the nmc took their most accurate data for @xmath214 and @xmath215 scattering , one can hardly use the gribov theory to calculate reliably @xmath211 .
however , the same mechanism works both for the deuteron and for heavier nuclei as strongly suggested by fig .
[ fig : adep ] .
hence we can employ the information on the @xmath216 ratios in order to determine the value of @xmath211 . combining the results of the present analysis with the results of our analysis of nuclear shadowing corrections to the deuteron @xmath207 @xcite
, we find that @xmath217 using this equation and the nmc nuclear data , we find that @xmath218 .
using the cteq5l fit to the nucleon pdfs @xcite , we observe that for this kinematics nuclear shadowing significantly changes the @xmath219 difference @xmath220 where @xmath221 ; @xmath222 is the neutron structure function extracted from the deuteron @xmath207 ignoring the shadowing correction . in this estimate
, we included only errors due to the experimental uncertainty in the @xmath223 ratio .
the main results of this paper can be summarized as follows : we explain the derivation of the leading twist theory of nuclear shadowing in dis on nuclei that relates nuclear shadowing in dis on nuclei to dis diffraction on the proton .
the theory enables us to predict nuclear shadowing for individual nuclear pdfs in a model - independent way at small @xmath5 , @xmath224 . at larger @xmath5 , other nuclear effects ( antishadowing , emc effect ) and details of the mechanism of diffraction at high
@xmath44 introduce a large model dependence and uncertainty .
nuclear shadowing corrections to npdfs are found to be large . in particular , we predict larger shadowing than given by the fits by eskola _
@xcite for gluons for all @xmath5 and for quarks for @xmath225 . in a stark disagreement with all other approaches ,
we predict larger nuclear shadowing for gluons than for quarks .
the presented formalism is applied to evaluate nuclear shadowing for npdfs at all impact parameters .
as one decreases the impact parameter , the effect of nuclear shadowing increases .
the results of our purely leading twist calculations for the @xmath226 , @xmath227 and @xmath228 ratios disagree with the corresponding fixed - target nmc data @xcite at low @xmath5 and low @xmath13 .
while we can not compare our prediction directly to the data at the @xmath13 values of the first five data points , we notice that the backwards qcd evolution is small and it does not seem to increase nuclear shadowing . hence , we conclude that the nmc data with @xmath229 gev@xmath2 are likely to contain a significant amount ( about 50% ) of higher twist effects .
this is supported by the explicit inclusion of the @xmath30 , @xmath32 and @xmath31 meson contributions to the shadowing correction in the spirit of the vector meson dominance model .
an alternative scenario would be the existence of a local duality pattern for diffraction at @xmath22 where so far no data were taken , see appendix [ sec : reggeon ] .
this implies that a leading twist qcd analysis of the low-@xmath5 and low-@xmath13 fixed - target data will not produce reliable results for the low-@xmath5 nuclear pdfs . using general features of the gribov theory and the data on @xmath230 nuclei , it is possible to develop a reliable procedure for the extraction of the neutron @xmath206 in the nmc small-@xmath5 kinematics . our predictions for npdfs , impact parameter - dependent npdfs and the structure function @xmath142 for the nuclei of @xmath143c , @xmath130ca , @xmath144pd , @xmath145au and @xmath146pb and for the kinematic range @xmath147 and @xmath148 gev@xmath2 have been tabulated .
they are available in the form of a simple fortran program from v. guzey upon request , [email protected] .
the qcd evolution was carried out using the qcdnum evolution package @xcite .
we must mention that there is a renewed interest in nuclear shadowing because of the recent surprising measurements of the suppression of production of hadrons with high @xmath231 in deuteron - gold collisions at rhic .
it was claimed that the observed suppression is a spectacular confirmation of the colored glass condensate model .
however , in the rhic kinematics npdfs are probed at relatively large values of bjorken @xmath5 , on average @xmath23 , which is beyond the domain of the colored glass condensate model .
since leading twist nuclear shadowing is rather weak for @xmath23 , nuclear shadowing can not be responsible for the dramatic effect of the suppression of the hadron spectra at forward rapidities at rhic , see @xcite .
however , shadowing in the forward rhic kinematics can be observed by selecting appropriate two jet production kinematics .
after the first version of this paper was released , there appeared an analysis @xcite which calculates higher twist effects in shadowing . similarly to us , the authors come to the conclusion that the higher twist effects in the fixed - target kinematics are large . within uncertainties of their analysis
, the higher twist effects could be even responsible for all shadowing observed at fixed target energies .
so far the connection of the approach of @xcite to the gribov theory is not clear . in particular , the diagrams , which correspond to vector meson production ( which dominates the higher twist small @xmath5 contribution in the gribov theory ,
see discussion in sect . [
sec : ht ] ) seem to be neglected as a very high twist effect .
this work was supported by the german - israel foundation ( gif ) , sofia kovalevskaya program of the alexander von humboldt foundation ( germany ) and the department of energy ( usa ) .
v.g . is grateful to ingo bojak for his help and explanations of qcdnum .
we also thank h. abramowicz for the discussion of the recent zeus diffractive data , and j. morfin for emphasizing the need for a more reliable procedure for the extraction of @xmath10 at small @xmath5 .
the nuclear density @xmath232 , which enters the calculation of nuclear shadowing in eq .
( [ eq : master2 ] ) , was parametrized in a two - parameter fermi form for @xmath130ca , @xmath144pd , @xmath145au and @xmath146pb @xcite @xmath233 } \ , \label{fermi}\ ] ] where @xmath234 and @xmath235 fm and the parameters @xmath236 and @xmath237 are presented in table [ table : fermi ] .
also note that @xmath238 was normalized as @xmath239 .
.the parameters entering the nuclear one - body density @xmath240 . [ cols="^,^,^",options="header " , ] for @xmath143c , we used @xmath241 with @xmath242 , @xmath243 and @xmath244 fm .
the analysis of the 1994 h1 data on hard diffraction was carried out with the assumption that the diffractive structure function @xmath168 is described by a sum of the effective pomeron ( leading ) and reggeon ( subleading ) contributions @xcite @xmath245 where @xmath246 and @xmath247 are the so - called pomeron and reggeon fluxes ; @xmath248 and @xmath249 are the pomeron and reggeon structure functions .
it is important to note that the both terms in eq .
( [ regge1 ] ) are leading twist contributions . in our analysis
we use fit 3 of model b of @xcite which assumed no interference between the leading and subleading contributions .
the latter contribution becomes important only for large values of measured @xmath44 , @xmath22 . in the h1 qcd analysis
, @xmath249 was assumed to be the pion @xmath250 @xcite multiplied by a free coefficient @xmath251 to be determined from the data .
unfortunately , the value of @xmath251 is not given in the h1 publication .
therefore , using eq .
( [ regge1 ] ) we performed a @xmath252 fit to a set of selected h1 data points with @xmath22 and found that @xmath253 .
the reggeon contribution to the nuclear shadowing correction to the @xmath142 structure function has the form similar to our master equation ( [ eq : master2 ] ) @xmath254 \ , .
\label{regge2 } \end{aligned}\ ] ] in this equation , the @xmath255 rescattering cross section is defined as ( compare to eq .
( [ eq : sigma ] ) ) @xmath256 where it is worth noting the absence of the reggeon flux since @xmath257 @xcite . the theoretical uncertainty associated with the reggeon contribution originates from the uncertainty in the choice of @xmath258 , which is the ratio of the real to imaginary parts of the subleading exchange amplitude , @xmath259 where @xmath260 is the signature factor .
the intercept of the reggeon trajectory , @xmath261 , was a fit parameter in the h1 analysis .
the fit to the h1 diffractive data gives @xmath262 @xcite , which leads to @xmath263 depending on the signature factor .
for the @xmath30 and @xmath32 meson exchanges , @xmath264 ; for the @xmath129 and @xmath265 meson exchanges , @xmath266 . since no attempt was made to separate the contributions with different signatures to the reggeon contribution in the analysis of @xcite ( inclusive diffraction is not sensitive to the signature of the exchange ) the value of @xmath258 is uncertain , @xmath267 .
ratio at @xmath111 gev@xmath2 for ca-40 .
the thick solid curves present our main result ; the thin solid curves present the uncertainty of our predictions ; the dashed curve presents the fit of @xcite ; the dot - dashed curves is the result of the calculation including both the pomeron and reggeon contributions to diffraction .
the lower dot - dashed curve corresponds to @xmath266 ; the upper one corresponds to @xmath264 .
, width=491,height=491 ] because of the large @xmath268 , the uncertainty in the choice of @xmath258 leads to a very significant uncertainty in the resulting shadowing correction .
this is illustrated in fig .
[ fig : reggeon_a40 ] presenting @xmath160 for ca-40 at @xmath111 gev@xmath2 . in fig .
[ fig : reggeon_a40 ] , the thick solid curves present the predictions of our model neglecting the subleading exchange contribution to hard diffraction ; the thin solid curves present the uncertainty of the predictions ; the dashed curve presents the result of @xcite ; the dot - dashed curves is the result of the calculation including both the pomeron and reggeon contributions to diffraction .
the lower dot - dashed curve corresponds to @xmath266 ; the upper one corresponds to @xmath264 . as seen from fig .
[ fig : reggeon_a40 ] , the variation of @xmath258 between its lower and upper limits leads to a dramatic change in the predicted nuclear shadowing .
the lower set of the solid curves is obtained by adding the vmd contribution using eq .
( [ eq : vmd ] ) .
the dot - dashed curves are the sums of the pomeron , reggeon and vmd contributions .
the lower dot - dashed curve corresponds to @xmath266 ; the upper one corresponds to @xmath264 .
, width=491,height=491 ] the influence of the reggeon contribution on the comparison of our predictions to the nmc fixed - target nuclear dis data is presented in figs . [
fig : reggeon_catod ] and [ fig : reggeon_pbtoc ] .
the upper set of solid curves and the associated dashed error bands is the result of the calculation using only the pomeron contribution to diffraction .
the lower set of the solid curves is obtained by adding the vmd contribution using eq .
( [ eq : vmd ] ) .
the dot - dashed curves are the results of the calculation taking into account the pomeron , reggeon and vmd contributions .
the lower dot - dashed curve corresponds to @xmath266 ; the upper one corresponds to @xmath264 .
the lower set of the solid curves is obtained by adding the vmd contribution using eq .
( [ eq : vmd ] ) .
the dot - dashed curves are the sums of the pomeron , reggeon and vmd contributions .
the lower dot - dashed curve corresponds to @xmath266 ; the upper one corresponds to @xmath264 .
, width=491,height=491 ] as seen from figs .
[ fig : reggeon_catod ] and [ fig : reggeon_pbtoc ] , varying @xmath258 in the @xmath267 range , one obtains a wide spectrum of predictions for @xmath269 and @xmath228 , which accommodate the nmc data .
this indicates the possibility of a duality between the higher twist vector meson contribution and the leading twist subleading contribution . | using the leading twist approach to nuclear shadowing , which is based on the relationship between nuclear shadowing and diffraction on a nucleon , we calculate next - to - leading order nuclear parton distribution functions ( npdfs ) and structure functions in the region @xmath0 and @xmath1 gev@xmath2 . the uncertainties of our predictions due the uncertainties of the experimental input and the theory are quantified .
we determine the relative role of the small ( @xmath3 ) and large ( @xmath4 ) diffractive masses in nuclear shadowing as a function of @xmath5 and find that the large mass contribution , which is an analog of the triple pomeron exchange , becomes significant only for @xmath6 . comparing our predictions to the available fixed - target nuclear dis data , we argue , based on the current experimental studies of the leading twist diffraction , that the data at moderately small @xmath7 and @xmath8 gev@xmath2 could contain significant higher twist effects hindering the extraction of npdfs from that data . also , we find that the next - to - leading order effects in nuclear shadowing in the ratio of the nucleus to nucleon structure functions @xmath9 are quite sizable . within the same formalism , we also present results for the impact parameter dependence of npdfs .
we also address the problem of extracting of the neutron @xmath10 from the deuteron and proton data .
we suggest a simple and nearly model - independent procedure of correcting for nuclear shadowing effects using @xmath11 ratios . |
quantum entanglement is the major resource in quantum information science and can be used as a potential source for quantum teleportation of unknown states [ 1 ] , quantum key distribution [ 2 ] , quantum cryptography [ 3 ] and quantum computation [ 4 , 5 ] .
entanglement sudden death for bipartite and multipartite systems has been the main focus of researchers during recent years [ 6 - 11 ] .
another important feature , entanglement sudden birth ( esb ) has also been investigated where the initially unentangled qubits can be entangled after a finite evolution of time [ 12 - 13 ] .
recently , entanglement behavior in non - inertial frames was investigated by alsing et al . [ 14 ] .
they studied the fidelity of teleportation between relative accelerated partners .
quantum information in a relativistic setup has become an interesting topic of research during recent years [ 15 - 27 ] . since
, quantum systems are influenced by their environment that may results in the non - unitary dynamics of the system .
therefore , the environmental effect on a quantum system gives rise to the phenomenon of decoherence that causes an irreversible transfer of information from the system to the environment [ 28 - 29 ] .
study of decoherence in non - inertial frames have been investigated for bipartite and multipartite systems by number of authors [ 30 - 33 ] , where it is shown that entanglement is degraded by the acceleration of the inertial observers .
recently , a qubit - qutrit system in non - inertial frames has been analyzed under decoherence [ 34 ] , where it is shown that esb does occur in case of depolarizing channel .
on the other hand , quantum channels with memory [ 35 - 39 ] provide a natural theoretical framework for the study of any noisy quantum communication .
the main focus of this work is to study the entanglement dynamics in the presence of correlated noise as it has not been studied yet in non - inertial frames . in this paper , decoherence and correlated noise effects
are investigated for @xmath0-type states in non - inertial frames by considering using amplitude damping , depolarizing and bit flip channels .
the two observers alice and rob share an @xmath0-type state in non - inertial frames .
alice is considered to be stationary whereas rob moves with a uniform acceleration .
two important features of entanglement , esd and esb are investigated .
no esd occurs in case of depolarizing channel in the presence of correlated noise .
the evolution of a system and its environment can be described by @xmath4where @xmath5 represents the evolution operator for the combined system and @xmath6 corresponds to the initial state of the environment . by taking trace over the environmental degrees of freedom ,
the evolution of the system can be obtained as@xmath7where @xmath8 represents the orthogonal basis of the environment and @xmath9 is the operator describing the evolution of the system .
the above equation can also be written as@xmath10where @xmath11 are the kraus operators as given in ref . [
the kraus operators satisfy the completeness relation@xmath12the decoherence process can also be represented by a map in terms of the complete system - environment state .
the dynamics of a @xmath13-dimensional quantum system can be represented by the following map [ 41]@xmath14where @xmath15 @xmath16 is the complete basis for the system and@xmath17 let alice and rob ( the accelerated observer ) share the following @xmath0-type initial state@xmath18where @xmath19 is the identity operator in a two - qubit hilbert space , @xmath20 and @xmath21 are the pauli operators of the alice s and rob s qubit and @xmath22 @xmath23 are real numbers satisfying the unit trace and positivity conditions of the density operator @xmath24 in order to study the entanglement dynamics , different cases for initial state are considered , for example , the general initial state @xmath25 , the werner initial state @xmath26 , and bell basis state @xmath27 .
let the dirac fields , as shown in refs .
[ 42 , 43 ] , from an inertial perspective , can be described by a superposition of unruh monochromatic modes @xmath28 and @xmath29and@xmath30where @xmath31 , @xmath32 is the acceleration of the observer , @xmath33 is frequency of the dirac particle and @xmath34 is the speed of light in vacuum .
the subscripts @xmath35 and @xmath36 of the kets represent the rindler modes in region @xmath35 and @xmath36 , respectively , as shown in the rindler spacetime diagram ( see ref . [ 31 ] , fig .
( 1 ) ) . by using equations
( 8) and ( 9 ) , equation ( 7 ) can be re - written in terms of minkowski modes for alice ( @xmath37 ) and rindler modes for rob @xmath38 .
the single - mode approximation is used in this study , i.e. a plane wave minkowski mode is assumed to be the same as a plane wave unruh mode ( superposition of minkowski plane waves with single - mode transformation to rindler modes ) . therefore , alice being an inertial observer while her partner rob who is in uniform acceleration , are considered to carry their detectors sensitive to the @xmath33 mode . to study the entanglement in the state from their perspective
one must transform the unruh modes to rindler modes .
hence , unruh states must be transformed into the rindler basis .
let rob detects a single unruh mode and alice detects a monochromatic minkowski mode of the dirac field .
considering that an accelerated observer in rindler region @xmath35 has no access to the field modes in the causally disconnected region @xmath36 and by taking the trace over the inaccessible modes , one obtains the following density matrix@xmath39where @xmath40 and @xmath41 since noise is a major hurdle while transmitting quantum information from one party to other through classical and quantum channels .
this noise causes a distortion of the information sent through the channel .
it is considered that the system is strongly correlated quantum system , the correlation of which results from the memory of the channel itself .
the action of a pauli channel with partial memory on a two qubit state can be written in kraus operator formalism as [ 35]@xmath42}\sigma _ { i}\otimes \sigma _ { j}\]]where @xmath43 ( @xmath44 are usual pauli matrices , @xmath45 ( @xmath46 ) represent the decoherence parameter and indices @xmath47 and @xmath48 runs from @xmath49 to @xmath50 the above expression means that with probability @xmath51 the channel acts on the second qubit with the same error operator as on the first qubit , and with probability @xmath52 it acts on the second qubit independently
. physically the parameter @xmath53 is determined by the relaxation time of the channel when a qubit passes through it .
the action of a two qubit pauli channel when both the qubits of alice and rob are streamed through it , can be described in operator sum representation as [ 46 ] @xmath54where @xmath55 represents the initial density matrix for quantum state and @xmath56 are the kraus operators as expressed in equation ( 11 ) .
a detailed list of single qubit kraus operators for different quantum channels with uncorrelated noise is given in table 1 .
whereas , the kraus operators for amplitude damping channel with correlated noise are given by yeo and skeen [ 36]@xmath57 , \ \ \
a_{11}^{c}=\left [ \begin{array}{llll } 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 \\ \sin \chi & 0 & 0 & 0% \end{array}% \right]\]]where , @xmath58 and is related to the quantum noise parameter as @xmath59the action of such a channel with memory can be written as @xmath60where the superscripts @xmath61 and @xmath34 represent the uncorrelated and correlated parts respectively .
the kraus operators are of dimension @xmath62 and are constructed from single qubit kraus operators by taking their tensor product over all @xmath63 combinations @xmath64where @xmath47 is the number of kraus operator for a single qubit channel .
the final state of the system after the action of the channel can be obtained as @xmath65where @xmath66 is the super - operator realizing the quantum channel parametrized by real numbers @xmath2 and @xmath51 . since , the entanglement dynamics of the bipartite subsystems ( especially the system - environment dynamics ) is of interest here , therefore , only bipartite reduced matrices are considered .
it is assumed that both alice and rob s qubits are influenced by the time correlated environment .
the entanglement of a two - qubit mixed state @xmath67 in a noisy environment can be quantified by the concurrence as defined by [ 45]@xmath68where @xmath69 are the square roots of the eigenvalues of the matrix @xmath70 with @xmath71 being the spin flip matrix of @xmath72 and is given by@xmath73where @xmath74 is the usual pauli matrix .
since the density matrix under consideration has @xmath0-type structure , therefore a simpler expression for the concurrence [ 46 ] can be used@xmath75where @xmath76 and @xmath77 the reduced - density matrix of the inertial subsystem @xmath37 and the non - inertial subsystem @xmath78 , can be obtained by taking the partial trace of @xmath79 over the degrees of freedom of the environment i.e. @xmath80which yields the concurrence for the @xmath0-state structure using equation ( 20 ) , under amplitude damping channel as@xmath81the concurrence of @xmath0-state system in case of depolarizing channel becomes@xmath82and the concurrence of the @xmath0-state system under the influence of bit flip channel reads@xmath83where the super - scripts ad , dep and bf correspond to amplitude damping , depolarizing and bit flip channels respectively .
the results are consistent with refs .
[ 46 , 47 ] and can be easily verified from the expressions ( equations 22 - 24 ) by setting @xmath84 and @xmath85 respectively .
analytical expressions for the concurrence are calculated for @xmath0-type initial state in non - inertial frames influenced by amplitude damping , depolarizing and bit flip channels . in figures 1 and 2 ,
the concurrence is plotted as a function of memory parameter @xmath51 for @xmath86 and @xmath87respectively , for amplitude damping , depolarizing and bit flip channels .
the first panel ( column - wise ) corresponds to bell states , whereas the second and third panels correspond to werner and general initial states , respectively .
it is seen that for bell basis states , entanglement sudden death can be avoided in case of amplitude damping and depolarizing channels in the presence of correlated noise .
however , it is possible to fully avoid esd for all the channels under consideration for @xmath1 ( which can be seen from the figure ) . on the other hand
, esd can also be avoided for werner and general initial states as well in case of amplitude damping channel at higher degree of correlations . as the value of acceleration @xmath88 increases@xmath89 the entanglement degradation
is enhanced which is more prominent for lower range of memory parameter @xmath51 .
it is also seen that bit - flip noise heavily influences the entanglement of the system as compared to the amplitude damping and depolarizing channels for lower level of decoherence . whereas at higher level of decoherence , damping effect of amplitude damping channel becomes more prominent ( see figure 2 ) .
furthermore , it can be seen that initial state plays an important role in the system - environment dynamics of entanglement in non - inertial frames . in figure 3 , the concurrence is plotted as a function of decoherence parameter @xmath2 for @xmath90 for amplitude damping , depolarizing and bit flip channels .
it can be seen that the concurrence is heavily damped by different environments .
this effect is much prominent for amplitude damping and bit flip channels .
it is seen that maximum entanglement degradation occurs at @xmath91 in case of bit flip channel and entanglement rebound process take place for @xmath92 .
furthermore , no esd behaviour is seen for depolarizing channel in the presence of correlated noise .
the degree of entanglement degradation enhances as one shifts from bell type initial state to the case of general initial state .
however , it saturates for infinite acceleration limit @xmath93 . in figures 4 and 5 ,
the concurrence is plotted as a function of rob s acceleration @xmath88 and decoherence parameter @xmath2 with @xmath94 and @xmath95 for amplitude damping , depolarizing and bit flip channels .
the upper panel ( row - wise ) corresponds to bell states , whereas the middle and lower panels correspond to werner and general initial states respectively . from figure 5
, it can be seen that esd can be fully avoided for all the three types of initial states under depolarizing and bit flip noises at 50% quantum correlations . on the other hand ,
esd behaviour is seen only in case of amplitude damping channel for @xmath96 .
therefore , different environments affect the entanglement of the system differently . in figure 6 ,
the concurrence is plotted as a function of memory parameter @xmath51 and decoherence parameter @xmath2 with @xmath97 for amplitude damping , depolarizing and bit flip channels .
it is shown that maximum esd occurs in case of amplitude damping channel .
however , esd can be avoided for depolarizing channel for @xmath98 even for maximum value of decoherence i.e. @xmath99
environmental effects on the entanglement dynamics of dirac fields in non - inertial frames is investigated by considering @xmath0-type initial state shared between the two partners .
it is assumed that the rob is in accelerated frame moving with uniform acceleration whereas alice is the stationary observer .
the concurrence is used to investigate the decoherence and correlated noise effects on the entanglement of the system .
different initial states are considered such as bell basis , werner type and general initial states .
it is seen that in case of bell basis states , entanglement sudden death can be avoided for amplitude damping and depolarizing channels in the presence of correlated noise . whereas , for werner like and general initial states , the entanglement sudden death occurs more rapidly as the value of decoherence parameter @xmath2 and rob s acceleration @xmath88 increase .
therefore , the initial state plays an important role in the system - environment dynamics of entanglement in non - inertial frames .
it is shown that bit - flip channel and amplitude damping channels heavily influence the entanglement of the system as compared to the depolarizing channel .
it is possible to avoid esd for all the channels under consideration for @xmath100 irrespective of the type of initial state considered .
the effect of environment is much stronger than that of rob s acceleration @xmath88 .
furthermore , no esd occurs for depolarizing channel for any value of @xmath2 and @xmath88 , in the presence of correlated noise . in conclusion ,
correlated noise compensates the loss of entanglement caused by the unruh effect .
99 bennett , c.h . ,
brassard , g. , crepeau , c. , jozsa , r. , peres , a. , wootters , w.k . : teleporting an unknown quantum state via dual classical and einstein - podolsky - rosen channels .
* 70 , * 18951 ( 993 ) figures captions**figure 1**. ( color online ) .
the concurrence is plotted as a function of memory parameter @xmath51 for @xmath86 for amplitude damping , depolarizing and bit flip channels , where the abbreviations ad , dep and bf correspond to amplitude damping , depolarizing and bit flip channels respectively.*figure 2*. ( color online ) .
the concurrence is plotted as a function of memory parameter @xmath51 for @xmath87 for amplitude damping , depolarizing and bit flip channels.*figure 3*. ( color online ) .
the concurrence is plotted as a function of decoherence parameter @xmath2 for @xmath90 for amplitude damping , depolarizing and bit flip channels.*figure 4*. ( color online ) .
the concurrence is plotted as a function of rob s acceleration , @xmath88 and decoherence parameter , @xmath2 with @xmath94 for amplitude damping , depolarizing and bit flip channels.*figure 5*. ( color online ) . the concurrence is plotted as a function of rob s acceleration , @xmath88 and decoherence parameter , @xmath2 with @xmath101 for amplitude damping , depolarizing and bit flip channels.*figure 6*. ( color online ) . the concurrence is plotted as a function of memory parameter , @xmath51 and decoherence parameter , @xmath2 with @xmath97 for amplitude damping , depolarizing and bit flip channels.*table 1*. single qubit kraus operators for amplitude damping , depolarizing and bit flip channels where @xmath2 represents the decoherence parameter . | effect of decoherence and correlated noise on the entanglement of @xmath0-type state of the dirac fields in the non - inertial frame is investigated .
a two qubit @xmath0-state is considered to be shared between the partners where alice is in inertial frame and rob in an accelerated frame .
the concurrence is used to quantify the entanglement of the @xmath0-state system influenced by time correlated amplitude damping , depolarizing and bit flip channels .
it is seen that amplitude damping and bit flip channels heavily influence the entanglement of the system as compared to the depolarizing channel . it is found possible to avoid entanglement sudden death ( esd ) for all the channels under consideration for @xmath1 for any type of initial state .
no esd behaviour is seen for depolarizing channel in the presence of correlated noise for entire range of decoherence parameter @xmath2 and rob s acceleration @xmath3 .
it is also seen that the effect of environment is much stronger than that of acceleration of the accelerated partner .
furthermore , it is investigated that correlated noise compensates the loss of entanglement caused by the unruh effect.pacs : 04.70.dy ; 03.65.ud ; 03.67.mnkeywords : quantum decoherence ; correlated noise ; non - inertial frames . |
controlling and preserving coherent quantum dynamics in the framework of quantum information processing is a challenging task.@xcite very recently , more and more experiments on implementing such ideas in mesoscopic systems based on solid state devices @xcite have been realized , e.g. josephson junctions,@xcite and also single electron spins in single defect centers.@xcite the electron spin in quantum dots has been recognized early as a potential carrier of quantum information,@xcite but experimental developments of suitable mesoscopic devices have only recently been pursued . in previous work
it was shown how quantum dots may serve as spin filters , or memory devices for electron spin.@xcite important progress was made in both theoretical and experimental research focusing on measurement schemes through charge currents.@xcite even a single - shot readout of the electron spin state has been realized @xcite and allows for the measurement of the relaxation time of a single spin . still ,
the decoherence time of a single electron spin in a quantum dot has not yet been determined experimentally .
in some of these experiments important quantities are the counting statistics of tunneling electrons.@xcite as for charge qubits , a measurement on the single electron level may be achieved through a single electron transistor ( set ) device @xcite or with a quantum point contact ( qpc ) close to the quantum dot.@xcite it is important to realize that the measurement through a charge current itself has dynamical implications for the measured qubit .
we are thus lead to the problem of noise and statistics induced by the measurement process in these mesoscopic systems.@xcite such problems have been tackled some time ago very elegantly through the concept of _ quantum trajectories _ in quantum optical applications.@xcite in particular , jump processes to describe the time evolution of open systems while counting emitted quanta are well established in the framework of systems that are described by a master equation of lindblad type .
such ideas have already been applied to measurement processes based on quantum point contacts in mesoscopic devices.@xcite in the context of quantum information processing , such quantum trajectory methods turn out to be essential for the design of active quantum error correcting codes @xcite and , more generally , of quantum feedback mechanisms.@xcite historically , a major driving force behind the development of quantum trajectory methods were the growing possibilities to experiment with single quantum systems in traps or cavities .
more recently , such experiments have been extended to mesoscopic solid state devices .
therefore , we expect a growing need for such methods in these fields . the aim of this paper is to show how quantum trajectories serve as a useful framework to discuss the physics of mesoscopic carriers of quantum information under continuous measurement . in particular , we determine counting statistics of electrons tunneling through a quantum dot , depending on the electron spin - state .
we show how a simple setup for state tomography can be achieved through a measurement of counting statistics in combination with a coherent esr field .
we display how decoherence and relaxation time scales can be extracted from the measured data , in the time domain .
we consider a quantum dot with spin-@xmath0 ground state in the coulomb blockade regime as in @xcite , see also fig.[qdot ] .
the quantum dot is subject to a constant magnetic field @xmath1 which leads to a zeeman splitting @xmath2 of the electronic states , where @xmath3 is the electron g factor and @xmath4 the bohr magneton ( throughout this paper we use @xmath5 for gaas and units such that @xmath6 ) .
two leads at chemical potentials @xmath7 and @xmath8 are coupled to the dot for charge transport .
further , as in ref . , we allow for an electron spin resonance ( esr ) field to drive coherent transition between the two spin states . leaving sources of uncontrollable environmental influences aside for a moment ( see below ) , the total hamiltonian consists of contributions from electrons on the dot , electrons in the leads and a tunneling interaction between dot and leads , @xmath9 here , @xmath10 contains contributions from charging and interaction energies of the electrons on the dot , the interaction energy @xmath11 with the static magnetic field , and the esr - hamiltonian @xmath12 of the interaction of the electron spin with a magnetic field @xmath13 , oscillating linearly in the @xmath14 direction .
the @xmath15 @xmath16 denote the usual pauli spin matrices .
the hamiltonian for the two leads @xmath17 reads @xmath18 , with @xmath19 the creation operator of an electron with orbital state @xmath20 , spin @xmath21 and energy @xmath22 in lead @xmath23 .
finally , the coupling between dot and leads is described by the standard tunneling hamiltonian @xmath24 h.c . , where we denote with @xmath25 a tunneling amplitude and with @xmath26 the annihilation operator of an electron on the dot in orbital state @xmath27 .
following ref .
, for the description of the dot dynamics in sect . [ mastereq ] we will also include further ( microscopically unspecified ) dissipative interactions between the dot states and their environment that are not among the known contributions to the total energy as they appear in eq .
( [ htot ] )
. closed dot ( left ) : chemical potentials are too small to allow an electron to tunnel onto the dot .
open dot ( right ) : after excitation of the dot electron , the chemical potential @xmath7 is large enough for an electron of lead one to tunnel onto the dot and form the singlet state with the dot electron.,title="fig:",scaledwidth=30.0% ] closed dot ( left ) : chemical potentials are too small to allow an electron to tunnel onto the dot .
open dot ( right ) : after excitation of the dot electron , the chemical potential @xmath7 is large enough for an electron of lead one to tunnel onto the dot and form the singlet state with the dot electron.,title="fig:",scaledwidth=30.0% ] in the following we give a qualitative picture of the relevant dot states , see fig.[qdot ] , more details may be found in ref . . for simplicity
we assume there is only one electron on the dot . with @xmath28 ,
the electron has the ground state @xmath29 with energy @xmath30 and the exited state @xmath31 with @xmath32 .
if an electron tunnels onto the dot , the two electrons can form the singlet state @xmath33 with energy @xmath34 or either of the three triplet states . as the triplet state @xmath35 has higher energy ( for suitable magnetic field @xcite ) , the singlet @xmath36 is the ground state for two electrons on the dot .
the chemical potentials are chosen such that @xmath37 . under these conditions
the dot can be opened and closed for a sequential tunneling current by a spin flip induced by an esr field:@xcite an electron at chemical potential @xmath7 in the left lead and a dot electron in state @xmath29 do not have sufficient energy to form the singlet state .
if due to an esr induced excitation the dot state is @xmath31 , however , less energy is required and an electron in lead one can tunnel onto the dot to form the singlet state . tunneling onto the dot from lead two is suppressed by several orders of magnitude if the thermal energy is much lower than the energy gap even if the dot electron is in the excited state @xmath31 . at higher temperatures , to assure that the singlet can only be formed with the excited dot electron , one can choose spin - polarized leads .
this may be achieved with several methods , see ref . and
references therein .
thus , within these constraints we see that the current in lead one is proportional to the probability for the dot being in the excited state , i.e. @xmath38 , while the current in lead two is proportional to the probability of the dot being in the singlet state , @xmath39 , see ref . .
the traditional description of the dynamics of the dot is based on a master equation for the reduced density operator of the dot , obtained from the total density matrix @xmath40 by tracing over the degrees of freedom of the leads : @xmath41 $ ] . as usual , we denote matrix elements with @xmath42 ( or @xmath43 ) and include only the three relevant dot states @xmath44 .
we assume the dot and the leads to be uncorrelated initially , @xmath45 .
starting from the von neumann equation for the full density operator @xmath46 $ ] , the master equation for @xmath47 was derived in ref . using standard methods within the markov approximation .
further , we will allow for an arbitrary ( fixed ) phase @xmath48 of the esr field which will play an important role in determining the spin state . in order to eliminate the explicit time dependence emerging from the esr field
, we here base our analysis on the dot state in a rotating frame , @xmath49 in fact , with the exception of @xmath50 , @xmath51 and the corresponding transposed expressions , this transformation leaves almost all matrix elements untouched . along the lines of the derivation in ref .
, one finds for the dot state in the rotating frame ( [ rotframe ] ) a master equation of lindblad form.@xcite it can be written as @xmath52 + \frac{1}{2}\sum_{nm } ( [ l_{nm } \tilde\rho_{dot } , l_{nm}^\dagger ] + [ l_{nm } , \tilde\rho_{dot } l_{nm}^\dagger ] ) \ ] ] with the time - independent hamiltonian ( in rotating wave approximation ) @xmath53 and the operators @xmath54 describing incoherent transitions between levels @xmath27 and @xmath20 with a rate @xmath55 . in particular , the four operators @xmath56 and @xmath57 describe transitions from and to the singlet state and hence correspond to the tunelling of an electron off or onto the dot .
these four contributions give rise to the current in the leads and are derived from the underlying hamiltonian ( [ htot ] ) .
the rates are @xmath58 with @xmath59 and @xmath60 with @xmath61 where @xmath62^{-1}$ ] is the fermi function of lead @xmath63 .
analogously we define the rates @xmath64 , @xmath65 , @xmath66 , @xmath67 with @xmath68 and @xmath69 . here
@xmath70 and @xmath71 are the transition rates with density of states @xmath72 and tunneling amplitude @xmath73.@xcite in the limit @xmath74 we have @xmath75 and @xmath76 , which resembles the sequential tunneling from lead one onto the dot and into lead two .
furthermore , we have @xmath77 and @xmath78 , because we choose @xmath79 . throughout this paper
we assume equal rates for both leads @xmath80 and @xmath81 .
finally , we set @xmath82 if the leads are not spin - polarized , and @xmath83 , @xmath84 in the case of spin - polarization .
we simulate stochastically all processes that could be observed in principle , but eventually extract the desired information from those quantities that correspond to the specific measurement scheme chosen . quantum transitions between the dot states
may be observed by monitoring the current through the dot , which is the starting point for our quantum trajectory analysis of the following sections .
by contrast , mechanisms for incoherent spin flips ( described by the operators @xmath85 and @xmath86 ) and dephasing mechanisms ( described by the projectors @xmath87 ) are introduced on phenomenological grounds and not contained in the hamiltonian ( [ htot ] ) .
the ( phenomenological ) spin flip rates are assumed to satisfy the condition of detailed balance : @xmath88 .
the rates @xmath89 are phenomenological dephasing rates : the effect of an operator @xmath90 in ( [ masterlind ] ) is to kill coherences between state @xmath91 and the remaining states ( at a rate @xmath89 ) , while leaving probabilities unaffected .
if the coupling to the leads is switched off ( by an appropriate choice of the chemical potentials ) , the dynamics as described by the master equation ( [ masterlind ] ) is that of a standard decaying two - state spin - system .
then the corresponding ( intrinsic ) relaxation and decoherence rates turn out to be @xmath92 let us now turn to the dot dynamics : in terms of its coefficients , the time evolution of the dot state @xmath93 given by the master equation ( [ masterlind ] ) reads @xmath94 with the effective rates @xmath95 note that eqs .
( [ rwa5 ] , [ rwa6 ] ) are decoupled from eqs .
( [ rwa1]-[rwa4 ] ) , and the latter are the only ones of relevance to us .
they enable us to determine easily counting statistics of tunneling electrons numerically by means of the quantum trajectory method which we describe in the following sections .
one major motivation behind the development of quantum trajectory methods were experiments with with single quanta . before these developments , naturally , ensemble experiments required simple ensemble theories . matters changed with experiments involving single atoms , electrons or ions in traps . continuously monitoring those systems ,
single quantum jumps became visible to the bare eye .
a theory of continuous quantum measurement taking into account continuous measurement records of the observed environment to update the quantum state accordingly , were developed , mainly with an eye on applications in quantum optics .
experiments on the single quantum level have reached solid state devices , as for instance electrons in quantum dots .
accordingly , the dynamics of such nano - scale quantum systems may be described adequately by quantum trajectories .
in fact , it may well turn out that these methods are even more useful in solid state devices since the sensitivity of electron detectors is typically far better than that for photon detectors , on the single quantum level .
as will be explained in the following , a quantum trajectory @xmath96 describes a subensemble of the full ( ensemble ) density operator @xmath97 , _ conditioned _ on a certain ( stochastic ) measurement record , here detection events at certain times . in this approach
we determine the dynamics of an electron on a quantum dot , conditioned on the measured ( stochastic ) tunneling current through the dot .
quantum trajectory methods have changed remarkably the way we think about open quantum system dynamics . while traditionally an open quantum system is described by its density operator @xmath97 as in the last section ,
quantum trajectories describe open system dynamics taking into account certain continuous , stochastic measurement outcomes . in other words , with quantum trajectories one determines a _ conditioned _ density operator @xmath96 , reflecting knowledge obtained from a continuous monitoring of the environment .
sampling over all these possible measurement records , in other words , ignoring the state of the environment , one recovers the usual full ensemble @xmath97 . we write @xmath98 $ ] where @xmath99 $ ] denotes the ensemble mean over all possible measurement records with corresponding probability ( see below ) .
the principle idea is to monitor the environment rather than ignoring , i.e. tracing over it . in quantum optics one
tries to detect photons emitted from the quantum system of interest , here we detect electrons in the leads coupled to the quantum dot . in order to illustrate this approach , we consider a simplified open quantum system the generalization to the quantum dot case will be obvious .
this model system consists of two levels and is coupled to a continuum of states .
excitation is done by some additional mechanism , included in the hamiltonian of the system @xmath100 .
we start with a master equation of type ( [ masterlind ] ) , and in this model with a single lindblad operator @xmath101 , @xmath102 + \frac{1}{2 } ( [ l \rho , l^\dagger ] + [ l , \rho l^\dagger ] ) .\ ] ] in the following we abbreviate the right hand side of the equation with the superoperator @xmath103 . for concreteness ,
consider @xmath101 to describe a spontaneous transition from level @xmath104 to level @xmath105 with rate @xmath106 , i.e. @xmath107 .
we introduce the superoperator @xmath108 such that @xmath109 - \frac{1}{2 } ( l^\dagger l \rho + \rho l^\dagger l ) \\
\mathcal{s}\rho & = & l\rho l^\dagger.\end{aligned}\ ] ] the latter is referred to as the _ jump _ operator since its describes an emission process accompanied by the replacement of the density operator @xmath110 with the ground state : @xmath111 . with @xmath108 such
defined , one obtains the _ quantum jump representation _ @xcite of the solution of ( [ masterlind2 ] ) in the form @xmath112 clearly , the solution @xmath97 is a sum ( or integral , respectively ) over any number @xmath27 of emission processes ( number of projections onto @xmath113 due to the application of the jump operator @xmath108 ) , appearing at any times @xmath114 between zero and the current time @xmath115 .
one has to integrate over all corresponding ( unnormalized ) density operators @xmath116 , as apparent from expression ( [ quantumjump ] ) .
thus , one particular _
quantum trajectory _ is the normalized density operator @xmath117 which describes the time evolution of the quantum system _ conditioned _ on the particular measurement record , i.e. conditioned on the number and times of emission processes .
the quantum trajectory @xmath118 occurs with probability @xmath119 .
the normalized quantum trajectory @xmath96 may be determined directly through the following prescription : at time @xmath120 the new density operator @xmath121 is obtained in one of two ways : first , the probability @xmath122 , to undergo a quantum jump , i.e. to emit a quantum during the time interval @xmath123 is equal to the jump rate times the length of the time interval times the probability to be in the excited state : @xmath124tr@xmath125tr@xmath126 . if a quantum is emitted ( and thus detected ) , i.e. a jump has occurred , the conditioned quantum state is the ground state : @xmath127 .
if however , no jump occurs , the new density operator is given by @xmath128 as is apparent from the representation ( [ quantumjump ] ) .
in practice therefore , a quantum trajectory is obtained by determining a random number @xmath129 between zero and one in each time step @xmath123 : if @xmath130 , we set @xmath131 , if , however @xmath132 , we set @xmath133 . the full ensemble of possible states is thus given by @xmath134 and indeed , one may easily verify that the right hand side equals @xmath135 as expected from the master equation ( [ masterlind2 ] ) for the full ensemble .
this branching may occur at any time step and a thus huge ensemble of different quantum trajectories may be obtained . as mentioned before , the usual reduced density operator is obtained by taking the ensemble mean . in order to obtain counting statistics as in the following sections , we simply average over many runs and
obtain numerically a distribution of jump times as in a real experiment involving a single quantum system .
an electron spin on a quantum dot has been found useful as a memory device or a qubit for quantum information processing .
readout of the spin state through a tunneling current was investigated using a rather restricted parameter regime for which analytical results were obtained in ref . .
first we want to show , how the analytical results emerge very easily and directly from the quantum trajectory approach .
now we consider a regime where we can neglect spin flips , i.e. @xmath136 . as in ref .
we choose spin - polarized leads , @xmath137 , and @xmath138 . in the limit @xmath139
we then have @xmath140 .
the initial state is @xmath31 and since no spin flip occurs on the time scale of interest the only processes that happen are transitions between @xmath31 and @xmath36 .
the _ quantum jump representation _ ( [ quantumjump ] ) of this particular solution then reads @xmath141 with @xmath142 and @xmath143 . in the regime
chosen we can write @xmath144 and then get @xmath145 here every operator @xmath146 describes an electron tunneling onto the dot from lead one and @xmath147 represents the hopping of an electron from the dot into lead two .
since the initial state is @xmath31 , the first transition is @xmath148 and with a second transition back to @xmath31 the first electron accumulates in lead two .
a third transition to the singlet does not change the number of electrons in lead two . for a particular @xmath149 ( number of electrons in lead two )
we have to consider @xmath150 and @xmath151 and the ( unnormalized ) density operator for a certain @xmath149 at time @xmath115 is @xmath152 therefore , the probability to find exactly @xmath149 electrons in lead two at time @xmath115 is @xmath153 confirming the findings in ref . . with the general quantum jump representation ( [ quantumjump ] )
, we can overcome the limitations of the analytical result , considering arbitrary regimes and investigating the dynamics numerically .
so far , the proposed measurement scheme allows one to deduce the probability to be in either of the two spin states from the current through the dot . a relative phase between @xmath29 and @xmath31 ,
however , can not be detected . in order to measure the full spin state , therefore ,
a tomographical measurement setup is required . here
, the freedom to apply the esr field comes into play .
we show that while applying an esr field , phase - sensitive counting statistics result , leading to clear identification of the qubit state on the bloch sphere . as in quantum optical setups , the full state could also be obtained with appropriate @xmath154-pulses , that effectively change the measurement axis . in this way ,
not only the @xmath155-component as in the original proposal , but also @xmath156 and @xmath157 and thus the full @xmath110 can be measured .
a simpler concept , not involving these precise pulses , is to measure the spin state via counting statistics of a current through the dot in conjunction with a constant esr field as we will show in the following . for this scheme to be successful
it is crucial to control the interaction between dot and leads .
we are not interested in the asymptotic , stationary distribution , but in the typical time between switching the coupling on and the first ( or second , or third and so on ) electron appearing in lead two .
also , it is not necessary to be able to measure the electrons in lead two with a high temporal resolution : one can switch off the coupling between dot and lead two after a certain time @xmath115 and has any time thereafter to collect the electrons in lead two .
we note that different measurement schemes are possible .
since an electron tunneling onto the dot already carries the information about the spin of the dot electron , one could abandon lead two altogether and try to monitor the number of electrons on the quantum dot , e.g. with a quantum point contact .
our proposal for quantum state tomography could be transferred to other setups as well , as for recent experiments.@xcite we assume that the dot is in a given initial state at @xmath158 , when the coupling to the leads is switched on
. then we measure the number of electrons tunneling into lead two . according to the quantum trajectory approach
we calculate the evolution of the density matrix .
every jump from @xmath36 to @xmath31 or @xmath29 indicates that an electron tunneled out of the dot . at very low temperatures ,
as assumed throughout this paper , the probability of tunneling into lead two is close to unity , while tunneling into lead one is very unlikely .
a single run of the stochastic evolution will display emission processes , i.e. contributions to the current , at certain random times .
counting the corresponding number of quanta in lead two as a function of time for a large ensemble of quantum trajectories allows us to determine the probability @xmath159 of finding exactly @xmath149 electrons in lead two at time @xmath115 for a given initial state of the dot .
such counting distributions are displayed in the following figures .
we stress again that experimentally , it is not required to be able to measure these arrival times with a high resolution .
one simply switches off the coupling between dot and lead two after a given time @xmath115 .
then one has any time to determine the number of electrons in lead two .
our numerical procedure can be applied to any parameter values and any time dependence of the driving esr field . for the regime chosen in ref .
, we recover the analytical results ( [ quantumjump5 ] ) to a high degree of precision , as will be shown below .
let us now turn to the probability distributions @xmath159 of finding exactly @xmath149 electrons in lead two at time @xmath115 for a given initial state @xmath110 .
as we will show , by employing the esr field , the counting statistics allows to clearly identify the full two - level state , including the relative phase . as usual
, we choose to parametrise the latter through the coordinates on the bloch sphere : the spin up state @xmath29 corresponds to the north pole with @xmath160 , while the spin down state @xmath31 has coordinates @xmath161 .
the full mixture @xmath162 corresponds to the center of the bloch sphere , @xmath163 while coherent superpositions @xmath164 live on the equator with @xmath165 .
counting statistics @xmath166 for the first electron with initial spin up state ( dashed , @xmath167 ) , spin down state ( dotted , @xmath168 ) and the totally mixed state ( full curve , @xmath163 ) .
the respective coordinates refer to the bloch sphere .
the inset shows the same curves with the counting statistics of the fully mixed state subtracted .
parameters chosen are @xmath169,scaledwidth=50.0% ] in order to be able to use these counting statistics as a method for spin state tomography , the right choice of parameters is crucial . from eqs .
( [ rwa1],[rwa2],[rwa3],[rwa4 ] ) it is obvious that coherences in the two - level state can only be transferred to measurable probabilities through the coupling introduced by the esr field of magnitude @xmath170 . on the other hand , a large value of @xmath170 leads to rabi oscillations and thus prevents us to distinguish clearly the two fundamental @xmath29 and @xmath31 states on a time scale large compared with the rabi frequency @xmath171 .
closer inspection of eqs .
( [ rwa1 ] , [ rwa2 ] , [ rwa4 ] ) and numerical evidence shows that a good phase sensitivity with preserved distinguishability of @xmath29 and @xmath31 is achieved through the choices @xmath172 physically , the first condition ( on the esr field strength ) means that the spin should not be flipped to fast ( compared with the measurement time scale @xmath173 ) but still , the esr field had time enough to make the coherences felt . the second condition ( on the esr field frequency )
ensures that the method is sensitive to all values of the phase angle @xmath174 . in fig .
[ downuposp ] we show counting statistics for the first electron @xmath166 to appear in lead two .
we choose the transition rate @xmath175 , an experimentally accessible magnetic field strength@xcite of the esr field @xmath176 , a slightly detuned esr field frequency @xmath177 , a temperature @xmath178 , and a static magnetic field of strength @xmath179 . for the esr field to start at zero we choose the fixed phase @xmath180 .
furthermore , we assume @xmath181 and @xmath182 for the intrinsic relaxation and decoherence times .
all figures are calculated with an ensemble of 50000 trajectories .
the spin down state only allows for electrons to tunnel through the dot , which is clearly visible in the counting statistics : if the spin starts off in the spin up state ( dotted curve ) the time to measure the first electron is delayed compared to the mixture and even more so compared to the spin down state .
eventually , however , due to the presence of the esr field , a sufficient spin down component will be established allowing electrons to tunnel through the dot . still
, both states are clearly distinguishable through their counting statistics .
counting statistics @xmath166 for eight coherent superpositions @xmath164 along the equator of the bloch sphere ( @xmath183 , various angles @xmath174 ) and the fully mixed state ( @xmath163 , full line ) .
the inset shows the same curves with @xmath166 of the full mixture subtracted .
same parameters as in fig .
[ downuposp].,scaledwidth=50.0% ] not only are counting statistics useful to distinguish between up and down state .
the arrival time distribution also differentiates between coherent superpositions and mixtures . in conjunction with the esr field one
may even determine the phase of coherent superpositions of type @xmath184 as displayed in fig .
[ downuposp ] .
the full line corresponds to @xmath166 of a fully mixed initial state ( @xmath163 ) , the dashed and dotted lines correspond to eight coherent superpositions along the equator of the bloch sphere .
clearly , @xmath166 shows different behavior for different angles @xmath174 and may thus be used to fully identify the initial state .
as we have seen , with these choices for the esr field , not only we can distinguish @xmath29 from @xmath31 through counting statistics as in fig .
[ downuposp ] .
we are in a position to fully determine the two - level state in particular , it is possible to clearly distinguish a coherent superposition of @xmath29 from @xmath31 from the mixture of the two , as shown in fig .
[ equatorosp ] .
the insets of figs .
[ downuposp ] and [ equatorosp ] reveal an interesting structure underlying the shapes of @xmath166 : once the counting distribution of the full mixture ( @xmath163 ) is subtracted , statistics of states corresponding to opposite points on the bloch sphere appear as mirror - images of each other , as highlighted in figs .
[ diff_paarweise0 ] and [ diff_paarweise ] . in these figures
we display the counting statistics @xmath166 for four pairs of opposite initial states along the equator of the bloch sphere and clearly confirm the observations just mentioned .
a linear combination of initial states leads to a linear combination of counting statistics in the ensemble and thus to this symmetry .
still , each curve in itself seems complicated enough to underline the importance of our numerical approach . using the quantum trajectory method , any time dependence of the fields and any choice of parameters is possible .
graphs taken from fig .
[ equatorosp ] for four pairs of opposite states along the equator of the bloch sphere.,scaledwidth=80.0% ] same as fig .
[ diff_paarweise0 ] with @xmath166 of the full mixture subtracted .
we clearly see the symmetry of the curves for opposite states on the bloch sphere.,scaledwidth=80.0% ] the more mixed the initial state , i.e. the smaller @xmath185 on the bloch sphere , the closer the curve to the curve of the fully mixed state .
it is also worth noting that we keep the initial phase @xmath48 of the esr pulse fixed for all calculations .
an average over all possible phases would indeed lead to the graph of the fully mixed state , irrespective of the phase @xmath174 of the initial quantum state .
we close this section by pointing out that also higher order counting statistics ( @xmath187 ) display state - sensitive behaviour if only less pronounced .
this is quite obvious since a delayed first tunneling event shifts the starting time for the following electrons . for @xmath186 , the difference between the counting statistics for various initial states is well pronounced . in this latter case , however , the curves do not cross which diminishes the distinguishability of states and the best choice for that is @xmath188 . as displayed in fig .
[ qdiagram ] , higher - order counting statistics @xmath159 still distinguishes between the fully mixed state ( @xmath163 ) and a coherent superposition ( @xmath189 ) .
counting statistics of exactly zero , one , five electrons tunneling through the dot . evidently ,
if less pronounced , all counting statistics @xmath190 carry information about the initial quantum state.,scaledwidth=60.0% ] the original proposal for the spin state readout was based on spin - polarized leads in order to clearly distinguish the two states @xmath29 , @xmath31 by a single shot measurement . as the counting statistics require an ensemble measurement
, our results suggest that spin polarization is not required for those not even advantageous , in fact . in fig .
[ equatormsp ] we display counting statistics @xmath166 for spin - polarized leads ( only spin - up electrons in the leads , i.e. @xmath84 ) .
we notice only marginal differences compared to the case of unpolarized leads ( fig .
[ equatorosp ] ) . for large times , it is more likely to observe precisely one electron in the case of unpolarized leads .
the reason for this behaviour is the fact that for unpolarized leads , there is also the possibility that the spin - up electron on the dot ( rather than the spin - down electron entering the dot ) may tunnel out of the dot .
then the dot is in the ground state and therefore closed for the tunneling of another electron .
it is only after the esr field had time to populate the excited state that a second electron may tunnel through the dot .
in fact , it turns out that this mechanism is the preferred tunneling event : for the parameters of fig .
[ downuposp ] we find @xmath191 and @xmath192 .
same graph as fig .
[ equatorosp ] . here
, however using spin - polarized leads with @xmath193 , @xmath84.,scaledwidth=50.0% ]
our proposed setup including the esr field may be used to determine the intrinsic relaxation time @xmath194 and decoherence time @xmath195 of the qubit in the time domain . tuning the tunneling rate over a wide range ( and adjusting the esr field strength and frequency according to conditions ( [ readoutconditions ] )
, one can easily see the effect of decoherence and relaxation . in the series of graphs in fig .
[ wscan ] we show counting statistics for the @xmath31 and @xmath29-state , for the full mixture and for two coherent superpositions ( states on the equator of the bloch sphere ) .
clearly , for large tunneling rate ( left graph , ( a ) ) , all states may be distinguished .
the third graph ( c ) shows a regime where decoherence has fully set in : while the states @xmath31 and @xmath29-state remain essentially unaffected , the counting statistics of the coherent superpositions collapse onto the curve of the full mixture . in other words , while no relaxation has set in yet , coherences between the states @xmath31 and @xmath29 have disappeared . decreasing the tunneling rate even further , counting statistics finally reveal the relaxation time : eventually , the initial states @xmath31 and @xmath29 may no longer be distinguished , i.e. the relaxation has taken place .
measurement of a variety of states ( @xmath29 , @xmath31 , coherent superposition ) facilitates an estimate of @xmath195 and @xmath194.,scaledwidth=90.0% ]
counting statistics was used in ref . to determine the measurement time and measurement efficiency . it was shown that after about ten times the tunneling time , the spin state on the dot could be determined to be either @xmath29 or @xmath31 with close to 100% efficiency , even if only a single measurement is made ( provided it was either @xmath29 or @xmath31 ) .
crucially , these results are based on spin - polarized leads . without spin - polarization
the determination of the dot state with a single measurement appears problematic .
the reason is that an electron may tunnel from the quantum dot back into lead one rather than into lead two .
in fact , since in the case of unpolarized leads there are three ways for an electron to leave the dot : it may tunnel from lead one onto the dot and further into lead two ( transition @xmath196 ) or the electrons interchange the roles and the one residing on the dot tunnels into either of the two leads ( @xmath197 ) .
the latter two possibilities are almost of same probability and therefore with a probability of about one third a tunneling process has taken place without having been observed in lead two . in other words , with a probability of two thirds only we can claim that without detecting an electron in lead two after sufficient time the spin state on the dot was @xmath29 . in order to overcome this problem we propose to measure the number of electrons on the dot with a quantum point contact and recent experiments @xcite
show that such a concept may work .
if an electron tunnels onto the dot it confirms that the state was @xmath31 and if it stays there sufficiently long the qpc as electrometer recognizes the charge . since the projection takes place
if an electron tunnels onto the dot , the second lead could even be omitted and then the dot electron would tunnel into lead one . for clarity in the following argument , however , we do not change our setup and leave lead two as before . with the help of the qpc it is possible to measure the state of a single electron spin with an efficiency of almost 100% after about ten times the tunneling time , as for polarized leads as shown in fig .
[ proj_nc ] .
note that for this scheme to work , it is important that the time resolution of the point contact measurement has to be better than the inverse tunneling rate @xmath198 to ensure the detection of the tunneled electron on the dot .
with spin - polarized ( s - p ) leads , a detection of an electron in lead two after about 10 times the tunneling time ensures to almost 100% that the initial dot state was @xmath31 . without spin polarization
, only a probability of about two thirds may be achieved due to the possibility of the electron to tunnel into lead one .
the numerical curve accounts for spin relaxation and therefore differs slightly from the analytical result . without spin relaxation
no difference can be seen .
parameters as in fig .
[ downuposp ] respectively fig .
[ equatormsp ] , but without esr - field.,scaledwidth=50.0% ] measuring the probability to find two electrons on the dot one may determine the dot state with 100% efficiency . clearly , the time resolution has to be better than the typical time the two electrons stay on the dot , i.e. better than @xmath198 . here
@xmath199}$ ] is the probability that no tunneling event occurred in the whole interval @xmath200 $ ] and @xmath201 is the probability to find exactly @xmath149 electrons on the dot at time @xmath115 .
the initial state was always @xmath31 , parameters as in fig .
[ proj].,scaledwidth=50.0% ]
we use quantum trajectory methods to investigate counting statistics of electrons tunneling through a quantum dot .
we show how an additional esr field may actively be used to perform a full `` state tomography '' .
applying the field during the measurement allows one to clearly identify the coherences between the two superposed states .
we discuss the relevance of our findings for determining intrinsic relaxation and decoherence times of electron spin states in quantum dots in the time domain . based on a quantum point contact
we propose a scheme for single - shot readout without the need for spin - polarized leads .
similarities of the investigated quantum dot to three - level systems in quantum optics ( e.g. so - called `` v''-systems ) are evident .
we underline these connections by applying the quantum jump method in order to _ unravel _ the dynamics of the full density operator @xmath97 into subensembles @xmath96 corresponding to certain measurement records in the leads .
thus , we describe the _ conditioned _ time evolution of the spin state , given a certain measurement record , as in actual experiments with single quanta .
we believe that such connections between methods of quantum optics and mesoscopic devices will prove more and more useful in the future as nanotechnology achieves further breakthroughs in the coherent manipulation of quantum dynamics in solid state devices .
l. m. k. vandersypen , r. hanson , l. h. willems van beveren , j. m. elzerman , j. s. greidanus , s. de franceschi , and l. p. kouwenhoven , in _ quantum computing and quantum bits in mesoscopic systems _ , edited by a. leggett , b. ruggerio , and p. silvestrini , ( kluwer academic / plenum , new york , 2003 ) ; see also quant - ph/0207059 , ( 2002 ) . | we investigate the dynamics of electron spin qubits in quantum dots .
measurement of the qubit state is realized by a charge current through the dot .
the dynamics is described in the framework of the quantum trajectory approach , widely used in quantum optics , and we show that it can be applied successfully to problems in condensed matter physics .
the relevant master equation dynamics is unravelled to simulate stochastic tunneling events of the current through the dot .
quantum trajectories are then used to extract the counting statistics of the current .
we show how , in combination with an electron spin resonance ( esr ) field , counting statistics can be employed for quantum state tomography of the qubit state .
further , it is shown how decoherence and relaxation time scales can be estimated with the help of counting statistics , in the time domain . finally , we discuss a setup for single shot measurement of the qubit state without the need for spin - polarized leads . |
the solar system today is embedded in a warm low density interstellar cloud ( t@xmath27000 k , n(h@xmath5+h@xmath6)@xmath20.3 @xmath3 ) , which flows through the solar system with a relative sun - cloud velocity of @xmath226 km s@xmath7 .
neutral interstellar gas penetrates the charged solar wind of the heliosphere
98% of the diffuse material in the heliosphere is interstellar gas , and the densities of neutral interstellar gas and the solar wind are equal at approximately the orbit of jupiter .
the galactic environment of the sun is regulated by the properties of the interstellar cloud surrounding the solar system . however , when the surrounding cloud is of low density , the solar wind prevents most interstellar gas and dust from reaching 1 au , the location of the earth .
the discovery of small scale structure with column densities @xmath8 10@xmath9 @xmath10 in cold interstellar matter ( @xcite ) , and the structured nature of the interstellar cloud surrounding the solar system , allow the possibility that the spatial density of the interstellar cloud surrounding the solar system may change within the next 10@xmath1110@xmath12 years ( frisch 1995,1997a,1997b,1998 ; hereafter referred to as fr ) . over the past century
, many conjectures have appeared in the scientific literature linking encounters with dense interstellar clouds to possible climate changes on earth ( e.g. shapely 1921 ; mccrea 1975 ; begelman and rees 1976 ; fahr 1968 ; reid et al .
1976 ; mckay and thomas 1978 ; scoville and sanders 1986 ; thaddeus 1986 ; @xcite ; @xcite , @xcite ) . for these suggestions to have substance , however , it must first be shown that the interplanetary environment of the earth varies with changing properties of the surrounding interstellar cloud .
it has been shown that in the past , the galactic environment of the sun has changed as a function of time , and that the cloud complex sweeping past the sun now has an order - of - magnitude more nearby interstellar gas in the upwind than the downwind directions ( @xcite , fr ) .
therefore the sensitivity of the heliosphere to variations in the boundary conditions imposed by the lism justify closer examination .
it is the purpose of this paper to show that even a moderate alteration in the density of the cloud surrounding the solar system can yield substantial variations to the interplanetary environment in the inner heliosphere .
early studies investigating a heliosphere embedded in a dense interstellar cloud considered the relative ram pressures of the solar wind and surrounding interstellar cloud to estimate the heliopause location ( e.g. @xcite ) .
contemporary models consider the interaction of the solar wind and interstellar medium ( ism ) self - consistently , by including the effects of resonant charge exchange between the ionized and neutral gases . in the supersonic solar wind itself
, charge - exchange can lead to a significant deceleration of the wind due to the freshly ionized interstellar neutrals extracting momentum from the solar wind .
the concomitant reduction in solar wind ram pressure can lead to a significant reduction in the size of the heliospheric cavity . in the boundary region separating the solar wind from the ism ( the `` heliosheath '' ) , neutral hydrogen charge exchange with decelerated interstellar plasma acts to partially divert , heat and filter the h@xmath0 before it enters the heliosphere .
this filtration of h@xmath0 in the heliosheath can reduce the number density of inflowing h@xmath0 by almost half .
the rather complicated nonlinear coupling of plasma and h@xmath0 in the vicinity of a stellar wind is now captured in modern models ( @xcite , zank et al .
, 1996a , or see e.g. , zank 1998a for a review ) .
the weak coupling of neutral hydrogen gas and plasma via resonant charge exchange affects both distributions in important ways .
this implies that the self - consistent coupling of plasma and neutral hydrogen is necessary for modelling the interaction of the solar wind with the ism .
we employ a self - consistent two - dimensional ( 2d ) numerical simulation to evaluate heliospheric structure and properties when the heliosphere is embedded in a neutral interstellar cloud whose number density is some thirty times greater than at present .
lcccc & & + + & `` lic '' & solar wind & model & model + & & & plasma & neutral h + & ism & ( 1 au ) & ism & ism + @xmath13 ( @xmath14 ) & 0.2 & & & 10 + @xmath15 ( @xmath14 ) & 0.10.25 & 5.0 & 0.1 & + @xmath16 ( @xmath14 ) & 0.10.25 & 5.0 & 0.1 & + @xmath17 ( km s@xmath7 ) & 26 & 400 & 26 & 26 + @xmath18(k ) & 7000 & @xmath19 & 8000 & 8000 + @xmath20 & & 7.6 & 1.75 & 2.48 + +
a multi - fluid model is used here to model the interaction of the solar wind with cloud of enhanced density .
the need for modelling the neutrals as a multi - fluid stems from the variation in the charge exchange mean - free - path for h in different regions of the heliosphere and ism .
large anisotropies are introduced in the neutral gas distribution by charge exchange with the solar wind plasma ( both sub- and supersonic regions ) and the multi - fluid approach represents an attempt to capture this characteristic in a tractable and computationally efficient manner .
we consider the interaction of the heliosphere with an interstellar cloud similar to the cloud now surrounding the solar system , but with neutral number densities increased to 10 @xmath3 .
the parameters of the enhanced density cloud , and the solar wind at 1 au , are given in table 1 .
the relatively high neutral density ensures that the h@xmath0 distribution is essentially collisional
. williams et al .
( 1997 ) , using the results of dalgarno ( 1960 ) , fitted the function @xmath21 to describe the cross - section for h@xmath0-h@xmath0 collisions ( @xmath22 is the neutral atom energy in electronvolts ) .
the collisional mean - free - path for h@xmath0 with the given input parameters ranges from less than 2 au in the ism to less than 1 au in the heliospheric boundary regions ( below ) to less than 2.5 au in the heliosphere itself .
this suggests that the multi - fluid description outlined below is suitable on scales larger than a few au for these moderate ism densities , and this , as illustrated below , is shown to be the case .
the heliosphere - ism environment can be described in terms of three thermodynamically distinct regions ; the supersonic solar wind ( region 3 ) , the very hot subsonic solar wind ( region 2 ) , and the ism itself ( region 1 ) .
each region acts a source of secondary h@xmath0 atoms whose distribution reflects that of the plasma distribution in the region .
accordingly , the neutral distribution resulting from the interaction of the solar wind with the surrounding interstellar cloud may be approximated by three distinct neutral components originating from each region ( zank et al .
each of these three neutral components is represented by a distinct maxwellian distribution function appropriate to the characteristics of the source distribution in the multi - fluid models .
this approximation allows the use of simpler production and loss terms for each neutral component .
the complete highly non - maxwellian h distribution function is then the sum over the three components , @xmath23 in principle , for each component , an integral equation must be solved ( hall 1992 ; zank et al .
1996b ) . instead , zank et al . use ( [ eq:3.1 ] ) to obtain three boltzmann equations corresponding to each neutral component
this is an extension of the procedure developed in pauls et al .
for component 1 , both losses and gains in the interstellar medium need to be included , but only losses are needed in the heliosheath and solar wind .
similarly for components 2 and 3 . thus , for each of the neutral hydrogen components @xmath24 ( @xmath25 or 3 ) @xmath26 and @xmath27 means that the production or source term @xmath28 is to be evaluated for the parameters of components 1 , 2 , or 3 respectively .
the @xmath29 and @xmath30 terms describe losses by either charge exchange or photoionization respectively . under the assumption that each of the neutral component distributions is approximated adequately by a maxwellian ,
one obtains immediately from ( [ eq:3.2 ] ) an isotropic hydrodynamic description for each neutral component , @xmath31 & = & { \bf q}_{mi } ; \label{eq:3.4 } \\
\frac{\partial}{\partial t } \left ( \frac{1}{2 } \rho_i u_i^2 + \frac{p_i}{\gamma - 1 } \right ) + \nabla \cdot \left [ \frac{1}{2 } \rho_i u_i^2 { \bf u}_i + \frac{\gamma}{\gamma - 1 } { \bf u}_i p_i \right ] & = & q_{ei } . \label{eq:3.5}\end{aligned}\ ] ] the source terms @xmath32 are listed in pauls et al .
( 1995 ) and zank et al .
( 1996a ) .
the subscript @xmath24 above refers to the neutral component of interest ( @xmath33 ) , @xmath34 , @xmath35 , and @xmath36 denote the neutral component @xmath24 density , velocity , and isotropic pressure respectively , @xmath37 the unit tensor and @xmath38 ( @xmath39 ) the adiabatic index .
the plasma is described similarly by the 2d hydrodynamic equations @xmath40 & = & { \bf q}_{m p } ; \label{eq:3.8 } \\ \frac{\partial}{\partial t } \left ( \frac{1}{2 } \rho u^2 + \frac{p}{\gamma - 1 } \right ) + \nabla \cdot \left [ \frac{1}{2 } \rho u^2 { \bf u } + \frac{\gamma}{\gamma - 1 } { \bf u } p \right ] & = & q_{e p } , \label{eq:3.9}\end{aligned}\ ] ] where @xmath41 denote the source terms for plasma density , momentum , and energy .
these terms are also defined in pauls et al .
( 1995 ) and zank et al .
( 1996a ) .
the remaining symbols enjoy their usual meanings .
the proton and electron temperatures are assumed equal in the multi - fluid models .
the coupled multi - fluid system of equations ( [ eq:3.3 ] ) ( [ eq:3.9 ] ) are solved numerically as described in pauls et al .
( 1995 ) and zank et al .
( 1996a ) .
the global structure of the heliosphere embedded in a high density environment is obtained by integrating the coupled time - dependent system of hydrodynamic equations ( [ eq:3.3 ] ) ( [ eq:3.5 ] ) and ( [ eq:3.7 ] ) ( [ eq:3.9 ] ) numerically in two spatial dimensions .
the integration begins at time @xmath42=0 using as an initial condition the heliosphere embedded in a low density cloud ( zank et al .
the lism neutral number density is increased to 10 @xmath14 to mimic the `` encounter '' of the heliosphere with a denser cloud , and it is shown below that the heliosphere then fails to settle into a steady - state or equilibrium configuration .
a time sequence of the 2d global plasma structure is illustrated in plate 1 .
the time dependence of both the plasma and neutrals is presented separately below .
four successive figures are shown , each separated from the preceding in time by @xmath43 days .
the color depicts the plasma temperature .
the sun is located at the origin and the interstellar wind is assumed to flow from the right boundary ( located at 1000 au ; a numerical semi - circle of radius 1000 au forms the computational domain ) with the parameters listed in table 1 . since the plasma interstellar mach number is supersonic ( see zank et al .
1996a for a discussion about a subsonic ism ) , a bow shock , labelled bs in plate 1a , forms , which decelerates , heats and diverts the interstellar plasma around the heliospheric obstacle .
the plasma is now no longer collisionally equilibrated with the interstellar h@xmath0 and charge exchange between the interstellar plasma and neutrals just downstream of the bow shock leads to an effective heating / compression and diversion of the neutral interstellar h in the heliosheath region ( between the bow shock and heliopause ) . as a result , a pile - up of decelerated , diverted and heated interstellar h@xmath0 is created in the region downstream of the bow shock , which is illustrated in plate 2 .
plate 2 shows the corresponding global distribution of h@xmath0 at the same times as used in plate 1 , except that the h@xmath0 density is plotted .
far upstream of the bow shock , the h@xmath0 number density is @xmath44 @xmath14 , increasing rapidly just downstream of the bow shock in the pile - up region with densities 22.5 times greater than the ambient ism density .
changes in the neutral number density , velocity , temperature and mach number are seen more easily in the one - dimensional ( 1d ) plots along the stagnation axis than in plate 2 ( discussed later in the context of figure 2 ) .
the number density of neutrals entering the heliosphere at @xmath45 au is @xmath46 @xmath3 , decreasing to @xmath47 @xmath3 at 1 au .
the enhanced - density ism case can be compared to simulations using the current low - density surrounding ism ( the lic column of table 1 ) .
the hydrogen wall of the high density cloud model possesses a more complicated structure than that of the low density case it possesses a weak double - peaked structure along the stagnation axis in the upstream direction , reverts to a single - peaked structure away from but in the neighbourhood of the stagnation axis ,
after which it again becomes double - peaked .
the h@xmath0 flow is strongly filtered in the vicinity of the wall , this due essentially to the divergence of the h@xmath0 flow through charge exchange with the bow shock - diverted interstellar plasma , thus leading to an h@xmath0 number density entering the heliosphere that is much lower than that of the ism . as is well - known ( e.g. holzer 1972 )
, the effect of interstellar h streaming through the supersonic solar wind , and experiencing charge exchange with solar wind protons , is to reduce the momentum of the solar wind significantly .
the relatively large h@xmath0 number density entering the heliosphere reduces the solar wind ram pressure significantly , 25%75% , from its nominal non - mediated value thereby reducing the global extent of the heliosphere . for the parameters of table 1 , the heliosphere shrinks dramatically to @xmath44 14 au in the upstream direction .
the density of interstellar neutral hydrogen at 1 au becomes @xmath2 2 @xmath3 ( figure 2 ) , unlike the current state for which no neutral interstellar hydrogen reaches earth orbit ( e.g. @xcite ; zank et al .
, 1996a ) .
in addition , the configuration becomes unstable ( see below ) .
the reduction in solar wind momentum is accompanied by an increase in the total temperature of the solar wind ( solar wind ions plus pickup ions ) , in comparison to an adiabatically expanding solar wind , since pickup ions acquire a large velocity perpendicular to both the solar wind velocity and magnetic field vectors and therefore have energies that are typically @xmath48 kev in the unmediated solar wind ( where @xmath49 km s@xmath7 ) . accordingly , beyond the ionization cavity ( a cavity created by the photoionization of interstellar neutrals by solar wind uv radiation ) , the solar wind temperature increases ( in the absence of pickup ions , the total solar wind temperature decays adiabatically , the polytropic index depending on the nature of the specific heat sources identified to heat the wind ) .
radial cuts along the stagnation axis for the plasma density , velocity , temperature , and mach number are illustrated in figure 1 .
the increase in the solar wind temperature is evident in figure 1c .
by contrast , in the absence of pickup ions or other heating processes , the solar wind temperature would cool adiabatically according to @xmath50 for an adiabatic index of 5/3 .
the very high number density of interstellar neutrals entering the heliosphere , even after filtration ( @xmath46 @xmath14 , plate 2 , figure 2a ) leads to a very strong time - dependent mediation of the solar wind . in the sequence of velocity profiles plotted in figure 1b ( the sequence corresponds to the four plot sequence of plates 1 and 2 ) ,
the solar wind at the time corresponding to the solid line is decelerated from 400 km s@xmath7 to @xmath51 km s@xmath7 at the termination shock .
some 66 days later ( the dashed line ) , the deceleration is considerably larger , now down to @xmath52 km s@xmath7 .
neutrals that were ionized and then picked up by the solar wind during the least mediated solar wind period will therefore have considerably more energy than those picked up by a strongly mediated wind ( a factor @xmath44 difference ) .
thus , the strongly mediated solar wind will possess a very large temperature gradient .
the temperature gradient can become so large that the pressure gradient , which normally decreases with increasing radial distance in weakly mediated solar wind , is now inward ( figure 3 ) , and it is this inwardly directed pressure gradient that is responsible for decelerating the outwardly flowing solar wind so strongly . as a result of the much stronger deceleration of the solar wind , the radial solar wind density profile departs from an @xmath53 dependence ( figure 1a ) on occasion .
once the solar wind is mediated strongly by pickup ions ( the dashed curve , figure 1 ) , the newly born ions are much less energetic and no longer contribute as strongly to the temperature and pressure build - up in the outer heliosphere . consequently , the outer heliosphere begins to cool , the inwardly directed pressure gradient is reduced , and the solar wind begins to revert to a less strongly mediated state . in the simulation ,
the periodic weak and strong mediation of the solar wind is very evident when the results are plotted in a movie format , and temperature fronts are seen to propagate continually out of the solar wind .
this can be seen in the temperature sequence of plate 1 .
the numerical simulations have been run for very long times , but no abatement of the temporal behaviour is observed . essentially , the number density of neutrals entering the heliosphere is far too large to allow the solar wind to ever settle into a steady or quasi - steady state . the large neutral number density forces the wind to deviate strongly from its zeroth - order unmediated state , which in turn drives strong variations in the heating effects introduced by pickup ions . associated with the periodic increase of plasma temperature in the outer heliosphere
is a decrease of the solar wind mach number .
the mach number , which decreases in a pickup ion mediated solar wind because of both the solar wind deceleration and the increased sound speed , normally remains larger than 1 ( supersonic ) .
however , the strongly mediated solar wind mach number decreases to 1 and less ( figure 1d ) . in this case , a termination shock is no longer necessary to bound the solar wind and a shock - free transition from supersonic to subsonic flow occurs ( plate 1b ) . as the outer heliospheric temperature begins to decrease again , the mach number increases and a termination shock is again present .
the sequence showing the presence of a termination shock , then a shock - free transition , followed by the reformation of the termination shock is exhibited in plate 1a , b , and c respectively .
finally , it should be noted that the heliopause region is unstable to a rayleigh - taylor - like instability .
this is discussed in more detail in zank ( 1998b ) but the effect is understood easily .
the coupling of neutral interstellar h to plasma introduces a frictional term proportional to @xmath54 in the momentum balance , where @xmath55 denotes the charge exchange cross - section , @xmath56 the neutral number density , @xmath57 the neutral velocity , and @xmath58 and @xmath59 the plasma variables density and velocity .
this expression is somewhat simplified compared to the full expressions used in the simulation .
the inwardly directed frictional term plays a similar role as the gravitational term in the classical rayleigh - taylor analysis ( e.g. chandrasekhar 1961 ) and the charge - exchange ion - neutral friction term can therefore destabilize the heliopause .
this leads to a somewhat unstable flow pattern in the upstream region , and some mixing of the solar wind and interstellar plasma . in plate 2 ,
the global neutral distribution is also time - dependent but much more weakly than the plasma .
plate 2 suggests that the overall structure of the heliospheric hydrogen wall does not change very significantly with time .
it is shown in this paper , using a coupled time - dependent system of hydrodynamic equations to describe the interaction between the heliosphere and a partially ionized interstellar cloud , that the configuration of the heliosphere and the interplanetary environment of the earth are sensitive to the density of neutrals in the interstellar cloud surrounding the sun . raising the neutral number density in the surrounding cloud to @xmath44 @xmath3 , and leaving the other parameters at current values ,
has a dramatic effect on the global spatial and temporal structure of the heliosphere .
the heliosphere in the upstream direction is reduced in extent to @xmath21014 au , compared to @xmath60 - @xmath2120 au currently .
interstellar h@xmath0 is filtered considerably as it enters the heliosphere , but some survives to penetrate successfully to 1 au where interstellar h@xmath0 densities are @xmath22 @xmath3 , unlike the current conditions for which interstellar hydrogen is ionized entirely a few au from the sun .
although we have changed only the density of the surrounding cloud in these simulations , denser interstellar clouds are expected to have somewhat higher magnetic field strengths than the low value found locally .
an average value of 7.3@xmath61 g has been found recently for neutral interstellar gas from zeeman splitting ( @xcite ) . were the interstellar magnetic field strength also to be increased to 7@xmath61 g in our simulations
, we would expect the heliopause radius to decrease possibly to a half to a third ( depending on magnetic field orientation ) of our computed radius , using a very crude balance estimate .
the heliospheric configuration becomes unstable ; properties of the solar wind in the outer heliosphere vary on timescales as short as @xmath43 days , and the termination shock disappears at times .
the formation , disappearance and reformation of the termination shock appears to be periodic with a timescale of @xmath62 days .
pickup ions dominate the thermodynamical character of the outer heliospheric solar wind completely .
solar wind temperatures , when pickup ions are included , can exceed @xmath19 k. an inward pressure gradient can develop , which causes the solar wind to decelerate far more than predicted by simple mass - loading arguments and , consequently , the density decrease with increasing heliocentric distance can deviate from the usual @xmath63 dependence .
the implications of high interstellar neutral densities at 1 au for the terrestrial environment are interesting ( e.g. , begelman and rees , 1976 ; @xcite ) .
is there evidence that the earth has passed through a dense interstellar cloud in the ( pre-)historical past , or might do so in the near future ?
a cloud condensation with density @xmath210 @xmath3 embedded in the interstellar cloud complex flowing past the solar system could have a thickness of up to 0.3 pc and not violate any known observational constraints .
such a condensation would pass over the solar system in less than 15,000 years .
there are several sets of data that indicate that the galactic environment of the sun may have changed recently .
the first set are geological records of nitrate layers @xmath2 11000 years before present , which have been attributed to energetic photons released by the vela supernova explosion ( @xcite ; also see @xcite ) .
the second set of data is absorption line measurements of a slightly denser and cooler interstellar cloud seen within 3 pc of the sun towards the stars @xmath64 cma and @xmath65 cma in the anti - apex direction ( see the discussion of the distribution of the nearby ism in frisch 1997b ) .
the second cloud at @xmath217 km s@xmath7 ( heliocentric velocity , sometimes referred to as the `` blue - shifted '' cloud ) has an electron density n(e@[email protected] @xmath3 , and temperature t=3600 k , in comparison with the values @xmath20.10.2 @xmath3 and 7000 k seen locally ( @xcite ) .
a third kind evidence is the cosmic ray record implied by the spikes in the @xmath67be record found in antarctic ice core samples at ages corresponding to 33 000 and 60 000 years ago ( the d1 and d2 events respectively ) .
raisbeck et al .
( 1987 ) and sonett et al .
( 1987 ) suggest that an increase in the cosmic ray flux on the earth s atmosphere can lead to an enhancement in the precipitated beryllium .
since supernova remnant shocks are thought to be responsible for accelerating galactic cosmic rays up to energies of @xmath68 ev per nucleii , sonett et al.suggested that the spikes might correspond to the passage of supernova shocks through the heliosphere .
however , frisch ( 1997a ) has argued that little observational evidence in the lism exists for the presence of a supernova shock at the times needed to explain the @xmath4be enhancements .
we suggest below an alternative interpretation of the @xmath4be enhancements .
there are two possible ways that the cosmic ray flux at 1 au can be increased by the interaction of the solar wind with a dense interstellar cloud .
the first is that the increased pickup ion population increases the anomalous cosmic ray population .
the second is that the reduction in the size of the heliospheric cavity indicates that galactic cosmic rays are no longer modulated significantly by an extended solar wind and so , at 1 au , the earth samples the full spectrum .
a more quantitative connection between the anomalous cosmic ray flux and a high density neutral interstellar cloud can be made using the model discussed above .
the pickup ion number density at the termination shock , located at @xmath45 14 au , may be estimated as @xmath69 @xmath3 ( after solving @xmath70 , under the assumptions that the solar wind number density @xmath71 on average , @xmath72 @xmath73 , and using figure 2a ) .
the injection of pickup ions into the anomalous cosmic ray component at a quasi - perpendicular termination shock has been considered by zank et al .
( 1996c ) and lee et al .
( 1996 ) . on the basis of such a shock `` surfing '' , or multiply reflected ion acceleration mechanism , zank et al . estimated an injection efficiency of @xmath74% for hydrogen , with correspondingly smaller values for the heavier pickup ions .
thus , the number density of anomalous cosmic rays at 10 au for a heliosphere embedded in a h@xmath0 interstellar cloud of density 10 @xmath3 is @xmath75 @xmath3 . to determine the cosmic ray density at 1 au
, we can use a simple convection - diffusion solution to the cosmic ray transport equation , @xmath76 $ ] where @xmath77 is the radial diffusion coefficient for cosmic rays in the heliosphere . using an estimate for
the radial mean free path @xmath78 au for 100 mv rigidity ions ( zank et al .
1998 ) yields @xmath79 @xmath14 .
this is a very high level indeed .
with such a large flux of anomalous cosmic rays incident on the atmosphere of the earth , one might therefore expect a noticeable enhancement in the @xmath4be flux measured on earth .
this effect will be further enhanced by the reduced modulation of galactic cosmic rays in the smaller heliosphere .
this alternative explanation for the @xmath67be spikes offers a natural explanation for the absence of @xmath4be enhancements prior to the d2 event in terms of the recent history of the galactic environment of the sun ( @xcite ) . during those periods
when the heliosphere is embedded in an interstellar cloud with appreciable densities of h@xmath0 , the anomalous and galactic cosmic ray fluxes incident on the earth s atmosphere will increase and enhance the level of precipitated @xmath4be . during the time prior to the d2 event ( 60 000 years ago ) , the sun was in the third - quadrant void where the heliosphere would have been at least as large as today ( @xcite ) , entering the local fluff cloud complex only within the last @xmath80 years .
the effect of a large population of h@xmath0 atoms filling interplanetary space at 1 au may well affect the interaction of the terrestrial magnetosphere with the solar wind , since interstellar he@xmath0 densities will increase to @xmath21 @xmath3 at 1 au for this increased density case and be comparable to the density of interstellar h@xmath0 atoms .
in addition , enhanced cosmic ray fluxes at 1 au may alter the global electric circuit , since the cosmic ray flux is the dominant source of conductivity in the lower atmosphere ( @xcite ) ; the global electrical circuit has been postulated to play a role in the terrestial climate ( tinsley 1994,1997 ) .
gpz is supported in part by an nsf young investigator award atm-9357861 , an nsf award atm-9713223 , a nasa award nag5 - 6469 , a nasa delaware space grant ngt5 - 40024 , and jpl contract 959167 .
pcf acknowledges the support of nasa grants nag 5 - 6405 and nag 5 - 7007 .
baranov , v.b . , and y.g .
malama , 1993 , j. geophys .
res . , 98 , 15 157 begelman , m.c . , and rees , m.j . , 1976 , nature , 261 , 198 brakenridge , g. r. 1981 , icarus , 46 , 81 bzowski , m. , h.j .
fahr , and d.rucinski , 1996 , icarus , 124 , 209 chandrasekhar , s. , 1961 , hydrodynamic and hydromagnetic stability , dover dalgarno , a. , 1960 , proc .
, 75 , 374 dieter , n. h. , welch , w. j. , romney , j. d. 1976 , , 206 , l113 fahr , h.j . , 1968 ,
space sci .
, 2 , 496 faison , m. d. , goss , w. m. , diamond , p. j. 1998 , , 192 , 4616 frail , d. a. , weisberg , j. m. , cordes , j. m. , mathers , c. 1994 , , 436 , 144 frisch , p.c .
, 1993 , apj , 407 , 198 frisch , p.c .
, 1995 , space sci .
, 72 , 499 frisch , p.c .
, 1996 , space sci .
, 78 , 213 frisch , p.c .
, 1997b , enrico fermi institute preprint no .
97 - 23 frisch , p.c .
, 1997a , space sci .
, in press .
frisch , p.c .
, 1998 , to be published in planetary systems the long view , eds . l. m. celnikier and j. tran than van , editions frontieres .
frisch , p.c . and
slavin , j. d. , 1996 , space sci .
, 78 , 223 frisch , p. c. , and york , d. g. 1986 , the galaxy and the solar system , ed.r .
smoluchowski , j. bahcall , and m. matthews , tucson , university of arizona press , 83 frisch , p. , dorschner , j. , geiss , j. , greenberg , m. , grn , e. , hoppe , p. , jones , a. , krtschmer , w. , landgraf , m. , linde , t. , morfill , g. , reach , w. , slavin , j. , svestka , j. , witt , a. , zank , g.p .
1998 , , submitted gry , c. , and dupin , o. 1996 , in science with the hubble space telescope - ii , p.472 , p.benvenuti , f.d.macchetto and e.j.schreier , eds.proceedings of a workshop held in paris dec 4 - 8 , 1995 .
hall , d.t .
, 1992 , phd thesis , univ .
arizona heiles , c. 1997 , , 481 , 193 holzer , t.e .
, 1972 , j. geophys .
res . , 77 , 5407 holzer , t.e . , 1989 , , 27 , 199 lallement , r. , bertin , p. , ferlet , r. , vidal - madjar , a. , and bertaux , j.l . 1994 , , 286 , 898 lee , m.a . , shapiro , v.d . , and sagdeev , r.z .
, 1996 , j. geophys .
res . , 101 , 4777 mccrea , w.h . , 1975 , nature , 255 , 607 mckay , c.p . , and thomas , g.e
. , 1978 , geophys .
lett . , 5 , 215 meyer , d. m. and blades , j. c. 1996 , , 464 , l179 murphy , e.m . , lockman , f.j . , 1997 ,
aas , 190 , 1703 pauls , h.l . , zank , g.p . , and williams , l.l . , 1995 , j. geophys .
, 100 , 21 595 raisbeck , g.m .
, 1987 , nature , 326 , 273 reid , g.c . , isaksen , i.s.a . ,
holzer , t.e . , and crutzen , p.j .
, 1976 , nature , 259 , 177 roble , r. g. 1985 , , 90 , 6000 scoville , n.z . , and sanders , d.b . , 1986 , the galaxy and the solar system , ed.r .
smoluchowski , j. bahcall , and m. matthews , tucson , university of arizona press , 69 shapley , h. 1921 , j. geology , 29 , 502 sonett , c.p . ,
morfill , g.e . , and jokipii , j.r . , 1987 , nature , 330 , 458 thaddeus , p. , 1986
, the galaxy and the solar system , ed.r .
smoluchowski , j. bahcall , and m. matthews , tucson , university of arizona press , 69 tinsley , b. a. 1997 , eos , 78 , 341 tinsley , b. a. 1994 , eos , 75 , 369 watson , j. k. and meyer , d. m. 1996 , , 473 , l127 williams , l.l . , hall , d.t . ,
pauls , h.l . , and zank , g.p .
, 1997 , apj , 476 , 366 zank , g.p .
, 1998a , space sci .
, in press zank , g.p . , 1998b , j. geophys .
, submitted zank , g.p . ,
pauls , h.l .
, williams , l.l . , and hall , d.t . , 1996a , j. geophys.res . ,
101 , 21 639 zank , g.p .
, pauls , h.l . , williams , l.l . , and hall , d.t . ,
1996b , in proc .
of solar wind 8 , 599 zank , g.p . , pauls , h.l . , cairns , i.h . , and webb , g.m . , 1996c , j. geophys.res . , 101 , 457 zank , g.p . ,
matthaeus , w.h . ,
bieber , j.w . , and moraal , h. , 1998 , j. geophys.res . , 103 , 2085
plate 1 : a time sequence of the plasma temperature distribution .
the color corresponds to the log[temperature ] .
the ordering a , b , c , d corresponds to a temporal separation from the preceding figure of @xmath43 days .
the four figures show an approximately full evolutionary cycle which is repeated on a @xmath62 day period .
see text for further details .
[ plate1 ] plate 2 : a time sequence ( corresponding to that of plate 1 ) of the global distribution of the neutral interstellar h density .
the color refers to the density measured in @xmath14 .
the hydrogen wall in the upstream direction is clearly visible , as is the effective filtration of h as it enters the heliosphere .
[ plate2 ] figure 1 : the plasma density ( a ) , velocity ( b ) , temperature ( c ) and mach number ( d ) along the stagnation axis .
four curves are presented on each graph , each corresponding to a cut through one of the figures of plate 1 . in each case
, the solid line corresponds to plate 1a , the dashed line to plate 1b , the dash - dotted line to plate 1c and the dotted line to plate 1d .
[ fig1 ] figure 2 : the interstellar neutral density ( a ) , velocity ( b ) , temperature ( c ) and mach number ( d ) along the stagnation axis .
four curves are presented on each graph , each corresponding to a cut through one of the figures of plate 2 . in each case
, the solid line corresponds to plate 2a , the dashed to plate 2b , the dash - dotted to plate 2c and the dotted to plate 2d .
[ fig2 ] figure 3 : the plasma pressure along the stagnation axis . as in figure 1 , four curves are plotted , each of which corresponds to a figure from plate 1 and the same line convention is followed as in figure 1 . | the interaction of the heliosphere with interstellar clouds has attracted interest since the late 1920 s , both with a view to explaining apparent quasi - periodic climate `` catastrophes '' as well as periodic mass extinctions . until recently , however , models describing the solar wind - local interstellar medium ( lism ) interaction self - consistently had not been developed . here
, we describe the results of a two - dimensional ( 2d ) simulation of the interaction between the heliosphere and an interstellar cloud with the same properties as currently , except that the h@xmath0 density is increased from the present value of @xmath1(h@xmath0)@xmath20.2 @xmath3 to 10 @xmath3 .
the mutual interaction of interstellar neutral hydrogen and plasma is included .
the heliospheric cavity is reduced considerably in size ( approximately 10 14 au to the termination shock in the upstream direction ) and is highly dynamical .
the interplanetary environment at the orbit of the earth changes markedly , with the density of interstellar h@xmath0 increasing to @xmath22 @xmath3 .
the termination shock itself experiences periods where it disappears , reforms and disappears again .
considerable mixing of the shocked solar wind and lism occurs due to rayleigh - taylor - like instabilities at the nose , driven by ion - neutral friction .
implications for two anomalously high concentrations of @xmath4be found in antarctic ice cores 33 kya and 60 kya , and the absence of prior similar events , are discussed in terms of density enhancements in the surrounding interstellar cloud .
the calculation presented here supports past speculation that the galactic environment of the sun moderates the interplanetary environment at the orbit of the earth , and possibly also the terrestrial climate . |
it is well known that t tauri stars ( tts ) vary on many timescales , both periodicially and irregularly .
while a few photometric or spectroscopic observations can give us valuable `` snapshot '' information about the characteristics of these objects , more extended monitoring is needed to observe their full range of behavior and , ultimately , to understand it . by analyzing changes in light curves over several years ,
we hope to gain an improved understanding of the physical mechanisms affecting the light of tts .
while this endeavor requires large amounts of telescope time , ccd s have rendered such monitoring feasible with relatively small telescopes .
for over a decade , the 0.6 m telescope at van vleck observatory , located on the campus of wesleyan university , has been used to monitor several nearby extremely young open clusters .
this consistent coverage facilitates a detailed study of a significant number of tts over time .
here we present results based on five years of monitoring the nearby young open cluster ic 348 .
ic 348 is an ideal target for a variability study for several reasons .
it is both nearby and extremely young .
its distance is 260@xmath325 pc as calculated from _
parallaxes @xcite or 316 pc as determined by @xcite using main sequence fitting .
the uncertainty in distance contributes to an uncertainty in age . according to the models of @xcite ,
the median age of the pre - main sequence stars in ic 348 is between 1.3 and 3 million years .
photometric and spectroscopic observations of this cluster ranging across the electromagnetic spectrum have been amassed over the last decade .
these include photometry and spectroscopy in the near - infrared by @xcite and @xcite , and in the visual wavelength range by @xcite and @xcite
. a wealth of x - ray data has been obtained from _
rosat _ @xcite and _ chandra _ @xcite .
membership probabilities have been determined by @xcite from a proper motion survey , and a search for binaries using adaptive optics has been performed by @xcite .
results of the first six - month observing season of photometric monitoring of ic 348 at van vleck observatory were presented by @xcite . here
we present the results of four additional seasons of monitoring and discuss all of the data , with the aim of clarifying the nature of several different types of tts variability over this timespan .
in addition to identifying new periodic and irregular variables , it is now possible to examine gradual changes in their light curves which may not be detected by observations which range over only one or two years .
also , the determination of periods and phasing of light curves using observations from each season separately allows the first consistent observational investigation of the stability of tts rotation periods and a search for spot cycles .
while @xcite has claimed that tts may exhibit differential rotation as dramatic as that of the sun , more recent studies , both observational @xcite and theoretical @xcite , suggest that tts rotate essentially as rigid bodies across their surfaces .
one of the goals of this study , made possible by its extended duration , is to empirically determine or constrain the degree to which tts rotate differentially .
in addition , our data are useful for characterizing stars as wtts and ctts and for probing the distinctions between them . in section 2
we discuss the observations and initial data reduction . in section 3
we identify periodic variables and describe their characteristics .
non - periodic variables , including the unusual eclipsing star hmw 15 , are discussed in section 4 . in section 5 , we summarize our findings and suggest some areas which would be useful for further study .
the observations were obtained between 10 december 1998 and 26 march 2003 with a 1024 @xmath4 1024 photometrics ccd attached to the 0.6 m telescope at van vleck observatory , located on the campus of wesleyan university .
each pixel covers 0.6@xmath5 so our field of view is 10.2@xmath6 on a side . on each clear night
, a sequence of 5 one - minute exposures was taken through the cousins _ i _ filter , as well as twilight flats , bias frames , and dark frames . when possible , this sequence was repeated more than once per night .
preliminary reductions were accomplished using standard iraf tasks , and each set of five images was added together and shifted to the same position to within less than one pixel , creating one combined image with an effective exposure time of five minutes and an increased dynamic range .
a log of our observations is presented in table 1 .
the sample of 151 stars is the same identified by @xcite , but the coordinates given there were erroneous and have been corrected in table 2 . before performing differential aperture photometry , all images with seeing worse than the chosen aperture radius of 6 pixels ( 3.6@xmath5 )
were rejected .
seeing in the remaining images ranges from 1.5@xmath5 to 3.5@xmath5 , with a median value of 2.5@xmath5 .
photometry was performed using the apphot package in iraf , and the median level of the sky background was determined using an annulus with inner and outer radii of 10 and 15 pixels respectively .
there are some stars in our field whose photometry may be suspect because of their proximity to other stars , and a list of these objects can be found in @xcite . they are also noted in table 2 . since our observations range over a five year period , it was desirable to determine a set of stable comparison stars which could be used over this entire timespan .
the comparison stars used by @xcite ) did not fulfill this criterion , as they found that two of their comparison stars actually varied over small ranges . to determine a single , consistent set of comparison stars for the entire observing interval , we began by finding a set of stable comparison stars for one season alone ( 1999 - 2000 ) .
all stars which could not be measured photometrically on every night of observation were eliminated from the list of potential comparison stars , and an initial reference magnitude was defined by the remaining stars .
the remaining stars with the greatest variability were eliminated , and this process was repeated until a final list emerged with nine stars .
this procedure was repeated independently for a second season ( 2001 - 2002 ) and the two seasons had six comparison stars in common .
these six stars were subjected to our selection method for the remaining three seasons , and fortunately showed the same level of stability . using this method , which is not biased towards stellar characteristics such as spectral type or cluster membership , we emerged with stars 3 , 7 , 8 , 10 , 13 , and 17 as our set of comparison stars for all five seasons .
while two of the comparison stars ( 13 and 17 ) show a slight trend in their magnitudes over the five - year course of the observations , their influence on the resulting differential magnitudes is negligible for two reasons .
first , the two stars change magnitudes in opposite directions , both remaining within a range of 0.05 mag .
second , when averaged together with the other four comparison stars , deviations from season to season are diluted to a level of below 0.01 mag .
it is unclear whether these drifts are due to real variations of the stars or small changes in the wavelength dependence of the sensitivity function of the telescope , filter , and detection system over time .
all six comparison stars display a standard deviation of under 0.01 magnitudes in each season , and their stability can be understood _ a posteriori _ as a result of their physical characteristics .
five of the six stars are cluster members with spectral types earlier then g1 , implying that they are on or close to the main sequence and can be expected to be relatively stable photometrically , and the sixth star ( 17 ) has been identified as a nonmember by @xcite .
this set of comparison stars should , therefore , prove useful for future variability studies in this cluster .
differential magnitudes on the instrumental system ( _ i _ ) were computed for each star on each night relative to the average magnitude of the comparison stars .
these were then transformed to a standard i magnitude ( _ i _ ) adopting the color term derived by @xcite .
the adopted transformation is : @xmath7 average _ i _ magnitudes over the entire five years of observation are given in table 2 , along with the r - i colors used for the transformation .
r - i values are taken from @xcite , whose photometric system matches ours best , when available and from @xcite otherwise . for the three stars with known spectral types from the literature but no r - i colors ( 53 , 69 , and 88 ) , r - i values were taken as an average of several stars in our sample with similar magnitudes and spectral types . for the remaining 22 stars , r
- i was assumed to be 2.00 , which ( while less than ideal ) will suffice as an approximation since most of these stars are relatively faint ( _ [email protected] ) . stars with assumed , as opposed to measured , r - i values are indicated by parentheses in table 2 . the designations used here are cross - referenced with those used in recent studies by @xcite .
in addition to the average magnitude , we also give in table 2 the standard deviation ( @xmath9 ) for each star based on its photometric scatter .
these values allow us to identify non - periodic variable stars ( see section 4 ) , and such objects are signified by the label `` var '' in the column labeled `` period '' .
the time series data were also searched for periodicity , as will be described in section 3 .
when significant periodicity was detected , the period is given in table 2 .
in the case of stars for which slightly different periods were found in different observing seasons , which will be discussed in detail in section 3 , the given period is an average over all five years of observation .
note that periodic stars are easier to detect as variables than stars which vary irregularly because their variations are concentrated in a narrow band in frequency space .
therefore , all stars are searched for periodicity , not just those whose @xmath9 values mark them as variable .
to search for periodicity , the periodogram function was calculated for each star in our sample using the method of @xcite , as formulated by @xcite .
since our data ranges over a timespan of 4.5 years , this function could be calculated for each observing season separately as well as for the entire duration of the observations .
we used the following criteria to assess whether a star should be regarded as periodic .
first , it must have a peak in its power spectrum with a power greater than 7.0 , corresponding to a false alarm probability ( fap , as defined by @xcite ) below 2% .
second , it must possess one or both of the following attributes : i ) a well - defined light curve characterized by small scatter and accompanied by a strong peak in the periodogram in at least one season , or ii ) the highest peak in the periodogram ( exceeding a power of 7.0 ) recurring in multiple seasons .
based on these criteria , we have detected periods for 28 stars in our sample .
nine of these represent new detections , and the remaining 19 were discovered by @xcite .
it is interesting to note that our sample is remarkably unambiguous with respect to periodicity .
all stars which were periodic in one season were periodic , with very similar periods , in at least one additional season except for one ( star 47 , which is discussed in section 4 ) , and half of the periodic stars were so - identified in all five seasons .
[ seashist ] shows a histogram of the number of seasons in which a star was detected as periodic .
a search was also performed for stars with smaller peaks in their power spectra , falling slightly below our criteria , but which might recur .
this was done by multiplying the power spectra from all five seasons at each frequency and searching for peaks in the result .
no additional periodic stars were found through this method , supporting our choice of selection criteria for periodicity .
it appears that periods of periodic variables in ic 348 are very stable , as is quantified below , and that our procedure for identifying them is effective and efficient . the distribution of periods in our sample is shown in fig .
[ perdisthist ] along with the period distributions of the orion nebula cluster ( onc ) from @xcite and taurus @xcite .
the distributions shown are , in all cases , for stars with masses above 0.25 m@xmath10 , a criterion fulfilled by all the stars in our sample except for one ( star 42 ) .
the significance of this division , which is equivalent to a spectral type of m2 according to the models of @xcite , is that in the onc stars below this mass have shorter periods and show a unimodal rather than a bimodal period distribution @xcite . while the number of periodic stars in ic 348 and taurus - auriga are too small to be clearly distinguishable as unimodal or bimodal , the ic 348 distribution shares three characteristics with the onc and taurus samples .
these are : i ) an absence of periods shorter than one day , ii ) a relative deficiency of periods between four and five days , and iii ) a small number of periods greater than ten days .
more quantitatively , a two - sided kolmogorov - smirnov test @xcite reveals that there is a 65% chance that the period distributions of ic 348 and the onc came from the same parent population , and a 53% chance for ic 348 and taurus .
given the dramatic differences between the environments of the onc , the taurus - auriga association , and ic 348 , this similarity in period distributions has an interesting implication .
it appears that environmental factors such as stellar density or the presence of hot stars and ionized gas have little or no effect on the distribution of rotation periods for pre - main sequence stars with masses above 0.25 m@xmath10 . in fig .
[ rangevspt ] we show the full range of variability for the periodic stars as a function of spectral type .
it is clear that most of the periodic stars have spectral classes from k to early m , which is typical of tts in general .
the absence of later - type periodic stars is a selection effect .
they are too faint for us to obtain sufficiently reliable photometry at wesleyan to find periods .
the relatively small number of g - type periodic variables and their small amplitudes is in basic agreement with what is found for pms variability in general @xcite . in fig .
[ pervspt ] we show that there is an interesting correlation between rotation period ( or frequency = 1/p ) and spectral type in the sense that more massive stars rotate faster .
@xcite suggested a possible correlation between mass and rotation period , which is equivalent to a correlation between spectral type and rotation period . while the sample may be too small to be definitive , it is nonetheless a striking correlation .
a non - parametric ( spearman ) rank - order test @xcite indicates that spectral type is correlated with rotation in these data at a significance level of 99.7% .
@xcite showed that in the onc , stars with spectral types later than m2 also rotate faster , as a group , than the mid - k to m2 stars .
it appears , therefore , that mid - k and early m stars define a minimum in rotation rate among pms stars of onc and ic 348 age .
both g stars and stars later than m2 ( at least in the onc ) appear to spin faster , on the whole , although a wide range in spin rate certainly exists among pms stars of all spectral classes .
since disk - locking @xcite is widely regarded as the braking mechanism for pms stars , it appears that it is most effective in the mid - k to early m spectral regime , i.e. for masses @xmath10.5 m@xmath2 .
since our data extend over 4.5 years , it is possible to examine how the light curves of periodic variables evolve on that timescale .
very little work has been done with regard to this question , a notable exception being the wtts v410 tau @xcite .
one might expect both evolution of light curve shapes and changes in period , assuming tts have some differential rotation .
light curve shapes should evolve as the spots change size , shape , temperature , or location , and a latitudinal migration of spots could cause changes in period .
it would , of course , be very interesting to find any cyclic pattern which could be interpreted as evidence for a magnetic cycle .
for these reasons , we have carefully examined the data on our 28 periodic variables with respect to changes in period and light curve shape .
we began the analysis by searching for a single period which could be used to phase the entire data set for each star .
this was done by evaluating a single periodogram for each star based on all five observing seasons .
as noted above , we found that , in most cases , there was not a single strong peak at one frequency but several closely - spaced smaller peaks . in no case
could we find a period that produced a single convincing light curve when used to phase the data over the entire time span .
evidently , the spots on all of these stars evolve on timescales of less than a few years , affecting either the period or the phase of minimum light or both . in every case
, we found that light curves phased using periods determined from individual seasons independently were clearly superior to any combined light curve spanning all five seasons .
for this reason , we present the data as individual light curves for each season , plotted with the best period ( as indicated by the periodogram ) for that season .
if a star did not have a peak in its periodogram meeting our criteria in a particular season , a period from another season was used to phase the light curve , and that period is given in parentheses in the figures .
light curves for 10 of the 28 periodic stars in our sample are shown in fig .
we have chosen to show the most well defined or otherwise interesting light curves .
the full set of data may be obtained electronically by anonymous ftp to ftp.www.astro.wesleyan.edu .
the files are in the directory bill / ic348/. considering the most stable stars first , we note that 4 of the 28 periodic stars in our sample ( 16 , 49 , 69 , and 134 ) had identical periods detected in all five seasons . however , even these variables have light curves which show changes in shape , amplitude , and average magnitude between observing seasons .
star 16 is the best example of stability in our sample ( see fig .
[ lcs ] ) , and even in its case there are obvious changes in the shape of the light curve and the phase of minimum light from season to season which no refinement in the period can eliminate .
not a single star showed a precisely repeating light curve even from one season to the next , implying that the longest time that a given starspot configuration can remain stable enough to produce a completely coherent light curve is between @xmath10.5 - 1.0 years .
for some stars , periods are detected in some seasons but not in others ( e.g. star 12 in fig .
[ lcs ] ) . in these cases , the spot patterns on the stars evidently do not always remain stable for even the length of an observing season .
the stars most likely remain spotted , but the spot pattern must evolve such that a single period can not be determined over six consecutive months of observation .
it is clear from these data that spots evolve in terms of size , shape or distribution on typical timescales of weeks to months ( @xmath1few to tens of rotation periods ) , and we will now analyze the changes in starspot configuration in more detail . the changes in the light curves of the periodic stars in our sample can be accounted for by changes in starspot coverage and/or configuration , and one aspect of this phenomenon which remains relatively unexplored is cyclic behavior .
since the sun displays cycles of magnetic activity on multiple timescales , the occurence or lack of occurence of such a phenomenon on pre - main sequence stars may shed some light on the physical mechanisms operating within these young solar analogs . to investigate the extent to which spots on periodic tts might exhibit cyclic behavior , the average magnitude was compared with the amplitude of each periodic star in each season in fig .
[ spots ] . the movement of stars in the range-@xmath11i@xmath12 plane can be used to address the question of whether the total area covered by spots changes with time , as in the solar cycle , or whether spot redistribution is more important in changing amplitude . in general , the paths of the periodic stars over five observing seasons on fig .
[ spots ] are horizontal .
this stability of average magnitude from season to season most likely indicates that spots on tts tend to redistribute themselves and undergo changes in size , temperature and location rather than appear or disappear _ en masse_. if the spot coverage were dramatically different from season to season we would expect the mean magnitude of the stars to fluctuate more than is observed .
there are two stars , 12 and 30 , which display potentially cyclic behavior based on their light curves and their movement in the range-@xmath11i@xmath12 plane . their paths in fig .
[ spots ] are shown in bold , and the possible cyclical behavior is also evident in their light curves ( see fig .
[ lcs ] ) . in both cases , there is a gradual decrease in amplitude of variations , and in the case of star 12 , there is no detectable period in the 2000 - 2001 season before its periodic variations return in the remaining two seasons .
interestingly , when star 12 loses its periodicity , it remains stable at the fainter end of its magnitude range , not the brighter portion as would be expected based on a disappearance of cool spots .
it is possible , of course , that this could also be due to a redistribution of spots .
star 30 exhibits the greatest change in range in our sample , although this is primarily caused by only two observations in the 2001 - 2002 season . as with star 12 ,
its range variations make sense when viewed in the context of its light curves .
however , having seen only one episode of spot disappearance and reappearance in each of these cases , neither constitutes solid evidence for a repeating cycle .
although it is entirely possible that spot cycles on wtts are either too long or too short to be detectable in our data set , our findings are consistent with those of @xcite , whose models of rotating t tauri stars were not capable of producing the type of magnetic dynamo which would generate cyclical changes in spot configuration . since the sun exhibits spot cycles on multiple timescales , the shortest of which lasts approximately eleven years , continued monitoring will be useful to confirm and extend these results .
as discussed previously , it was not possible to find a single period to coherently phase together all the data of any star in our sample . in all but four cases , we found that different periods in different seasons were definitely required ( in the other four cases one need only adjust the phase of the light curves , not the period ) .
on this basis we may claim to have detected real changes in periods for most of these stars .
however , as is now discussed , the size of these changes is very small and , in fact , on the edge of what can be considered significant .
we can not rule out the possibility that changes in light curve shapes during a season cause most or all of the apparent changes in period detected .
the uncertainty in period for any season may be estimated using the method of @xcite .
he relates the error in angular frequency , @xmath13 , to the amplitude ( _ a _ ) , the standard deviation of the noise ( @xmath14 ) , the number of observations ( _ n _ ) and the timespan over which the observations were made ( _ t _ ) by : @xmath15 application of this formula to our data suggests that our periods are accurate to between 0.0029 days and 0.26 days , or between 0.13% and 1.58% .
the changes in period seen between seasons range from 0 to 2.4% and their cumulative distribution is shown in fig .
[ dperdist ] .
while the two largest changes in period would be considered significant based on the error calculated using kovacs formula , we are not entirely convinced that we have detected any real period changes .
examination of the light curves suggests that it is possible that phase changes are responsible for apparent changes in period .
it is well known that , on the sun , the rotation period changes by @xmath120% from the equator to the poles . the median observed period change for tts in our sample is less than 0.8% , suggesting that , unless the latitudes of the spots are confined to a very small range , tts are significantly more rigid rotators than the sun .
this is in agreement with suggestions by @xcite , whose models of rotating tts displayed differential rotation of under 1% , as well as observational studies by @xcite and @xcite . considering this evidence
, it seems at least possible , if not likely , that during an observing season , spot configurations can change such that the period which best fits a series of observations shifts very slightly , causing a small shift in the peak seen in the power spectrum .
we conclude that there is not yet any incontrovertible evidence for a period change in any tts . on the contrary
, periods appear to be remarkably stable on times of @xmath15 yr .
several classes of non - periodic pms variability have been identified by various investigators over the years , including fuors , exors , type ii and type iii ( uxors ) variations @xcite . in some cases the time scales for these variations are quite long years or even decades .
in addition , we might expect to find at least some examples of eclipsing variables in a pms sample , and it would be very important to identify them because of their potential to yield a great deal of otherwise unknowable information including basic stellar parameters such as mass and radius . in this section
we describe a search for non - periodic variables in our sample and discuss the properties of those discovered .
most are `` normal '' ( i.e. type ii ) irregular variables , typical of ctts . a couple of possible uxors
are also found but no fuor or exor candidates .
two unusual variable stars ( 15 and 47 ) have been discovered which merit particular attention and are discussed separately in what follows . to characterize the degree of variability of the stars in our sample , we have used the standard deviation @xmath16 measured over the entire timespan of our data . in fig .
[ sigmas ] we show log ( @xmath9 ) as a function of brightness . following
@xcite we identify two regions on this diagram .
stars brighter than _
i_@xmath114 have a constant minimum error ( @xmath10.007 mag ) dominated by systematic effects which are probably related to the flat - fielding process .
fainter stars have an increasing minimum error due to the random effect of photon statistics .
the solid line represents a fit to the minimum errors taken from @xcite , which also describes our expanded data set well .
four of the bright comparison stars fall below the minimum error line because their differential magnitudes are slightly correlated with their measured brightness on each night . to identify real variables we have adopted the criterion that a star must vary by more than three times the minimum @xmath9 .
this condition is shown on fig .
[ sigmas ] as a dashed line .
we regard stars above the line as likely to be real variables .
stars below it , especially fainter ones , may of course also be variables , but we do not have sufficient precision to distingiush them as such , unless they happen to be periodic variables . in that case , they will have already been detected as variables by the more sensitive periodogram technique discussed above . in all , there are 26 stars above the line , of which 10 are periodic variables , leaving 16 irregular " variables , whose variations are discussed in this section .
these stars are identified in table 2 by a var " in the column labeled period " .
a couple of light curves of special interest will be displayed here ; all of the data are available through anonymous ftp from ftp.astro.wesleyan.edu , as noted above .
a large majority ( 11/16 ) of the irregular variables identified in ic 348 by this study have the spectral characteristics of ctts @xcite .
only 2 are classified as wtts by @xcite , one is not yet classified and 2 are of g - type .
one of the g - type irregular stars ( 18 ) is only barely above the line , but the other ( 15 ) is extraordinary and discussed in a separate section below .
the two wtts ( 56 and 101 ) clearly behave in somewhat anomalous fashion given their spectra ; all of the other wtts above the variability line are periodic variables .
we have no explanation for the anomaly and simply note that these two stars may be worth looking at more closely in the future .
finally , one of the periodic tts ( 47 ) shows large amplitude irregular variability as well .
first we discuss the common type of variability ( ctts ) and then the behavior of the more exotic stars 23 and 47 .
finally , we briefly describe the amazing star 15 , which has been the subject of a separate contribution @xcite .
irregular variability was a defining characteristic of tts ( now ctts ) as originally described by joy ( 1942,1945 , 1947 ) .
our study strongly supports the perhaps now canonical view that unsteady accretion , presumably from a circumstellar disk , drives these variations . of the 13 known tts in our field which could definitely be identified as irregular variables , 11 are ctts .
this is a clear confirmation of the correspondence which exists between an indicator of active accretion ( h@xmath17 equivalent width ) and irregular variability . for the brighter end of our sample ( i @xmath0 14.3 ) , where we can be reasonably certain that we have identified all of the variable stars present in the field ,
there are 5 ctts and every one of them is an irregular variable . the observed correspondence between the spectroscopic signature of a ctts ( equivalent width of h@xmath17 @xmath8 10 ) and irregular photometric variability is clear and compelling in this cluster .
infrared excess emission is also commonly used as an indicator of the presence of a disk and might also , therefore , be expected to correlate with irregular variability .
infrared photometry of ic 348 stars is given by @xcite in the _ j_,_h _ and _ k _ bands .
these data can be used in combination with the long - term photometric data which we have amassed to calculate a measure of infrared excess , following @xcite : @xmath18 the observed _ i - k _ color is computed from our average _
i _ magnitudes ( taken over all five observing seasons ) and the _ k _ magnitudes of @xcite .
the intrinsic _ i - k _ color is taken from @xcite based on the spectral types given by @xcite or @xcite . the extinction is calculated based on the _ v - i _ color excess determined using the _ v _ magnitudes of @xcite or @xcite otherwise . again , the calibrations of @xcite were used , and the extinction law adopted by @xcite was used .
infrared excesses were calculated for all stars in our sample which had a spectral type determined by @xcite or @xcite .
[ sigik ] shows variability , as characterized by @xmath9 , versus infrared excess , for all the tts in our sample , as well as for only the tts brighter than _ i_=14.3 , where our identification of irregular variables may be regarded as complete .
the large scatter in these plots , including the non - physical phenomenon of infrared deficiencies represented by negative values of @xmath19(i - k ) , may result primarily from the non - simultaneity of the optical and infrared measurements . despite the scatter , additional causes of which
will be mentioned below , some weak trends are visible in the full data set .
of the six stars with @xmath19(i - k)@xmath80.5 , five are ctts , as one might have expected . within the cluster as a whole ,
our data indicate that the ctts are more variable than the wtts .
these facts are consistent with the canonical view that ctts variability is due at least partially to unsteady accretion from a circumstellar disk .
this becomes more evident when considering only the magnitude - limited sample . in that case , all of the stars with @xmath19(i - k)@xmath80.5 are ctts .
it is interesting to note that the periodic wtts cluster around @xmath19(i - k)=0 .
they also appear to be somewhat less variable than the non - periodic wtts and less likely to have derived infrared excesses below zero .
however , these latter two distinctions can be understood as selection effects due to the difficulty in detecting periods in fainter stars , as discussed above .
in general , the correspondence between infrared excess and variability is weak and stands in contrast to the results for h@xmath17 equivalent width described above .
this led us to examine whether the two supposed disk indicators , h@xmath17 equivalent width and infrared excess were in fact correlated with each other in this sample . fig .
[ havik ] shows the log of the h@xmath17 equivalent width , as measured by @xcite versus the infrared excesses which we have computed for all the tts in our sample with the neccessary spectral information . while there is a fairly clear division between ctts and wtts , the correlation between infrared excess and hydrogen emission is weak at best and actually almost non - existent .
variability obviously correlates much better with h@xmath17 equivalent width than infrared excess , perhaps because it is more a measure of current accretion activity or perhaps because it can be determined with better precision . in any event , in this sample
, there is almost a one - to - one correspondence between irregular variability and ctts status based on h@xmath17 equivalent width and only weak correspondence between infrared excess emission and irregular variability ( or h@xmath17 equivalent width ) .
it is also possible that the large degree of scatter in the infrared excesses shown in figs .
[ sigik ] and [ havik ] can be attributed to the causes discussed by @xcite , which include disk inclination effects , inner holes in disks , and excess k - band radiation from large cool spots .
however , there is another complicating factor in this case , namely the fact that the photometric measurements used to compute the infrared excess were not observed simultaneously or , in the case of k , very recently .
the infrared observations of @xcite were made in september 1991 , and the photometric magnitudes of @xcite , needed for the extinction computation , were calculated as an average of three observations between november 1993 and october 1996 . based on the light curves of the ctts in our sample
it is clear that the photometric characteristics of a star can change from day to day and from year to year .
it is entirely possible that the general variability of the stars coupled with the non - simultaneity of the observations is a principle cause of the large scatter in the infrared excess measurement which then masks any correspondence with variability . only continued monitoring simultaneously in the optical and infrared could determine if this is the case .
this k6 ctts was noted in paper i as the most extreme variable star in our sample and that has continued to prove true .
its variations ( see fig . [ star23 ] ) , which approach 2 mag in amplitude , are also more reminiscent of uxor behavior than of typical ctts behavior
. in particular , the star tends to be near the bright end of its range most of the time and exhibits relatively brief , non - periodic excursions to fainter levels , as opposed to the more random sorts of fluctuations characteristic of type ii variables . on the other hand , most uxors are of earlier spectral type and do not have such strong h@xmath17 emission .
the star may be some sort of transition object or perhaps a close binary ( or both ) .
it clearly warrants further study as a potential uxor with interesting properties .
multi - color photometry and simultaneous spectral monitoring would be useful in elucidating its nature .
star 47 , a k8 wtts , is the only example of a star in our sample which could be classified as both a periodic and an irregular variable ( of substantial amplitude ) at different times .
it was periodic in the 1998 - 99 season with a period of 8.38 days ( see fig .
[ lcs ] ) , but in the following seasons there are no peaks in its power spectra meeting our criteria for periodicity . between 2000 and 2002 , its light curve is punctuated by several dramatic minima of increasing depth , the deepest occuring at jd 2452310 ( see fig .
[ star47 ] ) .
the star fades by about one magnitude in one day , and four days later it has returned to its normal brightness . as in the case of star 23
, this kind of variation is much more characteristic of uxor behavior than of typical ctts ( or wtts ) variability .
again , it may be some sort of transition object or a binary and clearly warrants follow - up study .
it is possible that this star , and star 23 , are additional ( to star 15 , see below ) examples of objects whose variability is caused by transient occultation events from dust clouds along the line of sight ( and probably circumstellar or circumbinary ) . within our sample ,
star 15 is the only example of a star which exhibits coherent behavior over 4.5 years of observation on a time scale longer than a single observing season .
this remarkable object apparently undergoes an eclipse lasting @xmath13.5 years , the longest known astronomical eclipse ever detected , in which it gradually fades by 0.66 mag in _
i _ and subsequently recovers to its out - of - eclipse magnitude .
such an amazing and unexpected behavior clearly deserves special attention in the literature and was given it by @xcite , to which the reader is referred for more detail .
here we present a brief summary of that work and some supplementary material .
table [ table3 ] gives our photometry for the star .
@xcite have classified star 15 as g8-k6 based on near - infrared spectroscopy .
it lacks infrared excess and hydrogen emission but is almost certainly a cluster member based on its locations on the sky and cm diagram .
the light curve of the star , based on our data and some from the literature , is shown in fig .
[ 15alldata ] .
the shape , consisting of a gradual ingress , a stable minimum , and a gradual egress , resembles that of an eclipsing binary with the obvious exception of its duration .
observations in the _ r _ band during minimum light show that the system became redder as it faded , although not as much as would be expected based on in interstellar reddening law .
@xcite argue that the star may be a binary system in which one member is occulted by a dust cloud , presumably of circumstellar or circumbinary origin . as fig .
[ 15alldata ] suggests , the eclipse may be recurrent on a time scale as short as 4 years .
obviously this is a star which should continue to be monitored photometrically and high resolution spectroscopy obtained near a brightness maximum should reveal whether it is , in fact , a binary system .
five years of monitoring stars in ic 348 has yielded a set of 28 periodic variables , almost all of which are wtts , and 16 irregular variables , the large majority of which are ctts .
our sample of variables is likely to be complete for stars brighter than i = 14.3 . of the 42 stars fulfilling that criteria ( which includes 25 variable stars ) , 10 are of earlier spectral class than g and none of those 10 was detected as a variable .
one is a late - type non - member and it was also not detected as a variable .
four of the remaining stars have no spectral information : one of these is a periodic variable and three are non - variable . that leaves 27 g , k or m - type pms members of the cluster within our field which are brighter than i=14.3 , and 24 of them ( i.e. @xmath190% ) were detected as variable stars in this study .
we believe this nicely illustrates the power and value of variability work for identifying pms stars in young clusters .
it is furthermore interesting that , where our sample is complete ( i @xmath0 14.3 ) , every one of the 5 ctts was identified as an irregular variable , and 14 of the 16 k and m - type wtts were identified as periodic variables .
in addition , 3 out of the 6 bright g - type pms stars showed periodic variations and 2 of the remaining 3 showed irregular variations ( including the unusual variable star 15 ) .
not only are variability studies excellent at identifying pms stars , they are also apparently capable of distinguishing ctts from wtts with rather good precision .
all of these results are quite similar , both qualitatively and quantitatively to what is found for the set of bright tts in tau / aur , orion , sco / cen and other well known associations analyzed by @xcite .
one particular fact worth noting is that not one of the 20 known ctts in our monitored sample revealed periodic variability during five seasons of monitoring .
by contrast , 23 of the 64 known wtts in the same area were detected as periodic and we probably would have found many more periodic variables among them if we had deeper images .
this means that rotation studies based on periods determined by photometric monitoring are probably strongly biased against stars which are currently actively accreting . since disk - locking is thought to be a primary controlling factor in the rotational evolution of pms stars , and since actively accreting stars ( ctts ) are , one might suppose , the ones most likely to be currently locked to their disks , this bias is important to consider when analyzing rotation data on clusters . in principle , one could use v sin i measurements to ascertain whether the rotational properties of ctts differ from wtts .
@xcite found no significant difference between their periodic and non - periodic stars samples in the orion nebula cluster , but they did not examine separately the definite irregular variables .
a v sin i study of ic 348 could be extremely useful in addressing this issue .
we thank g. herbig and c. jordi for providing us with unpublished data on hmw 15 .
we also thank the large number of wesleyan students who contributed to this project over the years by obtaining the data on our campus telescope .
this investigation was supported by grants from the nasa - origins program .
bouvier , j. 1990 , , 99 , 946 bouvier et al .
1995 , , 299 , 89 cohen , r. e. , herbst , w. , & williams , e. c. 2003 , , in press .
dantona , f. & mazzitelli , i. 1994 , apjs , 90 , 467 duchene , g. , bouvier , j. & simon , t. 1999 , , 343 , 831 grankin , k. n. 1999 , ast .
l. , 25 , 526 herbig , g. h. 1977 , , 217 , 693 herbig , g. h. 1989 , proceedings of the eso workshop on low mass star formation and pre - main sequence objects , held in garching bei munchen , f.r.g . , july 11 - 13 , 1989 .
editor , bo reipurth ; publisher , european southern observatory , garching bei munchen , f.r.g .
herbig , g. h. 1998 , , 497 , 736 herbst , w. , 1989 , , 98 , 2268 herbst , w. , herbst , d. k. , grossman , e. j. & weinstein , d. 1994 , , 108 , 1906 herbst , w. & shevchenko , v. s. 1999 , , 118 , 1043 herbst , w. , maley , j. a. & williams , e. c. 2000 , , 120 , 349 herbst , w. , rhode , k. l. , hillenbrand , l. a. & curran , g. 2000 , , 119 , 261 herbst , w. , bailer - jones , c. a. l. , mundt , r. 2001 , , 554 , 197 herbst , w. , bailer - jones , c. a. l. , mundt , r. , meisenheimer , k. , & wackerman , r. 2002 , , 396 , 513 hillenbrand , l. 1997 , , 113 , 1733 hillenbrand , l. et al .
1998 , , 116 , 1816 horne , j. h. & baliunas , s. l. 1986 , , 302 , 757 johns - krull , c. m. 1996 , , 306 , 803 kenyon , s. j. & hartmann , l. 1995 , , 101 , 117 konigl , a. 1991 , , 370 , 39 kovacs , g. 1981 , app .
space sci .
78 , 175 kuker , m. & rudiger , g. 1997 , a&a , 328 , 253 lada , e. a. & lada , c. j. 1995 , 109 , 1682 luhman , k. l. , rieke , g. h. , lada , c. j. & lada , e. a. 1998 , , 508 , 347 luhman , k. l. , stauffer , j. r. , muench , a. a. , rieke , g. h. , lada , e. a. , bouvier , j. & lada , c. j. 2003 , preprint ostriker , e. c. & shu , f. h. , , 447 , 813 preibisch , t. & zinnecker , h. 2002 , , 123 , 866 .
preibisch , t. , zinnecker , h. & herbig , g. h. 1996 , , 310 , 456 press , w. h. , teukolsky , s. a. , vetterling , w. t. , & flannery , b. p. 1992
, numerical recipes ( cambridge u. press , cambridge ) rice , j. b. and strassmeier , k. g. 1996 , , 316 , 164 ripepi , v et al . 2002 , , 391 , 587
rhode , k. l. , herbst , w. & mathieu , r. d. 2001 , 122 , 3258 scargle , j. d. 1982 , , 263 , 835 scholz , r .- d .
1999 , , 137 , 305 smith , m. d. 1994 , , 287 , 523 stelzer , b. et al . , 2003 , , in press trullols , e. & jordi , c. 1997 , , 324 , 549 vrba , f. j. , herbst , w. & booth , j. f. 1988 , , 96 , 1032 l|llllllll * 1998 - 1999 & 1157.687 & 1157.695 & 1157.824 & 1162.581 1162.589 & 1162.603 & 1162.611 & 1162.680 + & 1162.805 & 1162.813 & 1163.636 & 1163.644 1163.658 & 1163.666 & 1164.598 & 1164.607 + & 1166.505 & 1166.513 & 1166.581 & 1166.589 1166.610 & 1166.618 & 1170.748 & 1170.790 + & 1170.799 & 1178.561 & 1178.569 & 1178.603 1178.611 & 1178.702 & 1178.710 & 1183.496 + & 1183.522 & 1183.627 & 1183.669 & 1183.727 1184.464 & 1184.657 & 1184.782 & 1186.525 + & 1186.599 & 1186.745 & 1190.588 & 1196.564 1198.515 & 1198.683 & 1199.484 & 1199.690 + & 1209.515 & 1209.534 & 1209.592 & 1209.686 1210.569 & 1210.649 & 1213.526 & 1215.556 + & 1215.559 & 1215.623 & 1215.711 & 1221.556 1224.572 & 1224.658 & 1225.665 & 1226.620 + & 1232.513 & 1232.604 & 1243.521 & 1243.577 1243.601 & 1245.515 & 1245.567 & 1245.623 + & 1246.521 & 1246.573 & 1246.628 & 1263.532 & & & & + * 1999 - 2000 & 1513.622 & 1514.617 & 1518.647 & 1520.662 1521.631 & 1522.651 & 1525.632 & 1529.649 + & 1529.707 & 1529.803 & 1530.638 & 1532.645 1549.557 & 1549.629 & 1558.593 & 1558.690 + & 1563.521 & 1565.521 & 1570.554 & 1571.517 1572.529 & 1572.612 & 1577.560 & 1580.506 + & 1580.568 & 1582.526 & 1583.542 & 1583.647 1584.530 & 1586.554 & 1590.543 & 1590.606 + & 1592.512 & 1595.505 & 1596.524 & 1597.583 1598.514 & 1604.522 & 1604.599 & 1607.529 + & 1608.524 & 1610.573 & 1612.589 & 1617.546 1618.544 & 1618.576 & 1627.581 & 1630.572 + * 2000 - 2001 & 1837.689 & 1838.749 & 1840.651 & 1841.629
1842.762 & 1843.693 & 1850.693 & 1866.816 + & 1870.685 & 1884.560 & 1886.506 & 1886.609 1891.671 & 1897.545 & 1899.538 & 1913.551 + & 1913.692 & 1914.521 & 1921.524 & 1921.735 1922.538 & 1922.626 & 1923.551 & 1923.698 + * 2001 - 2002 & 2220.676 & 2221.686 & 2224.758 & 2225.762 2226.731 & 2227.621 & 2227.753 & 2230.640 + & 2231.609 & 2231.696 & 2231.838 & 2232.681 2240.618 & 2253.610 & 2255.560 & 2259.588 + & 2300.629 & 2300.741 & 2301.594 & 2301.686 2302.596 & 2302.690 & 2304.639 & 2309.548 + & 2309.701 & 2310.584 & 2314.596 & 2315.564 2315.635 & 2317.588 & 2320.512 & 2322.564 + & 2322.617 & 2324.535 & 2324.604 & 2327.530 2329.557 & 2329.628 & 2330.579 & 2334.553 + & 2335.586 & 2335.643 & 2341.592 & 2362.512 & & & & + * 2002 - 2003 & 2549.770 & 2553.815 & 2555.774 & 2562.744 2569.698 & 2576.730 & 2587.696 & 2602.700 + & 2603.705 & 2612.719 & 2617.693 & 2617.782 2621.631 & 2650.640 & 2658.536 & 2659.519 + & 2660.537 & 2662.566 & 2676.539 & 2676.627 2683.542 & 2684.526 & 2684.632 & 2691.576 + & 2696.597 & 2697.622 & 2702.596 & 2709.596 2714.595 & 2722.580 & 2723.567 & + * * * * * 0.455 & & a + 2 & 2 & 3 44 35.35 & 32 10 04.2 & 0.69 & 8.99 & 0.010 & 0.062 & & + 3 & 19 & 3 44 30.81 & 32 09 55.4 & 0.59 & 10.69 & 0.004 & 0.030 & & + 4 & 4 & 3 44 31.18 & 32 06 21.7 & 0.68 & 9.36 & 0.007 & 0.047 & & b + 5 & 8 & 3 44 09.12 & 32 07 09.0 & 0.38 & 9.35 & 0.020 & 0.105 & & + 6 & 7 & 3 44 08.44 & 32 07 16.2 & 0.36 & 9.30 & 0.018 & 0.095 & & + 7 & 28 & 3 44 21.04 & 32 07 38.3 & 0.47 & 10.67 & 0.006 & 0.033 & & + 8 & 30 & 3 44 19.11 & 32 09 30.8 & 0.53 & 10.86 & 0.005 & 0.027 & & + 9 & 38 & 3 44 23.97 & 32 10 59.5 & 0.83 & 11.86 & 0.019 & 0.100 & & + 10 & 10 & 3 44 24.65 & 32 10 14.7 & 0.71 & 10.29 & 0.005 & 0.030 & & + 11 & 6 & 3 44 36.94 & 32 06 45.1 & 1.03 & 10.80 & 0.017 & 0.094 & 1.687 & c , g + 12 & 29 & 3 44 31.52 & 32 08 44.6 & 0.83 & 12.15 & 0.032 & 0.181 & 2.237 & c + 13 & 12 & 3 44 31.99 & 32 11 43.5 & 1.15 & 12.01 & 0.011 & 0.048 & & g + 14 & 9 & 3 44 39.16 & 32 09 17.8 & 1.39 & 12.12 & 0.024 & 0.109 & 2.539 & c , g + 15 & 35 & 3 44 39.23 & 32 07 35.1 & 1.40 & 13.15 & 0.212 & 0.715 & var . & d + 16 & 36 & 3 44 38.45 & 32 07 35.3 & 1.30 & 13.17 & 0.085 & 0.330 & 5.221 & + 17 & 77 & 3 44 43.43 & 32 08 17.0 & 0.96 & 12.79 & 0.008 & 0.039 & & + 18 & 59 & 3 44 40.11 & 32 11 33.7 & 1.26 & 13.49 & 0.034 & 0.170 & var . & d + 19 & & 3 44 25.56 & 32 12 29.5 & 1.15 & 13.50 & 0.053 & 0.247 & 8.363 & c + 20 & 5 & 3 44 26.01 & 32 04 30.1 & 1.35 & 12.42 & 0.103 & 0.754 & var . & d + 21 & & 3 44 27.00 & 32 04 43.4 & 1.27 & 13.61 & 0.025 & 0.115 & & + 22 & & 3 44 44.71 & 32 04 02.2 & 1.52 & 13.17 & 0.047 & 0.287 & var . & d + 23 & 37 & 3 44 37.96 & 32 03 29.4 & 1.18 & 13.39 & 0.342 & 1.978 & var . & d + 24 & & 3 44 04.97 & 32 09 53.4 & 1.07 & 13.04 & 0.024 & 0.124 & & + 25 & & 3 44 05.74 & 32 12 28.4 & 1.05 & 13.84 & 0.014 & 0.071 & & + 26 & & 3 44 16.40 & 32 09 54.8 & 0.89 & 12.65 & 0.048 & 0.230 & 3.021 & c + 27 & 16 & 3 44 32.74 & 32 08 37.0 & 0.85 & 11.62 & 0.017 & 0.097 & 2.656 & c + 28 & 33 & 3 44 32.58 & 32 08 42.0 & 1.62 & 12.51 & 0.481 & 1.462 & & e + 29 & 32 & 3 44 37.88 & 32 08 03.7 & 1.64 & 14.04 & 0.067 & 0.377 & var .
& d + 30 & 58 & 3 44 38.54 & 32 08 00.2 & 1.63 & 14.10 & 0.068 & 0.490 & 7.570 & c + 31 & 62 & 3 44 26.61 & 32 03 58.0 & 1.82 & 14.03 & 0.048 & 0.204 & 3.087 & c + 32 & & 3 44 24.55 & 32 03 56.9 & 1.20 & 15.16 & 0.086 & 0.377 & 4.942 & c + 33 & 120 & 3 44 22.96 & 32 11 56.7 & 1.64 & 14.40 & 0.023 & 0.138 & & + 34 & & 3 44 22.30 & 32 12 00.2 & 1.27 & 14.52 & 0.087 & 0.501 & var . & d + 35 & & 3 44 21.28 & 32 11 55.9 & 1.87 & 14.66 & 0.025 & 0.144 & & + 36 & & 3 44 17.89 & 32 12 19.8 & 1.44 & 13.91 & 0.027 & 0.137 & & + 37 & 95 & 3 44 21.90 & 32 12 11.0 & 1.64 & 14.32 & 0.021 & 0.106 & & + 38 & & 3 44 21.75 & 32 12 30.9 & ( 2.00 ) & 14.96 & 0.024 & 0.155 & & f + 39 & & 3 44 23.65 & 32 06 46.2 & 1.74 & 14.12 & 0.029 & 0.251 & 9.667 & c , g + 40 & 125 & 3 44 21.63 & 32 06 24.5 & 1.64 & 14.49 & 0.036 & 0.175 & 8.363 & c + 41 & & 3 44 28.46 & 32 07 22.1 & 1.14 & 13.32 & 0.075 & 0.275 & 6.997 & c + 42 & 86 & 3 44 27.86 & 32 07 31.2 & 1.73 & 14.05 & 0.015 & 0.097 & 6.532 & c , g + 43 & 71 & 3 44 32.56 & 32 08 55.4 & 1.91 & 14.31 & 0.020 & 0.144 & & + 44 & 65 & 3 44 33.97 & 32 08 53.7 & 1.20 & 13.54 & 0.054 & 0.235 & 16.40 & c + 45 & 23 & 3 44 38.71 & 32 08 41.6 & 1.72 & 13.80 & 0.031 & 0.130 & 2.405 & c + 46 & 108 & 3 44 38.69 & 32 08 56.3 & 1.81 & 14.39 & 0.035 & 0.319 & & + 47 & 83 & 3 44 37.41 & 32 09 00.5 & 1.64 & 14.77 & 0.146 & 1.066 & 8.378 & d , i + 48 & 113 & 3 44 37.19 & 32 09 15.7 & 1.22 & 14.25 & 0.030 & 0.155 & & + 49 & & 3 44 37.40 & 32 06 11.3 & 1.29 & 13.77 & 0.018 & 0.103 & 6.221 & + 50 & & 3 44 34.86 & 32 06 33.3 & 1.18 & 13.27 & 0.048 & 0.494 & 5.483 & c + 51 & & 3 44 42.61 & 32 06 19.0 & 1.13 & 13.89 & 0.020 & 0.102 & 11.51 & c + 52 & & 3 44 11.22 & 32 06 11.7 & 1.59 & 14.21 & 0.039 & 0.175 & 10.77 & c + 53 & & 3 44 38.37 & 32 12 59.2 & ( 1.77 ) & 14.55 & 0.036 & 0.198 & 13.48 & c , h + 54 & 74 & 3 44 34.25 & 32 10 49.2 & 1.98 & 14.38 & 0.021 & 0.142 & & + 55 & & 3 44 49.61 & 32 09 11.5 & 1.09 & 14.98 & 0.022 & 0.164 & & + 56 & 75 & 3 44 43.77 & 32 10 29.9 & 1.59 & 14.36 & 0.229 & 1.358 & var . & d + 57 & & 3 44 18.14 & 32 04 56.7 & 1.95 & 15.24 & 0.062 & 0.478 & & + 58 & & 3 44 14.10 & 32 10 27.9 & 1.49 & 14.69 & 0.022 & 0.110 & & + 59 & 41 & 3 44 21.59 & 32 10 37.2 & 1.63 & 15.10 & 0.251 & 1.177 & var . & d + 60 & & 3 44 21.54 & 32 10 17.0 & 1.65 & 14.54 & 0.089 & 0.481 & 7.092 & c + 61 & 128 & 3 44 20.15 & 32 08 56.2 & 1.92 & 14.80 & 0.022 & 0.154 & & + 62 & & 3 44 04.07 & 32 07 16.7 & 1.64 & 14.94 & 0.035 & 0.224 & & + 63 & & 3 44 06.77 & 32 04 40.6 & 1.70 & 15.06 & 0.028 & 0.190 & & + 64 & & 3 44 21.23 & 32 05 02.2 & 1.74 & 15.06 & 0.053 & 0.283 & & + 65 & & 3 44 22.26 & 32 05 42.4 & 1.48 & 14.95 & 0.196 & 0.979 & var . & d + 66 & & 3 44 35.69 & 32 04 52.3 & 2.42 & 16.03 & 0.061 & 0.480 & & + 67 & & 3 44 38.59 & 32 05 06.0 & 2.15 & 14.80 & 0.019 & 0.164 & & + 68 & & 3 44 31.65 & 32 06 53.3 & 2.55 & 16.10 & 0.048 & 0.435 & & + 69 & & 3 44 37.77 & 32 12 17.7 & ( 1.88 ) & 15.39 & 0.078 & 0.355 & 3.256 & e , h + 70 & & 3 44 11.19 & 32 08 15.8 & 2.62 & 15.13 & 0.036 & 0.248 & & + 71 & & 3 44 6.75 & 32 07 53.7 & 2.24 & 15.29 & 0.040 & 0.305 & & + 72 & & 3 44 6.10 & 32 07 06.6 & 2.50 & 15.89 & 0.052 & 0.527 & & g + 73 & & 3 44 19.22 & 32 07 34.4 & 1.80 & 14.88 & 0.257 & 1.271 & 7.582 & c , e + 74 & & 3 44 18.24 & 32 07 32.2 & 2.27 & 16.04 & 0.058 & 0.662 & & + 75 & 40 & 3 44 29.71 & 32 10 39.4 & 1.44 & 14.18 & 0.054 & 0.337 & var . & d , g + 76 & 88 & 3 44 32.75 & 32 09 15.3 & ( 2.00 ) & 14.62 & 0.043 & 0.240 & 5.286 & c , f + 77 & 151 & 3 44 34.81 & 32 11 17.5 & 2.08 & 14.91 & 0.022 & 0.145 & & + 78 & 45 & 3 44 24.27 & 32 10 19.0 & 1.08 & 12.64 & 0.201 & 1.221 & & e + 79 & & 3 44 25.25 & 32 10 12.6 & 2.37 & 14.48 & 1.052 & 4.816 & & e + 80 & 96 & 3 44 34.83 & 32 09 52.6 & 2.11 & 13.63 & 1.724 & 7.644 & & e + 81 & 91 & 3 44 39.19 & 32 09 44.4 & 1.86 & 14.69 & 0.059 & 0.445 & 3.993 & c + 82 & 145 & 3 44 41.29 & 32 10 24.7 & 2.22 & 14.75 & 0.021 & 0.190 & & + 83 & & 3 44 43.04 & 32 10 14.6 & 2.55 & 16.09 & 0.059 & 0.626 & & + 84 & 42 & 3 44 42.05 & 32 09 00.2 & 1.98 & 14.31 & 0.110 & 0.549 & & e + 85 & & 3 44 42.07 & 32 09 00.7 & 1.87 & 14.34 & 0.114 & 0.550 & & e + 86 & & 3 44 31.40 & 32 11 28.9 & ( 2.00 ) & 15.82 & 0.043 & 0.318 & & f + 87 & 193 & 3 44 37.99 & 32 11 36.6 & 2.20 & 15.56 & 0.092 & 0.644 & & + 88 & & 3 44 37.38 & 32 12 23.6 & ( 1.96 ) & 15.85 & 0.106 & 0.955 & & e , h + 89 & & 3 44 40.78 & 32 13 06.1 & ( 2.00 ) & 15.97 & 0.054 & 0.357 & & f + 90 & & 3 44 41.43 & 32 13 09.2 & ( 2.00 ) & 15.72 & 0.041 & 0.257 & & f + 91 & & 3 44 48.82 & 32 13 21.3 & ( 2.00 ) & 14.94 & 0.035 & 0.190 & & f + 92 & & 3 44 41.72 & 32 12 01.8 & ( 2.00 ) & 16.13 & 0.067 & 0.430 & & f + 93 & 404 & 3 44 44.42 & 32 10 05.1 & 1.65 & 16.67 & 0.086 & 0.576 & & + 94 & 171 & 3 44 44.83 & 32 11 05.1 & 1.94 & 15.31 & 0.030 & 0.253 & & + 95 & 360 & 3 44 43.69 & 32 10 47.4 & 2.09 & 16.46 & 0.073 & 0.612 & & + 96 & 277 & 3 44 39.43 & 32 10 07.4 & 2.22 & 16.10 & 0.047 & 0.376 & & + 97 & 167 & 3 44 41.15 & 32 10 09.6 & 2.47 & 16.66 & 0.139 & 0.892 & & + 98 & 166 & 3 44 42.58 & 32 10 02.2 & 2.83 & 16.84 & 0.091 & 0.960 & & e + 99 & 221 & 3 44 40.25 & 32 09 32.4 & 2.41 & 16.60 & 0.157 & 2.136 & & g + 100 & 442 & 3 44 40.67 & 32 09 40.6 & 1.32 & 17.43 & 0.182 & 1.658 & & + 101 & & 3 44 18.17 & 32 09 58.9 & 2.54 & 15.78 & 0.129 & 0.749 & var . & d + 102 & & 3 44 19.19 & 32 09 59.8 & ( 2.00 ) & 18.92 & 0.512 & 2.350 & & f + 103 & & 3 44 18.53 & 32 09 46.3 & ( 2.00 ) & 18.10 & 0.830 & 2.798 & & f + 104 & 294 & 3 44 24.56 & 32 10 03.0 & 2.09 & 16.05 & 0.147 & 1.106 & & e + 105 & 237 & 3 44 23.55 & 32 09 33.4 & 2.30 & 15.70 & 0.041 & 0.385 & & + 106 & 90 & 3 44 33.31 & 32 09 39.6 & 1.83 & 14.07 & 0.255 & 2.430 & & e + 107 & 165 & 3 44 35.46 & 32 08 56.0 & 2.60 & 15.96 & 0.054 & 0.328 & & e + 108 & 149 & 3 44 36.98 & 32 08 33.8 & 2.41 & 15.58 & 0.062 & 0.597 & & + 109 & 153 & 3 44 42.76 & 32 08 33.4 & 2.25 & 15.78 & 0.067 & 0.458 & & + 110 & 103 & 3 44 44.58 & 32 08 12.1 & 1.82 & 15.35 & 0.163 & 0.910 & var . &
d + 111 & 52 & 3 44 43.52 & 32 07 42.3 & 1.85 & 14.82 & 0.029 & 0.187 & & e + 112 & & 3 44 34.43 & 32 06 24.6 & 2.59 & 15.91 & 0.054 & 0.354 & & + 113 & & 3 44 49.98 & 32 03 45.2 & ( 2.00 ) & 14.97 & 0.052 & 0.660 & & f + 114 & & 3 44 49.78 & 32 03 33.7 & ( 2.00 ) & 15.17 & 0.034 & 0.206 & & f + 115 & & 3 44 34.34 & 32 04 21.5 & 1.62 & 16.47 & 0.074 & 0.595 & & + 116 & & 3 44 32.79 & 32 04 12.9 & 1.24 & 16.09 & 0.051 & 0.474 & & + 117 & & 3 44 38.84 & 32 06 35.9 & 2.62 & 16.54 & 0.070 & 0.538 & & + 118 & & 3 44 41.22 & 32 06 26.6 & 2.50 & 16.73 & 0.089 & 0.805 & & + 119 & 141 & 3 44 30.53 & 32 06 29.3 & 1.75 & 14.68 & 0.140 & 1.016 & & e + 120 & & 3 44 29.68 & 32 05 51.9 & 1.56 & 15.97 & 0.047 & 0.386 & & + 121 & & 3 44 31.02 & 32 05 45.6 & 2.54 & 16.64 & 0.074 & 0.666 & & + 122 & & 3 44 27.27 & 32 07 17.4 & 2.24 & 16.60 & 0.086 & 0.685 & & + 123 & & 3 44 05.25 & 32 08 02.1 & 1.40 & 16.37 & 0.082 & 0.956 & & + 124 & & 3 44 08.03 & 32 06 56.0 & 1.77 & 16.18 & 0.111 & 0.911 & & + 125 & & 3 44 07.67 & 32 05 04.7 & 2.45 & 16.62 & 0.074 & 0.510 & & + 126 & & 3 44 07.46 & 32 04 08.5 & 1.96 & 15.53 & 0.036 & 0.330 & & + 127 & & 3 44 10.09 & 32 04 04.2 & ( 2.00 ) & 15.78 & 0.060 & 0.533 & & f + 128 & & 3 44 32.41 & 32 05 11.4 & ( 2.00 ) & 19.07 & 0.481 & 1.945 & & e , f + 129 & & 3 44 32.36 & 32 05 11.3 & ( 2.00 ) & 19.00 & 0.446 & 1.935 & & e , f + 130 & & 3 44 21.20 & 32 12 36.8 & ( 2.00 ) & 15.69 & 0.151 & 1.710 & & e , f + 131 & & 3 44 20.90 & 32 12 37.1 & ( 2.00 ) & 15.55 & 0.088 & 0.448 & & e , f + 132 & & 3 44 18.57 & 32 12 52.7 & ( 2.00 ) & 16.56 & 0.345 & 2.005 & var . &
d , f + 133 & 60 & 3 44 25.55 & 32 11 30.1 & 1.90 & 14.78 & 0.101 & 0.485 & var . & d + 134 & & 3 44 19.99 & 32 06 45.1 & ( 2.00 ) & 15.72 & 0.070 & 0.570 & 8.687 & f + 135 & & 3 44 25.53 & 32 06 16.7 & 2.24 & 15.86 & 0.068 & 0.354 & & + 136 & & 3 44 23.59 & 32 07 11.3 & 1.36 & 16.60 & 0.092 & 0.789 & & g + 137 & & 3 44 34.04 & 32 06 56.8 & 2.80 & 16.87 & 0.120 & 0.891 & & + 138 & & 3 44 04.20 & 32 09 38.1 & 0.95 & 15.02 & 0.029 & 0.275 & & + 139 & & 3 44 09.97 & 32 09 41.3 & 1.31 & 16.15 & 0.060 & 0.488 & & + 140 & & 3 44 25.58 & 32 10 44.8 & ( 2.00 ) & 19.21 & 0.441 & 1.040 & & e + 141 & 194 & 3 44 27.22 & 32 10 36.7 & 2.38 & 16.22 & 0.206 & 1.264 & var . & d + 142 & 230 & 3 44 35.51 & 32 08 04.1 & 2.54 & 16.15 & 0.059 & 0.485 & & + 143 & & 3 44 35.03 & 32 07 36.6 & 1.23 & 12.72 & 0.039 & 0.226 & & e + 144 & & 3 44 35.17 & 32 07 36.4 & ( 2.00 ) & 12.72 & 0.039 & 0.226 & & e , f + 145 & & 3 44 20.26 & 32 05 43.5 & 2.57 & 16.91 & 0.110 & 0.963 & & + 146 & & 3 44 19.14 & 32 05 59.5 & 2.63 & 16.83 & 0.098 & 0.630 & & + 147 & 169 & 3 44 17.73 & 32 04 47.4 & 2.57 & 15.69 & 0.048 & 0.413 & & + 148 & & 3 44 16.48 & 32 05 32.5 & 1.95 & 16.41 & 0.080 & 0.450 & & + 149 & & 3 44 46.93 & 32 05 36.5 & 1.13 & 15.89 & 0.048 & 0.432 & & + 150 & & 3 44 48.37 & 32 06 27.9 & ( 2.00 ) & 18.76 & 0.505 & 1.682 & & f + 151 & & 3 44 50.78 & 32 07 00.5 & ( 2.00 ) & 18.96 & 0.387 & 1.488 & & f + cccccccc 1157.687 & 12.96 & 1157.695 & 12.96 & 1162.581 & 12.96 & 1162.680 & 12.97 + 1163.644 & 12.97 & 1163.658 & 12.96 & 1163.666 & 12.97 & 1164.598 & 12.97 + 1166.505 & 12.96 & 1166.513 & 12.96 & 1170.799 & 12.96 & 1184.464 & 12.96 + 1210.569 & 12.94 & 1215.556 & 12.95 & 1215.623 & 12.96 & 1243.521 & 12.97 + 1263.532 & 12.97 & 1243.521 & 12.97 & 1166.581 & 12.96 & 1166.589 & 12.97 + 1166.610 & 12.95 & 1166.618 & 12.96 & 1170.748 & 12.95 & 1170.790 & 12.96 + 1170.799 & 12.96 & 1178.561 & 12.98 & 1178.569 & 12.98 & 1178.603 & 12.98 + 1178.611 & 12.96 & 1178.702 & 12.96 & 1178.710 & 12.95 & 1183.496 & 12.96 + 1183.522 & 12.97 & 1183.627 & 12.97 & 1183.669 & 12.97 & 1183.727 & 12.97 + 1184.464 & 12.96 & 1184.657 & 12.97 & 1184.782 & 12.98 & 1186.525 & 12.97 + 1186.599 & 12.97 & 1186.745 & 12.97 & 1190.588 & 12.97 & 1196.564 & 12.97 + 1198.515 & 12.97 & 1198.683 & 12.97 & 1199.484 & 12.96 & 1199.690 & 12.98 + 1209.515 & 12.96 & 1209.534 & 12.96 & 1209.592 & 12.96 & 1209.686 & 12.97 + 1210.569 & 12.97 & 1210.649 & 12.97 & 1213.526 & 12.96 & 1215.556 & 12.97 + 1215.559 & 12.97 & 1215.623 & 12.96 & 1215.711 & 12.97 & 1221.556 & 12.98 + 1224.572 & 12.97 & 1224.658 & 12.96 & 1225.665 & 12.98 & 1226.620 & 12.96 + 1232.513 & 12.97 & 1232.604 & 12.97 & 1243.521 & 12.97 & 1243.577 & 12.96 + 1243.601 & 12.97 & 1245.515 & 12.98 & 1245.567 & 12.97 & 1245.623 & 12.97 + 1246.521 & 12.97 & 1246.573 & 12.98 & 1246.628 & 12.98 & 1263.532 & 12.98 + 1513.622 & 13.16 & 1514.617 & 13.16 & 1518.647 & 13.17 & 1520.662 & 13.18 + 1521.631 & 13.17 & 1522.651 & 13.18 & 1525.632 & 13.19 & 1529.649 & 13.19 + 1529.707 & 13.19 & 1529.803 & 13.18 & 1530.638 & 13.20 & 1532.645 & 13.19 + 1549.557 & 13.22 & 1549.629 & 13.22 & 1558.593 & 13.24 & 1558.690 & 13.23 + 1563.521 & 13.25 & 1565.521 & 13.25 & 1570.554 & 13.25 & 1571.517 & 13.24 + 1572.529 & 13.26 & 1572.612 & 13.26 & 1577.560 & 13.25 & 1580.506 & 13.26 + 1580.568 & 13.27 & 1582.526 & 13.27 & 1583.542 & 13.27 & 1583.647 & 13.27 + 1584.530 & 13.26 & 1586.554 & 13.28 & 1590.543 & 13.28 & 1590.606 & 13.28 + 1592.512 & 13.28 & 1595.505 & 13.30 & 1596.524 & 13.29 & 1597.583 & 13.30 + 1598.514 & 13.29 & 1604.522 & 13.29 & 1604.599 & 13.29 & 1607.529 & 13.29 + 1608.524 & 13.31 & 1610.573 & 13.31 & 1612.589 & 13.33 & 1617.546 & 13.32 + 1618.544 & 13.33 & 1618.576 & 13.32 & 1627.581 & 13.33 & 1630.572 & 13.33 + 1837.689 & 13.64 & 1838.749 & 13.65 & 1840.651 & 13.64 & 1841.629 & 13.65 + 1842.762 & 13.64 & 1843.693 & 13.65 & 1850.693 & 13.65 & 1866.816 & 13.64 + 1870.685 & 13.65 & 1884.560 & 13.63 & 1886.506 & 13.63 & 1886.609 & 13.64 + 1891.671 & 13.62 & 1897.545 & 13.65 & 1899.538 & 13.65 & 1913.551 & 13.65 + 1913.692 & 13.64 & 1914.521 & 13.65 & 1921.524 & 13.65 & 1921.735 & 13.63 + 1922.538 & 13.63 & 1922.626 & 13.64 & 1923.551 & 13.65 & 1923.698 & 13.66 + 2220.676 & 13.25 & 2221.686 & 13.24 & 2224.758 & 13.24 & 2225.762 & 13.25 + 2226.731 & 13.25 & 2227.621 & 13.26 & 2227.753 & 13.24 & 2230.640 & 13.25 + 2231.609 & 13.26 & 2231.696 & 13.25 & 2231.838 & 13.24 & 2232.681 & 13.24 + 2240.618 & 13.24 & 2253.610 & 13.22 & 2255.560 & 13.22 & 2259.588 & 13.21 + 2300.629 & 13.18 & 2300.741 & 13.17 & 2301.594 & 13.18 & 2301.686 & 13.17 + 2302.596 & 13.17 & 2302.690 & 13.17 & 2304.639 & 13.18 & 2309.548 & 13.17 + 2309.701 & 13.17 & 2310.584 & 13.16 & 2314.596 & 13.14 & 2315.564 & 13.15 + 2315.635 & 13.16 & 2317.588 & 13.14 & 2320.512 & 13.15 & 2322.564 & 13.15 + 2322.617 & 13.14 & 2324.535 & 13.14 & 2324.604 & 13.15 & 2327.530 & 13.14 + 2329.557 & 13.13 & 2329.628 & 13.14 & 2330.579 & 13.13 & 2334.553 & 13.12 + 2335.586 & 13.11 & 2335.643 & 13.11 & 2341.592 & 13.09 & 2362.512 & 13.08 + 2549.770 & 13.02 & 2553.815 & 13.02 & 2555.774 & 13.02 & 2562.744 & 13.02 + 2569.698 & 13.02 & 2576.730 & 13.00 & 2587.696 & 13.01 & 2602.700 & 12.98 + 2603.705 & 12.99 & 2612.719 & 12.98 & 2617.693 & 12.98 & 2617.782 & 12.98 + 2621.631 & 12.98 & 2650.640 & 12.98 & 2658.536 & 12.97 & 2659.519 & 12.98 + 2660.537 & 12.97 & 2662.566 & 12.97 & 2676.539 & 12.98 & 2676.627 & 12.98 + 2683.542 & 12.98 & 2684.526 & 12.98 & 2684.632 & 12.98 & 2691.576 & 12.98 + 2696.597 & 12.99 & 2697.622 & 12.99 & 2702.596 & 12.99 & 2709.596 & 12.98 + 2714.595 & 12.99 & 2722.580 & 12.98 & 2723.567 & 12.99 & & + | the extremely young cluster ic 348 has been monitored photometrically over 5 observing seasons from dec 1998 to march 2003 in cousins _ i _ with a 0.6 m telescope at van vleck observatory .
twenty - eight periodic variables and 16 irregular variables have been identified .
the variability study is most sensitive for stars with i @xmath0 14.3 mag ; at that brightness level we find that 24 of the 27 known pms cluster members in the monitored field are variables , illustrating the value of photometric monitoring for identifying pms cluster members . among this brighter sample , 14 of the 16 known k or m - type wtts
were found to be periodic variables , while all 5 of the known ctts were found to be irregular variables . in the full sample , which includes 150 stars with i mag as faint as 18
, we find that 40% of the 63 wtts are detected as variables , nearly all of them periodic , while 55% of the 20 ctts are also detected as variable , with _
none of them periodic .
_ our study suggests that 80 - 90% of all wtts in young clusters will be detected as periodic variables given sufficiently precise and extended monitoring , whereas ctts will reveal themselves primarily or solely as irregular variables .
this has clear consequences for pms rotational studies based on photometric periods , suggesting that any such sample may be biased against stars which are currently actively accreting ( i.e. ctts ) .
we examine the stability of the periodic light curves from season to season .
all periodic stars show modulations of their amplitude , mean brightness and light curve shape on time scales of less than 1 yr , presumably due to changes in spot configurations and/or physical characteristics . in no case , however , can we find definitive evidence of a change in period , indicating that differential rotation is probably much less on wtts than it is on the sun . while some stars show a hint of what could be cyclic behavior analogous to the sunspot cycle , no clear cycles could be found .
it appears that most of the variation in light curve shape is caused by redistribution of spots on the surface rather than by an increase or decrease in the areal spot coverage .
while most of the variables are of k or m spectral class , we do confirm the existence of three , low amplitude , periodic g stars .
the rotation periods of these more massive stars are short compared to the bulk of the sample ; it appears that mid - k to early m ( i.e @xmath10.5 m@xmath2 ) represents a minimum in mean rotation rate for extremely young stars . among the non - periodic stars , we report the detection of two possible uxors as well as a pre - main sequence star , hmw 15 , which apparently undergoes an eclipse with a duration exceeding three years . |
the nature of the first excited states of conjugated polymers is an important and still unsolved question in condensed matter sciences @xcite . knowing if they are band to band excitations or exciton states , if polarons , bipolarons or solitons are stable quasiparticles in the doping case , are fundamental issues for the understanding of the electronic properties of these compounds .
the low - lying excited states are supposed to be suitably described by the well - known @xmath0-electron pariser - parr - pople ( ppp ) hamiltonian .
this model hamiltonian takes into account both strong electron - phonon and electron - electron interaction terms yielding only exact numerical solutions for the smallest oligomers@xcite . for the thermodynamic limit ,
the situation remains unclear since the calculations for the ground state and the excited states , including electron correlations , are uneasy to achieve and some drastic approximations are needed @xcite .
however , a first qualitative understanding of this complicated physics can be done by the use of some simplified hamiltonian .
for instance the rice - gartstein s molecular exciton model@xcite , similar to previous works @xcite , is useful for a qualitative description of the linear absorption of conjugated polymers . on the other hand
, the molecular holstein s model gives a simplified picture of the polaron states@xcite .
recently , an approximate scheme to build the ground and the first excited states has been proposed@xcite . with this method ,
starting from the ppp hamiltonian , one reaches a rice - gartstein like model ; the excitations relevant for linear absorption are then easy to obtain and the results are comparable with those from more tedious methods @xcite . in this paper , we will show that the same procedure permits to derive formally , from the very same ppp hamiltonian , the simple molecular holstein s model for the polaron state in conjugated polymers .
polarons are thought to be important for the understanding of the charge transport in these compounds , and the possibility to study these non - linear states at a correlated level in an easy formalism is needed and valuable .
we choose a simple dimerized linear chain as an effective model for conjugated polymers ; this chain is characterized by @xmath1 and @xmath2 , the double and single bond lengths , respectively .
extending our method to a realistic geometry is straightforward but the essential of the physics is reached with this simplified picture .
let us now briefly introduce the pariser - parr - pople hamiltonian which is our starting point @xmath3 where @xmath4 , ( @xmath5 ) is the creation ( destruction ) operator of an electron in site _
n _ with spin @xmath6 ; @xmath7 is the hopping term which includes via the electron - phonon interaction a linear dependence upon the length of the bond ( _ n_,_n+1_)@xcite . in the case of a dimerized linear chain ,
this dependence gives two distinct hopping integrals @xmath8 and @xmath9 for the double and the single bonds respectively ( @xmath10 ) ; they could be written as @xmath11 where @xmath12 is the hopping integral without dimerization , @xmath13 is the electron - phonon interaction and @xmath14 is a measure of the dimerization giving the difference of the lengths of single and double bonds@xcite .
the coulomb term is parametrized following ohno , where the effect of the @xmath6 electrons is considered via a phenomenologic screening , @xmath15 where @xmath16 is the distance ( in @xmath17 ) between two electrons localized on site _
n _ and _ m _ @xcite .
we also write this term as @xmath18 and @xmath19 for convenience . in view to link up the ppp hamiltonian and the molecular crystal models , the rice - gardstein and holstein models , we choose the monomer self - consistent orbitals as basis functions -
this is the so - called exciton - basis @xcite .
this choice is of course led by the dimerization . in our case ,
the monomers are the double - bonds and their self - consistent orbitals are associated with the following creation ( destruction ) operators for the bonding and anti - bonding orbitals : @xmath20 and @xmath21 ; here @xmath22 indexes the double bonds . with this specific choice of local basis operators ,
the electronic configurations are built by combining different kinds of local configurations ( lc)@xcite . in order to get a tractable model , we truncate the hilbert space by choosing a small set of different lc which will be the elemental building blocks for the electronic configurations @xcite .
these lc are the so - called generative local configurations ( glc ) in @xcite .
we may notice that this method shows some similarities with the valence bond method used efficiently for the studies of oligomers @xcite but with the important difference that atomic sites are replaced by monomer units with internal electronic structure ( double bonds here ) .
the configurations build from glc are diagonal with respect to the hopping term @xmath8 on contrary to the valence bond configurations which are diagonal with respect to the coulomb term .
each glc is a set of several valence bond diagrams , chosen to be the adequate ones for a reasonable description of polymer states . in this work ,
we first improve the proposed ground state of ref .
@xcite by enlarging the set of electronic configurations used to describe it ( section ii ) .
second , we consider the case with an extra electron on the chain and show that , if one authorizes small lattice distortions around the extra particle , our treatment allows to reach quite naturally a holstein like model but expressed in terms of many - body particle states ( section iii ) .
we keep as glc for the ground state the lc which appear the most relevant in calculations performed on small oligomers@xcite . in ref @xcite ,
only three lc were considered ; they are named f - lc , d - lc and ct@xmath23-lc and are schematically represented in figure ( [ lcgs].a ) .
this approximation could appear rather abrupt , but it is sufficient to get a correct qualitative picture of the linear absorption spectra as it was shown in @xcite ; moreover , even at this level of approximation , the results are quantitatively comparable with the results of more tedious calculations @xcite . in this work ,
we propose some natural improvements to this first approximation by extending the set of glc . in a first improvement , we add to the previous set of glc , the so - called triplet - triplet lc , tt - lc , shown in figure ( [ lcgs].b ) , where two nearest - neighbour ( n.n . ) localized triplets are combined into a singlet . together with the three first lc , they are the major constituents of the ground state wave function in small cluster calculations@xcite . in a second improvement , we enlarge again the set of glc by including in it the lc which interact directly with the four previous selected ones ( figure[lcgs].c ) . in the following
, only the first case is treated explicitely .
we develop in full detail our proposed way to get the ground state wave function with the four selected glc . the case with the complete set of lc represented in figure ( [ lcgs ] ) can be treated following the same scheme ; only the obtained results are then given .
first , we introduce the four glc , their associated creation operators and their energies . *
the named f - lc is associated with the creation operator @xmath24this define the lowest lc in the range of parameter of interest ; therefore we choose as reference state @xmath25 where @xmath26 denotes the state without any @xmath0 electron .
the state @xmath27 is the ground state considered in the molecular crystal approaches @xcite ; there , the linear dimerized chain is simply identified to an one - dimensional crystal of ethylene without any electronic correlations .
+ with respect to @xmath27 , @xmath28 which is simply the identity operator . in the following ,
all the creation operators and the energies are defined with respect to @xmath27 . *
the named d - lc is associated with the creation operator @xmath29and with energy given by @xmath30 .
+ the f and d - lc describe the dynamics of the @xmath0-electrons coupled by pairs into each monomer : the two electrons are independent in f - lc , whereas d - lc introduces intramonomer electronic correlation . in the strong dimerization limit , these two lc are sufficient to give a good approximation of the ground state ; the system is then very close to a true molecular crystal . for small or intermediate dimerization , it is however necessary to consider more extended lc or , in other words , some fluctuations around the molecular crystal limit .
this is done by introducing two more lc extended over two n.n
. monomers . *
the named ct@xmath23-lc is associated with the creation operator @xmath31and with energies given in the case of a linear dimerized chain by @xmath32 .
the last term , in bracket , is the attractive interaction between the electron and the hole due to the long - range part of the ohno potential .
+ the ct@xmath23-lc introduces n.n .
intermonomer charge fluctuations , reproducing the conjugation phenomenon in a minimal way . * last , the named tt - lc is associated with the creation operator @xmath33and with energy given by @xmath34 .
+ in this lc , two triplets appearing in n.n .
monomers are combined into a singlet ( figure([lcgs].b ) ) .
it was shown to be important for the first time in the work of schulten and karplus @xcite where it was recognized as a major constituent of the low - lying excitations , the famous @xmath35 state , optically forbidden . in the ground state , which is our interest here
, the importance of this lc can be comparable to the d - lc one @xcite .
we may notice that a similar treatment for the ppp hamiltonian was proposed a few years ago to study the spin - charge separation mechanism in the limit of strong dimerization @xcite . with our choice of four glc , all possible electronic configurations
are then build up .
they are characterized by the number of d , ct@xmath23 and tt - lc , @xmath36 , @xmath37 and @xmath38 respectively , and by the positions of these different glc .
the positions of the d , ct@xmath23 and tt - lc are labelled by the coordinates @xmath39 ( @xmath40 ) , @xmath41 ( @xmath42 ) and @xmath43 ( @xmath44 ) respectively . the necessary non - overlapping condition between lc
is supposed to be fulfilled all along the paper - the lc behave as hard core bosons .
the electronic configurations are then expressed as @xmath45 the glc are all neutral local configurations , therefore the energy of ( [ espacemodel ] ) is independent of the relative positions between lc and entirely determined by the number of each glc . @xmath46 at this point , we have to mention an incorrect statement in @xcite where it is saying that the energy of the configurations made of f , d and ct@xmath23-lc depends on the relative positions of the ct@xmath23-lc .
this statement is actually wrong , however , this simplification goes in favor of our treatment ( indeed , it was not possible to do calculations with this statement and finally the energy ( [ energiemodel ] ) was also adopted in @xcite ) . the way we choose to diagonalize the ppp hamiltonian in the reduced hilbert space spanned by the electronic configurations ( [ espacemodel ] ) follows from @xcite .
first , we reorganize the configurations ( [ espacemodel ] ) . we make linear combinations from the states with @xmath36 d - lc , @xmath38 tt - lc localized at sites @xmath47 and @xmath37 ct@xmath23-lc localized at sites @xmath48 .
since we are interested , at the end of the day , only by the lowest state in energy ( the ground state ) , we can consider only the linear combinations of highest symmetry @xmath49where the summation is carried over the @xmath50 possible configurations .
the energy of these combinations is still given by ( [ energiemodel ] ) .
the states ( [ excocorr ] ) interact between them by the following term @xmath51 the tri - diagonal matrix , where the diagonal part is given by ( [ energiemodel ] ) and the off - diagonal part by ( [ interacd ] ) can be divided into sub - matrices characterized by @xmath37 localized ct@xmath52-lc and @xmath38 localized tt - lc but with a variable number of d - lc , @xmath36 ( @xmath53 ) ; these sub - matrices can be separately diagonalized and it is easy to show that the resulting lowest states are given by the following expression @xmath54 with energy expressed as @xmath55 where @xmath56the coefficients @xmath57 and @xmath58 of ( [ diago1 ] ) are given by @xmath59 and @xmath60 . with these expressions ,
the double bonds free of ct@xmath52- and tt - lc are correlated independently .
the upper - script , @xmath61 , in ( [ diago1 ] ) is for correlated .
@xmath62 is called intramonomer correlation energy .
the next step toward the evaluation of the ground state is to retain , among all the states resulting from the previous sub - diagonalizations , only the lowest ones given by ( [ diago1 ] ) .
this approximation is well justified since the energy difference between these states and the corresponding lowest excited ones is given by the quantity @xmath63 which is rather high for usual parameters with a value around @xmath64 .
we then reorganized the states ( [ diago1 ] ) into collective excitations of highest symmetry @xmath65^{-\frac{1}{2}}\sum_{\{x(i),y(j)\}}\mid x(1), ... ,x(n_{tt}),y(1), ... ,y(n_{ct})>^{c}\]]still associated with the energy ( [ energiecor ] ) and where the summation runs over the @xmath66 possible configurations .
the ground state is then expressed as a linear combination @xmath67 where the coefficients @xmath68 are determined by solving the following secular equation @xmath69^{\frac{1}{2}}n_{tt}\frac{\sqrt{3}}{2}t_{s}x_{n_{tt}-1,n_{ct}+1}+ [ n_{ct}(n_{tt}+1)]^{-\frac{1}{2}}n_{ct}\frac{\sqrt{3}}{2}t_{s}x_{n_{tt}+1,n_{ct}-1}=0 \end{array}\ ] ] where @xmath70 the equation ( [ equationgs ] ) is not solvable with the interaction term ( [ i ] ) .
next , and last , we do an approximation on the term @xmath71 by assuming @xmath72 this is a very good approximation of ( [ i ] ) , if the number of glc extended over two monomers , @xmath73 , is small@xcite .
consequently , this treatment will be justified if in the final wave function , the most important configurations are the ones with a small value of @xmath74 ; this is actually the case as it can be seen from the work of ref .
@xcite and as it appears , at the end of the day , in this study . with the last simplification
, the problem is mapped onto @xmath75 independent three - level systems .
one write @xmath76 where @xmath77 takes the integer part , @xmath78 and @xmath79 are real constants to be determined . inserting this definition in ( [ equationgs ] ) and after some algebraic manipulations one finds that the problem goes back to calculate the lowest eigenvalue , @xmath80 , of the following 3 by 3 matrix @xmath81 the ground state energy is then simply divided into two different components @xmath82 the first part is the intramonomer correlation energy defined by the first subdiagonalization ; it is obtained by correlating independently the @xmath83 double bonds .
the second part is the intermonomer fluctuation energy defined by the second subdiagonalization ; it is obtained by considering @xmath75 identical and independent effective three level systems defined by the matrix ( [ 3levels ] ) .
finally , the ground state wave function is clarified by the following two equations @xmath84 the resulting wave function contains , as the energy , two different kinds of components : the first ones localize electrons by pairs in the double bonds ; the second ones introduce n.n .
intermonomer fluctuations , charge fluctuations by means of ct@xmath52-lc and spin fluctuation by means of tt - lc .
the ground state proposed above may be easily improved by adding new local configurations extended over two n.n .
double bonds .
for example , one can include the whole lc represented in the figure ( [ lcgs ] ) ; the lc of ( [ lcgs].c ) are the ones directly coupled to the others .
the strategy is then the same .
first , one takes care of the intramonomer correlation ; second , one builds the collective excitations of highest symmetry ; third , one approximates the part of the resulting interaction connecting configurations which differ by only one lc extended over two monomers in the way of ( [ approx ] ) .
the problem is then equivalent to consider @xmath75 independent seven - level systems ; @xmath80 is then the lowest eigenvalue of the associated 7 by 7 matrix . in order to test our assumptions
from which we propose several ground state wave functions in the form of ( [ wfgs ] ) , we do comparisons , first , for the su - schrieffer - heeger ( ssh ) model . for this model ,
similar to ( [ ppp ] ) but without the complicated coulomb term@xcite , the exact result is well known @xcite .
we compare this result with successively the results given by the model ground state of ref @xcite , the hereafter so - called model i , the one with in addition the tt - lc , the model ii , and , last , the model with all the glc represented in figure ( [ lcgs ] ) , the model iii .
we make comparison in function of the dimerization parameter @xmath85 .
the results are shown in table i where the percentage of the exact energy for our successive approximations are given .
for @xmath86 , the case of complete dimerization , the three models give obviously the exact result . for @xmath87 , the case without dimerization , one gets around 92@xmath88 of the total energy .
a - priori in this limit , one would expect less accurate results since the charge fluctuations of longer range than one play a role ; they contribute in fact only in the missing 6@xmath88 . for @xmath89 , a value often attributed to the polyacetylene
, one gets around 97@xmath88 of the total energy . in conclusion , our approximation seems rather good for realistic cases , within this independent electron model .
next , we do also comparisons for the hubbard model which is well known to be exactly solvable in one dimension @xcite ; this is the model ( [ ppp ] ) with @xmath90 and where only the on - site electron - electron interaction , @xmath91 , is retained . for @xmath92 ,
one gets the ssh model without dimerization for which we obtained around 92@xmath88 of the total energy ( see table i ) . starting from these values
, the agreement monotonically decreases when @xmath91 increases to finally get for infinite @xmath91 , between 77@xmath88 and 79@xmath88 of the total energy , depending on the model ( i , ii or iii ) under consideration .
this discrepancy shows that important lc are missing especially in the strong @xmath91 limit ; for instance , it is easy to see , just by energetic considerations , that , for large enough @xmath91 , the ttt - lc , which is singlet made by three localized triplets , the tttt - lc , which is singlet made by four localized triplets , and so on , may become important for the ground state wave function . with our specific choice of basis set completely localized on the double bonds , the dimerization parameter , @xmath93 , is crucial ; the more it will be important , the more our treatment will be relevant to become exact for a complete dimerization . in the hubbard model , the dimerization is simply missing .
if @xmath94 , the energy of the lc made from localized triplets increases making our approximations more and more reasonable .
last , we do comparison for the so - called extended peierls - hubbard model ; this is the model ( [ ppp ] ) but with only the hubbard term , @xmath91 , and the n.n .
interaction @xmath95 , with the assumption that @xmath96@xcite . on the contrary to the two previous models ,
this model is not integrable , also , we do comparisons with calculations performed with the density - matrix renormalization group ( dmrg ) technique @xcite ; a very recent review of the advances related to this method may be found in @xcite .
the dmrg calculations have been done by e. jeckelmann @xcite following the method developed in ref @xcite .
we compare our approximate results with an extrapolation of the energy per unit cell made from calculations for different lattice lengths up to two hundreds double bonds .
the calculations are performed for a reasonable choice of parameters , @xmath97 and @xmath98 .
the results for several values of the dimerization parameter are confined in table ii .
we see that the errors are always less than 20@xmath88 and are around 13 - 10@xmath88 for realistic parameters . in our opinion ,
the agreements obtained here are satisfactory considering the relative simplicity of the wave functions proposed in this work . moreover , with these approximate wave functions , some analytical insights are now possible which is very new in this range of parameters , appropriated for conjugated polymers
. we do not compare for the moment our results with calculations made for the complete ppp hamiltonian .
nevertheless , since the remaining long range terms of the coulomb potential are of smaller importance than the other terms of the hamiltonian , one can reasonably expect only small quantitative changes in the results obtained with the extended peierls - hubbard model by using the full ppp hamiltonian . before closing this section
, we may say that our wave functions are not variational since , in the way we choose to diagonalize the model , we do two successive sub - diagonalizations with some approximations .
however , it is possible to build some variational wave - functions very similar to ( [ wfgs ] ) . by the way
, works are already done to propose a variational version of the model ii @xcite and other are in progress for the model iii @xcite . in other way
, a very efficient matrix - product - ansatz is also proposed in @xcite .
compared to the work developed in @xcite , one can say that our proposed way to diagonalize the ppp - hamiltonian in the selected sub - hilbert space is a very good approximation for appropriated parameters .
in this part , we consider the situation with one additional charge .
we treat explicitely the case of an additional electron but , the case of the removal of one electron can be treated exactly in the same way .
we show that this problem can be describe , with some approximations , in terms of quasi - particles which obey to simple effective hamiltonian . for a rigid lattice
, we get a one dimensional tight binding hamiltonian .
if one authorizes some distortions of the lattice around the extra particle , we get at second order in the distortion coordinates , a holstein like model@xcite . in both cases , the parameters of these one - electron models are related to the ppp one s . in this work ,
we are not attempted to derive quantitative results .
our goal , based on semi - quantitative results , is to open up a way between a true many - body model given by the ppp - hamiltonian and more simple one - electron models as the holstein s model for polaronic states . because it is not possible to solve the ppp model and since the important physical ingredients for an understanding of conjugated polymers are still not fully recognized @xcite , the derivation of more effective models is needed in order to get some physical insight .
this work , and the very related one of ref .
@xcite , goes in this direction . for convenience ,
we choose in this part the simplest description for the ground state given by the model i , using f , d and ct@xmath23-lc .
since the model i already contains the most important local constituents for the ground state wave function , namely the f and ct@xmath23-lc , we believe the results would not changed dramatically with a better description - by using the model ii or iii .
then , if we define @xmath99^{-1/2}\sum_{\ { y(i),z(j)\ } } ct^{\dag}_{y(1)} ... ct^{\dag}_{y(n_{t})}d^{\dag}_{z(1)}\cdots d^{\dag}_{z(n_{d})}\mid 0>\ ] ] where the summation is over the @xmath100 possible configurations , the ground state wave function is simply written as @xmath101 where @xmath102 , @xmath103 , @xmath104 , @xmath105 and @xmath106 .
@xmath80 is then the lowest eigenvalue of the 2 by 2 matrix obtained from ( [ 3levels ] ) by suppressing the effective level corresponding to the tt - lc@xcite . for a typical choice of parameters relevant for conjugated polymers@xcite ,
the most probable lc is the f - lc ( @xmath107 and @xmath108 ) ; typical values for the energies are given by @xmath109 and @xmath110 .
an additional charge disturbs the electronic cloud more or less strongly depending on the system under consideration .
it could be a local distortion where the extra particle rearranges the system in short distances to create around it what it is called polarization cloud ; this is the case for usual semi - conductors . on the contrary
, it could be a complete rearrangement of the system as for strongly correlated systems@xcite . in our case ,
the first behaviour is concerned and a quasi - particle picture is reached .
we describe the perturbations caused by the extra - particle - the polarization cloud - by introducing a new set of lc more or less extended , which we call charged local configurations ( c - lc ) ; the term `` charged '' means that they contain explicitely the extra - particle .
some example of c - lc , extended over one , two and three double bonds are shown in figure ( [ lcp ] ) where the extra - electron is represented by the thick arrow . in the case of a `` macroscopic '' rearrangement of the electronic structure - as it could be the case for strongly correlated systems - the maximum extension of the relevant c - lc would be of the order of the system size . in our case , this critical size is of the order of some monomer units only . all around these c - lc , we assume the electronic structure unchanged with respect to the ground state ; therefore , we consider such charged configurations - strictly speaking , these are linear combinations of electronic configurations but we adopt the proposed terminology for convenience - @xmath111where , @xmath112 , @xmath113 and @xmath114 are some c - lc extended over one , two and three nearest - neighbour double bonds respectively , @xmath115 ( @xmath116 , @xmath117 , @xmath118 ) is the part on the left ( right ) of the c - lc , described in the same way as @xmath119 . with this crude description , a c - lc acts as a dramatic boundary which simply interrupts the chain :
the system is separated into two chains both described exactly as the ground state ; the boundary contains explicitely the extra - particle within a defined c - lc .
the more extended c - lc are inserted in the ground state in the same way as ( [ excitations ] ) . with our approximation , the energy of each charged configuration ( as ( [ excitations ] ) ) is given by the addition of two different terms .
the energy of the isolated c - lc and the energy of the external parts to the left and to the right of the c - lc .
since @xmath115 and @xmath116 are neutral , the external parts do nt interact via the coulomb potential with the extra - particle .
however , the configurations ( [ excitations ] ) must be improved for more quantitative results . indeed , in a better description , because of the presence of the p - lc , the relative weight of the f , d and ct@xmath23-lc , controlled by the coefficients @xmath120 , @xmath121 , @xmath122 and @xmath123 should depend on their positions on the chain .
moreover , with an additional particle , the electron - hole symmetry is broken .
all the lc used in @xmath115 and @xmath116 are in the same sector of symmetry - the proper one for the building of the ground state .
this is the case , for instance , of the ct@xmath23-lc where the charge transfer on the right and on the left are of the same importance . in the presence of the p - lc
these two charge transfers are no longer equivalent ; the symmetry is broken and this implies a coulombic interaction between the p - lc and the external parts .
these effects , not considering in this work , would certainly modify the polarisation cloud in a sensitive way .
one can say , in other words , that the embedding " of the c - lc , due to the part @xmath124 and @xmath116 of ( [ excitations ] ) , are not treated efficiently in this work .
we believe this is here the main point to be improved in the future for more quantitative results . for usual values of the ppp model , one kind of charged configurations is smaller in energy than the other and in such way that a perturbative treatment is possible to do .
these configurations are due to the c - lc , referred as the p - lc hereafter ( p stands for particle ) , associated with the following creation operator @xmath125 and represented in figure ( [ lcp].a ) .
the extra - particle is immersed in the reference vacuum and gives the following charged configurations @xmath126 where @xmath22 referred to the position of the p - lc , and with an energy given by @xmath127 , with @xmath128 and @xmath129 , the energy of the isolated p - lc . by comparing with ( [ egs ] )
, we see that there is a loss of intramonomer correlation energy and a loss of intermonomer fluctuation energy with respect to the ground state ; indeed , the additional electron occupies a site in which one can not place d and ct@xmath23-lc .
this loss of energy is more important for the more extended c - lc . in the following ,
we consider explicitely only the charged configurations ( [ particle ] ) since the effects of the other charged configurations can be taken into account by perturbation . with our approximation , because of the n.n . hopping integral
, the p - lc can hop on the lattice with the help of the f - lc or the ct@xmath23-lc . with the former
, the p - lc can hop from site to site on the monomer lattice ( see figure [ nnhopping ] ) .
@xmath130 in this expression , the product @xmath131 gives the probability to find a f - lc in the wave function ( [ pf ] ) ; the factor @xmath132 in ( [ hopping ] ) comes from our choice to work with the monomer orbitals .
moreover , with our approximation , there exists also a n.n.n . hopping process with the help of the more extended glc , the ct@xmath23-lc ( see figure [ nnnhopping ] ) .
@xmath133the additional factor 2 in the denominator comes from the fact that only one term from the ct@xmath23-lc ( see equation ( [ ctlc ] ) ) is involved during the transfer ; the coefficient @xmath134 gives the probability to find a ct@xmath23-lc in the ground state wave function ( [ pf ] ) .
the n.n.n .
transfer is of course less important than the n.n .
one s . with the values for the parameters we use here ,
the values of these two hopping processes differ by one order of magnitude .
therefore , we neglect the n.n.n .
effective hopping term in this work .
the extra - particle ( p - lc ) can be dressed by perturbation .
some effects of the other c - lc appear then in renormalized energy and n.n . hopping term for the extra - particle .
this dressing of the p - lc can be simply done by a second order perturbative treatment , giving , in one hand , the so - called polarization energy @xmath135and , in the other hand , some corrections for the n.n . hopping integral @xmath136 @xmath137 in these expressions @xmath138 and @xmath139
are some interacting terms between the p - lc and other c - lc .
the inequalities @xmath140 are respected for the values of the parameters we use which guaranty the relevance of a perturbative treatment .
after that , we have reached a quasi - particle picture , the quasi - particle being represented by the p - lc . in principle , many c - lc give some contributions to the perturbative series ( [ epolarisation ] ) and ( [ effhopping ] ) . however , because the states ( [ excitations ] ) ignore many effects due to an inappropriate embedding , as we already mentioned , we believe it is not useful to carry out the full calculation . consequently , we do here a simplified treatment for the dressing of the extra - particle which we believe contains anyhow the most important contributions to ( [ epolarisation ] ) and ( [ effhopping ] ) .
this simplified treatment consists to consider the c - lc not embedded in the ground state defined by ( [ pf ] ) but in a simplified vacuum made of only f - lc .
since , the f - lc is the very most important lc in the ground state ( [ pf ] ) , we believe this simplified treatment sufficient to capture the most important parts of the polarisation energy and the effective hopping term . moreover , among the remaining charged configurations only a few are incorporated in the perturbative treatment ; they are shown in the figure ( [ lcp].b ) . by this last simplification
we neglect all the c - lc shown in the figure ( [ lclongrangepolar ] ) which take into account some long range polarisation effects ; these c - lc are numerous but their total effect on ( [ epolarisation ] ) and ( [ effhopping ] ) are small and they do nt participate sensitively to the binding energy of the polaron state which is the main quantity we are looking for here . with this treatment ,
the corrections for the hopping term remain always negligible in the range of parameters of interest ; we will therefore neglect these last corrections , @xmath141 .
after the dressing operation , we obtain formally a one particle like problem with two characteristic energy terms @xmath142 , the site energy of the additional electron with respect to the ground state , and @xmath136 , the hopping term , which are functions of the ppp parameters : @xmath143 and @xmath144 . if we suppose a rigid lattice , the problem can obviously be diagonalized , giving a band centred at @xmath142 , with a bandwidth of @xmath145 . in the case of the ssh hamiltonian@xcite and by neglecting @xmath80 and @xmath146 , the bottom of the band is given by @xmath147 , the exact result ; with inclusion of these corrective terms , this energy becomes slightly overestimated .
the effective mass associated with the p - lc is given by @xmath148 ( @xmath57 is the unit cell length ) which is of course higher than the effective mass of a free particle on the bottom of the conduction band .
with the coulomb term , @xmath142 increases and @xmath149 stays unchanged . in conclusion , for a rigid lattice , we have reached a simple tight binding hamiltonian - the so - called hckel model .
last , one may say that such approach is quite close in spirit of a recent work of j. grafenstein et al . , where an effective tight - binding model is derived at ab - initio level by means of an incremental method @xcite .
now we allow a relaxation of the lattice . for simplicity ,
we choose a model displacement where the two atoms of the same double bond move with the same amplitude @xmath150 but in opposite directions ( cf . figure ( [ deformation ] ) ) .
the two parameters @xmath142 and @xmath136 depend now on the lattice coordinates via mainly the linear dependence of the two hopping terms , @xmath151 and @xmath152 . on the contrary ,
the coulombic terms remain almost unchanged after a small displacement .
the contributions due to these displacements to @xmath153 and @xmath136 are small so we make a linear expansion with respect to @xmath154 of these two quantities @xmath155 @xmath156 where @xmath157 , @xmath158 and @xmath159 are functions of ppp parameters and @xmath13 is the electron - phonon interaction term@xcite .
the extra elastic constraint of the dimerized chain due to the lattice relaxation in presence of an additional charge is expressed as @xmath160\ ] ] where @xmath161 , the spring constant , is defined relatively to the dimerized equilibrium structure .
the coefficients @xmath162 of the holstein polaron wave function@xcite , @xmath163 , are determined by minimization of the corresponding total energy , @xmath164 , with respect to the lattice coordinate @xmath165 . at the second order in @xmath165 and taking into account that @xmath166 in conjugated polymers @xcite , we obtain the characteristic equations of the molecular holstein s model @xmath167c_{n}+jc_{n+1}+jc_{n-1}=0\ ] ] @xmath168 where the coefficients are expressed in function of the ppp parameters : @xmath169 , @xmath170 and @xmath144 . by injecting ( [ relaxation ] ) into ( [ holstein ] )
we obtain the non - linear schrdinger equation which gives the coefficients of the wave function ; the second equation connects in a simple manner these coefficients and the lattice deformation .
the analytical solution of these two equations in the continuum limit@xcite , valid for the `` large '' polaron case , gives the well known polaronic wave function @xmath171 with @xmath172 , @xmath173 and @xmath174 ; the polaron state is localized around @xmath175 , an undetermined quantity because of the translational invariance of the system .
the associated binding energy of the polaron state is given by @xmath176 .
we evaluate these quantities for several choices of parameters by the following sequence of calculations .
first we optimize the dimerized geometry referring to a spring constant , k , relative to a hypothetical undimerized geometry@xcite ; then , we evaluate @xmath161 , calculating the second derivative of @xmath177 with respect to the dimerization coordinate at the geometrical equilibrium .
second , we solve the equations ( [ holstein ] ) and ( [ relaxation ] ) . in the continuum version of the ssh hamiltonian limit@xcite , analytical expressions have been given .
our results always overestimate the reported values . for example , with @xmath178 , @xmath179 and @xmath180 , we get @xmath181 in place of @xmath182 . in the same manner ,
our method also overestimate the value of the dimerization .
these overestimations occur naturally from our starting point which relies on a molecular description . besides it has been shown that the ssh hamiltonian is never equivalent to the holstein s model for the dimerized linear chain@xcite , so that the approximations of our model can not be expected to lead to a good agreement in this case .
however our approximations will cope better when the coulomb interaction is taken into account ; then the energies of charge fluctuation components decrease with their extensions , due to the long range part of the potential .
this fact is in favour of our approximation .
furthermore the value of @xmath183 used in this example is the appropriated one for the ssh hamiltonian@xcite , but seems not to be in agreement with the experimental results obtained for small oligomers@xcite .
an higher value must be taken , favouring again our description .
if one adds the ohno potential , the binding energy decreases : as example , for the same choice of parameters and @xmath184 , we get @xmath185 .
finally , taking the same parameters but with a more appropriate value for @xmath183 , @xmath186 , we get a reasonable equilibrium geometry characterized by @xmath187 and @xmath188 .
moreover we get the following values @xmath189 , @xmath190 , @xmath191 and the binding energy for the polaron decreases to @xmath192 . in any case , we stay around traditionally adopted values . before closing this section ,
note that with such a low binding energy , expected for conjugated polymers , the quantum fluctuations of the lattice should be explicitely considered .
however , it is for the moment totally hopeless to introduce additional bosonic variables in the full ppp hamiltonian .
in conclusion , we have proposed a simplified treatment of the ppp hamiltonian which is typically a diagonalization of this hamiltonian in a restricted hilbert space .
the method adopted , using monomer orbitals , is a natural way to bridge the gap between small cluster and polymer calculations@xcite .
the ground state is composed of intermonomer nearest - neighbour fluctuation components in the background of coupled electrons by pairs localized on monomers .
comparisons with dmrg results for the extended peierls - hubbard model show satisfactory agreements considering the simplicity of our proposed wave functions .
the electronic excitations are then described as local perturbations moving in this `` vacuum '' . for an appropriate set of parameters ,
this description gives rather good values for the dimerisation and for the energies of the excited states active in one photon spectroscopy@xcite . in the doped case ( 2n+1 particles )
studied here , following the adiabatic scheme proposed by holstein@xcite , we show that our model leads naturally to a holstein s polaron like problem .
however our description differs drastically from the holstein s polaron image in the sense that it is able to describe the behaviour of a strong correlated ( 2n+1 ) particle state whereas the holstein s model considers only the additional particle in interaction with a deformable medium .
the obtained binding energy of the polaron is of the correct order of magnitude .
some improvements are suitable concerning , first , the ground state where more extended glc must be considered in order to reproduce more accurately the delocalization proper to @xmath0-systems . in other hand , variational calculations based on the very same ideas are possible @xcite . for the doped case , we believe the first point would be to improve the description of the vacuum in presence of the extra - particle .
even if it is difficult to test our derivation in part owing to a lack of accurate calculations including correlation effects , we think that our formulation keeps the essential behaviour of the considered physical phenomenon and believe that it could be useful for future more advanced studies in part because of its relative conceptual simplicity and of its ability to give analytical expressions .
for example , the behaviour of polaron states in presence of a strong electric field @xcite , which corresponds to a common situation in electroluminescence studies , would be to consider taking into account the effects of the strongly correlated n particle system .
, edited by n.s .
sariciftci ( world scientific publishing , singapore , 1997 ) .
m. chandross , y. shimoi and s. mazumdar , phys .
b 59 * , 4822 ( 1999 ) .
d. baeriswyl , d.k .
campbell and s. mazumdar , in conjugated conducting polymers , edited by h. kiess ( springer - verlag , heidelberg , 1992 ) , pp 7 - 133 .
rice and y.n .
gartstein , phys .
* 73 * , 2504 ( 1994 ) ; y.n .
gartstein , m.j .
rice and e. conwell , phys . rev .
* b52 * , 1683 ( 1995 ) .
simpson , j. am .
soc . * 77 * , 6164 ( 1955 ) ; h.c .
longuet - higgins and j.n .
murrel , proc .
a68 * , 602 ( 1955 ) ; j.a .
pople and s.h .
walmsley , trans .
faraday soc .
* 58 * , 441 ( 1962 ) .
t. holstein , ann . of phys .
* 8 * , 325 ( 1959 ) .
bussac , j. dorignac and l. zuppiroli , phys . rev . *
b55 * , 8207 ( 1997 ) .
s. pleutin and j.l .
fave , j. phys .
matt . * 10 * , 3941 ( 1998 ) .
wu , x. sun and k. nasu , phys .
* b52 * , 4849 ( 1995 ) .
p. tavan and k. schulten , phys .
* b36 * , 4337 ( 1987 ) .
k. ohno , theor .
acta * 2 * , 219 ( 1964 ) .
s. pleutin , thesis , paris 7 university , 1997 .
z.g soos and s. ramasesha , in valence bond theory and chemical structure , edited by d.j .
klein and n. trinajstic , elsevier , amsterdam , 1990 .
k. schulten and m. karplus , chem .
* 14 * , 305 ( 1972 ) .
d. mukhopadhyay , g.w .
hayden and z.g .
soos , phys . rev .
* b51 * , 9476 ( 1995 ) .
. schrieffer and a.j .
heeger , phys .
rev * b22 * , 2099 ( 1980 ) .
l. salem , _ molecular orbital theory of conjugated systems _
( benjamin , london 1966 ) . e.h .
lieb and f.y .
wu , phys .
lett . * 20 * , 1445 ( 1968 ) .
p. fulde , _ electron correlations in molecules and solids _ , 3rd edn .
springer series in solid - state sciences ( springer , berlin , 1995 ) .
e. jeckelmann , phys . rev . *
b57 * , 11838 ( 1998 ) .
e. jeckelmann , private communication .
white , phys .
rev . * b48 * , 10345 ( 1993 ) .
, lecture notes in physics , eds i. peschel , x. wang and k. hallberg , springer - verlag , 1999 .
s. pleutin , e. jeckelmann , m.a .
martin - delgado and g. sierra , preprint cond - mat/9908062 , submitted to prog .
s. pleutin , in preparation .
martin - delgado , g. sierra , s. pleutin and e. jeckelmann , preprint cond - mat/9908066 , submitted to phys .
* b*. j. grafenstein , h. stoll and p. fulde , phys . rev . *
b55 * , 13 588 ( 1997 ) .
campbell , a.r . bishop and k. fesser , phys
. rev . * b26 * , 6862 ( 1982 ) .
.percentage of the exact energy obtained with the different models studied here for the s.s.h .
hamiltonian . the model i contains the f , d and ct@xmath52-lc
; the tt - lc are added for the model ii and the whole set of lc shown in figure ( [ lcgs ] ) are considered for the model iii . [ cols="^,^,^,^ " , ] | starting from the @xmath0-electron pariser - parr - pople ( ppp ) hamiltonian which includes both strong electron - phonon and electron - electron interactions , we propose some strongly correlated wave functions of increasing quality for the ground state of conjugated polymers .
these wavefunctions are built by combining different finite sets of local configurations extended at most over two nearest - neighbour monomers . with this picture ,
the doped case with one additional particle is expressed in terms of quasi - particle .
thus , the polaron formation problem goes back to the study of a holstein like model . |
a _ spherical curve _ is a smooth immersion of the circle into the sphere where the self - intersections are transverse and double points ( we call the double point _ crossing _ ) . in this paper
we assume every spherical curve is oriented , and has at least one crossing .
we represent , if necessary , the orientation of a spherical curve by an arrow as depicted in the left - hand side of fig .
[ ex - reductivity ] .
a spherical curve @xmath0 is _ reducible _ and has a _
reducible crossing _
@xmath1 if @xmath0 has a crossing @xmath1 as shown in fig .
[ red ] , where @xmath2 and @xmath3 imply parts of the spherical curve .
@xmath0 is _ reduced _ if @xmath0 is not reducible such as the spherical curves in fig .
[ ex - reductivity ] .
note that around a reducible ( resp .
non - reducible ) crossing , there are exactly three ( resp .
four ) disjoint regions , where a _ region _ of a spherical curve is a part of the sphere divided by the spherical curve .
+ a _ half - twisted splice _ is the local transformation on spherical curves as depicted in fig .
[ halftwisted ] ( @xcite ) .
then the inverse is the transformation depicted in fig .
[ halftwisted - i ] . in this paper
we call the inverse of the half - twisted splice _ inverse - half - twisted splice _ , and denote by @xmath4 .
we remark that the half - twisted splice and the inverse - half - twisted splice do not preserve the orientation of spherical curves .
then we give an orientation again to the spherical curve we obtain .
we also remark that the half - twisted splice and the inverse - half - twisted splice do not depend on the orientations of spherical curves , but depend only on the relative orientations .
now we define the reductivity .
the _ reductivity _
@xmath5 of a spherical curve @xmath0 is the minimal number of @xmath4 which are needed to obtain a reducible spherical curve from @xmath0 .
for example , a reducible spherical curve has the reductivity 0 , and the spherical curves @xmath0 , @xmath6 and @xmath7 in fig
. [ ex - reductivity ] have the reductivity 1 , 2 and 3 , respectively ( see fig .
[ eight - two ] for @xmath6 .
note that we can not obtain a reducible curve by any single @xmath4 from @xmath6 . ) is 2.,width=377 ] in this paper , we show the following theorem : every spherical curve has the reductivity four or less .
[ main ] this implies that we can obtain a reducible spherical curve from any spherical curve by four or less @xmath4 .
we have the following question .
is it true that every spherical curve has the reductivity three or less ?
[ main - q ] in other words , is it true that there are no spherical curve with reductivity four ?
the rest of this paper is organized as follows : in section 2 , we discuss the properties of reductivity by considering chord diagrams , and prove theorem [ main ] . in section 3 , we study the unavoidable sets of tangles for spherical curves as an approach to question [ main - q ] .
in this section we show theorem [ main ] by using chord diagrams . we consider a spherical curve @xmath0 as an immersion @xmath8 of the circle into the sphere with some double points ( crossings ) .
a _ chord diagram _ for a spherical curve is an oriented circle considered as the preimage of the immersed circle with chords connecting the preimages of each crossing ( @xcite ) .
a chord diagram is for a reducible spherical curve if and only if the chord diagram could have a chord without crossings .
for example , the chord diagram in fig . [ chord - ex ]
has the chord labeled 4 without crossings .
the move @xmath4 at a crossing @xmath1 on a spherical curve corresponds to the move shown in fig .
[ a - on - chord ] on the chord diagram ; cut the chord diagram along the chord labeled @xmath1 ( then @xmath1 vanishes ) , turn over the semicircle , and join the two semicircles again . on a chord diagram.,width=491 ] for example , @xmath4 at the crossing labeled @xmath9 on the spherical curve in fig .
[ chord - ex ] is represented on the chord diagram as shown in fig .
[ a - on - chord - ex ] . at 2 on a chord diagram.,width=302 ] a region of a spherical curve is _ coherent _ ( resp . _
incoherent _ ) if the boundary of the region is coherent ( resp .
incoherent ) ( see , for example , fig . [ bigons ] ) . a coherent bigon and an incoherent bigon are represented by chord diagrams as shown in fig .
[ chord - bigon ] . for coherent and incoherent bigons , we have the following lemmas : if a spherical curve @xmath0 has an incoherent bigon , then @xmath10 . by applying @xmath4 at one of the crossing on an incoherent bigon
, we obtain a reducible spherical curve as shown in fig .
[ c - non - c-2 ] .
[ non - c-2 ] if a spherical curve @xmath0 has a coherent bigon , then @xmath11 .
if @xmath0 is reducible , the reductivity is zero .
if @xmath0 is reduced , there is a crossing @xmath1 as shown in the chord diagram in fig .
[ c - c-2 ] . by applying @xmath4 at @xmath1
, we obtain a spherical curve which has an incoherent bigon .
hence @xmath0 has the reductivity two or less .
[ c-2 ] a trigon of a spherical curve is one of the types a , b , c and d in fig .
[ abcd ] with respect to the outer connections .
these trigons are represented in chord diagrams as shown in fig .
[ chord - trigon ] .
we have the following lemmas for trigons of type a and b. if a spherical curve @xmath0 has a trigon of type a , then @xmath11 . by applying @xmath4 at @xmath1 in fig . [ c-3a ]
, we have a spherical curve which has an incoherent bigon . [ 3a ]
if a spherical curve @xmath0 has a trigon of type b , then @xmath12 . by applying @xmath4 at @xmath1 in fig .
[ c-3b ] , we have a spherical curve which has a coherent bigon . [ 3b ] a _ connected sum _ @xmath13 of two spherical curves @xmath0 and @xmath6 is a spherical curve as depicted in fig .
[ conn ] . for trigons of type c , we have the following lemma : if a spherical curve @xmath0 has a trigon of type c , then @xmath12 . if the arcs @xmath14 , @xmath15 , and @xmath16 in fig .
[ lmn - c ] around a trigon of type c have no mutual crossings , then @xmath0 is a connected sum of a trefoil ( _ trefoil _ is the spherical curve depicted in the left - hand side of fig .
[ ex - reductivity ] ) and some ( trivial or non - trivial ) spherical curves as depicted in fig .
[ conn - trefoil ] , whose reductivity is one or zero . if @xmath0 has a mutual crossing @xmath1 of @xmath14 , @xmath15 and @xmath16 , we obtain a spherical curve which has a trigon of type a by applying @xmath4 at @xmath1 ( see fig .
[ c-3c ] ) .
[ 3c ] for type d , we have the following lemma : if a spherical curve @xmath0 has a trigon of type d , then @xmath17 .
if @xmath0 is reducible , the reductivity is zero .
if @xmath0 is reduced , @xmath0 has a crossing @xmath1 as shown in the chord diagram in fig .
apply @xmath4 at @xmath1 , and we obtain a trigon of type b. hence the reductivity of @xmath0 is four or less .
[ c-3d ] adams , shinjo and tanaka showed the following lemma @xcite : every reduced spherical curve has a bigon or trigon .
[ ast - lem ] they proved this lemma by considering the equality @xmath18 which is obtained by the euler characteristic of @xmath19 , where @xmath20 is the number of @xmath16-gons .
we prove theorem [ main ] .
_ proof of theorem [ main ] .
_ if @xmath0 is reducible , then @xmath21 . if @xmath0 is reduced , then @xmath17 because @xmath0 has a bigon or trigon .
in this section , we discuss the unavoidable sets of tangles for reduced spherical curves .
a _ tangle _ of a spherical curve is a part of the spherical curve . in this section
we do not distinguish the mirror image of tangles .
a set @xmath23 of tangles is an _
unavoidable set for spherical curves _ if any spherical curve has at least one tangle in @xmath23 .
for example , the set @xmath24 in fig .
[ s1 ] is an unavoidable set for _ reduced _ spherical curves because of lemma [ ast - lem ] .
is an unavoidable set.,width=226 ] we have the following proposition : the sets @xmath25 , @xmath26 and @xmath27 in fig .
[ s11 ] are unavoidable sets for reduced spherical curves . ,
@xmath26 and @xmath27 are unavoidable sets.,width=491 ] we use the discharging method which is used in graph coloring theory ( see , for example , @xcite ) .
we assume there exists a reduced spherical curve @xmath0 which has none of @xmath25 .
then @xmath0 has trigons by lemma [ ast - lem ] , and all the regions around each trigon are 5-gons or more .
we assign a charge of @xmath28 to each @xmath29-gon .
then each trigon receives 1 , 4-gon receives 0 , 5-gon receives -1 , @xmath30 .
the total charge @xmath31 on @xmath0 is @xmath32 where @xmath20 is the number of @xmath16-gons . by the equality @xmath33 for lemma [ ast - lem ] and that @xmath34 , we have @xmath35 .
now we do discharging ; at each trigon , we take off the charge 1 , and distribute it to the 6 regions around the trigon by @xmath36 ( see fig . [ discharging ] ) .
after the discharging , each trigon has the charge 0 because it has been taken the initial charge , and receive no charge from other trigons because they are not adjacent .
each 4-gon has also 0 because the initial charge was 0 and receive no charge from trigons .
next we consider 5-gons .
if a 5-gon @xmath37 has @xmath15 trigons around it , then @xmath37 has the charge @xmath38 after the discharging .
remark @xmath39 because if @xmath40 , some of the trigons around @xmath37 are adjacent .
hence each 5-gon has a negative charge after the discharging . for @xmath16-gons ( @xmath41 )
, they have @xmath42 after the discharging , where @xmath15 is the number of trigons around them .
we have @xmath43 because if @xmath44 , then @xmath45 for @xmath46 , and this means all the @xmath47 regions around the @xmath16-gon are trigons , and they are adjacent . therefore , after discharging , the total charge turns a negative number despite that the discharging preserves the total charge 8 .
hence @xmath25 is an unavoidable set .
it is similarly shown for @xmath26 and @xmath27 by considering the discharging @xmath48 and @xmath49 in fig .
[ discharging2 ] , respectively . and @xmath49.,width=491 ]
we have the following question : is the set @xmath50 in fig . [ s2 ] an unavoidable set for reduced spherical curves ? if the answer to this question is yes , then the answer to question [ main - q ] is also yes .
a set @xmath2 of tangles is an _
avoidable set _ for spherical curves if @xmath2 is not an unavoidable set , i.e. , there is a spherical curve which has no tangles in @xmath2 . since we have the spherical curve in fig .
[ counter ] , the sets @xmath51 and @xmath52 in fig .
[ t1 ] are avoidable sets for reduced spherical curves .
remark that @xmath52 is the set consisting of the tangles of the bigon and the _ incoherent _ trigons .
the author is deeply grateful to professors tetsuya abe , reiko shinjo and kokoro tanaka for valuable discussions and suggestions .
she is also grateful to the members of the friday seminar on knot theory at osaka city university and kobe topology seminar for helpful advice and encouragements .
she is supported by grant - in - aid for research activity start - up ( 24840030 ) .
99 c. c. adams , r. shinjo and k. tanaka , complementary regions of knot and link diagrams , _ ann .
_ * 15 * ( 2011 ) , 549563 .
j. a. calvo , knot enumeration through flypes and twisted splices , _ j. knot theory ramifications _ * 6 * ( 1997 ) , 785798 . m. goussarov , m. polyak and o. viro , finite type invariants of classical and virtual knots , _ topology _ * 39 * ( 2000 ) , 10451068
. n. ito and a. shimizu , the half - twisted splice operation on reduced knot projections , _
j. knot theory ramifications _ * 21 * , 1250112 ( 2012 ) [ 10 pages ] .
r. wilson , _ four colors suffice _ , princeton science library ( 2002 ) . | we show that we can obtain a reducible spherical curve from any non - trivial spherical curve by four or less inverse - half - twisted splices , i.e. , the reductivity , which represents how reduced a spherical curve is , is four or less . we also discuss unavoidable sets of tangles for spherical curves . |
until the mid-1970s , it was believed that black holes ( bhs ) are the super absorbent objects in the universe that nothing can come out from them . in 1974 , hawking @xcite , showed that a bh can thermally create and emit virtual particles until it snuffs out its energy .
it means that if a bh does not gain mass from any other source , eventually it could shrink itself out of existence , and evaporate completely : hawking radiation ( hr ) . on the other hand
, hr is an intriguing puzzle resulting from the amalgamation of general relativity and quantum mechanics .
as is the case with black - body radiation , hr quanta do not carry information .
thermal radiation is quantum mechanically characterized by the density matrix , not with the unitary operator of the quantum wave function . since mathematically there is no transformation between the unitary operator and the density matrix , the information encoded in the wave function
is irretrievably lost , which is unacceptable in quantum mechanics : the so - called information loss paradox @xcite .
the efforts for resolving this problem have been continuing even today .
a reader , who wants to learn the details and recent developments about the information loss paradox may refer to @xcite . since the original studies of bh emission @xcite , hr is studied and verified by various methods for many different types of particles having spin-@xmath0 ( @xmath1 etc . )
hr studies include the lower and higher dimensional bhs , wormholes , and black strings @xcite .
recent studies @xcite have claimed that hr has been observed in the laboratory environment .
those experiments about the hr were conducted by steinhauer .
he used a sonic ( or the so - called analogue ) bh in an atomic bose - einstein condensate @xcite so that a state , where gas bosons are cooled to @xmath2 kelvin .
thus , the sonic bh could mimic a real cosmic event horizon . at this stage , steinhauer managed to observe the particles of sound ( phonons ) at the bh s event horizon : sound waves in the bose - einstein condensate obey the hawking s theory @xcite .
we should also add that many physicists are still cautious about these results .
there is a need for other experiments to support the experiment of steinhauer . in particle physics ,
spin-1 particles are called vector particles .
the most known massive spin-1 particles , which are described by a 3 dimensional spin vector are the weak intermediate vector bosons @xmath3 and @xmath4 , and the mesons @xmath5 @xmath6/@xmath7 @xmath8 and @xmath9 @xcite .
photons are the massless spin-1 particles , which can only be directed parallel or anti - parallel to their direction of motion .
free massive spin-1 fields are governed by the proca field equation , and massless spin-1 fields by the maxwell field equation .
there is compelling evidence that the spin-1 fields are the potential dark matter candidates @xcite . in the recent years
, hr of the spin-1 particles have gained a great interest @xcite . in the present study
, we consider the rindler modified schwarzchild bh ( rmsbh ) @xcite , which was initially proposed by grumiller to explain the mysterious attractive constant radial force acting on the pioneer spacecrafts @xcite .
the force law obtained from the rindler s acceleration term has a considerable effect on the gravity at very long distances . on the other hand , turyshev et al .
@xcite perceptibly showed that the reason of the pioneer anomaly is the thermal heat loss of the satellites .
nevertheless , rmsbh is still on the agenda since it supplies a basis for the theoretical explanation of the following issues : explanation of the rotation curves of spiral galaxies , gravitational redshift , and perihelion shift in planetary orbits .
former studies of the rmsbh , which include the subjects of spectroscopy of area / entropy , quantum tunneling , and geodesics can be seen in @xcite . in the present study , we study the hr of the massive spin-1 particles tunneling from the rmsbh . to this end , we employ the proca equation @xcite . next , we apply the quantum tunneling method @xcite to the proca equation and obtain a set of differential equations . following @xcite , we set the determinant of coefficient matrix of the equation set to zero in order to have a non - trivial solution .
thus , we get the leading order term of the classical action ( @xmath10 ) of the vector particles outgoing / ingoing from the horizon .
we finally derive the tunneling rate of the spin-1 particles in the rmsbh and read the hawking temperature of the rmsbh .
the phenomenon of gravitational lensing , predicted by einstein s theory of general relativity , is a side effect of light moving along the curvature of spacetime , where the light passing nearby a massive object is deflected slightly toward the mass .
first time , in 1919 , it was observed by eddington and dyson during a solar eclipse @xcite . since then
, the gravitational lensing is one of the important tools to be used in astronomy and astrophysics . for more details and recent contributions about the gravitational lensing
, the reader may refer to @xcite .
we shall also study the gravitational lensing problem of the rmsbh .
for this purpose , we follow the geometrical method of gibbons and werner @xcite . by this way
, we explore the effect of the rindler acceleration on the deflection of light moving in the ir region of the rmsbh .
the paper is organized as follows : in sec .
ii we introduce the physical features of the rmsbh geometry .
section iii is devoted to the computation of the hr of the massive spin-1 particles from the rmsbh . in sec .
iv , we study the deflection of light from the rmsbh at the ir region via the method of gibbons and werner @xcite .
our results are summarized and discussed in sec . v.
grumiller @xcite constructed an effective model for gravity of a central object at large scales : the rmsbh geometry .
rindler term in the rmsbh spacetime causes to an anomalous acceleration in geodesics of test particles .
rmsbh is the solution to the generic effective theory of gravity described by the following action @xmath11 , \label{1}\ ] ] where @xmath12 is the determinant of the metric tensor matrix , @xmath13 denotes the scalar field , @xmath14 is the cosmological constant , @xmath15 is the ricci scalar , and @xmath16 stands for the rindler acceleration . after applying the variational principle to action ( 1 ) and solving the corresponding field equations , one can obtain the following spherically symmetric line element that models the ir gravity : @xmath17 where @xmath18 metric ( 2 ) is nothing but the rmsbh spacetime . in eq .
( 3 ) , the quantity @xmath19 is an integral constant .
when @xmath20 , one easily recovers the schwarzschild solution , with @xmath19 being the bh mass . moreover , if @xmath21 , line element ( 2 ) is the 2-dimensional rindler metric @xcite .
since the value of @xmath14 is approximated to @xmath22 @xcite , we shall set it to zero ( @xmath23 ) for simplicity .
besides , at the ir region , the value of the rindler acceleration is estimated as @xmath24 @xcite .
metric function ( 3 ) can be rewritten as @xmath25 in which @xmath26 it is clear that @xmath27 can not be interpreted as the horizon since it has a negative value .
so , rmsbh spacetime possesses an event horizon ( @xmath28 ) , which is given by @xmath29 the bekenstein - hawking entropy @xcite of the rmsbh is given by @xmath30 where @xmath31 is the surface area of the rmsbh . using the conventional definition of the surface gravity defined for a spherically static spacetime @xcite : @xmath32 where a prime on a function denotes differentiation with respect to @xmath33 . from here
on , we read the hawking temperature of the rmsbh as follows @xmath34 figure 1 shows how the hawking temperature changes with rindler acceleration in the rmsbh geometry . versus @xmath16 .
the plots are governed by eq .
each color represents a rmsbh with different mass . ]
in this section , we compute the hr of massive vector particles , which quantum mechanically tunnel from the rmsbh .
we consider the proca equation @xcite : @xmath35 which corresponds to the wave equation of a spin-1 field @xmath36 of mass @xmath37 in eq .
( 10 ) , the second rank tensor is defined by @xmath38 considering the wkb method , the ansatz of the spin-1 field can be given by @xmath39 where @xmath40 represents some arbitrary constants .
furthermore , taking cognizance of the hamilton - jacobi method which takes the advantage of the symmetry of the spacetime , one can set the leading order of the action to @xmath41 where @xmath42 and @xmath43 denote the energy and angular momentum of the vector particles , respectively .
@xmath44 is a complex constant . after manipulating the proca equation ( 10 ) with eqs .
( 12 ) and ( 13 ) , we get a quadruple equation set , to the lowest order in @xmath45 , as follows ( in the following equations , for ease of typing , every functional term is shown with a simple letter , e.g. @xmath46 ) : @xmath47 c_{4}}{r^{2}f\sin^{2}\theta } & = 0\\ -\frac{\left [ -\sin^{2}\theta f(\partial_{\theta}r)^{2}-f(m^{2}r^{2}\sin ^{2}\theta+j^{2})+e^{2}r^{2}\sin^{2}\theta\right ] c_{1}}{r^{2}\sin^{2}\theta } -\frac{fr^{\prime}(\partial_{\theta}r)c_{2}}{r^{2}}-\frac{jfr^{\prime}c_{3}}{r^{2}\sin\theta}-er^{\prime}c_{4 } & = 0\\ -\frac{fr^{\prime}(\partial_{\theta}r)c_{1}}{r^{2}}-\frac{\left [ -\sin ^{2}\theta f(r^{\prime})^{2}-f(m^{2}r^{2}\sin^{2}\theta+j^{2})+e^{2}r^{2}\sin^{2}\theta\right ] c_{2}}{r^{2}\sin^{2}\theta}-\frac{j(\partial_{\theta } r)c_{3}}{r^{4}\sin^{2}\theta}-\frac{(\partial_{\theta}r)ec_{4}}{r^{2}f } & = 0\\ -\frac{jfr^{\prime}c_{1}}{r^{2}\sin^{2}\theta}-\frac{j(\partial_{\theta } r)c_{2}}{r^{4}\sin^{2}\theta}+\frac{\left [ f(\partial_{\theta}r)^{2}-r^{2}(-f^{2}(r^{\prime})^{2}-m^{2}f+e)\right ] c_{3}}{r^{4}f\sin^{2}\theta } -\frac{jec_{4}}{r^{2}\sin^{2}\theta f } & = 0\end{aligned}\ ] ] after straightforward calculation one obtains a @xmath48 coefficient matrix @xmath49 of @xmath50 , in which the superscript @xmath51 represents the transition to the transposed vector
. the non - zero components of the matrix of @xmath49 are given by @xmath52 } { r^{2}f\sin^{2}\theta},\\ \aleph_{21 } & = -\frac{\left [ -\sin^{2}\theta f(\partial_{\theta}r)^{2}-f(m^{2}r^{2}\sin^{2}\theta+j^{2})+e^{2}r^{2}\sin^{2}\theta\right ] } { r^{2}\sin^{2}\theta},\\ \aleph_{22 } & = \aleph_{31}=-\frac{fr^{\prime}(\partial_{\theta}r)}{r^{2}},\end{aligned}\ ] ] @xmath53 } { r^{2}\sin ^{2}\theta},\\ \aleph_{33 } & = \aleph_{42}=-\frac{j(\partial_{\theta}r)}{r^{4}\sin^{2}\theta } , \\
\aleph_{43 } & = \frac{\left [ f(\partial_{\theta}r)^{2}-r^{2}(-f^{2}(r^{\prime})^{2}-m^{2}f+e)\right ] } { r^{4}f\sin^{2}\theta}.\end{aligned}\ ] ] the non - trivial solution of the quadruple equation set is conditional on @xmath54 .
thus , we have @xmath55 ^{3}}{r^{10}\sin\theta f^{3}}=0,\ ] ] which yields the integral solution of the radial function as @xmath56 @xmath57 and @xmath58 correspond to the outgoing ( i.e. emission ) and ingoing ( i.e. absorption ) vector particles from / to the rmsbh , respectively .
moreover , there exists a pole located at horizon , and the imaginary part of @xmath59 can be calculated by using the complex path integration method @xcite . therefore , one can evaluate the integral ( 29 ) in the vicinity of the horizon as @xmath60 the probabilities of the massive vector particles , tunneling from the horizon of rmsbh , become @xmath61 } , \label{32}\\ p_{absorption } & = e^{-\frac{2}{\hbar}ims_{-}}=e^{\left [ -\frac{2}{\hbar } ( imw_{-}+im\bbbk)\right ] } .
\label{27}\ ] ] according the classical concept of bh physics , the ingoing vector particles must be fully absorbed : @xmath62 .
fulfillment of this requirement is possible with @xmath63 .
recalling @xmath64 , we can thus compute the quantum tunneling rate of the massive vector particles of the rmsbh as @xmath65 @xmath66 is also equivalent to the boltzmann factor @xcite .
the latter remark enables us to compute the surface temperature of the rmsbh : @xmath67 the above result completely overlaps with the hawking temperature ( 9 ) of the rmsbh .
in the present section , we shall study the deflection of null geodesics from the rmsbh : the gravitational lensing problem . to this end
, we shall suppose that rmsbh consists of a perfect fluid , which can be thought of as the stellar fluid of a cluster of galaxies .
such stellar fluids act as a gravitational lens , which is capable of bending the light from the source as the light travels towards the observer . using the static and spherically symmetric feature of metric ( 2 ) , without loss of generality
, one can assume that null geodesics ( @xmath68 ) lie in the equatorial plane : @xmath69 .
this means that all images are collinear with the lens center @xcite .
light rays are the spatial projection of the null geodesics of the line element of the optical metric @xcite , which is given by @xmath70 where @xmath71 is originated from the fermat s
principle @xcite : @xmath72 thus , metric ( 35 ) becomes @xmath73 where @xmath74 is the radial regge - wheeler tortoise coordinate : @xmath75 and @xmath76 one immediately deduce from eq .
( 37 ) that the optical metric is a surface of revolution .
the intrinsic or the so - called gaussian curvature @xmath77 @xcite of the optical metric can be expressed as follows @xmath78 .
\label{40}\ ] ] the area element of the optical metric ( 35 ) is given by @xmath79 we now consider a lens model of perfect fluid , which is characterized by the following energy momentum tensor @xcite @xmath80 , \label{42}\ ] ] where @xmath81 and @xmath82 denote the pressure and density , respectively .
if we set @xmath83 and in sequel the einstein field equations are considered , we obtain the following integral solution : @xmath84 after comparing eqs .
( 3 ) and ( 43 ) , it can be easily seen that @xmath85 hence , tolman - oppenheimer - volkoff equation @xcite , which ensues from the conservation of the energy momentum tensor ( @xmath86 ) reads @xmath87 \left [ \mu(r)+4\pi r^{3}p(r)\right ] } { r^{2}f(r)}. \label{46}\ ] ] equation ( 46 ) , in fact , shows the hydrodynamical behavior of a massive compact astrophysical object in terms of the mass , pressure and energy density @xcite .
we also have @xmath88 .
\label{47}\ ] ] inserting eqs .
( 46 ) and ( 47 ) into eq .
( 40 ) , one can get the gaussian optical curvature as @xmath89 \right\ } .
\label{48}\ ] ] our main motivation is to study the weak gravitational lensing phenomenon , which is more realistic in the astrophysical observation .
for this reason , we shall focus on the non - relativistic kinetic theory ( e.g. the collisionless boltzmann equation @xcite of the non - relativistic stellar fluids ) .
this means that the pressure is ignored in eq .
so , the multiplication of the gaussian optical curvature with the area element of the optic metric results in @xmath90 \right\ } drd\phi . \label{49}\ ] ] suppose that the region where the observed gravitational lensing is too far from the bh ( i.e. , ir region ) .
furthermore , let us also assume that in the ir region the radial coordinate @xmath33 has almost a constant value ( @xmath91constant ) .
therefore , we can infer from eqs .
( 44 ) and ( 45 ) that @xmath92 whence eq .
( 49 ) reduces to @xmath93 since we are presently interested in the ir region , by using the taylor series , the above equation can be recast into the following form @xmath94 the deflection angle ( @xmath95 ) is calculated by integrating the gaussian curvature in the circular boundary of domain @xmath96 @xcite . in the weak deflection limit , it is assumed the light rays are governed by the function ( at zeroth order ) of @xmath97 where @xmath98 is known as the impact parameter with @xmath99 @xcite .
thus , the deflection angle is computed via the following integral @xcite @xmath100 the above result shows the rmsbh s gravitational lensing deflection angle at the ir region .
it is obvious that the deflection angle is inversely proportional with the rindler acceleration .
namely , gravitational lensing of an accelerated rmsbh is less than the almost non - accelerated rmsbh .
the latter remark implies that observing ( via the gravitational lensing ) a rmsbh will be more difficult than the observation of schwarzschild bh .
meanwhile , one should not overlook a point that since the impact parameter @xmath99 , in the denominator of eq .
( 54 ) @xmath101 would have a significant value .
this means that the deflection angle possesses a small value at the ir region .
in this paper , mainly two problems , hr of massive spin-1 particles and gravitational lensing of the rmsbh , have been thoroughly examined . in order to investigate the hr of the massive vector particles , we have employed the proca equation ( 10 ) in the rmsbh background within the concept of semiclassical wkb approximation . by taking into account the hamilton - jacobi and the complex path integration methods @xcite ,
we have computed the tunneling rate ( 28 ) of the rmsbh , which is ruled by the well - known boltzmann factor .
then , the temperature obtained from the resulting tunneling rate is shown to be same with the original hawking temperature ( 9 ) of the rmsbh .
the hawking temperature of the rmsbh proportionally increases with the rindler acceleration , which can be best seen from fig .
we have also applied the geometrical approach of the gravitational lensing theory to the rmsbh .
the optical metric ( 37 ) in which the geodesics are the spatial light rays of the rmsbh has been derived to calculate the gaussian curvature in the weak deflection limit . at the ir region
, it has been shown that deflection angle ( 54 ) decreases with the increasing rindler acceleration value .
namely , the gravitational lensing of the rmsbh is always less than the schwarzschild bh .
finally , it would be interesting to explore the gravitational lensing of a rotating rmsbh , which can be easily obtained by the newman - janis algorithm @xcite . in a rotating geometry ,
images are no longer collinear with the lens center @xcite
. therefore , such a problem may reveal more information compared to the present case .
this is the next stage of study that interests us . | we investigate the hawking radiation of massive spin-1 vector particles , which are coupled to vacuum fluctuations of a quantum field , from rindler modified schwarzschild black hole .
rindler acceleration is used to produce the post - general relativistic theory of gravity for the distant field of a point mass .
the gravitational lensing problem of the rindler modified schwarzschild black hole is also studied .
we compute the deflection angle for the ir region ( large distance limit as _ infrared _ ) by using the gaussian curvature of the optical metric of this back hole .
our investigations clarify how the rindler acceleration plays a role on the hawking radiation and gravitational lensing . |
the concentration of particles to high spatial densities promotes the formation of planetesimals , the super - kilometer scale building blocks of planets .
drag forces on pebbles and rocks in disks lead to spontaneous particle clumping @xcite .
the discovery of a linear streaming instability @xcite shows that clumping is a robust consequence of particles drifting in and gas flowing out in disks with some radial pressure support @xcite .
@xcite studied the non - linear saturation of the streaming instability , neglecting vertical gravity and self - gravity .
those simulations showed that groups of boulders accelerate the gas around them towards the keplerian velocity , reducing the radial drift locally and leading to temporary concentrations of boulders ( see also * ? ? ? * ) .
@xcite and @xcite recently performed three - dimensional ( 3d ) simulations of vertical shear instabilities in keplerian disks in the single fluid limit where particles and gas have exactly the same velocities .
these studies confirmed expectations that mid - plane turbulence develops when the richardson number @xmath0 .
while perfect coupling is a good approximation for small grains , it can not include vertical settling or in - plane streaming motions .
in this letter we present 3d simulations of the motion of gas and pebbles in sub - keplerian disks , including vertical gravity and particle sedimentation .
thus , we can study the combined effect of vertical shearing and streaming instabilities , as particles self - consistently settle towards and are stirred from the mid - plane .
we exclude external sources of turbulence , including magnetorotational instabilities ( which can actually promote clumping , see * ? ? ?
* ; * ? ? ?
our hydrodynamical simulations offer a first approximation to dead zones with low ionization @xcite where turbulent surface layers drive only weak motions in the mid - plane @xcite . in this non - magnetized limit ,
we investigate the clumping of smaller particles than considered in @xcite , which increases the likelihood of coagulation up to the initial sizes .
we find that clumping of pebbles in the mid - plane layer increases sharply above a threshold mass fraction of solids roughly consistent with solar metallicity .
thus planetesimal formation may help explain the high probability of finding giant exoplanets around stars rich in heavy elements @xcite .
we perform 3d hybrid simulations . they model gas on a fixed grid and solids with superparticles , each representing a swarm of actual particles .
we solve the standard shearing sheet dynamical equations for a frame rotating at the keplerian frequency @xmath1 at a fixed orbital distance @xmath2 from the star .
the axes are oriented such that @xmath3 points radially outwards , @xmath4 points in the orbital direction , while @xmath5 points vertically out of the disk .
the gas is subject to a global radial pressure gradient that reduces the gas orbital speed by @xmath6ms@xmath7 @xcite .
the sound speed @xmath8 , gas scale height @xmath9 and mid - plane gas density @xmath10 are the natural units of the simulation .
the motion of gas and particles are coupled through momentum - conserving drag forces with particle friction time @xmath11 .
our dynamical equations are identical to those of @xcite , with the addition of a vertical gravitational acceleration @xmath12 affecting both gas and particles .
the superparticles are evenly distributed in mass and number into four bins of normalized friction time @xmath13 .
these friction times are characteristic of compact solids with radius @xmath14 cm at @xmath15 in the minimum mass solar nebula @xcite .
rescaling to @xmath16 yields @xmath174 cm . we colloquially refer to this range of particle sizes as pebbles to contrast with larger @xmath18 boulders .
the total pebble mass is fixed by setting the pebble - to - gas column density ratio @xmath19 , where @xmath20 and @xmath21 are the mean particle and gas column densities , taking into account that most of the gas resides beyond the vertical extent of the simulation box .
this pebble abundance turns out to be the crucial parameter for triggering particle clumping .
the total abundance of condensable materials beyond the ice line was estimated by @xcite to be @xmath22 , while more up - to - date models give a somewhat lower value of @xmath23 at temperatures less than 41 k @xcite . for our models
a greater uncertainty is the efficiency of conversion from dust grains to pebbles . assuming that a majority ( @xmath24 ) of the condensable solids are bound in pebbles
, @xmath25 corresponds to solar metallicity .
we also experiment with higher values of @xmath26 and @xmath27 , which are motived both by stars with super - solar metallicities and by mechanisms that enrich the solids - to - gas ratio in disks ( see [ sec : conc ] ) .
a given pebble abundance would correspond to higher values of the metallicity if pebbles make up a smaller fraction of the condensable material .
we use a box size of @xmath28 and resolutions of @xmath29 zones with 125,000 particles , and @xmath30 zones with 1,000,000 particles .
this relatively small box size is chosen to capture typical wavelengths of streaming and kelvin - helmholtz instabilities .
the gas density is in vertical hydrostatic equilibrium .
particle positions are initialized to give a gaussian density distribution around the mid - plane with scale height @xmath31 , while gas and particle velocities are initially set to match the drag force equilibrium solution of @xcite .
since the disk is initially laminar , particles settle to the disk mid - plane . as particles collect in the mid - plane , they accelerate gas there towards the keplerian orbital speed .
this generates vertical shear that can drive kelvin - helmholtz instabilities .
the velocity difference between gas and solids also triggers streaming instabilities .
the resulting turbulence halts particle sedimentation and can lead to clumping . in fig .
[ f : sigmapmx_t ] we show results for pebble abundances @xmath25 ( roughly solar ) , @xmath26 ( super - solar ) and @xmath27 ( strongly super - solar ) .
the @xmath25 simulation shows minimal clumping , but reveals instead a surprising interaction between settling and stirring .
the particle layer is not centered uniformly on the mid - plane , as usually assumed .
instead the solids organize in a wave - like pattern ( top left panel of fig . [
f : sigmapmx_t ] ) produced by coherent vertical oscillations combined with radial drift .
we note that the streaming instability is known to drive coherent vertical motions in clumps of boulders @xcite .
the characteristic particle density @xmath32 is high enough to exert feedback on the gas .
however the average mid - plane density @xmath33 is diluted by the voids in the wave - like pattern , and is too small to trigger strong radial clumping . increasing the pebble abundance to @xmath26 and @xmath27 triggers strong overdensities in the particle layer ( see fig . [
f : sigmapmx_t ] ) .
initially , streaming instabilities produce many azimuthally extended clumps in the simulation box .
since the denser clumps drift inward more slowly , mergers result in a single dominant clump .
the characteristic particle density increases due to the strong clumping , with @xmath34 for @xmath35 , respectively . the maximum density in the box increases from less than ten to
more than 2,000 times the gas density when the pebble abundance is increased from @xmath25 to @xmath36 .
so why does the clumping increase so sharply for order unity changes to the pebble abundance @xmath37 ?
[ f : sigmapmx_t ] ( top panels ) shows that the vertical extent of the wave - shaped particle layer decreases with @xmath37 , because less kinetic energy is released in particle - dominated flows @xcite . consequently the average mid - plane particle density increases , with @xmath38 for @xmath39 .
strong clumping is thus expected for the higher metallicity cases , because the streaming instability is more powerful for linear growth and nonlinear clumping when the average @xmath40 @xcite .
a runaway develops as clumping limits stirring , giving higher average densities and in turn more clumping .
we can estimate the transitional value @xmath41 above which particles strongly clump .
assume that the sub - layer thickness is set by vertical shear turbulence to be @xmath42 as appropriate for small particles @xcite .
then the clumping threshold @xmath43 requires @xmath44 for @xmath45 this yields a critical pebble abundance of @xmath46 .
the slight overestimate of @xmath41 may arise because our pebbles settle more effectively than small grains .
the result that clumping happens for lower @xmath37 when @xmath47 is smaller ( weaker radial pressure support ) is seen in the simulations of @xcite .
the same metallicity threshold arises in particle layers with constant richardson number @xcite . above @xmath41 a high density cusp forms in the mid - plane because the shear turbulence can not lift solids with mass exceeding the local gas mass @xcite .
vertical self - gravity causes the analytic cusps to diverge , while streaming instabilities are responsible for the strong clumping in our dynamical simulations .
the growth of solids into planetesimals depends on collision speeds , which affect the rates of fragmentation and coagulation @xcite .
the threshold speed for catastrophic destruction of a meter - scale body , with the material strength of compact rock , is around 10ms@xmath7 @xcite .
porous aggregates , with less material strength , have an even lower threshold for destruction .
sticking of equal - sized bodies @xmath48 mm is not observed .
however , small grains can adhere at speeds less than @xmath49ms@xmath7 @xcite . from our simulation snapshots we measured relative speeds of all particle pairs within the same grid cell , giving @xmath50 unique collisions in a grid cell with @xmath51 particles .
all collisions are assumed to be head on ; for a random distribution of impact parameters the normal velocity would be reduced by 50% compared to the relative speed .
we averaged over snapshots taken between 30 and 50 orbits of the @xmath29 and @xmath30 runs with @xmath26 .
the distribution of collision speeds between particles of various sizes is shown in fig . [
f : vcoll ] .
we weight the occurrence of relative speeds by the speed itself to reflect that high speed collisions take place more frequently than low speed collisions .
the typical collision speed lies in the interval 0.55ms@xmath7 for a sound speed @xmath52ms@xmath7 .
[ f : vcoll ] also shows that the collision speed is systematically lower in the denser particle clumps where the typical speed falls to less than 2ms@xmath7 . averaging collision speeds without weighting by the speed itself leads to much lower typical speeds , around 0.22ms@xmath7 .
the low collision speeds are surprising given the sub - keplerian speed of 25ms@xmath7 .
the differential radial drift alone is 12.7ms@xmath7 between @xmath53 and @xmath54 test particles .
however , the large bodies reside in dense clumps in the mid - plane layer where the differential radial drift is much reduced and the turbulence is weakened by mass loading ( see insert in fig . [
f : vcoll ] ) . in simulations of particles in magnetorotational turbulence ,
collision speeds are measured to be much higher , of order @xmath5510ms@xmath7 ( and well beyond for different - sized particles , see * ? ? ?
* ; * ? ? ?
the more benign collision environment in the current simulations is more resistant to destruction and may even allow for growth by sticking .
low random velocities in high density clumps also favor gravitational collapse on a much shorter time - scale .
collective gravity can act to assemble solid particles into planetesimals @xcite .
the clumps in our @xmath26 and @xmath27 simulations reach densities beyond the roche density , @xmath56 so self - gravity will be dynamically relevant .
we activate the self - gravity in a @xmath26 simulation that already developed for @xmath55@xmath57 orbits , starting at a chosen time that displayed relatively weak clumping .
a few orbits later a dense filament forms due to the streaming instability , but the concurrent action of the self - gravity leads to the condensation of seven gravitationally bound objects ( see fig .
[ f : planetesimals ] ) .
initially the planetesimals actively accrete pebbles from their surroundings .
however , the accretion slows down after 5 orbits as 20% of the solid mass is locked in seven clumps with masses corresponding to solid bodies of radii between 100 and 200 km .
further growth can occur both by accreting dust and pebbles and by planetesimal - planetesimal impacts , since the modest gas density fluctuations in mid - plane layer turbulence can not pump collision speeds of large planetesimals into the erosive regime @xcite .
in this letter we present numerical evidence that the streaming instability , which arises in the coupled motion of gas and solids , can alone provide the necessary ingredients for the formation of planetesimals . with a high enough abundance of pebble - sized objects
, the streaming instability drives the formation of high - density clumps in the sedimented mid - plane layer .
collision speeds inside those clumps are relatively low , around a few meters per second .
the clumps get dense enough to contract gravitationally on a short time - scale into gravitationally bound clusters containing mass equivalent to solid planetesimals with radii of order 100 km .
our simulations display a threshold dependence of clumping on the global value of the pebble abundance . at a pebble abundance of @xmath25 , corresponding roughly to solar metallicity , powerful turbulence develops and puffs up the mid - plane layer to such widths that clumping is strongly suppressed .
this environment is not conducive to planetesimal growth , with low particle densities and high collision speeds .
however , a doubling of the pebble abundance to @xmath26 leads to strong clumping with local peak particle densities exceeding a thousand times the gas density .
this threshold behavior arises because the streaming instability triggers strong clumping when the particle density exceeds the gas density in the mid - plane .
if the fraction of condensables contained in pebbles is less than we assumed in this discussion ( approximately @xmath58 ) , then the threshold we describe corresponds to higher values of the bulk metallicity .
coagulation - fragmentation models @xcite and observations of pebbles in protoplanetary discs @xcite will be crucial for putting tighter constraints on the pebble abundance and the location of the metallicity switch .
the pebble abundance may increase as solid particles drift radially inward , converging in the inner few dozen aus of the protoplanetary disk @xcite .
a higher @xmath37 also will occur if the gas in the disk preferentially evaporates .
for example , photoevaporation removes gas from the disk surface , leaving behind larger solids that have sedimented to the mid - plane @xcite .
this opens a window for late planet formation with planetesimals forming as the gas dissipates .
the planets resulting from planetesimals formed so late will not have time enough to accrete a large gaseous envelope before the gas dissipates and would also suffer less severe type i migration @xcite .
this could explain why super - earths and exo - neptunes seem able to form equally well around stars of low and high metallicity @xcite .
the correlation between host star metallicity and giant exoplanets can be partly attributed to the efficiency of forming several - earth - mass cores from planetesimals @xcite .
our findings show that high metallicity already facilitates planetesimal formation , giving further support to the core accretion model for giant planet formation @xcite .
computer simulations were performed at the huygens cluster in amsterdam , funded by the nwo , and at the pia cluster of the max planck institute for astronomy .
ay and aj are grateful to chris thompson for stimulating discussions at c.i.t.a .
we thank the referee for careful reading of the manuscript .
m - mml was partially supported by the nasa origins of solar systems program under grant nnx 07-ai74 g , and by the nsf cyberenabled discovery initiative under grant ast08 - 35734 .
alexander , r. d. , & armitage , p. j. 2007 , , 375 , 500 allende prieto , c. , lambert , d. l. , & asplund , m. 2001 , , 556 , l63 balsara , d. s. , tilley , d. a. , rettig , t. , & brittain , s. a. , 2009 , , 397 , 24 barranco , j. a. 2009 , , 691 , 907 benz , w. 2000 , space science reviews , 92 , 279 blum , j. , & wurm , g. 2000 , icarus , 143 , 138 blum , j. , & wurm , g. 2008 , , 46 , brauer , f. , dullemond , c. p. , & henning , t. 2008 , , 480 , 859 carballido , a. , stone , j. m. , & turner , n. j. 2008 , , 386 , 145 chiang , e. i. , & goldreich , p. 1997
, , 490 , 368 chiang , e. 2008 , , 675 , 1549 cuzzi , j. n. , hogan , r. c. , & shariff , k. 2008 , , 687 , 1432 fischer , d. a. , & valenti , j. 2005 , , 622 , 1102 fleming , t. & stone , j. m. 2003 , , 585 , 908 goldreich , p. , & ward , w. r. 1973 , , 183 , 1051 gonzalez , g. 1997 , , 285 , 403 goodman , j. , & pindor , b. 2000 , icarus , 148 , 537 hayashi , c. 1981 , prog .
phys . , 70 , 35 ida , s. , & lin , d. n. c. 2004 , , 616 , 567 ida , s. , guillot , t. , & morbidelli , a. 2008 , , 686 , 1292 johansen , a. , henning , t. , & klahr , h. 2006 , , 643 , 1219 johansen , a. , & youdin , a. 2007 , , 662 , 627 johansen , a. , oishi , j. s. , mac low , m .-
m . , klahr , h. , henning , t. , & youdin , a. 2007 , , 448 , 1022 johnson , j. a. , & apps , k. 2009 , , 699 , 933 lodders , k. 2003 , , 591 , 1220 mordasini , c. , alibert , y. , benz , w. , & naef , d. 2009 , , 501 , 1161 nakagawa , y. , sekiya , m. , & hayashi , c. 1986 , icarus , 67 , 375 oishi , j. s. , mac low , m .- m . , & menou , k. 2007 , , 670 , 805 pollack , j. b. , hubickyj , o. , bodenheimer , p. , lissauer , j. j. , podolak , m. , & greenzweig , y. 1996 , icarus , 124 , 62 safronov , v. s. 1969 , evoliutsiia doplanetnogo oblaka .
( english transl . :
evolution of the protoplanetary cloud and formation of earth and the planets , nasa tech .
f-677 , jerusalem : israel sci . transl .
1972 ) sano , t. , miyama , s. m. , umebayashi , t. , & nakano , t. 2000 , , 543 , 486 santos , n. c. , israelian , g. , & mayor , m. 2001 , , 373 , 1019 sekiya , m. 1998 , icarus , 133 , 298 stepinski , t. f. , & valageas , p. 1996
, , 309 , 301 throop , h. b. , & bally , j. 2005 , , 623 , l149 udry , s. , mayor , m. , benz , w. , bertaux , j .- l . , bouchy , f. , lovis , c. , mordasini , c. , pepe , f. , queloz , d. , & sivan , j .-
2006 , , 447 , 361 weidenschilling , s. j. 1977a , , 180 , 57 weidenschilling , s. j. 1977b , , 51 , 153 wilner , d. j. , dalessio , p. , calvet , n. , claussen , m. j. , & hartmann , l. 2005 , , 626 , l109 yang , c .- c . ,
mac low , m .-
m . , & menou , k. 2009 , , submitted ( arxiv:0907.1897 ) youdin , a. n. , & shu , f. h. 2002 , , 580 , 494 youdin , a. n. , & chiang , e. i. 2004 , , 601 , 1109 youdin , a. n. , & goodman , j. 2005 , , 620 , 459 youdin , a. , & johansen , a. 2007 , , 662 , 613 | we present three - dimensional numerical simulations of particle clumping and planetesimal formation in protoplanetary disks with varying amounts of solid material .
as centimeter - size pebbles settle to the mid - plane , turbulence develops through vertical shearing and streaming instabilities .
we find that when the pebble - to - gas column density ratio is 0.01 , corresponding roughly to solar metallicity , clumping is weak , so the pebble density rarely exceeds the gas density .
doubling the column density ratio leads to a dramatic increase in clumping , with characteristic particle densities more than ten times the gas density and maximum densities reaching several thousand times the gas density .
this is consistent with unstratified simulations of the streaming instability that show strong clumping in particle dominated flows .
the clumps readily contract gravitationally into interacting planetesimals of order 100 km in radius .
our results suggest that the correlation between host star metallicity and exoplanets may reflect the early stages of planet formation .
we further speculate that initially low metallicity disks can be particle enriched during the gas dispersal phase , leading to a late burst of planetesimal formation . |
if the solar galactic environment is to have a discernible effect on events on the surface of the earth , it must be through a subtle and indirect influence on the terrestrial climate .
the scientific and philosophical literature of the 18th , 19th and 20th centuries all include discussions of possible cosmic influences on the terrestrial climate , including the effect of cometary impacts on earth ( @xcite ) , and the diminished solar radiation from sunspots , which herschel attributed to `` holes '' in the luminous fluid on the surface of the sun ( @xcite ) .
the discovery of interstellar material in the 20th century led to speculations that encounters with dense clouds initiated the ice ages ( @xcite ) , and many papers appeared that explored the implications of such encounters , including the influence of interstellar material ( ism ) on the interplanetary medium and planetary atmospheres ( e.g. @xcite ) .
the ism - modulated heliosphere was also believed to affect climate stability and astrospheres ( e. g. @xcite ) .
recent advances in our understanding of the solar wind and heliosphere ( e. g. @xcite ) justify a new look at this age - old issue .
this book addresses the underlying question : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ how does the heliospheric interaction _ _ with the interstellar medium affect the heliosphere , interplanetary medium , and earth ? _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the heliosphere is the cavity in the interstellar medium created by the dynamic ram pressure of the radially expanding solar wind , a halo of plasma around the sun and planets , dancing like a candle in the wind and regulating the flux of cosmic rays and interstellar material at the earth .
neutral interstellar gas and large interstellar dust grains penetrate the heliosphere , but the solar wind acts as a buffer between the earth and most other interstellar material and low energy galactic cosmic rays ( gcr ) .
together the solar wind and interstellar medium determine the properties of the heliosphere . in the present epoch
the densities of the solar wind and interstellar neutrals are approximately equal outside of the jupiter orbit .
solar activity levels drive the heliosphere from within , and the physical properties of the surrounding interstellar cloud constrain the heliosphere from without , so that the boundary conditions of the heliosphere are set by interstellar material .
figure [ fig:1 ] shows the sun and heliosphere in the setting of the milky way galaxy .
the answer to the question posed above lies in an interdisciplinary study of the coupling between the interstellar medium and the solar wind , and the effects that ism variations have on the 1 au environment of the earth through this coupling .
the articles in this book explore different viewpoints , including _
experiments , as well as data - rich summaries of variations in the solar environment and paleoclimate data on cosmic ray flux variations at earth .
the book begins with the development of theoretical models of the heliosphere that demonstrate the sensitivity of the heliosphere to the variations in boundary conditions caused by the passage of the sun through interstellar clouds ( @xcite ) .
a series of _ gedanken _ experiments then yield the response of planetary magnetospheres to encounters with denser ism ( @xcite )
. variations in the galactic environment of the sun , caused by the motions of the sun and clouds through the galaxy , are shown to occur for both long and short timescales ( @xcite ) .
the heliosphere acts as a buffer between the earth and interstellar medium , so that dust and particle populations inside of the heliosphere , which have an interstellar origin , vary as the sun traverses interstellar clouds .
these buffering mechanisms determine the interplanetary medium13.6 ev .
the high energy galactic cosmic ray population incident on the heliosphere is also modulated . ] .
the properties of these buffering interactions are evaluated for heliosphere models that have been developed using boundary conditions appropriate for when the sun traverses different types of interstellar clouds ( @xcite ) .
the consequences of sun - cloud encounters are then discussed in terms of the accretion of ism onto the terrestrial atmosphere for dense cloud encounters , and the possibly extreme variations expected for cosmic ray modulation when interstellar densities vary substantially ( @xcite ) .
radioisotope records on earth extending backwards in time for over @xmath00.5 myrs , together with paleoclimate data , suggest that cosmic ray fluxes are related to climate .
the galactic environment of the sun must have left an imprint on the geological record through variations in the concentrations of radioactive isotopes ( @xcite ) .
the selection of topics in this book is based partly on scientific areas that have already been discussed in the literature .
the authors who were invited to contribute chapters have previously studied the heliosphere or terrestrial response to variable ism conditions or cosmic rays .
figure [ fig:1 ] shows the heliosphere in our setting of the milky way galaxy . a postscript at the end of this chapter lists basic useful information .
i introduce the term `` paleoheliosphere '' to represent the heliosphere in the past , when the boundary conditions set by the local interstellar material ( lism ) may have differed substantially from the boundary conditions for the present - day heliosphere .
the `` paleolism '' is the local ism that once surrounded the heliosphere .
the solar wind drives the heliosphere from the inside , with the properties of the solar wind varying with ecliptic latitude and the phase of the 11-year solar activity cycle .
the global heliosphere is the volume of space occupied by the supersonic and subsonic solar wind .
interstellar material forms the boundary conditions of the heliosphere , and the windward side of the heliosphere , or the `` upwind direction '' , is defined by the interstellar velocity vector with respect to the sun .
the leeward side of the heliosphere is the `` downwind direction '' .
figure [ fig:1 ] shows a cartoon of the present - day heliosphere , with labels for the major landmarks such as the termination shock , heliopause , and bow shock . in the present - day heliosphere ,
the transition from solar wind to interstellar plasma occurs at a contact discontinuity known as the `` heliopause '' , which is formed where the total solar wind and interstellar pressures equilibrate ( @xcite ) . for a non - zero interstellar cloud velocity in the solar rest frame ,
the solar wind turns around at the heliopause and flows around the flanks of the heliosphere and into the downwind heliotail . before reaching the heliopause
, the supersonic solar wind slows to subsonic velocities at the `` termination shock '' , where kinetic energy is converted to thermal energy .
the subsonic solar wind region between the termination shock and heliopause is called the inner `` heliosheath '' .
the outer heliosheath lies just beyond the heliopause , where the pristine ism is distorted by the ram pressure of the heliosphere . a bow shock , where the interstellar gas becomes subsonic , is expected to form ahead of the present - day heliosphere in the observed upwind direction of the ism flow through the solar system . large interstellar dust grains and interstellar atoms that remain neutral inside of the orbit of earth , such as he , are gravitationally focused in the downwind direction .
this `` focusing cone '' is traversed by the earth every year in early december , and extends many au from the sun in the leeward direction ( e.g. @xcite ) .
the heliotail itself extends @xmath1 au from the sun in the downwind direction , forming a cosmic wake for the solar system . of significance
when considering the interaction of the heliosphere with an interstellar cloud is that neutral particles enter the heliosphere relatively unimpeded , after which they are ionized and convected outwards with the solar wind .
ions and small charged dust grains are magnetically deflected in the heliosheath around the flanks of the heliosphere ( see figure [ fig:1 ] ) .
space and astronomical data now confirm the basic milestones of the outer heliosphere .
voyager 1 crossed the termination shock at 94 au on 16 december , 2004 ( ut ) , and observed the signature of the termination shock on low - energy particle populations , the solar wind magnetic field , low - energy electrons and protons , and langmuir radio emission ( @xcite ) .
the present - day termination shock appears to be weak , with a solar wind velocity jump ratio ( the ratio of upstream to downstream values ) of @xmath02.6 and a magnetic field compression ratio of @xmath03 .
the magnetic wall that is predicted for the heliosphere ( @xcite , chapter 3 by pogorelov and zank ) appears to have been detected through observations of magnetically aligned dust grains ( @xcite ) , and the offset between upwind directions of interstellar and ( @xcite ) .
the compressed and heated in the hydrogen wall region of the outer heliosheath has now been detected around a number of stars ( @xcite ) .
the present - day solar wind is the baseline for evaluating the heliosphere response to ism variations in the following articles , so a short review of the solar wind is first presented .
the remaining part of @xmath21.2 introduces the topics in the following articles in terms of the underlying query of the book .
the solar wind originates in the million degree solar corona that expands radially outwards , with a density @xmath3 where @xmath4 is the distance to the sun , and contains both features that corotate with the sun , and transient structures ( e. g. @xcite ) .
the properties of the solar wind vary with the phase of the solar magnetic activity cycle and with ecliptic latitude .
the best historical indicator of solar magnetic activity levels is the number of sunspots , first detected by galileo in 1610 , which are magnetic storms in the convective zone of the sun .
sunspot numbers indicate that the magnetic activity levels fluctuate with a @xmath011 year cycle , or the `` solar cycle '' , and solar maximum / minimum corresponds to the maximum / minimum of sunspot numbers .
the magnetic polarity of the sun varies with a @xmath022 year cycle . during solar maximum , a low - speed wind , with velocity @xmath0300600 and density @xmath5 particles at 1 au
, extends over most of the solar disk .
open magnetic field lines700 , solar wind .
] are limited to solar pole regions .
a neutral current sheet @xmath00.4 au thick forms between the solar wind containing negative magnetic polarity fields and the solar wind that contains positive magnetic polarity fields .
the neutral current sheet reaches its largest inclination ( @xmath6 ) during solar maximum . during the conditions of solar minimum , a high speed wind with velocity @xmath0600800 and density @xmath05
is accelerated in the open magnetic flux lines in coronal holes . during mininum ,
the high speed wind and open field lines extend from the polar regions down to latitudes of @xmath7@xmath8(@xcite ) .
the higher solar wind momentum flux associated with solar minimum conditions produces an upwind termination shock that is @xmath0540 au more distant in the upwind direction than during solar maximum conditions ( e.g. @xcite ) .
the alignment and strength of the solar magnetic multipoles depend on the phase of the solar cycle ( @xcite ) . during solar minimum conditions ,
the magnetic field is dominated by the dipole and hexapole moments , and the dipole moment is generally aligned with the solar rotation axis .
sunspots migrate from high to low heliographic latitudes .
the magnetic poles follow the coronal holes to the solar equator as solar activity increases . during the solar maximum period ,
the galactic cosmic rays undergo their maximum modulation , the dipole component of the magnetic field is minimized , the quadrapole moment dominates , and the polarity of the solar magnetic field reverses ( @xcite , figure [ fig:2 ] ) .
over historic times , the cosmic ray modulation by the heliosphere correlates better with the open magnetic flux line coverage than with sunspot numbers ( @xcite ) .
variable cosmic ray modulation produced by a variable heliosphere may be a primary factor in both solar and ism forcing of the terrestrial climate .
the heliosphere modulation of cosmic rays is well established .
john simpson , to whom this book is dedicated , initiated a program 5 solar cycles ago in 1951 to monitor cosmic ray fluxes on earth using high - altitude neutron detectors ( @xcite ) .
the results show a pronounced anticorrelation between cosmic ray flux levels and solar sunspot numbers , which trace the 11-year schwabe magnetic activity cycle , and which also show that the polarity of the solar magnetic field affects cosmic ray modulation ( see figure [ fig:2 ] ) .
the articles in this book show convincingly that the ism also modulates the heliosphere , and the effect of the solar wind on the heliosphere must be differentiated from the influence of interstellar matter .
variations in solar activity levels are also seen over @xmath9 year timescales , such as the absence of sunspots during the maunder minimum in the 17th century .
modern climate records show that the maunder minimum corresponded to extremely cold weather , and radioisotope records show that the flux of cosmic rays was unusually high at this time ( see kirkby and carslaw , chapter 12 ) .
similar effects will occur from the modulation of galactic cosmic rays by the passage of the sun through an interstellar cloud .
these temporal and latitudinal variations in the solar wind momentum flux produce an asymmetric heliosphere , which varies with time .
any possible historical signature of the ism on the heliosphere must first be distinguished from variations driven by the solar wind itself .
the ism forms the boundary conditions of the heliosphere , so that encounters with interstellar clouds will affect the global heliosphere , the interplanetary medium , and the inner heliosphere region where the earth is located .
today an interstellar wind passes through the solar system at 26.3 ( @xcite ) .
an entering parcel of ism takes about 20 years to reach the inner heliosphere , so that ism near the earth is constantly replenished with new inflowing material .
this warm gas is low density and partially ionized , with temperature @xmath106,300 k , and densities of neutral and ionized matter of @xmath11 , and @xmath12 .
an elementary perspective of the response of the heliosphere to interstellar pressures is given by an analytical expression for the heliopause distance based on the locus of positions where the solar wind ram pressure , @xmath13 , and the total interstellar pressure equilibrate ( @xcite ) .
the solar wind density @xmath14 falls off as @xmath15 , where @xmath16 is the distance to the sun , while the velocity @xmath17 is relatively constant . at 1
au the solar wind ram pressure is @xmath18 so the heliosphere distance , @xmath19 , is given by : @xmath20 the interstellar pressure terms include the magnetic pressure @xmath21 , the thermal , @xmath22 , and the ram , @xmath23 , pressures of the charged gas , and the pressures of dust grains , @xmath24 , and cosmic rays , @xmath25 , which are excluded by heliosphere magnetic fields and plasma .
some interstellar neutrals convert to ions through charge exchange with compressed interstellar proton gas in heliosheath regions , adding to the confining pressure .
an important response characteristic is that , for many clouds , the encounter will be ram - pressure dominated , where @xmath26 for interstellar cloud mass density @xmath27 and relative sun - cloud velocity @xmath17 , so that variations in the cloud velocity perturb the heliosphere even if the thermal pressures remain constant .
the multifluid , magnetohydrodynamic ( mhd ) , hydrodynamic and hybrid approaches used in the following chapters provide much more substantial models for the heliosphere , and include the coupling between neutrals and plasma , and field - particle interactions .
these sophisticated models predict variations in the global heliosphere in the face of changing interstellar boundary conditions , and for a range of different cloud types .
although impossible to model a solar encounter with every type of interstellar cloud , the following articles include discussions of many of the extremes of the interstellar parameter space , including low density gas with a range of velocities , very tenuous plasma , high velocity clouds , dense ism , and magnetized material for a range of field orientations and strengths .
the discussions in these chapters extrapolate from our best theoretical understanding of the heliosphere boundary conditions today to values that differ , in some cases dramatically , from the boundary conditions that prevailed at the beginning of the third millennium in the gregorian calendar .
the sun has been , and will be , subjected to many different physical environments over its lifetime .
theoretical heliosphere models yield the properties of the solar wind - ism interaction for these different environments , which in turn determine the nature and properties of interstellar populations inside of the heliosphere for a range of galactic environments .
these models form the foundation for understanding the significance of our galactic environment for the earth .
the interstellar parameter space is explored by zank et al .
( chapter 2 ) , where 28 sets of boundary conditions are evaluated with computationally efficient multifluid models .
moebius et al .
( chapter 8) , fahr et al .
( chapter 9 ) , florinski and zank ( chapter 10 ) , and yeghikyan and fahr ( chapter 11 ) also develop heliosphere models for a range of interstellar conditions .
together these models evaluate the heliosphere response to interstellar density , temperature , and velocity variations of factors of @xmath28 , @xmath29 , and @xmath30 , respectively .
the interstellar magnetic field introduces an asymmetric pressure on the heliosphere , affecting the heliosphere current sheet and cosmic ray modulation .
pogorelov and zank ( chapter 3 ) use mhd models to probe the heliosphere response to the interstellar magnetic field , including charge exchange between the neutrals and solar wind .
the resulting asymmetry provides a test of the magnetic field direction , and shows strong differences between cases where the interstellar flow is parallel , instead of perpendicular , to the interstellar magnetic field direction .
since the random component of the interstellar magnetic field is stronger , on the average , than the ordered component , particularly in spiral arm regions where active star formation occurs , a range of interstellar magnetic field strengths and orientations are expected over the solar lifetime ( shaviv , chapter 5 , and frisch and slavin , chapter 6 ) .
the earth s magnetosphere acts as a buffer between the solar wind and atmosphere , and as such is an ingredient in understanding the effect of our galactic environment on the earth .
the decreasing solar wind density in the outer heliosphere results in an interplanetary medium around outer planets that is more sensitive to ism variations than for inner planets , with implications for the magnetospheres of jupiter , neptune , and uranus .
most topics in this book are already considered in the scientific literature , but questions about magnetosphere variations from an ism - modulated heliosphere have received scant attention . in a quintessential _ gedanken _
experiment , parker explores the interaction between magnetospheres and the solar wind for variations in the interstellar density , and for inner versus outer planets ( chapter 4 ) .
there is every reason to expect that the galactic environment of the sun varies over geological timescales .
the sun moves through space at a velocity of 1320 , and interstellar clouds have velocities ranging up to hundreds of .
the arecibo millennium survey showed that @xmath025% of the mass contained in interstellar , including both warm and cold ism , is in clouds traveling with velocities @xmath3110 through the local standard of rest ( @xcite ) .
thus sun - cloud encounters with relative velocities exceeding 25 are quite likely , and for a typical cloud length of @xmath01 pc the cloud transit time would be @xmath040,000 years .
the many types of ism traversed by the sun during the past several million years have affected the heliosphere , the inner solar system , and the flux of anomalous and galactic cosmic rays at earth ( @xcite ) . for the past @xmath03
myrs the sun has been in a nearly empty region of space , the `` local bubble '' , with very low densities of @xmath3210@xmath33 gr . within the past 44,000140,000 years
the sun entered a flow of tenuous , partly neutral ism , nick - named the `` local fluff '' , with density @xmath060 times higher ( chapter 6 ) .
this transition was accompanied by the appearance of interstellar dust and neutrals in the heliosphere , along with the pickup ion and anomalous cosmic ray populations .
galactic cosmic ray modulation was affected , providing a possible link between our galactic environment and climate .
intriguingly , the averaged cosmic ray flux at earth , as traced by records , was lower in the past @xmath0135 kyrs than for earlier times ( chapter 12 ) .
was the decrease in the galactic cosmic ray flux @xmath0135 kyrs ago caused by an increase in modulation as the sun entered the local fluff ?
the galactic environment of the sun also varies quite dramatically over long time scales , as discussed by shaviv ( chapter 5 ) . over its 4.5 billion year lifetime ,
the sun traverses spiral arm and interarm regions , with atomic densities varying from less than @xmath34 g to over @xmath35 g , and temperatures ranging over 7 orders of magnitude , 1010@xmath36 k. the sun is now in low density space between the perseus and sagittarius spiral arms , and on the inner edge of what is known as the orion spur on the local arm .
the local arm is not shown in figure [ fig:1 ] , as is consistent with the usual galaxy depictions .
the local arm does not appear to be a grand design spiral shock ( bochkarev , 1984 ) .
the sun has a systematic motion of 1320 km s@xmath37 with respect to the nearest stars , corresponding to @xmath034 au per year .
the local interstellar cloud ( lic ) now surrounding the sun traverses the heliosphere at @xmath05.5 au per year .
the sun oscillates vertically through the galactic plane once every @xmath034 myrs , and orbits the center of the milky way galaxy once per @xmath0220 myrs .
shaviv evaluates variations in the galactic environment of the sun over long timescales .
this bold discussion compares various geologic records of cosmic ray flux variations , based on radioisotope data that sample timescales of @xmath38 years , with models of the milky way galaxy spiral arm pattern to reconstruct the timing of the sun s passage through spiral arms .
the chapter concludes that star formation in spiral arms leaves a signature on the radioisotope records of the solar system .
frisch and slavin ( chapter 6 ) reconstruct short - term variations of the galactic environment of the sun using observations of interstellar matter towards nearby stars and inside of the solar system .
radiative transfer models of the lic show that ionization varies across this low density cloud , so that the heliosphere boundary conditions vary from radiative transfer considerations alone as the sun traverses the lic .
cloud transitions are predicted during the past @xmath03 myrs , including the departure of the sun from the local bubble interior 44,000140,000 years ago , and entry into the surrounding cloud 100040,000 years ago .
the particle populations formed by the interactions between the solar wind and interstellar dust , gas , and cosmic rays are emissaries between the cosmos and inner heliosphere , varying as the sun moves through clouds . about @xmath01% of the mass of the cloud surrounding
the sun is contained in interstellar dust grains .
the largest of these charged grains , mass @xmath39 g , have large magnetic larmor radii of @xmath40500 au at the heliopause for an interstellar field of @xmath01.5 @xmath41 g , and flow into the solar system .
the earth passes through the gravitational focusing cone formed by these grains early each december .
the smallest charged grains , mass@xmath42 g and radii@xmath43 m , have larmor radii of @xmath020 au , depending on the magnetic field strength and radiation field , and are deflected around the heliosheath ( @xcite ) .
interstellar dust grains are measured in the inner heliosphere within @xmath05 au of the sun , and over the solar poles , by satellites such as ulysses , galileo and cassini .
landgraf ( chapter 7 ) reviews the properties of the interaction between interstellar dust and the solar wind , and speculates on the changes that might be expected from an encounter with a dense interstellar cloud .
should it some day be possible to compare the ratio of large to small interstellar dust grains on the surfaces of the inner versus outer planets , it would become possible to disentangle cloud encounters from solar activity effects . at the very large end of the dust population mass spectrum we find interstellar micrometeorites , with masses @xmath03 @xmath44 10@xmath45 g , open orbits , and inflow velocities greater than the 42 escape velocity from the solar system at 1 au . these interstellar objects , detected by radar as they impact the atmosphere , evidently originate in circumstellar disks such as that around @xmath46 pictoris , and in the interior of the local bubble ( @xcite ) .
these objects do not collisionally couple to the interstellar gas ( @xcite ) , and should not vary with the type of ism surrounding the sun .
presently , low energy interstellar neutrals , high energy galactic cosmic rays , and interstellar dust all enter the heliosphere .
the characteristics of each of these populations and their secondary products are modified as the sun transits the ism , or the cloud ionization changes .
the first ionization potential ( fip ) of is 13.6 ev .
neutral interstellar atoms with [email protected] ev are ionized in nearly all interstellar clouds because the main source of interstellar opacity is .
interstellar ions are deflected around the heliosheath , so the result is that only interstellar atoms with [email protected] ev enter the heliosphere where they are then destroyed , primarily by charge exchange with solar wind ions .
the density of interstellar neutrals in the inner heliosphere depends on the density and ionization of the surrounding cloud , the ionization ( or `` filtration '' ) of those neutrals by the heliosheath , and the subsequent interactions with the solar wind inside of the heliosphere .
secondary products produced by solar wind interactions with interstellar neutrals inside of the heliosphere include pickup ions,1 gev , in the solar wind and at the termination shock , and then subjected to the same modulation and propagation processes as galactic cosmic rays ( @xcite ) . ] energetic neutral atoms , the gravitational focusing cone formed by helium ( also seen in dust ) , and the anomalous cosmic ray population with energies @xmath321 gev .
interstellar neutrals inside of the heliosphere , and the heliosphere itself , form a coupled system that together respond to variations in the heliosphere boundary conditions .
moebius et al .
( chapter 8) model the heliosphere for several different conditions , and then probe the response of the inner heliosphere to the density of interstellar neutrals flowing into this ism - modified heliosphere . at 1 au , the neutral densities ,
particle populations derived from interstellar neutrals , and characteristics of the helium focusing cone all respond to variations in the interstellar boundary conditions . for some cases ,
increased neutral fluxes fall on the atmosphere of earth ( also see yeghikyan and fahr , chapter 11 ) .
the velocity structure of the ism appears to vary on subparsec scale lengths ( frisch and slavin , chapter 6 ) , and these variations may in some cases result in significant modifications of the inner heliosphere , particularly the gravitational focusing cone , when all other interstellar parameters such as thermal pressure are invariant ( zank et al . , chapter 2 , moebius et al .
, chapter 8) .
the most readily available diagnostics of the paleoheliosphere are radioisotopes , formed by cosmic ray spallation on the atmosphere , interplanetary and interstellar dust , and meteorites .
thus , the evaluation of cosmic ray modulation for various types of interstellar cloud boundary conditions is a key part of understanding the paleoclimate records that might trace the solar journey through the milky way galaxy .
fahr et al .
( chapter 9 ) and florinski and zank ( chapter 10 ) use our understanding of galactic cosmic ray modulation in the modern - day heliosphere as a basis for making detailed calculations of the response of the paleoheliosphere , or the heliosphere as it once was , to the paleolism , or the local interstellar medium that once surrounded the sun .
the predictions of these calculations are quite intriguing . both the termination shock compression ratio and the solar wind turbulence spectrum may vary dramatically with different environments , as mass - loading by pickup ions and the heliosphere properties vary .
the problem of galactic cosmic ray modulation in an ism - forced heliosphere is extremely important to understanding the paleoheliosphere signature in the terrestrial isotope record . today
, galactic cosmic rays ( gcr ) with energies @xmath47 gev penetrate the solar system , and anomalous cosmic rays ( energies @xmath48 gev ) are formed from accelerated pickup ions .
the cosmic ray flux at earth is sampled by geological radioisotope records , as reviewed kirkby and carslaw ( chapter 12 , also see florinski and zank , chapter 10 ) .
astronomical data indicates that the sun has emerged from a region of space with virtually no neutral ism within the past @xmath00.41.5 10@xmath49 years , and entered the local fluff ( chapter 6 ) . the gcr modulation discontinuity that accompanied this transition may be in the geologic record , which show lower cosmic ray fluxes at earth , on the average , for the past 135 kyrs years than the 135 kyrs before that ( @xcite ) .
harlow shapley ( 1921 ) suggested that an encounter between the sun and giant dust clouds in orion may have perturbed the terrestrial climate and caused ice ages .
the discovery of interstellar and inside the heliosphere was soon followed by studies of the ism influence on the atmosphere for dense cloud conditions ( @xcite ) .
yeghikyan and fahr ( chapter 11 ) , evaluate the density of ism at the earth based on models describing the heliosphere inside of an dense cloud , and the interactions between the solar wind and ism for these dense cloud conditions ( also see chapter 9 , by fahr et al . ) .
these models then yield the concentration of interstellar hydrogen at the earth , and the flow of water downward towards the earth s surface , as a function of the dense cloud density .
significant atmosphere modifications are predicted in some cases . enhanced neutral populations at 1 au for a somewhat lower interstellar cloud density regime are discussed in chapter 8 , by moebius et al . both solar activity cycles ( figure [ fig:2 ] ) and
ism variations modulate the cosmic ray flux in the heliosphere , and kirkby and carslaw ( chapter 12 ) compare galactic cosmic ray records with paleoclimate archives .
they examine sources of climate forcing such as solar irradiance and cosmic ray fluxes , and conclude that arguments in favor of cosmic ray climate forcing are strong although the mechanism is uncertain .
this relation between cosmic ray flux levels and the climate is shown by radioisotope records and climate archives , such as ice cores , stalagmites , and ice - rafted debris , and for modern times , by historical records .
paleoclimate archives include terrestrial records of cosmic ray spallation in the atmosphere , as traced by radioisotopes with short half - lives ( ) , e. g. ( = 5,730 yrs ) and ( = 1.6 myrs ) .
possible mechanisms linking the cosmic ray flux at 1 au and the climate include cloud nucleation by cosmic rays , and the global electrical circuit ( see chapter 12 and @xcite ) .
the discussion in chapter 12 provides persuasive evidence linking the surface temperature to cosmic ray fluxes at earth .
the anticorrelation between sunspot number and cosmic ray fluxes in figure [ fig:2 ] shows the heliosphere role in cosmic ray modulation ; this mechanism must have also been a prominent mechanism for relating the ism - modulated heliosphere with the climate .
fortunately this hypothesis is also verifiable by comparing paleoclimate data with astronomical data on the timing of cloud transitions .
the radioisotope records also indicate that cosmic ray fluctuations have occurred over longer timescales of many @xmath50 years .
shaviv compares the ( @xmath00.3 myrs ) and ( @xmath01.3 gyr ) cosmic ray exposure records in iron meteorites ( chapter 5 ) , but in this case to obtain cosmic ray flux increases due to the sun s location in spiral arms where active star formation occurs .
a number of studies , none convincing , have invoked the geological record , as a proxy for cosmic ray fluxes at earth , to infer historical encounters with interstellar clouds . as a way of dating the loop i supernova remnant , it was suggested that the relative constancy of in sea sediments precluded a strong nearby x - ray source within the past @xmath02 myrs ( @xcite ) .
sonett ( 1992 ) suggested that peaks in layers 35,000 and 65,000 years ago resulted from a compressed heliosphere caused by the passage of a high - velocity interstellar shock .
this extreme heliosphere compression expected for a rapidly moving cloud is supported by heliosphere models ( chapter 2 ) .
structure in the peaks has also been related to spatial structure in the local ism ( @xcite ) , and solar wind turbulence caused by mass - loading of interstellar neutrals may supply the required mechanism .
global geomagnetic excursions such as the events @xmath032 kyr and @xmath040 kyr ago also affect the record , and can not be ignored ( @xcite ) .
indeed , figure [ fig:2 ] shows the sensitivity of galactic cosmic ray fluxes on earth to geomagnetic latitude .
this brief summary of the scientific question motivating this book does not relay the full significance of the galactic environment of the sun to the heliosphere and earth ; the following chapters provide deeper insights into this question .
historical and paleoclimate data show a correspondence between high cosmic ray flux levels and cool temperatures on earth ( @xcite ) .
the disappearance of sunspots for extended periods of time , such as the maunder minimum in the years 1645 to 1715 , shows up in terrestrial radioisotope records such as in ice cores ( chapter 12 ) .
the solar magnetic activity cycle was present during this period , and cosmic ray modulation by the heliosphere was still evident ( @xcite ) .
the record now extends to @xmath010@xmath49 years before present , raising the hope that encounters between the sun and interstellar clouds can be separated from solar activity effects , and from the global signature of geomagnetic pole wandering .
sunspots have long been controversial as an influence on the terrestrial climate .
sir william herschel carefully observed them , and postulated that diminished solar radiation at earth during sunspot maximum affected the terrestrial climate ( 1801 ) .
langley ( 1876 ) measured the radiative heat from sunspot umbral and penumbral regions , and concluded the @xmath320.1% solar radiation decrease associated with sunspots was inadequate to affect the climate .
climate records show that the maunder minimum and other periods of low solar activity levels have been exceptionally cold , which implicates high cosmic ray fluxes with cold climate conditions .
solar activity levels have returned to historic highs in the past few decades ( @xcite ) , and the historic correlations indicate these high levels also yield warm climate conditions .
unfortunately , these scientific conclusions also impact the politically loaded issue of global warming .
the possibility that the cosmos has affected the terrestrial climate is a longtime source of speculation , with many of the first discussions focused on explaining the universal deluge `` . in 1694
edmond halley presented his thoughts to the royal society as to whether the ' ' casual shock of a comet , or other transient body `` might instantly alter the axis orientation or diurnal rotation of the earth , thus disturbing sea levels , or whether the impact of a comet could explain the presence of ' ' vast quantities of earth and high cliffs upon beds of shells , which once were the bottom of the sea " ( @xcite ) .
halley s speculation has resurfaced in the hypothesis that the impact of a comet led to the extinction of dinosaurs 65 myrs ago at the cretaceous - tertiary boundary ( @xcite ) .
the common sense disclaimer that accompanied halley s discussion is timeless : _
`` ... the almighty generally making us of natural means to bring about his will , i thought it not amiss to give this honourable society an account of some thoughts that occurrd to me on this subject ; wherein , if i err , i shall find myself in very good company . '' _ the articles in this volume show firmly that the interaction between the heliosphere and ism depends on the detailed boundary conditions set for the heliosphere by each type of interstellar cloud encountered by the sun , and that the galactic environment of the sun changes over both geologically short time scales of @xmath51 years , and long time scales of @xmath52 years .
this interaction , in turn , affects the flux of gas , dust , and energetic particles in the inner heliosphere .
the discussions in this book also apply to the study of astrospheres around cool stars , which are expected to have similar properties as the heliosphere .
is the historical astrosphere of a star a factor in climate stability for planetary systems ?
i think so ( @xcite ) .
if so , then the sample of @xmath0100 detected extrasolar planetary systems can be narrowed to those that are the most likely to harbor technological civilizations by evaluating the astrosphere characteristics suitable to the space trajectory of each star ( @xcite ) .
astrospheres have now been detected towards @xmath060% of the observed cool stars within 10 pc ( @xcite ) , and extensive efforts to detect earth - sized exoplanets are underway . perhaps some day these questions will be answered .
_ acknowledgments : _ the author thanks dr .
clifford lopate , of the university of new hampshire , for providing figure [ fig:2 ] , and dr .
lopate thanks nsf grant 03 - 39527 for supporting the research displayed in this figure .
the author thanks nasa for supporting her research , including grants nag5 - 13107 and nng05gd36 g .
additional support has been provided by grants nag 5 - 13558 and nag5 - 11999 .
this article will appear in the book
solar journey : the significance of our galactic environment for the heliosphere and earth , springer , in press ( 2006 ) , editor p. c. frisch .
ll object & description + _ interstellar : _ & + interstellar material , ism & atoms in the space between stars + local fluff or clic & ism within @xmath030 pc , density @xmath53 g + & clic = cluster of local interstellar clouds + local interstellar cloud , lic & the cloud feeding ism into the solar system + local bubble , lb & nearby ism with density @xmath54 g + & + _ heliosphere : _ & + solar wind , sw & solar plasma expanding to form heliosphere + & density@xmath05 ions , velocity @xmath0450 km s@xmath37 + & at earth + neutral current sheet & thin neutral region separating sw + & with opposite magnetic polarities + heliosphere , hs & region of space containing the solar wind + termination shock , ts & shock where solar wind becomes subsonic + & ts at @xmath094 au on 16 december , 2004 + heliosheath & subsonic solar wind , outside ts + heliosphere bow shock & shock where lic becomes subsonic + focusing cone & gravitationally focused ism dust + & and helium gas downwind of the sun + & + + pickup ions , pui & ions from sw - ism charge exchange + energetic neutral atoms & enas , energetic atoms formed by + & charge exchange with ions + cosmic rays : & + anomalous , acr & accelerated pickup ions , energy @xmath321 gev + galactic , gcr & from supernova , energy @xmath401 gev at earth +
the nine planets of the solar system ( including pluto as a planet ) extend out to 39 au , compared to the distance of the solar wind termination shock in the upwind direction of 94 au .
the earth is 8.3 light minutes from the sun , versus the @xmath00.5 light day distance to the upwind termination shock of the solar wind .
the ecliptic and galactic planes are tilted with respect to each other by @xmath060@xmath8 , and the north ecliptic pole points towards the galactic coordinates @xmath55=96.4@xmath8 and @xmath56=+29.8@xmath8 .
this tilt allows the separation of large scale ecliptic and large scale galactic phenomena by geometric considerations .
acronyms are used throughout this book , and some of these are listed in table 1 . for those new to this subject , an astronomical unit , au , is the distance between the earth and sun .
a parsec , pc , is 206,000 au , 3.3 light years ( ly ) , or 3.1 @xmath44 10@xmath57 cm . for comparison ,
the earth is 8.3 light minutes from the sun , and the nearest star , @xmath58 cen , is 1.3 pc from the sun .
the planet pluto is at 39 au , versus the 94 au termination shock distance .
fahr , hans j. , fichtner , horst , scherer , klaus , and stawicki , olaf ( 2006 ) . .
in _ solar journey : the significance of our galactic environment for the heliosphere and earth , in press _ , page 000 .
springer .
, p. c. , dorschner , j. m. , geiss , j. , greenberg , j. m. , grn , e. , landgraf , m. , hoppe , p. , jones , a. p. , krtschmer , w. , linde , t. j. , morfill , g. e. , reach , w. , slavin , j. d. , svestka , j. , witt , a. n. , and zank , g. p. ( 1999 ) . .
, 525:492516 . ,
e. , bzowski , m. , chalov , s. , fahr , h .- j . ,
gloeckler , g. , izmodenov , v. , kallenbach , r. , lallement , r. , mcmullin , d. , noda , h. , oka , m. , pauluhn , a. , raymond , j. , ruci ' nski , d. , skoug , r. , terasawa , t. , thompson , w. , vallerga , j. , von steiger , r. , and witte , m. ( 2004 ) . .
, 426:897907 .
mbius , eberhard , bzowski , maciek , mller , hans - reinhard , and wurz , peter ( 2006 ) . .
in _ solar journey : the significance of our galactic environment for the heliosphere and earth , in press _
, page 000 .
springer .
zank , gary p. , mller , hans - r . ,
florinski , vladimir , and frisch , priscilla c. ( 2006 ) . .
in _ solar journey : the significance of our galactic environment for the heliosphere and earth , in press _
, page 000 . springer . | speculations that encounters with interstellar clouds modify the terrestrial climate have appeared in the scientific literature for over 85 years .
the articles in this volume seek to give substance to these speculations by examining the exact mechanisms that link the pressure and composition of the interstellar medium surrounding the sun to the physical properties of the inner heliosphere at the earth .
heliosphere , interstellar clouds , interstellar medium , cosmic rays , magnetosphere , atmosphere , climate , solar wind , paleoclimate |
the yttrium aluminum perovskite ( yap ) is an inorganic material well known for its interesting mechanical and chemical properties .
its high light yield and good resolution have made it suitable for a variety of applications in optical physics , ranging from solid state lasers @xcite and scintillators @xcite to medical apparatus @xcite and recording media @xcite .
this rare - earth ( re ) aluminate has also demonstrated very good performance in the microwave range , with a relative permittivity value of order 16 and low loss@xcite , which are valuable for the design of dielectric resonators . also , due its high heat conductivity , it has found use as a substrate material for thin films of high temperature superconductors with a wide range for microwave component applications , such as use in microstrip lines @xcite .
recently , experimentalists have shown that re electron spin defects in yap have great potential for the development of crucial components for quantum computers and quantum communication devices @xcite .
in fact , strong coupling has been demonstrated in circuit qed experiments using erbium doped yap , er@xmath1:yalo@xmath2@xcite at microwave frequencies , with the aim to implement quantum memories using hybrid architecture . however , the microwave properties of yap are not well characterized and such experiments would greatly benefit from a rigorous characterization of the material properties , especially at dilution fridge temperatures .
the cavity - resonator method has been implemented previously@xcite to measure the relative permittivity and dielectric losses of yap .
this work presented a single permittivity component at a frequency of 7.767 ghz equal to 16 at room temperature .
this determined value decreased no more than 4% when the crystal was cooled down to 20 k. however , yap is a biaxial crystal with orthorhombic symmetry ; which means it has a diagonal permittivity tensor with three distinct elements as in eq .
[ eq1 ] @xcite .
therefore , full determination of its permittivity tensor components is desirable and will allow more accurate design in applications using this material .
@xmath3 for those reasons , this work presents for the first time a thorough characterization of real permittivity tensor of yap using the split - post dielectric resonator ( spdr ) and whispering gallery mode ( wgm ) techniques .
temperature dependence of the tensor components and the crystal losses down to 20 mk were also investigated and temperature coefficients of permittivity ( tcp ) were calculated . following this we performed spectroscopy of impurity ions in the sample using the wgm technique @xcite at 20 mk .
new impurity ion transitions were measured with zero - field splittings of 16.72 and 9.92 ghz .
their correspondent interaction with the microwave photonic modes were also analyzed with respective coupling strengths determined .
c2.7 cm c1.5 cm c1.5 cm c1.5 cm temperature ( k ) & @xmath4 & @xmath5 & @xmath6 + 0.020 & 16.316 & 15.535 & 15.128 + 1 & 16.316 & 15.535 & 15.128 + 4 & 16.321 & 15.531 & 15.126 + 10 & 16.321 & 15.531 & 15.126 + 23 & 16.323 & 15.530 & 15.126 + 29 & 16.324 & 15.530 & 15.126 + 39 & 16.327 & 15.531 & 15.128 + 50 & 16.328 & 15.535 & 15.133 + 58 & 16.333 & 15.537 & 15.139 + 65 & 16.341 & 15.538 & 15.145 + 69 & 16.342 & 15.542 & 15.149 + 295 & 16.582 & 15.997 & 15.673 + room temperature measurement of the mode frequency versus azimuthal mode number ( _ m_).,scaledwidth=42.0% ] through the spdr technique @xcite the three components of the real part of the complex permittivity tensor of the yap were initially determined only at room temperature and with low precision . to perform this measurement three flat laminar crystal samples of 0.52 mm thickness with 1% variation between samples were used .
such samples had the remaining dimensions equal to 10 mm @xmath7 10 mm and each square face perpendicular to one of the yap principal axes .
the spdr technique uses the transverse electric mode to probe the permittivity of the two crystal axes perpendicular to the thickness direction . in this way ,
an initial estimate of the permittivity tensor may be provided . for the yap crystal
the permittivity components were measured as @xmath8 , @xmath9 and @xmath10 . in order to have a more accurate measure of the permittivity tensor for a bulk yap sample , we then implemented the more precise wgm technique @xcite , allowing us to refine the preliminary results .
the wgm method uses a cylindrical bulk sample placed within a cylindrical metallic cavity .
multiple electromagnetic modes are then excited within the dielectrically loaded cavity , with the wgms families are selected to perform the permittivity characterization . a cylindrical yap sample of @xmath11 mm height and diameter equal to @xmath12 mm made by scientific materials corporation was used .
it had a concentric hole @xmath13 mm diameter parallel to its longitudinal axis , also parallel to the crystal crystalographic a - axis .
the sample was placed into a copper cavity and supported by a teflon holder to position it at near the center of the cavity ( cavity design in detail in @xcite ) .
two electromagnetic probes were inserted into the cavity and connected to the input / output signal of a vector network analyzer ( vna ) operating in transmission mode .
the fundamental wgm photonic mode families in the yap crystal were recorded in a frequency range from 6 to 20 ghz at room temperature as shown in fig .
[ fig1 ] . in this frequency range
typically the real part of permittivity of dielectric crystals have negligible frequency dependence @xcite .
so to calculate the permittivity tensor components a finite element method ( fem ) model of the resonator was implemented , first using the initial estimates of the permittivity tensor components from the spdr technique .
the permittivity values were then incremented around this value with steps of @xmath14 until the best match between the experimental ( shown in fig . [ fig1 ] ) and
simulated modes was reached .
the real permittivity tensor at room temperature for yap was obtained as shown at the last line of table [ tab1 ] .
the discrepancy between the experimental mode frequencies and the simulated data were typically 10 mhz on average for the fundamental wgm families , with the uncertainties mainly due the irregularities of the crystal dimensions .
to perform a temperature dependent determination of the real permittivity , two wgms of different polarization must be selected @xcite .
we chose the wgh@xmath15 and wge@xmath16 modes as they had well distinguished doublet modes required to determine the biaxial nature and were of high enough order to ensure enough mode confinement in the crystal@xcite .
measurement of the yap real permittivity tensor components as a function of temperature.,scaledwidth=38.0% ] measured temperature dependence of the electrical q - factor of the modes wge@xmath15 and wgh@xmath16 and their respective doublets pairs ( marked with an asterisk).,scaledwidth=45.0% ] tcp data of the biaxial yap , yso @xcite and uniaxial sapphire ( sap ) @xcite crystal permittivity elements.,scaledwidth=42.0% ] the cavity containing the sample was placed into a bluefors ld-250 dilution refrigerator and cooled down to 20 mk .
all the selected mode frequencies and electrical quality factors ( q - factors ) were tracked and recorded over the warming up process .
the fem model was implemented for an additional 11 different temperatures measurements to determine the three components of permittivity as a function of temperature .
the thermal contraction of the copper cavity and the sample were taken into consideration , whereas the teflon contribution was determined to be negligible from the fem model .
table [ tab1 ] shows the calculated permittivity tensor over the measured temperature range .
they are also presented in the fig .
[ fig2 ] , alongside with their respective second order polynomial fit .
these results are consistent with the results published by @xcite and show no significant variation below 40 k. under 1 k , small frequency shifts with temperature are observed , which are below the precision of the permittivity determination . this effect has been well documented in cryogenic sapphire resonators and was shown to be due to residual paramagnetic impurity defects within the crystal lattice @xcite . this outcome has been used to annul the temperature coefficient of frequency ( tcf ) of high - q microwave wgms , by operating at the temperature where the temperature coefficient of magnetic susceptibility of the impurity ion cancels the shift due to the tcp @xcite .
this is essential for the realization of the ultra stable cryogenic sapphire oscillator , which attains exquisite frequency stability of order parts in @xmath17 @xcite .
similar paramagnetic defects in sapphire have also been used to create high stability masers @xcite , precision frequency conversion @xcite and the generation of high stability frequency combs due to four wave mixing @xcite .
[ fig3 ] presents how the crystal losses change in cryogenic temperatures .
the q - factor of the resonance modes was measured at the same temperatures the permittivity components were probed .
the results show the losses decrease by about three to four times from 300 k to 40 k and do not vary much more for colder temperatures .
the tcp was also calculated and in fig .
the yap coefficients are compared to other two other widely used single crystals .
the data points are only presented for temperatures above 30 k , away from effects due to residual paramagnetic impurities .
an immediate analysis shows that yap has a stronger temperature dependence of permittivity than sapphire , however sapphire is not a good host for re elements @xcite , which is an undesirable feature for several applications . comparing to the yttrium orthosilicate ( yso )
, yap s tcp does not have significant difference , but still has the advantage of having relatively lower magnetic anisotropy , which make it more suitable for operation in superconducting cavities and circuits @xcite .
the evidence of paramagnetic impurity defects was originally observable from the frequency temperature dependence of the microwave modes at low temperature .
thus , due to the potential of the yap crystal for microwave and optical quantum information applications , we performed electron spin resonance ( esr ) spectroscopy close the lowest temperature of our measurements , where spins in the lattice may be coupled strongly with photons and manipulated with qubits @xcite . moreover , recently iron group impurity ion spin states were discovered in single crystal yso with strong coupling to whispering gallery photons@xcite , and in this work we confirm similar spin ensembles in yap .
energy level splitting of two ionic impurity transitions with applied magnetic field .
each dot corresponds to an avoided level crossing ( alc ) , the horizontal shaded line is representing the wgm at 15.18 ghz and the angled solid and broken lines are showing the zeeman effect of the @xmath18 and @xmath19 zero - field splittings , respectively .
the effective g - factors of the transitions are also included.,scaledwidth=49.5% ] magnetic field dependence of the cavity transmission from 0 to 0.25 t. three interactions between the 15.18 ghz photonic mode and the crystal spins ensembles are labeled by the letters a , b and c.,scaledwidth=49.5% ] the experiment consisted of placing the copper cavity containing the yap cylinder into a superconducting magnet in such a way that a dc magnetic field would be parallel to the crystal axis ( more details of the experimental apparatus can be found in @xcite ) .
transmission measurements identified several wgms generated in the crystal between 6 and 20 ghz and at 20 mk in temperature .
these modes were then monitored in a narrow frequency span and at low input power while the magnet performed a sweep from 0 to 0.1 t. the data collected from the vna showed the signatures of several interactions between spin states and photonic modes as indicated by avoided level crossings ( alcs ) , i.e. , the normal mode splitting resultant from the hybridization of two harmonic oscillators .
[ fig5 ] presents a map where the relevant alcs are represented by dots and are connected by lines to emphasize the zeeman effect .
the two indicated zero - field splittings ( @xmath18 and @xmath19 ) at 16.720 and 9.917 ghz , demonstrate the presence of two ionic impurity transitions in the crystal lattice .
they are characterized by the indicated effective g - factors , @xmath20 , which are related to the nature of unintentional dopants and are calculated from @xmath21 , where @xmath22 is the slope of line fitting the transition and @xmath23 is the bohr magneton .
particularly interesting due to apparently higher spin - photon couplings , the wgm at 15.18 ghz ( highlighted by the shaded horizontal line in fig .
[ fig5 ] ) was subject to further investigation .
[ fig6 ] shows the transmission spectroscopy plot of the spin - photon interactions when the magnetic field is changed from 0 to 0.25 t. in this plot , the alcs a and b are explained by the hybridization of the photonic mode with the @xmath24 and the @xmath25 spins , respectively . in contrast , the mode splitting c indicates the coupling with an re impurity , responsible for the additional splitting caused by the hyperfine ( hf ) transitions between states . [
cols="^,^,^,^,^,^,^",options="header " , ] in reference @xcite , the resonant photon interaction with the electron spin and corresponding hf structure of the @xmath26 isotope with a nuclear spin of i = 7/2 was rigorously researched . in fig .
[ fig6 ] a similar prominent alc is recognised , due to the transition between @xmath27 spin states , surrounded by eight hf transitions .
in contrast , the transitions associated with the a and b alcs have never before been identified , and are thus of potential interest for application in hybrid quantum architectures . to determine the properties of the 15.18 ghz wgm interacting with
the measured spins ensembles , the alcs were fitted using a two - mode coupled harmonic oscillators model .
this model assumes two representative lcr oscillators ( labeled with subscripts 1 and 2 ) coupled reactively through the magnetic field ( mutual inductance ) with a dimensionless coupling coefficient of @xmath28 @xcite .
the calculated characteristic equation for the coupled mode system is written in eqn .
[ eq2 ] , ignoring all terms of order @xmath29 .
this is a very good approximation for any system of relatively high q - factor ( i.e. above 10 ) .
@xmath30 to solve for the normal mode frequencies , and hence coupling between the normal modes , all q - values were set to infinity and the normal mode frequencies , @xmath31 and @xmath32 , were solved for by setting the characteristic equation to zero .
the only unknown to determine was the coupling strengths , which in hz is given by @xmath33 , and was determined by fitting the experimental data ( assigning subscript 1 to the wgm and subscript 2 to the spin ensemble ) .
results of the fits for the tree photon - spin interactions are shown in fig .
[ fig7 ] ( a)-(c ) , with the numeric values of @xmath34 listed in table [ tab2 ] .
the hybridized spin - photon system clearly shows a coupling greater than the wgm photon linewidth .
however due to the losses introduced by the spin ensembles in the interaction , it was difficult to estimate the spin linewidth ( or effective spin q - factor ) as the cavity coupling was reduced to below the sensitivity of our experiment , and it could not be determined if we were in the strong coupling regime .
thus , to solve for the spin linewidths ( or effective spin q - factor ) we used a similar technique as developed in reference @xcite for a coupled n - mode system of mechanical oscillators . to do this we write the characteristic equation for the normal mode system as
; @xmath35 then we equate odd powers of @xmath36 between eqn .
[ eq2 ] and eqn .
[ eq3 ] ( note , equating even coefficients verifies the same normal mode frequencies as already determined ) .
subsequently the following matrix equation may be derived to solve for normal mode q - factors .
@xmath37 the photonic q - factor , @xmath38 ( @xmath39 ) was easily measured at zero field and by measuring the normal mode q - factors during the mode hybridization , the effective q - factor of the spin ensemble , @xmath40 ( @xmath41 ) , may be estimated by fitting the data to the matrix equation above ( given by eqn .
[ eq4 ] ) .
the data along with the fitting curves are shown in fig .
[ fig7 ] ( a)-(f ) and the extracted parameters shown in table [ tab2 ] .
although the coupling strengths of the three interactions exceed the photonic half linewidths ( @xmath42 ) , the spins half linewidths ( @xmath43 ) surpassed the couplings by a factor of three , which means the spins transitions losses prevented the strong coupling regime ( @xmath44 ) to be reached .
the collective coupling strength , nevertheless , is related to the spins concentration , @xmath45 , by @xmath46 @xcite , where @xmath47 and @xmath48 are the frequency of the wgm and the sample transverse magnetic filling factor , respectively . for the 15.18 ghz resonance , @xmath48 is about 84% , corresponding to the simulated @xmath49 mode structure .
thus the estimated average spin concentration is on the order of 10@xmath50 @xmath51 and increasing the concentration of impurities in the crystal by intentional doping would allow strong coupling between the photonic mode and the spins ensembles .
in conclusion , this work presented an investigation of the yap microwave properties . in the first part
, a precise measurement of the three non - zero components of the yap real permittivity tensor was performed and the values of @xmath52 , @xmath53 and @xmath54 were obtained with the sample at 20 mk .
mode frequencies predicted from the found permittivity tensor differ on average by only 10 mhz from the measured data , with the uncertainties mainly due the imprecision in the determination of the sample s dimensions .
the crystal losses were also assessed through the measurement of the q - factors of four photonic modes and a study of the temperature dependence of these properties were performed between room and sub - kelvin temperatures , including the measurement of the temperature coefficient of permittivity . in the second part , esr spectroscopy of the same sample was performed at 20 mk and allowed the identification of two new electronic spin impurity transitions with zero - field splitting at 16.72 and 9.92 ghz and spin - photon couplings strengths of @xmath55 and @xmath56 mhz , respectively , exceeding the wgm linewidth of 116 khz .
spin ensembles coupled to a microwave resonator is a promising architecture for the reversible transfer of quantum states between systems in the ghz range @xcite , in the sense the presented results represent a valuable information for the design of experiments aiming to implement the use of yap crystals in hybrid quantum systems .
this research is supported by the australian research council , ce110001013 and by the conselho nacional de desenvolvimento cientfico e tecnolgico ( cnpq brazil ) .
the authors would like to thank warrick farr and nikita kostylev with some help with data acquisition and processing .
35ifxundefined [ 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 * * , ( ) link:\doibase 10.1103/physrevb.87.094412 [ * * , ( ) ] link:\doibase 10.1103/physrevb.88.224426 [ * * , ( ) ] link:\doibase 10.1103/physrevb.89.224407 [ * * , ( ) ] link:\doibase 10.1103/physrevb.92.060406 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | low loss crystals with defects due to paramagnetic or rare earth impurity ions is a major area of investigation for quantum hybrid systems at both optical and microwave frequencies . in this work
we examine the single crystal yttrium aluminium perovskite , yalo@xmath0 using the whispering gallery mode technique .
multiple resonant microwave modes were measured from room temperature to 20 mk allowing precise characterization of the permittivity tensor at microwave frequencies .
we show that it is biaxial and characterize the tensor as a function of temperature with estimated uncertainties below 0.26% .
electron spin resonance spectroscopy was also performed at 20 mk , with new transitions identified with zero - field splittings of 16.72 and 9.92 ghz .
spin - photon couplings of order 4.2 and 8.4 mhz were observed for residual levels of concentration , which are stronger than the photon cavity linewidths of 116 khz but the same order of the linewidths of the discovered spin transitions . |
macromolecules containing ionizable groups when dissolve in a polar solvent such as water , dissociate into charged macromolecules and counterions ( ions of opposite charge ) . depending on acidic or basic property of their monomers ,
ionizable polymers in solution can be classified into polyelectrolytes and polyampholytes .
polyelectrolytes contain a single sign of charged monomers and polyampholytes bear charged monomers of both signs .
these macromolecules are often water - soluble and have numerous industrial and medical applications .
many biological macromolecules such as dna , rna , and proteins are charged polymers . in polymer science
, charged polymers has been an important subject during last several decades @xcite .
contrary to a polyelectrolyte chain in which the intra - chain electrostatic interactions are repulsive and tend to swell the chain , in a polyampholyte chain attractive interactions between charged monomers of opposite sign tend to decrease the chain size .
oppositely charged monomers can be distributed randomly along a polyampholyte chain or charges of one sign can be arranged in long blocks . with
the same ratio of positively and negatively charged monomers ( isoelectric condition ) , behaviors of a single polyampholyte and the solution of polyampholytes depend noticeably on the sequence of charged monomers on the chains .
for example , it has been shown that the sequence of charged amino acids ( charge distribution ) along ionically complementary peptides affect the aggregation behavior and self - assembling process in the solution of such peptides @xcite .
also , using monte carlo simulations it has been shown that charged monomers sequence of charge - symmetric polyampholytes affect their adsorption properties to a charged surface @xcite .
the properties of the system of polymers anchored on a surface are of great interest both in industrial and biological applications and academic research . with a sufficiently strong repulsion between the polymers ,
the chains become stretched and the structure obtained is known as a polymer brush .
planar and curved brushes formed by grafted homopolymers have extensively been investigated by various theoretical methods @xcite .
the anchored polymers of a brush may be consisting of charged monomers . in this case
, the brush is known as a polyelectrolyte or a polyampholyte brush depending on the charged monomers of the chains being of the same sign or being composed of both signs . in a brush of charged polymers ,
electrostatic interactions introduce additional length scales such as bjerrum length and debye screening length to the system . in a polyelectrolyte brush , the repulsion of electrostatic origin between the chains can be sufficiently strong even at low grafting densities , making it easy for the system to access the brush regime .
polyelectrolyte brushes have been investigated extensively using both theoretical @xcite and computer simulation methods @xcite . at high enough grafting densities and charge fractions of polyelectrolyte chains ,
most of counterions are trapped inside the polyelectrolyte brush and competition between osmotic pressure of the counterions and elasticity of the chains determines the brush thickness .
this regime of a polyelectrolyte brush is known as the osmotic regime in which some theoretical scaling methods predict no dependence of the brush thickness to the grafting density @xcite .
however , other scaling method that takes into account the excluded volume effects and nonlinear elasticity of polyelectrolyte chains and is in agreement with experiment and simulation , predicts a linear dependence of the brush thickness on the grafting density @xcite .
also , it has been shown that diffusion of a fraction of counterions outside the polyelectrolyte brush leads to a logarithmic dependence of the average brush thickness on the grafting density @xcite .
electrostatic interactions in a polyelectrolyte brush cause most of the counterions to be trapped inside the brush and help the chains to be more stretched and the brush to be more aligned . however , in a brush of overally neutral polyampholyte chains , most of counterions are outside the brush and the electrostatic correlations tend to decrease the chains size and the brush thickness . at a given value of the grafting density , the average thickness and equilibrium properties of such a polyampholyte brush are mainly determined by the chains properties such as fraction and sequence of charged monomers and the bending energy .
brushes formed by grafted diblock polyampholytes have been investigated by lattice mean field modeling @xcite and computer simulation @xcite .
the effect of chain stiffness , charge density and grafting density on spherical brushes of diblock polyampholytes and interaction between colloids with grafted diblock polyampholytes have been studied using monte carlo simulations @xcite .
also , using molecular dynamics ( md ) simulations , the effects of various parameters such as charged monomers sequence , grafting density and salt concentration on the average thickness and equilibrium conformations of planar semiflexible polyampholyte brushes have been investigated @xcite . in this paper , we study planar brushes of flexible , semiflexible and rodlike diblock polyampholytes using md simulations in a wide range of the grafting density .
we find that in all cases the average brush thickness linearly depends on the grafting density regardless of the chains different flexibility .
our results also show that the strength of this dependence is considerably weaker in the case of the brush of flexible polyampholytes than two other cases . despite mentioned same functionality obtained for the average thickness versus the grafting density for brushes of different polyampholytes
we find that histograms of their equilibrium conformations are noticeably different .
inter - chain correlations are too weak in the brush of flexible polyampholytes and the brush properties are dominantly determined by single chain behavior . in this case ,
dependence of the equilibrium conformation of the brush on the grafting density is very weak . in the cases of the brushes of semiflexible and rodlike polyampholytes however , because of the combination of electrostatic correlations and strong excluded volume effects , collective behavior of the chains is dominant and dependence of equilibrium conformations on the grafting density is strong . in these cases ,
we also observe separation of the anchored chains into two coexisting fractions . using a simple scaling method which is consistently applicable for the brush of flexible chains ,
we describe theoretically the linear dependence of the brush thickness on the grafting density .
the rest of the paper is organized as follows . in section [ simulation ]
we describe our model and simulation method in detail and present the results of md simulations .
our scaling analysis to describe dependence of the average thickness of flexible diblock polyampholyte brushes on the grafting density is presented in section [ theory ] . in section [ discussion ]
we conclude the paper and present a short discussion .
in our simulations which are performed with the md simulation package espresso @xcite , each brush is modeled by @xmath0 diblock polyampholyte bead - spring chains of length @xmath1 ( 24 spherical monomers ) which are end - grafted onto an uncharged surface at @xmath2 .
the positions of anchored monomers which are fixed during the simulation , form an square lattice on the grafting surface ( @xmath3 plane ) with lattice spacing @xmath4 in which @xmath5 is the grafting density of the chains .
the fraction @xmath6 of the monomers of each chain are charged and the chains consist of an alternating sequence of charged and neutral monomers .
each chain contains the same number of positively and negatively charged monomers with charges @xmath7 and @xmath8 respectively ( see fig .
[ fig1 ] ) . excluded volume interaction between particles
is modeled by a shifted lennard - jones potential , @xmath9 in which @xmath10 and @xmath11 are the usual lennard - jones parameters and the cutoff radius is @xmath12 .
successive monomers of each chain are bonded to each other by a fene ( finite extensible nonlinear elastic ) potential @xcite , @xmath13 with bond strength @xmath14 and maximum bond length @xmath15 . bending elasticity of the chains
is modeled by a bond angle potential , @xmath16 in which @xmath17 is the angle between two successive bond vectors and @xmath18 is the bending energy of the chains .
the value of the persistence length , @xmath19 , of the chains depends on the value of @xmath18 as @xmath20 . to model brushes of flexible , semiflexible and rodlike chains , we use four different values of @xmath18 namely @xmath21 , @xmath22 , @xmath23 and @xmath24 respectively .
the simulation box is of volume @xmath25 in which @xmath26 is the box width in @xmath27 and @xmath28 directions and @xmath29 is its height in @xmath30 direction and the grafting density is given by @xmath31 .
we consider @xmath32 monovalent counterions to neutralize the chains charge .
positive and negative monovalent counterions are modeled by equal number of spherical lennard - jones particles of diameter @xmath11 with charges @xmath7 and @xmath8 respectively .
all the particles interact repulsively with the grafting surface at short distances with the shifted lennard - jones potential introduced in eq .
in addition , a similar repulsive potential is applied at the top boundary of the simulation box and in our simulations @xmath33 .
all the charged particles interact with each other with the coulomb interaction @xmath34 in which @xmath35 and @xmath36 are charges of particles @xmath37 and @xmath38 in units of elementary charge @xmath7 and @xmath39 is separation between them .
the bjerrum length , @xmath40 , which determines the strength of the coulomb interaction relative to the thermal energy , @xmath41 , is given by @xmath42 , where @xmath43 is the dielectric constant of the solvent and we set @xmath44 in our simulations .
periodic boundary conditions are applied only in two dimensions ( @xmath27 and @xmath28 ) . to calculate coulomb forces and energies
, we use the so - called @xmath45 technique introduced by strebel and sperb @xcite and modified for laterally periodic systems ( @xmath46 ) by arnold and holm @xcite . the temperature in our simulations
is kept fixed at @xmath47 using a langevin thermostat . for each value of the bending energy , @xmath18 , we do simulations of the brush at dimensionless grafting densities @xmath48 .
in the beginning of each simulation , all of the chains are straight and perpendicular to the grafting surface and all the ions are randomly distributed inside the simulation box .
we equilibrate the system for @xmath49 md time steps which is enough for all values of the grafting density mentioned above and then calculate thermal averages over @xmath50 independent configurations of the system selected from @xmath51 additional md steps after equilibration .
md time step in our simulations is @xmath52 in which @xmath53 is the md time scale and @xmath54 is the mass of the particles .
we calculate the average brush thickness which can be measured by taking the first moment of the monomer density profile @xmath55 in which @xmath56 is the number density of monomers as a function of the distance from the grafting surface . for a better monitoring of the statistics of the chains conformations ,
we calculate the histogram of the mean end - to - end distance of the chains , @xmath57 , in which @xmath58 and @xmath59 is the end - to - end vector of chain @xmath37 .
we also calculate the histogram of the average distance of the end monomers of the chains from the grafting surface , @xmath60 , in which @xmath61 and @xmath62 is the @xmath30 component of the end monomer of chain @xmath37 . with the same method discussed in ref .
@xcite , it has been checked that our results are not affected by finite - size effects ( see sec .
[ discussion ] ) .
the average thickness versus the grafting density for brushes of flexible , semiflexible and rodlike diblock polyampholytes are shown in fig .
dependence of the average thickness on the grafting density can be described well by a linear function for all values of the bending energy , @xmath18 , that we use in our simulations .
also , it can be seen that at all values of the grafting density the average brush thickness decreases with increasing the flexibility of the chains .
linear fits to the average thickness of the brushes of semiflexible and rodlike chains for which the persistence length , @xmath19 , exceeds their contour length , @xmath63 , are of approximately the same slope ( see solid and dashed lines in fig .
[ fig2 ] ) .
the slopes of the linear dependence in cases of two brushes of flexible chains are also approximately the same and differ by a factor of @xmath64 from those of two other cases ( dotted and dash - dotted lines in fig .
[ fig2 ] ) .
to analyze such dependencies of the brushes thickness on the grafting density , we look at the equilibrium conformations statistics of the chains at different values of @xmath5 . in figs .
[ fig3 ] and [ fig4 ] the histograms @xmath57 and @xmath60 at three different grafting densities are shown for brushes of flexible , semiflexible and rodlike chains . because of very similar behaviors of the histograms in @xmath65 and @xmath66 cases , the histograms of the brush with @xmath65 are not shown for clarity of the figures . as it can be seen in fig .
[ fig3 ] , in the case of the brush of flexible diblock polyampholytes ( @xmath67 ) , dependence of the histogram profile on the grafting density is very weak .
also , in this case , the contribution of large values of @xmath39 ( @xmath68 ) in the histogram is negligible which shows that the chains are mostly coiled at all grafting densities ( see a sample configuration of the brush at @xmath69 in fig .
[ fig1]a ) . in the case of the brush of semiflexible chains ( @xmath70 ) , with increasing the grafting density ,
the values of the histogram corresponding to smaller values of @xmath39 become nonzero showing that polyampholyte chains take buckled conformations at high grafting densities @xcite .
as it is expected , the histogram @xmath57 of the brush of rodlike chains ( @xmath71 ) exhibit no noticeable buckling of the chains .
the histograms @xmath60 in fig .
[ fig4 ] show that in the brush of flexible chains at all grafting densities the end monomers are mostly distributed near the grafting surface showing that the positive blocks of the chains are mostly turned back towards the anchored negative blocks .
the profile of this histogram also does nt depend noticeably on the grafting density .
the histograms @xmath60 for brushes of semiflexible and rodlike chains show that two maxima appear and their height increase with increasing the grafting density . by combining the information obtained from histograms @xmath57 and @xmath60 for the brush of rodlike chains
it can be understood that at high grafting densities a fraction of the chains which are perpendicular to the grafting surface coexist with the remaining fraction which fluctuate in the vicinity of the grafting surface . in the case of the brush of semiflexible chains also ,
the fraction of perpendicular chains to the brush surface coexist with the chains which are buckled towards the grafting surface @xcite .
histograms @xmath57 and @xmath60 show that conformations of the brushes of semiflexible and rodlike chains change noticeably with changing the grafting density despite the case of the brush of flexible chains .
as mentioned in the previous section , our md simulations show that the brush thickness is a linear function of the grafting density regardless of different values of grafted diblock polyampholytes bending energy . also , as it is shown in fig .
[ fig2 ] , in cases of the brushes of flexible chains the slope of the fitted line to the average thickness versus the grafting density is considerably smaller than that in two other cases .
we use here a simple scaling theory similar to that of the solution of charge - symmetric diblock polyampholytes @xcite to describe the linear dependence of the brush thickness on the grafting density .
this scaling analysis is applicable to the brushes of flexible diblock polyampholytes .
consider @xmath72 flexible diblock polyampholyte chains end - grafted to a flat surface of area @xmath73 .
the degree of polymerization of each chain and the fraction of charged monomers are @xmath74 and @xmath75 respectively .
electrostatic attractive interactions between oppositely charged blocks of the chains lead them to form a dense layer of positive and negative monomers of average thickness @xmath76 near the grafting surface ( see fig .
[ fig5 ] ) . inside the layer , to use the blob concept we define correlation length @xmath77 as a length scale that electrostatic interactions do nt perturb the chains statistics at smaller length scales and are dominant over thermal fluctuations at larger length scales .
let suppose that the number of monomers inside the correlation blob is @xmath78 and we have @xmath79 where @xmath80 and @xmath81 are the kuhn length and the flory exponent respectively .
accordingly , the layer is a melt of correlation blobs in which positive blobs with a high probability are surrounded by negative blobs .
electrostatic interaction between any two neighboring blobs is of the order of the thermal energy , @xmath41 @xmath82 thus we obtain @xmath83 . as a result ,
the local monomer concentration inside the blob is @xmath84 considering that correlation blobs are space filling , the global monomer concentration , @xmath85 , approximately equals to the local monomer concentration , @xmath86 . thus , dependence of the thickness of a brush of diblock polyampholytes on the grafting density can be obtained as @xmath87 the main prediction of this equation in the range of its validity is as follows .
the brush thickness , @xmath76 , is a linear function of the grafting density , @xmath5 , irrespective of the values of the system parameters .
only the slope of this linear function , @xmath88 , depends on the values of the system parameters .
one should note here that despite the bulk solution of flexible diblock polyampholytes , in a polyampholyte brush the fact that the chains are end grafted to the surface introduces an additional length scale to the system , namely @xmath89 . in the scaling analysis presented here ,
if the size of the correlation length , @xmath77 , exceeds the separation between the grafting points , @xmath90 , the scaling approach becomes inconsistent .
although linear dependence of the brush thickness on the grafting density is observed at all values of the bending energy used in our simulations , the condition @xmath91 is valid only in the case of the brush of flexible chains with @xmath65 . in this case
the values of @xmath92 corresponding to lowest and highest values of the grafting density we have used in our simulations are 0.19 and 0.45 respectively .
for the set of parameters used in our simulations , in the case of the brush with @xmath65 the upper limit of the validity of the scaling method ( @xmath93 ) corresponds to the grafting density @xmath94 which is quite far from our range of the grafting density . for a brush of flexible chains with @xmath65 using the values
@xmath95 and @xmath96 gives the slope @xmath97 for the brush thickness versus the grafting density for the used set of system parameters . as it is shown in fig .
[ fig2 ] , the value of @xmath98 obtained from our simulations of flexible chains is @xmath99 which is in reasonable agreement with prediction of the scaling method .
our scaling method is not consistently applicable for the brush of flexible chains with @xmath100 because the range of the grafting density used in our simulations is higher than the validity range of the condition @xmath91 in this case .
this method is not also applicable for the brushes of semiflexible and rodlike chains because in these cases the kuhn length exceeds the contour length of the chains .
the fact that the average thickness linearly depends on the grafting density at all values of @xmath18 used in our simulations shows that although the scaling method presented here can not be used to describe all the simulation results , this linear dependence persists over a wide range of the system parameters .
brushes of flexible , semiflexible and rodlike diblock polyampholytes have been studied using md simulations and a scaling analysis has been presented to describe the results of the simulation of flexible chins brush .
the average thickness as a function of the grafting density and histograms of equilibrium conformations of the brushes are obtained .
strong dependence of the system conformations on the grafting density and separation of the chains into two coexisting fractions at high grafting densities have been observed in cases of the brushes of semiflexible and rodlike chains . in cases of the brushes of flexible chains ,
single - chain behavior is dominant and dependence of the brush conformations on the grafting density is very weak . in spite of above mentioned differences , it has been observed that dependence of the average brush thickness on the grafting density is linear for brushes of all different chains .
this linear dependence resulted from our md simulations has been described well using a simple scaling method in the case of the brush of flexible chains .
brushes of polyelectrolytes and polyampholytes are dense assembly of these macromolecules in which the interplay between electrostatic correlations , strong excluded volume effects and bending elasticity of the chains determine equilibrium properties of the system
. the main differences between brushes of polyelectrolyte and polyampholyte chains originates from opposite trends of inter- and intra - chain electrostatic interactions and different rules of counterions osmotic pressure in these brushes . in a brush of polyelectrolyte chains
most of counterions are contained inside the brush and their osmotic pressure tends to increase the brush thickness . in a brush of overally neutral polyampholytes
however counterions are outside the brush and have no effect on the brush thickness .
linear dependence of polyelectrolyte brush thickness on the grafting density has theoretically been described @xcite .
the results of a recent simulation of semiflexible polyampholytes @xcite and our simulations and theoretical analysis here show that linear dependence of the average thickness on the grafting density is also the case in the brush of polyampholyte chains .
amount of the flexibility of the chains which control the strength of excluded volume effects and inter - chain correlations , causes the equilibrium conformations of the brushes of three different polyampholytes to be different . in brushes of semiflexible and rodlike polyampholytes , strong dependence of the equilibrium conformations on the grafting density and separation of the chains into two coexisting fractions at high grafting densities
are resulted from strong electrostatic and excluded volume correlations between the chains .
similar phenomenon has been observed in the brush of rodlike polyelectrolytes @xcite . to pay attention to possible finite - size effects in our simulations
, we have calculated the average size of the chains lateral fluctuations as well as the histograms @xmath60 and @xmath57 for a larger brush containing @xmath101 semiflexible chains at the grafting density @xmath102 @xcite .
the average size of lateral fluctuations is defined as @xmath103 , where @xmath104 and @xmath105 denotes averaging over equilibrium configurations .
we have found that the histograms @xmath57 and @xmath60 do not change noticeably with increasing the system size @xcite .
also , we have found that the value of @xmath106 is smaller than the lateral size , @xmath26 , of the simulation box of a brush containing @xmath0 chains at the same grafting density .
this result shows that in simulation of the brush of @xmath0 polyampholytes , the chains do not overlap with their own images .
accordingly , to avoid time consuming simulations of larger brushes , we concentrated on the simulations of the brushes containing @xmath0 chains .
j. rhe , m. ballauff , m. biesalski , p. dziezok , f. grhn , d. johannsmann , n. houbenov , n. hugenberg , r. konradi , s. minko , m. motornov , r. r. netz , m. schmidt , c. seidel , m. stamm , t. stephan , d. usov , h. zhang , adv .
sci . * 165 * , 79 ( 2004 ) . | planar brushes of flexible , semiflexible and rodlike diblock polyampholytes are studied using molecular dynamics simulations in a wide range of the grafting density .
simulations show linear dependence of the average thickness on the grafting density in all cases regardless of different flexibility of anchored chains and the brushes different equilibrium conformations .
slopes of fitted lines to the average thickness of the brushes of semiflexible and rodlike polyampholytes versus the grafting density are approximately the same and differ considerably from that of the brushes of flexible chains .
the average thickness of the brush of flexible diblock polyampholytes as a function of the grafting density is also obtained using a simple scaling analysis which is in good agreement with our simulations . |
the light flavor scalar mesons present a remarkable exception for the naive quark model , and the structures of those mesons have not been unambiguously determined yet .
the numerous candidates with @xmath1 below @xmath2 can not be accommodated in one @xmath3 nonet , some are supposed to be glueballs , molecular states and tetraquark states ( or their special superpositions ) @xcite .
the @xmath4 and @xmath5 are good candidates for the @xmath6 molecular states @xcite , however , their cousins @xmath7 and @xmath8 lie considerably higher than the corresponding thresholds , it is difficult to identify them as the @xmath9 and @xmath10 molecular states , respectively
. there may be different dynamics which dominate the @xmath11 mesons below and above @xmath12 respectively , and result in two scalar nonets below @xmath13 @xcite .
the strong attractions between the diquark states @xmath14 and @xmath15 in relative @xmath16-wave may result in a nonet tetraquark states manifest below @xmath12 , while the conventional @xmath17 @xmath3 nonet have masses about @xmath18 , and the well established @xmath19 and @xmath20 @xmath3 nonets with @xmath21 and @xmath22 respectively lie in the same region . furthermore , there are enough candidates for the @xmath17 @xmath3 nonet mesons , @xmath23 , @xmath24 , @xmath25 , @xmath26 and @xmath27 @xcite . in the tetraquark scenario ,
the structures of the nonet scalar mesons in the ideal mixing limit can be symbolically written as @xcite @xmath28 the four light isospin-@xmath29 @xmath30 resonances near @xmath31 , known as the @xmath8 mesons , have not been firmly established yet , there are still controversy about their existence due to the large width and nearby @xmath30 threshold @xcite . in general , we may expect constructing the tetraquark currents and studying the nonet scalar mesons below @xmath12 as the tetraquark states with the qcd sum rules @xcite .
for the conventional mesons and baryons , the `` single - pole + continuum states '' model works well in representing the phenomenological spectral densities , the continuum states are usually approximated by the contributions from the asymptotic quarks and gluons , the borel windows are rather large and reliable qcd sum rules can be obtained . however , for the light flavor multiquark states , we can not obtain a borel window to satisfy the two criteria ( pole dominance and convergence of the operator product expansion ) of the qcd sum rules @xcite . in ref.@xcite , t. v. brito et al take the quarks as the basic quantum fields , and study the scalar mesons @xmath7 , @xmath8 , @xmath5 and @xmath4 as the diquak - antidiquark states with the qcd sum rules , and can not obtain borel windows to satisfy the two criteria , and resort to a compromise between the two criteria .
for the heavy tetraquark states and molecular states , the two criteria can be satisfied , but the borel windows are rather small @xcite .
we can take the colored diquarks as point particles and describe them as the basic scalar , pseudoscalar , vector , axial - vector and tensor fields respectively to overcome the embarrassment @xcite . in this article
, we construct the color singlet tetraquark currents with the scalar diquark fields , parameterize the nonperturbative effects with new vacuum condensates besides the gluon condensate , and perform the standard procedure of the qcd sum rules to study the nonet scalar mesons below @xmath12 .
the qcd sum rules are `` new '' because the interpolating currents are constructed from the basic diquark fields instead of the quark and gluon fields .
whether or not the colored diquarks can be taken as basic constituents is of great importance , because it provides a new spectroscopy for the mesons and baryons @xcite .
the article is arranged as follows : we derive the new qcd sum rules for the nonet scalar mesons in sect.2 ; in sect.3 , we present the numerical results and discussions ; and sect.4 is reserved for our conclusions .
in the following , we write down the interpolating currents for the nonet scalar mesons below @xmath12 , @xmath32 where @xmath33 the @xmath34 are color indices , the @xmath35 is the charge conjugation matrix , the @xmath36 , @xmath37 and @xmath38 are basic scalar diquark fields , while the @xmath39 , @xmath40 and @xmath41 are the corresponding scalar two - quark currents . in this article , we take the isospin limit for the @xmath42 and @xmath43 quarks , and denote the fields @xmath36 and @xmath37 as @xmath44 . for the general color antitriplet bilinear quark - quark fields @xmath45 , @xmath40 and @xmath41
are recovered . ] and color singlet bilinear quark - antiquark fields @xmath46 , where the flavor , color and spin indexes are not shown explicitly for simplicity , we can project them into a local and a nonlocal part , after bosonization , the two parts are translated into a basic quantum field and a bound state amplitude , respectively , @xmath47 where the @xmath48 and @xmath49 denote the diquark and meson fields respectively , the @xmath50 and @xmath51 denote the corresponding bethe - salpeter amplitudes respectively @xcite . in ref.@xcite , we study the structures of the pseudoscalar mesons @xmath52 , @xmath53 and the scalar diquarks @xmath54 , @xmath55 , @xmath56 in the framework of the coupled rainbow schwinger - dyson equation and ladder bethe - salpeter equation using a confining effective potential , and observe that the dominant dirac spinor structure of the bethe - salpeter amplitudes of the scalar diquarks is @xmath57 . if we take the local limit for the nonlocal bethe - salpeter amplitudes , the dimension-1 scalar diquark fields @xmath54 , @xmath55 and @xmath56 are proportional to the dimension-3 scalar two - quark currents @xmath58 , @xmath59 and @xmath60 , respectively . a dimension-1 quantity @xmath61 can be introduced to represent the hadronization @xmath62 , @xmath63 and @xmath64 .
the attractive interaction of one - gluon exchange favors formation of the diquarks in color antitriplet @xmath65 , flavor antitriplet @xmath66 and spin singlet @xmath67 @xcite .
lattice qcd studies of the light flavors indicate that the strong attraction in the scalar diquark channels favors the formation of good diquarks , the weaker attraction ( the quark - quark correlation is rather weak ) in the axial - vector diquark channels maybe form bad diquarks , the energy gap between the axial - vector and scalar diquarks is about @xmath68 of the @xmath69-nucleon mass splitting , i.e. @xmath70 @xcite , which is also expected from the hypersplitting color - spin interaction @xmath71 @xcite . on the other hand ,
the studies based on the random instanton liquid model indicate that the instanton induced quark - quark interactions are weakly repulsive in the vector and axial - vector channels , strongly repulsive in the pseudoscalar channel , and strongly attractive in the scalar and tensor channels @xcite .
so it is sensible to use the scalar diquark fields to construct the tetraquark currents .
the two - point correlation functions @xmath72 can be written as @xmath73 denotes @xmath74 , @xmath75 , @xmath76 and @xmath77 .
we can insert a complete set of intermediate hadronic states with the same quantum numbers as the current operators @xmath78 into the correlation functions @xmath79 to obtain the hadronic representation @xcite . isolating the ground state contributions from the pole terms of the nonet scalar mesons
, we obtain the results , @xmath80 where the @xmath81 are the ground state masses , the @xmath82 are corresponding pole residues defined by @xmath83 , the thresholds @xmath84 in the channels @xmath85 , @xmath8 , @xmath7 respectively , the @xmath86 are the thresholds for the higher resonances and continuum states @xmath87 , and the @xmath88 are the corresponding hadronic spectral densities . we introduce the following new lagrangian @xmath89 , @xmath90 where @xmath91 , and carry out the operator product expansion with the basic diquark fields @xmath92 , @xmath93 and @xmath16 instead of the quark fields @xmath42 , @xmath43 and @xmath94 , and we have neglected the terms concerning the heavy diquark fields in the lagrangian . in the qcd , the basic quantum fields are the quark and gluon fields , the attractive interactions in the color antitriplet @xmath65 , flavor antitriplet @xmath95 and spin singlet @xmath67 quark - quark channels favor formation of the scalar diquarks , we can absorb some effects of the quark - gluon interactions into the effective diquark masses , which are characterized by the correlation length @xmath96 . at the distance @xmath97 , the @xmath98 diquark state combines with one quark or one @xmath99 antidiquark to form a baryon state or a tetraquark state , while at the distance @xmath100 , the @xmath98 diquark states dissociate into asymptotic quarks and gluons gradually ,
the strength of the quark - quark correlations is very weak .
just like the quarks , the diquarks have three colors , and can be gauged with the same @xmath0 color group to embody the residual quark - gluon interactions . in calculations ,
we take into account all diagrams like the typical ones shown in fig.1 , introduce new vacuum diquark condensates @xmath101 and @xmath102 besides the gluon condensate to parameterize the nonperturbative qcd vacuum , and consider the vacuum condensates up to dimension four . in the qcd sum rules ,
the high dimensional vacuum condensates are usually suppressed by large denominators or additional powers of the inverse borel parameters @xmath103 , the net contributions are very small . for example
, in the present case the contributions of the dimension-6 vacuum condensates can be estimated as @xmath104 , which is a tiny quantity .
if additional suppressions originate from the denominators are taken into account , the contributions are even smaller , and can be safely neglected .
once the analytical results are obtained , then we can take the dualities below the thresholds @xmath105 and perform the borel transform with respect to the variable @xmath106 , finally we obtain the following sum rules , @xmath107 where the @xmath108 denote the qcd spectral densities @xmath109 , @xmath110 and @xmath111 , @xmath112\delta(s-\widehat{m}_{q / s}^2)\ , , \end{aligned}\ ] ] @xmath113 , @xmath114 , @xmath115 in the channels @xmath85 , @xmath8 , @xmath7 respectively , and the @xmath116 is the borel parameter .
the threshold parameters @xmath117 are different from the corresponding @xmath118 in eq.(6 ) , because we absorb some qcd interactions into the effective diquark masses .
differentiate eq.(8 ) with respect to @xmath103 , then eliminate the pole residues @xmath119 , we can obtain the sum rules for the masses of the nonet scalar mesons , @xmath120
we estimate the vacuum diquark condensates @xmath101 and @xmath102 with the assumption of the vacuum saturation , which works well in the large @xmath121 limit , @xmath122 where the @xmath61 is a quantity has dimension of mass and can be taken as the confinement energy scale @xmath123 . at the energy scale @xmath124 , @xmath125 , @xmath126 , @xmath127 @xcite , and @xmath128 .
the quark condensates play a special role being responsible for the spontaneous breaking of the chiral symmetry , and relate with the masses and decay constants of the light pseudoscalar mesons through the gell - mann - oakes - renner relation @xcite .
the values of other vacuum condensates , such as the mixed condensates , the four quark condensates and the gluon condensates , can not be obtained from the first principles , we usually calculate them with the lattice qcd , the instanton models , or determine them empirically by fitting the qcd sum rules to the experimental data . in this article
, we introduce the diquark condensates @xmath101 and @xmath102 to parameterize the nonperturbative qcd vacuum , and assume that they relate with the four quark condensates ( therefore they have implicit relations with the spontaneous breaking of the chiral symmetry ) , and take the dimension - one parameter @xmath61 to be the confinement energy scale , as the scalar mesons @xmath5 , @xmath4 , @xmath8 and @xmath7 are bound states which consist of the confined quarks and gluons , and some parameters are needed to embody the confinement . on the other hand , we can understand the parameter @xmath123 as a fitted value , which happens to be the confinement energy scale , the crude estimation works well .
we take the updated values of the diquark masses from the qcd sum rules for consistency , where the interpolating currents @xmath39 , @xmath40 and @xmath41 are used @xcite , @xmath129 and @xmath130 ; the scalar diquarks were originally studied with the qcd sum rules about twenty years ago @xcite .
there have been several theoretical approaches to estimate the diquark masses , for example , the simple constituent diquark mass plus hyperfine spin - spin interaction model @xcite .
the @xmath5 and @xmath4 are well established , and the existence of the @xmath7 meson is confirmed , although there are controversy about its mass and width , the values listed in the review of particle physics are @xmath131 and @xmath132 respectively @xcite .
as far as the @xmath8 are concerned , there are still controversy about their existence , we take them as the @xmath16-wave isospin-@xmath29 @xmath30 resonance with the breit - wigner mass about @xmath133 . the e791 collaboration observed a low - mass scalar @xmath30 resonance with the breit - wigner mass @xmath134 and width @xmath135 respectively in the decay @xmath136 @xcite , and the bes collaboration observed a clear low mass enhancement in the invariant @xmath30 mass distribution in the decay @xmath137 with the breit - wigner mass @xmath138 and width @xmath139 , respectively @xcite .
recently , the bes collaboration reported the charged @xmath8 in the decay @xmath140 with the breit - wigner mass @xmath141 and width @xmath142 , respectively @xcite .
it is sensible to estimate @xmath143 . on the other hand ,
the qcd sum rules for the tetraquark states indicate that @xmath144 and @xmath145 @xcite . assuming the energy gap between the ground and first radial excited tetraquark states is about @xmath146
, we can tentatively determine the threshold parameters @xmath147 , @xmath148 , and @xmath149 .
the convergence behavior of the operator product expansion is very good , the contributions from the different terms have the hierarchy : perturbative - term @xmath150 . in calculation , we take uniform minimum value for the borel parameters @xmath151 .
the perturbative continuum @xmath152 is suppressed by the factor @xmath153 , the contributions from the pole terms are very large , see fig.2 . in this article , we take uniform maximum value for the borel parameters , @xmath154 , the contributions from the pole terms are about @xmath155 , @xmath156 and @xmath157 in the channels @xmath158 , @xmath8 and @xmath7 , respectively .
the two criteria ( pole dominance and convergence of the operator product expansion ) of the qcd sum rules are well satisfied .
if the conventional quark currents are chosen , the multiquark states i.e. tetraquark states , pentaquark states , hexaquark states , etc , have the spectral densities @xmath159 with the largest @xmath160 , the integral @xmath161 converges slowly @xcite .
if one do nt want to release the criterion of pole dominance , we have to either postpone the threshold parameter @xmath105 to very large value or choose very small value for the borel parameter @xmath162 .
with large value of the threshold parameter @xmath105 , for example , @xmath163 , here @xmath164 stands for the ground state , the contributions from the high resonance states and continuum states are included in , we can not use single - pole ( or ground state ) approximation for the spectral densities ; on the other hand , with very small value of the borel parameter @xmath162 , the borel window @xmath165 shrinks to zero or very small values . the numerical values of the masses and pole residues are presented in table 1 and figs.3 - 4 . from table 1
, we can see that the present predictions are compatible with ( or not in conflict with ) the experimental data @xcite and theoretical estimations @xcite .
the scalar tetraquark currents @xmath166 , @xmath76 and @xmath77 maybe have non - vanishing couplings with the scattering states @xmath9 , @xmath167 , @xmath30 , @xmath168 , @xmath169 , @xmath170 , etc , for example , @xmath171 if the couplings denoted by the @xmath172 and @xmath173 are strong enough , the contaminations from the continuum states are expected to be large . in the following , we study the contributions of the intermediate pseudoscalar meson loops to the correlation function @xmath174 in details as an example , @xmath175 where @xmath176\left [ ( p - q)^2-m_{\pi}^2\right ] } \ , , \nonumber\\ \sigma_{kk}(p)&=&\int~{d^4q\over(2\pi)^4}\frac{1}{\left [ q^2-m_{k}^2\right]\left [ ( p - q)^2-m_{k}^2\right ] } \ , , \end{aligned}\ ] ] the @xmath177 , @xmath178 are the strong coupling constants between the @xmath5 and the pseudoscalar meson pairs @xmath9 , @xmath167 respectively , and the @xmath179 , @xmath180 , @xmath181 , @xmath182 are the scattering amplitudes among the pseudoscalar meson pairs @xmath9 and @xmath167 . the couplings @xmath183 and @xmath173 are complicated functions of the @xmath184 , @xmath185 , @xmath186 , @xmath179 , @xmath187 , @xmath181 , @xmath188 , @xmath177 and @xmath178 , the explicit expressions are difficult to obtain .
we should bear in mind that the intermediate meson loops contribute a self - energy to the scalar meson @xmath5 , and therefore the scalar meson @xmath5 develops a breit - wigner width .
in fact , the scalar mesons @xmath85 , @xmath8 and @xmath7 below @xmath12 can be generated dynamically from the unitaried scattering amplitudes of the pseudoscalar mesons @xcite .
we can take into account those meson loops effectively by taking the following replacement for the hadronic spectral density , @xmath189 here we neglect the complicated renormalization procedure , take the physical values , and ignore the energy scale dependence of the mass @xmath184 and pole residue @xmath186 for simplicity ; the approximation works well .
the qcd sum rules for other scalar mesons are treated with the same routine . in ref.@xcite ,
dai et al perform the renormalization procedure in details to take into account the contributions from the continuum states .
the widths listed in the review of particle physics are @xmath190 , @xmath191 , @xmath192 and @xmath193 , respectively @xcite . taking into account the finite widths , we can obtain the modified masses from the corresponding qcd sum rules , which are shown in table 1 and fig.5 . in calculation
, we observe that the narrow width @xmath194 modifies the mass @xmath195 slightly and the effect can be neglected safely , while the broad widths @xmath196 and @xmath197 reduce the masses @xmath198 and @xmath199 about @xmath200 and @xmath201 , respectively .
comparing with the experimental data @xcite , the modified masses are better .
the scalar tetraquark currents @xmath166 , @xmath76 and @xmath77 maybe also have non - vanishing couplings with the higher resonances @xmath23 , @xmath24 , @xmath25 , @xmath26 and @xmath27 , which are supposed to be the @xmath202 @xmath3 states or glueballs , the couplings should be very small .
furthermore , the threshold parameters are @xmath203 , @xmath204 and @xmath205 in the channels @xmath85 , @xmath8 and @xmath7 , respectively , the contaminations should be very small . .
the masses and the pole residues of the nonet @xmath11 tetraquark states .
the star denotes the modified masses from the sum rules where the finite widths shown in the bracket are taken into account in the hadronic spectral densities . [ cols="^,^,^",options="header " , ] , @xmath206 and @xmath35 correspond to the channels @xmath158 , @xmath8 and @xmath7 respectively .
, width=302 ] , @xmath206 and @xmath35 correspond to the channels @xmath158 , @xmath8 and @xmath7 respectively .
, width=302 ] , @xmath206 and @xmath35 correspond to the channels @xmath158 , @xmath8 and @xmath7 respectively .
, width=302 ] , @xmath206 and @xmath35 correspond to the channels @xmath158 , @xmath8 and @xmath7 respectively .
the values in the bracket denote the finite widths in the hadronic spectral densities .
, width=302 ]
in this article , we take the scalar diquarks as the point particles and describe them as the basic quantum fields , then introduce the @xmath0 color gauge interactions , and construct the tetraquark currents which consist of the scalar fields to study the nonet scalar mesons as tetraquark states with the new qcd sum rules .
the numerical values are compatible with ( or not in conflict with ) the experimental data and theoretical estimations . comparing with the conventional quark currents , the diquark currents have the outstanding advantage to satisfy the two criteria of the qcd sum rules more easily
, the new sum rules can be extended to study other multiquark states .
this work is supported by national natural science foundation of china , grant number 11075053 , and the fundamental research funds for the central universities .
j. weinstein and n. isgur , phys .
* 48 * ( 1982 ) 659 ; f. e. close , n. isgur and s. kumana , nucl . phys .
* b389 * ( 1993 ) 513 ; n. n. achasov , v. v. gubin and v. i. shevchenko , phys . rev .
* d56 * ( 1997 ) 203 .
z. g. wang , eur .
j. * c62 * ( 2009 ) 375 ; z. g. wang , eur .
j. * c63 * ( 2009 ) 115 ; r. d. matheus , s. narison , m. nielsen and j. m. richard , phys .
* d75 * ( 2007 ) 014005 ; z. g. wang , eur . phys .
j. * c67 * ( 2010 ) 411 . | in this article , we take the scalar diquarks as point particles and describe them as basic quantum fields , then introduce the @xmath0 color gauge interaction and new vacuum condensates to study the nonet scalar mesons as tetraquark states with the qcd sum rules . comparing with the conventional quark currents , the diquark currents have the outstanding advantage to satisfy the two criteria of the qcd sum rules more easily . + zhi - gang wang + department of physics , north china electric power university , baoding 071003 , p. r. china pacs numbers : 12.38.lg ; 13.25.jx ; 14.40.cs * key words : * nonet scalar mesons , qcd sum rules |
chains of amino acids are important components of biological cells , and for such chains the specific ordering of the amino acids is often so fundamental to the resulting function and stability of the folded chain that if major deviations from the correct ordering were to occur , the final chain could fail to perform its requisite function within the cell , proving fatal to the organism . more specifically , we see the relevance of correct ordering in the study of protein structure , which is often divided into the protein folding and protein design problem . while the protein _ folding _ problem concerns finding the three - dimensional structure associated with a given amino acid sequence , the protein _ design _ problem ( also termed the inverse - folding problem ; see figure [ fig : design ] ) concerns finding the correct amino acid sequence associated with a given protein structure . + an aspect of one solution to the protein design problem is to maximize the energy difference between the low - energy folded native structure and the higher energy misfolded / denatured structures . in doing so
, one takes native structure as fixed and then determines the sequence yielding the minimum energy , under the assumption ( termed the `` fixed amino - acid composition '' assumption ) that only certain quantities of amino - acids appear in the chain @xcite . in this resolution ( specifically termed heteropolymer models @xcite @xcite )
the correct amino acid sequence is found by implementing an mc algorithm in sequence space given a certain fixed amino acid composition .
this entails assuming the number of various types of amino acids does not change , and distinct states in sequence space are permutations of one another .
for example , for a polypeptide chain with @xmath0 residues , rather than searching over the entire sequence space ( of size @xmath1 ) , one searches over a space of sequences ( of size @xmath2 ) which are defined by a fixed number of each amino acid .
this aspect of the protein design problem alerts one to a gap in the statistical mechanics literature .
namely , there do not seem to be any simple and analytically soluble statistical mechanics models where the space of states is defined by permutations of a list of components .
we can take steps toward constructing such a model by considering reasonable general properties it should have .
if we assume there was a specific sequence of components which defined the lowest energy sequence and was thermodynamically stable in the model , then deviations from this sequence would be less stable . because of the role sequences of molecules play in biological systems
, it is worth asking what features we expect such sequences to have from the perspective of modeling in statistical mechanics . in section
ii we introduce the model , and compute an exact partition function which displays what we term
quasi "- phase transitions a transition in which the sequence of lowest energy becomes entropically disfavored above a certain temperature . in section iii , we extend the previous model by adding a quadratic mean field interaction term and show that the resulting system displays two transition temperatures , a triple point , and a quadruple point . in section iv , we discuss various ways we can extend this model in theoretical or more phenomenological directions .
our larger goal is to study equilibrium thermodynamics for a system defined by permutations of a set of @xmath0 components where each unique permutation is defined by a specific energy . in general
, we should consider the case where the set of @xmath0 components consists of @xmath3 types of components for which if @xmath4 is the number of repeated components of type @xmath5 , then @xmath6 . for simplicity
, however , we will take @xmath7 for all @xmath5 so that each component is of a unique type and @xmath8 . to study the equilibrium thermodynamics of such a system with a fixed @xmath0 at a fixed temperature @xmath9
, we need to compute its partition function .
for example , for a sequence with @xmath0 components ( with no components repeated ) , there are @xmath10 microstates the system can occupy and assuming we label each state @xmath11 , and associate an energy @xmath12 with each state , then the partition function would be @xmath13 where @xmath14 for each state @xmath5 could be reasoned from a more precise microscopic theory of how the components interact with one another .
phenomenologically , eq . would be the most precise way to construct a model to study the equilibrium properties of permutations , but because it bears no clear mathematical structure , it is unenlightening from a theoretical perspective .
+ instead we will postulate a less precise , but theoretically more interesting model . for
most ordered chains in biological cells , there is a single sequence of components which is the
correct " sequence for a particular macrostructure .
deviations from this correct sequence are often disfavored because they form less stable macrostructures or they fail to perform the original function of the correct " sequence . with the general properties of such sequences in mind , we will abstractly represent our system as consisting of @xmath0 sites which are filled with particular coordinate values denoted by @xmath15 .
that is , we have an arbitrary but fixed coordinate vector @xmath16 expressed in component form as @xmath17 we will take the collection of components @xmath18 as intrinsic to our system , and thus take the state space of our system to be the set of all the vectors whose ordering of components can be obtained by permuting the components of @xmath16 , i.e. , all permutations of @xmath19 .
we represent an arbitrary state in this state space as @xmath20 , where the @xmath21 are drawn without repeat from @xmath18 .
formally , we would say our space of states is isomorphic to the symmetric group on @xmath16 ( @xcite ) .
we will thus denote our state space as @xmath22 and then an arbitrary state @xmath23 is just an element element of this set .
+ as a first formulation of the model , we will take @xmath24 ( the correct sequence ) to represent the zero energy state in the system , and for each component @xmath25 of an arbitrary vector @xmath23 which differs from the corresponding component @xmath26 in @xmath16 , there is an energy cost of @xmath27 .
the hamiltonian is then @xmath28 where @xmath29 and @xmath30 are components of vectors @xmath23 and @xmath16 respectively , and @xmath31 is defined by @xmath32 we note that although we label our general state as @xmath20 , the components @xmath33 can only take on mutually - exclusive values from the set @xmath34 . + we want to explore the equilibrium thermodynamics of a system with a hamiltonian of eq .. this amounts to calculating the partition function @xmath35 where @xmath36 is again the set of all permutations of the components of @xmath37 . to find a closed form expression for the partition function , we group terms in eq .
according to the number of ways to completely reorder @xmath38 components in @xmath16 while keeping the remaining components fixed .
each such reordering ( i.e. , each value of @xmath38 ) is associated with a sum over products of @xmath39 terms with @xmath38 factors of @xmath39 ( for various @xmath40 ) in each term .
the total partition function is a sum of all such reorderings for all @xmath38s from @xmath41 to @xmath0 . as can be seen from a direct expansion of eq . , we have @xmath42 where @xmath43 , termed the number of derangements of a list of @xmath38 ( @xcite ) , is the number of ways to completely reorder a list of @xmath38 elements .
the quantity @xmath44 , termed the @xmath38th elementary symmetric polynomial on @xmath45 ( @xcite ) , is the sum of all ways to multiply @xmath38 elements out of the @xmath0 term set @xmath46 .
for example , @xmath47 . by definition @xmath48_{q
= 0}.\ ] ] by the definition of the incomplete gamma function as @xmath49 and its relation to derangements ( i.e. , @xmath50 , see @xcite ) , we then find @xmath51 , \label{eq : partfunc001}\end{aligned}\ ] ] which is the desired closed - form expression for the partition function of this system .
+ with eq . , the problem of abstractly studying a thermal system of permutations with hamiltonian eq .
is , from the perspective of equilibrium statistical mechanics , now complete .
however , there are still some physical and theoretical results which can be teased from this formalism .
specifically , we can ask whether this system exhibits phase transitions . to answer this question ,
it would prove more analytically tractable to take @xmath52 for all @xmath40 . with this condition , our partition function simplifies to @xmath53 where we transformed our hamiltonian as @xmath54 with @xmath38 , defined as @xmath55 the number of components of @xmath23 which are not equal to the corresponding component in @xmath16 .
we call @xmath38 the number of _ incorrect _ components of @xmath23 , and if @xmath56 we say @xmath23 is _ completely disordered_. the factor @xmath57 in eq .
is defined as @xmath58 is the number of ways to reorder a list of @xmath0 elements so that @xmath38 elements are no longer in their original position .
this combinatorial definition of @xmath57 will prove useful when we explore the phase behavior of more complex models of permutations . from the form of eq .
, it is clear that , physically , its associated hamiltonian is not realistic as it places distinct permutations ( which in any true physical system most likely have quite different energy properties ) in the same degenerate energy state .
still , from a theoretical perspective , the simplicity of this model makes it a suitable starting point for studying the general properties of systems of permutations .
we can investigate the phase - like behavior of the system defined by the hamiltonian eq .
( for constant @xmath59 across @xmath40 ) , by applying steepest descent to eq . in the @xmath60 limit .
doing so , we find we can approximate the free energy of the system to be @xmath61 where @xmath62 is a @xmath63 independent constant . noting that @xmath64 , we find the average number of incorrect components satisfies the following equation of state : @xmath65 by eq . , we can infer that @xmath66 must be greater than or equal to 0 .
however , the right - hand - side of eq .
exhibits no such explicit constraint .
thus we can infer there is a phase - like transition in our system at the temperature @xmath67 below this temperature , we must have @xmath68 and thus the correct permutation " has the lowest free energy and is thermodynamically favored ; above this temperature , @xmath69 and the system is in a disordered phase where the previous lowest energy correct - permutation " is energetically disfavored .
+ interestingly , this transition arises from the naively non - interacting hamiltonian @xmath70 we say naively non - interacting " because eq .
consists of a sum over linear functions of a single index @xmath40 , and thus does not suggest any coupling between terms of differing index .
however , statistical mechanics tells us that the energy of a system is nt the only thing which determines the thermodynamic behavior of a system .
indeed we have to consider entropic contributions as well , and in this system the entropy ( as it is a function of @xmath38 ) can drive thermodynamic behavior . in other words , although the hamiltonian is depicted as non - interacting and can set - theoretically be represented as @xmath71 our system really exhibits interactions between components because our total space of states @xmath72 can not be factorized : @xmath73 thus the non - interacting " system exhibits a transition at eq . due to the coupled nature of the state space . as discussed in the subsequent section , we term this transition a quasi - phase transition " because it does not bear all of the standard properties we expect in phase transitions . , the landau free energy of the system as a function of @xmath38 , the number of incorrect components in @xmath23 , is always convex with a single global minimum . because @xmath74 , the @xmath75 domain of each plot ( dashed section ) is inaccessible . for sufficiently low temperatures ,
the minimum is at @xmath76 , but as we increase the temperature beyond eq .
, the free energy curve moves to the right ( but retains its functional form ) and the new minimum is at a @xmath77 value .
actual plots of eq . for
@xmath75 require us to replace the combinatorial argument of the logarithm with its corresponding gamma function expression . ]
we claim the system does not exhibit true phase transition behavior because many of these results are not consistent with the traditional thermodynamic definition of phase transitions .
for one , phase transitions are associated with divergences in the derivatives of the free energy , but there is no divergence in the free energy associated with the partition function eq . for possible parameter values .
also , the result eq
. arises from the steepest descent approximation which makes @xmath78 s temperature dependence near @xmath79 appear non - differentiable when by eq .
it is actually differentiable over its entire domain . finally , with eq .
we can define a landau free energy @xmath80 for this system according to @xmath81 , and what we may ordinarily label as a phase transition ( i.e. , going from @xmath82 to @xmath83 ) arises , not from changes in the functional form of the landau free energy as we see in real phase transitions , but from changes in the excluded region of the landau free energy ( see figure [ fig : perm_nonint_freeenergy ] ) . because the functional form of the free energy remains the same we observe no true phase transition .
alternatively , a heuristic argument for the non - existence of phase transitions in our permutation model is mathematically very similar to the landau argument ( @xcite ) for the non - existence of transitions in 1d ising models . for our permutation system with @xmath0 lattice sites ,
the state of zero energy and zero entropy consists of every site being occupied by its correct component . to increase the energy of this system
, we can choose @xmath38 sites to contain incorrect components , thus giving us an energy @xmath84 .
the number of ways we can choose these @xmath38 components is given by eq .
thus , upon introducing @xmath85 incorrect components , the change in the landau free energy of our system is @xmath86 \simeq j ( \lambda_0 - k_bt \ln n ) , \label{eq : freeen_0}\end{aligned}\ ] ] where we took these results in the @xmath87 limit and used @xmath88 . in the thermodynamic ( @xmath89 )
limit , we find that @xmath90 meaning there is no non - zero @xmath9 at which the entropic contribution becomes subdominant to the energy .
thus the system exhibits no phase transition .
when we first considered a model of thermal permutations , we began with a hamiltonian where sites did not interact with one another and each had a site - dependent energy cost for being incorrectly occupied : @xmath91 more generally , we can consider hamiltonians with an arbitrary number of multiple - site interaction terms .
such a hamiltonian could be written as @xmath92 the first term in eq .
associates an energy cost of @xmath59 with incorrectly occupying the component at position @xmath40 .
the second term models interactions between sites where the correct ( or incorrect ) occupation of a single site determines the energy of another .
the exact values of @xmath59 and @xmath93 could be chosen to ensure the
correct " state ( @xmath94 for all @xmath40 ) is non - degenerate as in the non - interacting model .
the ellipses represent higher order interactions in this framework .
hamiltonians such as eq . should be more physically relevant as they would correspond to systems where the energy cost for deviating from the lowest energy permutation is not simply linear but could be represented as a tensor valued fitting function .
we can make progress in studying the thermodynamics of more general hamiltonians like eq . by first only considering first- and second - order interaction terms and taking the interactions to be constants : @xmath95 for all @xmath40 ; @xmath96 for all @xmath97 .
the factor of @xmath98 is chosen so that the second term matches the extensive scaling of the first term .
the partition function for such parameter selections is then @xmath99 where @xmath100 and @xmath101 are interaction parameters with units of energy .
we can also write this partition function in the eq .
basis as @xmath102 where @xmath57 is defined in eq .
and @xmath103 is the energy function for the system .
our goal is to analyze the quasi "- phase behavior of this system in a way analogous to our analysis for the non - interacting system .
to do so we begin with the landau free energy function @xmath104 alternative starting points for this derivation are presented in appendix [ app : alt_deriv ] .
our system is constitutively discrete , so it is not precisely correct to discuss our free energy in the language of analysis , but given our expression for eq .
we can map this system to a continuous one which bears the same thermodynamic properties and for which analysis is appropriate . specifically , if we take @xmath38 to be continuous and use the identity @xmath105 , we can write @xmath106 with the approximation @xmath107 and the substitution eq . , eq
. then becomes @xmath108 where we defined our approximated free energy as @xmath109 and collected the @xmath38 independent constants into @xmath110 .
now eq . is fully continuous and amenable to analysis . to find the thermodynamic equilibrium of this system , we need to find the value of @xmath38 for which @xmath111 and @xmath112 .
for the first condition we have @xmath113 for @xmath114 we have @xmath115 as can be affirmed by taylor expansion or plots of each side . since the argument of @xmath116 is bounded below by 1 , the approximation in eq
can be applied to eq .. with this substitution , and setting the result to be valid for the equilibrium value @xmath117 , we then find the constraint @xmath118 which has the solution @xmath119 where @xmath120 is the ( branch unspecified ) lambert w function , defined by @xmath121 to specify the branch of the @xmath120 which corresponds to a stable equilibrium we compute the second derivative of our free energy at this derived critical point .
doing so yields @xmath122 thus eq .
( for @xmath123 ) yields a free energy minimum for @xmath124 and yields a maximum for the inverse condition .
this amounts to stating that the stable equilibrium for @xmath125 is defined by the principal branch of the lambert w function where @xmath126 , and the unstable equilibrium for @xmath125 is defined by the negative branch where @xmath127 .
thus , the order parameter for this system is @xmath128 we note that taking @xmath129 and using @xmath130 for @xmath131 returns us to the non - interacting result eq .. for completeness , we define the value of @xmath38 which yields a free energy maximum as @xmath132 ; it is related to eq . by replacing the principal branch function @xmath133 with @xmath134 . in the previous section
, we found that the order parameter for this system was given by eq .. we noted that this solution represented a local minimum of the free energy as long as the lambert w function satisfied @xmath135 .
thus when this condition is violated , @xmath136 is no longer a valid stable equilibrium , and our system has undergone a quasi
"- phase transition or simply a transition .
moreover , our values of @xmath38 are bounded below by @xmath76 and bounded above by @xmath137 , neither conditions of which are naturally constrained by eq .. thus these two conditions are associated with two other transitions . in all , then , there are three conditions which define the quasi - phase boundaries in this system .
while there are three conditions which define transitions in this system , there are in fact five distinct regimes of parameter space .
we can obtain a qualitative sense of these regimes by creating schematic plots of the free energy eq .
for various parameter values of @xmath100 and @xmath101 .
the possible plots can be placed into five categories according to the plot s stable or metastable @xmath38 values .
we depict these possible plots in figure [ fig : free_en ] .
we note that only the free energy plots with valid values of @xmath136 contain what we normally consider a thermodynamic equilibrium ; the other plots have stable " values of @xmath38 arising only from the @xmath76 and/or @xmath56 boundary conditions .
qualitatively , we can name the states according to the sequence space to which their equilibrium values of @xmath38 correspond .
we know for @xmath76 , our system is in a state with zero incorrect components in @xmath23 and hence the system is perfectly ordered `` or just ' ' ordered " .
conversely for @xmath56 our system has @xmath0 incorrect components and hence the system is completely disordered " or just disordered " .
the in - between case of @xmath117 where @xmath138 can be given the related label of partially - ordered " .
thus , the regime names associated with our possible values of the order parameters are * * ordered regime * * ( @xmath139 stable ) * : neither @xmath140 or @xmath132 exist ; @xmath141 .
* * disordered regime * * ( @xmath142 stable ) * : neither @xmath140 or @xmath132 exist ; @xmath143 . * * partially - ordered regime * * ( @xmath144 stable ) * : only @xmath136 exists . * * order and disorder metastable regime * * ( @xmath145 and @xmath146 stable ) * : only @xmath132 exists . * * order and partial - order metastable regime * * ( @xmath147 and @xmath148 stable ) * : both @xmath136 and @xmath132 exist .
we note that it seems to be a fundamental feature ( or a lack of one ) of this system , that the free energy eq .
does not admit a metastability between partial - order and disorder . with these regime definitions
, we can depict the parameter space graphically . in figure
[ fig : free_en ] we showed the possible forms of the free energy for this system where each was categorized according to the existence of the local minimum critical point @xmath136 , the existence of the local maximum critical point @xmath132 , and the sign of the quantity @xmath109 .
we can extrapolate this categorization to @xmath149 parameter space , by determining which regions of parameter space correspond to specific plots in eq .. doing so through the monte carlo procedure described in appendix [ app : mcpro ] , we generated 10,000 points of the parameter space diagram in figure [ fig : phase_diagram5 ] for @xmath150 set to 1 .
we note that the parameter space exhibits five regimes separated by three lines cited in table [ tab : table1 ] each of which correspond to the three conditions mentioned at the beginning of this section .
these lines can be derived analytically ( as shown in appendix [ app : phase_bound ] ) by considering the conditions in turn and which regimes they serve to connect .
.[tab : table1]functions defining boundaries between parameter regimes . [ cols= "
< , < " , ] from figure [ fig : phase_diagram5 ] we see that our system is characterized by two points where there is a coexistence between multiple regimes
. given that @xmath151 , we have that the ordered , partially - ordered , and order - partially - ordered metastability coexistence point is characterized by the condition @xmath152 these conditions characterize the system s triple point .
similarly , for @xmath60 , the partially - ordered , disordered , order - partially - ordered metastability , and order - disorder metastability coexistence point is characterized by the condition @xmath153 this condition characterizes the quadruple point of the system .
figure [ fig : phase_diagram5 ] also depicts the possible regimes of our system for a given temperature and various hamiltonian parameters @xmath100 and @xmath101 .
more physically , we may be interested in knowing what are the
quasi " phase properties of a system with a fixed @xmath100 and @xmath101 and a variable temperature .
that is , what are the temperatures which define the various transitions between regimes in the system ?
an arbitrary permutation system for a fixed @xmath0 and at a variable temperature is characterized by a specific energy function eq .. such a system is therefore defined by a specific @xmath100 and @xmath101 , and the system can be associated with a particular point ( and hence region ) in the parameter space of figure [ fig : phase_diagram5 ] .
as we vary the temperature of this system , the temperature dependent regime - coexistence lines change and if they change in such a way as to extend the region of a regime to newly encompass our original point then our system has undergone a transition . in this way
, we can define the temperatures which characterize various possible transitions of this system .
first , from figure [ fig : phase_diagram5 ] and eq .
we note the regime - coexistence line between the partially - ordered and disordered regime is independent of temperature , and so there is no critical temperature defining a partial - order to disorder transition . from eq .
, we can infer that the partial - order to order transition is characterized by moving below the temperature @xmath154 contingent on which region of parameter space the system lies , this temperature also characterizes the disorder to order - disorder metastability transition and the partial - order to order - partial - order metastability transition . and from eq .
, we can solve for the associated transition temperature given fixed @xmath100 and @xmath101 to find @xmath155^{-1}\ ] ] where this expression is only relevant for @xmath156 and @xmath157 . moving above
this temperature leads to the order to order - partial - order metastability transition .
in this work , motivated by an abstraction of a foundational problem in protein design , we posited and analyzed the basic properties of a statistical physics model of permutations .
formally , we considered a simple statistical physics model where the space of states for @xmath0 lattice sites was isomorphic to the symmetric group of degree @xmath0 @xcite , and where the energy of each permutation was a function of how much the permutation deviates from the identity permutation . in this model , we found that due to a state space which could not be factorized in a basis defined by lattice sites , even the superficially non - interacting system can exhibit phase - like transitions , i.e. , temperature dependent changes in the value of the order parameter which do note exhibit the properties typically associated which phase transitions in infinite systems .
when interactions are introduced through a quadratic mean field term , the system is capable of exhibiting five regimes of thermal behavior , and is characterized by two transition - temperatures corresponding to various quasi - phase transitions .
the introduced model provides us with a basic exactly soluble system for certain interaction assumptions and thus provides a concrete model - based understanding of a system with a non - factorizable state space .
because of its utility and the type of results obtained , the model deserves to be subject to the standard extensions of typical canonical models in statistical mechanics .
in particular we hope to extend it to non - trivial site dependent interactions .
for example , a nearest neighbor interaction hamiltonian of the kind which characterize the ising model , @xmath158 would be an alternative physical extreme to the mean - field interactions considered in section [ sec : three ] .
we could also consider a generalized chain of components where the interactions between sites or the cost for an incorrectly filled site is not constant but is drawn from a distribution of values .
such a system of quenched disorder would characterize a permutation glass which may contain interesting results due to the unique nature of the state space .
supposing it is possible to define more interesting interactions models , a natural investigation would concern the renormalization group properties of the system . specifically , we would be interested in how would one sum over specific states ( as characteristic of a renormalization group transformation ) when the state space of a system looks like , @xmath159 i.e. , is not factorizable along lattice sites . finally , to connect this model of permutations to problems more relevant to protein design it would prove necessary to incorporate the possibility of repeated components or the background geometry of a lattice chain .
mw thanks verena kaynig - fittkau for her unpublished work that inspired this investigation and amy gilson , abigail plummer , vinothan manoharan , and michael brenner for helpful discussions .
we re - derive eq . using the hubbard - stratonovich method .
we start this derivation assuming @xmath160 ; we will later see our resulting free energy can be analytically continued to the @xmath161 case . for @xmath162 ,
the partition function is @xmath163 then , applying the identity @xmath164 we have @xmath165 where @xmath166 . in a previous set of notes
, we found @xmath167 so eq .
becomes @xmath168 the function to which we apply steepest descent is then @xmath169 computing the conditions for @xmath170 and @xmath171 we obtain , respectively , @xmath172 solving for @xmath173 in the first equation we have @xmath174 and with the second equation we have the condition @xmath175 substituting the second condition into the first yields @xmath176 or @xmath177 thus the solution for @xmath173 ( conveniently written ) can be expressed in terms of the lambert w function as @xmath178 for @xmath123 , we can employ the complex version of the hubbard - stratonovich identity @xmath179 working through an analogous steepest descent procedure , we find that the equilibrium value for @xmath173 is @xmath180 which could have been extrapolated from eq . by taking @xmath181 . from this expression for the equilibrium condition , and as an analogy with the non - interacting case , it turns out the order parameter in this case is not @xmath173 but rather @xmath182 .
we re - derive eq . using the gibbs - bogoliubov inequality@xcite .
the inequality is @xmath183 \leq f[{\cal h}_0 ] + \langle { \cal h } - { \cal h}_0\rangle_0.\ ] ] the hamiltonian which defines our system is @xmath184 our variational hamiltonian is instead @xmath185 from our study of the non - interacting case we know @xmath186 & = - \frac{1}{\beta } \ln z_{n}(\beta\lambda_0){\nonumber \\ } & = - \frac{1}{\beta } \ln \left\ { \int^{\infty}_{0 } ds\ , e^{-s } \left(1+(s-1)e^{-\beta\lambda_0}\right)^n\right\}.\end{aligned}\ ] ] we also define @xmath187 thus our variational equation becomes @xmath188 & \leq - \frac{1}{\beta } \ln z_{n}(\beta\lambda_0 ) + ( \lambda_1 - \lambda_0 ) \langle j \rangle_{0,n } + \frac{\lambda_2}{2n}\langle j^2 \rangle_{0,n}{\nonumber \\ } & \equiv f(\lambda_0)\end{aligned}\ ] ] differentiating @xmath189 with respect to @xmath190 allows us to compute the maximum of this quantity .
given @xmath191 , we then find @xmath192 which if we take to be zero at some @xmath193 gives us @xmath194 to compute these derivatives we make use of various identities .
first we note @xmath195 so @xmath196 this last equality implies @xmath197 and so eq . becomes @xmath198 \frac{\partial } { \partial \lambda_0}\langle j \rangle_{0,n}\big|_{\lambda_0 = { { \overline{\lambda}_0}}}{\nonumber \\ } & \quad - \frac{\lambda_2}{2n\beta}\frac{\partial^2\langle j \rangle_{0,n}}{\partial \lambda_0 ^ 2}\big|_{\lambda_0 = { { \overline{\lambda}_0 } } } \label{eq : lamb_eq2}\end{aligned}\ ] ] to compute these quantities we need to approximate the partition function for our variational system . using the method of steepest descent @xmath199 and so we have @xmath200 computing the relevant derivatives in eq .
we find @xmath201 ( - e^{\beta\lambda_0 } + { \cal o}(1/n ) ) { \nonumber \\ } & \quad - \frac{\lambda_2}{2n}(- e^{\beta\lambda_0 } + { \cal o}(1/n ) ) = 0\end{aligned}\ ] ] @xmath202e^{\beta\lambda_0 } + { \cal o}(1/n ) = 0\end{aligned}\ ] ] neglecting subleading terms of @xmath203 ( a choice only valid for @xmath204 ) , solving for @xmath205 above , and using eq .
we obtain the equilibrium constraint @xmath206 which when solved for @xmath207 gives us @xmath208 or , given our approximations and limiting expressions , the result @xmath209
to generate figure [ fig : phase_diagram5 ] , we implemented the following mc algorithm : 1 . uniformly sample two points for @xmath100 and @xmath101 separately from within a certain bounded domain .
2 . draw the free energy curve eq .
corresponding to the sampled values @xmath210 .
3 . label the curve according to which schematic curve in figure [ fig : free_en ] it corresponds ( i.e. , according to its @xmath136 , @xmath211 and @xmath212 properties ) .
4 . color the point to signify the label .
we repeated this procedure for @xmath213 points with @xmath214 .
the regime separation lines were included after the mc procedure from the analytic forms cited in the text .
or @xmath217 from a plot of @xmath218 for real @xmath219 , we see that @xmath220 if @xmath221 and @xmath222 if @xmath223 .
thus this order to partial - order transition is defined by the condition @xmath224 , or @xmath225 the regime boundary which separates the ordered regime from the order and partial - order metastability regime is defined by the condition @xmath226 .
this condition is where the @xmath136 and @xmath211 begin coexisting @xcite , thus creating the mutual existence of a local maxima and local minima in figure [ fig : free_en ] .
thus for this regime boundary we have the condition @xmath227 this condition is valid so long as the argument of @xmath120 satisfies @xmath228 this inequality can only possibly be satisfied for @xmath229 and if @xmath230 the right inequality is automatically true .
so , our transition condition is given by @xmath231 in the order phase we automatically have @xmath232 , so the lhs of eq .
is greater than or equal to @xmath233 .
moreover , since @xmath101 is exclusively negative , at @xmath234 , @xmath235 is at a maximum value of @xmath236 . for @xmath237 , the solution to eq .
is then @xmath238 the regime boundary which separates the partially ordered regime from the disordered regime is defined by the condition @xmath239 .
we set the maximum value of @xmath136 to @xmath240 rather than @xmath0 because eq .
is associated with a free energy which diverges at @xmath241 and this approximate result is thus only physical up to @xmath240 .
alternatively we could see the maximum condition @xmath242 as respecting the fact that eq .
is only valid up to @xmath243 .
for this regime boundary we have the condition @xmath244 or @xmath245 again , using the condition that @xmath246 if @xmath221 , we find that the critical condition for this transition is @xmath247 or @xmath248 11ifxundefined [ 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 _ _ , vol .
( , ) @noop _ _ , vol .
( , ) @noop _ _ , vol .
( , ) @noop ( ) @noop ( ) @noop _ _ ( , ) @noop ( ) | ordered chains ( such as chains of amino acids ) are ubiquitous in biological cells , and these chains perform specific functions contingent on the sequence of their components . using the existence and general properties of such sequences as a theoretical motivation , we study the statistical physics of systems whose state space is defined by the possible permutations of an ordered list , i.e. , the symmetric group , and whose energy is a function of how certain permutations deviate from some chosen correct ordering .
such a non - factorizable state space is quite different from the state spaces typically considered in statistical physics systems and consequently has novel behavior in systems with interacting and even non - interacting hamiltonians .
various parameter choices of a mean - field model reveal the system to contain five different physical regimes defined by two transition temperatures , a triple point , and a quadruple point .
finally , we conclude by discussing how the general analysis can be extended to state spaces with more complex combinatorial properties and to other standard questions of statistical mechanics models . |
our test uses satellites that are diametrically opposite each other around their host to quantify the incidence of rotating planar alignments .
the signature of coherent rotation is an enhancement in the number of anti - correlated satellites . using a small tolerance angle ( fig .
1a ) and a minimum velocity difference , samples can be generated with a higher probability of containing edge - on planar structures , if they are present .
we first test this method on a simple toy model , to show the expected behaviour to choices of the tolerance angle parameter @xmath7 ( fig .
1b ) : the contrast of the planar component is seen to decrease with increasing @xmath7 , suggesting that small values of @xmath7 should preferably be used for the tests . to construct a somewhat more realistic model , we select galaxies and their satellites from the millennium ii cosmological simulation , and reassign some of the satellites to planar structures .
the selection process for hosts and satellites is kept as close as possible to the selections applied to the observed sdss sample .
10 url # 1`#1`urlprefix[2]#2 [ 2][]#2 , & . _ _ * * , ( ) . , & .
_ _ * * , ( ) . , & .
_ _ * * , ( ) . & .
_ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
et al . _ .
_ _ * * , ( ) .
. _ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) . , &
_ _ * * , ( ) . , & .
_ _ * * , ( ) . , & .
_ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
_ et al . _ .
_ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
_ _ * * , ( ) . , , & .
_ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) . .
_ _ ( ) . , .
_ _ * * , ( ) . , , , & .
_ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
_ et al . _ . _
_ * * , ( ) . & .
_ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
_ et al . _ . _ _ ( ) . .
_ et al .
_ _ * * , ( ) . .
_ _ * * , ( ) .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. the millennium - ii simulation databases used in this paper and the web application providing online access to them were constructed as part of the activities of the german astrophysical virtual observatory ( gavo ) .
all authors assisted in the development and writing of the paper .
primarily contributed to the development of the test for planar alignments , while r.i .
implemented this test on the sdss galaxy catalogue .
the authors have no competing financial interests .
reprints and permissions information is available at www.nature.com/reprints .
correspondence and requests for materials should be addressed to n.i .
( [email protected] ) . * table 1 : * the redshift @xmath32 , positions ( right ascension and declination in degrees ) , absolute magnitudes @xmath33 and radial velocities @xmath34 ( in @xmath35 ) of the hosts ( superscript `` h '' ) and the satellites ( superscript `` s1 '' or `` s2 '' ) , for the sample selected with a tolerance angle of @xmath21 .
the final column lists the sum of the angular momentum of the stellar component of both satellites @xmath36 , multiplied by a sign flag @xmath31 , where @xmath37 implies that the pair have anti - correlated velocity and @xmath38 that the velocity is correlated .
[ cols= " > , > , > , > , > , > , > , > , > , > , > , > , > " , ] .
however , the simple toy model ( methods ) shows a decline of the ratio with increasing @xmath7 .
* c * , fraction of anti - correlated galaxy pairs as a function of the fraction of satellites in the rotating planar population ( using @xmath21 , the most significant peak in fig .
2c ) . in the absence of a planar component ,
equal numbers of correlated and anti - correlated satellites should be detected .
however , the ratio increases as expected as the planar component is made more significant . ] at @xmath39 , which we consider the maximum useful opening angle given the low number of satellite pairs in the sdss .
* c * , the significance ( in units of standard deviation ) of the excess of anti - correlated satellite pairs .
the most significant peak has significance @xmath22 at an opening angle of @xmath40 . ] ) that surround the hosts of the satellite pairs with anti - correlated velocities ( using @xmath24 ) .
each field is rotated so that the receding satellite lies on the positive abscissa .
a clear horizontal feature is found out to @xmath41 ; this result remains robust for various subsamples and parameter choices .
( the black disk shows a radius of @xmath1 ) . *
b * , the angular distribution of the galaxies in ( a ) , rejecting galaxies within @xmath42 .
the significance of the peaks for the @xmath43 samples are @xmath44 , respectively .
* c * , applying the same procedure to the region around sdss correlated pairs ( red line , using @xmath45 to build up better statistics ) shows minimal correlation , as does the environment around anti - correlated pairs in ms2 ( purple ) . ]
* methods *
the simple statistical test we have developed is devised to allow us to quantify the frequency of satellites belonging to disk - like structures .
we use primarily the distinctive property of a rotating disk - like structure that objects on opposing sides have anti - correlated velocities .
the expectation from observations of m31 is that any such structures are superposed on a `` contaminating '' population of `` normal '' satellites which appear , to first approximation , to have a spherically symmetric distribution around the host .
the presence of such a contaminating population , together with the fact that most galaxies beyond the local group have only a small number of satellites with well - measured velocities , means that at present we can only test for the alignments in a statistical manner on a sample of hosts .
thus the challenge is to devise a means to enhance the contrast of the putative disk over a potentially dominant spherical population . since our viewing direction on these distant systems
can not be special , on average , the median inclination of any disk - like structures will be @xmath46 ( if we define an edge - on configuration to have inclination @xmath47 ) .
this naturally suggests using a test that makes use of the resulting elongation .
however , we can bias a sample towards being more edge - on by selecting those systems with satellites that have radial velocities significantly different from their host galaxy .
( clearly face - on disk - like alignments will have zero velocity difference , as viewed along our line of sight ) . as sketched in figure 1a
, we consider systems consisting of a massive host galaxy harbouring at least one pair of satellites . picking each satellite in turn
, we determine whether another satellite lies on the opposite side of the host within a tolerance angle @xmath7 .
if both satellites possess a velocity that is significantly different from that of their host , the pair is retained for study .
motivated by the galaxy velocity uncertainties in the sdss , we selected this minimum velocity difference parameter to be @xmath48 in all calculations presented here ( the results are qualitatively very similar for @xmath49 ) . to explore how the method works ,
we constructed first a very simple test configuration , containing 50% of satellites in a spherical population , and 50% in a disk - like structure .
both structures were populated with satellites with uniform probability between @xmath50 and @xmath1 .
the satellites in the disk rotate at @xmath51 , independent of radius , while the spherical population has an isotropic velocity dispersion of @xmath52 .
this toy model is then viewed from a random direction , and two satellites are selected at random beyond a projected radius of @xmath50 . if the two satellites lie on opposite sides of the host to within the chosen tolerance angle , and if both satellites have a velocity difference with respect to their host of more than @xmath53 , the pair is retained , and we count whether the velocities are correlated or anti - correlated .
the procedure is repeated 2000 times for each tolerance angle value .
the open triangles in fig .
1b show the ratio of the number of anti - correlated to correlated pairs as a function of the tolerance angle @xmath7 .
as this simple model shows , we expect the highest contrast to be found at small tolerance angles .
the selection on diametrically opposite galaxies together with the minimum velocity criterion ensures that close to edge - on configurations are preferentially selected ; for this toy model , the average inclination angle for the @xmath54 sample is @xmath55 .
in order to explore further the reliability of our method to uncover genuine planar satellite alignments , we decided to construct artificial galaxy systems that we could run and test our algorithm on .
for this purpose the millennium ii simulation ( in particular with the semi - analytic modelling by guo et al.@xcite ) provides an ideal view of the expected distribution of galaxies and their satellites in a very large ( @xmath56 ) volume in a @xmath8cdm universe .
the catalogue lists the absolute r - band magnitudes , total galaxy masses , positions and velocities that are necessary for our comparison to observations . to create the random views of galaxies derived from the simulation we proceed as follows .
we first choose a random direction from which we will view the galaxies in the millennium ii volume .
a list of candidate host galaxies is generated by selecting those objects with absolute magnitudes in the range @xmath11 ( identical to the selection on the real sdss data presented in the main text ) .
we examine each of the candidate hosts in turn , placing the host to be studied at @xmath57 ( the mean velocity of the sdss sample ) , and then making sure that it appears isolated in projection with no brighter neighbour within @xmath58 , with velocity differences less than @xmath10 .
we make a list of all the neighbouring galaxies within a projected distance of @xmath9 and a velocity of @xmath10 that are at least one magnitude fainter than the host , but that are brighter than @xmath15 ; we will refer to these objects as `` satellites '' .
( we reject so - called `` orphan '' galaxies which are systems whose parent subhalo is no longer resolved in the millennium ii simulation but as figs 1b and 1c show , our results remain qualitatively identical if these objects are included ) . for each host
we then randomly draw a vector to define the normal to the planar population , and we go through the list of satellites randomly assigning them to the planar population , according to the desired planar component fraction that we wish to test for .
clearly , when testing for a planar fraction of zero , the millennium ii positions and velocities remain unaltered . for those satellites that are assigned thus to the planar component
, we keep the galactocentric distance that they had in the millennium ii simulation , but place them onto the plane with a random azimuthal angle .
the space velocities of the planar satellites are devised to give circular motions in the plane of the alignment , with the total velocity chosen from the circular velocity of a universal halo model@xcite of total mass given by the virial mass of the host .
a gaussian random velocity of @xmath59 ( a representative value for the sdss velocity errors ) is added to the radial velocity of the host and all its satellites .
having thus reordered the three - dimensional positions of some of the satellites , we filter the sample to keep those objects that lie at projected distances in the range @xmath60 , and that have velocity difference of less than @xmath18\kms$ ] with respect to the host ( extended data fig .
the brightest two satellites within the @xmath60 annulus are selected for study .
if the two satellites lie on opposite sides of the host to within the chosen tolerance angle , and if both satellites have a velocity difference with respect to their host of more than @xmath53 , the pair is retained , and we count whether the velocities are correlated or anti - correlated .
the entire process is then repeated for all other candidate hosts .
we rerun the entire procedure , selecting new initial viewing angles , as many times as necessary until a total of 2000 satellite pairs have been generated .
with the parameter selections detailed in the text , and setting @xmath48 , there are 380 galaxy systems in the sdss . using @xmath21 , we find @xmath61 pairs to have anti - correlated velocities , and @xmath62 to have correlated velocities , i.e. @xmath63% of all pairs are found with this tolerance angle . with the unaltered millennium ii simulation ( 0% in a disk )
we find 4.7% of pairs with @xmath21 ; this fraction rises to 4.9% with a 50% disk component , and to 6.7% with a 100% disk component .
this suggests that the fraction of satellites in a planar component within the sdss is greater than 50% , consistent with the estimate given in the text , but the simplicity of the disk model for the planar component prevents us from drawing strong conclusions from this comparison .
1 url # 1`#1`urlprefix[2]#2 [ 2][]#2 * extended data figure 1 * adopted velocity envelope relation .
dots mark the distance - velocity distribution of satellites in the ms2 simulation that surround isolated host galaxies of similar luminosity and mass to the milky - way@xcite .
the empirical envelope relation shown in red ( @xmath18\kms$ ] ) is used in our analysis as a means to reduce contamination from velocity outliers . | recent work has shown that both the milky way and the andromeda galaxies possess the unexpected property that their dwarf satellite galaxies are aligned in thin and kinematically coherent planar structures .
it is now important to evaluate the incidence of such planar structures in the larger galactic population , since the local group may not be a sufficiently representative environment .
here we report that the measurement of the velocity of pairs of diametrically opposed galaxy satellites provides a means to determine statistically the prevalence of kinematically coherent planar alignments . in the local universe ( redshift @xmath0 ) , we find that such satellite pairs out to a galactocentric distance of @xmath1 are preferentially anti - correlated in their velocities ( 99.994% confidence level ) , and that the distribution of galaxies in the larger scale environment ( beyond @xmath1 and up to @xmath2 ) is strongly elongated along the axis joining the inner satellite pair ( @xmath3 confidence ) .
our finding may indicate that co - rotating planes of satellites , similar to that seen around the andromeda galaxy , are ubiquitous in nature , while their coherent motion also suggests that they are a significant repository of angular momentum on @xmath4 scales .
lyce international des pontonniers , 1 rue des pontonniers , f-67000 strasbourg , france . observatoire astronomique de strasbourg , universit de strasbourg , cnrs , umr 7550 , 11 rue de luniversit , f-67000 strasbourg , france sydney institute for astronomy , school of physics , a28 , the university of sydney , nsw 2006 , australia the satellite galaxies of the milky way have long been known to be preferentially located close to a plane@xcite , but this observation could be dismissed as a mere coincidence . however , as faint galaxies were uncovered in the sloan digital sky survey ( sdss)@xcite , it became clear that our galaxy hosts a planar structure of satellites with a close to polar orientation@xcite .
the complications due to spatial incompleteness of satellite samples that complicate analyses in the milky way are largely alleviated when observing the next nearest giant galaxy , andromeda ( m31 ) .
the presence of a vast plane of co - rotating dwarf galaxies was recently detected in that galaxy thanks to new photometric@xcite and spectroscopic@xcite surveys of its halo .
a full 50% of the dwarf galaxies around m31 belong to this structure@xcite .
the satellites in the plane extend to @xmath5 , yet they display very small scatter in the direction perpendicular to the plane ( @xmath6 ) , and they also possess coherent kinematics , suggestive of common rotation about their host .
recent analyses have also uncovered possible galaxy alignments in the m81 system@xcite , and in the ngc 3109 association .
such satellite alignments may arise naturally if dwarf galaxies formed out of tidal debris left over by ancient galaxy mergers@xcite , but this scenario remains difficult to reconcile with the high dark matter content deduced for these objects@xcite .
although the reality of the planar structures in the local group is now firmly established , they may represent a fossil of the particular dynamical formation history of the milky way and andromeda systems@xcite .
it is therefore necessary to investigate more distant systems to ascertain the true significance of these local detections .
we devised a test ( methods ) to quantify the incidence of planar systems of satellites .
beyond a few mpc , reliable and accurate relative distance measurements are inaccessible ; this means that one has to deal with two - dimensional projections of galactic systems , possessing only the radial component of velocity .
we take the m31 system as a template for the search of satellite alignments , since its global structure and dynamics is currently best understood .
half of that system shows coherent rotation , which means that for orientations that are not exactly face - on , satellites on either side of the galaxy as seen on the sky will in general have opposite velocities relative to the host ( i.e. the velocities of the satellites will be anti - correlated ) .
this motivates the following simple detection method : for each satellite around a given host , we check whether it possesses a counterpart that is located on the opposite side of the galaxy to within a certain tolerance angle @xmath7 ( sketch in fig .
1a ) , and if it does , we determine whether the pair has correlated or anti - correlated velocities
. clearly , with circular orbits , no contamination and perfect data , all pairs will be anti - correlated if they all belong to co - rotating planes . as a control , we first applied this test to the large millennium ii simulation ( ms2 ) of structure formation and evolution@xcite , which reflects our best theories of galaxy formation in @xmath8 cold dark matter ( @xmath8cdm ) cosmology@xcite .
we find that diametrically opposite pairs of bright satellites selected from that simulation ( methods ) show no kinematic coherence , with roughly equal numbers of correlated and anti - correlated pairs for all @xmath7 ( filled circles in fig .
1b ) .
note however that with a very different satellite selection strategy , a slight preference for co - rotating satellites can be found@xcite . to analyse the behaviour of our statistic in the presence of a contaminating background
, we forced different fractions of satellites around m31-like hosts in ms2 ( methods ) to lie within a randomly - chosen rotating plane .
fig .
1c shows the anti - correlation of the satellite pairs as a function of the dominance of planar configurations ; evidently a measure in real galaxies of the fraction of anti - correlated pairs has the potential to reveal whether planar satellite alignments are common .
we therefore applied this test to the sdss , which currently gives the most complete view of the nearby universe . since we wish to investigate the environment around galaxies similar to the milky way and andromeda , we select from the `` nyu value - added galaxy catalog''@xcite isolated host galaxies ( no brighter neighbour within a distance of @xmath9 and within a velocity difference of @xmath10 ) with magnitudes in the range @xmath11 .
cosmological parameters from the planck mission are assumed@xcite . to ensure a clean sample of satellites , we select hosts up to a redshift @xmath12 ( beyond this , few faint satellites are detected ) , remove hosts closer than @xmath13 to avoid noisy measurements , and remove all galaxies with velocity uncertainties greater than @xmath14 .
the satellites themselves are any galaxy one magnitude or more fainter than the host , but brighter than @xmath15 , within the radial range @xmath16 ( again to be similar to the m31 analysis@xcite ) .
we further require that the satellites at projected distance @xmath17 lie within a velocity of @xmath18\kms$ ] ( see methods , extended data ) . as for the ms2 analysis ,
we retain only those satellites whose direction of motion with respect to their hosts is well - resolved ; since we imposed an upper velocity uncertainty of @xmath14 for both hosts and satellites , we require a minimum velocity difference of @xmath19 .
there are @xmath20 galaxy systems in the sdss that pass these requirements .
various choices for @xmath7 are examined in fig .
2 . as our toy model shows ( methods , fig .
1b ) , the highest contrast between anti - correlated and correlated satellite pairs should be found for small @xmath7 .
there is an inevitable trade - off between the number of satellite pairs that pass the selection criteria and the contamination fraction suffered by the sample .
strict selection gives good contrast , but poor statistics ; lenient selection gives poor contrast , but good statistics .
optimal significance will therefore lie at an intermediate tolerance angle , but given the unknown density and kinematic properties of both the normal and putative disk - like satellite populations , we believe the best strategy is to allow the data themselves to guide our choice of @xmath7 .
fig .
2a shows that with @xmath21 , 20 out of 22 pairs are anti - correlated , implying 99.994% ( @xmath22 ) significance ; these systems are listed in table 1 .
( with a less strict velocity cutoff of @xmath23 , 21 out of 23 pairs are anti - correlated ) .
high significance is found out to @xmath24 , although as expected in the presence of non - planar `` contaminants '' , the significance decreases with increasing @xmath7 . by comparison to the simulations where a disk population was added to ms2 ( fig .
1c ) , the observed ratio of anti - correlated to correlated pairs ( @xmath25 at 99% confidence ) at @xmath21 suggests that a fraction of @xmath26% of satellites reside in planes , although we stress that this constraint is weak as this disk model is simplistic .
thus we have found that the average giant galaxy in the sdss is consistent with our m31 template . while the sdss spectroscopic observing strategy produced certain spatial biasses@xcite
, it seems extremely improbable that such biasses could artificially cause an over - abundance of anti - correlated satellite pairs as found here .
fig .
3 highlights a possible correlation between the direction defined by the satellite alignment and the large - scale structure surrounding the hosts .
an elongated overdensity of galaxies appears aligned along the axis of the satellite pair , extending out to @xmath27 times the distance of the selected pair ( figs .
3a , b ) .
this is consistent with what we see in the local group , where the m31 satellite alignment points within @xmath28 of the milky way . although these filaments are much thicker than the planes around galaxies , it is possible that this reveals the influence of large scale structures on the dynamics of the smaller satellite system .
furthermore , in ms2 the larger - scale environment around anti - correlated pairs shows no strong preferential direction , and neither does the environment around sdss correlated pairs ( fig .
3c ) .
while it remains possible that the large - scale elongation of the galaxy distribution along the direction of the galaxy pairs is an artefact of the sdss target selection , this seems unlikely given the random orientation of the satellite pairs on the sky .
table 1 also lists the angular momentum of the satellite pairs , calculated using the projected distance , line of sight velocity and estimated stellar masses@xcite of these galaxies . for comparison
, the total angular momentum of our galaxy in stars is @xmath29 ( approximating it as an exponential disk@xcite ) .
thus the angular momentum contained in the stellar component of the aligned satellites we have identified ( mean of the @xmath21 sample : @xmath30 , where @xmath31 is a sign flag explained in table 1 ) is comparable to the angular momentum in a giant galaxy s stellar disk .
this suggests that these coherent structures make a significant contribution to the angular momentum budget on galaxy halo scales ( @xmath4 ) , although a better understanding of their incidence and physical properties is required to quantify their importance .
our tests were constructed using ms2 as a control sample to predict what should have been a priori expected in @xmath8cdm cosmology . just as this paradigm did not predict the planes observed in the local group@xcite , it did not a priori predict the velocity correlations presented here
. it should be noted , however , that ms2 contains only dark matter , and future large cosmological simulations that include detailed baryonic physics should be performed to see if the discrepancies can be alleviated . while it is as yet uncertain whether the pairs of satellites detected here actually form part of kinematically coherent planes , their velocity anti - correlation , alignment with larger scale structures , and high angular momentum are all unexpected properties of the universe that will require explanation . |
one of the formidable tasks in heavy - ion physics is to identify a precise understanding of the jet - medium dynamics , the jet - medium interactions , and the jet - energy loss formalism .
below , we study the influence of the details of the jet - medium coupling and the medium background on the simultaneous description of the nuclear modification factor ( @xmath1 ) and the high-@xmath0 elliptic flow ( @xmath2 ) measured at rhic and lhc @xcite for a radiative pqcd energy - loss ansatz @xcite .
we contrast media determined via the viscous hydrodynamic approach vish2 + 1 @xcite with the parton - cascade bamps @xcite as well as a jet - medium coupling depending on the collision energy with a jet - medium coupling influenced by the energy of the jet , the temperature of the medium and non - equilibrium effects around the phase transition .
besides this , we compare the jet - energy loss based on radiative pqcd @xcite with the hybrid ads energy - loss ansatz of ref .
we contrast the pion nuclear modification factor obtained via the radiative pqcd - energy loss @xcite and the hybrid ads energy - loss ansatz with a parton - jet nuclear modification factor that can be considered as an idealized lo jet @xmath1 at rhic and lhc energies .
the pqcd - based energy loss model studied is parametrized as @xcite @xmath3 with the jet - energy dependence @xmath4 , the path - length dependence @xmath5 , and the energy dependence @xmath6 . in the following ,
the jet - medium coupling @xmath7 will depend either on the collision energy @xmath8 or the energy of the jet and the temperature of the background medium considered @xmath9 .
the jet - energy loss fluctuations are distributed via @xmath10 , allowing for an easy interpolation between non - fluctuating ( @xmath11 ) , uniform dirac distributions and distributions increasingly skewed towards small @xmath12 .
the jets are spread according to a transverse initial profile specified by the bulk flow fields given by the vish2 + 1 and bamps backgrounds considered @xcite . on the other hand ,
the jet - energy loss of the hybrid ads energy - loss ansatz @xcite is based on falling strings @xcite where @xmath13 the initial jet energy is given by @xmath14 and the string stopping distance for quark and gluon jets is determined via @xmath15 with the jet - medium coupling @xmath16 for quarks and @xmath17 for gluons , including the respective casimir operators @xmath18 and @xmath19 .
this energy loss ansatz has been integrated into our existing model @xcite .
please note that ref .
@xcite uses natural units , @xmath20 . for a direct comparison , we quote our results below using a dimensionless coupling .
the main differences between the two energy - loss descriptions is the square - root dependence that leads to the formation of a bragg peak with the explosive burst of energy close to the end of the jet s evolution .
there have been discussions in literature @xcite on the impact of the bragg peak . in line with previous findings
@xcite we will show below that there is a difference between the hybrid ads energy - loss ansatz featuring a bragg peak and the pqcd model without a bragg peak , however , this difference is only marginal .
fig . [ fig01 ] shows the pion nuclear modification factor ( @xmath1 ) for central ( left panel ) and mid - central ( middle panel ) collisions at rhic ( black ) and lhc ( red ) as well as the high-@xmath0 elliptic flow ( @xmath2 ) for mid - central events ( right ) .
the measured data @xcite is compared to the pqcd - based energy loss of eq .
( [ eq1 ] ) with @xmath21 .
jet - energy loss fluctuations ( @xmath22 ) and the transverse expansion of the background flow ( @xmath23 ) are included , as well as a running jet - medium coupling that depends on the energy of the collision , @xmath8 .
[ fig01 ] demonstrates that there is a surprising similarity between the results that can not be expected a priori given the fact that the two background media are so different : while the hydrodynamic description of vish2 + 1 @xcite assumes an equilibrated system , the parton cascade bamps @xcite also includes non - equilibrium effects in the bulk medium evolution .
in addition , the figure exhibits the so - called high-@xmath0 @xmath2-problem @xcite : the high-@xmath0 elliptic flow below @xmath24 gev is about a factor of two below the measured data @xcite .
this effect has been discussed in literature @xcite and recently it has been suggested by cujet3.0 @xcite that a temperature and energy - dependent jet - medium coupling @xmath9 , which includes non - perturbative effects around the phase transition of @xmath25 mev , can overcome this problem .
this jet - medium coupling was derived from the dglv gluon number distribution @xcite and is given by the analytic formul @xmath26 it includes a running coupling @xmath27 with @xmath28 , the polyakov - loop suppression of the color - electric scattering @xcite via @xmath29 with pre - factors @xmath30 for quarks and gluons , and the polyakov loop @xmath31 as parametrized from lattice qcd , as well as an enhancement of scattering due to the magnetic monopoles near the critical temperature @xmath32 also derived from lattice qcd @xcite .
this temperature and energy - dependent jet - medium coupling shows an effective running as it decreases with temperature .
we included the above jet - medium coupling @xmath9 in our jet - energy loss approach @xcite .
the result is shown in fig .
[ fig02 ] , again for the hydrodynamic background vish2 + 1 ( solid lines ) and a medium determined via the parton cascade bamps ( dashed lines ) . for comparison
, we depict the results from cujet3.0 @xcite . as in fig .
[ fig01 ] , the ion nuclear modification factor is well described both at rhic and lhc . the high-@xmath0 elliptic flow , however , increases drastically below @xmath24 gev as compared to fig .
[ fig01 ] , especially for the bamps background which already includes non - equilibrium effects @xcite . finally , we compare results from the linear pqcd approach of eq .
( [ eq1 ] ) with the highly non - linear hybrid ads holographic model of jet - energy loss , see eq .
( [ eq2 ] ) .
we compare the pion nuclear modification factor and an idealized lo jet @xmath1 given by @xmath33 naturally , this lo jet @xmath1 represents a reconstructed jet with vanishing cone radius and is only a lower bound for the nlo jet @xmath1 with jet - cone radii @xmath34 .
[ fig03 ] shows this comparison at lhc ( left ) and rhic ( right ) energies for two different jet - medium couplings that are treated as constants : a larger one ( red ) fitted to the pion @xmath1 data ( dashed - dotted lines ) at rhic and a lower one ( blue ) fitted the pion @xmath1 data at lhc . to guide the eye
, we include the reconstructed jet @xmath1 from cms @xcite with @xmath35 in fig .
[ fig03 ] .
the solid blue lines for the lo jet @xmath1 in the left panels of fig .
[ fig03 ] lie in the same ballpark as the experimental data . fragmenting this result to pions ( dashed - dotted lines ) leads to an @xmath1 that reproduces the measured pion nuclear modification factor at lhc .
a straight extrapolation of this results to rhic energies shows that the lo jet @xmath1 for _ the same _ jet medium couplings lie on top of the measured _ pion _ nuclear modification factor .
however , fragmenting this result to pions leads to a @xmath36 that is larger than the measured data at rhic .
larger jet - medium couplings ( red lines ) , on the other hand , describe the _ pion _ nuclear modification factor at rhic for the pqcd scenario and the lo jet @xmath1 at lhc is again close to the experimental data .
the pion nuclear modification factor at lhc , however , only touches the lower bound of present error bars . in case of the hybrid ads energy loss
the results always only touch the lower end of the experimental error bars .
[ fig03 ] demonstrates that the results for the pqcd and the hybrid ads energy - loss including a bragg peak are remarkably similar .
thus , unfortunately , neither the pion nor a lo jet @xmath1 are sensitive to the difference in the path - length between pqcd and ads models .
we compared the measured data on the nuclear modification factor for pions and reconstructed jets as well as on the high-@xmath0 elliptic flow at rhic and lhc energies to results obtained by a linear pqcd and a highly non - linear hybrid ads holographic model of jet - energy loss .
we found that the simultaneous description of the @xmath1 and @xmath2 requires a jet - medium coupling that depends on the energy of the jet , the temperature of the medium @xcite , and non - equilibrium effects around the phase transition .
we also contrasted a hydrodynamic background ( vish2 + 1 ) @xcite with a medium obtained from the parton cascade bamps @xcite and showed that the influence of the underlying bulk medium considered is suprisingly small .
unfortunately , neiter the pion nor the lo jet @xmath1 are sensitive to the difference in the path - length between pqcd and ads models .
this work was supported in part through the bundesministerium fr bildung und forschung , the helmholtz international centre for fair within the framework of the loewe program ( landesoffensive zur entwicklung wissenschaftlich - konomischer exzellenz ) launched by the state of hesse , the us - doe nuclear science grant no .
de - ac02 - 05ch11231 within the framework of the jet topical collaboration , and the us - doe nuclear science grant no.de-fg02-93er40764 .
numerical computations have been performed at the center for scientific computing ( csc ) . | the measured data on the nuclear modification factor for pions and reconstructed jets as well as on the high-@xmath0 elliptic flow at rhic and lhc energies are compared to results from a linear pqcd and a highly non - linear hybrid ads holographic model of jet - energy loss .
we find that the high-@xmath0 ellitic flow requires to include realistic medium transverse flow fields and a jet - medium coupling including the effects of the energy of the jet , the temperature of the bulk medium , and non - equilibrium effects close to the phase transition .
we extend our jet - energy loss model that is coupled to state - of - the - art hydrodynamic prescriptions to backgrounds generated by the parton cascade bamps .
we demonstrate that the results for the hydrodynamic and the parton - cascade backgrounds show a remarkable similarity .
unfortunately , the results for both the pion and a parton - jet nuclear modification factor are insensitive to the jet - path dependence of the models considered .
jet quenching , viscous hydrodynamics , transport model , jet holography |
the 4d ( non-)abelian 1-form gauge theories are at the heart of standard model ( sm ) of particle physics where there is a stunning degree of agreement between theory and experiment .
one of the weak links of sm is connected with the very existence of the esoteric higgs particle which is responsible for the mass generation of gauge bosons and fermions . in view of the fact that higgs particle has not yet been observed experimentally with a hundred percent certainty ,
other theoretical tools for the mass generation of gauge bosons ( in various dimensions of spacetime ) have become important and they have generated a renewed interest in the realm of theoretical physics . in the context of the above
, it may be mentioned that the 4d topologically massive ( non-)abelian gauge theories have been studied in the past [ 1 - 4 ] where there is merging and mixing of 1-form and 2-form ( non-)abelian gauge fields through the celebrated topological @xmath4 term . in such models
, it has been shown that the ( non-)abelian 1-form gauge field acquires a mass in a very natural fashion without taking any recourse to the higgs mechanism .
however , these models suffer from problems connected with renormalizability , consistency and unitarity .
we have studied [ 5 - 10 ] these models , within the framework of superfield and becchi - rouet - stora - tyutin ( brst ) formalisms in the hope that we would be able to propose a model that would be free of the drawbacks of the earlier models [ 1 - 4 ] .
however , it remains still an open problem to construct a 4d consistent , unitary and renormalizable non - abelian 2-form gauge theory ( where the 1-form and 2-form non - abelian gauge fields are incorporated together ) . in the above scenario ,
it is an interesting idea to propose and study some lower dimensional models which are free of the problems of 4d topologically massive theory and where mass and gauge - invariance co - exist together .
one such _ massive _ model , that has been a topic of theoretical interest , is the jackiw - pi ( jp ) model in three ( 2 + 1)-dimensions of spacetime where the non - abelian gauge invariance and parity are respected together due to the introduction of a 1-form vector field , endowed with a parity , that is opposite of the usual non - abelian 1-form vector field [ 11 ] .
in fact , the 3d gauge theories , in general , have been topic of theoretical research in the recent past because of the novel and attractive properties associated with them [ 12,13 ] .
furthermore , it has already been shown that , for sufficiently strong vector coupling , the gauge invariance does _ not _ necessarily imply the _ masslessness _ of gauge particles [ 14,15 ] . in the backdrop of the above statements ,
the jp model of 3d massive gauge theory has been studied from different theoretical angles .
for instance , the hamiltonian formulation and its constraint analysis have been carried out in [ 16 ] .
the jp model is also endowed with some interesting continuous symmetries . in this
context , mention can be made of the usual yang - mills ( ym ) symmetry transformation and a symmetry that is different ( i.e. nym ) from the ym .
the brst symmetry and corresponding slavnov - taylor identity of this model have also been recently found [ 17 ] .
however , the off - shell nilpotent and absolutely anticommuting anti - brst symmetry transformations of this model have not been discussed [ 17 ] which are essential for the completeness of the theory as their very existence is theoretically backed , guided and governed by the concept of mathematical objects called gerbes ( see , e.g. [ 18,19 ] for details ) .
the local and continuous gauge symmetry , generated by the first - class constraints of a given gauge theory , is generalized to the nilpotent brst and anti - brst symmetry transformations in the brst formalism .
the anti - brst symmetry is a new kind of symmetry transformation [ 20 ] that is satisfied by the yang - mills theory .
it has also been shown [ 21 ] that the anti - brst symmetry has not been a matter of choice rather it has real fundamental importance in providing necessary additional conditions for the ghost freedom that is essential for a consistent quantization .
both the nilpotent symmetries have been formulated in a completely model independent way in [ 22 ] . in our recent works [ 18,19 ] , we have demonstrated the relevance of gerbes in the context of brst formalism through the existence of cf - type restrictions .
we have claimed that the latter is the hallmark of a gauge theory within the framework of brst formalism .
thus , for the sake of completeness of the brst analysis of the jp model , it is essential to derive a proper anti - brst symmetry corresponding to the brst symmetry of [ 17 ] . the main motivation behind our present investigation is to derive the full set of proper [ i.e. off - shell nilpotent ( @xmath5 ) and absolutely anticommuting ( @xmath1 ) ] ( anti-)brst symmetry transformations @xmath2 corresponding to the usual ym gauge symmetry transformation for the jp model by exploiting the augmented " superfield approach to brst formalism [ 23 - 27 ] .
this geometrical approach leads to the derivation of ( anti-)brst invariant curci - ferrari ( cf ) condition [ 28 ] which ensures the absolute anticommutativity of @xmath2 and derivation of the coupled ( but equivalent ) lagrangian densities that respect the above ( anti-)brst symmetry transformations . in our present endeavor ,
we have purposely concentrated _ only _ on the usual ym gauge symmetries for the brst analysis within the framework of superfield formalism .
this is due to the fact that we plan to understand the jp model step - by - step so that we can gain deep insights into the key aspects of this model .
this understanding , perhaps , would enable us to propose an accurate model for the 4d theory and would make us confident about the limiting cases of the general brst analysis of this model where ym and nym gauge symmetries would be combined together for their thorough discussions ( within the framework of superfield formalism ) .
exactly similar kind of programme , we have followed for the brst analysis of the 4d dynamical non - abelian gauge theory [ 3,2 ] within the framework of superfield approach ( see , e.g. [ 10 ] for details ) .
the prime factors that are responsible for our present investigations are as follows .
first , the jp model is free of the problems encountered in the 4d topologically massive models with @xmath4 term .
second , this 3d model does not invoke any higher form gauge field ( like the 2-form @xmath6 field of 4d theory ) for the mass generation .
third , the understanding of this 3d theory might provide some insights that would show the correct path to construct a renormalizable and consistent 4d massive theory .
fourth , we study jp model within the framework of brst formalism where the renormalizability and unitarity could be proven with the help of slavnov - taylor identities and _ nilpotency _ of the _ conserved _ brst charge .
finally , the 3d jp model is very special model because it generates mass for the gauge field without violating the parity symmetry .
this feature is drastically different from the mass generation by incorporating the chern - simons term in the lagrangian density of 3d gauge theory where parity is violated .
the contents of our present investigation are organized as follows . in section 2 , we discuss two sets of local gauge symmetry transformations associated with the jp model .
our section 3 incorporates the derivation of ( anti-)brst symmetry transformations for the gauge field ( @xmath7 ) and ( anti-)ghost fields [ @xmath8 with the help of bonora - tonin s superfield formalism [ 23,24 ] . in section 4
, we deal with the derivation of ( anti-)brst symmetry transformations for the vector field @xmath9 and scalar field @xmath3 within the framework of augmented " superfield formalism [ 25 - 27 ] .
our section 5 is fully devoted to the derivation of coupled lagrangian densities that respect the above ( anti-)brst transformations .
we show the conservation of ( anti-)brst current ( and corresponding charges ) in section 6 .
our section 7 contains the discussion of ghost symmetry transformations and the derivation of the algebra satisfied by _ all _ the symmetry generators .
finally , we made a few concluding remarks in section 8 . in our appendix
a , we capture the off - shell nilpotency and anticommutativity of the ( anti-)brst charges and the brst invariance ( as well as equivalence ) of the coupled lagrangian densities within the framework of `` augmented '' superfield formalism .
we provide a geometrical interpretation for the existence of the ( anti-)brst invariant cf restriction [ 28 ] in our appendix b with a few clear and cogent physical arguments .
_ conventions and notations : _ we adopt here the conventions and notations such that the background spacetime minkowskian flat metric has the signature @xmath10 , totally antisymmetric levi - civita tensor @xmath11 satisfies @xmath12 etc . , and @xmath13 . in the above
, the greek indices @xmath14 correspond to the 3d time and space directions .
we take the dot and cross products @xmath15 in the @xmath16 lie algebraic space where the generators @xmath17 of the @xmath16 lie algebra satisfy the commutator @xmath18 = i f^{abc}\ , t^c$ ] with @xmath19 .
the structure constants @xmath20 are chosen to be totally antisymmetric in @xmath21 for the semi -
simple @xmath16 lie algebra [ 29 ] .
we begin with the lagrangian density of the three ( 2 + 1)-dimensional ( 3d ) _ massive _ non - abelian 1-form gauge theory proposed by jp where the @xmath16 yang - mills ( ym ) gauge invariance and parity are respected _
together_. furthermore , this theory respects a non - yang - mills gauge symmetry transformation as well .
the lagrangian density of the theory , in its full blaze of glory , is ( see , e.g. [ 11 ] ) @xmath22 where @xmath7 and @xmath9 are the vector fields with opposite parity , @xmath3 is a scalar field , @xmath23 is a coupling constant and @xmath24 is the mass parameter .
the 2-form [ @xmath25 curvature tensor @xmath26 , corresponding to the 1-form @xmath27 field @xmath7 , is derived from the maurer - cartan equation @xmath28 .
similarly , the field strength tensor @xmath29 , corresponding to the 1-form @xmath30 field @xmath9 , is obtained from @xmath31 \equiv \frac { 1}{2 ! } ( dx^\mu \wedge dx^\nu)\ , g_{\mu\nu } \cdot t$ ] where the covariant derivative is defined as : @xmath32 .
the above lagrangian density ( 1 ) respects the following usual local , continuous and infinitesimal yang - mills ( ym ) gauge transformations @xmath33 , as @xmath34 this theory also respects a non - yang - mills ( nym ) gauge transformation @xmath35 .
the infinitesimal version of this transformation is : @xmath36 in the above , @xmath37 and @xmath38 are the @xmath16 valued infinitesimal ( lorentz - scalar ) gauge parameters .
it is straightforward to check that the lagrangian density ( 1 ) transforms , under the local , continuous and infinitesimal transformations ( 2 ) and ( 3 ) , as @xmath39.\end{aligned}\ ] ] we note that the usual ym symmetry is a perfect symmetry for @xmath40 because we have perfect invariance ( i.e. @xmath41 ) .
however , the lagrangian density @xmath40 remains quasi - invariant under @xmath35 because it transforms to a total spacetime derivative . in our present investigation
, we shall concentrate only on the _ usual _ ym gauge symmetry transformations @xmath42 and derive the corresponding ( anti-)brst symmetry transformations that are off - shell nilpotent and absolutely anticommuting in nature .
we apply the well - known bonora - tonin s superfield approach [ 23,24 ] to derive the nilpotent ( anti-)brst symmetry transformations corresponding to the ym gauge transformations ( @xmath33 ) for the 1-form gauge field @xmath7 and ( anti-)ghost fields @xmath43 . in this approach ,
first of all , we generalize the 3d bosonic vector field @xmath44 and fermionic ( anti-)ghost fields @xmath45 to their corresponding superfields .
the latter are defined on the @xmath46-dimensional supermanifold parametrized by the superspace variables @xmath47 where @xmath48 are the spacetime variables and @xmath49 are a pair of grassmannian variables ( with @xmath50 ) .
these superfields can be expanded along the directions of the grassmannian directions @xmath51 and @xmath52 ( of the ( 3 , 2)-dimensional supermanifold ) , as [ 23,24 ] ; @xmath53 where the local secondary fields
@xmath54 $ ] are fermionic ( @xmath55 , etc . ) and @xmath56 $ ] are bosonic in nature .
these secondary fields can be determined in terms of the basic and auxiliary fields of the 3d local brst invariant quantum field theory through the application of the celebrated horizontality condition ( hc ) .
we note that the kinetic term @xmath57 , corresponding to the 1-form gauge field @xmath7 , remains invariant under the gauge transformations ( 2 ) .
the hc condition implies that the gauge invariant kinetic term remains invariant when we generalize the 3d ordinary non - abelian gauge theory onto ( 3 , 2)-dimensional supermanifold .
the above statement of the gauge invariance can be , mathematically , expressed as : @xmath58 where the super curvature @xmath59 , defined on the @xmath46-dimensional supermanifold , is derived from the maurer - cartan equation : @xmath60 . here
@xmath61 is the super exterior derivative ( with @xmath62 ) and @xmath63 is the super 1-form connection which are the generalizations of the ordinary exterior derivative @xmath64 and 1-form connection @xmath65 as @xmath66 where @xmath67 and @xmath68 are the superfields on the ( 3 , 2)-dimensional supermanifold and @xmath69 .
the celebrated hc condition ( 6 ) leads to the following relationships amongst the basic , auxiliary and secondary fields [ 23,24 ] @xmath70 where we have made the choices @xmath71 and @xmath72 which are , finally , identified with the nakanishi - lautrup type auxiliary fields of the 3d local quantum field theory .
it is to be noted that , to satisfy the hc ( 6 ) , one sets equal to zero the grassmannian components of the super tensor @xmath73 in super 2-form @xmath74 . the equation @xmath75 [ that is quoted in ( 8) ] is the curci - ferrari ( cf ) restriction which is one of the key _
hallmarks _ of the non - abelian 1-form gauge theory .
this condition is derived from hc when one sets equal to zero the @xmath76 component of the super curvature tensor @xmath73 .
the cf condition plays an important role in providing the proof for the absolute anticommutativity of the ( anti-)brst transformations .
furthermore , the cf condition is instrumental in obtaining a coupled set of lagrangian densities [ cf .
( 30 ) , ( 31 ) below ] that respect the ( anti-)brst transformations ( see , also appendix b ) . substituting the above relationships ( 8) into the super - expansions of the superfields [ cf .
( 5 ) ] , we obtain the following explicit expansions : @xmath77 ( x ) \nonumber\\ \fl & \equiv a_\mu ( x ) + \theta\ ; ( s_{ab } a_\mu ( x ) ) + \bar \theta\ ; ( s_b a_\mu ( x ) ) + \theta\ ; \bar\theta\ ; ( s_b s_{ab } \;a_\mu ( x ) ) \nonumber\\ \fl \qquad \tilde f^{(h ) } ( x , \theta , \bar\theta ) & = c ( x ) + \theta\ ; ( i \bar b ( x ) ) + \bar \theta\ ; \bigl [ \frac{g}{2}\ ; ( c \times c ) ( x ) \bigr ] + \theta \bar\theta \ ; [ -i g\;(\bar b \times c)(x ) ] \nonumber\\ \fl & \equiv c ( x ) + \theta\ ; ( s_{ab } c ( x ) ) + \bar \theta\ ; ( s_b c ( x ) ) + \theta\ ; \bar\theta\ ; ( s_b s_{ab } \;c ( x))\nonumber\\ \fl \qquad \tilde { \bar f}^{(h ) } ( x , \theta , \bar\theta ) & = \bar c ( x ) + \theta\ ; \bigl [ \frac{g}{2 } \;(\bar c \times \bar c ) ( x ) \bigr ] + \bar \theta\ ; ( i b ( x ) ) + \theta \bar\theta \ ; [ ( + ig\;(b \times \bar c ) ( x ) ] \nonumber\\ \fl & \equiv \bar c ( x ) + \theta\ ; ( s_{ab } \bar c ( x ) ) + \bar \theta\ ; ( s_b \bar c ( x ) ) + \theta\ ; \bar\theta\ ; ( s_b s_{ab}\ ; \bar c ( x))\end{aligned}\ ] ] where @xmath78 , as the superscript on the superfields , denotes the expansions of the superfields after the application of hc .
the super 2-form curvature tensor can be expressed as @xmath79(x ) -
\;\bar\theta \;[g \;(f_{\mu\nu } \times c)](x ) \nonumber\\ & + \theta\ ; \bar\theta\ ; \bigl [ g^2\;(f_{\mu\nu } \times c ) \times \bar c - ig \;(f_{\mu\nu } \times b ) \bigr ] ( x)\nonumber\\ & \equiv f_{\mu\nu}(x ) + \theta \;(s_{ab } f_{\mu\nu}(x ) ) + \bar\theta \;(s_b f_{\mu\nu}(x ) ) + \theta\ ; \bar\theta\ ; ( s_bs_{ab}\;f_{\mu\nu}(x)).\end{aligned}\ ] ] it is clear , from the above expressions , that the kinetic term of @xmath7 ( i.e. @xmath80 ) remains invariant ( i.e. independent of the grassmann variables @xmath81 ) under the application of hc . in other words , we obtain @xmath82 .
before we wrap up this section , we would like to state that the equations ( 9 ) and ( 10 ) imply that : @xmath83 .
this mapping establishes a relationship between the ( anti-)brst transformations and the translational generators along the grassmannian directions of the @xmath46-dimensional supermanifold .
this key relationship entails upon the ( anti-)brst transformations , emerging from our superfield formalism , to be _ always _ nilpotent of order two ( @xmath0 ) and absolutely anticommuting in nature because of the fact that @xmath84 and @xmath85 .
it can be checked that the composite fields @xmath86 and @xmath87 remain invariant under the usual ym gauge transformations @xmath42 because @xmath88 since @xmath86 and @xmath87 are the gauge - invariant quantities , therefore , these are physical quantities ( in some sense ) .
these quantities must remain unaffected by the presence of the grassmannian variables when the former entities are generalized onto the ( 3 , 2)-dimensional supermanifold .
thus , we have the following gauge invariant restrictions ( girs ) on the ( super)fields : @xmath89 in the above , the expansion for @xmath90 is quoted in equation ( 10 ) .
the bosonic superfields @xmath91 and @xmath92 can be , in general , expanded on the ( 3 , 2)-dimensional supermanifold along the grassmannian directions ( @xmath93 ) , as @xmath94 where the secondary fields @xmath95 are fermionic and @xmath96 are bosonic in nature .
these secondary fields can be determined with the help of girs [ cf .
in fact , the equality ( 12 ) , leads to the following relationships : @xmath97\nonumber\\ \fl \bar p_\mu = - g\,(\phi_\mu \times \bar c ) \qquad \bar p = - g\,(\rho \times \bar c ) \qquad q = - i\ , \bigl [ g^2\,(\rho \times c ) \times \bar c - i\ , g\,(\rho \times b ) \bigr ] .\end{aligned}\ ] ] substituting the above values into ( 13 ) , we obtain @xmath98 ( x ) - \bar\theta\ ; [ g\;(\phi_\mu \times c ) ] ( x ) \nonumber\\ & + \theta \;\bar\theta\ ; [ g^2(\phi_\mu \times
c ) \times \bar c - ig\ ; ( \phi_\mu \times b ) ] ( x)\nonumber\\ & \equiv \phi_\mu ( x ) + \theta \;(s_{ab}\ ; \phi_\mu ( x ) ) + \bar \theta \;(s_b \;\phi_\mu ( x ) ) + \theta \;\bar \theta \;(s_b \;s_{ab } \;\phi_\mu ( x))\nonumber\\ \fl \qquad \tilde \rho^{(h , g ) } ( x,\theta,\bar\theta ) & = \rho ( x ) - \theta \ ; [ g\;(\rho \times \bar c ) ] ( x ) - \bar\theta\ ; [ g\;(\rho \times c ) ] ( x ) \nonumber\\ & + \theta \;\bar\theta \;[g^2\;(\rho \times c ) \times \bar c - ig \;(\rho \times b ) ] ( x)\nonumber\\ & \equiv \rho ( x ) + \theta \;(s_{ab}\ ; \rho ( x ) ) + \bar \theta \;(s_b \;\rho ( x ) ) + \theta \;\bar \theta \;(s_b \;s_{ab } \;\rho ( x))\end{aligned}\ ] ] where the superscripts @xmath99 on the above superfields refers to the super - expansions of the superfields obtained after the application of celebrated hc condition and gauge invariant restrictions .
thus , we have already obtained the ( anti-)brst symmetry transformations in the above for the fields @xmath9 and @xmath3 in view of the mappings : @xmath100 . within the framework of superfield formalism
, we can also calculate the nilpotent ( anti-)brst transformations for the field strength tensor @xmath101 and composite field ( @xmath102 ) .
with the inputs from section 3 and section 4 , we have the following : @xmath103 the substitution of expansions from ( 9 ) and ( 15 ) leads to the following expansion @xmath104 + \bar \theta \ ; [ -g \ ; ( g_{\mu\nu } \times c)](x ) \nonumber\\ & + \theta \bar \theta \ ; [ g^2 \ ; ( g_{\mu\nu } \times c ) \times \bar c - i g \ ; ( g_{\mu\nu } \times b)](x)\end{aligned}\ ] ] which results in the following off - shell nilpotent ( anti-)brst transformations for the tensor field @xmath101 , namely ; @xmath105 it is also interesting to check explicitly that , @xmath106 which establishes the grassmannian independence of the l.h.s .. as a consequence , we infer from this observation that @xmath107 is an ( anti-)brst invariant quantity . in an exactly similar fashion
, it is straightforward to note that [ cf .
( 10 ) and ( 15 ) ] : @xmath108 ( x ) - \bar \theta \ , [ g ( f_{\mu\nu } \times \rho ) \times c ] ( x ) \nonumber\\ & + \theta \bar \theta \ , [ g^2 \ , \{(f_{\mu\nu } \times \rho ) \times c \ } \times \bar c - i g \ , ( f_{\mu\nu } \times \rho ) \times b ] ( x).\end{aligned}\ ] ] the above expansion implies that the ( anti-)brst symmetry transformations of the composite field @xmath109 are as given below : @xmath110 furthermore , it is elementary to show that the following is correct , namely ; @xmath111 \cdot \big [ \big(\tilde f^{\mu\nu ( h ) } \times \tilde \rho^{(h , g)}\big ) ( x , \theta , \bar \theta ) \big ] = ( f_{\mu\nu } \times \rho ) ( x ) \cdot ( f^{\mu\nu } \times \rho ) ( x)\end{aligned}\ ] ] which establishes the grassmannian independence of the l.h.s .. as a consequence , we conclude that @xmath112 is an ( anti-)brst invariant quantity ( i.e. @xmath113 = 0 $ ] ) . before we close this section ,
it is interesting to note that the following equation is correct , viz ; @xmath114 it verifies the grassmannian independence of the l.h.s .. this observation , in turn , implies the ( anti-)brst invariance of @xmath115 ( i.e. @xmath116 = 0 $ ] ) .
finally , it is clear that @xmath117 of the lagrangian density ( 1 ) is an ( anti-)brst invariant quantity .
furthermore , it is true to state that @xmath118 which proves the ( anti-)brst invariance of the last term of @xmath40 because the l.h.s . of ( 24 ) is actually independent of the grassmannian variables .
in the previous section ( cf . section 4 ) , we have derived the ( anti-)brst symmetry transformations for the relevant fields of the theory .
this can be seen from a close look at equations ( 9 ) , ( 10 ) and ( 15 ) . in explicit form , these ( anti-)brst transformations are @xmath119 @xmath120 the above transformations are off - shell nilpotent of order two @xmath121 and absolutely anticommuting @xmath122 in nature in their operator form .
the ( anti-)brst symmetry transformations for the nakanishi - lautrup type auxiliary fields ( @xmath123 ) have been obtained from the requirements of the nilpotency and absolute anticommutativity properties of @xmath2 .
in fact , the above requirements lead to the following proper ( anti-)brst symmetry transformations : @xmath124 the absolute anticommutativity of the ( anti-)brst symmetry transformations ( @xmath2 ) , applied onto the following fields , namely ; @xmath125 is _ true _ only when cf - condition @xmath126 $ ] is satisfied .
thus , we note that our superfield formalism leads to ( i ) the derivation of the off - shell nilpotent and absolutely anticommuting ( anti-)brst transformations , and ( ii ) the cf - condition .
the latter defines a hypersurface in the 3d minkowski spacetime manifold on which the proper ( i.e. nilpotent and absolutely anticommuting ) ( anti-)brst transformations are defined .
furthermore , it can be checked that the cf - condition is ( anti-)brst invariant ( i.e. @xmath127 = 0 $ ] ) . in other words ,
the key results of our superfield formalism are the derivation of the _ proper _
( anti-)brst transformations and ( anti-)brst invariant cf - condition for the brst invariant theory .
the expressions for the coupled ( anti-)brst - invariant lagrangian densities of the 3d massive non - abelian gauge theory can be written as @xmath128\nonumber\\ { \cal l}_{\bar b } = { \cal l}_0 - s_{ab } \ ; s_b \left[\frac { i}{2 } \ ; a_\mu \cdot a^\mu + c \cdot \bar c + \frac { 1}{2}\ ; \phi_\mu \cdot \phi^\mu \right].\end{aligned}\ ] ] we note that the terms in the square brackets are chosen in such a way that each term has a mass dimension one and ghost number equal to zero .
furthermore , all these terms ( in the bracket ) are lorentz scalar .
this is due to the fact that the proper nilpotent ( anti-)brst symmetries increase the mass dimension of fields by _ one _ on which they operate . as a consequence , we have the following expressions for the 3d ( anti-)brst invariant coupled ( but equivalent ) lagrangian densities : @xmath129 @xmath130 we emphasize that there are no gauge - fixing and faddeev - popov ghost terms for the vector field @xmath9 in the lagrangian densities ( 30 ) and ( 31 ) .
the reason , behind this observation , is that the field @xmath9 transforms covariantly [ i.e. @xmath131 under the ( anti-)brst transformations [ cf .
( 25 ) , ( 26 ) ] .
thus , the term @xmath132 remains invariant under the ( anti-)brst transformations and , as a consequence , there is no contribution from this term to the ( anti-)brst invariant coupled ( but equivalent ) lagrangian densities quoted in ( 30 ) and ( 31 ) .
it can be checked that the above lagrangian densities @xmath133 and @xmath134 transform , under the off - shell nilpotent ( anti-)brst transformations , as @xmath135 \qquad s_{ab } { \cal l}_b = \partial_\mu(b \cdot \partial^\mu \bar c ) - d_\mu [ b + \bar b + i\ , g\ , ( c \times \bar c ) ] \cdot \partial^\mu \bar c
\nonumber\\ \fl s_{ab } { \cal l}_{\bar b } = - \partial_\mu[\bar b\cdot ( d^\mu \bar c)]\ ; \ ; \;\ ; s_b { \cal l}_{\bar b } = - \partial_\mu(\bar b \cdot \partial^\mu c ) + d_\mu [ b + \bar b + i\ , g ( c \times \bar c ) ] \cdot \partial^\mu c.\end{aligned}\ ] ] thus , the corresponding actions ( i.e. @xmath136 and @xmath137 ) remain invariant on the constrained hypersurface in the 3d spacetime manifold where the cf - condition is satisfied .
the explicit expressions in ( 32 ) explain that @xmath133 and @xmath134 are equivalent and both of them respect the nilpotent ( anti-)brst symmetry transformations .
the invariance of the lagrangian density ( or action ) , under any arbitrary continuous symmetry transformation , leads to the derivation of the conserved current according to the noether s theorem . as a consequence ,
the local , continuous and off - shell nilpotent ( anti-)brst symmetry transformations ( @xmath2 ) lead to the derivation of the following noether s conserved currents @xmath138 : @xmath139 \cdot ( d_\nu c)\nonumber\\ & + g \;[g^{\mu\nu } + g\ ; ( f^{\mu\nu } \times \rho ) ] \cdot ( \phi_\nu \times c ) + \frac { i}{2 } \ ; g \ ; \partial^\mu \bar c \cdot ( c \times c)\end{aligned}\ ] ] @xmath140 \cdot ( d_\nu \bar c)\nonumber\\ & + g\ ; [ g^{\mu\nu } + g ( f^{\mu\nu } \times \rho ) ] \cdot ( \phi_\nu \times \bar c ) - \frac { i}{2 } \ ; g \ ; \partial^\mu c \cdot ( \bar c \times \bar c).\end{aligned}\ ] ] the above expressions for the norther s currents can be re - expressed ( for the algebraic convenience ) , in the following form @xmath141\end{aligned}\ ] ] @xmath142.\end{aligned}\ ] ] now the proof of conservation law ( @xmath143 ) becomes easier and it can be confirmed by exploiting the following euler - lagrange equations of motion , derived from the lagrangian densities @xmath133 and @xmath134 , respectively , namely ; @xmath144 + m \ ; \varepsilon^{\mu\eta\nu } \ ; d_\mu \phi_\eta - \partial^\nu b \nonumber\\ \fl \quad \quad + g \,[(g^{\mu\nu } + g\ , f^{\mu\nu}\times \rho ) \times \phi_\mu ] - i \ ; g\ ; ( \partial^\nu \bar c \times c ) = 0 \nonumber\\ \fl \quad\quad d_\mu [ \;g^{\mu\nu } + g \,(f^{\mu\nu}\times \rho ) ] + \frac { m}{2 } \ ; \varepsilon^{\mu\eta\nu } \ ; f_{\mu\eta } = 0 \nonumber\\ \fl \quad \quad ( g^{\mu\nu } + g\ , f^{\mu\nu}\times \rho ) \times f_{\mu\nu } = 0 \qquad \partial_\mu ( d^\mu c ) = 0 \qquad d_\mu ( \partial^\mu \bar c ) = 0\end{aligned}\ ] ] @xmath145 + m \ ; \varepsilon^{\mu\eta\nu } \ ; d_\mu \phi_\eta + \partial^\nu \bar b
\nonumber\\ \fl \quad\quad + g \,[(g^{\mu\nu } + g\ , f^{\mu\nu}\times \rho ) \times \phi_\mu ] + i \ ; g\ ; ( \bar c \times \partial^\nu c ) = 0 \nonumber\\ \fl \quad\quad d_\mu [ g^{\mu\nu } + g \,(f^{\mu\nu}\times \rho ) ] + \frac { m}{2 } \ ; \varepsilon^{\mu\eta\nu } \ ; f_{\mu\eta } = 0 \nonumber\\ \fl \quad\quad ( g^{\mu\nu } + g\ , f^{\mu\nu}\times \rho ) \times f_{\mu\nu } = 0 \qquad \partial_\mu ( d^\mu \bar c ) = 0 \qquad d_\mu ( \partial^\mu c ) = 0.\end{aligned}\ ] ] the above conserved currents [ cf . (
35 ) , ( 36 ) ] lead to the derivation of the following conserved ( anti-)brst charges @xmath146 , namely ; @xmath147\end{aligned}\ ] ] @xmath148.\end{aligned}\ ] ] the conserved ( @xmath149 ) and nilpotent ( i.e. @xmath150 ) ( anti-)brst charges @xmath151 are the generators of the transformations ( 25 ) and ( 26 ) , respectively . for instance , it can be checked that @xmath152_{\pm}$ ] ( @xmath153 ) . here
the field @xmath154 is the generic field of the theory and @xmath155 signs , on the square bracket , stand for the ( anti)commutator for the generic field @xmath154 of the theory being ( fermionic)bosonic .
it is interesting to point out that the nilpotent generators @xmath151 , even though produce the nilpotent ( anti-)brst transformations for the basic fields , they are _ unable _ to generate the ( anti-)brst transformations for the auxiliary field @xmath3 of the theory .
even the requirements of the nilpotency and absolutely anticommuting properties of the proper ( anti-)brst symmetry transformations do not generate the ( anti-)brst transformations for the auxiliary field @xmath3 .
this is a novel observation in this theory ( within the framework of brst formalism ) . for the usual 4d ( non)-abelian 1-form gauge theory , there are two inputs that lead to the derivation of _ all _ the ( anti-)brst symmetry transformations of _ all _ the relevant fields of the specific theory .
these are ( i ) the ( anti-)brst charges @xmath151 as the generators of the nilpotent ( anti-)brst symmetry transformations , and ( ii ) the requirements of nilpotency and absolute anticommutativity which lead to the derivation of the ( anti-)brst symmetry transformations for the auxiliary fields of the theory .
thus , the auxiliary field @xmath3 is very special .
the lagrangian densities ( 30 ) and ( 31 ) remain invariant under the following scale transformations for the ( anti-)ghost and other basic as well as auxiliary fields , namely ; @xmath156 where @xmath157 is a global infinitesimal scale parameter .
the @xmath158 signs , in the exponents , represent the ghost number of the fields @xmath159 and the ghost number for the rest of the fields ( i.e. @xmath160 ) is equal to zero . as a consequence , the latter fields do not transform at all under the ghost - scale transformations .
it is straightforward to check that the following infinitesimal transformations @xmath161 , obtained from ( 41 ) , namely ; @xmath162 are the _ symmetry _
transformations for the lagrangian densities ( 30 ) and ( 31 ) because it is straightforward to check that @xmath163 .
the above ghost symmetry transformations @xmath164 lead to the derivation of noether s conserved current and charge as given below : @xmath165\quad \qquad\qquad q_g = i \int d^2 x \ ; \big[\bar c \cdot d^0 c - \dot{\bar c } \cdot c \big].\end{aligned}\ ] ] it can be proven that the above ghost charge @xmath166 is the generator of ( 42 ) .
the nilpotent ( anti-)brst charge @xmath167 and the ghost charge @xmath166 satisfy the following standard brst algebra , namely ; @xmath168 = q_b\;\ ; \qquad i\ ; [ q_g , \ ; q_{ab } ] = - q_{ab } \nonumber\\ & & \{q_b , \;q_{ab}\ } \,=\ , q_b\ ; q_{ab } + q_{ab}\;q_b = 0\;\ ; \qquad q_g^2 \neq 0\end{aligned}\ ] ] which shows that the ghost number of the brst charge is @xmath169 and that of the anti - brst charge is @xmath170 .
the above statements about the ghost numbers can be checked explicitly by starting with a state @xmath171 that has the ghost number equal to @xmath172 ( i.e. @xmath173 ) . with this input and the above algebra ( 44 )
, we can check that the ghost numbers of states @xmath174 and @xmath175 are @xmath176 and @xmath177 , respectively . thus , the brst charge increases the ghost number by one when it operates on a quantum state . on the other hand ,
the anti - brst charge decreases the above number by one .
in our present investigation , we have exploited the _ usual _ classical yang - mills gauge symmetry of the jp model of 3d massive non - abelian gauge theory and generalized it to the ( anti-)brst symmetry transformations at the quantum level that are off - shell nilpotent and absolutely anticommuting in nature . in this endeavor ,
the augmented " superfield formalism ( where the hc and girs blend together beautifully ) [ 25 - 27 ] has played a decisive role in the derivation of the full set of appropriate transformations .
one of the important features of our superfield formulation is the derivation of the cf condition that enables us to obtain the absolutely anticommuting ( anti-)brst symmetry transformations .
thus , in addition to the brst symmetry for the jp model [ 17 ] , we have been able to derive the _
proper _ anti - brst symmetry transformations for the sake of completeness .
furthermore , the celebrated cf condition has been able to help us in deducing the coupled lagrangian densities [ cf .
( 30 ) , ( 31 ) ] that respect the above proper ( anti-)brst symmetry transformations for our present theory . in the context of 4d non - abelian 1-form gauge theory ,
it is well - known that the existence of anti - brst symmetry transformations ( @xmath178 ) is _ non - trivial_. the off - shell nilpotent ( @xmath179 ) ( anti-)brst symmetries _ anticommute _ @xmath122 with each other only on a constrained hypersurface , described by the curci - ferrari field equations , on a 4d minkowskian spacetime manifold [ 28 ] .
it can be also checked , using the appropriate equations [ cf .
( 25 ) , ( 26 ) , ( 39 ) , ( 40 ) ] , that for our present theory : @xmath180 = 0\end{aligned}\ ] ] @xmath181 = 0\end{aligned}\ ] ] are absolutely true on the 3d minkowski spacetime manifold where the cf restriction @xmath182 is satisfied .
the absolute anticommutativity of the ( anti- ) brst symmetries imply the linear independence of brst and anti - brst symmetry transformations .
the mathematical basis for the independence of the brst and anti - brst symmetries is encoded in the concept of gerbes that has been discussed in our earlier works [ 18,19 ] and illustrated geometrically in our appendix b. to obtain the full set of proper ( anti-)brst symmetry transformations , we are theoretically compelled to go beyond the hc and exploit the suitable girs to deduce the proper ( anti-)brst symmetries .
this is a novel feature of this model .
furthermore , as it turns out , the auxiliary field @xmath3 is _ not _ like the other auxiliary ( e.g. nakanishi - lautrup ) fields of the theory because its ( anti-)brst symmetry transformations do _ not _ arise from the requirements of nilpotency and absolute anticommutativity of the ( anti-)brst symmetry transformations .
this observation is _ also _ a novel feature of our present theory .
the good thing about our present augmented superfield formalism [ 25 - 27 ] is that it leads to the precise derivation of the proper ( anti-)brst symmetry transformations associated with this _ special _ auxiliary field @xmath3 , too .
it would be very nice endeavor to exploit the nym gauge transformations ( 3 ) within the framework of our superfield formalism and obtain the novel results connected with the nilpotent ( anti-)brst symmetry transformations that emerge from it .
in fact , two of us ( rk and sg ) , have already made significant progress in obtaining the proper ( anti-)brst symmetry transformations corresponding to nym symmetries by taking into account the following restrictions on the ( super)fields [ 30 ] : @xmath183 where the symbols carry their standard meanings ( as discussed in our present text ) .
it would also be exciting to take the combination of the local ym and nym gauge transformations _ together _ and obtain the full set of proper ( anti-)brst symmetry transformations , relevant coupled lagrangian densities and exact ( conserved , nilpotent and absolutely anticommuting ) ( anti-)brst charges that generate the proper and full set of ( anti- ) brst symmetry transformations for the jp model .
our present work and the forthcoming paper [ 30 ] would be the limiting cases of this general approach to the jp model ( where the ym and nym symmetries would blend together ) .
sg and rk would like to gratefully acknowledge the financial support from csir and ugc , new delhi , government of india , respectively .
the nilpotency and anticommutativity of the ( anti-)brst symmetry transformations and corresponding generators can be proven , in a simple and elegant manner , by exploiting the potential and power of superfield formalism .
as has been pointed out earlier , it can be checked , from the expansions in ( 9 ) and ( 15 ) , that @xmath184 where @xmath185 is the generic 3d field of the theory and @xmath186 are the corresponding superfields ( defined on the @xmath46-dimensional supermanifold and expanded after the application of hc and/or girs ) . as a consequence
, the nilpotency of the ( anti-)brst transformations is captured in the nilpotency ( @xmath187 ) of the grassmannian directions ( @xmath188 ) . in an exactly similar fashion
, it can be checked that @xmath189 encodes the absolute anticommutativity @xmath190 of the ( anti-)brst symmetry transformations in 3d spacetime for the jp model of massive theory .
that the ( anti-)brst charges are _ also _ nilpotent of order two , can be captured in the following expressions within the framework of superfield formalism , namely ; @xmath191 \nonumber\\ & \equiv \int d^2 x \int d \bar\theta \ ;
\big[b(x ) \cdot \tilde b_0^{(h ) } ( x , \theta , \bar\theta ) + i \ ; { \dot{\tilde { \bar f}}}^{(h)}(x , \theta , \bar\theta ) \cdot \tilde f^{(h ) } ( x , \theta , \bar\theta ) \big].\end{aligned}\ ] ] written in the ordinary 3d spacetime , the above expansions imply : @xmath192.\end{aligned}\ ] ] as a consequence , it is clear that @xmath193 because of the nilpotency ( @xmath194 ) of the brst transformations ( @xmath195 ) . in the language of superfield and superspace variables [ with the inputs from the nilpotency @xmath196 of derivative @xmath197 : @xmath198 there is an alternative way to express brst charge ( that is valid only on the constrained surface in 3d spacetime manifold where cf - condition @xmath199 $ ] is satisfied ) .
this is given , within the framework of superfield formalism , as @xmath200 \nonumber\\ & \;\equiv\ ; & \int d^2 x \ ; s_b\ , s_{ab } \ ; \big(i a_0 ( x ) \cdot c(x)\big).\end{aligned}\ ] ] the off - shell nilpotency of the brst charge is , once again , proven by the nilpotency of grassmannian derivative @xmath201 ( i.e. @xmath202 ) and the off - shell nilpotency ( @xmath194 ) of the brst transformations @xmath195 .
thus , we note that the nilpotency of brst charge is encoded in @xmath203 and @xmath194 when it is expressed in terms of grassmann derivative @xmath201 on the @xmath46-dimensional supermanifold and in terms of the proper brst transformations @xmath195 existing in 3d ordinary spacetime .
as we have expressed the brst charge @xmath204 in terms of the superfields ( obtained after the application of hc and/or girs ) , we can also express the anti - brst charge @xmath205 in the following two different ways , namely ; @xmath206 \nonumber\\ \fl \qquad \quad q_{ab } \ ; = i \ ; \frac { \partial}{\partial \theta } \frac { \partial}{\partial \bar \theta } \ ; \int d^2 x \ ; \big [ \tilde b_0^{(h ) } ( x , \theta , \bar\theta ) \cdot \tilde { \bar f}^{(h ) } ( x , \theta , \bar\theta ) \big]\end{aligned}\ ] ] where the second expression is true only on the 3d constrained surface where cf - condition @xmath207 $ ] is satisfied .
the nilpotency ( @xmath208 ) of the grassmannian derivative @xmath209 ensures that @xmath210 because it can be seen that @xmath211 .
the above expression can be expressed , in the 3d ordinary space , as given below @xmath212 \nonumber\\ & \equiv & i\ ; \int s_{ab } s_b \ ; \big ( a_0 ( x ) \cdot \bar c(x ) \big ) .
\end{aligned}\ ] ] the nilpotency of @xmath205 is captured in the equation @xmath213 because we know , from the anti - brst symmetry transformations ( 25 ) , that @xmath214 . the ( anti-)brst invariance and the equivalence of the coupled lagrangian densities can also be captured in the language of superfield formalism . to this end in mind
, it can be checked that the starting lagrangian density @xmath40 can be generalized onto the @xmath46-dimensional supermanifold as illustrated below @xmath215 the above expression for @xmath216 is , however , independent of the grassmannian variables because of our clear discussion in section 3 and section 4 . as a consequence , we have @xmath217 thus , we conclude that the above equation captures the ( anti-)brst invariance of the starting lagrangian density @xmath40 in terms of the superfield and superspace variables .
the ( anti-)brst invariant coupled lagrangian densities can be generalized onto the @xmath46-dimensional supermanifold as : @xmath218\nonumber\\ \fl \qquad { \cal l}_{b } \longrightarrow \tilde { \cal l}_{b } \ ; = \ ; \tilde { \cal l}_0 + \frac { \partial } { \partial \bar \theta}\ , \frac { \partial } { \partial \theta}\ ; \big [ \frac { i } { 2 } \ ; \tilde b_\mu^{(h ) } \cdot \tilde b^{\mu ( h ) } + \tilde f^{(h ) } \cdot \tilde { \bar f}^{(h ) } + \frac { 1 } { 2 } \ ; \tilde \phi_\mu^{(h , g ) } \cdot \tilde \phi^{\mu ( h , g ) } \big].\end{aligned}\ ] ] the nilpotency ( @xmath219 ) of the grassmannian derivatives ( @xmath220 ) and grassmannian independence of @xmath221 , lead to the following : @xmath222 the equivalence of the coupled lagrangian densities @xmath133 and @xmath134 is captured in the proof that @xmath223 because of the nilpotency ( @xmath224 ) and anticommutativity of the grassmannian derivatives ( i.e. @xmath225 ) .
in this appendix , we provide a clear geometrical interpretation for the existence of cf condition in the language of ( anti-)brst symmetry transformations on the non - abelian 1-form gauge field and associated ( anti-)ghost fields .
it is evident , from a close look at the off - shell nilpotent and absolutely anticommuting ( anti-)brst transformations ( 25 ) , ( 26 ) and ( 27 ) , that the following schematic diagram ( fig .
b1 ) captures these transformations geometrically .
we note that , in the whole diagram , there is a _ single point _ , at the ghost number _
zero _ , where there is a clustering of the auxiliary fields @xmath226 and @xmath227 which emerge from the ( anti-)brst symmetry transformations on the ghost fields @xmath228 and anti - ghost fields @xmath229 , respectively .
this is the place where the cf condition exists as it connects these auxiliary nakanishi - lautrup fields @xmath6 and @xmath230 with the ghost number _
zero _ object constructed from the fermionic ( anti-)ghost fields @xmath43 of the theory [ i.e. @xmath231 . in our earlier works on the 4d abelian 2-form and 6d 3-form gauge theories [ 18,19 ] ,
we have explained the existence and emergence of the ( anti-)brst invariant cf - type restrictions by exploiting the fundamental notions of geometry and group theory from pure mathematics . as a warm - up exercise in [ 19 ] , we have also considered the existence of cf condition for the 4d non - abelian 1-form gauge theory within the framework of brst formalism by taking the help of concepts from pure mathematics . in our present endeavor , we claim that , for any arbitrary @xmath232-form gauge theory within the framework of brst formalism , it is the symmetry transformations that would provide us the clue for the existence of cf - type conditions in the theory .
in fact , wherever , in the above type of diagram ( cf .
b1 ) , there is clustering of different variety of fields at a particular ghost number , there will emerge a cf type condition which will be connected with the idea of gerbes [ 18,19 ] .
thus , physically , we interpret the mathematical object gerbes as some artifact that connects the fields which cluster at a particular ghost number ( in the diagram like fig .
b1 for a given theory ) and they basically prove the linear independence ( i.e. the absolute anticommutativity ) of brst and anti - brst symmetries .
the hallmark of a gauge theory , at the classical level , is the existence of first - class constraints on them in the language of dirac s prescription for classification scheme . on the other hand ,
one of the decisive features of a gauge theory , at the quantum level , is the existence of cf - type conditions which are mathematically backed by the idea of gerbes in the realm of brst approach to gauge theories . | in the available literature , only the becchi - rouet - stora - tyutin ( brst ) symmetries are known for the jackiw - pi model of three ( 2 + 1)-dimensional ( 3d ) massive non - abelian gauge theory .
we derive the full set of off - shell nilpotent ( @xmath0 ) and absolutely anticommuting ( @xmath1 ) ( anti-)brst transformations @xmath2 corresponding to the _ usual _ yang - mills gauge transformations of this model by exploiting the augmented " superfield formalism where the horizontality condition and gauge invariant restrictions blend together in a meaningful manner .
this superfield formalism leads to the derivation of ( anti-)brst invariant curci - ferrari restriction which plays a key role in the proof of absolute anticommutativity of @xmath2 .
the derivation of the proper anti - brst symmetry transformations is important from the point of view of geometrical objects called gerbes .
a novel feature of our present investigation is the derivation of ( anti-)brst transformations for the auxiliary field @xmath3 from our superfield formalism which is _ neither _ generated by the ( anti-)brst charges _ nor _ obtained from the requirements of nilpotency and/or absolute anticommutativity of the ( anti-)brst symmetries for our 3d non - abelian 1-form gauge theory .
_ keywords _ : jackiw - pi model ; 3d massive gauge theory ; superfield formalism ; ( anti- ) brst symmetries ; nilpotency and absolute anticommutativity ; curci - ferrari condition |
in this paper , a graph is understood to be undirected and have no loops or multiple edges .
while graphs with multiple edges could be taken under consideration , it is not necessary to do so as multiple edges do not add any difficulty or important properties . a set of vertices in a graph @xmath0
is said to be independent if no two vertices in the set are joined by an edge .
an independent set @xmath1 of @xmath0 is called maximal if no independent set of @xmath0 properly contains @xmath1 .
the largest ( in terms of cardinality ) maximal independent set ( or sets ) of @xmath0 is called a maximum independent set of @xmath0 , and a graph is said to be well - covered if every maximal independent set of @xmath0 is also maximum . a well - covered graph could also be defined by the property of all maximal independent sets having the same cardinality .
this notion was introduced by plummer in @xcite . in @xcite ,
brown and nowakowski defined a well - covered weighting of a graph @xmath0 as a function @xmath2 that assigns values to the vertices of @xmath0 in such a way that @xmath3 is a constant for all maximal independent sets @xmath1 of @xmath0 .
it is immediate from the latter definition that one could re - define well - coveredness by saying that a well - covered graph is a graph that admits the constant function equal to @xmath4 as a well - covered weighting of @xmath0 .
we will use brown and nowakowski s presentation ( notation , nomenclature , etc ) , although this problem was originally introduced by caro , ellingham , ramey , and yuster in @xcite and @xcite .
it is easy to show that , once a field @xmath5 is fixed , the set of all well - covered weightings of a graph @xmath0 is an @xmath5-vector space , which is called _ the well - covered space of @xmath0_. the dimension of this vector space over @xmath5 is called the _ well - covered dimension of @xmath0 _ and is denoted by @xmath6 .
if @xmath7 does not depend on the field used then the well - covered dimension of @xmath0 is instead denoted as @xmath8 .
the well - covered dimension of a graph @xmath0 can therefore be considered to be the cardinality of a set of vertices @xmath9 of @xmath0 whose weights are independent of each other and on which the weights of the remaining vertices of @xmath0 are dependent , provided that the weighting is a well - covered weighting .
most of the results obtained in this paper are independent of the characteristic of @xmath10 , and when the characteristic becomes something to consider we will be careful to remark on it .
our graph theoretic notation , algebraic notation , and matrix theoretic notation is standard ; the reader can look at @xcite for any concepts we fail to define .
the vertex set of a graph @xmath0 is denoted by @xmath11 .
the cardinality of a set of vertices @xmath12 is denoted by @xmath13 . a field with @xmath14 ( @xmath15 prime ) elements
is denoted by @xmath16 .
the @xmath17 identity matrix is denoted by @xmath18 .
the @xmath17 matrix where each entry is a @xmath4 is denoted by @xmath19 .
an @xmath20 column vector where each entry is a @xmath4 is denoted by @xmath21 .
an @xmath20 column vector where each entry is a @xmath22 is denoted by @xmath23 .
it is relatively simple to calculate the well - covered dimension of a graph @xmath0 , provided @xmath0 is not too large .
one first needs to find all the independent sets of @xmath0 , which can be done using a greedy algorithm .
suppose that the maximal independent sets of @xmath0 are @xmath24 for @xmath25 .
then a well - covered weighting @xmath26 of @xmath0 is determined by a solution of the linear system of equations formed by selecting a maximal independent set , in this particular instance @xmath27 , and setting the system @xmath28 for @xmath29 .
this system is homogeneous , and can therefore be written in the form @xmath30 .
note that @xmath31 is an @xmath32 matrix where @xmath33 and @xmath34 .
as this system is homogeneous , the nullity of @xmath31 ( here is where @xmath35 could be relevant ) is equivalent to @xmath6 .
so , @xmath36 . in the case when @xmath37 , then @xmath38 , which implies that in this case the only possible well - covered weighting is the @xmath22 function . for the remainder of this paper , we shall concern ourselves only with the determining of the well - covered dimensions for various individual graphs and graph families .
we start by recalling a lemma from @xcite , as it will allow us to focus only on connected graphs .
let @xmath0 and @xmath39 be graphs . then @xmath40 the family of ( connected ) graphs that is possibly the easiest to attack when looking for well - covered dimensions is the family of complete graphs . by simply looking at the maximal independent sets of @xmath41 we get that @xmath42 .
similarly , using the previous lemma , we get @xmath43 .
only using the technique mentioned above , it is easy to find the well - covered dimension of several families of graphs .
\(a ) @xmath44 , where @xmath45 is the petersen graph .
+ ( b ) the two graphs in figure [ sg1andsg2 ] below are he smallest graphs such that their well - covered dimension depends on the characteristic of the base field ( @xmath46 ) or it is zero ( @xmath47 ) .
in fact , @xmath48 ( left ) and @xmath47 ( right).,title="fig:",height=144 ] ( left ) and @xmath47 ( right).,title="fig:",height=144 ] a similar result is obtained for crown graphs .
recall that , for any @xmath49 , the crown graph @xmath50 is formed by removing a perfect matching from @xmath51 .
though not specifically stated as such , it was proven in @xcite that @xmath52 , if @xmath53 , and @xmath54 if both @xmath55 and @xmath56 are even .
we shall extend this result to allow us to calculate the well - covered dimensions of all crown graphs over all fields .
let @xmath50 denote a crown graph , for all @xmath57 .
then , @xmath58 let @xmath59 be the complete bipartite graph with @xmath60 and @xmath61 and let @xmath62 be the perfect matching that is removed from @xmath59 to form @xmath50 .
the maximal independent sets of @xmath50 are @xmath63 for @xmath64 , and @xmath65 and @xmath66 . setting the sum of each of the weights on the maximal independent sets equal to that of the weights on the vertices of @xmath66
, we find that the linear system corresponding to the well - covered weightings is @xmath67 , where @xmath68 an @xmath69 matrix . subtracting the top @xmath56 rows from the bottom yields @xmath70 it follows that we have two possibilities depending on whether or not @xmath71 divides @xmath72 .
the theorem follows after finding the rank of this matrix in either case .
[ k - partite ] @xmath73 , where @xmath74 is the ( obvious ) complete @xmath75-partite graph .
let @xmath76 be a well - covered weighting of @xmath77 .
the maximal independent sets of @xmath78 are @xmath79 for @xmath80 . setting the sum of each of the weights on the maximal independent sets equal to that of the weights on the vertices of @xmath81
, we find that the linear system corresponding to the well - covered weightings is @xmath67 , where @xmath82 a @xmath83 matrix .
@xmath31 has rank @xmath84 .
hence the nullity is @xmath85 , which implies that @xmath86 .
[ turan ] @xmath87 , where @xmath88 denotes a turn graph .
we can see immediately that if @xmath56 is divisible by @xmath89 , then corollary [ turan ] reduces to @xmath90 .
the behaviors , in terms of well - covered weightings , of paths and cycles are very similar .
hence , we will study these two families simultaneously .
consider @xmath0 to be an @xmath56-path or an @xmath91-cycle , for @xmath92 .
label six ` consecutive ' vertices @xmath93 and @xmath76 as in the picture below .
let @xmath26 be a well - covered weighting of @xmath0 , and let @xmath94 and @xmath95 be two maximal independent sets of @xmath0 that contain the same vertices except from @xmath94 containing @xmath96 and @xmath95 containing @xmath97 instead .
locally , these two independent sets are represented in the figure below . and @xmath95 on six consecutive vertices . ] since @xmath94 and @xmath95 just ` interchange ' @xmath98 and @xmath99 , then @xmath100 .
it is now immediate that all vertices of @xmath101 , for @xmath102 , have the same weight for all well - covered weightings of this graph .
hence , @xmath103 for all @xmath102 . now consider two maximal independent sets @xmath104 and @xmath105 of @xmath101 , with @xmath102 , that contain the same vertices outside of a string of seven consecutive vertices , where @xmath104 and @xmath105 contain four and three vertices respectively .
these seven vertices , with the vertices contained in @xmath104 and @xmath105 are represented in figure [ differentnumberofelements ] below .
it follows that this graph admits maximal independent sets with different cardinalities , and thus @xmath106 for all @xmath102 .
similarly , from the argument associated to figure [ sixconsecutivevertices ] , if @xmath92 and @xmath107 ( edges connecting @xmath108 with @xmath109 ) then vertices @xmath110 must have the same weight for all well - covered weightings of @xmath111 .
moreover , for small values of @xmath56 it is easy to see that these weights must be zero . for larger values of @xmath56 figure [ differentnumberofelements ] provides a way to construct maximal independent sets with different cardinality , which forces @xmath112 . finally , we can construct two maximal independent sets of @xmath111 that share all but one vertex , which is @xmath113 for one of them and @xmath114 for the other .
this can be seen in the figure below .
it follows that @xmath115 , and symmetrically that @xmath116 , for all well - covered weightings @xmath26 of @xmath111 .
lastly , we want to remark that @xmath117 is independent of @xmath118 , and thus , adding simple computations to the arguments above we obtain the following theorem : if @xmath26 is a well - covered weighting of @xmath119 and @xmath120 , then @xmath121 , @xmath122 , and @xmath123 .
moreover , @xmath124 and @xmath125 if @xmath126 .
also , 1 .
@xmath127 , if @xmath102 , 2 .
@xmath128 , if @xmath129 , 3 .
@xmath130 , 4 .
@xmath131 .
the well - covered dimensions of paths had been already computed in @xcite using methods different from the ones used in this paper .
now we look at the family of gear graphs . a gear graph over @xmath132 vertices , denoted
@xmath133 is the graph with vertex set @xmath134 where : 1 .
@xmath135 is adjacent to @xmath136 and @xmath137 for @xmath138 .
2 . if @xmath139 , then @xmath135 is adjacent to @xmath140 . we can compute the well - covered dimensions of the gear graphs using the same methods we used to compute the well - covered dimensions of the cycle graphs .
let @xmath141 be the gear graph in @xmath132 vertices , then @xmath142
in this section we look at the well - covered dimension of graphs that can be constructed from known ones by using various techniques . we begin with a definition .
let @xmath0 be a graph and @xmath143 .
a @xmath144-blowup of a vertex @xmath145 is an independent set @xmath146 that ` takes the place ' of @xmath147 .
more precisely , wherever there was an edge joining @xmath147 to @xmath148 there is an edge joining @xmath108 with @xmath26 .
the graph obtained by the @xmath144-blowup of @xmath147 will be denoted @xmath149 .
similarly , for @xmath150 and @xmath151 we denote a ` double blowup ' @xmath152 as @xmath153 . for multiple blowups
we extend in the natural way the notation set of double blowups .
we begin by noticing that a maximal independent set of @xmath0 not containing @xmath113 is also a maximal independent set of @xmath39 , and if @xmath161 is a maximal independent set of @xmath0 then @xmath162 is a maximal independent set of @xmath39 .
moreover , it is easy to see that every maximal independent set of @xmath39 must be of one of these two types .
let @xmath1 and @xmath163 be the matrices associated to the systems of equations arising from looking for well - covered weightings of @xmath0 and @xmath39 respectively .
we notice that @xmath163 has @xmath164 more columns than @xmath1 but that it has exactly the same number of rows , and in fact the same rank as @xmath1 , which is @xmath165 .
the result follows .
the lexicographic product of @xmath0 and @xmath39 , denoted @xmath170 , is the graph with vertex set @xmath171 and edges joining @xmath172 and @xmath173 if and only if @xmath174 or @xmath175 and @xmath176 .
[ linalg1 ] let @xmath1 be an @xmath181 matrix and let @xmath182 be the @xmath183 matrix obtained by subtracting the first row @xmath184 of @xmath1 from all the other rows of @xmath1 , and then deleting @xmath184 .
assume @xmath185 , then , @xmath186 [ linalg2 ] let @xmath188 and @xmath189 be matrices obtained by using the construction described in lemma [ linalg1 ] from matrices @xmath190 , and @xmath187 respectively , we will re - arrange rows in the matrices if necessary to get , when possible , the first row to be dependent of the others .
then , @xmath191 whenever there is a row that is dependent of others in @xmath31 or @xmath1 , as in these cases we can always choose a row of @xmath187 that depends on the other rows of this matrix . on the other hand ,
if both @xmath1 and @xmath31 have linearly independent rows , then @xmath187 also has this property ( using @xmath192 ) , and thus @xmath193 .
[ lex ] let @xmath0 and @xmath39 be graphs with @xmath196 , @xmath197 , @xmath198 , and @xmath199 .
then , @xmath200 where @xmath201 represents the kronecker delta , and @xmath202 , @xmath203 are the number of maximal independent sets of @xmath39 and @xmath0 respectively .
we first notice that if @xmath204 is a maximal independent set of @xmath0 then @xmath205 is a maximal independent set of @xmath206 , where @xmath207 is a maximal independent set of @xmath39 for all @xmath208 .
moreover , it is easy to see that every maximal independent set of @xmath206 must be obtained this way
. set the weight - sums of each of the independent sets of @xmath0 equal to zero .
let @xmath1 be the matrix representing that homogeneous system of equations .
note that the matrix @xmath182 needed to find @xmath7 is obtained from @xmath1 by using the construction described in lemma [ linalg1 ] .
similarly , by repeating this process with @xmath39 we obtain @xmath209 ( needed for finding @xmath210 ) out of @xmath31 ( found by setting the weight - sums of the maximal independent sets of @xmath39 equal to zero ) .
now we notice that ( because of the first paragraph in this proof ) @xmath211 is the matrix associated to the homogeneous system given by setting the weight - sums of all the independent sets of @xmath206 equal to zero .
it follows that we are interested in finding the rank of the matrix @xmath189 obtained from @xmath211 by using the construction described in lemma [ linalg1 ] . since @xmath212 , @xmath213 , and @xmath214 , then using that a matrix has linearly dependent rows if and only if its rank is not equal to its number of rows , and remark [ linalg2 ] , we get @xmath215 where @xmath202 , @xmath203 represent the number of maximal independent sets of @xmath39 and @xmath0 respectively ( which are the number of rows of @xmath31 and @xmath1 respectively ) . | the well - covered dimension of various individual graphs and graph families are computed .
special families , such as the crown graphs , are studied and they yield well - covered dimensions that change depending on the characteristic of the field chosen as the field of scalars . also , formulas to find the well - covered dimension of graphs obtained by vertex blowups on a known graph , and to the lexicographic product of two known graphs are given . |
the classical schur horn theorem@xcite , @xcite , relates the digonal and the eigenvalue list of a hermitian matrix : let @xmath9 be a positive semidefinite element of @xmath10 and let @xmath11 and @xmath12 be the lists of diagonal entries and eigenvalues respectively , both sorted in non - increasing order .
then , the schur - horn theorem says that we must have @xmath13 the above condition on the lists is denoted by saying that the diagonal list is majorized by the eigenvalue list , written @xmath14 .
the schur - horn theorem states that further , given two positive lists @xmath15 with @xmath14 , then there is a positive semi - definite matrix @xmath9 with eigenvalues @xmath16 and diagonal @xmath17 .
majorization can also be defined for matrices .
given two positive operators @xmath18 in @xmath10 , we say that @xmath19 if the eigenvalue sequence of @xmath9 is majorized by the eigenvalue sequence of @xmath20 . the schur - horn theorem
can then be stated as saying that if @xmath9 is a diagonal positive matrix and @xmath20 a positive matrix so that @xmath21 , then there is a unitary operator @xmath22 so that the diagonal of @xmath23 is @xmath9 .
majorization for matrices has the following alternate description due to hardy , littlewood and polya@xcite , given two self - adjoint operators @xmath18 in @xmath24 , @xmath9 is majorized by @xmath20 iff @xmath25 for every continuous convex real valued function @xmath26 defined on a closed interval @xmath27 $ ] containing the spectra of both @xmath9 and @xmath20 .
majorization in type @xmath1 factors@xcite is described analogously , with the trace on @xmath24 in the definition replaced by the canonical trace @xmath28 .
let @xmath0 be a maximal abelian sefladjoint subalgebra(masa , in short ) in a type @xmath1 factor @xmath2 ; there is a unique trace preserving normal(weak * to weak * continuous ) conditional expectation @xmath29 that is in many ways analogous to the restriction mapping onto the diagonal for elements of @xmath10 .
arveson and kadison@xcite showed that if @xmath20 is a positive operator in @xmath2 , then @xmath30 .
this fact can also be deduced from hiai s work on stochastic maps on von neumann algebras@xcite .
there are two natural generalizations of the schur - horn theorem to type @xmath1 factors .
the first originates in the standard interpretation of the schur - horn theorem as characterizing the set of all possible diagonals of a positive matrix .
let @xmath31 be the set of unitary operators in @xmath2 and given an operator @xmath20 , let @xmath32 be the norm closure of the unitary orbit of @xmath20 , i.e @xmath33 two positive operators @xmath9 and @xmath20 in a type @xmath1 factor @xmath2 are said to be _
equimeasurable _ , denoted @xmath34 , if @xmath35 for @xmath36 .
it is routine to see that the following are equivalent .
1 . @xmath34 .
2 . @xmath37 .
the following result characterizing possible diagonals of positive operators , is the first main theorem in this paper .
let @xmath2 be a type @xmath1 factor and let @xmath38 be positive operators with @xmath19 .
then , there is some masa @xmath0 in @xmath2 such that @xmath39 the second generalization was conjectured by arveson and kadison in @xcite .
the second main theorem in this paper is the proof of their conjecture , let @xmath0 be a masa in a type @xmath1 factor @xmath2 . if @xmath3 and @xmath40 are positive operators with @xmath19 .
then , there is an element @xmath41 such that @xmath42 one can not escape having to take the norm closure of the unitary orbit of @xmath20 , see the paper _ loc.cit .
_ for a discussion on the necessity . in infinite dimensions
, unitary equivalence can not be determined from spectral data alone . on another note ,
it is trivial to see that the above theorems about diagonals for positive operators immediately yield identical theorems for hermitian operators , by adding a suitable constant to make them positive . a special case of the above theorem , namely , that given any positive contraction @xmath9 in @xmath0 , there is a projection @xmath43 in @xmath2 so that @xmath44 , had been conjectured earlier by kadison in @xcite , see also @xcite , who referred to it as the carpenter problem in type @xmath1 factors .
neither of the two theorems directly implies the other .
it is however , easy to see that theorem([conj2 ] ) implies theorem([conj1 ] ) when @xmath20 has finite spectrum .
there has been methodical progress towards the resolution of arveson and kadison s conjecture ( [ conj2 ] ) : argerami and massey@xcite showed that @xmath45 this was improved by bhat and ravichandran@xcite , who showed that it is enough to take the norm closure .
they also showed that the conjecture holds when both the operators @xmath9 and @xmath20 have finite spectrum .
dykema , hadwin , fang and smith@xcite gave a natural way to approach the problem and reduced the conjecture to a question involving kernels of conditional expectations . using this approach
, they were able to show that the conjecture holds , among other cases , for the radial and generator masas in the free group factors .
however , it is unclear if their strategy can be be used to settle the conjecture in full .
it must be pointed out that approximate schur - horn theorems are easier to obtain than exact ones .
further , it is possible that one might lose much fine structure : for instance , kadison characterised the diagonals of projections in @xmath46 and discovered an index obstruction to a sequence arising as the diagonal of a projection .
this subtlety is lost when ones passes to the norm closure of the set of diagonals , see @xcite for a discussion
. there has been a great deal of progress towards characterising the diagonals of hermitian operators in @xmath46 ; unlike the finite dimensional case , the situation in infinite dimensions is highly subtle .
bownick and jasper have given a complete description of the possible diagonals of hermitian operators in @xmath46 in @xcite and @xcite .
in another direction , the diagonals of compact operators have been characterised by loreaux and weiss in @xcite , extending the earlier work of kaftal and weiss in @xcite .
i end with a recent development ; thompson s theorem in matrix algebras characterises operators which can realised to have a prescribed diagonal in terms of their singular values .
kennedy and skoufranis in @xcite have recently extended thompson s theorem to the setting of type @xmath1 factors .
this paper has six sections apart from the introduction . in section
( 2 ) , i collect some standard facts in noncommutative measure theory .
section ( 3 ) exploits the useful observation that once we can solve the problem locally , theorem ( [ conj1 ] ) follows using transfinite induction .
section ( 4 ) contains the proof of theorem ( [ conj1 ] ) .
section ( 5 ) then builds upon this result to prove theorem ( [ conj2 ] ) .
we then use the schur - horn theorem for type @xmath1 factors to deduce theorems for type @xmath7 factors in section ( 6 ) . in the last section ,
namely section ( 7 ) for sake of completeness ,
i explain the situation both in the case of general finite von neumann algebras and type @xmath8 factors .
i would like to thank junsheng fang for telling me about this problem and for several useful discussions .
there is a more concrete description of majorization in type @xmath1 factors that is more convenient to work with , that we now describe .
we will use the following nonstandard definition repeatedly : given two subsets @xmath47 and @xmath48 of @xmath49 , say that @xmath50 if @xmath47 is to the right of @xmath48 , i.e. @xmath51 .
we similarly define the relation @xmath52 . also , given a self - adjoint operator @xmath20
, we will use @xmath53 to denote @xmath54 .
let @xmath9 be a positive operator in type @xmath1 factor . by the spectral theorem
, there is a borel measure with compact support , @xmath55 on @xmath49 so that @xmath56 define the real valued function @xmath57 on @xmath58 by @xmath59 this function @xmath57 is non - increasing and right continuous .
we have the identity @xmath60 where @xmath61 denotes lebesgue measure .
the values of this function were denoted the generalised @xmath62 numbers of @xmath9 by fack and kosaki@xcite .
we however , choose to call the function @xmath57 the spectral scale of @xmath9 .
it is a standard fact@xcite that one can find a projection valued measure which we denote by @xmath63 on @xmath64 $ ] so that @xmath65 for any borel measurable set @xmath66 $ ] and so that @xmath67 this projection valued measure is not unique when there are atoms in the spectrum of @xmath9 .
however , given a positive operator @xmath9 , we will fix a measure once and for all and use @xmath63 to denote this . throughout this paper , it will be evident that the results will not depend on the particular choice of measure in this degenerate case . given two positive operators @xmath9 and @xmath20 inside a type @xmath1 factor @xmath68 with spectral scales @xmath57 and @xmath69 repectively , it can be shown that @xmath20 majorizes @xmath9 , written @xmath70 if @xmath71 when we do not have the last trace equality , we say that @xmath20 submajorizes @xmath9 and denote this by @xmath72 .
there are two concise ways of representing these inequalities in type @xmath1 factors .
the first uses the ky fan norm functions are defined by @xmath73 the function @xmath74 is continuous and we have that , @xmath72 iff @xmath75 for @xmath76 $ ] .
if we also have that @xmath77 , then , @xmath19 .
alternately , @xmath19 if @xmath78 ) ) \leq \tau(s \mu_s([0,t ] ) ) \text { for } 0 < t < 1 \quad \text { and } \quad \tau(a ) = \tau(s ) \end{aligned}\ ] ] given two positive operators @xmath9 and @xmath20 in a type @xmath1 factor @xmath2 , define the quantity @xmath79 , also denoted simply by @xmath80 when the ambient algebra is clear , by @xmath81 we have that @xmath72 exactly when @xmath82 and the function @xmath80 measures how far @xmath20 is from submajorizing @xmath9 .
we record some facts about the quantity @xmath80 .
[ llemma ] let @xmath18 be positive operators in a type @xmath1 factor @xmath2 .
then , 1 . @xmath83 .
if @xmath9 and @xmath20 commute with a set of orthogonal projections @xmath84 which sum up to @xmath85 , then , @xmath86 3 .
suppose additionally that @xmath87 and @xmath88 then , letting @xmath89 and @xmath90 for @xmath91 , @xmath92 it is easy to see that @xmath93 and that @xmath94 .
for the first assertion , we have @xmath95 for the second , it is easy to see that once we have proved the assertion for @xmath96 , the general case follows by induction .
assume then , that @xmath96 .
let @xmath97 be arbitrary ; we may write @xmath98 ) = \mu^{p_1\mathcal{m}p_1}_{sp_1}([0,a ] ) \oplus \mu^{p_2\mathcal{m}p_2}_{sp_2}([0,b])\]]and @xmath99 ) = \mu^{p_1\mathcal{m}p_1}_{ap_1}([0,c ] ) \oplus \mu^{p_2\mathcal{m}p_2}_{ap_2}([0,d]),\ ] ] for some @xmath100 . here
, the notation @xmath101 means that we calculate the relevant spectral projection for the operator @xmath102 considered as an operator inside the @xmath1 factor @xmath103 .
the expressions @xmath104 and @xmath105 when referring to the spectral scale and the ky fan norm function will be used similarly .
suppose that @xmath106 - the complementary case is handled similarly .
we have that @xmath107 ) ) = \tau(p_1)\tau_{p_1\mathcal{m}p_1}(\mu^{p_1\mathcal{m}p_1}_{sp_1}([0,a ] ) ) + \tau(p_2)\tau_{p_2\mathcal{m}p_2}(\mu^{p_2\mathcal{m}p_2}_{sp_2}([0,b ] ) ) = \tau(p_1)a + \tau(p_2)b,\ ] ] and similarly , @xmath108 . together
, with the assumption above , this implies that @xmath109 .
we also have that @xmath110 .
further , @xmath111 we make a simple calculation , @xmath112 ) ) - \tau(p_2 ) \tau_{p_2\mathcal{m}p_2}(sp_2\mu_{ap_2}^{p_2\mathcal{m}p_2}([b , d ] ) ) \\ & = & \tau(p_1)[f_{sp_1}^{p_1\mathcal{m}p_1}(a ) - f_{sp_1}^{p_1\mathcal{m}p_1}(c ) ] + \tau(p_2)[f_{sp_2}^{p_2\mathcal{m}p_2}(d ) - f_{sp_2}^{p_2\mathcal{m}p_2}(b)]\\ & = & \tau(p_1)(a - c ) \dfrac{\int_c^a
f_{sp_1}^{p_1\mathcal{m}p_1}(x)dm(x)}{a - c } - \tau(p_2)(d - b)\frac{\int_b^d f_{sp_2}^{p_2\mathcal{m}p_2}(x)dm(x)}{d - b}\\ & \geq & 0\end{aligned}\ ] ] another simple calculation shows that @xmath113 ) \oplus \mu_{sp_2}([0,b ] ) ) ] - \tau[a(\mu_{ap_1}^{p_1\mathcal{m}p_1}([0,c ] ) \oplus \mu_{ap_2}([0,d ] ) ) ] \\ & = & \tau(p_1 ) \tau_{p_1\mathcal{m}p_1}[sp_1\mu_{sp_1}^{p_1\mathcal{m}p_1}([0,a ] ) - ap_1\mu^{p_1\mathcal{m}p_1}_{ap_1}([0,c])]\\ & + & \tau(p_2 ) \tau_{p_2\mathcal{m}p_2}[sp_2\mu_{sp_2}^{p_2\mathcal{m}p_2}([0,b ] ) - ap_2\mu_{ap_2}^{p_2\mathcal{m}p_2}([0,d])]\\ & = & \tau(p_1 ) \tau_{p_1\mathcal{m}p_1}[sp_1\mu_{sp_1}^{p_1\mathcal{m}p_1}([0,c ] ) - ap_1\mu_{ap_1}^{p_1\mathcal{m}p_1}([0,c])]\\ & + & \tau(p_2 ) \tau_{p_2\mathcal{m}p_2}[sp_2\mu_{sp_2}^{p_2\mathcal{m}p_2}([0,d ] ) - ap_2\mu_{ap_2}^{p_2\mathcal{m}p_2}([0,d])]\\ & + & \tau(p_1 ) \tau_{p_1\mathcal{m}p_1}(sp_1\mu_{sp_1}^{p_1\mathcal{m}p_1}([c , a ] ) ) - \tau(p_2 ) \tau_{p_2\mathcal{m}p_2}(sp_2\mu_{ap_2}^{p_2\mathcal{m}p_2}([b , d ] ) ) \\ & \geq & \tau(p_1 ) l_{p_1 \mathcal{m } p_1}(ap_1,sp_1 ) + \tau(p_2 ) l_{p_2 \mathcal{m } p_2}(ap_2,sp_2)\end{aligned}\ ] ] we conclude that @xmath114 for the last assertion , given the hypotheses , it is easy to see that @xmath115 and thus , @xmath116 and hence , @xmath117\ } ) = \tau((s - a)q_{m-1 } ) + \tau(p_m)l_{p_m \mathcal{m } p_m}(ap_m , sp_m)\ ] ] the assertion follows .
we are also interested in schur - horn theorems in type @xmath7 factors .
approximate results in this setting were recently obtained by argerami and massey [email protected] @xmath118 be a @xmath119 finite type @xmath7 factor and let @xmath28 be a faithful normal semifinite trace on @xmath2 .
we will restrict our attention to masas @xmath0 that admit a normal trace preserving conditional expectation
. we will refer to them as atomic masas ; such masas are generated by their finite projections .
let @xmath9 be a positive trace class operator . then ,
as in the case of positive operators in type @xmath1 factors , there exists a spectral scale @xmath57 , this time on @xmath120 and a projection valued measure @xmath63 , this time on @xmath120 so that @xmath121 given , two trace class operators @xmath9 and @xmath20 , we say that @xmath20 majorizes @xmath9 , again written @xmath19 if inequalities analogous to ( [ maj21 ] ) hold . for trace class operators ,
it is more natural to take the closure of the unitary orbit in the trace norm than in the operator norm ; we thus define @xmath122 when @xmath20 is trace class in a type @xmath7 factor .
when the operators considered are not trace class , one needs to be more careful while considering majorization . as pointed out by neumann@xcite , one needs to consider both the upper and lower spectral scales defined as @xmath123 when @xmath9 is trace class , @xmath124 becomes zero . for two positive operators @xmath9 and @xmath20
, we say that @xmath20 majorizes @xmath9 if 1 .
we have the inequalities @xmath125 2 .
additionally , if there is a @xmath16 such that @xmath126 is trace class , then so is @xmath127 and @xmath128 .
recall , see ( [ em ] ) , that two positive operators @xmath9 and @xmath20 in a type @xmath1 factor @xmath2 are said to be equimeasurable if @xmath35 for @xmath129 .
this is equivalent to saying that the spectral measures and hence the spectral scales of @xmath9 and @xmath20 are identical .
it is also routine to see that this is also equivalent to the existence of a sequence of unitary operators @xmath130 so that @xmath131 .
an example of popa@xcite shows that equimeasurable operators need not be unitarily equivalent .
let @xmath9 lie inside a masa @xmath0 .
the same example of popa also shows that we can not hope to even locally conjugate @xmath20 into @xmath9 , i.e , it is not possible to find a unitary @xmath22 and a projection @xmath43 in @xmath0 so that @xmath132 and @xmath133 .
however , i show in proposition([prop1 ] ) that this can be accomplished whenever @xmath19 but @xmath9 is not equimeasurable to @xmath20 .
we start off with some elementary lemmas .
[ 2matlem ] let @xmath43 be a projection of trace @xmath134 inside a masa @xmath0 in a type @xmath1 factor @xmath2 .
let @xmath9 be a positive operator in @xmath0 and let @xmath20 be a positive operator in @xmath2 that commutes with @xmath43 .
with respect to the decomposition @xmath135 , we write(using an arbitrary partial isometry @xmath136 with @xmath137 and @xmath138 as the matrix unit @xmath139 ) , @xmath140 where @xmath141 and @xmath142 are masas in @xmath103 , the operators @xmath143 and @xmath144 are in @xmath141 and @xmath145 and @xmath146 are in @xmath103 .
assume that @xmath147 then , there is a unitary @xmath22 so that @xmath148 with @xmath149 .
we will automatically have that , @xmath150 = \sigma_{p\mathcal{m}p}(y ) \subset \operatorname{conv}(\sigma(s ) ) = [ \alpha(s_2),||s_1||]\ ] ] let @xmath151 be the positive operator in @xmath141 determined by the formula @xmath152 the operators @xmath143 , @xmath153 and @xmath154 form a commuting set and the condition([dom ] ) , it is easy to see that @xmath151 is a positive contraction
. now , let @xmath22 be the unitary given by @xmath155 conjugating @xmath20 by @xmath22 , we have that @xmath156 another calculation shows that @xmath157 & = & h^2 e_{\mathcal{a}_1}(s_1 ) + ( i - h^2)e_{\mathcal{a}_1}(s_2)\\ \nonumber & = & h^2 ( e_{\mathcal{a}_1}(s_1 ) - e_{\mathcal{a}_1}(s_2 ) ) + e_{\mathcal{a}_1}(s_2)\\ & = & a_1 \end{aligned}\ ] ] recall that @xmath158 is the smallest point in the spectrum of @xmath146 , @xmath159 we conclude that @xmath160 \subset [ \alpha(s_2),||s_1|| ] = \operatorname{conv}(\sigma(s))\end{aligned}\ ] ] [ greater ] in the setup of lemma ( [ 2matlem ] ) , suppose we have that @xmath161 , we will have that @xmath150 \subset ( \alpha(s_2),||s_1||)\ ] ] this is because , in line ( [ tech ] ) , we will have a strict inequality instead of mere inequality .
[ lem1 ] let @xmath19 be positive operators and suppose @xmath162 where the decomposition is with respect to @xmath163 where @xmath164 are orthogonal projections commuting with @xmath9 and @xmath20 and @xmath165 .
suppose that 1 .
@xmath166 and @xmath167 , 2 .
@xmath168 .
then , for any positive operator @xmath169 in @xmath170 with the same trace as @xmath146 and so that @xmath171 , we have the majorization relation , @xmath172 if in addition , we have that @xmath173 is strictly positive on @xmath174 , then @xmath175 is strictly positive on @xmath174 as well .
it is easy to see that we have @xmath176 \cup [ \tau(p+q),1]\ ] ] which is positive by hypothesis off @xmath177 and by assertions @xmath178 and @xmath179 of ( [ llemma ] ) that for @xmath180$]@xmath181 the assertion follows .
the following proposition is the critical step in the proof of the first schur - horn theorem , namely theorem ( [ conj1 ] ) .
it shows that the problem can be locally solved .
more precisely , [ prop1 ] let @xmath0 be a masa in a type @xmath1 factor @xmath2 .
let @xmath3 and @xmath40 be positive operators with @xmath19 .
assume that @xmath182 .
then , there are projections @xmath43 in @xmath0 and @xmath183 in @xmath2 with @xmath184 and @xmath185 and a unitary @xmath22 in @xmath2 satisfying @xmath186 so that @xmath187 let @xmath188 be the spectral scales of @xmath18 respectively and let @xmath189 be the ky fan norm functions associated to @xmath26 and @xmath190 respectively , see ( [ deff ] ) .
since @xmath19 , we have that @xmath191 assume for now that @xmath192 almost everywhere and that @xmath193 on @xmath174 . once we have proved the proposition under this assumption , the general case will follow using routine arguments - see the last paragraph of the proof .
let @xmath194\ , \mid f_a(x ) < f_s(x)\}$ ] and let @xmath195\ , \mid f_a(x ) > f_s(x)\}$ ] .
pick @xmath196 so that @xmath197 $ ] and @xmath198 $ ] have positive measure for every @xmath199 .
this is possible because of the condition ( [ majdef ] ) .
now , choose numbers @xmath200 with @xmath201 and define the number @xmath202 @xmath203 since @xmath74 and @xmath204 are continuous and @xmath173 is strictly positive on @xmath174 , we have that @xmath205 .
pick @xmath206 and pick subsets @xmath47 and @xmath48 of positive measure in @xmath197 $ ] and @xmath207 $ ] .
let @xmath208 be the sets @xmath209 we may further arrange , by passing to subsets , if needed , that the following are satisfied : 1 .
2 . @xmath211 see the figure below for a schematic description : ) , width=302 ] let @xmath212 be the projection @xmath213 and @xmath214 the projection @xmath215 and let @xmath216 .
choose a unitary @xmath136 such that @xmath217 it is easy to see that the unitary @xmath136 may be chosen so that @xmath218 .
with respect to the decomposition @xmath219 , we may write @xmath220 we have that @xmath221 by lemma(1 ) , there is a unitary @xmath222 satisfying @xmath223 inside @xmath2 so that @xmath224 where @xmath225 and further , @xmath226 $ ] .
let @xmath22 be the unitary @xmath227 and @xmath169 the operator @xmath228 .
we note that @xmath229 letting @xmath230 , the equation ( [ local ] ) implies that @xmath231 we have the trace inequality @xmath232 we now show that we also have the other required majorization condition , namely @xmath233 let us decompose @xmath9 and @xmath169 using the projection decompositions @xmath234 ) \oplus \mu_a(x ) \oplus [ \mu_{a}([a-\epsilon , a+\epsilon ] ) - \mu_a(x ) - \mu_a(y ) ] \oplus \mu_a(y ) \oplus \mu_{a}([a+\epsilon,1])\ ] ] and @xmath235 ) \oplus \mu_a(x ) \oplus [ \mu_{t}([a-\epsilon , a+\epsilon ] ) - \mu_a(x ) - \mu_a(y ) ] \oplus \mu_a(y ) \oplus \mu_{t}([a+\epsilon,1])\ ] ] respectively .
note that here we use the fact that @xmath236 to assure ourselves of that fact that @xmath237 and @xmath238 are sub projections of @xmath239 ) $ ] , a consequence of the special form of the averaging unitary @xmath22 .
@xmath240 we have that @xmath241 and @xmath242 and further , @xmath243)]}{\tau[\mu_a([0,a-\epsilon])]}\end{aligned}\ ] ] let @xmath244 be the conditional expectation given by compression to the diagonal followed by applying @xmath4 to the second diagonal entry .
we have that @xmath245 by the calculations ( [ hyp1 ] ) and ( [ hyp2 ] ) , the operators @xmath9 and @xmath246 satisfy the hypothesis of lemma(@xmath247 ) and thus @xmath248 , namely @xmath249 this implies that @xmath250 or , in other words , @xmath251 this completes the proof when @xmath192 a.e . and @xmath252 on @xmath174 .
now , we look at the general case , dropping the assumption that @xmath192 almost everywhere and that @xmath193 on @xmath174 .
define @xmath253 \mid f_a(x ) = f_s(x)\}$ ] , which may now have positive measure .
pick a unitary @xmath22 that conjugates @xmath254 into @xmath237 .
we may write , under the decomposition @xmath255 , @xmath256 where @xmath143 and @xmath145 have the same spectral measure , hence @xmath257 by @xcite[theorem 5.4 ] and @xmath258 with the property that the spectral scales @xmath259 and @xmath260 inside @xmath261 satisfy @xmath262 almost everywhere . since @xmath182 , @xmath144 and @xmath146 are non - zero .
since the ky fan norm functions are continuous , we may find two points @xmath263 so that @xmath264 and @xmath265 on @xmath266 . pick a unitary @xmath136 that conjugates @xmath267)$ ] into @xmath268)$ ] and commutes with @xmath237 .
we may write , under the decomposition @xmath269 ) \oplus [ i- \mu_a(x ) - \mu_a([a_1,a_2])]$ ] , @xmath270 where @xmath271 and whose respective spectral scales are non - equal a.e .
further , the ky fan norm functions satisfy @xmath272 on @xmath174 .
note that we also have that @xmath273 .
the proposition applies to @xmath274 and yields the desired conclusion for @xmath9 and @xmath275 .
given positive operators @xmath3 and @xmath40 as above , we say that @xmath276 is a _ partial solution _
if @xmath22 is a unitary , @xmath43 is a projection in @xmath0 , @xmath277 and @xmath278 . with this notation
, we have the following corollary , [ cor1 ] let @xmath0 be a masa in a type @xmath1 factor @xmath2 .
let @xmath3 and @xmath40 be positive operators with @xmath19 and let @xmath279 be a partial solution .
assume that @xmath280 inside @xmath281 .
then , there are projections @xmath282 and @xmath283 so that @xmath284 and @xmath285 and a unitary @xmath136 in @xmath2 so that @xmath286 is a partial solution and @xmath287 apply the previous proposition ( [ prop1 ] ) to @xmath288 and @xmath289 inside @xmath281 to get a unitary @xmath222 with @xmath290 and a projection @xmath291 in @xmath281 and a projection @xmath292 in @xmath288 so that ( inside @xmath281 ) we have @xmath293 and also , @xmath294 now , let @xmath295 , let @xmath296 and let @xmath297 . here , we interpret @xmath298 which are operators in @xmath281 , in the natural fashion inside @xmath2 . it is clear that @xmath136 is a unitary , that @xmath285 and that @xmath299 further , @xmath300 we conclude that @xmath286 is a partial solution with the desired properties .
we deduce a schur - horn theorem in type @xmath1 factors from corollary([cor1 ] ) using an induction argument .
[ shm1 ] let @xmath0 be a masa in a type @xmath1 factor @xmath2 .
if @xmath3 and @xmath40 are positive operators with @xmath19 . then , there is a unitary @xmath22 and a projection @xmath43 in @xmath0 such that @xmath301 if @xmath302 , there is nothing to prove ; just set @xmath43 to be zero and the unitary to be the identity .
let us therefore assume that @xmath182 .
let @xmath303 and @xmath40 be positive operators so that @xmath19 .
let @xmath304 be the collection of all tuples @xmath305 where @xmath22 is a unitary in @xmath2 and @xmath43 is a projection in @xmath0 with @xmath132 and @xmath133 .
define an ordering @xmath306 on the set @xmath304 by @xmath307 if 1 .
@xmath308 , i.e @xmath309 , 2 .
there is a projection @xmath183 , with @xmath310 and satisfying @xmath311 so that @xmath312 .
the set @xmath304 with the given ordering is a poset . to see this ,
suppose @xmath307 and @xmath313 ; while showing that @xmath314 , property @xmath178 is immediate .
let @xmath315 be the projections that ensure condition @xmath179 in the inequalities @xmath307 and @xmath313 respectively .
take @xmath316 . by definition
, we have that @xmath317 and @xmath318 .
this yields that @xmath319 and since @xmath320 is orthogonal to @xmath321 , we have @xmath322 .
similarly , @xmath323 .
we see that @xmath324 and further , @xmath325 thus , @xmath326 , giving us that @xmath327 next , @xmath328 in the third line , we used the fact that @xmath329
. we conclude from ( [ comp1 ] ) and ( [ ind ] ) that @xmath330 is a poset .
in what follows , we consider @xmath2 in its standard form , sitting inside @xmath331 .
let @xmath332 be a chain in @xmath304 .
since the projections @xmath333 are increasing , they have a strong operator limit , which we denote by @xmath43 .
fix operators @xmath169 and @xmath20 in @xmath2 .
we claim that @xmath334 exists .
fix @xmath199 .
since @xmath333 converge in the sot , there is an @xmath202 so that if @xmath335 , then @xmath336 .
let @xmath183 be the projection that witnesses @xmath337 , i.e. we have that @xmath338 we have that @xmath339 it is now routine to see that the sesquilinear forms on @xmath340 , @xmath341 converge pointwise and we denote the limit by @xmath342 .
it is easy to see that @xmath343 is a sesquilinear form too and that @xmath344 . by the riesz representation theorem
there is a contraction @xmath22 in @xmath345 such that @xmath346 this means in particular that @xmath347 converges to @xmath22 in the wot and thus , @xmath22 is in @xmath2 .
we now how that we actually have sot convergence - this will imply that @xmath22 is in fact a unitary .
since the operators are bounded , it is enough to check for sot convergence on the dense set @xmath348 .
let @xmath169 be in @xmath2 .
since @xmath349 converges in the wot to @xmath22 , we have that @xmath350 for any @xmath351 , let @xmath352 be the projection that witnesses @xmath337 .
we have that , @xmath353 it follows that @xmath347 converges to @xmath22 in the sot .
a similar calculation shows that @xmath354 converges in the sot to @xmath355 . since the @xmath347
are uniformly bounded in norm(by @xmath356 ) , we have that @xmath357 converges in the sot to @xmath358 and thus @xmath359 .
we conclude that @xmath22 is a unitary .
the strong @xmath360 convergence of the @xmath347 to @xmath22 implies that the automorphisms @xmath361 converge in the point @xmath247 norm topology to @xmath362 .
now , we have that @xmath363 concerning majorization , @xmath364 for every @xmath202 and hence , passing to the strong operator limit ,
@xmath365 we conclude that @xmath279 is in @xmath304 .
we now show that for every @xmath202 , we have that @xmath366 .
pick a sequence @xmath367 , @xmath368 in @xmath85 with @xmath369 so that @xmath370 is increasing and converges to @xmath43 in the sot .
as above , let @xmath371 for @xmath372 be the projection that witnesses @xmath373 .
let @xmath183 be the projection @xmath374 for any @xmath375 , we have that @xmath376 since @xmath377 is larger than @xmath378 and a fortiori larger than @xmath379 for @xmath380 , we have that @xmath381 .
further , @xmath382 . as a result ,
@xmath383 the sum telescopes to yield @xmath384 taking the limit as @xmath385 , we get that @xmath386 we conclude that @xmath366 for every @xmath202 .
thus , every chain has an upper bound and now , zorn s lemma gives us that there is a maximal element in @xmath304 .
let @xmath279 be this maximal element .
if @xmath288 is not equimeasurable to @xmath289 , corollary ( [ cor1 ] ) applies and yields us a larger element in @xmath304 , yielding a contradiction .
we conclude that there is a unitary @xmath22 and a projection @xmath387 so that @xmath301 we now prove the first of the two generalizations of the schur - horn theorem to type @xmath1 factors .
we repeat the statement of the theorem for the convenience of the reader .
[ conj1][the schur - horn theorem in type @xmath1 factors i ] let @xmath2 be a type @xmath1 factor and let @xmath38 be positive operators with @xmath19 .
then , there is some masa @xmath0 in @xmath2 such that @xmath39 choose a masa @xmath141 such that @xmath9 belongs to @xmath141 .
theorem ( [ shm1 ] ) yields that there is a unitary @xmath22 in @xmath68 and a projection @xmath43 in @xmath141 such that @xmath388 choose a masa @xmath389 in @xmath281 that contains @xmath289 .
then , we have that @xmath390 note that we have that @xmath391 .
let @xmath0 be the masa @xmath392 .
we then get by applying the automorphism @xmath393 to the equation ( [ fc ] ) that @xmath394 since @xmath395 , we are done .
the above theorem , as remarked in the introduction , is a natural generalization of the schur - horn theorem to type @xmath1 factors .
this generalization does not directly imply the alternative conjecture of arveson and kadison from @xcite .
we prove the conjecture in full in the next section .
we now turn to the second natural generalization of the schur - horn theorem .
the theorem of the last section characterizes the spectral distributions of operators that arise as the `` diagonal '' of a given positive operator @xmath20 . on the other hand ,
the conjecture of arveson and kadison complements the abovementioned theorem by characterizing the spectral distributions of operators which have a prescribed diagonal @xmath9 .
the calculations in this section are straightforward but technical .
perhaps a few words about the idea of the proof might be helpful .
let @xmath0 be a masa in type @xmath1 factor @xmath2 and let @xmath3 and @xmath40 be positive elements so that @xmath19 . theorem ( [ shm1 ] ) in the last section says that there is a unitary @xmath22 and a projection @xmath43 so that if we write write out @xmath9 and @xmath23 in block matrix form with diagonal @xmath396 , @xmath397 then @xmath398 and @xmath399 inside @xmath103 and @xmath281 respectively .
even though @xmath144 and @xmath146 are approximately unitarily equivalent inside @xmath281 , we can not expect to use these to implement an approximate unitary equivalence between @xmath23 and an operator of the form @xmath400 .
nevertheless , there is a workaround ; let us look closely at what @xmath19 means in terms of the spectral scales @xmath57 and @xmath69 .
roughly speaking , @xmath401 is larger than @xmath402 for @xmath403 close to @xmath404 and smaller for @xmath403 close to @xmath356 . rather than work with @xmath9 and @xmath20 directly
, we will work with @xmath405 and @xmath406 where @xmath43 and @xmath183 are carefully chosen spectral projections supported away from the extreme points of the spectra
. we will apply theorem ( [ shm1 ] ) to @xmath405 and @xmath406 and might end up with pieces that are equimeasurable as above .
we will then use the reserved head and tail of the spectra that we have hitherto left untouched to massage the equimeasurable parts carefully , in order to achieve the desired diagonal .
the main result in this section is the proof of theorem ( [ conj2 ] ) .
we first prove a couple of lemmas .
we will use the notation @xmath407 to mean that @xmath408 for every @xmath76 $ ] .
[ sht2lem ] let @xmath0 be a masa in a type @xmath1 factor @xmath2 and let @xmath3 and @xmath40 be two positive operators commuting with a projection @xmath43 of trace @xmath134 in @xmath0 , written with respect to the decomposition @xmath409 , @xmath410 where @xmath411 and @xmath412 and further , @xmath413 . then , for any @xmath414 , there is a unitary @xmath22 and projections @xmath415 and @xmath416 , both in @xmath0 , with @xmath417 for @xmath418 such that @xmath419 and @xmath420 .
let @xmath421 be fixed .
it is easy to see(using the fact that the spectral scales @xmath57 and @xmath69 are right continuous and non increasing ) that we may find a natural number @xmath422 , a number @xmath199 and disjoint intervals @xmath423 in @xmath64 $ ] with @xmath424 such that 1 .
@xmath425 . and , 2 .
@xmath426 for @xmath427 for @xmath428 .
define the projections @xmath429 pick a unitary @xmath430 in @xmath103 that conjugates @xmath431 onto @xmath432 and a unitary @xmath433 in @xmath281 that conjugates @xmath434 onto @xmath435 for @xmath428 .
let @xmath436 and let @xmath437 . then , @xmath169 commutes with the projections @xmath432 and @xmath435 . for @xmath438
, let @xmath439 and @xmath440 .
inside @xmath441 , we may write @xmath442 where @xmath443 and since , @xmath444 , we have that @xmath445 we see that for each @xmath428 , the pair of operators @xmath446 and @xmath447 inside @xmath441 satisfy the hypothesis of the remark following lemma([2matlem ] ) and we may thus find unitaries @xmath448 in @xmath441 such that @xmath449 let @xmath222 be a unitary in @xmath281 that conjugates @xmath450 onto @xmath451 for @xmath452 , i.e. @xmath453 . the second fact above gives us that @xmath454 let @xmath22 be the unitary @xmath455 . also , let @xmath456 and @xmath457 .
the two facts , ( [ fac1 ] ) and ( [ fac2 ] ) give us that @xmath458 finally , we have that @xmath459 and @xmath460 and both are greater than @xmath461 . we are done .
[ eqmlem ] let @xmath0 be a masa in a type @xmath1 factor @xmath2 and let @xmath3 and @xmath40 be two positive operators commuting with a projection @xmath43 in @xmath0 , written with respect to the decomposition @xmath409 , @xmath462 assume @xmath411 and @xmath463 and @xmath413 . then , there is a projection @xmath183 in @xmath0 with @xmath464 and a unitary @xmath22 such that @xmath465 we may assume that @xmath466 for otherwise there is nothing to prove .
let @xmath422 be the natural number such that @xmath467 ; note that @xmath468 .
now , choose a positive @xmath469 such that @xmath470 .
next , define the sequence of numbers @xmath471 using the following prescription : @xmath472 is such that @xmath473 and for @xmath474 , the number @xmath475 is such that @xmath476 now , define the sequence of projections @xmath477 by @xmath478 and then for @xmath474 , define @xmath479 we interpret these projections as lying in @xmath2 .
note that by ( [ pcal1 ] ) and ( [ pcal2 ] ) , we have , @xmath480 pick a unitary @xmath136 such that @xmath481 as well as @xmath482 and which commutes with @xmath43 , i.e. @xmath483 .
we then have , @xmath484 where @xmath485 for @xmath428 and @xmath486 . also , @xmath487 . applying lemma([sht2lem ] ) to @xmath488 and @xmath489 inside @xmath490 ,
we conclude that we may find a unitary @xmath430 commuting with @xmath491 and projections @xmath212 and @xmath492 of trace @xmath493 in @xmath0 with @xmath494 and @xmath495 such that letting @xmath496 , we have that @xmath497 and further , @xmath498 .
inductively , for @xmath499 , do the following : note that @xmath500 and apply lemma([sht2lem ] ) together with the remark ( [ greater ] ) following it to @xmath501 and @xmath502 inside @xmath503 .
the lemma yields a unitary @xmath504 commuting with @xmath505 and projections @xmath451 and @xmath506 of trace @xmath507 in @xmath0 with @xmath508 and @xmath509 such that letting @xmath510 , we have 1 .
@xmath511 2 . since @xmath512 , we have @xmath513 and hence , @xmath514 for @xmath515 as well .
3 . @xmath516 .
putting it all together , we have that @xmath517 where @xmath518 .
we have that @xmath519 and thus , @xmath520 we now turn to the main theorem of the paper , the proof of the conjecture ( [ conj2 ] ) of arveson and kadison in @xcite .
we start off with some preliminary remarks .
let @xmath57 and @xmath69 be the spectral scales of @xmath9 and @xmath20 respectively .
define @xmath521 choose a unitary that conjugates @xmath522 onto @xmath523 .
with respect to the decomposition @xmath524 , we may write @xmath525 then , @xmath526 and @xmath258 inside @xmath527 . it is now easy to see that if we can prove the theorem for @xmath144 and @xmath146 inside @xmath528 , the result for @xmath9 and @xmath20 inside @xmath2 would follow .
we may therefore assume that @xmath192 almost everywhere on @xmath64 $ ] .
let @xmath74 and @xmath204 be the ky fan norm functions .
the relation @xmath19 gives us that @xmath529 on @xmath64 $ ] .
define @xmath530 since we assume that @xmath192 almost everywhere on @xmath64 $ ] , @xmath531 can not contain any intervals .
we may write @xmath532 as a union of disjoint intervals @xmath533 ; pick a unitary @xmath22 that conjugates @xmath534 onto @xmath535 for every @xmath202 . then , @xmath536 where @xmath537 and further the corresponding ky fan norm functions are strictly positive on @xmath174 for every @xmath202 .
it is routine to see that if we can solve the problem for every @xmath202 , the general theorem follows .
therefore , we may assume , additionally to @xmath192 a.e . on @xmath64 $ ] , that @xmath252 on @xmath174
. [ honmaj ] we use the following notation : @xmath538 if we have that @xmath539 and also @xmath540 , we say that @xmath541 .
[ sh1part ] let @xmath3 and @xmath40 be positive operators with @xmath541 .
then , there is a projection @xmath43 in @xmath0 with @xmath542 and a unitary @xmath22 in @xmath2 such that @xmath132 and @xmath543 . by assumption , we have that @xmath544 on @xmath174
. choose numbers @xmath545 such that @xmath546 this can be done as follows ; recall that the functions @xmath204 and @xmath74 are continuous on @xmath64 $ ] .
let @xmath547\})$ ] .
now , let @xmath548 the function @xmath549 is greater than @xmath550 on @xmath551 $ ] and @xmath552 $ ] by the choice of @xmath553 and @xmath554 and is at least @xmath202 on @xmath555 $ ] .
assertion ( [ calc1 ] ) follows .
now , choose a unitary @xmath556 so that @xmath557 with respect to the decomposition @xmath558 , we write @xmath559 note that @xmath560 and @xmath561 .
now , @xmath562 it is easy to see that @xmath563 and thus , @xmath564dm(y)}{\tau(q_2)}\\ & = & \int_{\tau(q_1)}^{\tau(q_1)+x\tau(q_2 ) } [ f_{s}(y ) -f_{a}(y)]dm(y)\\ & = & [ f_s(\tau(q_1)+x\tau(q_2))-f_a(\tau(q_1)+x\tau(q_2))]-[f_s(a)-f_a(a)]\\ & > & 0\end{aligned}\ ] ] the last inequality is because for every @xmath76 $ ] , we have @xmath565 $ ] and thence because of ( [ calc1 ] ) .
we therefore have that @xmath566 .
similarly , we can prove that @xmath567 note further that @xmath568 apply theorem ( [ conj1 ] ) to the pair @xmath144 and @xmath146 inside @xmath569 .
we get a unitary @xmath570 of @xmath2 so that letting @xmath571 , we have @xmath572 noting that @xmath573 , we may choose a number @xmath574 in @xmath575 such that @xmath576 . recall that the spectral scales are right continuous .
thus , we may find an interval @xmath577 $ ] such that @xmath578 on @xmath85 .
now , choose a @xmath414 satisfying @xmath579,\end{aligned}\ ] ] as well as(by passing to a smaller @xmath469 if needed ) , @xmath580 let us define @xmath581 now choose a unitary @xmath222 that is the identity on @xmath582 such that @xmath583 now , with respect to @xmath584 , we may write @xmath585 compress to @xmath586 , and let @xmath587 , @xmath588 , @xmath589 note that we have the following , @xmath590 the first and fourth assertions follow from ( [ 1imp2 ] ) and the second and third from ( [ delcho ] ) . applying lemma([eqmlem ] ) to @xmath591 and @xmath246 , we get a projection @xmath43 in @xmath0 with @xmath592 and a unitary @xmath593 that is the identity on @xmath594 such that letting @xmath595 , we have , @xmath596 and also , @xmath597 let @xmath598 .
we have with respect to @xmath599 , @xmath600 where @xmath601 and also(again from the conclusion of lemma([eqmlem ] ) ) , @xmath602 let us write the operator @xmath603 as @xmath604 and @xmath605 as @xmath606 , that is , @xmath607 the last step in the proof is to show that @xmath608 the condition ( [ delcho2 ] ) , @xmath609 says that @xmath610 note further that @xmath611 . the condition on the spectra required for lemma ( [ lem1 ] ) is easily verified as well .
the lemma(the last line in the statement of the lemma ) gives us that @xmath543 .
[ cor2 ] let @xmath3 and @xmath40 be positive operators and suppose we have a projection @xmath43 in @xmath0 and a unitary @xmath22 in @xmath2 such that @xmath612 then , there is a projection @xmath183 in @xmath0 such that @xmath613 with @xmath614 and a unitary @xmath136 in @xmath2 such that @xmath615 this is proved in the same way that corollary([cor1 ] ) is deduced from proposition ( [ prop1 ] ) and we omit the proof , using proposition([sh1part ] ) in place of proposition ( [ prop1 ] ) .
corollary ( [ cor2 ] ) will imply the main schur - horn theorem .
the passage from a partial solution to the full solution of the problem can be done exactly as in the proof of theorem ( [ conj1 ] ) .
[ conj2][the schur - horn theorem in type @xmath1 factors ii ] let @xmath0 be a masa in a type @xmath1 factor @xmath2 .
if @xmath3 and @xmath40 are positive operators with @xmath19 . then , there is an element @xmath41 such that @xmath42 assume first that @xmath541 . using proposition ( [ sh1part ] ) and corollary ( [ cor2 ] )
, we may pick a sequence of projections @xmath616 in @xmath0 and a sequence of unitaries @xmath130 of @xmath2 such that @xmath617 as well as @xmath618 we may choose the @xmath619 so that @xmath620 yielding that @xmath621 .
it is now routine to see that the unitaries @xmath622 converge in the strong operator topology to a unitary that we denote @xmath22 and that we have @xmath623 for the general case , as in the discussion preceding proposition ( [ sh1part ] ) , we can find a unitary @xmath136 and a projection @xmath183 in @xmath0 so that with respect to @xmath624 , @xmath625 where @xmath526 and @xmath626
. then , there is a unitary @xmath22 of the form @xmath627 such that @xmath628 .
it is routine to see that the operator @xmath169 defined by @xmath629 is such that @xmath630 and that @xmath169 is in @xmath32 .
we record one consequence that emerged in the above proof separately .
[ uthm ] let @xmath0 be a masa in a type @xmath1 factor @xmath2 .
if @xmath3 and @xmath40 are positive operators with @xmath19 .
assume further that @xmath193 for all @xmath631 then , there is a untary @xmath22 so that @xmath632 this last theorem has a nice consequence ; there is no need to take the norm closure of the unitary orbit to achieve a desired diagonal when the diagonal has finite spectrum . to prove this theorem , we need choquet s notion of comparison of measures : given two regular borel measures @xmath55 and @xmath633 on @xmath49
, we say that @xmath634 if for every tuple of positive borel measures @xmath635 such that @xmath636 , there are positive borel measures @xmath637 such that @xmath638 and such that @xmath639 for @xmath640 .
let @xmath9 and @xmath20 be two positive operators in a type @xmath1 factor @xmath2 , with spectral measures @xmath63 and @xmath641 .
we let @xmath642 denote the scalar measure on @xmath49 given by @xmath643 where @xmath47 is any borel set and similarly for @xmath20 .
it is a basic fact that the following are equivalent for positive operators @xmath9 and @xmath20 in type @xmath1 factors , see @xcite , 1 .
@xmath19 2 .
@xmath644 . interpreting statement ( 2 ) above operator algebraically
, we see that @xmath19 is equivalent to saying that for every partition into projections , @xmath645 commuting with @xmath9 we have a partition into projections @xmath646 commuting with @xmath20 so that @xmath647 let @xmath2 be a type @xmath1 factor , @xmath0 a masa in @xmath2 , @xmath20 a positive operator in @xmath2 and @xmath9 a positive operator in @xmath0 with finite spectrum so that @xmath19
. then , there is a unitary @xmath22 so that @xmath648 .
let us first assume that @xmath9 is a scalar , that is , @xmath649 . if @xmath20 was a scalar as well , there is nothing to prove . let us therefore assume that @xmath650 .
since @xmath69 is non - increasing , we see that for any @xmath651 , @xmath652 with equality precisely when @xmath653 .
hence , @xmath654 on @xmath174 . theorem([uthm])gives us a unitary @xmath22 so that @xmath655 now suppose @xmath9 has finite spectrum ; there are projections @xmath656 summing up to @xmath85 and numbers @xmath657 so that @xmath658 . by choquet
s comparison of measures , there are projections @xmath659 commuting with @xmath20 and summing up to @xmath85 so that @xmath660 and @xmath661 for @xmath662 .
choose a unitary @xmath22 so that @xmath663 for @xmath662 .
we see that for every @xmath61 , @xmath23 commutes with @xmath664 and that we have @xmath665 . by the result for scalar diagonals
, we have projection @xmath666 in @xmath667 so that @xmath668 .
let @xmath669 ; it is easy to check that @xmath670
the schur - horn theorem in type @xmath1 factors allows us to quickly prove an analogous theorem for trace class operators in type @xmath7 factors .
one thing to note is that not all masas in type @xmath7 factors admit normal conditional expectations .
it is a result of takesaki@xcite that if all masas in a von neumann algebra admit normal conditional expectations , then the von neumann algebra is finite . masas in type @xmath7 factors that do admit normal conditional expectations
are generated by their finite projections - we will refer to these as atomic masas in analogy to @xmath46 . in @xcite ,
arveson and kadison proved a schur - horn theorem for trace class operators in @xmath46 ; we prove an exact analogue of their result here .
the proof follows from a routine reduction to the @xmath1 factor case , which we accomplish by let @xmath0 be a atomic masa in a type @xmath7 factor @xmath2 and let @xmath3 and @xmath40 be positive trace class operators so that @xmath19
. then , there is a unitary and a finite projection @xmath43 in @xmath0 so that @xmath23 commutes with @xmath43 and @xmath671 the proof is identical to the first part of the proof of theorem([conj1 ] ) and we omit it .
the lemma yields a straighforward corollary [ corti ] let @xmath0 be an atomic masa in a type @xmath7 factor @xmath2 and let @xmath3 and @xmath40 be positive trace class operators so that @xmath19 .
then , there is a unitary @xmath22 and a countable set of orthogonal finite projections @xmath672 in @xmath0 so that @xmath23 commutes with each projection @xmath673 and @xmath674 this a routine induction argument and we omit it .
recall that for trace class operators in type @xmath7 factors , we have defined @xmath675 as the closure of the unitary orbit in the trace norm , see ( [ tco ] ) .
it is wasy to see that for a positive operator inside a type @xmath1 factor , the closures in the operator norm and the trace norm coincide(with the set of operators equimeasurable to the given one ) .
the schur horn theorem for trace class operators in type @xmath7 factors is as follows [ shtin ] let @xmath0 be an atomic masa in a type @xmath7 factor @xmath2 and let @xmath3 and @xmath40 be positive trace class operators so that @xmath19 .
then , there is an operator @xmath41 so that @xmath676 where @xmath4 is the canonical @xmath28 preserving conditional expectation onto @xmath0 .
corollary([corti ] ) yields us a unitary @xmath22 and a countable set of orthogonal projections , @xmath672 so that @xmath677 and so that @xmath678 . applying theorem([conj2 ] ) to each of the @xmath1 factors @xmath679 yields us a set of operators @xmath680 such that @xmath681 .
now , let @xmath682 and fix an @xmath199 . since @xmath683 belongs to @xmath684 , for each @xmath685 , we can find a unitary @xmath686 in @xmath687 so that @xmath688 letting @xmath689 , we see that @xmath690 .
thus , @xmath169 belongs to @xmath675 and we are done .
another problem in this context is that of characterizing the images of operators , for instance projections , under the conditional expectation onto an atomic masa . in the case of @xmath691
, there are subtle index type obstructions that pop up@xcite .
the work of kadison was recently extended from projections to hermitians with finite spectrum by bownik and jasper in @xcite , @xcite and @xcite .
the complete characterisation that they obtain , while pleasing , is extremely subtle . in the type @xmath7 factor case , however , the situation is completely transparent .
we first show that any reasonable `` diagonal '' can be lifted to a projection .
let @xmath0 be an atomic masa in a type @xmath7 factor @xmath2 and let @xmath3 be a positive contraction .
then , there is a projection @xmath43 in @xmath2 so that @xmath44 .
write @xmath692 where @xmath693 are a family of orthgonal finite projections in @xmath0 .
then , @xmath694 is a type @xmath1 factor and we may find a projection @xmath333 in @xmath695 so that @xmath696
. then , letting @xmath697 , we have that @xmath44 .
we now turn things around and ask for a characterization of all possible diagonals of a given projection as well as that of positive operators in general .
we use the convention that if a positive operator is not trace class , then it s trace is @xmath698 .
argerami and massey in a recent paper@xcite proved approximate theorems in this context , which i am able to improve .
first , the result for projections .
let @xmath43 be a projection in a type @xmath7 factor @xmath2 and let @xmath3 be a positive contraction where @xmath0 is an atomic masa .
then , there is a unitary @xmath22 such that @xmath699 iff @xmath700 and @xmath701 .
if either @xmath702 or @xmath703 is finite , the theorem follows from theorem([shtin ] ) . for the other case , pick an orthogonal family of finite projections @xmath704 in @xmath0 summing up to the identity .
decompose @xmath705 and @xmath706 so that @xmath333 and @xmath693 are finite projections for every @xmath202 , and such that @xmath707 and @xmath708 .
pick a unitary @xmath22 that conjugates @xmath709 onto @xmath710 for every @xmath202 and by theorem([conj2 ] ) , pick unitaries @xmath711 in @xmath712 so that @xmath713 for every @xmath202 . then , if we let @xmath714 , we have that @xmath715 i now extend the above analysis to general positive operators .
let @xmath40 and @xmath3 be positive operators .
for there to exist a @xmath169 in @xmath675 such that @xmath716 , it is necessary that @xmath19(see ( [ maj2i ] ) for the definition of majorization between general positive operators in type @xmath7 factors ) .
however , this is not enough .
for example , let @xmath9 be a projection such that both @xmath9 and @xmath717 have infinite trace .
let @xmath718 be a sequence of trace @xmath356 projections indexed by the rationals in @xmath719 summing upto @xmath85 and let @xmath20 be the operator @xmath720 .
then , for both @xmath9 and @xmath20 , the upper and lower spectral scales are the constant functions @xmath356 and @xmath404 respectively .
it is easy to see that if there is a positive operator @xmath169 such that @xmath716 , then @xmath169 must equal @xmath9 .
however , @xmath9 is not in @xmath675 .
let @xmath721 be the ideal of @xmath28 finite rank operators , @xmath722 and let @xmath723 be the norm closed two sided ideal of @xmath28 compact operators@xcite .
let @xmath724 be the generalized calkin algebra @xmath725 and let @xmath726 and @xmath727 be the essential spectra of @xmath20 and @xmath9 , namely the spectra when projected down into @xmath724 . the majorization relation @xmath19 will force @xmath728 .
the above example shows that we need additional constraints on the essential point spectra of @xmath9 and @xmath20 .
we have the following theorem , whose proof is not too hard - it involves a standard cut and paste argument and a use of theorem([conj2 ] ) and we omit it .
let @xmath729 be a positive operator in a type @xmath7 factor @xmath2 and let @xmath3 be a positive operator where @xmath0 is an atomic masa .
then , there is a @xmath169 in @xmath675 such that @xmath716 iff 1 .
we have that @xmath19 . and
further , 2 .
if @xmath730 and if @xmath731 belongs to the essential point spectrum of @xmath9 , then it belongs to the essential point spectrum of @xmath20 as well . and , 3
if @xmath732 and if @xmath733 belongs to the essential point spectrum of @xmath9 , then it belongs to the essential point spectrum of @xmath20 as well .
it is routine to extend the schur - horn theorem to general finite von neumann algebras .
let @xmath2 be a type @xmath1 von neumann algebra and let @xmath0 be a masa in @xmath2 . instead of working with a tracial state
, we must now work with the center valued trace @xmath28 .
majorization is defined analogously to the case of type @xmath1 factors .
the schur - horn theorem in this case is let @xmath0 be a masa in a type @xmath1 von neumann algebra @xmath2 .
if @xmath3 and @xmath40 are positive operators with @xmath19 . then , there is an element @xmath41 such that @xmath734 alternately , we have that @xmath735 this can be proved exactly as in the factor case by first getting a local version and then using induction .
the proof is a standard application of the direct integral decomposition of @xmath2 into type @xmath1 factors and an argument analogous to the proof of theorem([conj2 ] ) and we omit it .
the situation when it comes to type @xmath8 factors is far simpler than that for semifinite factors .
one point to be noted is that the norm and sot closures of the unitary orbits of a hermitian operator in this case are different , unlike the type @xmath1 case .
for instance , the norm closure of the unitary orbit of a non - trivial projection is the set of all non - trivial projections , while the sot closure contains in addition , the identity projection @xmath85 and the zero projection @xmath404 .
we will reserve the term @xmath32 for the norm closure of the unitary orbit .
the proof of the following is again a simple adaptation of the proof of theorem ( [ conj2 ] ) and i omit it . 1 .
there is an operator @xmath41 so that @xmath736 .
2 . the following spectral conditions are satisfied 1 . @xmath737 .
2 . if @xmath738 is in the point spectrum of @xmath9 , then it is also in the point spectrum of @xmath20 .
similarly for @xmath739 .
william arveson and richard v. kadison , _ diagonals of self - adjoint operators _ , operator theory , operator algebras , and applications , contemp .
414 , amer .
soc . , providence , ri , 2006 , pp . 247263 .
fumio hiai , _ spectral majorization between normal operators in von neumann algebras _ , operator algebras and operator theory , pitman res .
notes math .
vol 271 , longman sci . tech . , harlow , 1992 , pp . | a few years ago , richard kadison thoroughly analysed the diagonals of projection operators on hilbert spaces and asked the following question : let @xmath0 be a masa in a type @xmath1 factor @xmath2 and let @xmath3 be a positive contraction .
letting @xmath4 be the canonical normal conditional expectation from @xmath2 to @xmath0 , can one find a projection @xmath5 so that @xmath6 in a later paper , kadison and arveson , as an extension , conjectured a schur - horn theorem in type @xmath1 factors . in this paper ,
i give a proof of this conjecture of arveson and kadison .
i also prove versions of the schur - horn theorem for type @xmath7 and type @xmath8 factors as well as finite von neumann algebras .
dedicated to the memory of william arveson(1934 - 2011 ) |
quarks can be associated in ads / cft with strings stretching between a probe d7 brane and the d3 branes from which the @xmath3 sym gauge fields originate @xcite . in the probe approximation
the d7 branes lie in the @xmath12 field theory directions , @xmath13 and @xmath14 coordinates of the transverse space in the ads@xmath4 metric @xmath15 a flat d7 brane that intersects the d3 at the origin would lie at @xmath16 and describe massless quarks .
the dbi action is @xmath17 where the dilaton factor is constant in pure ads but is non - zero in the presence , for example , of a supersymmetry breaking u(1)@xmath5 magnetic field @xcite or one could imagine it as some running gauge coupling @xcite .
one might now substitute the metric , integrate over the inert @xmath14 and expand in small @xmath18 to study the instability of the brane from the flat massless state @xcite .
one essentially finds @xcite . here
the scale is set by the value of @xmath19 at which the bf bound is violated ( an approximation to @xmath20 being set to 1 ) . , width=226 ] @xmath21 , \label{daq } \end{array}\ ] ] @xmath22 is just the expansion coefficient from the dilaton term .
there are some sleights of hand here to generate a consistent phenomenological model : * one should , formally , expand the factors of @xmath23 ( which contain @xmath18 ) but they are retained for two reasons .
if we simply write @xmath24 then if this term trips us through the bf bound in some range of small @xmath13 then the instability would exist however large @xmath25 were to grow .
if we instead promote @xmath13 in this term to @xmath23 then at sufficiently large @xmath26 the instability is no longer present and a full solution for @xmath26 is possible . in addition
when one looks at fluctuations about the background embedding those states only know about the non - trivial embedding through the factors of @xmath26 in the metric - these terms communicate the existence of a gap to the bound states .
we are of course being inconsistent about higher order terms in @xmath18 but the key point is to keep the @xmath27 term that triggers the bf bound violation and the brutality of our other assumptions will be judged by the success of the model below .
* we have added an axial gauge field in symmetry with the vector field whilst in the full d3/d7 system this state is more complex .
* we have allowed a free 5d gauge coupling @xmath28 which we expect to become a phenomenological parameter with the breaking of supersymmetry .
we will impose appropriate gauge dynamics on our model by fixing the form of @xmath29 using the one loop qcd runing with @xmath10 and @xmath11 .
@xmath30 the one loop result for the anomalous dimension is @xmath31 we will identify the rg scale @xmath32 with the ads radial parameter @xmath23 in our model .
working perturbatively from the ads result @xmath33 @xcite we have @xmath34 we call the scale where the bf bound is first violated @xmath35 and it is a rough measure of the traditional qcd scale @xmath20 the vacuum structure of the theory is found by setting all fields except @xmath25 to zero .
the euler - lagrange equation for the determination of @xmath26 , in the case of a constant @xmath22 , is @xmath36 - \rho \delta m^2 l = 0\,.\ ] ] we can now ansatz the @xmath23 dependent @xmath22 above to describe the running of the dimension of @xmath37 ( we do this at the level of the equation of motion ) . to find numerical solutions
we need an ir boundary condition . in top down models @xmath38 is the condition for a regular solution .
since we do not wish to describe ir physics below the quark mass ( where the quark contribution to the running coupling will decouple ) we use a very similar on - shell condition - we shoot from points @xmath39 with @xmath40 . in the uv
the solution ( neglecting @xmath22 which falls close to zero ) takes the form @xmath41 where @xmath42 in interpreted as the uv quark mass and @xmath43 as the quark condensate .
we show sample embeddings in fig [ embedding ] for different choices of @xmath44 .
the spectrum of the theory is determined by looking at linearized fluctuations of the fields about the vacuum where fields generically take the form @xmath45 .
a sturm - louville equation results for @xmath46 leading to a discrete spectrum . by substituting the wave functions back into the action and integrating over @xmath13 the decay constants can also be determined .
the normalizations of the fluctuations are determined by matching to the gauge theory expectations for the vv , aa and ss correlators in the uv of the theory .
this full procedure is described in detail in @xcite . with @xmath8 and
@xmath9 fixed the free parameters in the theory are the overall scale @xmath35 , the uv quark mass and the 5d coupling @xmath28 .
for example one can fix @xmath35 by scaling to give the correct @xmath47 ; the remaining parameters can then be fitted to the data .
we choose to minimize the maximum deviation @xmath48 in any observable and find a good fit at @xmath49 at a uv scale of @xmath50 and @xmath51 : [ cols="^,^,^",options="header " , ] all the mesonic variables lie within @xmath52 of the experimental centrepoints shown except for the @xmath53 meson mass that lies very close to the first excited @xmath54 state .
the lighter @xmath55 is thought to potentially be a mesonic molecule @xcite which might explain the discrepancy . in anycase our model provides a sufficiently close spectrum match to begin an analysis of njl dynamics in the model . with @xmath56 from bottom to top .
the addition of an njl term generates a minimum of the potential that tracks to @xmath57 at @xmath58 .
all dimensionful objects are expressed in terms of @xmath35 .
, width=226 ]
consider a free fermion with a four fermion interaction @xmath59 . in the standard njl approximation
there are two contributions to the effective potential @xcite .
first there is the one loop coleman weinberg potential @xcite for the free quarks @xmath60 this falls with growing @xmath42 and is unbounded , although normally one treats @xmath42 as a fixed parameter so one would not seek to minimize this potential .
when we add the four fermion term we allow @xmath42 to become dynamically determined but there is the extra term from the four fermion interaction evaluated on @xmath61 @xmath62 this makes the effective potential bounded and ensures a minimum . for small @xmath63
the extra term is large and the minimum is at @xmath57 .
when @xmath63 rises above @xmath64 the minimum lies away from @xmath57 .
the phase transition is second order . in witten s prescription for `` multi - trace '' operators @xcite we add the equivalent of the extra potential term ( [ njlextra ] ) as a boundary term at the uv cut off @xmath1 . for large @xmath1 where @xmath65
the term we add is @xmath66 the effective potential from the background model is computed by evaluating minus the action ( [ daq ] ) evaluated on the vacuum solution as a function of the uv mass term .
we extract the values of @xmath42 and @xmath43 in the uv by fitting to the form ( [ uvform ] ) near the cut off . in fig [ potential ]
we plot the effective potential in the holographic description of the @xmath10 and @xmath11 gauge theory showing that like ( [ colemanw ] ) it is unbounded with @xmath42 .
when we add the potential term in ( [ bound ] ) a minimum at non - zero @xmath42 is found .
the minimum tracks to @xmath57 at @xmath58 indicating the crossover nature of the transition . and
@xmath67 ) and @xmath53 meson ( @xmath68 ) masses against the njl coupling constant @xmath63 for choices of uv cut off @xmath69 .
, title="fig:",width=302 ] + and @xmath67 ) and @xmath53 meson ( @xmath68 ) masses against the njl coupling constant @xmath63 for choices of uv cut off @xmath69 .
, title="fig:",width=302 ] + and @xmath67 ) and @xmath53 meson ( @xmath68 ) masses against the njl coupling constant @xmath63 for choices of uv cut off @xmath69 . , title="fig:",width=302 ] we can also understand witten s prescription in terms of a change to the uv boundary condition on the solution of the embedding equation . varying the action gives @xmath70 normally in the uv one would require the mass to be fixed and @xmath71 to satisfy the boundary condition but now we allow @xmath26 to change and instead impose @xmath72 where we have included the variation of the surface term . for our action @xmath73 . assuming ( [ uvform ] ) we find that we need @xmath74 this condition is simpler to apply to the solutions of the euler lagrange equation than constructing and minimizing the effective potential but equivalent .
we can now study the mesonic variables of our holographic model in the presence of an njl interaction . in fig [ mvvsg ] and
fig [ fvvsg ] we display respectively the meson masses ( @xmath75 ) and the decay constants ( @xmath76 ) of the su(3 ) gauge theory with @xmath11 .
when the njl coupling @xmath58 the model generates the output of the table above and corresponds to our description of qcd .
results are shown in the figures for cut offs @xmath69 .
as @xmath63 grows at the cut off the njl interaction enhances the mass gap of the theory . at larger @xmath63 the njl interaction dominates and generates a much larger gap that sharply grows to the cut off scale
this clearly shows the anticipated cross over behaviour .
as the cut off is taken large relative to @xmath20 the transition becomes sharper and moves closer to the second order njl only behaviour .
the holographic model reproduces the expected physics well .
interestingly there is a small dip in the sigma meson mass before the njl interaction dominates although it is not a large effect .
meson decay constants against the njl coupling constant @xmath63 for choices of uv cut off @xmath69 . ,
title="fig:",width=302 ] + meson decay constants against the njl coupling constant @xmath63 for choices of uv cut off @xmath69 .
, title="fig:",width=302 ] + meson decay constants against the njl coupling constant @xmath63 for choices of uv cut off @xmath69 .
, title="fig:",width=302 ] the success in finding a holographic description of the gauged njl model opens up the possibility of doing a range of beyond the standard model physics including extended techniolour @xcite and top condensation @xcite interactions .
we will explore the phenomenology of these models in detail in an upcoming paper @xcite .
k. yamawaki , hep - ph/9603293 .
y. nambu and g. jona - lasinio , phys .
* 122 * ( 1961 ) 345 .
doi:10.1103/physrev.122.345 d. j. gross and f. wilczek , phys .
d * 8 * ( 1973 ) 3633 .
doi:10.1103/physrevd.8.3633 s. catterall , r. galvez , j. hubisz , d. mehta and a. veernala , phys .
d * 86 * ( 2012 ) 034502 doi:10.1103/physrevd.86.034502 [ arxiv:1112.1855 [ hep - lat ] ] ; s. catterall and a. veernala , phys .
d * 87 * ( 2013 ) no.11 , 114507 doi:10.1103/physrevd.87.114507 [ arxiv:1303.6187 [ hep - lat ] ] .
v. a. miransky , t. nonoyama and k. yamawaki , mod .
lett . a * 4 * ( 1989 ) 1409 .
doi:10.1142/s021773238900160x t. takeuchi , phys .
d * 40 * ( 1989 ) 2697 .
doi:10.1103/physrevd.40.2697 s. f. king and d. a. ross , phys .
b * 236 * ( 1990 ) 327 .
doi:10.1016/0370 - 2693(90)90991-e j. rantaharju , v. drach , c. pica and f. sannino , phys .
d * 95 * ( 2017 ) no.1 , 014508 doi:10.1103/physrevd.95.014508 [ arxiv:1609.08051 [ hep - lat ] ] .
t. appelquist , m. einhorn , t. takeuchi and l. c. r. wijewardhana , phys .
b * 220 * ( 1989 ) 223 .
doi:10.1016/0370 - 2693(89)90041 - 5 j. m. maldacena , adv .
* 2 * ( 1998 ) 231 [ hep - th/9711200 ] ; e. witten , adv .
* 2 * ( 1998 ) 253 [ hep - th/9802150 ] ; s. s. gubser , i. r. klebanov and a. m. polyakov , phys .
b * 428 * ( 1998 ) 105 [ hep - th/9802109 ] .
a. karch and e. katz , jhep * 0206 * ( 2002 ) 043 [ arxiv : hep - th/0205236 ] ; m. grana and j. polchinski , phys .
d * 65 * ( 2002 ) 126005 [ arxiv : hep - th/0106014 ] ; m. bertolini , p. di vecchia , m. frau , a. lerda and r. marotta , nucl .
b * 621 * ( 2002 ) 157 [ arxiv : hep - th/0107057 ] ; m. kruczenski , d. mateos , r. c. myers and d. j. winters , jhep * 0307 * ( 2003 ) 049 [ hep - th/0304032 ] ; j. erdmenger , n. evans , i. kirsch , and e. threlfall , _ eur
. phys . j. _ * a35 * ( 2008 ) 81133 , [ arxiv:0711.4467 ] .
r. alvares , n. evans and k. y. kim , phys .
d * 86 * ( 2012 ) 026008 doi:10.1103/physrevd.86.026008 [ arxiv:1204.2474 [ hep - ph ] ] .
m. jarvinen and e. kiritsis , jhep * 1203 * ( 2012 ) 002 [ arxiv:1112.1261 [ hep - ph ] ] ; d. arean , i. iatrakis , m. jarvinen and e. kiritsis , phys .
b * 720 * ( 2013 ) 219 doi:10.1016/j.physletb.2013.01.070 [ arxiv:1211.6125 [ hep - ph ] ] ; d. aren , i. iatrakis , m. jarvinen and e. kiritsis , jhep * 1311 * ( 2013 ) 068 doi:10.1007/jhep11(2013)068 [ arxiv:1309.2286 [ hep - ph ] ] ; m. jarvinen , jhep * 1507 * ( 2015 ) 033 doi:10.1007/jhep07(2015)033 [ arxiv:1501.07272 [ hep - ph ] ] .
d. kutasov , j. lin and a. parnachev , nucl .
b * 858 * ( 2012 ) 155 [ arxiv:1107.2324 [ hep - th ] ] .
m. jarvinen and f. sannino , jhep * 1005 * ( 2010 ) 041 [ arxiv:0911.2462 [ hep - ph ] ] .
p. breitenlohner and d. z. freedman , annals phys .
* 144 * ( 1982 ) 249 .
t. alho , n. evans and k. tuominen , phys .
d * 88 * ( 2013 ) 105016 doi:10.1103/physrevd.88.105016 [ arxiv:1307.4896 [ hep - ph ] ] .
n. evans and k. tuominen , arxiv:1302.4553 [ hep - ph ] ; n. evans and m. scott , phys .
d * 90 * ( 2014 ) no.6 , 065025 doi:10.1103/physrevd.90.065025 [ arxiv:1405.5373 [ hep - ph ] ] ; j. erdmenger , n. evans and m. scott , phys .
d * 91 * ( 2015 ) no.8 , 085004 doi:10.1103/physrevd.91.085004 [ arxiv:1412.3165 [ hep - ph ] ] .
e. witten , hep - th/0112258 .
n. evans and k. y. kim , phys . rev .
d * 93 * ( 2016 ) no.6 , 066002 doi:10.1103/physrevd.93.066002 [ arxiv:1601.02824 [ hep - th ] ] . j. t. londergan , j. nebreda , j. r. pelaez and a. szczepaniak , phys .
b * 729 * ( 2014 ) 9 doi:10.1016/j.physletb.2013.12.061 [ arxiv:1311.7552 [ hep - ph ] ] .
g. jona - lasinio , nuovo cim .
* 34 * ( 1964 ) 1790 .
doi:10.1007/bf02750573 s. r. coleman and e. j. weinberg , phys . rev .
d * 7 * ( 1973 ) 1888 .
doi:10.1103/physrevd.7.1888 w. a. bardeen , c. t. hill and m. lindner , phys .
d * 41 * ( 1990 ) 1647 .
doi:10.1103/physrevd.41.1647 e. farhi and l. susskind , phys .
* 74 * ( 1981 ) 277 .
doi:10.1016/0370 - 1573(81)90173 - 3 | the nambu jona - lasinio model of chiral symmetry breaking predicts a second order chiral phase transition .
if the fermions in addition have non - abelian gauge interactions then the transition is expected to become a crossover as the njl term enhances the ir chiral symmetry breaking of the gauge theory .
we study this behaviour in the holographic dynamic ads / qcd description of a non - abelian gauge theory with the njl interaction included using witten s multi - trace prescription .
we study the behaviour of the mesonic spectrum as a function of the njl coupling and the ratio of the uv cut off scale to the dynamical scale of the gauge theory .
the gauged nambu jona - lasinio ( njl ) model @xcite @xmath0 with a uv cut off @xmath1 , provides an interesting study of two competing chiral symmetry breaking interactions .
the njl term triggers symmetry breaking at a second order transition if the coupling exceeds a critical value of @xmath2 @xcite .
the qcd - like non - abelian gauge interactions run from asymptotic freedom to strong coupling in the infrared ( ir ) and again trigger chiral symmetry breaking @xcite . in combination the two mechanisms enhance each other and the njl transition is changed to a cross - over as the dynamics change from being dominated by the ir gauge theory to the ultraviolet ( uv ) njl term @xcite . over the years there has been interest in the model for possible roles in strongly coupled beyond the standard model physics but it remains an interesting system in itself .
it has been studied previously using gap equation @xcite and lattice techniques @xcite . here
we wish to study the model holographically @xcite .
quarks with couplings to conformal @xmath3 gauge dynamics can be described , in the quenched limit , by probe d7 branes in ads@xmath4 @xcite .
the dirac born infeld ( dbi ) action provides a very simple description of the mesonic spectrum . in a very small number of cases supersymmetry
can be broken by , for example , a background u(1)@xmath5 magnetic field @xcite or a running dilaton @xcite and chiral symmetry breaking is described .
the mechanism is simplest to understand if one expands around the chirally symmetric vacuum @xcite ( see also the work in @xcite ) .
the action becomes that of a scalar in ads ( the brane embedding ) with a mass squared that runs with radial distance in ads ( ie with energy scale ) and the instability to chiral symmetry breaking sets in when the breitenlohner freedman ( bf ) bound @xcite is violated in the ir .
using the ads / cft dictionnary @xcite this translates to saying that the anomalous dimension of the quark condensate , @xmath6 , runs into the ir until @xmath7 , at which point chiral symmetry breaking occurs .
it is irrisitible to take this very simple model of the quark dynamics and replace the running @xmath6 with a sensible guess for another theory where a full holographic description of the glue background does not exist ( note such a background might include backreaction of the quarks themselves ) .
the dynamic ads / qcd model @xcite is such a model with the perturbative running for an su(@xmath8 ) gauge theory with @xmath9 quarks inserted . if one simply takes the one - loop expressions for @xmath6 for @xmath10 and @xmath11 naively extended to the scale where @xmath7 , a very sensible description of qcd is obtained ( inspite of the very simplified choice of operators involved in the dynamics and the neglect of any stringy physics of the dual ) .
this is the holographic description of the gauge theory we will use here .
the njl interaction may be included using witten s double trace prescription @xcite and it has recently been shown how the basic njl second order transition can be achieved holographically in non - supersymmetric duals @xcite . thus we will be able to display here the behaviour in a number of mesonic variables of the parameter space of the gauged njl model .
below we will introduce dynamic ads / qcd , witten s prescription and then present our computations for the model . |
ime - domain simulation is an important approach for power system dynamic analysis .
however , the complete system model , or interchangeably the long - term stability model , typically includes different components where each component requires several differential and algebraic equations ( dae ) to represent , at the same time , these dynamics involve different time scales from millisecond to minute . as a result
, the total number of dae of a real power system can be formidably large and complex such that time domain simulation over long time intervals is expensive@xcite .
these constraints are even more stringent in the context of on - line stability assessment .
intense efforts have been made to accelerate the simulation of long - term stability model .
one approach is to use a larger time step size to filter out the fast dynamics or use automatic adjustment of step size according to system behavior in time - domain simulation @xcite@xcite@xcite from the aspect of numerical method .
another approach is to implement the quasi steady - state ( qss ) model in long - term stability analysis @xcite@xcite from the aspect of model approximation .
nevertheless , the qss model suffers from numerical difficulties when the model gets close to singularities which were addressed in @xcite-@xcite .
moreover , the qss model can not provide correct approximations of the long - term stability model consistently as numerical examples shown in @xcite@xcite .
in addition , sufficient conditions of the qss model were developed in @xcite which pointed to a direction to improve the qss model . as a result
, the qss model requires improvements in both model development and numerical implementation .
this paper contributes to the latter one . in this paper
, we apply pseudo - transient continuation ( @xmath0 ) which is a theoretical - based numerical method in power system long - term stability analysis .
pseudo - transient continuation method can be implemented directly in the long - term stability model to accelerate simulation speed compared with conventional implicit integration method . on the other hand
, the method can also be applied in the qss model to overcome possible numerical difficulties due to good stability property .
this paper is organized as follows .
section [ sectiondyptc ] briefly reviews general pseudo - transient continuation method in dae system .
section [ sectionptcinpowersystem ] includes a introduction about power system models followed by implementation of pseudo - transient continuation method in the long - term stability model and the qss model respectively .
section [ sectionnumerical ] presents three numerical examples to show the feasibility of the method . and
conclusions are stated in section [ sectionconclusion ] .
pseudo - transient continuation is a physically - motivated method and can be used in temporal integration .
the method follows the solution of dynamical system accurately in early stages until the steady state is approaching .
the time step is thereafter increased by sacrificing temporal accuracy to gain rapid convergence to steady state @xcite .
if only the steady state of a dynamical system instead of intermediate trajectories is of interest , pseudo - transient continuation method is a better choice than accurate step - by - step integration . on the other hand ,
compared with methods that solve nonlinear equations for steady state such as line - search and trust region methods , pseudo - transient continuation method can avoid converging to nonphysical solutions or stagnating when the jacobian matrix is singular .
this is particularly the case when the system has complex features such as discontinuities which exist in power system models .
therefore , @xmath0 method can be regarded as a middle ground between integrating accurately and calculating the steady state directly .
@xmath0 method can help reach the steady state quickly while maintain good accuracy for the intermittent trajectories .
for ode dynamics , sufficient conditions for convergence of @xmath0 were given in @xcite .
the results were further extended the semi - explicit index - one dae system in @xcite .
we recall the basic algorithm here .
we consider the following semi - explicit index - one dae system : @xmath1 with initial value @xmath2 . here
@xmath3 , @xmath4 , @xmath5^t \in \re^{n_1+n_2}$ ] , and @xmath6 where @xmath7 is a nonsingular scaling matrix .
we assume the initial condition for ( [ dae ] ) is consistent , i.e. @xmath8 and seek to find the equilibrium point @xmath9 such that @xmath10 and satisfies @xmath11 . as stated before , the step - by - step integration is too time consuming if the intermediate states are not of interest . on the other hand ,
newton s method for @xmath12 alone usually fails as the initial condition is not sufficiently near the equilibrium point .
the @xmath0 procedure is defined by the iteration : @xmath13 where @xmath14 is adjusted to efficiently find @xmath9 rather than to enforce temporal accuracy .
the convergence results in @xcite@xcite assume that the time step is updated with `` switched evolution relaxation''(ser ) : @xmath15 the algorithm is shown as below : set @xmath16 and @xmath17 .
evaluate @xmath18 . while @xmath19 is too large .
solve @xmath20 .
set @xmath21 .
evaluate @xmath18 .
update @xmath22 according to ( [ delta ] ) .
step 2.a is a newton step which is typically solved by an iterative method which terminates on small linear residuals while it may also be solved by inexact newton iteration .
note that the failure of @xmath0 usually can be well signaled by reaching the bound on the total number of iterations @xcite .
the convergence of @xmath0 for smooth @xmath23 was proved in @xcite under the assumptions that the dae has index one in a certain uniform sense , that it has a global solution in time , and that the solution converges to a steady state .
the result were further extended to nonsmooth @xmath23 in @xcite with @xmath24 in ( [ ptc ] ) replaced by a generalized derivative .
next we explain why @xmath0 has a better stability property .
firstly , conventional integration methods insist on a small norm of the linear residual at each step and will either converge , diverge to infinity , or stagnate at a point where iteration matrix is singular .
however , @xmath0 method will accept an increase in the residual , responding to that increase by decreasing @xmath22@xcite .
in addition , @xmath0 stems from backward euler method which is an attractive choice when stability is the desired property instead of accuracy .
one may think of @xmath0 method as a predictor - corrector method where the simple predictor is from previous time step and the corrector is backward euler with newton iteration @xcite . to see this , consider the implicit euler step from @xmath25 with @xmath26 , @xmath27 thus , @xmath28 is the root of the following equation : @xmath29 the newton s method to find the root of the above equation is : @xmath30 if we take @xmath31 , then by binomial inverse theorem , the first newton iterate is : @xmath32 which is exactly @xmath0 step .
as @xmath0 method has a better stability property , it can be applied in the qss model when conventional integration method fails to converge .
in this section , we firstly introduce power system models in long - term stability analysis .
then we apply @xmath0 method to the long - term stability model with modifications .
finally , we present an algorithm of @xmath0 method in the qss model to overcome possible numerical difficulties . the long - term stability model for calculating system dynamic response relative to a disturbance can be described as : @xmath33 equation ( [ algebraic eqn ] ) describes the transmission system and the internal static behaviors of passive devices , and ( [ fast ode ] ) describes the internal dynamics of devices such as generators , their associated control systems , certain loads , and other dynamically modeled components . @xmath34 and @xmath35 are continuous functions , and vector @xmath36 and @xmath37 are the corresponding short - term state variables and algebraic variables . besides , equations ( [ slow ode ] ) and ( [ slow dde ] ) describe long - term dynamics including exponential recovery load , turbine governor , load tap changer ( ltc ) , over excitation limiter ( oxl ) , etc . @xmath38 and @xmath39 are the continuous and discrete long - term state variables respectively , and @xmath40 is the maximum time constant among devices .
note that shunt compensation switching and ltc operation are typical discrete events captured by ( [ slow dde ] ) and @xmath41 is shunt susceptance and the transformer ratio correspondingly .
transitions of @xmath41 depend on system variables , thus @xmath41 change values from @xmath42 to @xmath43 at distinct times @xmath44 where @xmath45 , otherwise , these variables remain constants . since short - term dynamics have much smaller time constants compared with those of long - term dynamics , @xmath46 and @xmath41 are also termed as slow state variables , and @xmath47 are termed as fast state variables . if we represent the long - term stability model and the qss model in @xmath48 time scale , where @xmath49 , and we denote @xmath50 as @xmath51 , then the long - term stability model of power system can be represented as : @xmath52 where the study region @xmath53 , and @xmath54 , @xmath55 , @xmath56 , @xmath57 . at the same time , the qss model can be represented as : @xmath58 moreover , the long - term stability model ( [ complete ] ) can be regarded as two decoupled systems ( [ couple1 ] ) and ( [ couple2 ] ) showed as below when @xmath41 jump from @xmath42 to @xmath43 : @xmath59 and @xmath60 discrete variables @xmath41 are updated first and then system ( [ couple2 ] ) works with fixed parameters @xmath41 .
similarly , when @xmath41 jump from @xmath42 to @xmath43 , the qss model ( [ qss ] ) can be decoupled as : @xmath61 and @xmath62 assuming @xmath63 is nonsingular , then the long - term stability model ( [ couple2 ] ) with @xmath41 fixed as parameters is a semi - explicit index-1 dae system .
@xmath0 method requires initial condition to satisfy the algebraic equations , however , the discrete equation ( [ couple1 ] ) will violate this condition whenever it works . as a result
, we need to modify the original @xmath0 method for its implementation in long - term stability model .
in power system long - term stability model , @xmath64^t$ ] , @xmath65^t\in\re^{p+m+n}$ ] , @xmath66 , where @xmath67 is the identiy matrix of size @xmath68 . in order to make the initial condition of @xmath0 consistent , we switch back to implicit integration method whenever discrete variables jump and set the step length to be @xmath69 .
moreover , @xmath0 method is implemented for the post - fault system starting from @xmath70several seconds after the contingency . in examples of this paper
, @xmath70 was set to be @xmath71 .
the proposed algorithm is shown as below .
run the long - term stability model up to @xmath70 by implicit integration method .
set the value @xmath72 at @xmath70 as the initial condition @xmath73 of @xmath0 , and set @xmath17 . while @xmath74 is too large .
if discrete variables jump at @xmath44 update @xmath41 according to ( [ couple1 ] ) .
set @xmath75 , @xmath17 .
solve the newton step @xmath76 .
set @xmath77 .
evaluate @xmath78 .
otherwise set @xmath79 .
solve the newton step @xmath80 .
set @xmath77 .
evaluate @xmath78 .
update @xmath22 according to ( [ delta ] ) .
note that @xmath81 and @xmath82 depend on the specific integration method used .
for instance , if implicit trapezoidal method is used , then @xmath83 @xmath84 assuming @xmath85 is nonsingular , then the qss model ( [ coupleqss2 ] ) with @xmath41 fixed as parameters is a semi - explicit index-1 dae system .
@xmath0 method requires initial condition to satisfy the algebraic equations , however , the discrete equation ( [ coupleqss1 ] ) will violate this condition whenever it works . as a result
, we need to modify the original @xmath0 method for its implementation in the qss model . in the qss model ,
@xmath64^t$ ] , @xmath65^t\in\re^{p+m+n}$ ] , @xmath86 , where @xmath67 is the identity matrix of size @xmath87 .
in order to make the initial condition of @xmath0 consistent , we switch back to implicit integration method whenever discrete variables jump and set the step length to be @xmath69 . besides , the qss model is implemented at @xmath88when short - term dynamics settle down after the contingency .
usually , @xmath88 can be set as 30s .
the proposed algorithm is shown as below .
run the long - term stability model up to @xmath88 by implicit integration method . set the value @xmath72 at @xmath88 as the initial condition @xmath73 of the qss model , and set @xmath17 .
start to run the qss model .
if the qss model has a numerical difficulty by using implicit integration method , then go to step 3 , otherwise , continue with the qss model . while @xmath74 is too large .
if discrete variables jump at @xmath44 update @xmath41 according to ( [ coupleqss1 ] ) .
set @xmath75 , @xmath17 .
solve the newton step @xmath76 .
set @xmath77 .
evaluate @xmath78 .
otherwise set @xmath79 .
solve the newton step @xmath80 .
set @xmath77 .
evaluate @xmath78 .
update @xmath22 according to ( [ delta ] ) . similarly , @xmath81 and @xmath82 depend on the specific integration method used .
if implicit trapezoidal method is used , then @xmath89 @xmath90
in this section , three examples are to be presented .
the first two examples were the same 145-bus system in which the qss model met numerical difficulties during simulation while the long - term stability model was stable in long - term time scale .
firstly , @xmath0 method was implemented in the long - term stability model and the speed was more than 7 times faster than the trapezoidal method . secondly ,
when the qss model by trapezoidal method met difficulty , @xmath0 method was implemented in the qss model and provided correct approximations and the speed was still more than 5 times faster than the long - term stability model by trapezoidal method . and in the last example which was a 14-bus system , the long - term stability model was unstable and @xmath0 method successfully captured the instability which was signaled by reaching the bound of maximum iteration in the newton step .
all simulations were done in psat-2.1.6@xcite .
the system was a 145-bus system @xcite .
there were exciters and power system stabilizers for each of generator 1 - 20 .
and there were turbine governors for each of generator 30 - 40 .
besides , there were three load tap changers at lines between bus 73 - 74 , bus 73 - 81 and bus 90 - 92 respectively .
the contingency was a line loss between bus 1 - 6 . in this case , the post - fault system was stable after the contingency in the long - term time scale and it took 122.39s for the time domain simulation of the long - term stability model by implicit trapezoidal method .
however , @xmath0 method only took 16.12s for the simulation of the long - term stability model which was about @xmath91 of the time consumed by trapezoidal method .
[ tdsptc ] shows that the trajectories by @xmath0 method followed closely to the trajectories by trapezoidal method , and finally both converged to the same long - term stable equilibrium point .
thus @xmath0 method provided correct approximations for the long - term stability model in terms of trajectories and stability assessment . method .
@xmath0 method provided correct approximations.,width=172 ] method .
@xmath0 method provided correct approximations.,width=172 ] method .
@xmath0 method provided correct approximations.,width=172 ] method .
@xmath0 method provided correct approximations.,width=172 ] in this example , the system was the same as the last case . however , the qss model met numerical difficulties at 40s when implicit trapezoidal method was used , thus @xmath0 method was implemented in the qss model starting from 40s .
[ tdsptcqss ] shows that the trajectories by @xmath0 method converged to the long - term stable equilibrium point which the long - term stability model converged to , and also provided good accuracy for the intermittent trajectories .
it took 21.75s for @xmath0 method which was about @xmath92 of the time consumed by the long - term stability model using implicit trapezoidal method . method .
@xmath0 method overcame numerical difficulties and provided correct approximations.,width=172 ] method .
@xmath0 method overcame numerical difficulties and provided correct approximations.,width=172 ] method .
@xmath0 method overcame numerical difficulties and provided correct approximations.,width=172 ] method .
@xmath0 method overcame numerical difficulties and provided correct approximations.,width=172 ] in this case , the 14-bus system was long - term unstable due to wild oscillations of fast variables .
the system was modified based on the 14-bus test system in psat-2.1.6@xcite .
there were three exponential recovery loads at bus 9 , 10 and 14 respectively and two turbine governors at generator 1 and 3 .
besides , there were over excitation limiters at all generators and three load tap changers at lines between bus 4 - 9 , bus 12 - 13 and bus 2 - 4 .
the system suffered from long - term instabilities and simulation by implicit trapezoidal method could not continue after 101.22s .
@xmath0 method also stopped at 103.34s when the bound on the total number of iterations for the newton step was reached .
thus @xmath0 method was able to capture instabilities of the long - term stability model . method .
@xmath0 method was able to capture instabilities of the long - term stability model.,width=172 ] method .
@xmath0 method was able to capture instabilities of the long - term stability model.,width=172 ] method .
@xmath0 method was able to capture instabilities of the long - term stability model.,width=172 ] method .
@xmath0 method was able to capture instabilities of the long - term stability model.,width=172 ]
in this paper , modified @xmath0 methods for the long - term stability model and the qss model are given for power system long - term stability analysis with illustrative numerical examples .
we make use of the fast asymptotic convergence of @xmath0 method in the long - term stability model to achieve fast simulation speed . on the other hand , we take advantage of good stability property of @xmath0 method in the qss model to overcome numerical difficulties .
numerical examples show that @xmath0 can successfully provide correct stability assessment for the long - term stability model and overcome numerical difficulties in the qss model , as well as offer good accuracy for the intermediate trajectories .
@xmath0 can be regarded as a good in - between method with respect to integration and steady state calculation , thus serves as an alternative method in power system long - term stability analysis .
this work was supported by the consortium for electric reliability technology solutions provided by u.s .
department no .
de - fc26 - 09nt43321 .
kurita , a. , h. okubo , k. oki , s. agematsu , d. b. klapper , n. w. miller , w. w. price jr , j. j. sanchez - gasca , k. a. wirgau , and t. d. younkins , _ multiple time - scale power system dynamic simulation _ , ieee transactions on power systems , vol
216 - 223 , 1993 .
t. v. cutsem , m. e. grenier , d. lefebvre , _ combined detailed and quasi steady - state time simulations for large - disturbance analysis_. international journal of electrical power and energy systems , vol .
28 , issue 9 , pp .
634 - 642 , november 2006 .
x. z. wang , h. d. chiang , _ numerical investigations on quasi steady - state model for voltage stability : limitations and nonlinear analysis_. submitted to international transactions on electrical energy systems .
x. z. wang , h. d. chiang , _ analytical studies of quasi steady - state model in power system long - term stability analysis_. to appear in ieee transactions on circuits and systems i : regular papers .
doi 10.1109/tcsi.2013.2284171 . c. t. kelley , d. e. keyes _ convergence analysis of pseudo - transient continuation_. siam journal on numerical analysis , vol .
508 - 523 , april 1998 .
v. vittal , d. martin , r. chu , j. fish , j. c. giri , c. k. tang , f. eugenio villaseca , and r. g. farmer .
_ transient stability test systems for direct stability methods_. ieee transactions on power systems , vol . 7 , no .
37 - 43 , feb . 1992 | in this paper , pseudo - transient continuation method has been modified and implemented in power system long - term stability analysis .
this method is a middle ground between integration and steady state calculation , thus is a good compromise between accuracy and efficiency .
pseudo - transient continuation method can be applied in the long - term stability model directly to accelerate simulation speed and can also be implemented in the qss model to overcome numerical difficulties .
numerical examples show that pseudo - transient continuation method can provide correct approximations for the long - term stability model in terms of trajectories and stability assessment .
pseudo - transient continuation , long - term stability model , quasi steady - state model , long - term stability . |
presently planned @xmath0 linear collider ( lc ) projects will operate at an initial center of mass system ( cms ) energy of about 500 gev , with upgrades to higher energies designed in from the start .
the tev class colliders tesla @xcite and nlc / jlc @xcite target 800 gev and 1 - 1.5 tev , respectively , as their maximum cms energies . increasing the energy
further would require either a change in acceleration technology or an extension in accelerator length beyond the presently foreseen 30 - 40 km @xcite .
this would also increase the number of active elements , which will likely decrease the overall efficiency of such a facility .
the nature of the new physics which will hopefully be discovered and studied at the lhc and a tev class lc will determine the necessity and importance of exploring the multi - tev range with a precision machine such as an @xmath2 collider .
this paper summarizes the work of the e3 subgroup 2 on multi - tev colliders of the snowmass 2001 workshop ` the future of particle physics ' . based on our knowledge today
, the case for the multi - tev collider rests on the following physics scenarios : ( 200,1 ) ( 25,-357)_presented at the aps / dpf / dpb summer study on the future of particle physics ( snowmass 2001 ) , _ ( 120,-369)_30 june - 21 july 2001 , snowmass , colorado , usa _ *
the study of the higgs + for a light higgs , a multi - tev @xmath2 collider can access with high precision the triple higgs coupling , providing experimenters with the opportunity to measure the higgs potential .
the large event statistics will allow physicists to measure rare higgs decays such as @xmath3 .
for heavy higgses , predicted by e.g. supersymmetric models , the range for discovery and measurement will be extended for masses up to and beyond 1 tev .
* supersymmetry + in many susy scenarios only a subset of the new sparticles will be light enough to be produced directly at a tev class lc .
some of the heavier sparticles will be discovered at the lhc , but a multi - tev lc will be needed to complete the spectrum and to precisely measure the heavy sparticles properties ( flavor , mass , width , couplings ) . furthermore , polarized beams will help disentangle mixing parameters and aid cp studies . ultimately we _ will _ need to measure all sparticles as precisely as possible to fully pin down and test the underlying theory .
* new resonances + many alternative theories and models for new physics predict new heavy resonances with masses larger than 1 tev .
if these new resonances ( e.g. new gauge bosons , kaluza - klein resonances , or @xmath4 resonances ) have masses in the 1 tev - 5 tev range , a multi - tev collider becomes a particle factory , similar to lep for the @xmath5 .
the new particles can be produced directly and their properties can be accurately determined . *
no new particles + if _ no _ new particles are observed directly , apart from perhaps one light higgs particle , then a multi - tev collider will probe new physics indirectly ( extra dimensions , @xmath6 , contact interactions ) at scales in the range of 30 - 400 tev via precision measurements . * unexpected phenomena +
this is probably the most exciting of all : perhaps nature has chosen a road as yet not explored ( extensively ) by our imagination .
recent examples of new ideas are string quantum gravity effects , non - commutative effects , black hole formation , nylons , and split fermions .
.event rates for several processes in the multi - tev range , for 1 ab@xmath7 integrated luminosity . [ cols="^,^,^ " , ] [ tab : sum ]
linear @xmath0 colliders operating in the multi - tev energy range are likely to be based on the clic two - beam acceleration concept . to achieve a large luminosity
, such an accelerator would need to operate in the high beamstrahlung region , rendering experimentation at such a collider more challenging .
studies so far indicate that this is not a substantial handicap , and the precision physics expected from an @xmath0 collider will be possible .
the two - beam accelerator technology is not yet available today for use at a large scale collider .
r&d on this technology will continue until 2006 at least , after which if no bad surprises emerge one can plan for a full technical design of such a collider . from the physics program side
, a multi - tev collider has a large potential to push back the high energy horizon further , up to scales of 1000 tev , where if the higgs is light new physics can no longer hide from experiment . if no new scale is found by then we have to revise our understanding of nature
. a multi - tev collider with high luminosity can be used for precision measurements in the higgs sector .
it can precisely measure the masses and couplings of heavy sparticles , thereby completing the susy spectrum .
if extra dimensions or even black holes pop up in the multi - tev range , such a collider will be a precision instrument to study quantum gravity in the laboratory .
the physics reach , as envisioned today , for a multi - tev collider is summarized in table [ tab : sum ] . in short a collider with @xmath8 3 - 5 tev is expected to break new grounds , beyond the lhc and a tev class lc . , desy 2001 - 011 .
, slac - r-571 .
jlc-1 , jlc group , s matsumoto et al .
, kek report 92 - 16 .
p. burrows and r. patterson , _ lc expandability and upgradability _ , these proceedings .. m. battaglia , hep - ph/0103338 .
the clic study team , _ a 3 tev e@xmath9e@xmath10 linear collider based on clic technology _ , cern 2000 - 008 .
braun et al . , proc . of 18th international conference on high energy accelerators ( heacc2001 ) , 26 - 30 march 2001 , tsukuba , japan , and clic note 473
; see also http://geschonk.home.cern.ch/geschonk/ r. corsini et al . , cern / ps 2001 - 030 ( ae ) , and 2001 particle accelerator conference ( pac2001 ) , chicago , illinois , usa , june 18 - 22 , 2001 d. asner et al .
, hep - ex/0111056 v. telnov , _ photon colliders at multi - tev energies _ , these proceeedings .
r. assmann and f. zimmerman , _ polarization at clic _ , these proceeedings .
g.guignard , _ the clic study _ , these proceedings .
d. schulte , _ machine - detector interface at clic _ , these proceedings + d. schulte , cern - ps-99 - 066 .
r. settles , _ detector requirements at multi - tev lc _ , these proceedings .
m. battaglia , in proceedings of the lcws2000 workshop , p831 .
m. pohl and h.j .
schreiber , _ simdet - a parametric monte carlo for a tesla detector _
, desy 99 - 030
. m. battaglia , adaption of the delphi pvec package .
m. battaglia , e. boos , and w. yao , _ studying the higgs potential at the linear collider _ , hep - ph/0111276 .
m. battaglia , a.de roeck , _ determination of the muon yukawa coupling at high energy @xmath0 linear colliders _ , these proceedings , hep - ph/0111307 .
j. f gunion ( 1997 ) , hep - ph/9703203 .
t. plehn and d. rainwater , phys . lett .
* b520 * , 108 ( 2001 ) , hep - ph/017180 .
marco battaglia et al .
, hep - ph/0106204 . m. battaglia , a. kiiskinen and a. ferrari,_study of charged higgs bosons _ , these proceedings , hep - ex/0112015 .
g. blair , w. porod , p.m. zerwas phys . rev . *
d63 * ( 2001 ) 017703 , hep - ph/0007107 .
g. moortgat - pick , _ physics opportunities with polarized @xmath11 and @xmath12 beams at a linear collider _ , these proceedings
. g. wilson in proceedings of the lcws2000 workshop , p485 .
m. battaglia , private communication .
m. klasen , talk at the workshop on gamma - gamma colliders , chicago 2001 .
a. ferrari , _ study of majorana neutrinos at clic _ ,
these proceedings .
m. battaglia , s. riemann , s. de curtis and d. dominici , _ probing new scales at e+e- linear collider _ , these proceedings ; + battaglia et al .
, hep - ph/0101114 .
e. eichten , k lane and m. peskin , phys .
lett * 50 * ( 1983 ) 811 ; h. kroha , phys . rev . * d46 * ( 1992 ) 58 . t. hambye and k. riesselmann , phys .
* d55 * ( 1997 ) 7255 , hep - ph/9708416 .
t. barklow , private communication .
d. barger et al .
, phys rev . * d52 * ( 1995 ) 3815 , phys rev .
* d55 * ( 1997 ) 142 .
m. butterworth , b.e .
cox and j.r .
forshaw , in preparation .
a. dobado et al . , phys rev * d62 * ( 2000 ) 05501 .
a. de roeck , _ ww scattering at clic _
, these proceedings
. t. rizzo , hep - ph/0108235 .
n. arkani - hamed , s. dimopoulos , and g. dvali , b429 263 1998 , and d59 086004 1999 ; i. antoniadis , n. arkani - hamed , s. dimopoulos , and g. dvali , b436 257 1998 .
see , for example , i. antoniadis , b246 377 1990 ; i. antoniadis , c. munoz and m. quiros , b397 515 1993 ; i. antoniadis and k. benalki , b326 69 1994 and a15 4237 2000 ; i. antoniadis , k. benalki and m. quiros , b331 313 1994 .
l. randall and r. sundrum , 83 3370 1999 . for an introduction to add phenomenology , see g.f .
giudice , r. rattazzi and j.d .
wells , b544 3 1999 ; t. han , j.d .
lykken and r. zhang , phys . rev .
* d59 * , 105006 ( 1999 ) , e.a .
mirabelli , m. perelstein and m.e .
peskin , phys .
lett . * 82 * , 2236 ( 1999 ) ; j.l .
hewett , phys .
rev . lett . * 82 * , 4765 ( 1999 ) ; t.g .
rizzo , d60 115010 1999 .
m. battaglia , a. de roeck and t. rizzo , _ graviton production at clic _ , these proceedings , hep - ph/0112169 .
h. davousdiasl and t. rizzo , hep - ph/0104199 .
j hewett , f. petriello , and t. rizzo , hep - ph/0010354 .
j. lykken , talk at lcws2000 .
n. arkani - hamed , y. grossman , m. schmaltz , phys .
* d61 * ( 2000 ) 115004 , hep - ph/9909411 .
r. barbieri , l. j. hall , y. nomura , phys .
* d63 * ( 2001 ) 105007 , hep - ph/0011311 . | the physics at an @xmath0 linear collider with a center of mass energy of 3 - 5 tev is reviewed .
the following topics are covered : experimental environment , higgs physics , supersymmetry , fermion pair - production , @xmath1 scattering , extra dimensions , non - commutative theories , and black hole production . |
in recent years , the problem of anisotropic quantum scattering in two spatial dimensions ( 2d ) attracts increasing interest .
it is stimulated by the spectacular proposals for prospects to create exotic and highly correlated quantum systems with dipolar gases @xcite .
particularly , there were considered anisotropic superfluidity @xcite , 2d dipolar fermions @xcite , and few - body dipolar complexes @xcite .
the recent experimental production of ultracold polar molecules in the confined geometry of optical traps @xcite has opened up ways to realize these phenomena .
noteworthy also is a rather long history of research of 2d quantum effects in condensed matter physics .
one can note superfluid films @xcite , high - temperature superconductivity @xcite , 2d materials , such as graphene @xcite , and even possibilities for topological quantum computation @xcite .
unique opportunities for modeling these 2d effects in a highly controlled environment have recently appeared with the development of experimental techniques for creating quasi-2d bose and fermi ultracold gases @xcite .
interest in the processes and effects in 2d - geometry has stimulated the theory of elementary quantum two - body systems and processes in the plane .
special consideration should be given to the anisotropy and long - range character of the dipole - dipole interaction .
actually , usual partial - wave analysis becomes inefficient for describing the dipole - dipole scattering due to the strong anisotropic coupling of different partial - waves in the asymptotic region @xcite .
recently , considerable progress in the analysis of the 2d and quasi-2d ( q2d ) scattering of dipoles has been achieved @xcite .
thus , the 2d dipolar scattering in the threshold and semiclassical regimes was studied in the case of the dipole polarization directed orthogonally to the scattering plane @xcite .
an arbitrary angle of polarization was considered in @xcite . in this work ,
we develop a method for quantitative analysis of the 2d quantum scattering on a long - range strongly anisotropic scatterer .
particularly , it permits the description of the 2d collisions of unpolarized dipoles .
our approach is based on the method suggested in @xcite for the few - dimensional scattering which was successfully applied to the dipole - dipole scattering induced by an elliptically polarized laser field in the 3d free - space @xcite .
the key elements of the method are described in section ii . in section iii , we apply the method to the 2d scattering on the cylindrical potential with the elliptical base and the 2d dipole - dipole scattering of unpolarized dipoles .
we reproduce the threshold formula @xcite for the scattering amplitude on the cylinder potential with the circular base and the results of @xcite for the 2d scattering of polarized dipoles .
high efficiency of the method has been found in all problems being considered .
the last section contains the concluding remarks .
some important details of the computational scheme and illustration of the convergence are given in appendices .
the quantum scattering on the anisotropic potential @xmath0 in the plane is described by the 2d schrdinger equation in polar coordinates @xmath1 @xmath2 with the scattering boundary conditions @xmath3 in the asymptotic region @xmath4 and the hamiltonian of the system @xmath5 the unknown wave function @xmath6 and the scattering amplitude @xmath7 are searched for the fixed momentum @xmath8 defined by the colliding energy @xmath9 ( @xmath10 and the direction @xmath11 of the incident wave ( defined by the angle @xmath12 and for the scattering angle @xmath13 . here
@xmath14 is the reduced mass of the system . in the polar
coordinates , the angular part of the kinetic energy operator in @xmath15 has a simple form @xmath16 .
the interaction potential @xmath17 can be anisotropic in the general case , i.e. to be strongly dependent on @xmath13 .
it is clear that varying the direction of the incident wave @xmath11 can be replaced by the rotation @xmath18 of the interaction potential by the angle @xmath19 for the fixed direction of the incident wave , which we choose to be coincident with the x - axis .
thus , in the case of anisotropic potential @xmath17 the task is to solve the problem ( [ eq1 ] ) with the interaction potential @xmath20 for all possible @xmath19 and fixed @xmath9 with the scattering boundary conditions @xmath21 if the scattering amplitude @xmath7 is found , one can calculate the differential scattering cross section @xmath22 where @xmath23 , as well as the total cross section @xmath24 by averaging over all possible orientations @xmath19 of the scatterer and integration over the scattering angle @xmath13 . to integrate the problem ( [ eq1]),([eq2 ] ) , we use
the method suggested in @xcite to solving a few - dimensional scattering problem and applied in @xcite for the dipole - dipole scattering in the 3d free - space . following the ideas of these works we choose the eigenfunctions @xmath25 of the operator @xmath26 as a fourier basis for the angular - grid representation of the searched wave - function @xmath27 .
we introduce the uniform grid @xmath28 ) over the @xmath13 and @xmath19-variables and search the wave function as expansion @xmath29 where @xmath30 is the inverse matrix to the @xmath31 square matrix @xmath32 defined on the angular grid , we use the completeness relation for the fourier basis @xmath33 , which in our grid representation reads @xmath34 . ] . in the representation ( [ eq7 ] ) the unknown coefficients
@xmath35 are defined by the values of the searched wave function on the angular grid @xmath36 , any local interaction is diagonal @xmath37 and the angular part @xmath38 of the kinetic energy operator has a simple form @xmath39 note that the presence in the interaction potential of the `` nonlocal '' angular part ( i.e. the integration or differentiation over angular variable ) leads to destroying the diagonal structure in ( [ eq8 ] ) .
thus , the 2d schrdinger equation ( [ eq1 ] ) is reduced in the angular - grid representation ( [ eq7 ] ) to the system of coupled ordinary differential equations of the second order : @xmath40 since the wave function @xmath41 must be finite at the origin @xmath42 , the `` left - side '' boundary condition for the functions @xmath35 reads as @xmath43 in the asymptotic region @xmath4 the scattering boundary condition ( [ eq3 ] ) accepts the form @xmath44 by using the fourier expansion for the plane wave @xmath45 and the scattering amplitude @xmath46 are the first kind bessel functions of integer order .
their asymptotic behavior @xcite : @xmath47{}\sqrt { \frac{2}{\pi z } } \cos \left ( { z-\frac{m\pi } { 2}-\frac{\pi } { 4 } } \right)+e^{\left| { imz } \right|}{\rm o}\left ( { \left| z \right|^{-1 } } \right),\\ \left ( { \left| { \arg ( z ) } \right|<\pi } \right)\end{aligned}\ ] ] ] @xmath48 @xmath49 we eliminate the angular dependence from the asymptotic equation ( [ eq12 ] ) and represent the `` right - side '' boundary condition for the functions @xmath50 in the form @xmath51 to solve the boundary - value problem ( [ eq10]),([eq11 ] ) and ( [ eq15 ] ) , we introduce the grid over the @xmath52 and reduce the system of differential equations ( [ eq10 ] ) by using the finite - difference approximation of the sixth order to the system of @xmath53 algebraic equations @xmath54 with the band - structure of the matrix @xmath55 with the width @xmath56 of the band . by using the asymptotic equations ( [ eq15 ] ) in the last two points @xmath57 and @xmath58 one can eliminate the unknown vector @xmath59 from equation ( [ eq15 ] ) and rewrite the `` right - side '' boundary condition in the form @xmath60 analogously
, one can eliminate unknown constant from expression ( [ eq11 ] ) by considering asymptotic equations ( [ eq11 ] ) at the first points @xmath61 and @xmath62 .
the acquired `` left - side '' boundary condition reads @xmath63 thus , the scattering problem is reduced to the boundary value problem ( [ my_eq15]-[my_eq17 ] ) @xmath64 which can be efficiently solved with standard computational techniques such as the sweeping method @xcite or the lu - decomposition @xcite .
the detailed structure of the matrix of the coefficients @xmath65 is discussed in appendix a. after the solving of eq.([my_eq19 ] ) and finding the wave function @xmath66 the scattering amplitude @xmath67 is constructed according to eqs.([eq15 ] ) and ( [ eq14 ] ) .
first , we have analyzed the 2d scattering on the cylindrical potential barrier with the elliptical base @xmath68 the case of the circular base @xmath69 was considered in @xcite , where analytic formula for the scattering amplitude @xmath70+i\frac{\pi } { 2}}\ ] ] was obtained at the zero - energy limit @xmath71 . here and @xmath72 is the euler constant .
we have analyzed the scattering on the potential barrier with circular base @xmath73 for arbitrary momentum @xmath74 .
the results of calculation presented in figs .
[ fig1 ] and [ fig3 ] confirm the convergence of the scattering amplitude @xmath67 to the analytical value ( [ eq17 ] ) at @xmath75 . in this subsection
all the calculations were performed in the units @xmath76 . in the limiting case of the infinitely high potential barrier ( [ eq16 ] ) with the circular base @xmath77
the asymptotic formula ( [ eq17 ] ) becomes exact for arbitrary @xmath74 .
this is confirmed by investigation presented in table [ tab1 ] which illustrates the convergence of the numerical values @xmath78 with increasing ( @xmath79 ) and narrowing ( @xmath80 ) of the potential barrier to the analytic result ( [ eq17 ] ) . in the limit case @xmath79 and @xmath81
we obtain @xmath82 for the scattering length @xmath83 extracted from the calculated amplitude @xmath84 by the formula ( [ eq17 ] ) , what is in agreement to the estimate given in @xcite . the range of applicability of eq .
( [ eq17 ] ) was investigated recently in @xcite . [ cols="^,^,^,^,^ " , ]
the boundary - value problem ( [ eq10]),([eq11 ] ) and ( [ eq15 ] ) , obtained in section ii in the angular - grid representation ( [ eq7 ] ) , reads in a matrix form as : @xmath85\bm { \psi } ( \rho ) \,+\\ \hspace{5 cm } + \,\hat{h}^{(0 ) } \bm { \psi } ( \rho ) = 0 \\ \bm { \psi } \xrightarrow[{\rho \to 0}]{}const\cdot \sqrt \rho \\ \frac{2\pi}{(2 m + 1)\sqrt \rho } \sum\limits_{j=0}^{2 m } { e ^{-im\phi_j } \psi _ j ( \rho ) } = i^mj_m ( q\rho ) \sqrt { 2\pi } + \\
\hspace{5 cm } + \,\frac{f_m(\phi_q)}{\sqrt { -i\rho } } e^{iq\rho } \,,\\
\end{array } } \right.\ ] ] where and @xmath86 . in this representation
the angular dependence is built into the matrix @xmath87 and the interaction is included into the diagonal matrix @xmath88 of values of the potential in the angular grid nodes .
the constant matrix @xmath87 couples all equations in a system and does not depend on the radial variable .
there is no need to compute any matrix elements of the potential , what essentially minimizes the computational costs . for solving boundary value problem ( [ eq23 ] ) the seven - point finite - difference approximation for second derivatives of sixth order @xmath89
is applied in the points @xmath90 of the radial grid , where @xmath91 . as a result
, the system ( [ eq23 ] ) reduces to the system of linear algebraic equations ( [ my_eq16 ] ) with the matrix @xmath55 whose band structure reads @xmath92 where the @xmath93 coefficients are the square @xmath94 matrices , and the elements @xmath95 of the right - side of ( [ eq25 ] ) are the @xmath96 dimensional vectors . after employing the `` right - side '' boundary condition in the form ( [ my_eq16 ] ) in the last three grid points @xmath97 and , analogously , the `` left - side '' boundary condition in the form ( [ my_eq17 ] ) in the first two points @xmath98 , the detailed structure of the matrix @xmath55 is represented as @xmath99 the block structure of the system ( [ eq26 ] ) provides several significant advantages .
the block matrix can be stored in a packaged form , which allows the use of optimal resource .
the system ( [ eq26 ] ) can be efficiently solved by a fast implicit matrix algorithm based on the idea of the block sweep method @xcite .
in the table [ tab4 ] we illustrate the convergence of the calculated scattering amplitude @xmath67 over the number of angular grid points @xmath96 for the scatterers with weak and essential anisotropy at @xmath100 in the potential barrier ( [ eq16 ] ) . for the case @xmath101 we reach the accuracy of four significant digits in the scattering amplitude on the angular grids with @xmath102 . for stronger anisotropy @xmath103 the accuracy of two significant digits
was reached at @xmath104 .
33 m. a. baranov , phys .
rep . * 464 * , 71 ( 2008 ) .
i. bloch , j. dalibard and w. zwerger , rev .
phys . * 80 * , 885 ( 2008 ) . c. ticknor , r. m. wilson and j. l. bohn , phys .
lett . * 106 * , 065301 ( 2011 ) .
g. m. brunn and e. taylor , phys .
* 101 * , 245301 ( 2008 ) .
j. c. cremon , g. m. brunn and s. m. reimann , phys .
lett . * 105 * , 255301 ( 2010 ) .
al . , science * 322 * , 231 ( 2008 ) ; s.ospelkaus et .
al . , science * 327 * , 853 ( 2010 ) .
l. d. carr et .
new j. phys * 11 * , 055049 ( 2009 ) . m. h. g. de miranda et .
al . , nat . phys . * 7 * , 502 ( 2011 ) .
p. minnhagen , rev .
phys . * 59 * , 1001 ( 1987 ) .
a. lee , n. nagaosa and x .- g .
wen , rev .
mod . phys . * 78 * , 17 ( 2006 ) .
k. s. novoselov , rev .
* 83 * , 837 ( 2011 ) . c. nayak , s. h. simon , a. stern , m. freedman and s. d. sarma , rev .
80 * , 1083 ( 2008 ) .
k. martiyanov , v. makhalov and a. turlapov , phys .
* 105 * , 030404 ( 2010 ) ; a. turlapov , jetp letters * 95 * , 96 ( 2012 ) . m. marinescu and l. you , phys .
* 81 * , 4596 ( 1998 ) ; b. deb and l. you .
a * 64 * , 022717 ( 2001 ) v. s. melezhik and chi - yu hu , phys .
lett . * 90 * , 083201 ( 2003 ) . c. ticknor , phys .
a * 80 * , 052702 ( 2009 ) . c. ticknor , phys .
a * 84 * , 032702 ( 2011 ) . c. ticknor , phys .
a * 81 * , 042708 ( 2010 ) .
j. p. dincao and c. h. greene , phys .
a * 83 * , 030702 ( 2011 ) .
z. li , s. v. alyabishev and r. v. krems , phys .
* 100 * , 073202 ( 2008 ) .
v. s. melezhik , j. comput . phys . * 92 * , 67 - 81 ( 1991 ) .
l. d. landau and e. m. lifshitz , in _ quantum mechanics : non - relativistic theory _ , vol .
3 ( pergamon press , 1977 ) 3rd ed . , chap . | [ txt : abstract ] we study the quantum scattering in two spatial dimensions ( 2d ) without the usual partial - wave formalism .
the analysis beyond the partial - wave approximation allows a quantitative treatment of the anisotropic scattering with a strong coupling of different angular momenta nonvanishing even at the zero - energy limit .
high efficiency of our method is demonstrated for the 2d scattering on the cylindrical potential with the elliptical base and dipole - dipole collisions in the plane .
we reproduce the result for the 2d scattering of polarized dipoles in binary collisions obtained recently by ticknor [ phys .
rev .
a * 84 * , 032702 ( 2011 ) ] and explore the 2d collisions of unpolarized dipoles . |
within a decade of being identified as a supernova remnant ( snr ) , [email protected] ( also called g65.3 + 5.7 ) was the subject of several observational papers published in quick succession .
however , once its radio and optical images and its optical and x - ray spectra were recorded , the remnant was ignored .
now , a quarter of a century later , [email protected] has once again become interesting , this time because it helps to explain an unusual class of supernova remnants called thermal composite snrs ( also known as mixed morphology snrs ) .
both of these terms describe the combination of a shell - like radio continuum morphology with a centrally - bright x - ray morphology in which the x - ray emission is due to thermal , rather than synchrotron , processes .
compared with shell - type and plerionic snrs , these remnants have more mysterious origins .
various workers have suggested that thermal composite snrs lack x - ray shells because they have evolved into the radiative phase ( and thus have cool , x - ray dim shells ) , or that their centers are x - ray bright because thermal conduction , other forms of entropy mixing , or cloudlet evaporation has enhanced the central densities .
ejecta enrichment or dust destruction could increase the metallicities in the centers , also enhancing the x - ray emission .
alternatively , a thermal composite morphology may be due to collisions with molecular clouds .
interested readers are directed to @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , and @xcite for additional discussions of these processes
. currently , there are few observations of the `` smoking gun '' for the radiative phase evolution hypothesis : h shells .
the h shell on w44 is the only possible example ( @xcite , with interpretations in @xcite and @xcite ) . here
we demonstrate that another thermal composite supernova remnant , [email protected] , is clearly in the radiative phase .
ironically , this demonstration is not performed by discovering that a previously known thermal composite snr is in the radiative phase , but is done by demonstrating that a previously known radiative phase snr is a thermal composite . in section 2 , we discuss published optical observations , indicating that [email protected] has entered the shell formation phase ; we also discuss published radio and low spatial resolution x - ray observations , showing that [email protected] has a shell - like radio continuum morphology and emits thermal x - rays . with these characteristics ,
[email protected] meets two of the three requirements for the thermal composite classification . the third requirement ,
that the remnant has a centrally filled x - ray morphology , is established in section 3 , through the analysis of archival _ rosat _ observations .
furthermore , model fits to the _ rosat _ data imply that the temperature peaks in the center and decreases mildly with radius , as is found for other thermal composite supernova remnants . in section 4
, we summarize the results .
the first observations of [email protected] , those made with the red prints of the palomar sky survey @xcite , only found two bright filaments , s91 and s94 , but failed to recognize the entire outline of the remnant .
the true extent of the remnant remained unknown until @xcite s optical emission line survey of the galactic plane revealed filaments circumscribing a 4@xmath5@xmath6 3.3@xmath5 ellipse .
@xcite s radio frequency data also suggested a similarly sized shell - type remnant . even before its true extent was revealed
, @xcite used their h@xmath7 , [ n@xmath8 , [ s ] ( 6717 ) , and [ s ] ( 6731 ) measurements to show that the s91 and s94 filaments were dense ( @xmath9 @xmath10 ) shell fragments belonging to a large supernova remnant .
@xcite made additional optical emission line measurements of the s91 filament and of one of the brightest [ o@xmath11 filaments .
both regions have large [ o / h@xmath12 ratios , which , when compared with the shock models of @xcite , indicate that the photons were emitted by thick ( @xmath13 @xmath14 ) postshock cooling regions behind moderate velocity ( @xmath15 to 120 km s@xmath16 ) shocks
. given its [ o / h@xmath12 ratio , s91 s relatively modest [ o@xmath11/h@xmath12 ratio indicates that the cooling and recombination behind the shock is `` complete '' @xcite .
the cooling zone along @xcite s second pointing direction appears to be well developed , but incomplete .
given that the second pointing direction was chosen for its anomalously large [ o@xmath11/h@xmath12 ratio , this portion may be less complete than most of the remnant s shell .
similarly , @xcite found regions of incomplete cooling as well as regions of nearly complete cooling .
the shell temperature found from the older [ o@xmath11 and [ s measurements is @xmath1738,000 k ( fesen , blair & kirshner 1985 ; sabbadin & dodorico 1976 ) .
@xcite calculated shock velocities of 90 to 140 km sec@xmath16 , which are roughly consistent with @xcite s kinematically determined expansion velocity of @xmath18 km s@xmath16 . in summary ,
[email protected] is sufficiently evolved for the shock to be weak and the gas behind it to be relatively cool and recombined .
multiple techniques have been employed to estimate the remnant s size and distance .
@xcite applied the @xmath19-d relation to their 1420 mhz data obtaining a diameter of @xmath20 pc and a distance of @xmath21 pc .
@xcite applied the galactic kinematic relationship to her fabry - perot interferometric h@xmath7 data in order to estimate the distance as 800 pc and the minor and major axes of this elliptically shaped remnant as 56 and 64 pc , respectively . with a distance of @xmath21 pc and galactic latitude of 5.7@xmath5 ,
the remnant lies @xmath22 pc above the galactic disk .
furthermore , [email protected] s age has been estimated as @xmath23 yr ( @xcite , using radiative phase snr relations ) .
+ while others were observing the remnant with optical and radio - frequency detectors , @xcite and @xcite were observing it with x - ray detectors .
@xcite s _ heao 1 _ ( bandpass = 0.2 to 2.5 kev , fwhm = 3@xmath5 ) spectrum exhibits line emission , the hallmark of thermal emission from hot plasma . the model which best fits the _ heao 1 _
spectrum is a raymond and smith thermal model with a temperature between @xmath24 and @xmath25 k @xcite .
although they were not able to resolve the emission spatially with the _ heao 1 _ data , they were able to show that the emission region is not point - like . in the following section ,
we extend upon this work .
we use archival _ rosat _ data to map the remnant , confirming that the emitting gas is extended and finding that it fills the remnant s outline , lacks a bright shell , and exhibits a slowly decreasing temperature gradient .
between august 1995 and september 1997 , the remnant was mapped with a series of overlapping _ rosat _
pspc observations .
we processed the _ rosat _ pspc data ( removed the contamination , scattered solar x - rays , long - term enhancements , after - pulses , particle background , and bright point sources ) and mosaiced the pointings according to the extended object analysis procedures described in @xcite and @xcite .
the resulting images in the _ rosat _ 1l2 band ( sensitive to 0.11 to 0.284 kev photons ) , 45 band ( sensitive to 0.44 to 1.21 kev photons ) , and 67 band ( sensitive to 0.73 to 2.04 kev photons ) are displayed in figure [ rosatimages.fig ] , along with a 1l2 band image overlayed with contours from the 67 band map . with the exception of a ridge of band 1l2 emission along the northeast edge and roughly coincident with bright optical filaments ,
the periphery is x - ray dim .
most of the projected interior shows x - ray emission . because of
its generally `` centrally filled '' x - ray morphology , thermal x - ray emission , and `` shell - type '' radio continuum morphology , the remnant should now be classified as a thermal composite .
comparison of the soft and hard band images shows that the x - ray emission originating in the snr interior is slightly harder than that originating nearer to the edges .
we quantify this trend with a spectral model .
the gas is assumed to be near collisional ionizational equilibrium , as is observed both in similar remnants such as w44 ( harrus et al .
1997 ; shelton , kuntz & petre 2004 ) , 3c391 ( chen & slane 2001 ) , and 0045 - 734 in the smc ( yokogawa et al 2002 ) , and in simulated remnants which have evolved long enough to have centrally - peaked x - ray morphologies @xcite . given
s spectral resolution , the differences between equilibrium and nearly equilibrium spectra are inconsequential , as are differences between newer and older versions of the @xcite spectral code . in general , the model spectra are calculated and handled according to the techniques described in @xcite .
thus , we used the @xcite spectral code to create a suite of collisional ionizational equilibrium spectral models .
the electron temperature ( @xmath26 ) ranged from @xmath27 k to @xmath28 k , with increments in log(@xmath26 ) of 0.05 . in order to account for the absorption by the interstellar medium , we multiplied our suite of model spectra by @xmath29 , where @xmath30 is the product of the absorption column density ( @xmath31 ) and the absorption cross section @xcite .
@xmath31 ranged from @xmath32 @xmath14 to @xmath33 @xmath14 , with increments in log(@xmath31 ) of 0.10 .
the resulting spectra were convolved with the _ rosat _ pspc response matrix , yielding the countrates in each _ rosat _ pspc band .
we prepared the data for comparison with the calculated models by excluding counts from regions of the x - ray image that were contaminated by bright point sources or dominated by the galactic background emission , dividing the the remaining snr area into ten , concentric , nearly equal area annuli .
these annuli were centered on ra = 19@xmath2 33@xmath3 41@xmath4 , dec = 31@xmath5 15 57 in j2000 coordinates and had outer radii of 26 , 39 , 48 , 56 , 64 , 71 , 79 , 86 , 94 , and 122 arcmin , respectively .
regions beyond the snr and regions contaminated by point sources ( including the slightly extended hard source centered on ra = 19@xmath2 36@xmath3 46@xmath4 , dec = 30@xmath5 40 07 ) were excluded from the annuli areas . as a result , in order to create approximately equal area masks , the outermost annuli were extended to relatively large radii .
we then calculated the countrates in each of the _ rosat _ pspc bands for each annulus .
the background countrate for each band was calculated from the cleaned data for the portion of the observed sky that lies outside the snr s footprint and subtracted from the countrate for each annulus , resulting in the `` observed countrates '' .
the standard chi - squared test was used to compare the observed countrates in the _ rosat _ 1l , 4 , 5 , and 6 bands for each annulus with the countrates predicted for the spectral models .
band 2 was omitted , owing to complications of an absorption edge in the detector and band 7 was omitted owing to its low countrate .
the best fitting models were then selected for each annulus .
figure [ spectrum.fig ] displays the observed and model spectra for the seventh annulus ( spanning the region between 79 and 86 arcmin from the remnant s center ) .
the best - fit temperatures range from 2.5 to @xmath34 k and decrease slightly with radius ( see figure [ temp.fig ] ) .
the shallow temperature gradient is a common feature of thermal composite snrs ( i.e. , 3c400.2 @xcite , kes 27 @xcite , w44 @xcite ) . in the best fit models for the various regions , the column densities ranged from @xmath35 to @xmath36 @xmath14 , but did not exhibit a radial trend .
the snr s apparent flux of 0.11 to 2.04 kev photons is @xmath37 ergs @xmath14 s@xmath16 .
if there had been no interstellar scattering , the flux would have been @xmath38 ergs @xmath14 s@xmath16 . assuming that the remnant is 900 pc distant ,
its luminosity is @xmath39 ergs s@xmath16 and the electron density of the x - ray emitting gas is @xmath40 @xmath10 .
based on the theoretical predictions and hydrocode simulations of the archetypical thermal composite supernova remnant , w44 ( @xcite , @xcite ) , we expect the ambient density to be about 10 times the density within the remnant s center .
thus , the ambient density would be roughly 0.1 @xmath10 , a reasonable value at [email protected] s location 90 pc above the galactic plane .
the electron density is lower than that of w44 , causing the luminosity to be lower as well .
in addition to displaying a centrally concentrated snr morphology , the band 45 and band 67 x - ray images also reveal a `` slightly extended point source '' centered on ra = 19@xmath2 36@xmath3 46@xmath4 , dec = 30@xmath5 40 07 and extending approximately 6.5 arcmin in radius in the harder band map and less far in the band 45 map .
the 1l2 band map does not show excess emission at or near this location .
the observed spectrum is too noisy for meaningful fitting to model spectra and there is no known pulsar at this location .
although the princeton - arecibo survey for millisecond pulsar s did not find a pulsar coincident with the observed bright spot , it did find a pulsar , j1931 + 30 , within the remnant s outline @xcite .
we did not find excess x - ray emission at its location , ra = 19@xmath2 31@xmath3 28@xmath4 , dec = 30@xmath5 35.
these _ rosat _ observations , in combination with existing radio observations , establish [email protected] s membership in the thermal composite ( mixed morphology ) class of supernova remnants .
the radio continuum shells and thermal x - ray bright centers of this puzzling class of remnants have inspired several theoretical explanations . in one scenario ,
these remnants have evolved beyond the adiabatic phase into the radiative phase . as a result , their peripheries are now cool and not x - ray emissive .
this scenario can explain why these remnants lack x - ray bright shells , while other explanations are required in order to account for the extreme brightnesses of these remnants centers .
here , we note that crucial support for the evolutionary hypothesis is provided by optical evidence of nearly complete cooling behind [email protected] s relatively slow shockfront .
+ if @xcite s age estimate is correct , then [email protected] is the oldest known thermal composite supernova remnant .
the fact that several other thermal composite supernova remnants neighbor molecular clouds has led to the suggestion that collisions with molecular clouds cause the characteristic thermal composite morphology . in order to test this suggestion on [email protected] , we searched the @xcite co survey data and the @xcite h survey data near @xmath41 at the remnant s systemic velocity ( @xmath42 km s@xmath16 , @xcite ) .
confusion with galactic material made the search inconclusive .
the possibility of a snr - cloud interaction could potentially be addressed with a targeted search for co emission features outlining the remnant .
+ r.l.s . wishes to thank the national research council and nasa s long term space astrophysics grant ( nag5 - 10807 ) for financial support and the anonomous referee for his / her helpful comments .
k.d.k . wishes to thank gsrp for financial support .
this research has made use of data obtained from the high energy astrophysics science archive research center ( heasarc ) , provided by nasa s goddard space flight center . | this paper presents archival _ rosat _ pspc observations of the [email protected] supernova remnant (
also known as g65.3 + 5.7 ) .
little material obscures this remnant and so it was well observed , even at the softest end of _ rosat _ s bandpass ( @xmath1 to 0.28 kev ) .
these soft x - ray images reveal the remnant s centrally - filled morphology which , in combination with existing radio frequency observations , places [email protected] in the thermal composite ( mixed morphology ) class of supernova remnants .
not only might [email protected] be the oldest known thermal composite supernova remnant , but owing to its optically revealed cool , dense shell , this remnant supports the proposal that thermal composite supernova remnants lack x - ray bright shells because they have evolved beyond the adiabatic phase .
these observations also reveal a slightly extended point source centered on ra = 19@xmath2 36@xmath3 46@xmath4 , dec = 30@xmath5 40 07 and extending 6.5 arcmin in radius in the band 67 map .
the source of this emission has yet to be discovered , as there is no known pulsar at this location . |
the mrwd ( magnetic rotator wind - disk ) model for the formation of keplerian disks involves two processes .
the first is a fill - up process in which material from a star flows along magnetic field lines and passes through a shock surface to form a pre - keplerian disk region .
the magnetic field supplies angular momentum to the wind and assists the flow of material towards the disk region .
initially , the density in the pre - keplerian region is small and it increases significantly during this process . the second process occurs after the magnetic force in the disk region becomes small compared to the centrifugal force .
then , super - keplerian material in the pre - keplerian region expands to form a quasi - steady keplerian disk .
if there is no viscosity or inflow into the disk region , angular momentum will be conserved .
the time at which the second process starts depends on the rotational speed , magnetic field strength and rate of fill - up of the disk region . in some stars
, sub - keplerian wind material may continue to flow into the super - keplerian disk region . also , as the angular velocity of the expanding disk approaches a keplerian distribution , magnetorotational instability ( mri ) and viscosity can become important .
when there is inflow or viscosity , the radial extent of the quasi - steady keplerian disk region will be larger . in situations where the magnetic force in the disk is stronger than the centrifugal force at the end of the first process , the disk will continue to maintain its pre - keplerian structure .
figure [ fig1 ] gives a schematic meridional cross - section of a pre - keplerian disk during the fill - up stage .
it shows the magnetic field lines from the stellar surface to the inner and outer end points @xmath2 and @xmath3 , respectively , of the pre - keplerian region . in a meridional plane ,
streamlines of flow from the star to the disk coincide with magnetic field lines .
the _ limiting _ streamline from @xmath4 to @xmath3 separates streamlines that flow into the disk region from the open streamlines that flow away from the star and the disk .
let @xmath5 be a point on the equatorial plane in the pre - keplerian region .
let @xmath6 be the point on the stellar surface that is connected to @xmath5 by a magnetic field line that loops back to a point @xmath7 on the star .
if @xmath8 is the angular velocity at @xmath6 and @xmath7 , the angular velocity at @xmath5 during the pre - keplerian stage will also be equal to @xmath8 @xcite .
the ratio of the angular velocity at @xmath6 to the critical angular velocity at the same point is @xmath9^{1/2}$ ] , where @xmath10 and @xmath11 are the stellar mass and radius , respectively .
@xmath12 is the ratio of the continuum radiation force to gravity . for points on the equatorial plane we put @xmath13 where @xmath14 are spherical polar coordinates .
let @xmath15 and @xmath16 , respectively , be the points in the pre - keplerian region at which the rotational velocities are equal to the keplerian and escape velocities .
then , @xmath17 and @xmath18 .
when the meridional component @xmath19 of the magnetic field at the point @xmath4 on the stellar surface is increased , the value of @xmath20 also increases .
the mrwd model requires that a super - keplerian region be present in the pre - keplerian disk .
hence , @xmath3 must be further away from the star than @xmath15 so that @xmath21 .
thus , @xmath19 must be larger than a minimum value @xmath22 , which is the meridional field strength when @xmath3 coincides with @xmath15 .
figure [ fig2 ] schematically shows the different possible locations of @xmath2 and @xmath3 in relation to @xmath15 and @xmath16 . during the second process
, material between @xmath15 and @xmath16 will move into keplerian orbits .
when there is no viscosity or inflow of wind material , the material beyond @xmath16 will flow away .
if viscosity or inflow becomes important , material in the region between @xmath16 and @xmath3 can become part of a quasi - steady keplerian disk .
a pre - keplerian region having configuration 1 in figure [ fig2 ] will yield a quasi - steady keplerian disk region with maximal radial extent .
the _ optimal model _ corresponds to configuration 5 , where @xmath3 coincides with @xmath16 and @xmath2 coincides with @xmath15 or is closer to the stellar surface than @xmath15 .
this model requires the least magnetic field strength at the stellar surface to form a keplerian disk with the largest radial extent in the absence of inflow or viscosity .
the keplerian region formed when @xmath19 is only slightly larger than @xmath22 will be in the shape of a ring . for larger values of @xmath19
the radial extent of the keplerian region will be larger . in an inviscid model without inflow
, the keplerian region has its largest radial extent when @xmath19 equals @xmath23 , which is the field strength required for the optimal model . for values of @xmath24 larger than @xmath23
there is no increase in the radial extent of the quasi - steady keplerian region as long as the material beyond @xmath16 moves away from the disk region . when the rotation rate @xmath25 of the sector @xmath26 is approximately constant , the inner radius of a quasi - steady keplerian disk is @xmath27 .
let @xmath28 be a point on the outer boundary of the quasi - steady keplerian disk . when there is no viscosity or inflow , we have @xmath29 when @xmath30 and @xmath31 when @xmath32 .
however , if there is flow of wind material into the disk region or if there is an onset of mri during the second process and viscosity becomes significant , the value of @xmath33 will be much larger because of a contribution from material between @xmath16 and @xmath3 .
let @xmath34 be the lower bound derived by @xcite for the meridional field to withstand the effects of meridional circulation near the stellar photosphere and let @xmath35 be the turnover time of this circulation .
let @xmath36 be the minimum surface magnetic field required in the mtd model of @xcite . for be stars with realistic rotation rates we have @xmath37 .
also , @xmath38 , where @xmath39 is the fill - up time of the pre - keplerian disk .
table [ table1 ] gives the critical values of magnetic fields for the formation of disks in different stellar models .
here , we consider models that are appropriate for stars with different values of @xmath0 . although the azimuthal component may also play a role , we do not include it in this discussion .
\(a ) if @xmath40 , quasi - steady keplerian disks satisfying the mrwd model can be formed .
the surface magnetic field will be affected by meridional circulation near the photosphere .
this can lead to changes in the disk over time scales of order @xmath35 .
\(b ) when @xmath41 , the mrwd model is appropriate .
the surface magnetic field will be able to withstand the effects of circulation and persist over time periods that are long compared with @xmath35 .
\(c ) if @xmath42 , the disks will be of mtd type .
if the magnetic field lines from the star thread the disk region and loop back to the star , like field lines of a dipole - type field , the angular velocity distribution in this disk will be the same as that in the region @xmath43 of the stellar surface from which the field lines emerge .
on the other hand , if the field lines from the star pass through the disk region and travel outwards to large distances from the star without looping back , the rotation law for the disk will be similar to the empirical formula specified in the mtd model of @xcite .
the requirement that the disk density should be positive gives the condition @xmath44 for disk formation @xcite .
here , all quantities are evaluated on the appropriate sides of a point @xmath45 on @xmath46 and @xmath47 is the normal component of the wind velocity .
thus , for given values of the equatorial rotation speed and surface magnetic field strength , the wind velocity must be larger than a critical value for a disk to be formed .
since this critical value depends on the stellar rotation speed , a faster rotation rate does not ensure the formation of a shock - compressed disk region unless @xmath47 is correspondingly larger .
the mrwd model does not require that the magnetic field be uniformly strong across the entire stellar surface .
the field must have the required strength on the sector @xmath43 , whose spread in latitude is only a few degrees .
the basic processes in this model do not depend on whether the field is axially symmetric .
the results obtained for symmetrical systems will qualitatively apply to stars possessing magnetic fields with flux loops that thread the disk region .
observational evidence ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) indicates that several stars of the type we consider for disk formation are oblique magnetic rotators with dipole fields .
@xcite find that disk - like structures can form along the magnetic equatorial planes of models with strong magnetic fields in which the alfvn speed is faster than the rotation speed of the magnetosphere .
thus , when the mrwd model is applied to oblique rotators , a pre - keplerian region can be formed along a plane or surface that is determined by gravity , centrifugal force and magnetic force acting on the disk material . during the second process ,
if the magnetic force in the disk becomes small , we expect that super - keplerian material will move into keplerian orbits in the rotational equatorial plane . in the case of oblique rotators with @xmath48 , the results of preuss et al
imply that an mtd disk will be located in the magnetic equatorial plane rather than the rotational equatorial plane . because the magnetic force in the disk region of the mrwd model becomes small at the end of the fill - up stage , the one - armed spiral pattern of the global disk oscillation model ( e.g. , * ? ? ?
* ) can be used to explain the v / r variability in disks of be stars .
cassinelli , j. p. , brown , j. c. , maheswaran , m. , miller , n. a. , & telfer , d. c. 2002 , , 578 , 951 donati , j .- f . ,
wade , g. a. , babel , j. , henrichs , h. f. , de jong , j. a. , harries , t. j. 2001 , , 326 , 1265 donati , j .- f .
, babel , j. , harries , t. j. , howarth , i. d. , petit , p. , semel , m. 2002 , , 333 , 55 maheswaran , m. , & cassinelli , j. p. 1988
, , 335 , 931 maheswaran , m. , & cassinelli , j. p. 1992
, , 386 , 695 maheswaran , m. 2003 , , 592 , 1156 neiner , c. 2002 , ph.d .
thesis , university of amsterdam neiner , c. 2004 , in asp conf .
the nature and evolution of disks around hot stars , ed .
r. ignace & k. gayley ( san francisco : asp ) okazaki , a. t. 1997 , , 318 , 548 preuss , o. , schssler , m. holzwarth , v. , & solanki , s. k. 2004 , , 417 , 987 | we consider a magnetic rotator wind - disk(mrwd ) model for the formation of keplerian disks around be stars .
material from low latitudes of the stellar surface flows along magnetic flux tubes and passes through a shock surface near the equatorial plane to form a _
pre - keplerian _ disk region .
initially , the density in this region is small and the magnetic field helps to maintain super - keplerian rotation speeds . after a fill - up time , the density of the disk is significantly larger and the magnetic force in this region becomes negligible compared with the centrifugal force .
the material then moves outwards to form a quasi - steady keplerian disk . during the fill - up stage ,
the meridional component @xmath0 of the magnetic field at the stellar surface must be larger than a minimum value @xmath1 .
the radial extent of the quasi - steady keplerian region will be larger when @xmath0 is larger or when viscosity plays a role . in b - type stars , the values of @xmath1 are of order 1 g to 10 g. we find that a condition for the formation of shock - compressed disk regions is that in faster rotating stars , the wind speed must be correspondingly larger . |
permutations of the integers @xmath5 = \{1,\ldots , n\}$ ] are a basic building block for the succinct encoding of integer functions @xcite , strings @xcite , and binary relations @xcite , among others .
a permutation @xmath0 is trivially representable in @xmath6 bits , which is within @xmath7 bits of the information theory lower bound of @xmath8 bits . and @xmath9=\{1,\ldots , x\}$ ] . ] in many interesting applications , efficient computation of both the permutation @xmath2 and its inverse @xmath3
is required .
the lower bound of @xmath8 bits yields a lower bound of @xmath10 comparisons to sort such a permutation in the comparison model . yet , a large body of research has been dedicated to finding better sorting algorithms which can take advantage of specificities of each permutation to sort .
trivial examples are permutations sorted such as the identity , or containing sorted blocks @xcite ( e.g. @xmath11 or @xmath12 ) , or containing sorted subsequences @xcite ( e.g. @xmath13 ) : algorithms performing only @xmath7 comparisons on such permutations , yet still @xmath14 comparisons in the worst case , are achievable and obviously preferable .
less trivial examples are classes of permutations whose structure makes them interesting for applications : see mannila s seminal paper @xcite and estivil - castro and wood s review @xcite for more details .
each sorting algorithm in the comparison model yields an encoding scheme for permutations : it suffices to note the result of each comparison performed to uniquely identify the permutation sorted , and hence to encode it . since an adaptive sorting algorithm performs @xmath15 comparisons on many classes of permutations , each adaptive algorithm yields a _ compression scheme _ for permutations , at the cost of losing a constant factor on some other `` bad '' classes of permutations .
we show in section [ sec : applications ] some examples of applications where only `` easy '' permutations arise . yet
such compression schemes do not necessarily support in reasonable time the inverse of the permutation , or even the simple application of the permutation : this is the topic of our study .
we describe several encodings of permutations so that on interesting classes of instances the encoding uses @xmath15 bits while supporting the operations @xmath2 and @xmath3 in time @xmath16 .
later , we apply our compression schemes to various scenarios , such as the encoding of integer functions , text indexes , and others , yielding original compression schemes for these abstract data types .
the _ entropy _ of a sequence of positive integers @xmath17 adding up to @xmath1 is @xmath18 . by convexity of the logarithm , @xmath19 .
[ def : entrop ] [ [ sec : sequences ] ] * succinct encodings of sequences * + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + let @xmath20 $ ] be a sequence over an alphabet @xmath21 $ ] .
this includes bitmaps when @xmath22 ( where , for convenience , the alphabet will be @xmath23 ) .
we will make use of succinct representations of @xmath24 that support operations @xmath25 and @xmath26 : @xmath27 gives the number of occurrences of @xmath28 in @xmath29 $ ] and @xmath30 gives the position in @xmath24 of the @xmath31th occurrence of @xmath28 . for the case
@xmath22 , @xmath24 requires @xmath1 bits of space and @xmath25 and @xmath26 can be supported in constant time using @xmath32 bits on top of @xmath24 @xcite .
the extra space is more precisely @xmath33 for some parameter @xmath34 , which is chosen to be , say , @xmath35 to achieve the given bounds . in this paper
, we will sometimes apply the technique over sequences of length @xmath36 ( @xmath1 will be the length of the permutations ) .
still , we will maintain the value of @xmath34 as a function of @xmath1 , not @xmath37 , which ensures that the extra space will be of the form @xmath38 , i.e. , it will tend to zero when divided by @xmath37 as @xmath1 grows , even if @xmath37 stays constant .
all of our @xmath39 terms involving several variables in this paper can be interpreted in this strong sense : asymptotic in @xmath1 .
thus we will write the above space simply as @xmath40 .
raman _ et al . _
@xcite devised a bitmap representation that takes @xmath41 bits , while maintaining the constant time for the operations . here
@xmath42 , where @xmath43 is the number of occurrences of symbol @xmath28 in @xmath24 , is the so - called _ zero - order entropy _ of @xmath24 . for the binary case this simplifies to @xmath44 , where @xmath45 is the number of bits set in @xmath24 .
grossi _ et al . _
@xcite extended the result to larger alphabets using the so - called _ wavelet tree _ , which decomposes a sequence into several bitmaps . by representing those bitmaps in plain form ,
one can represent @xmath24 using @xmath46 bits of space , and answer @xmath47 $ ] , as well as @xmath25 and @xmath26 queries on @xmath24 , in time @xmath48 .
by , instead , using raman _
s representation for the bitmaps , one achieves @xmath49 bits of space , and the same times .
et al . _
@xcite used multiary wavelet trees to maintain the same compressed space , while improving the times for all the operations to @xmath50 .
[ [ sec : meas - disord - perm ] ] * measures of disorder in permutations * + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + various previous studies on the presortedness in sorting considered in particular the following measures of order on an input array to be sorted . among others ,
mehlhorn @xcite and guibas _ et al . _
@xcite considered the number of pairs in the wrong order , knuth @xcite considered the number of ascending substrings ( runs ) , cook and kim @xcite , and later mannila @xcite considered the number of elements which have to be removed to leave a sorted list , mannila @xcite considered the smallest number of exchanges of arbitrary elements needed to bring the input into ascending order , skiena @xcite considered the number of encroaching sequences , obtained by distributing the input elements into sorted sequences built by additions to both ends , and levcopoulos and petersson @xcite considered shuffled upsequences and shuffled monotone sequences .
estivil - castro and wood @xcite list them all and some others .
we first introduce a compression method that takes advantage of ( ascending ) runs in the permutation . then we consider a stricter variant of the runs , which allows for further compression in applications when those runs arise , and in particular allows the representation size to be sublinear in @xmath1 .
next , we consider a more general type of runs , which need not be contiguous .
one of the best known sorting algorithm is merge sort , based on a simple linear procedure to merge two already sorted arrays , resulting in a worst case complexity of @xmath14 . yet , checking in linear time for _ down - step _ positions in the array , where an element is followed by a smaller one , partitions the original arrays into ascending runs which are already sorted .
this can speed up the algorithm when the array is partially sorted @xcite .
we use this same observation to encode permutations .
a _ down step _ of a permutation @xmath0 over @xmath5 $ ] is a position @xmath51 such that @xmath52 .
a _ run _ in a permutation @xmath0 is a maximal range of consecutive positions @xmath53 which does not contain any down step . let @xmath54 be the list of consecutive down steps in @xmath0 .
then the number of runs of @xmath0 is noted @xmath55 , and the sequence of the lengths of the runs is noted @xmath56 .
for example , permutation @xmath11 contains @xmath57 runs , of lengths @xmath58 . whereas previous analyses @xcite of adaptive sorting algorithms considered only the number @xmath59 of runs , we refine them to consider the distribution @xmath60 of the sizes of the runs .
there is an encoding scheme using at most @xmath61 bits to encode a permutation @xmath0 over @xmath5 $ ] covered by @xmath59 runs of lengths @xmath60 .
it supports @xmath2 and @xmath3 in time @xmath62 for any value of @xmath63 $ ] .
if @xmath51 is chosen uniformly at random in @xmath5 $ ] then the average time is @xmath64 .
[ thm : main ] the hu - tucker algorithm @xcite ( see also knuth @xcite ) produces in @xmath65 time a prefix - free code from a sequence of frequencies @xmath66 adding up to @xmath1 , so that ( 1 ) the @xmath51-th lexicographically smallest code is that for frequency @xmath67 , and ( 2 )
if @xmath68 is the bit length of the code assigned to the @xmath51-th sequence element , then @xmath69 is minimal and moreover @xmath70 ( * ? ? ?
* , eq . ( 27 ) ) .
we first determine @xmath60 in @xmath7 time , and then apply the hu - tucker algorithm to @xmath60 .
we arrange the set of codes produced in a binary trie ( equivalent to a huffman tree @xcite ) , where each leaf corresponds to a run and points to its two endpoints in @xmath0 . because of property ( 1 ) , reading the leaves left - to - right yields the runs also in left - to - right order .
now we convert this trie into a wavelet - tree - like structure @xcite without altering its shape , as follows . starting from the root , first process
recursively each child . for the leaves
do nothing .
once both children of an internal node have been processed , the invariant is that they point to the contiguous area in @xmath0 covering all their leaves , and that this area of @xmath0 has already been sorted .
now we merge the areas of the two children in time proportional to the new area created ( which , again , is contiguous in @xmath0 because of property ( 1 ) ) . as we do the merging , each time we take an element from the left child we append a 0 bit to a bitmap we create for the node , and a 1 bit when we take an element from the right list .
when we finish , we have the following facts : ( 1 ) @xmath0 has been sorted , ( 2 ) the time for sorting has been @xmath71 plus the total number of bits appended to all bitmaps , ( 3 ) each of the @xmath67 elements of leaf @xmath51 ( at depth @xmath68 ) has been merged @xmath68 times , contributing @xmath68 bits to the bitmaps of its ancestors , and thus the total number of bits is @xmath72 .
therefore , the total number of bits in the hu - tucker - shaped wavelet tree is at most @xmath73 . to this
we must add the @xmath74 bits of the tree pointers .
we preprocess all the bitmaps for @xmath25 and @xmath26 queries so as to spend @xmath75 extra bits ( [ sec : sequences ] ) . to compute @xmath3 we start at offset @xmath51 at the root bitmap @xmath76 , with position @xmath77 , and bitmap size @xmath78 .
if @xmath79 = 0 $ ] we go down to the left child with @xmath80 and @xmath81 .
otherwise we go down to the right child with @xmath82 , @xmath83 , and @xmath84 . when we reach a leaf , the answer is @xmath85 . to compute @xmath2
we do the reverse process , but we must first determine the leaf @xmath86 and offset @xmath31 within @xmath86 corresponding to position @xmath51 : we start at the root bitmap @xmath76 , with bitmap size @xmath78 and position @xmath87 . if @xmath88 we go down to the left child with @xmath89 .
otherwise we go down to the right child with @xmath90 and @xmath91 .
we eventually reach leaf @xmath86 , and the offset within @xmath86 is @xmath31 .
we now start an upward traversal using the nodes that are already in the recursion stack ( those will be limited to @xmath92 soon ) .
if @xmath86 is a left child of its parent @xmath93 , then we set @xmath94 , else we set @xmath95 , where @xmath76 is the bitmap of @xmath93
. then we set @xmath96 until reaching the root , where @xmath97 . in both cases
the time is @xmath98 , where @xmath37 is the depth of the leaf arrived at .
if @xmath51 is chosen uniformly at random in @xmath5 $ ] , then the average cost is @xmath99 .
however , the worst case can be @xmath100 in a fully skewed tree .
we can ensure @xmath101 in the worst case while maintaining the average case by slightly rebalancing the hu - tucker tree : if there exist nodes at depth @xmath102 , we rebalance their subtrees , so as to guarantee maximum depth @xmath103 .
this affects only marginally the size of the structure .
a node at depth @xmath37 can not add up to a frequency higher than @xmath104 ( see next paragraph ) .
added over all the possible @xmath59 nodes we have a total frequency of @xmath105 .
therefore , by rebalancing those subtrees we add at most @xmath106 bits .
this is @xmath107 if @xmath108 , and otherwise the cost was @xmath109 anyway . for the same reasons the average time stays @xmath64 as it increases at most by @xmath110 .
the bound on the frequency at depth @xmath37 is proved as follows .
consider the node @xmath86 at depth @xmath37 , and its grandparent @xmath93 .
then the uncle of @xmath86 can not have smaller frequency than @xmath86 .
otherwise we could improve the already optimal hu - tucker tree by executing either a single ( if @xmath86 is left - left or right - right grandchild of @xmath93 ) or double ( if @xmath86 is left - right or right - left grandchild of @xmath93 ) avl - like rotation that decreases the depth of @xmath86 by 1 and increases that of the uncle of @xmath86 by 1 .
thus the overall frequency at least doubles whenever we go up two nodes from @xmath86 , and this holds recursively .
thus the weight of @xmath86 is at most @xmath111 .
the general result of the theorem can be simplified when the distribution @xmath60 is not particularly favorable .
there is an encoding scheme using at most @xmath112 bits to encode a permutation @xmath0 over @xmath5 $ ] with a set of @xmath59 runs .
it supports @xmath2 and @xmath3 in time @xmath62 for any value of @xmath63 $ ] .
[ cor : mainbal ] as a corollary , we obtain a new proof of a well - known result on adaptive algorithms telling that one can sort in time @xmath113
@xcite , now refined to consider the entropy of the partition and not only its size .
we can sort an array of length @xmath1 covered by @xmath59 runs of lengths @xmath60 in time @xmath114 , which is worst - case optimal in the comparison model among all permutations with @xmath59 runs of lengths @xmath60 so that @xmath115 . [ cor : mainsort ]
some classes of permutations can be covered by a small number of runs of a stricter type .
we present an encoding scheme which uses @xmath107 bits for encoding the permutations from those classes , and still @xmath116 bits for all others .
a _ strict run _ in a permutation @xmath0 is a maximal range of positions satisfying @xmath117 .
the _ head _ of such run is its first position .
the number of strict runs of @xmath0 is noted @xmath118 , and the sequence of the lengths of the strict runs is noted @xmath119 .
we will call @xmath120 the sequence of run lengths of the sequence formed by the strict run heads of @xmath0 . for example
, permutation @xmath12 contains @xmath121 strict runs , of lengths @xmath122 .
the run heads are @xmath123 , and contain 2 runs , of lengths @xmath124 .
instead , @xmath11 contains @xmath125 strict runs , all of length 1 .
there is an encoding scheme using at most @xmath126 bits to encode a permutation @xmath0 over @xmath5 $ ] covered by @xmath118 strict runs and by @xmath127 runs , and with @xmath120 being the @xmath59 run lengths in the permutation of strict run heads .
it supports @xmath2 and @xmath3 in time @xmath62 for any value of @xmath63 $ ] .
if @xmath51 is chosen uniformly at random in @xmath5 $ ] then the average time is @xmath128 .
[ thm : strict ] we first set up a bitmap @xmath129 marking with a 1 bit the beginning of the strict runs .
set up a second bitmap @xmath130 such that @xmath131 = r[\pi^{-1}(i)]$ ] .
now we create a new permutation @xmath132 of @xmath133 $ ] which collapses the strict runs of @xmath0 , @xmath134 .
all this takes @xmath7 time and the bitmaps take @xmath135 bits using raman _
s technique , where @xmath25 and @xmath26 are solved in constant time ( [ sec : sequences ] ) .
now build the structure of thm .
[ thm : main ] for @xmath132
. the number of down steps in @xmath0 is the same as for the sequence of strict run heads in @xmath0 , and in turn the same as the down steps in @xmath132 .
so the number of runs in @xmath132 is also @xmath59 and their lengths are @xmath120 .
thus we get at most @xmath136 bits to encode @xmath132 , and can compute @xmath132 and its inverse in @xmath137 worst case and @xmath128 average time . to compute @xmath2 , we find @xmath138 and then compute @xmath139 .
the final answer is @xmath140 . to compute @xmath3
, we find @xmath141 and then compute @xmath142 .
the final answer is @xmath143 .
this adds only constant time on top of that to compute @xmath132 and its inverse .
once again , we might simplify the results when the distribution @xmath120 is not particularly favorable , and we also obtain interesting algorithmic results on sorting .
there is an encoding scheme using at most @xmath144 bits to encode a permutation @xmath0 over @xmath5 $ ] covered by @xmath118 strict runs and by @xmath127 runs .
it supports @xmath2 and @xmath3 in time @xmath62 for any value of @xmath63 $ ] .
[ cor : strictbal ] we can sort a permutation of @xmath5 $ ] , covered by @xmath118 strict runs and by @xmath59 runs , and @xmath120 being the run lengths of the strict run heads , in time @xmath145 , which is worst - case optimal , in the comparison model , among all permutations sharing these @xmath59 , @xmath118 , and @xmath120 values , such that @xmath146 .
[ cor : strictsort ] levcopoulos and petersson @xcite introduced the more sophisticated concept of partitions formed by interleaved runs , such as _ shuffled upsequences _
we discuss here the advantage of considering permutations formed by shuffling a small number of runs .
a decomposition of a permutation @xmath0 over @xmath5 $ ] into _ shuffled upsequences _ is a set of , not necessarily consecutive , subsequences of increasing numbers that have to be removed from @xmath0 in order to reduce it to the empty sequence .
the minimum number of shuffled upsequences in such a decomposition of @xmath0 is noted @xmath147 , and the sequence of the lengths of the involved shuffled upsequences , in arbitrary order , is noted @xmath148 .
for example , permutation @xmath13 contains @xmath149 shuffled upsequences of lengths @xmath150 , but @xmath151 runs , all of length 2 .
whereas the decomposition of a permutation into runs or strict runs can be computed in linear time , the decomposition into shuffled upsequences requires a bit more time .
fredman @xcite gave an algorithm to compute the size of an optimal partition , claiming a worst case complexity of @xmath14 .
in fact his algorithm is adaptive and takes @xmath152 time .
we give here a variant of his algorithm which computes the partition itself within the same complexity , and we achieve even better time on favorable sequences @xmath148 .
[ lem : partitioninsus ] given a permutation @xmath0 over @xmath5 $ ] covered by @xmath147 shuffled upsequences of lengths @xmath148 , there is an algorithm finding such a partition in time @xmath153 .
initialize a sequence @xmath154 , and a splay tree @xmath155 @xcite with the node @xmath156 , ordered by the rightmost value of the sequence contained by each node . for each further element @xmath2 ,
search for the sequence with the maximum ending point smaller than @xmath2 . if any , add @xmath2 to this sequence , otherwise create a new sequence and add it to @xmath155 .
fredman @xcite already proved that this algorithm computes an optimal partition .
the adaptive complexity results from the mere observation that the splay tree ( a simple sorted array in fredman s proof ) contains at most @xmath147 elements , and that the node corresponding to a subsequence is accessed once per element in it .
hence the total access time is @xmath153 ( * ? ? ?
* thm . 2 ) .
the complete description of the permutation requires to encode the computation of both the partitioning algorithm and the sorting one , and this time the encoding cost of partitioning is as important as that of merging . [
thm : sus ] there is an encoding scheme using at most @xmath157 bits to encode a permutation @xmath0 over @xmath5 $ ] covered by @xmath147 shuffled upsequences of lengths @xmath148 .
it supports the operations @xmath2 and @xmath3 in time @xmath158 for any value of @xmath63 $ ] .
if @xmath51 is chosen uniformly at random in @xmath5 $ ] the average time is @xmath159 .
partition the permutation @xmath0 into @xmath147 shuffled upsequences using lemma [ lem : partitioninsus ] , resulting in a string @xmath24 of length @xmath1 over alphabet @xmath160 $ ] which indicates for each element of the permutation @xmath0 the label of the upsequence it belongs to . encode @xmath24 with a wavelet tree using raman _ et al .
_ s compression for the bitmaps , so as to achieve @xmath161 bits of space and support retrieval of any @xmath47 $ ] , as well as symbol @xmath25 and @xmath26 on @xmath24 , in time @xmath158 ( [ sec : sequences ] ) . store also an array @xmath162 $ ] so that @xmath163 $ ] is the accumulated length of all the upsequences with label less than @xmath37 .
array @xmath164 requires @xmath165 bits .
finally , consider the permutation @xmath132 formed by the upsequences taken in label order : @xmath132 has at most @xmath147 runs and hence can be encoded using @xmath166 bits using thm .
[ thm : main ] , as @xmath148 in @xmath0 corresponds to @xmath60 in @xmath132 .
this supports @xmath167 and @xmath168 in time @xmath158 .
now @xmath169+rank_{s[i]}(s , i))$ ] can be computed in time @xmath158 .
similarly , @xmath170)$ ] , where @xmath37 is such that @xmath163 < ( \pi')^{-1}(i ) \le a[\ell+1]$ ] , can also be computed in @xmath158 time .
thus the whole structure uses @xmath171 bits and supports @xmath2 and @xmath3 in time @xmath158 .
the obstacles to achieve the claimed average time are the operations on the wavelet tree of @xmath24 , and the binary search in @xmath164 .
the former can be reduced to @xmath172 by using the improved wavelet tree representation by ferragina _
et al . _
( [ sec : sequences ] ) .
the latter is reduced to constant time by representing @xmath164 with a bitmap @xmath173 $ ] with the bits set at the values @xmath163 + 1 $ ] , so that @xmath163 = select_1(a',\ell)-1 $ ] , and the binary search is replaced by @xmath174 . with raman _
_ s structure ( [ sec : sequences ] ) , @xmath175 needs @xmath176 bits and operates in constant time .
again , we might prefer a simplified result when @xmath148 has no interesting distribution , and we also achieve an improved result on sorting , better than the known @xmath152 . [ cor : susbal ]
there is an encoding scheme using at most @xmath177 bits to encode a permutation @xmath0 over @xmath5 $ ] covered by @xmath147 shuffled upsequences .
it supports the operations @xmath2 and @xmath3 in time @xmath158 for any value of @xmath63 $ ] .
[ cor : sortsus ] we can sort an array of length @xmath1 , covered by @xmath147 shuffled upsequences of lenghts @xmath148 , in time @xmath153 , which is worst - case optimal , in the comparison model , among all permutations decomposable into @xmath147 shuffled upsequences of lenghts @xmath148 such that @xmath178 .
consider a full - text inverted index which gives the word positions of any word in a text .
this is a popular data structure for natural language text retrieval @xcite , as it permits for example solving phrase queries without accessing the text . for each different text word ,
an increasing list of its text positions is stored .
let @xmath1 be the total number of words in a text collection @xmath179 $ ] and @xmath59 the vocabulary size ( i.e. , number of different words ) .
an uncompressed inverted index requires @xmath180 bits .
it has been shown @xcite that , by @xmath181-encoding the differences between consecutive entries in the inverted lists , the total space reduces to @xmath182 , where @xmath183 is the zero - order entropy of the text if seen as a sequence of words ( [ sec : sequences ] ) .
we note that the empirical law by heaps @xcite , well accepted in information retrieval , establishes that @xmath59 is small : @xmath184 for some constant @xmath185 depending on the text type . several successful methods to compress natural language text
take words as symbols and use zero - order encoding , and thus the size they can achieve is lower bounded by @xmath186 @xcite .
if we add the differentially encoded inverted index in order to be able of searching the compressed text , the total space is at least @xmath187 .
now , the concatenation of the @xmath59 inverted lists can be seen as a permutation of @xmath5 $ ] with @xmath59 runs , and therefore thm .
[ thm : main ] lets us encode it in @xmath188 bits . within the same space we can add @xmath59 numbers telling where the runs begin , in an array
@xmath189 $ ] .
now , in order to retrieve the list of the @xmath51-th word , we simply obtain @xmath190 ) , \pi(v[i]+1 ) , \ldots , \pi(v[i+1]-1)$ ] , each in @xmath62 time .
moreover we can extract any random position from a list , which enables binary - search - based strategies for list intersection @xcite .
in addition , we can also obtain a text passage from the ( inverse ) permutation : to find out @xmath191 $ ] , @xmath192 gives its position in the inverted lists , and a binary search on @xmath193 finds the interval @xmath194
\le \pi^{-1}(j ) < v[i+1]$ ] , to output that @xmath191 = i$]th word , in @xmath137 time .
this result is very interesting , as it constitutes a true word - based _ self - index _
@xcite ( i.e. , a compressed text index that contains the text ) .
similar results have been recently obtained with rather different methods @xcite .
the cleanest one is to build a wavelet tree over @xmath155 with compression @xcite , which achieves @xmath195 bits of space , and permits obtaining @xmath196 $ ] , as well as extracting the @xmath31th element of the inverted list of the @xmath51th word with @xmath197 , all in time @xmath198 . yet
, one advantage of our approach is that the extraction of @xmath37 consecutive entries @xmath199)$ ] takes @xmath200 time if we do the process for all the entries as a block : start at range @xmath201 $ ] at the root bitmap @xmath76 , with position @xmath77 , and bitmap size @xmath78 .
go down to both left and right children : to the left with @xmath201 \leftarrow [ rank_0(b , i),rank_0(b , i')]$ ] , same @xmath202 , and @xmath89 ; to the right with @xmath201 \leftarrow [ rank_1(b , i),rank_1(b , i')]$ ] , @xmath83 , and @xmath91 .
stop when the range @xmath201 $ ] becomes empty or when we reach a leaf , in which case report all answers @xmath203 , @xmath204 . by representing the inverted list as @xmath205
, we can extract long inverted lists faster than the existing methods .
[ cor : invfile ] there exists a representation for a text @xmath179 $ ] of integers in @xmath206 $ ] ( regarded as word identifiers ) , with zero - order entropy @xmath207 , that takes @xmath208 bits of space , and can retrieve the text position of the @xmath31th occurrence of the @xmath51th text word , as well as the value @xmath191 $ ] , in @xmath137 time .
it can also retrieve any range of @xmath37 successive occurrences of the @xmath51th text word in time @xmath209 .
we could , instead , represent the inverted list as @xmath0 , so as to extract long text passages efficiently , but the wavelet tree representation can achieve the same result . another interesting functionality that both representations share , and which is useful for other list intersection algorithms @xcite ,
is that to obtain the first entry of a list which is larger than @xmath210 .
this is done with @xmath25 and @xmath26 on the wavelet tree representation . in our permutation representation , we can also achieve it in @xmath62 time by finding out the position of a number @xmath210 within a given run .
the algorithm is similar to those in thm .
[ thm : main ] that descend to a leaf while maintaining the offset within the node , except that the decision on whether to descend left or right depends on the leaf we want to arrive at and not on the bitmap content ( this is actually the algorithm to compute @xmath25 on binary wavelet trees @xcite ) . finally , we note that our inverted index data structure supports in small time all the operations required to solve conjunctive queries on binary relations .
suffix arrays are used to index texts that can not be handled with inverted lists . given a text @xmath179 $ ] of @xmath1 symbols over an alphabet of size @xmath59 , the _ suffix _ array @xmath211
$ ] is a permutation of @xmath5 $ ] so that @xmath212,n]$ ] is lexicographically smaller than @xmath213,n]$ ] . as suffix arrays
take much space , several compressed data structures have been developed for them @xcite .
one of interest for us is the _ compressed suffix array ( csa ) _ of sadakane @xcite .
it builds over a permutation @xmath214 of @xmath5 $ ] , which satisfies @xmath215 = ( a[i]~\textrm{mod}~n ) + 1 $ ] ( and thus lets us move virtually one position forward in the text ) @xcite .
it turns out that , using just @xmath214 and @xmath74 extra bits , one can @xmath216 _ count _ the number of times a pattern @xmath217 $ ] occurs in @xmath155 using @xmath218 applications of @xmath214 ; @xmath219 _ locate _ any such occurrence using @xmath220 applications of @xmath214 , by spending @xmath221 extra bits of space ; and @xmath222 _ extract _ a text substring @xmath223 $ ] using at most @xmath224 applications of @xmath214 . hence this is another self - index , and its main burden of space is that to represent permutation @xmath214 .
sadakane shows that @xmath214 has at most @xmath59 runs , and gives a representation that accesses @xmath225 $ ] in constant time by using @xmath226 bits of space .
it was shown later @xcite that the space is actually @xmath227 bits , for any @xmath228 and constant @xmath229 . here
@xmath230 is the @xmath231th order empirical entropy of @xmath155 @xcite . with thm .
[ thm : main ] we can encode @xmath214 using @xmath232 bits of space , whose extra terms aside from entropy are better than sadakane s
. those extra terms can be very significant in practice .
the price is that the time to access @xmath214 is @xmath137 instead of constant . on the other hand ,
an interesting extra functionality is that to compute @xmath233 , which lets us move ( virtually ) one position backward in @xmath155 .
this allows , for example , displaying the text context around an occurrence without having to spend any extra space .
still , although interesting , the result is not competitive with recent developments @xcite .
an interesting point is that @xmath214 contains @xmath234 strict runs , for any @xmath231 @xcite .
therefore , cor .
[ cor : strictbal ] lets us represent it using @xmath235 bits of space . for @xmath231 limited as above ,
this is at most @xmath236 bits , which is similar to the space achieved by another self - index @xcite , yet again it is slightly superseded by its time performance .
munro _ et al . _
@xcite described how to represent a permutation @xmath0 as the concatenation of its cycles , completed by a bitvector of @xmath1 bits coding the lengths of the cycles .
as the cycle representation is itself a permutation of @xmath5 $ ] , we can use any of the permutation encodings described in [ sec : compr - techn ] to encode it , adding the binary vector encoding the lengths of the cycles .
it is important to note that , for a specific permutation @xmath0 , the difficulty to compress its cycle encoding @xmath132 is not the same as the difficulty to encode the original permutation @xmath0 . given a permutation @xmath0 with @xmath28 cycles of lengths @xmath237 , there are several ways to encode it as a permutation @xmath132 , depending on the starting point of each cycle ( @xmath238 } n_i$ ] choices ) and the order of the cycles in the encoding ( @xmath239 choices ) . as a consequence , each permutation @xmath0 with @xmath28 cycles of lengths @xmath237
can be encoded by any of the @xmath238 } i\times n_i$ ] corresponding permutations .
any of the encodings from theorems [ thm : main ] , [ thm : strict ] and [ thm : sus ] can be combined with an additional cost of at most @xmath240 bits to encode a permutation @xmath0 over @xmath5 $ ] composed of @xmath28 cycles of lengths @xmath237 to support the operation @xmath4 for any value of @xmath241 , in time and space function of the order in the permutation encoding of the cycles of @xmath0 .
the space `` wasted '' by such a permutation representation of the cycles of @xmath0 is @xmath242 bits . to recover some of this space ,
one can define a canonical cycle encoding by starting the encoding of each cycle with its smallest value , and by ordering the cycles in order of their starting point .
this canonical encoding always starts with a @xmath243 and creates at least one shuffled upsequence of length @xmath28 : it can be compressed as a permutation over @xmath244 $ ] with at least one shuffled upsequence of length @xmath245 through thm [ thm : sus ] .
munro and rao @xcite extended the results on permutations to arbitrary functions from @xmath5 $ ] to @xmath5 $ ] , and to their iterated application @xmath246 , the function iterated @xmath231 times starting at @xmath51 .
their encoding is based on the decomposition of the function into a bijective part , represented as a permutation , and an injective part , represented as a forest of trees whose roots are elements of the permutation : the summary of the concept is that an integer function is just a `` hairy permutation '' . combining the representation of permutations from @xcite with any representation of trees supporting the level - ancestor operator and an iterator of the descendants at a given level yields a representation of an integer function @xmath247 using @xmath248 bits to support @xmath246 in @xmath249 time , for any fixed @xmath250 , integer @xmath241 and @xmath63 $ ] .
janssen _ et al . _
@xcite defined the _ degree entropy _ of an ordered tree @xmath155 with @xmath1 nodes , having @xmath67 nodes with @xmath51 children , as @xmath251 , and proposed a succinct data structure for @xmath155 using @xmath252 bits to encode the tree and support , among others , the level - ancestor operator .
obviously , the definition and encoding can be generalized to a forest of @xmath231 trees by simply adding one node whose @xmath231 children are the roots of the @xmath231 trees . encoding the injective parts of the function using janssen _ et al . _
@xcite succinct encoding , and the bijective parts of the function using one of our permutation encodings , yields a compressed representation of any integer function which supports its application and the application of its iterated variants in small time .
[ cor : fun ] there is a representation of a function @xmath253\rightarrow[n]$ ] that uses @xmath254 bits to support @xmath246 in @xmath255 time , for any integer @xmath231 and for any @xmath63 $ ] , where @xmath155 is the forest representing the injective part of the function , and @xmath59 is the number of runs in the bijective part of the function .
bentley and yao @xcite , when introducing a family of search algorithms adaptive to the position of the element searched ( aka the `` unbounded search '' problem ) , did so through the definition of a family of adaptive codes for unbounded integers , hence proving that the link between algorithms and encodings was not limited to the complexity lower bounds suggested by information theory . in this paper
, we have considered the relation between the difficulty measures of adaptive sorting algorithms and some measures of `` entropy '' for compression techniques on permutations . in particular , we have shown that some concepts originally defined for adaptive sorting algorithms , such as runs and shuffled upsequences , are useful in terms of the compression of permutations ; and conversely , that concepts originally defined for data compression , such as the entropy of the sets of sizes of runs , are a useful addition to the set of difficulty measures that one can consider in the study of adaptive algorithms .
it is easy to generalize our results on runs and strict runs to take advantage of permutations which are a mix of up and down runs or strict runs ( e.g. @xmath256 , with only a linear extra computational and/or space cost .
the generalization of our results on shuffled upsequences to sms @xcite , permutations containing mixes of subsequences sorted in increasing and decreasing orders ( e.g. @xmath257 ) is sligthly more problematic , because it is np hard to optimally decompose a permutation into such subsequences @xcite , but any approximation scheme @xcite would yield a good encoding .
j. barbay and g. navarro .
compressed representations of permutations , and applications .
technical report tr / dcc-2008 - 18 , department of computer science ( dcc ) , university of chile , december 2008 . | we explore various techniques to compress a permutation @xmath0 over @xmath1 integers , taking advantage of ordered subsequences in @xmath0 , while supporting its application @xmath2 and the application of its inverse @xmath3 in small time .
our compression schemes yield several interesting byproducts , in many cases matching , improving or extending the best existing results on applications such as the encoding of a permutation in order to support iterated applications @xmath4 of it , of integer functions , and of inverted lists and suffix arrays .
jrmy barbay gonzalo navarro |
the interplay between dixmier traces , @xmath0functions and heat kernel formulae is a cornerstone of noncommutative geometry @xcite .
these formulae are widely used in physical applications . to define these objects ,
let us fix a hilbert space @xmath1 and let @xmath2 be the algebra of all bounded operators on @xmath1 with its standard trace @xmath3 let @xmath4 and @xmath5 be positive operators from @xmath6 consider the following @xmath7$]-valued functions @xmath8 and , for fixed @xmath9 @xmath10 when these functions are finitely valued , they are frequently referred to as @xmath0functions and heat kernel functions associated with the operators @xmath4 and @xmath11 when these functions are bounded , a particular interest is attached to their asymptotic behavior when @xmath12 which is usually measured with the help of some generalised limit @xmath13 yielding the following functionals @xmath14 and , @xmath15 a natural class of operators for which the formulae and are well defined ( respectively , and ) is given by the set @xmath16 ( respectively , @xmath17 ) of compact operators from @xmath6 more precisely , denote by @xmath18 @xmath19 the singular values of a compact operator @xmath20 ( the singular values are the eigenvalues of the operator @xmath21 arranged with multiplicity in decreasing order , ( @xcite )
. then @xmath22 defines a banach ideal of compact operators .
we set @xmath23 it is important to observe that the subset @xmath17 is not dense in @xmath16 ( see e.g. @xcite ) .
it should also be pointed out that our notation here differs from that used in @xcite .
it follows from ( * ? ? ?
* theorem 4.5 ) that the functions defined in are bounded if and only if @xmath24 it also follows from @xcite and @xcite that the functions defined in are bounded if and only if @xmath25 in fact the last result is a strong motivation to consider the following modification of formulae .
let us consider a cesaro operator on @xmath26 given by @xmath27 it follows from @xcite and @xcite that the functions @xmath28 are bounded if and only if @xmath24 therefore , for a given generalised limit @xmath29 let us set @xmath30 and instead of the functions given in consider the functions @xmath31 the class of dilation invariant states @xmath32 as above was introduced by a. connes ( see @xcite ) and it is natural to refer to this class as connes states .
we prove in section [ heatk lin ] that if @xmath33 in is dilation invariant , then @xmath34 is a linear functional on @xmath35 in fact , we also show in proposition [ coincidence ] that if @xmath33 in is such that @xmath34 is linear on @xmath36 then necessarily there exists a dilation invariant generalised limit @xmath37 such that @xmath38 there is a deep reason to require that the functionals @xmath34 and @xmath39 be defined on @xmath16 and be linear ( and thus , by implication , to consider connes states ) .
important formulae in noncommutative geometry @xcite and its semifinite counterpart @xcite then connect these functionals with dixmier traces on @xmath35 recall that in @xcite , j. dixmier constructed a non - normal semifinite trace ( a dixmier trace ) on @xmath2 using the weight @xmath40 where @xmath33 is a dilation invariant state on @xmath41 the interplay between positive functionals @xmath42 @xmath39 and @xmath34 on @xmath16 makes an important chapter in noncommutative geometry and has been treated ( among many other papers ) in @xcite .
we now list a few most important known results concerning this interplay and explain our contribution to this topic . in @xcite ,
the equality @xmath43 was established for every @xmath44 under very restrictive conditions on @xmath45 these conditions are dilation invariance for both @xmath33 and @xmath46 and @xmath47invariance of @xmath45 in @xcite , for the special case @xmath48 the assumption that @xmath33 is @xmath47invariant has been removed .
however , the case of an arbitrary @xmath5 appears to be inaccessible by the methods in that article . in section [ zeta ]
, we prove the general result which implies , in particular , that the equality holds without requiring @xmath47invariance of @xmath45 in @xcite , the equality @xmath49 was established under the same conditions on @xmath33 and @xmath46 as above . in @xcite , in the special case @xmath50 the equality was established under the assumption that @xmath33 is @xmath47invariant .
however , again the case of an arbitrary @xmath5 appears to be inaccessible by the methods in that article . here , we are able to treat the case of a general operator @xmath11 in @xcite a more general approach to the heat kernel formulae is suggested .
it consists of replacing the function @xmath51 with an arbitrary function @xmath52 from the schwartz class .
the following equality was proved in @xcite @xmath53 for @xmath54 and @xmath47invariant @xmath45 in @xcite , m. benameur and t. fack have asked whether the result above continues to stand without the @xmath47invariance assumption on @xmath45 in theorem [ xihk red ] below , we answer this question affirmatively for a much larger class of functions than the schwartz class and for any @xmath55 finally , it is important to emphasize the connection between our results with the theory of fully symmetric functionals . recall that a linear positive functional @xmath56 is called _ fully symmetric _ if @xmath57 for every positive @xmath58 such that @xmath59 the latter symbol means that @xmath60 it is obvious that every dixmier trace @xmath61 is a fully symmetric functional .
however , the fact that every fully symmetric functional coincides with a dixmier trace is far from being trivial ( see @xcite and theorem [ fs is dixmier ] below ) .
it is therefore quite natural to ask whether a similar result holds for the sets of all linear positive functionals on @xmath16 formed by the @xmath34 and @xmath39 respectively . to this end
, we establish results somewhat similar to those of @xcite .
firstly , in theorem [ xiomega linearity ] we prove that if @xmath33 in is dilation invariant , then the functional @xmath34 extends to a fully symmetric functional on @xmath35 secondly , in theorem [ also main ] we show that in fact every normalized fully symmetric functional on @xmath16 coincides with some @xmath62 where @xmath33 is dilation invariant .
thus , in view of @xcite , we can conclude that the set @xmath63 coincides with the set @xmath64 ( up to a norming constant ) . at the same time , a natural question , namely , whether the equality @xmath65 holds for every dilation invariant generalised limit @xmath33 is answered in the negative in theorem [ counterexample ] .
finally , we note that the question on the relationship between the sets @xmath66 @xmath67 and @xmath68 remains open .
the theory of singular traces on operator ideals rests on some classical analysis which we now review for completeness . as usual
, @xmath26 is the set of all bounded lebesgue measurable functions on the semi - axis equipped with the uniform norm @xmath69 given a function @xmath70 one defines its decreasing rearrangement @xmath71 by the formula ( see e.g. @xcite ) @xmath72 let @xmath1 be a hilbert space and let @xmath2 be the algebra of all bounded operators on @xmath1 equipped with the uniform norm @xmath69 let @xmath73 be a semi - finite von neumann algebra with a fixed faithful and normal semi - finite trace @xmath74 for every @xmath75 the generalised singular value function @xmath76 is defined by the formula ( see e.g. @xcite ) @xmath77 if , in particular , @xmath78 then @xmath79 is a step function and , therefore , can be identified with the sequence @xmath80 of singular numbers of the operators @xmath4 ( the singular values are the eigenvalues of the operator @xmath81 arranged with multiplicity in decreasing order ) .
equivalently , @xmath79 can be defined in terms of the distribution function @xmath82 of @xmath83 that is , setting @xmath84 we obtain @xmath85 here , @xmath86 denotes the spectral measure of the operator @xmath87 the following formula follows directly from the von neumann definition of trace ( see the definition at ( * ? ? ?
* definition 15.1.1 ) ) @xmath88 using the jordan decomposition , every operator @xmath89 can be uniquely written as @xmath90 here , @xmath91 ( respectively , @xmath92 ) for any operator @xmath89 and @xmath93 ( respectively , @xmath94 ) for any self - adjoint operator @xmath95 recall that @xmath96 for every @xmath97 and @xmath98 for every self - adjoint @xmath99 let @xmath100 be an increasing concave function such that @xmath101 as @xmath102 the marcinkiewicz function space @xmath103 ( see e.g. @xcite ) consists of all @xmath104 satisfying @xmath105 the marcinkiewicz operator space @xmath106 ( see e.g. @xcite ) consists of all @xmath97 satisfying @xmath107 we are especially interested in marcinkiewicz spaces @xmath16 and @xmath108 that arise when @xmath109 @xmath110 in the literature , the ideal @xmath16 is sometimes referred to as the dixmier ideal .
we recommend the recent paper of a. pietsch , @xcite , discussing the origin of @xmath16 in mathematics . for @xmath111 dilation operators @xmath112
are defined by the formula @xmath113 clearly , @xmath114 ( see also ( * ? ? ?
* theorem ii.4.4 ) ) .
further , we need to recall the important notion of hardy - littlewood majorization .
let @xmath115 @xmath5 is said to be majorized by @xmath4 and written @xmath116 if and only if @xmath117 we have ( see @xcite ) @xmath118 one of the most widely used ideals in von neumann algebras is @xmath119 usually called the schatten - von neumann ideal of @xmath120-summable operators .
using hardy - littlewood majorization , it is very easy to see ( e.g. ( * ? ? ?
* lemma 2.1 ) ) that @xmath121 for all @xmath122 a linear functional @xmath56 is said to be _ symmetric _ if @xmath123 for every positive @xmath58 such that @xmath124 a linear functional @xmath56 is said to be _ fully symmetric _ if @xmath57 for all @xmath125 such that @xmath116 @xcite .
every fully symmetric functional is symmetric and bounded .
the converse fails @xcite .
a positive normalised linear functional @xmath13 is called a _ generalised limit _ if @xmath126 for every @xmath127 such that @xmath128 a linear functional @xmath13 is called _ dilation invariant _
if @xmath129 for every @xmath127 and every @xmath130 let @xmath131 we denote by @xmath132 the set of all positive operators from @xmath133 let @xmath134 be a dilation invariant generalised limit . define a functional @xmath135 on @xmath136 by the formula
@xmath137 the functional @xmath135 is additive and unitarily invariant on @xmath138 thus , @xmath135 extends to a fully symmetric functional on @xmath35 one usually refers to it as to a _ dixmier trace_. we refer the reader to @xcite for details .
further , we use the following properties of dixmier traces .
let @xmath139 and let @xmath99 we have ( see @xcite ) @xmath140 suppose that @xmath141 it follows from that @xmath142 suppose that the trace @xmath143 on the von neumann algebra @xmath144 is infinite and the algebra @xmath144 is either diffuse ( that is with no minimal projections ) or else is @xmath6 given any finite sequence @xmath145 of operators , we can construct a sequence of operators @xmath146 such that @xmath147 for all @xmath148 s and @xmath149 for all @xmath150 further , we refer to any such sequence @xmath146 as a sequence of disjoint copies of @xmath145 .
cesaro operator @xmath151 is defined on @xmath26 by the formula @xmath27
in this section , for the reader s convenience , we collect a number of key known results , which will be used throughout this paper .
the following important theorem is proved in ( * ? ? ?
* theorem 11 ) for general marcinkiewicz spaces .
[ fs is dixmier ] every fully symmetric functional on @xmath16 is a dixmier trace . the following theorem is an analog of lidskii formula ( see @xcite ) for dixmier traces .
it is proved in ( * ? ? ?
* theorem 33 ) for a large subclass of marcinkiewicz spaces which contains @xmath35 [ lidskii formula ] let @xmath139 and let @xmath135 be an arbitrary dixmier trace on @xmath35 we have @xmath152 the following @xmath33-variant of the classical karamata theorem is established in @xcite .
[ karamata ] let @xmath153 be a continuous increasing function . set @xmath154 we have @xmath155 for any dilation invariant generalised limit @xmath45 consider the ideal @xmath156 of @xmath143-compact operators in @xmath157 ( that is the norm closed ideal generated by the projections @xmath158 with @xmath159 ) .
the following result is not new ( see ( * ? ? ?
* chapter ii , lemma 3.4 ) ) .
we present a short proof for convenience of the reader .
[ t equals majorization ] let @xmath160 be positive @xmath161compact operators .
we have @xmath116 if and only if @xmath162 fix @xmath163 it follows from the definition of generalised singular value function that @xmath164}.$ ] applying ( * ? ? ?
* proposition 2.7 ) to the operator @xmath165 we have @xmath166 and hence @xmath167 the function @xmath168 attains its maximum at @xmath169 if @xmath170 then @xmath171 inequality follows now from .
suppose now that holds .
fix @xmath172 and set @xmath173 it follows that @xmath174 @xmath175 hence , @xmath176 since @xmath177 is arbitrary , we have @xmath59
we begin by showing that the functionals given in are well defined on @xmath138 [ zeta correctness ] if @xmath13 is a generalised limit , then @xmath178 and @xmath179 for any @xmath180 it is clear that @xmath181 therefore , @xmath182 hence , @xmath183 it follows from @xmath184 that @xmath185 [ chei remark ] let @xmath186 for any generalised limit @xmath187 such that @xmath188 we have @xmath189 indeed , @xmath190 [ elementary inequality ] for any @xmath191 we have @xmath192 in the special case when @xmath78 the first inequality can be found in ( * ? ? ? * ( 2.9 ) ) . in the general case
, it follows directly from proposition 4.6(ii ) of @xcite when @xmath193 @xmath194 the second inequality follows from the same proposition by setting there @xmath195 let @xmath55 for a functional @xmath196 defined on @xmath136 by ( see lemma [ zeta correctness ] ) , we set @xmath197 the following theorem shows that functionals @xmath196 defined by are fully symmetric on @xmath35 [ zeta linearity ] if @xmath13 is a generalised limit , then @xmath196 is a fully symmetric linear functional on @xmath35 to verify that @xmath196 is linear , it is sufficient to check that @xmath198 for any @xmath199 it follows from the left hand side inequality of lemma [ elementary inequality ] that @xmath200 noting that @xmath201 it follows from the right hand side inequality of lemma [ elementary inequality ] and remark [ chei remark ] that @xmath202 therefore , we have @xmath203 the homogeneity of @xmath196 follows from remark [ chei remark ] . finally , if @xmath204 then @xmath205 and @xmath206 hence , @xmath207 and so @xmath208 let @xmath209 we extend the functional @xmath210 on @xmath36 similarly to . observe that @xmath211 [ continuity lemma ] if @xmath139 and @xmath212 in @xmath213 then @xmath214 it is sufficient to prove the assertion for @xmath180 since @xmath215 we obtain @xmath216 the following lemma follows immediately from ( * ? ? ?
* lemma 3.3 ) .
[ cps lemma ] let @xmath217 and let @xmath130 we have a. @xmath218 if @xmath219 b. @xmath220 if @xmath221 the result below significantly strengthens ( * ? ? ?
* proposition 3.6 ) by removing all extra assumptions on the generalised limit @xmath222 [ main reduction lemma ] if @xmath13 is a generalised limit , then @xmath223 it is sufficient to prove the assertion for @xmath180 suppose first that there are constants @xmath224 such that @xmath225 applying lemma [ cps lemma ] to the operators @xmath4 and @xmath226 ( respectively , @xmath227 ) , we have @xmath228 therefore , @xmath229 since @xmath230 and @xmath231 it follows from remark [ chei remark ] that @xmath232 for an arbitrary @xmath233 we set @xmath234,$ ] @xmath235 from the first part of the proof , we have @xmath236 since @xmath237 in @xmath36 we have by theorem [ zeta linearity ] @xmath238 on the other hand , by lemma [ continuity lemma ] we have @xmath239 the following is our main result on the @xmath0function .
[ zeta reduction ] if @xmath13 is a generalised limit , then @xmath240 it is sufficient to prove the assertion for @xmath241 by theorems [ zeta linearity ] and [ fs is dixmier ] , we know that @xmath196 is a dixmier trace on @xmath35 hence , by , we have @xmath242 the assertion follows now from proposition [ main reduction lemma ] .
our remaining objective in this section is to provide strengthening of several formulae linking dixmier traces and @xmath243-functions from @xcite .
[ convexity lemma ] let @xmath244 the mapping @xmath245 is convex and , therefore , continuous . for all @xmath246 we have @xmath247 therefore , for every @xmath248 @xmath249 let @xmath250 and let @xmath251 since the mapping @xmath252 is convex for every @xmath253 it follows from the spectral theorem that the map @xmath254 is also convex .
therefore , for all positive real numbers @xmath255 and @xmath256 we have @xmath257 the assertion follows immediately .
let @xmath187 be a generalised limit on @xmath258 below , we will formally apply the notation @xmath259 introduced in to some unbounded positive operators @xmath5 on @xmath260 [ simple technical estimate ] let @xmath97 be a positive @xmath161compact operator and let @xmath261 be an unbounded operator commuting with @xmath83 if ( the closure of ) the product @xmath262 and @xmath263 for every @xmath19 then @xmath264 it follows from @xmath265 and @xmath266 that @xmath267 the inequality @xmath268 follows immediately .
set @xmath269 @xmath270 and observe that @xmath271 setting @xmath272,$ ] we obtain @xmath273\leq(c_na)^{1/2n}e^a[0,c_n^{-1}]\leq 1.\ ] ] it follows from that @xmath274 thus , @xmath275 since @xmath4 is @xmath161compact , then @xmath276 is bounded operator with finite support . due to the linearity with respect to @xmath277 we have @xmath278
the assertion follows now from lemma [ convexity lemma ] .
the following result is mainly known ( see @xcite ) .
our proof is however much simpler than the arguments used there .
[ zeta dix ] if @xmath33 is a dilation invariant generalised limit such that the generalised limit @xmath46 is still dilation invariant , then @xmath279 it is sufficient to verify the equality @xmath280 on positive operators @xmath281 such that @xmath282 define a continuously increasing function @xmath283 by @xmath284 let @xmath285 be as in theorem [ karamata ] as applied to the above @xmath286 define an operator @xmath266 by the formula @xmath287 and set @xmath288 we have @xmath289 the conditions of lemma [ simple technical estimate ] are valid for @xmath5 and @xmath290 indeed , @xmath5 commutes with @xmath291 @xmath292 and @xmath293 for every @xmath294 by lemma [ simple technical estimate ] , we have @xmath295 by theorem [ lidskii formula ] , we have @xmath296 we can now conclude @xmath297 the following corollary strengthens and extends the results of ( * ? ? ?
* theorem 4.11 ) and ( * ? ? ?
* theorem 3.8 ) .
it follows immediately from theorems [ zeta dix ] and [ zeta reduction ] . if @xmath33 is a dilation invariant generalised limit such that the generalised limit @xmath46 is still dilation invariant , then @xmath298
in this section we focus on functionals @xmath300 defined in .
it was implicitly proved in ( * ? ? ?
* theorem 5.2 ) that @xmath301 and therefore , @xmath302 is finite for every @xmath303 and every generalised limit @xmath187 on @xmath258 we note , in passing that a stronger result than ( * ? ? ?
* theorem 5.2 ) is established in theorem [ mhk bounded ] below .
let @xmath55 for a functional @xmath304 we set @xmath305 it is probably a difficult task to describe the set of all generalised limits @xmath187 for which yields a linear functional @xmath306 however , the class of linear functionals @xmath307 is an easier object .
below in proposition [ coincidence ] , we show that the sets of linear functionals @xmath308 and linear functionals @xmath309 coincide .
[ stupid m remark ] for every locally integrable @xmath310 with @xmath311 we have @xmath312 here , @xmath313 is the space of all bounded continuous functions tending to @xmath314 at @xmath315 fix @xmath130 the assertion follows by writing @xmath316 and noting that the assumption @xmath317 easily implies that @xmath318 [ coincidence ] suppose that a generalised limit @xmath187 on @xmath26 is such that @xmath307 is a linear functional on @xmath35 then , there exists a dilation invariant generalised limit @xmath33 on @xmath26 such that @xmath319 fix @xmath248 and observe that @xmath320 therefore , @xmath321 by the assumption , we have @xmath322 and appealing to lemma [ stupid m remark ] , we obtain @xmath323 let @xmath324 be the linear span of the functions @xmath325 and let @xmath326 we claim that the space @xmath327 is dilation invariant .
indeed , it follows from lemma [ stupid m remark ] and that every function @xmath328 belongs to the set @xmath329 it follows from that @xmath330 on @xmath331 by the invariant form of the hahn - banach theorem ( see @xcite ) applied to the group of dilations @xmath332 we see that @xmath333 can be extended to a dilation invariant generalised limit @xmath33 on @xmath41 the following lemma can be found in @xcite .
we present a shorter proof for convenience of the reader .
[ xiomega repr0 ] if @xmath33 is a dilation invariant generalised limit on @xmath334 then @xmath335 it follows from that @xmath336 setting @xmath337 multiplying both sides of by @xmath338 and applying theorem [ karamata ] to @xmath339 ( which is dilation invariant , see @xcite ) , we obtain .
[ xiomega repr ] let @xmath340 and let @xmath33 be a dilation invariant generalised limit on @xmath258 we have @xmath341 in view of lemma [ xiomega repr0 ] , it is sufficient to show that right hand sides of and coincide
. this easily follows from the following computation , where we use integration by parts @xmath342 @xmath343 [ full symmetricity for hk ] let @xmath33 be a dilation invariant generalised limit on @xmath26 and let @xmath125 be such that @xmath59 we have @xmath344 the assertion follows from lemma [ xiomega repr ] and theorem [ t equals majorization ] .
the following is the main result of this section .
[ xiomega linearity ] for any dilation invariant generalised limit @xmath33 on @xmath334 the functional @xmath345 given by is linear and fully symmetric on @xmath35 the assertion follows from lemma [ full symmetricity for hk ] provided we have shown that @xmath346 to this end , we observe first that since @xmath33 and @xmath339 are dilation invariant , it follows from lemma [ full symmetricity for hk ] and that @xmath347 now , let @xmath348 and @xmath349 be disjoint copies of @xmath4 and @xmath5 ( see section [ definot ] ) .
thus , we have @xmath350 however , the equality @xmath351 for positive operators @xmath348 and @xmath349 such that @xmath352 follows immediately from the definition . since the equalities @xmath353 @xmath354 are obvious , we arrive at .
it follows from theorem [ xiomega linearity ] and theorem [ fs is dixmier ] , that the functional @xmath345 is a fully symmetric functional on @xmath16 whenever @xmath33 is a dilation invariant generalised limit @xmath33 on @xmath258 in this section , we show the converse .
define a ( non - linear ) operator @xmath355 by the formula @xmath356 we need some properties of the operator @xmath357 firstly , we show that it is additive on certain pairs of @xmath358 [ disjoint linearity ] let @xmath359 be such that @xmath360 it follows that @xmath361 it follows immediately from the assumption that @xmath362 next , we explain the connection of the operator @xmath20 with fully symmetric functionals on @xmath35 [ fs ab estimate ] let the operators @xmath125 be such that @xmath363 for every fully symmetric functional @xmath364 on @xmath36 we have @xmath365 it follows immediately from the definition that @xmath366 applying theorem [ t equals majorization ] we obtain @xmath116 and so @xmath365 [ razn1 ] let @xmath367 for every fully symmetric functional @xmath364 on @xmath36 we have @xmath368 without loss of generality , @xmath369 denote the right hand side by @xmath370 and suppose that @xmath371 ( the case when @xmath372 is treated similarly ) .
fix @xmath373 we have @xmath374 for all sufficiently large @xmath375
let @xmath348 be an operator with @xmath376 we have @xmath377 for all sufficiently large @xmath375 let @xmath378 and @xmath379 be disjoint copies of @xmath4 and @xmath291 respectively .
it follows from lemma [ disjoint linearity ] that @xmath380 for all sufficiently large @xmath375 choose @xmath381 small enough to guarantee @xmath382 for all @xmath383 where @xmath384 by corollary [ fs ab estimate ] , we have @xmath385 or equivalently @xmath386 since @xmath387 is arbitrarily small , we are done .
[ ro est ] let @xmath388 and let @xmath389 for some @xmath235 for every fully symmetric functional @xmath364 on @xmath16 we have @xmath390 both sides of the inequality depend continuously on the @xmath391 s . without loss of generality , we may assume that all @xmath392 multiplying both sides by the common denominator , we may assume that all @xmath393 writing @xmath394 we see that it is sufficient to prove only for the case when @xmath395 for every @xmath396 let @xmath397 be a disjoint copy sequence of @xmath398 both sides of the inequality do not change if we replace @xmath399 with @xmath400 without loss of generality , the operators @xmath401 @xmath402 by lemma [ razn1 ] we have @xmath403 @xmath404 since @xmath405 for all @xmath406 we have by lemma [ disjoint linearity ] that @xmath407 and the assertion follows .
let @xmath324 be the linear span of @xmath408 and @xmath409 for every @xmath248 we have @xmath410 it follows from the definition that for every @xmath111 we have @xmath411 let @xmath364 be a normalised fully symmetric functional on @xmath35 we need the following linear functional on @xmath412 [ ro defi ] for every @xmath413 such that @xmath414 we set @xmath415 that @xmath416 is well - defined is proved below .
[ rho estimate ] the linear functional @xmath417 is well - defined .
for every @xmath418 we have @xmath419 let @xmath413 be such that @xmath420 we have @xmath421 it follows from lemma [ ro est ] that @xmath422 so that @xmath416 is well - defined .
the second assertion directly follows from lemma [ ro est ] .
[ main fs lemma ] let @xmath364 be a normalised fully symmetric functional on @xmath35 there exists a dilation invariant generalised limit @xmath33 on @xmath26 such that @xmath423 for every @xmath180 for every @xmath424 we have @xmath425 therefore , @xmath416 is @xmath426invariant on @xmath412 it follows from lemma [ rho estimate ] that @xmath427 by the invariant form of the hahn - banach theorem ( see @xcite ) applied to the group of dilations @xmath332 we can extend @xmath416 to a dilation invariant generalised limit on @xmath258 the following assertion is the main result of this section .
it permits representation of a fully symmetric functional @xmath364 via heat kernel formulae .
[ also main ] let @xmath364 be a fully symmetric functional on @xmath35 there exists dilation invariant generalised limit @xmath33 on @xmath26 such that @xmath428 it follows from lemma [ main fs lemma ] that there exists a dilation invariant generalised limit @xmath33 such that @xmath429 the assertion follows now from lemma [ xiomega repr ] .
it is known ( see ( * ? ? ? * theorem 33 ) and the more general result in corollary [ corollary 51 ] below ) that the equality @xmath430 holds for every @xmath47invariant generalised limit @xmath33 on @xmath26 ( see also earlier results with more restrictive assumptions on @xmath33 in ( * ? ? ?
* theorem 4.1 ) and ( * ? ? ? * theorem 5.2 ) ) . in view of theorem
[ also main ] and theorem [ fs is dixmier ] , it is quite natural to ask whether the equality above holds for every dilation invariant generalised limit @xmath45 in this section we prove that this is not the case . [ est1 ] let @xmath33 be a dilation invariant generalised limit on @xmath258 for every @xmath431 we have @xmath432 @xmath433})=0.\ ] ] denote the left hand side of by @xmath434 due to the dilation invariance of @xmath29 we have @xmath435 since @xmath52 is monotone and bounded , we have @xmath436 denote the left hand side of by @xmath437 due to the dilation invariance of @xmath29 we have @xmath438})=g(st)-g(t),\quad s , t>1.\ ] ] since @xmath439 is monotone and bounded , we have @xmath440 [ est12 ] let @xmath33 be a dilation invariant generalised limit on @xmath258 we have a. @xmath441}(t))=0.\ ] ] b. @xmath442}(t))=0.\ ] ] we only prove the first assertion .
proof of the second one is similar .
fix @xmath443 we have @xmath444 and , therefore , @xmath445}(t)\leq\frac2s+\sum_k\chi_{[e^{k+e^k}/s , e^{k+e^k}]}(t)+\ ] ] @xmath446}(t).\ ] ] clearly , @xmath447}(t))=0.\ ] ] it follows from the lemma [ est1 ] that @xmath441}(t))\leq\frac2s.\ ] ] since @xmath448 is arbitrarily large , we have @xmath441}(t))=0.\ ] ] [ omega construct ] there exists a dilation invariant generalised limit @xmath33 on @xmath26 such that @xmath449 define a positive , homogeneous functional @xmath450 on @xmath26 by the formula @xmath451 it is verified in ( * ? ? ?
* lemma 4 ) that every @xmath452 satisfying @xmath453 is dilation invariant .
observing that @xmath454 let us select @xmath452 satisfying @xmath453 and such that @xmath455 therefore , @xmath456 define a function @xmath457 by the formula @xmath458}.\ ] ] fix @xmath459 for every @xmath460,$ ] we have @xmath461 which guarantees @xmath462 [ est2 ] let @xmath457 be as in and let @xmath33 be as in lemma [ omega construct ] .
we have @xmath463 fix @xmath460.$ ] we have @xmath464 it follows that @xmath465}(t))+\ ] ] @xmath466}(t)).\ ] ] by lemma [ est12 ] , the second generalised limit above vanishes .
we claim that the first generalised limit above is @xmath467 indeed , @xmath468}(t)\geq(1+o(1))\sum_k\chi_{[e^{e^k},e^{k+e^k}]}(t)\ ] ] and @xmath468}(t)\leq\sum_k\chi_{[e^{e^k},e^{k+e^k}]}(t)+e\sum_k\chi_{[e^{k-1+e^{k-1}},e^{e^k}]}.\ ] ] the claim follows from lemma [ omega construct ] .
[ est3 ] let @xmath457 be as in and let @xmath33 be as in lemma [ omega construct ] .
we have @xmath469 fix @xmath470 we have @xmath471 this estimate and lemma [ xiomega repr ] yield @xmath472}(t))-\ ] ] @xmath473}(t)).\ ] ] it follows from lemma [ est12 ] that the second generalised limit is @xmath474 we claim that the first generalised limit is @xmath467 indeed , @xmath475}(t)\geq(1+o(1))\sum_k\chi_{[e^{e^k},e^{k+e^k}]}\ ] ] and @xmath475}(t)\leq 1.\ ] ] the claim follows from lemma [ omega construct ] .
the following theorem delivers the promised counterexample .
[ counterexample ] there exists @xmath139 and dilation invariant generalised limit @xmath33 on @xmath26 such that @xmath476 for brevity , we assume that the von neumann algebra @xmath144 is of type @xmath477 ( the argument can be easily adjusted when @xmath144 is of type @xmath478 ) .
let @xmath457 be as in and let @xmath340 be such that @xmath479 the assertion follows from lemmas [ est2 ] and [ est3 ] .
let @xmath33 be a dilation invariant generalised limit on @xmath26 and let @xmath99 following @xcite , we consider the functionals on @xmath136 defined by the formula @xmath480 the main result of this section , theorem [ mhk bounded ] , shows that the function @xmath481 is bounded , and so the formula is well - defined .
[ operator estimate1 ] let @xmath180 we have @xmath482)=o(t^{-1}\log(t))$ ] as @xmath483 let @xmath484 we have @xmath485 fix @xmath163 define decreasing function @xmath486 by setting @xmath487 define a decreasing function @xmath488 by setting @xmath489 we claim that @xmath490 indeed , @xmath491 for @xmath492 and @xmath493 for @xmath494 it follows that @xmath495)\leq\int_0^{\infty}y_t^2(s)ds\leq\int_0^{\infty}x_t^2(s)ds.\ ] ] we have @xmath496 [ square estimate ] let @xmath497}(t)$ ] and let @xmath180 we have @xmath498 for fixed
@xmath383 we have @xmath499)ds=\frac1{\log(t)}\tau(a^2\int_1^te^a[0,\frac1s]ds).\ ] ] integrating by parts , we obtain @xmath500ds = se^a[0,\frac1s]|_1^t-\int_1^tsde^a[0,\frac1s]=se^a[0,\frac1s]|_1^t+\int_{1/t}^1u^{-1}de^a[\frac1t , u]=\ ] ] @xmath501+te^a[0,\frac1t].\ ] ] therefore , @xmath502)+o(\frac1{\log(t)}).\ ] ] it follows from the definitions of @xmath503 and @xmath504 that for every @xmath139 and every @xmath383 we have @xmath505 clearly , @xmath506)=\frac1{\log(t)}\int_0^{d_a(1/t)}\mu(s , a)ds\leq\frac{\log(d_a(1/t))}{\log(t)}\|a\|_{1,\infty}\in l_{\infty}.\ ] ] the assertion follows now from the lemma [ operator estimate1 ] .
[ mhk bounded ] let a bounded function @xmath507 be such that @xmath508 .
let @xmath340 and let @xmath99 we have @xmath509 due to the well known inequality @xmath510 it suffices to prove the theorem only when @xmath511 in this case , for the function @xmath512}(t),$ ] the assertion follows from lemma [ square estimate ] .
if @xmath513 then it holds trivially .
thus , it holds for the function @xmath514 finally , observe that the assumptions on @xmath52 guarantee that there exists a constant @xmath515 such that @xmath516 since the function @xmath517 satisfies the assumptions of theorem [ mhk bounded ] we obtain the following corollary , which was implicitly proved in ( * ? ? ?
* theorem 5.2 ) . for every @xmath518 and every @xmath424 we have @xmath519
the results of this section extend and generalise those of ( * ? ? ?
* theorem 4.1 ) and ( * ? ? ? * theorem 5.2 ) .
we also give an answer to the question asked in @xcite .
we explicitly prove that the functional @xmath520 ( extended to @xmath16 as in ) is linear on @xmath35 by theorem [ mhk bounded ] , for every @xmath527 we have @xmath528)^2)\right)\in l_{\infty}(0,\infty).\ ] ] since @xmath33 is dilation invariant , we conclude @xmath529)^2))=\varepsilon(\omega\circ m)(\frac1t\tau((tae^a[0,\frac1t])^2)).\ ] ] the assertion follows immediately .
since @xmath52 is increasing on @xmath536 $ ] and @xmath542 we have @xmath543 therefore , @xmath544 and @xmath545 we have @xmath546 for all @xmath547 we have @xmath548 therefore , @xmath549 hence , @xmath550
@xmath551 @xmath552 both coefficients in the latter formula tend to @xmath553 let @xmath52 satisfy the assumptions above .
observe that the assertion of lemma [ seredina ] holds for the function @xmath555},$ ] where @xmath556 indeed , every such function is a function of bounded variation and therefore may be written as a difference of two monotone functions .
now the assertion follows from lemmas [ golova k nulyu],[hvost k nulyu],[seredina ] by setting @xmath557 and @xmath558 and letting @xmath559 it follows from the definition of cesaro operator @xmath151 that @xmath562 integrating by parts , we obtain @xmath563 @xmath564 @xmath565 evidently , @xmath566 therefore , @xmath567 the assertion follows now from lemma [ sborka ] .
the following theorem extends the results of @xcite and gives an affirmative answer to the question stated in @xcite .
it also shows that the functionals @xmath575 are linear functionals on @xmath16 for a wide class of functions @xmath576 thus , we have the equality valid for every @xmath5 bounded from below by a strictly positive constant . set @xmath585.$ ] it follows that equality holds with @xmath5 replaced with @xmath586 throughout . by lemma [ hk continuity ] , we have @xmath587 since @xmath588 in @xmath16 and since @xmath345 is bounded on @xmath36 we have @xmath589 the assertion follows immediately .
100 m. benameur , t. fack , _ type ii noncommutative geometry . i. dixmier trace in von neumann algebras , _ adv.math . * 199 * ( 2006 ) , 29 - 87 j. bourin , _ convexity or concavity inequalities for hermitian operators , _ math .
* 7 * ( 2004 ) , no .
4 , 607620 . l. brown , h. kosaki , _ jensen s inequality in semi - finite von neumann algebras , _ j. operator theory
* 23 * ( 1990 ) , no .
a. l. carey , v. gayral , a. rennie , f. sukochev , _ integration on locally compact noncommutative spaces _ ,
a. carey , j. phillips , f. sukochev , _ spectral flow and dixmier traces , _ adv.math . * 173 * ( 2003 ) , no .
1 , 68113 .
a. carey , a. rennie , a. sedaev and f. sukochev , _ the dixmier trace and asymptotics of zeta functions , _ j.funct.anal .
* 249 * ( 2007 ) , no .
2 , 253283 .
a. carey , f. sukochev , _ dixmier traces and some applications to noncommutative geometry , _ ( russian ) uspekhi mat .
nauk * 61 * ( 2006 ) , no . 6 ( 372 ) , 45110 ; translation in russian math .
surveys * 61 * ( 2006 ) , no . 6 ,
10391099 a. connes , _ noncommutative geometry , _ academic press , san diego 1994
. j. dixmier , _ existence de traces non normales , _ c. r. acad .
paris * 262 * ( 1966 ) , a1107-a1108 .
p. dodds , b. de pagter , a. sedaev , e. semenov and f. sukochev , _ singular symmetric functionals , _ zap .
nauchn . sem .
otdel . mat .
inst . steklov .
( pomi ) * 290 * ( 2002 ) issled .
po linein .
30,42 - 71 ( russian ) .
english translation in j. math .
( n. y. ) * 124 * ( 2 ) ( 2004 ) , 48674885 .
p. dodds , b. de pagter , a. sedaev , e. semenov and f. sukochev , _ singular symmetric functionals with additional invariance properties _ , ( russian ) izv .
nauk ser . mat .
* 67 * ( 6 ) ( 2003 ) , 111136 .
english translation in izvestiya : mathematics * 67 * ( 2003 ) , 1187 - 1213 .
p. dodds , b. de pagter , e. semenov and f. sukochev , _ symmetric functionals and singular traces _ , positivity * 2 * ( 1998 ) , no .
1 , 4775 .
r. edwards , _ functional analysis , _ holt , rinehart and winston , new york , 1965 .
t. fack , h. kosaki , _
generalized @xmath448-numbers of @xmath143-measurable operators , _ pacific j. math .
* 123 * ( 1986 ) , no .
2 , 269300 . i. gohberg , m. krein , _ introduction to the theory of linear nonselfadjoint operators _ , translations of mathematical monographs , vol . 18 american mathematical society , providence , r.i .
koplienko , _ trace formula for nontrace - class perturbations , _ sibirsk .
zh . * 25 * ( 1984 ) , no . 5 , 6271 ( russian ) .
english translation in sib .
j. * 25 * ( 1984 ) , no . 5 , 735743 . s. krein , ju .
petunin and e. semenov , _ interpolation of linear operators , _ nauka , moscow , 1978 ( russian ) .
english translation in translation of mathematical monographs , amer .
* 54 * ( 1982 ) .
n. kalton and f. sukochev , _ rearrangement - invariant functionals with applications to traces on symmetrically normed ideals , _ canad .
math . bull .
* 51 * ( 2008 ) , 6780 .
n. kalton , a. sedaev , f. sukochev , _ fully symmetric functionals on a marcinkiewicz space are dixmier traces , _ submitted .
murray , j. von neumann , _ on rings of operators , _ ann .
* 37 * ( 1936 ) , no . 1 , 116 - 229 .
pietsch a. _ about the banach envelope of @xmath596 _ , rev .
* 22 * ( 1 ) ( 2009 ) 209226 .
sedaev a. _ generalized limits and related asymptotic formulas _ , math
. notes . * 86*:4 ( 2009 ) , 612 - 627 . b. simon , _ trace ideals and their applications , _ ams 2005 .
a. sedaev , f. sukochev , d. zanin , _ lidskii - type formulae for dixmier traces , _ int.eq.oper.th .
( to appear ) http://arxiv.org/pdf/1003.1817 | we present a systematic study of asymptotic behavior of ( generalised ) @xmath0functions and heat kernels used in noncommutative geometry and clarify their connections with dixmier traces .
we strengthen and complete a number of results from the recent literature and answer ( in the affirmative ) the question raised by m. benameur and t. fack @xcite . zeta function , heat kernel formulae , dixmier trace , noncommutative geometry .
primary : 58b34,46l51 , 46l52,58j42 |
on a lattice left - handed weyl fermions @xmath0 are always accompanied by their right - handed counterparts @xmath1 , unless certain mild conditions for the action are broken @xcite .
such a doubling is present both in the wilson @xcite and in the domain wall formulations @xcite ( for a review , see @xcite ) , being the main obstacle to the non - perturbative definition of chiral gauge theories .
so far all attempts to make the chiral counterpart @xmath1 sterile , in a way that does not use a hard breaking of gauge symmetry and leaves an interacting chiral theory , have failed ( see , for example , reviews @xcite and references therein ) . another conceivable way to define a theory with only fermion fields
@xmath0 is to decouple @xmath1 by giving them majorana masses of the order of the inverse lattice spacing .
it can be done directly if the fermions belongs to real representation of the gauge group .
if they belong to complex representation , a generalized higgs mechanism is to be employed in order the majorana mass not to break the gauge invariance . in this case model
must have strong coupling paramagnetic ( pms ) phase , where fermions acquire masses of the order of the inverse lattice spacing , while the gauge symmetry is not spontaneously broken , i.e. no chiral condensate or no vacuum expectation value of the higgs fields arise ( for a review and further references , see @xcite ) . in the wilson formulation ,
the @xmath0 and @xmath1 are coupled through the wilson term .
therefore the introduction of a majorana mass for @xmath1 generates a majorana coupling for @xmath0 .
so in the gauge theory fine tuning becomes necessary not only for dirac mass but also for majorana mass of @xmath0 .
furthermore , in the case of complex representations , there arises a serious problem with the properties of the model within the pms phase , where its spectrum either becomes vectorlike , or consists only of neutral , i.e. singlet under the gauge group , chiral fermions , whose gauge interactions very likely vanish in the continuum limit @xcite . in the domain wall formulation the chiral fermions @xmath0 and @xmath1 appear as collective states of coupled five - dimension fermions .
these states are localized at two surfaces formed by mass defects in the five - dimensional system @xcite or by free boundaries of the five - dimensional space @xcite .
these surfaces are separated in the fifth dimension and the overlap between these states is suppressed exponentially with this distance .
this gives rise to a hope that the above problems can be avoided in such a formulation .
this , in fact , is underlying idea of the recent proposal @xcite for lattice formulation of the standard model . in order to generate the majorana mass for the collective state @xmath1
, it has been suggested to introduce on the surface at which @xmath1 is localized certain gauge invariant four - fermion interactions motivated by the approach @xcite .
a similar idea is followed by the proposal @xcite for lattice formulation of @xmath2 supersymmetric theories . in this case
the majorana mass is introduced on the same surface directly , for the fermions belong to real representation .
thus , yet more questions , which should be answered first , arise in such an approach : ( @xmath3 ) whether the generation of the majorana mass on one of the surfaces leads to the chiral spectrum of the model ?
this question is common to both proposals , @xcite and @xcite , and requires special investigation , since the chiral states in the domain wall formulation do not present in the action explicitly .
( @xmath4 ) whether the pms phase exists in the systems employing the higgs mechanism ?
although some indirect arguments in favour of the presence of the pms phase has already been given in @xcite , it seems interesting to look at the problem from a more general point of view .
( @xmath5 ) does the model has chiral spectrum in the pms phase , and if so , what fermions , charged or neutral , form it ?
this is a crucial question to all formulations of chiral gauge theories that employ the pms phase .
the aim of this paper is to answer these questions .
we consider the variant of the domain wall formulation with free lattice boundaries @xcite and the majorana mass or higgs - majorana coupling introduced on one of the boundaries .
these models are introduced in section 2 . in section 3
we analyze the fermion propagators in the model with the majorana mass and gauge fields switched off , and show that the low - lying spectrum of such a model is chiral . in section 4 using a mean - field technique we demonstrate the existence of the pms phase in the systems with the higgs - majorana coupling . in section 5
we consider such systems within the pms phase and argue that they may have chiral spectrum only for the fermions that are singlets under the gauge group .
section 6 contains a summary and a discussion of possible applications of such models .
our conventions are the following .
we consider euclidean hypercubic @xmath6-dimensional lattice with spacing @xmath7 , which is set to one unless otherwise indicated , and volume @xmath8 with @xmath9 even .
the lattice sites numbered by @xmath6-dimensional vectors @xmath10 , where @xmath11 , and @xmath12 ; @xmath13 are unit vectors along positive directions in four - dimensional space .
fermion ( boson ) variables obey antiperiodic ( periodic ) boundary conditions for the first four directions and free boundary conditions in the fifth direction .
we consider the variant of the lattice domain wall fermions proposed in @xcite .
the action of such a model can be written in the form : @xmath14 = \sum_{m , n , s , t}[\overline{\psi}_{m , s } \delta_{s t } d_{m n } \psi_{n , t } + \overline{\chi}_{m , s } \delta_{s t } \overline{d}_{m n } \chi_{n , t } \cr & & \quad + \overline{\psi}_{m , s } ( \delta_{s-1 \ , t}\delta_{m n } - \delta_{s t } w_{m n } ) \chi_{n , t } + \overline{\chi}_{m , s } ( \delta_{s+1 \ , t}\delta_{m n } - \delta_{s t } w_{m n } ) \psi_{n , t } ] , \label{eq : dmf}\end{aligned}\ ] ] where @xmath15 and @xmath16 are two - component weyl fermions in the five - dimensional space , transforming under the four - dimensional rotations as left- and right - handed spinors , respectively , @xmath17 @xmath18 are the pauli matrices , @xmath19 are four - dimensional gauge variables , and @xmath20 , but without loss of generality we can restrict it as indicated above . ] is intrinsic mass parameter of the formulation ( ` domain wall ' mass ) . in ( [ eq : dmf ] ) it is implied that @xmath21 . both @xmath0 and @xmath1 belong to the same representation @xmath22 of the gauge group , so the action ( [ eq : dmf ] ) is gauge invariant . without loss of generality
we consider unitary groups .
the chiral fermions in such a formulation arise as surface modes : the left- and right - handed fermions are low momentum states localized at the @xmath23 and @xmath24 boundaries , respectively .
so these states are separated in the fifth dimension and overlap between them is suppressed exponentially with @xmath25 . the idea of refs .
@xcite is to introduce on the surface @xmath24 certain gauge invariant terms which might generate majorana masses @xmath26 for right - handed fermions @xmath1 .
such terms can always be represented in the form bilinear in the fermions , so we consider the following action @xmath27 = a_0[u ] + a_m[h ] , \cr & & a_m[h ] = \sum_{n } ( \chi_{n , n_s}^{t } h_{n } \chi_{n , n_s } + \overline{\chi}_{n , n_s } h^{\dagger}_{n } \overline{\chi}_{n , n_s}^{t } ) .
\label{eq : dmf1}\end{aligned}\ ] ] if representation @xmath22 is ( pseudo)real , one can simply put @xmath28 where @xmath29 is a certain constant symmetric matrix whose form depends on the group and ensures the gauge invariance of the mass term requires more than one fermion generations . ] . if @xmath22 belongs to complex representation , @xmath30 has the form @xmath31 where @xmath32 is the higgs - majorana coupling , and @xmath33 is the higgs field transforming under the gauge group as @xmath34 .
so the action ( [ eq : dmf1 ] ) is gauge invariant . in this case
the majorana mass @xmath26 is to be generated without spontaneous breaking of the gauge symmetry .
it is achieved in the pms phase , provided the system has such a phase .
let us examine first the case of the explicit mass ( [ eq : mr ] ) .
complete information about euclidean system is contained in the correlators , or propagators , of the fields involving in the action .
however , before we shall analyse them , it seems to be instructive to get a qualitative idea about its spectrum that can be easily obtained in a certain limiting case from the equations of motion . for our system with the gauge interactions switched off ( @xmath35 ) these equations can be written in the momentum space as @xmath36 , where @xmath37 is @xmath38 matrix , and @xmath39 is the column constructed from the fields @xmath0 , @xmath1 , @xmath40 , and @xmath41 .
the solutions to this equations are determined by zeros of det@xmath37 . since the formulation is reflection positive
, one can safely continue the equations to the minkowski space , so that @xmath42 , where @xmath43 is three - vector . then these zeros will determine the energy spectrum of the system after its quantization . to make the analysis as simple as possible , consider four dimensional space continuous and take the limit @xmath44 .
then @xmath45 , and the above equation can be reduced to two independent sets of equations : @xmath46 , and @xmath47 , where @xmath48 are diagonal @xmath49 matrices @xmath50 from these expressions we immediately can read off that in the @xmath0 sector there are one massless left - handed fermion ( @xmath51 ) localized on the surface @xmath23 , and @xmath25 massive solutions with the unit masses ( @xmath52 ) , that generally are not localized . in the @xmath1 sector
there are no massless fermions . instead , in addition to non - localized solutions with the unit masses , which form with their @xmath0 counterparts @xmath25 massive dirac fermions , we have right - handed fermion with majorana mass @xmath53 localized on the surface @xmath24 . in the limit @xmath54 it turns to the mirror massless mode of the domain wall formulation .
thus , we can conclude that such simplified system has desirable chiral spectrum .
of course , one can easily find also the fermion propagators in this limiting case . returning to the euclidean space
we get @xmath55 , \cr & & \langle \chi_s \overline{\chi}_t\rangle = - d\left[\delta_{s t } \frac{1}{p^2 + 1 } + \delta_{s n_s } \delta_{t n_s}\left(\frac{1}{p^2 + 4 m^2 } - \frac{1}{p^2 + 1}\right)\right ] , \cr & & \langle \psi_s \overline{\chi}_t\rangle = \delta_{s \ , t+1}\frac{1 } { p^2 + 1 } , \quad s \neq 0 , \ , t \neq n_s , \cr & & \langle \psi_0 \overline{\chi}_t\rangle = \langle \psi_s \overline{\chi}_{n_s}\rangle = 0.\end{aligned}\ ] ] let us now consider the fermion propagators of our lattice model .
the most simple way to do that is to express them in terms of the propagators determined by the original action ( [ eq : dmf ] ) .
we shall use for them notation @xmath56 with @xmath57 and @xmath58 .
they have been calculated in @xcite and in the four - dimensional momentum space have the following form : @xmath59 where @xmath60 , @xmath61 , and @xmath62 and @xmath63 are fourier transforms of the operators @xmath64 and @xmath65 in ( [ eq : def1 ] ) at @xmath35 : @xmath66 , @xmath67 . in the approximation
specified bellow one has @xmath68 with @xmath69 . here
@xmath70 , @xmath71 and @xmath72 are functions of @xmath73 , but only @xmath70 has a pole at @xmath74 : @xmath75 where @xmath76 .
( [ eq : g ] ) and ( [ eq : ab ] ) are given in the approximation where all terms @xmath77 are neglected .
since the @xmath78 is positive definite , the larger @xmath25 , the better such an approximation is justified . at @xmath79
one has @xmath80 , so this quantity can be considered as the accuracy with which this formulation defines chiral fermions at finite @xmath25 .
now the propagators for the system ( [ eq : dmf1 ] ) with the majorana mass can be represented as @xmath81\right ) \overline\psi_b \rangle_0,\ ] ] and similarly for @xmath82 and @xmath83 .
thus after some algebra we arrive at the following expressions @xmath84 consider now the low - momentum structure of the propagator @xmath85 . in the approximation specified above we have : @xmath86 , @xmath87 , and @xmath88 .
then , from ( [ eq : prop ] ) it follows that the introduction of the majorana mass modifies only function @xmath70 [ cf .
( [ eq : prop0 ] ) , ( [ eq : g ] ) ] : @xmath89 , \cr & & a^m_l = a_l - \frac{4m^2 \overline p^2 ( a_l+b)^3}{1 + 4m^2 \overline p^2 ( a_l+b)^2},\end{aligned}\ ] ] and this modification exactly cancels the pole in @xmath90 : @xmath91 note that this result matches well with our analysis in the beginning of this section , for in the limit considered there @xmath92 .
on the other hand , in the same approximation the propagator @xmath93 takes the form @xmath94,\end{aligned}\ ] ] with @xmath95 since @xmath96 , we get @xmath97 therefore , the pole structure of the propagator of the field @xmath0 is not affected by the majorana mass acquired by the field @xmath1 . finally , in order to see the effect of the fermion number violation in the physical left - handed sector @xmath0 , caused by the majorana mass term for @xmath1 , we consider the propagator @xmath98 . in the above approximation
we get @xmath99 thus , for physical left - handed fermions the effect is suppressed exponentially with @xmath25 .
we should note that the action of the original domain wall model @xcite can be taken as the action @xmath100 in ( [ eq : dmf ] ) .
the analysis of the spectrum and the propagators in this case , though is more complicated , leads essentially to the same conclusions .
let us now consider the question of the existence of the pms phase in the systems with the higgs - majorana coupling .
whether a system has the pms phase can be examined within mean - field approximation .
we consider the case of radially frozen higgs field , @xmath101 , using the technique developed in @xcite .
for that , it is sufficient to consider the gauge interactions turned off .
we write down the field @xmath102 as @xmath103 , with @xmath104 , and @xmath105 matrices determined by the group and its representation .
then the action for the @xmath33 takes the form @xmath106 where @xmath107 is a hopping parameter . not to break the reflection positivity of the model
, we shall consider it at @xmath108 .
the partition function of the system reads as @xmath109 ) \cr & & \quad = \int \prod_n d \phi_n \exp ( -a_{\phi } + \ln z_f[y \phi ] ) , \label{eq : sys1}\end{aligned}\ ] ] where @xmath110 is the haar measure on the group , and @xmath111 $ ] is the fermion partition function in the external field @xmath33 .
according to @xcite , the critical lines separating symmetric ( paramagnetic ) phases , where the vacuum expectation value of the higgs field @xmath112 , from the broken ( ferromagnetic ) phases , where @xmath113 , are determined by the expression @xmath114 \rangle_h \right| _ { h = 0 } , \label{eq : kkr}\ ] ] where @xmath115^{1/2}$ ] is the mean field , @xmath116 \rangle_h = \frac{\int \prod_n d \phi_n \ln z_f[y \phi ] \exp \sum_{n , a } h^a \phi^{a}_{n}}{\int \prod_n d \phi_n \exp \sum_{n , a } h^a \phi^{a}_{n } } , \label{eq : exv}\ ] ] and @xmath117 and @xmath118 are group dependent positive numbers .
for instance , for u(1 ) one has @xmath119 . integrating in ( [ eq : sys1 ] ) over @xmath0 , and then over @xmath1 successively slice by slice in the fifth direction , we find @xmath120 = \mbox{det}^{n_s+1}[d ] \ : \prod_{s = 0}^{n_s-1 } \mbox{det}[b_s ] \ : \mbox{det}^{1/2 } [ 4 y^2 + \phi^{\dagger } \sigma_2 b_{n_s}^{t } \phi \sigma_2 b_{n_s } ] , \label{eq : z1}\ ] ] where det means the determinant in the space of four - dimensional lattice and spinors indices , and the operators @xmath121 are determined by the recursive relations @xmath122 with @xmath123 , and @xmath124 . since one is interested in the paramagnetic phase at strong coupling @xmath32 , it is sufficient to know expectation value ( [ eq : exv ] ) in the leading order in @xmath125 and @xmath126 . from ( [ eq : exv ] ) , ( [ eq : z1 ] ) , and ( [ eq : rea ] ) , using formulae of ref .
@xcite , we get @xmath116 \rangle_h = - \frac{c_3}{32 y^2 } h^2 \mbox{tr } [ \sigma_2 b_{n_s}^{t } \sigma_2 b_{n_s } ] + o(h^0 ) + o(\frac{1}{y^4 } , h^3 ) , \label{eq : exp2}\ ] ] where tr means the trace over the spatial and spinor indices , and @xmath127 is group dependent positive number ; in the case of u(1 ) it is equal to one , as well .
finally , from ( [ eq : kkr ] ) , ( [ eq : rea ] ) , and ( [ eq : exp2 ] ) , we get : @xmath128 where @xmath129 and @xmath130 .
we find @xmath131 numerically practically independent of @xmath25 .
for instance at @xmath132 we have : @xmath133 , @xmath134 , @xmath135 .
thus , the system is in the pms phase at @xmath136 , and @xmath137 .
the corresponding phase diagram is shown in fig . 1 .
this confirms the indirect arguments of ref .
@xcite .
let us assume now that the system is within such a phase and consider what may happen to it there .
since the higgs field fluctuates strongly in this phase , and @xmath138 , standard methods of examination of such systems , working in the broken phases , is inapplicable here .
we however still can get some idea of what may happen in this phase making use the method proposed in @xcite .
let us represent the higgs field in ( [ eq : dmf1 ] ) as @xmath139 where the group valued field @xmath140 , @xmath141 , transforms under the gauge group as @xmath142 .
following the arguments of @xcite , we assume that adequate nonzero parameter in the pms phase is the link expectation value @xmath143 further results will depend on the dynamical variables in terms of which the consideration is performed .
therefore we consider all the three possibilities .
introduce the gauge group singlet fields @xmath144 then , with taking into account ( [ eq : vev ] ) , the system ( [ eq : dmf1 ] ) in the pms phase at tree level can be described by the action @xmath145 where we omit summation over four - dimensional indices .
we see that in this case the massive neutral fermion @xmath146 is decoupled , for the wilson term at @xmath24 vanishes .
this made @xmath146 and @xmath147 naive , and although at the first glance this should not affect the chiral properties of the system , since both these fermions still have masses of the order of the cutoff , it turns out that it gives rise to species doubling of a massless mode .
it is clearly seen from the structure of the fermion determinant , which in this case takes the form [ cf .
( [ eq : z1 ] ) ] @xmath148 \ : \prod_{s = 0}^{n_s-1 } \mbox{det}[b_s ] \ : \mbox{det}^{1/2 } [ 4 \frac{y^2}{z^4 } - d \overline{d } ] , \label{eq : zn1}\ ] ] where the functions @xmath121 are determined in ( [ eq : rea ] ) . since @xmath149 , the determinant has zero at the corners of the brillouine zone .
so this scenario leads to the failure of the model .
* 2 . * together with the neutral fields ( [ eq : chin ] ) one can also introduce the neutral fields @xmath150 in terms of these variables the system in the pms phase takes the form @xmath151 where in the four - dimensional momentum space @xmath152 in this case the slice @xmath24 decouples from the rest of the lattice , so that one has massive neutral fermions at @xmath24 and the original massless model on the rest of the lattice .
the neutral fermions @xmath153 , @xmath146 can still be made massless , by setting the domain wall mass to @xmath154 , but this does not rescue the situation with the rest of the lattice .
so in this case the model fails , too .
* 3 . * finally , in terms of the singlet fields like ( [ eq : chin ] ) and ( [ eq : psin ] ) introduced at all @xmath155 , the action reads as @xmath156 where @xmath157 we see that these system is equivalent to that considered in the previous section , only if @xmath158 .
however one can hardly expect that it is so on the ground of the estimations of @xmath159 in pure scalar models @xcite .
for instance , in @xmath160 and @xmath161 models it has been found @xmath162 .
therefore our previous analysis is not directly applicable to this system .
however we can made some qualitative conclusions about the effects of such renormalization .
an analysis similar to that of the beginning of the previous section , shows that in the continuum case , when the wilson term is equal to zero , @xmath163 simply renormalizes the masses of the massive excitations : now they become @xmath164 , rather than 1 .
the situation on the lattice is more complicated .
indeed , making the small @xmath73 expansion in ( [ eq : w3 ] ) one can see that the role of the factor @xmath165 , which provides the localization of the massless states at the boundaries , now is played by @xmath163 dependent combination @xmath166 , that may considerably shift the range of admissible values of @xmath167 .
this requires a priori knowledge of the @xmath163 , that complicates considerably the numerical study of such models . but
the main problem in this variant is that all the fermions in the system now are neutral .
the question of which of these scenarios is realized actually in the system requires special investigation , and the answer may depend on the strength of the wilson parameter @xmath168 , set to 1 in this paper , as well as on the couplings @xmath32 and @xmath107 .
however none of them leads to the chiral spectrum of charged , i.e. gauge non - singlet , fermions which one tends to describe .
so we can conclude that in this respect the domain wall fermions have no appreciable advantages compared with the four dimensional wilson fermions .
the lattice boundary formulation of chiral fermions with majorana mass introduced for the field @xmath1 on the boundary @xmath24 indeed yields the low - lying chiral spectrum at the tree level : there exists only one massless left - handed fermion @xmath0 localized at the boundary @xmath23 .
our results also show that effects caused by the introduction of such majorana mass is strongly suppressed in physical sector .
for instance , the fermion number violation for the @xmath0 is suppressed exponentially as the size of the fifth dimension @xmath25 increases .
since the majorana mass does not break the gauge invariance for real representations of the gauge group , the most immediate implementation of such systems is the lattice formulation of the @xmath2 supersymetric models @xcite .
our results speaks in favour of this idea , since the exponential suppression of the undesirable effects in the physical sector is an indication that the problem of fine tuning can be avoided in such a formulation , at least in the perturbation theory .
this however should be demonstrated explicitly , and this work now is in progress .
the prospect for chiral gauge theories is less certain , for they have to employ the generalized higgs mechanism in the pms phase .
although we have demonstrated that the pms phase can exist in these models , their possible dynamics within such a phase does not give a ground for optimism .
indeed , we argued that the spectrum of charge fermions within the pms phase is vectorlike , and that only spectrum of the neutral states , by a certain tuning of the domain wall mass , can be made chiral .
thus the crucial question is what is the gauge interactions of such neutral chiral states .
previous studies of similar states appearing in the models with the wilson - yukawa couplings give a strong evidence that such states in the continuum limit become non - interacting @xcite .
it is this point that leads to very plausible failure of those models , as well as the models with multifermion couplings @xcite .
such screening of the chiral charges appears to be an universal phenomenon pursuing any models within the pms phase .
therefore we consider this point as the main problem on the way of implementation of the domain wall fermions to exactly gauge invariant nonperturbative formulation of chiral gauge theories .
+ * acknowledgements * + s.v.z . is grateful to the staff of the center for computational physics , where this work has been done , and particularly to y. iwasaki and a. ukawa , for their kind hospitality .
s.a . was supported in part by the grants - in - aid of the ministry of education ( no.08640350 ) ; s.v.z .
was supported by `` coe foreign researcher program '' of the ministry of education of japan and by the russian basic research fund under the grant 95 - 02 - 03868a .
99 h. b. nielsen and m. ninomiya , nucl .
b 185 ( 1981 ) 20 ; b 195 ( 1981 ) 541 ( e ) ; b 193 ( 1981 ) 173 ; phys . lett . 105b ( 1981 ) 219 . k. g. wilson , in _ new phenomena in subnuclear physics _ , edited by a. zichichi ( plenum press , new york , 1996 ) , p. 69 .
d. b. kaplan , phys .
b 288 ( 1992 ) 342 ; nucl .
b ( proc . suppl . ) 30 ( 1993 ) 597 .
y. shamir , nucl . phys .
b 406 ( 1993 ) 90 .
k. jansen , phys .
( 1996 ) 1 .
r. narayanan and h. neuberger , phys .
b 302 ( 1993 ) 62 ; + s. a. frolov and a. a. slavnov , nucl . phys .
b 411 ( 1994 ) 647 .
r. narayanan and h. neuberger , nucl . phys .
b 412 ( 1994 ) 574 ; b 443 ( 1995 ) 305 .
d. n. petcher , nucl .
b ( proc . suppl . )
30 ( 1993 ) 50 .
y. shamir , nucl .
b ( proc . suppl . ) 47 ( 1996 ) 212 .
j. shigemitsu , nucl .
b ( proc . suppl . ) 20 ( 1991 ) 515 .
m. f. l. golterman , d. n. petcher , and j. smit , nucl .
phys b 370 ( 1992 ) 51 .
w. bock , a. k. de , and j. smit , nucl .
b388 ( 1992 ) 243 . m. f. l. golterman , d. n. petcher , and e. rivas , nucl .
b 377 ( 1992 ) 405 .
s. aoki , h. hirose , and y. kikukawa , int .
j. modern phys .
a9 ( 1994 ) 4129 .
m. creutz , m. tytgat , c. rebbi , and s - s .
xue , hep - lat/9612017 .
e. eichten and preskill , nucl .
b 268 ( 1986 ) 179 .
j. nishimura , hep - lat/9701013 .
s. aoki and h. hirose , phys .
d 49 ( 1994 ) 2604 .
s. tominaga and s. v. zenkin , phys .
d 50 ( 1994 ) 3387 ; + s. v. zenkin , phys .
d 53 ( 1996 ) 6674 .
j. smit , nucl .
b ( proc . suppl . ) 9 ( 1989 ) 579 ; 17 ( 1990 ) 3 .
lee , nucl .
b ( proc . suppl . ) 17 ( 1990 ) 457 ; + s. aoki , i - h .
lee , and r.e .
shrock , nucl .
b355 ( 1991 ) 383 . m. f. l. golterman , d. n. petcher , and e. rivas , nucl . phys .
b 395 ( 1993 ) 596 . | we examine the lattice boundary formulation of chiral fermions with either an explicit majorana mass or a higgs - majorana coupling introduced on one of the boundaries .
we demonstrate that the low - lying spectrum of the models with an explicit majorana mass of the order of an inverse lattice spacing is chiral at tree level . within a mean - field approximation
we show that the systems with a strong higgs - majorana coupling have a symmetric phase , in which a majorana mass of the order of an inverse lattice spacing is generated without spontaneous breaking of the gauge symmetry .
we argue , however , that the models within such a phase have a chiral spectrum only in terms of the fermions that are singlets under the gauge group .
the application of such systems to nonperturbative formulations of supersymmetric and chiral gauge theories is briefly discussed . -0.6
cm uthep-360 + ut - ccp - p21 + april 1997 + 1.0truecm s. aoki , institute of physics , university of tsukuba tsukuba , ibaraki 305 , japan k. nagai , and s. v. zenkin center for computational physics , university of tsukuba tsukuba , ibaraki 305 , japan 1.0truecm 1 |
the phenomenon of color superconductivity @xcite is of general interest , in particular , in studies of the qcd phase structure @xcite and applications in the astrophysics of compact stars @xcite .
observable consequences are expected for , e. g. , the cooling behavior @xcite .
different aspects have been investigated so far , whereby models of the njl type have been widely employed @xcite in studies of the phase structure in the vicinity of the hadronization transition .
recently , it has been shown in these investigations that for low temperatures ( @xmath2 ) and not too large chemical potentials ( @xmath3 ) the two - flavor color superconductivity ( 2sc ) phase is favored over alternative color superconducting phases @xcite . according to @xcite , the color - flavor - locked ( cfl ) phase occurs only at @xmath4 mev .
it is generally agreed that at low temperatures the transition of the matter from the phase with broken chiral symmetry to the color superconducting phase is of the first order ( see e. g. @xcite ) . from the point of view of phenomenological applications , as e.g. in compact star physics , the order of the phase transition to quark superconducting matter plays an important role .
the conclusion about the first order phase transition was drawn within models without vector interaction channels taken into account ; the vector interaction has been considered in few papers @xcite .
it was found that the presence of quark interaction in the vector channel moves the critical line in the in @xmath5 plain to larger @xmath3 @xcite .
recently it has been demonstrated @xcite that the critical line of first order phase transition in the @xmath5 plane can have a second end - point at low temperatures , besides the well known one at high temperatures .
the latter one could even be subject to experimental verification in heavy - ion collisions @xcite whereas the former could be of relevance for neutron stars .
while in ref .
@xcite this feature of the phase diagram was a consequence of the presence of interaction in the vector channel , we would like to investigate in the present work the sensitivity of the phase diagram to the choice of model parameters without interaction in the vector channel .
we will demonstrate that in the absence of the vector channel interaction the phase transition is not necessarily of the first order , thus revising statements in refs .
@xcite .
it is worth noting that some progress has recently been done in lattice calculations .
there are methods being developed that allow to extend lattice results to the case of finite chemical potentials @xcite
. however , these methods are valid only for small chemical potentials ( see e. g. @xcite ) , below the conditions at which the color superconductivity phase is expected to form .
the structure of our paper is as follows . in sect . 2 ,
a chiral quark model is introduced , its lagrangian is given and the model parameters are fixed from the vacuum state in two different schemes .
temperature and chemical potential are introduced into the quark model in sect .
3 , using the matsubara formalism .
the conclusions and a discussion of the obtained results are given in sect .
in order to study the quark matter phase diagram including color superconductivity , one should generalize the concept of the single order parameter related to the quark - antiquark condensate in the case of chiral symmetry breaking to a set of order parameters when condensation can occur in other interaction channels too .
the simplest extension is the scalar diquark condensate @xmath6 for @xmath7 and @xmath8 quarks @xmath9 which is an order parameter characterizing the domain where the color symmetry is spontaneously broken and the quark matter finds itself in the ( two - flavor ) color superconducting ( 2sc ) state .
this quantity is the most important one among other possible condensates that can be constructed in accordance with the pauli principle @xcite . in ( [ delta ] )
the matrix @xmath10 is the charge conjugation matrix operator for fermions @xmath11 the matrices @xmath12 and @xmath13 are pauli and gell - mann matrices , respectively . the first one acts on the flavor indices of spinors while the second one acts in the color space . if the electroweak interaction is discarded and only the strong coupling is in focus , the resulting quark matter phase diagram is essentially determined by nonperturbative features of the qcd vacuum state .
one therefore has to resort to nonperturbative approaches to describe the behavior of particles at various conditions , ranging from cold and dilute matter up to the hot and dense one .
a reliable and widely tested model to nonperturbative strong coupling qcd is provided by the dyson - schwinger equations @xcite , however , for qualitative studies like the one we attempt here it proves to be too complex .
therefore , we will use here a simple and tractable nonperturbative model of quark interaction , the nambu jona - lasinio ( njl ) model @xcite , which has been extensively exploited for the description of the properties of the light meson sector of qcd ( also to describe the color superconductivity phase @xcite ) and proved to be a model respecting the low - energy theorems . before we proceed to the case of finite temperature and density , the model parameters that determine
the quark interaction should be fixed .
this shall be done for the vacuum state where hadronic properties are known .
we will assume , according to common wisdom that , once fixed , these parameters ( originating from the nonperturbative gluon sector of qcd ) will not change , even in the vicinity of the transition to the quark matter .
this transition is thus caused by medium effects in the quark sector only . in the present paper
we restrict ourselves to the two - flavor case , leaving the strange quark and effects related to it beyond our consideration .
as we constrain ourselves to only two order parameters , the quark and scalar diquark condensates , during our investigation , the interaction of quarks will be represented in the lagrangian by @xmath14 symmetric scalar , pseudoscalar quark - antiquark , and scalar diquark vertices : @xmath15,\\ { { \mathcal l } } _ { qq}&=&\frac{h}{2}(\bar\psi i\gamma_5\tau_2\lambda_2c\bar\psi^t)(\psi^{t } c i\gamma_5\tau_2\lambda_2\psi),\end{aligned}\ ] ] where @xmath16 is the diagonal current quark mass matrix @xmath17 , @xmath18 and @xmath19 are constants describing the interaction of quarks in the scalar , pseudoscalar , and scalar diquark channels , respectively .
we work in the isospin symmetric case @xmath20 , thus @xmath21 . by the standard hubbard - stratonovich procedure ,
we introduce auxiliary scalar ( @xmath22 ) , pseudoscalar triplet ( @xmath23 ) , and diquark ( @xmath24 ) fields together with yukawa - like terms in the lagrangian density instead of the four - quark vertices : @xmath25 in order to integrate out the quark degrees of freedom by gaussian path integration , it is appropriate to represent the quark fields by the bispinor @xmath26 and to introduce the matrix propagator @xmath27 : @xmath28 integrating over @xmath29 and @xmath30 , we then obtain an effective lagrangian in terms of collective scalar and pseudoscalar quark - antiquark and scalar diquark excitations . here , we restrict ourselves to the mean - field approximation , leaving the next - to - leading order corrections in the @xmath31 expansion beyond our model .
finally , the effective lagrangian density reads @xmath32 the trace in ( [ invpropagator ] ) is taken in the dirac , color , and flavor space .
the matrix @xmath33 contains @xmath22 and @xmath23 fields : @xmath34 the sum over @xmath35 is assumed , and @xmath36 .
as it was mentioned above , we are working in the mean - field approximation and the quark condensates are of interest .
therefore , further study can be performed in terms of the effective potential @xmath37 where @xmath38 is 4-dimensional volume .
the vacuum expectation values of the collective variables @xmath22 , @xmath23 , @xmath39 , and @xmath40 determine the absolute minimum of @xmath41 .
they are given by the equation @xmath42 _ a priori _ it is known that in the vacuum only the @xmath22 field acquires a nonvanishing expectation value .
the diquark fields @xmath39 , @xmath40 are expected to have nonzero mean values only in dense matter .
the mean value of the pseudoscalar isotriplet field @xmath23 is always equal to zero , therefore we omit it hereafter .
having solved eq .
( [ vmin ] ) for the field @xmath22 , one can work in terms of the constituent quark mass @xmath43 , connected with the current quark mass by the gap equation @xmath44 in the chiral limit ( @xmath45 ) , the constituent quark mass is proportional to the quark condensate and thus can be treated as the order parameter . in the njl model
the quark condensate is @xmath46 where @xmath47 the divergence in @xmath48 is eliminated by means of a sharp 3d cut - off at the scale @xmath49 .
in our model we have four parameters : the four - quark interaction constants @xmath18 and @xmath19 , cut - off @xmath49 , and the current quark mass @xmath50 . without diquarks , there are only three : @xmath18 , @xmath49 , and @xmath50 .
they are fixed by the following relations : 1 . the goldberger - treiman relation ( gtr )
: @xmath51 where @xmath52 mev is the pion weak coupling constant and @xmath53 describes the coupling of a pion with quarks @xmath54 @xmath55 2 .
the quark condensate ( qc ) from qcd sum rules @xmath56 3 .
the decay constant @xmath57 for the @xmath58 ( r2pd ) process @xmath59 the @xmath60 transitions are omitted here .
the current quark mass @xmath50 is fixed from the gmor relation : @xmath61 in the chiral limit @xmath62 , @xmath45 .
5 . with the diquark channel
included , there is an additional parameter @xmath19 which can be fixed as @xmath63 from the fierz transformation ( as e. g. in@xcite ) .. it turned out that within our model the resulting phase diagram is not much affected if one makes the choice in favor of @xmath64 .
however , it would be preferable to fix the constant @xmath19 from some observable , e. g. from the nucleon mass . ] in the item 2 , we have given two alternatives : one can either use the value of the quark condensate taken from qcd sum rule estimates or demand from the model that it should describe the @xmath58 decay .
the latter is well observable in experiment contrary to the quark condensate . for simplicity , we perform all calculations in the chiral limit @xmath45 . in this case , when investigating the hot and dense quark matter , the borders between phases turn out to be sharp and the critical temperature and chemical potential are well defined . with the finite current quark mass ,
the transitions from one phase to the other become smooth . as a result ,
one obtains two different parameter sets shown in table [ paramset ] . in the type
i parameter set the interaction of quarks is stronger , the uv cut - off is smaller , and the constituent quark mass is greater .
one can calculate the dimensionless constant @xmath65 .
it equals 4.6 for the type i and 3.72 for the type ii , respectively . as we will see further , these two parametrizations result in qualitatively different phase diagrams
we extend the njl model to the case of finite temperatures @xmath2 and chemical potentials @xmath3 , applying the matsubara formalism , and restrict ourselves to the isospin symmetric case where up and down quark chemical potentials coincide .
the thermodynamical potential per volume is @xmath66 where @xmath67 are matsubara frequencies for fermions , and the chemical potential is included into the definition of inverse quark propagator @xmath68 the expression in ( [ thpot ] ) can be simplified using the equations @xmath69\nonumber\\ & + & 2\left[\ln\left(\frac{(\omega_n^2+{\epsilon^+}^2)(\omega_n^2 + { \epsilon^-}^2)}{t^4}\right)\right]\end{aligned}\ ] ] and @xmath70.\ ] ] @xmath71\right.\right.\nonumber\\ & + & \left.2 t \ln\left[1+\exp\left(-\frac{\epsilon^-}{t}\right)\right]\right)\nonumber\\ & + & 4\left(\left(e^++e^-\right)\theta(\lambda^2-\vec{p}^2)+2 t \ln\left[1+\exp\left(-\frac{e^+}{t}\right)\right]\right.\nonumber\\ & + & \left.\left.2 t \ln\left[1+\exp\left(-\frac{e^-}{t}\right)\right]\right)\right\ } + \frac{m^2}{2g}+\frac{|\delta|^2}{2 h}~,\end{aligned}\ ] ] where @xmath72 @xmath73 the cold matter limit @xmath74 looks as follows : @xmath75 \nonumber\\ & \times&\theta(\lambda^2-\vec{p}^2)+\frac{m^2}{2g}+\frac{|\delta|^2}{2 h}~.\end{aligned}\ ] ] the thermodynamical potential can not be calculated in closed form for arbitrary @xmath2 and @xmath3 .
however , in the cold matter limit one can easily obtain analytic expressions for the thermodynamical potential or its derivatives if only one of the collective variables @xmath22 or @xmath76 has a nonvanishing average value .
this allows to find what kind of phase transition is to be expected for different parameter choices .
we evaluate the remaining 3d momentum integrals numerically and calculate the value of thermodynamical potential at different @xmath2 and @xmath3 for the two types of model parameter sets .
the equilibrium state for each @xmath2 and @xmath3 is determined by @xmath77 and @xmath78 corresponding to the minimum of @xmath79 .
it is quite illustrative to look at the contour plots of the thermodynamical potential . for several values of @xmath3 at @xmath74
they are shown in figs .
[ thdp_i_1][thdp_i_4 ] where one can follow the appearance and disappearance of local minima , maxima , and saddle points of the thermodynamical potential with increasing chemical potential . for zero temperature and chemical potential
we have , as expected , a nonzero constituent quark mass ( quark condensate ) corresponding to the absolute minimum of the thermodynamical potential at @xmath80 mev and @xmath81 in fig .
[ thdp_i_1 ] . at a certain chemical potential , a new local minimum related to the diquark condensate near @xmath82110 mev and @xmath83 ( fig .
[ thdp_i_2 ] ) , but it does not yet give the absolute minimum .
there is also a local maximum around @xmath84 mev and @xmath81 .
as the matter becomes more dense , the second minimum lowers until it becomes degenerate with the first minimum while the average value of @xmath22 ( or @xmath85 ) remains almost unchanged ( see fig .
[ thdp_i_3 ] ) . above the
corresponding ( critical ) chemical potential @xmath86 mev , the second minimum becomes the absolute one and a first order transition occurs , during which @xmath87 discontinuously changes to zero while the diquark condensate acquires nonzero value breaking the color symmetry of the strong interaction .
this characterizes the color superconducting phase transition in quark matter .
furthermore , the local minimum on the @xmath43 axis merges the saddle point ( see fig .
[ thdp_i_4 ] ) and , at still higher @xmath3 , only the local minimum on the @xmath39 axis near @xmath88 130 mev and @xmath83 remains . at a fixed chemical potential above @xmath89 , with the temperature rising ,
the average value of @xmath76 decreases until it reaches zero at the critical temperature @xmath90 which can be roughly estimated using the bcs theory formula @xmath91 above this temperature quark matter is in the symmetric phase where the chiral and color symmetries are restored .
finally , we obtain the phase diagram shown on fig .
[ phdi ] with three phases : the hadron phase , 2sc phase , and symmetric phase .
all three phases coexist at the triple point : @xmath92 mev and @xmath93 mev . as for the type
i parameter set , at zero @xmath2 and @xmath3 only the constituent quark mass @xmath43 , being the order parameter for the chiral condensate , is nonzero , whereas the diquark gap @xmath39 vanishes .
however , the vacuum value of @xmath43 is lower than that for the type i and , with the chemical potential increasing , @xmath3 becomes equal to the vacuum value of @xmath43 before the second local minimum , corresponding to the diquark condensate , appears . at further increase of @xmath3 the constituent quark mass decreases , and it would vanish at @xmath94 , @xmath95 if the diquark condensate did not appear .
actually , at the critical value @xmath96 both the quark condensate and the diquark condensate are small but nonzero .
the changes of the local extrema for increasing chemical potential are similar to those shown in figs .
[ thdp_i_1][thdp_i_4 ] .
the cases of dilute ( @xmath97 ) and very dense matter ( @xmath98 mev ) are qualitatively analogous , only the absolute values of @xmath43 and @xmath76 at which the local minima are found are different . at intermediate densities , however , there is a qualitative difference . within a very narrow range of values of the chemical potential
, there exists a new phase of massive superconducting matter .
one can see this in fig .
[ thdp_ii ] for @xmath99 mev . at higher @xmath3 the chiral symmetry
is restored and the quark matter is in the pure superconducting phase .
a possibility of the chiral diquark condensates to coexist at certain condition has been already noticed in ref .
@xcite thus , for the type ii parameter set , the transition from the hadronic to the superconducting phase is of the second order . in this case
there are no degenerate local minima in the thermodynamical potential separated by a barrier .
this behavior is unlike to what is commonly expected for a cold and dense matter but it parallels the findings of ref .
@xcite where vector interactions are responsible for this behavior .
the average value of @xmath76 is much smaller than for the type i parameter set . as a consequence ,
the border between 2sc and the symmetric phases of quark matter lies at noticeably lower temperatures .
the phase diagram obtained in our model for the type ii parameter set is shown in fig .
[ phdii ] .
in the framework of the simple njl model for two flavors , a phase diagram is obtained for @xmath74200 mev and @xmath97450 mev .
three phases are found for the type i parameter set and four phases for the type ii parameter set .
the critical temperature and chemical potential obtained in the type i scheme differ from those obtained with the type ii parameter set . at @xmath74 , @xmath100 mev for the type i parameter set and @xmath101 mev for the type ii .
the corresponding quark densities differ by a factor 1.5 1.7 .
the critical temperature for the type ii parameter set is as low as @xmath102 mev and thus much closer to critical temperatures for the paring instability in nuclear matter systems ( see @xcite ) whereas for the type i parameter set the critical temperatures are an order of magnitude larger .
this striking difference in the critical parameters obtained within the same model calls for a more detailed investigation of the question of model parametrization . in our work
, the constant @xmath19 was not obtained from a fit to observable data . instead
, fierz transformation arguments have been used to fix the ratio @xmath103 .
a parameterization would be favourable where ( in the spirit of the type ii model ) experimentally measured quantities , like the @xmath104 meson width , are used rather than non - observable ones ( quark condensate etc . ) .
it would therefore be more consistent to fit the constant @xmath19 from baryon properties , see @xcite and also to go beyond the mean field level of description .
these investigations shall be performed in future work where it remains to be clarified which critical parameters for the color superconducting phase transition can be considered more realistic and of which order the phase transition is .
the authors are grateful to m. buballa , t. kunihiro , and m. kitazawa for useful discussions .
this work has been supported by daad and heisenberg - landau programs .
v.y . and m.k.v .
acknowledge support by rfbr grant no .
02 - 02 - 16194 .
99 b. barrois , nucl .
b * 129 * , 390 ( 1977 ) .
d. bailin and a. love , phys
. rep . * 107 * , 325 ( 1984 ) .
m. iwasaki and t. iwado , phys .
b * 350 * , 163 ( 1995 ) . m. iwasaki and t. iwado , prog .
phys . * 94 * , 1073 ( 1995 ) .
m. alford , k. rajagopal , and f. wilczek , phys .
b * 422 * , 247 ( 1998 ) .
r. rapp , t. schfer , e. v. shuryak , and m. velkovsky , phys .
lett . * 81 * , 53 ( 1998 ) .
m. a. halasz , a. d. jackson , r. e. shrock , m. a. stephanov , and j. j. verbaarschot , phys .
d * 58 * , 096007 ( 1998 ) ; m. a. stephanov , k. rajagopal , and e. v. shuryak , phys .
lett . *
81 * , 1998 ( 4816 ) .
j. c. r. bloch , c. d. roberts , and s. m. schmidt , phys .
rev . c * 60 * , 065208 ( 1999 ) .
m. alford , k. rajagopal , and f. wilczek , nucl .
b * 573 * , 443 ( 1999 ) .
r. d. pisarski and d. h. rischke , phys . rev . lett .
* 83 * , 37 ( 1999 ) .
d. t. son , phys . rev .
d * 59 * , 094019 ( 1999 ) .
t. schfer and f. wilczek , phys .
d * 60 * , 114033 ( 1999 ) .
j. berges and k. rajagopal , nucl .
b * 538 * , 215 ( 1999 ) .
r. rapp , t. schfer , e. v. shuryak , and m. velkovsky , ann .
phys . * 280 * , 35 ( 2000 ) .
r. d. pisarski and d. h. rischke , phys . rev .
d * 61 * , 051501 ( 2000 ) .
k. rajagopal and f. wilczek , arxiv : hep - ph/0011333 and references therein , in : shifman , m. ( ed . ) : _ at the frontier of particle physics _ , vol .
3 , world scientific , singapore ( 2002 ) , pp .
. m. alford , ann .
nucl . part .
sci . * 51 * , 131 ( 2001 ) .
m. alford , k. rajagopal , s. reddy , and f. wilczek , phys .
d * 64 * , 074017 ( 2001 ) .
m. buballa and m. oertel , nucl .
phys . a * 703 * , 770 ( 2002 ) .
m. oertel and m. buballa , in : _ ultrarelativistic heavy - ion collisions _ , proc .
workshop xxx on gross properties of nuclei and nuclear excitations , hirschegg / austria , m. buballa , w. nrenberg , b .-
schfer , and j. wambach ( eds . ) , darmstadt ( 2002 ) .
f. gastineau , r. nebauer , and j. aichelin , phys .
c * 65 * , 045204 ( 2002 ) .
m. harada and s. takagi , progr .
phys . * 107 * , 561 ( 2002 ) .
a. steiner , s. reddy , and m. prakash , phys .
d * 66 * , 094007 ( 2002 ) .
f. neumann , m. buballa , and m. oertel , nucl .
phys . a * 714 * , 481 ( 2003 ) .
j. bowers and k. rajagopal , phys .
d * 66 * , 065002 ( 2002 ) .
d. blaschke , n. k. glendenning , a. sedrakian ( eds . ) , _ physics of neutron star interiors _ , ( springer , lnp 578 , 2001 ) .
d. blaschke , t. klhn , and d. n. voskresensky , astrophys .
j. * 533 * , 406 ( 2000 ) .
d. blaschke , h. grigorian , and d. n. voskresensky , astron . & astrophys .
* 368 * , 561 ( 2001 ) .
d. ebert and h. reinhardt , nucl .
phys . b * 271 * , 188 ( 1986 ) .
m. k. volkov , sov .
j. part and nucl .
* 17 * , 186 ( 1986 ) .
d. ebert , yu .
l. kalinovsky , l. mnchow , and m. k. volkov , int .
j. mod .
a * 8 * , 1295 ( 1993 ) .
d. b. blaschke , g. r. g. burau , m. k. volkov , and v. l. yudichev , eur .
j. a * 11 * , 319 ( 2001 ) . k. langfeld and m. rho , nucl .
phys . a * 660 * , 475 ( 1999 ) . m. buballa , j. hoek , and m. oertel , phys .
d * 65 * , 014018 ( 2001 ) s. klimt , m. lutz , and w. weise , phys .
lett . b * 249 * , 249 ( 1990 ) m. asakawa and k. yazaki , nucl .
phys . a * 504 * , 668 ( 1989 ) . m. kitazawa , t. koide , t. kunihiro , and y. nemoto , prog . theor . phys .
* 108 * , 929 ( 2002 ) .
t. kunihiro , nucl .
b * 351 * , 593 ( 1991 ) . c. r. allton , s. ejiri , s. j. hands , o. kaczmarek , f. karsch , e. laermann , c. schmidt , l. scorzato , phys .
d * 66 * , 074507 ( 2002 ) .
s. d. katz and z. fodor , jhep * 0203 * , 014 ( 2002 ) .
f. csikor , g. i. egri , z. fodor , s. d. katz , k. k. szabo , a. i. toth , hep - lat/0301027 . c. d. roberts and s. m. schmidt , prog .
* 45 * ( 2000 ) s1 [ arxiv : nucl - th/0005064 ] .
s. p. klevansky , rev .
phys . * 64 * , 649 ( 1992 ) .
t. hatsuda and t. kunihiro , phys . rep . *
247 * , 221 ( 1994 ) .
d. ebert , ch .
v. zhukovsky , a. s. vshivtsev , int . j. mod
a * 13 * , 1723 ( 1998 ) phys .
c * 61 * , 055209 ( 2000 ) .
t. m. schwarz , s. p. klevansky , and g. papp , phys .
c * 60 * , 055205 ( 1999 ) .
m. kitazawa , t. koide , t. kunihiro , and y. nemoto , phys .
d * 65 * , 091504 ( 2002 ) .
m. sadzhikowski , mod .
lett . a * 16 * , 1129 ( 2001 ) .
a. sedrakian and u. lombardo , phys .
* 84 * , 602 ( 2000 ) and references therein .
s. pepin , m. c. birse , j. a. mcgovern , and n. r. walet , phys .
c * 61 * , 055209 ( 2000 ) .
w. bentz , t. horikawa , n. ishii , and a. w. thomas , arxiv : nucl - th/0210067 .
.the model parameters for two different schemes of parameter fixing .
the first row corresponds to the case where the quark condensate value is used , while in the second row the parameters correspond to the case where the decay width of the process @xmath105 is used . [ cols="<,^,^,^,^,^,^ " , ]
and @xmath39 at zero temperature and the chemical potential @xmath97 mev . ] | the phase diagram for quark matter is investigated within a simple nambu jona - lasinio model without vector correlations .
it is found that the phase structure in the temperature density plane depends sensitively on the parametrization of the model .
we present two schemes of parametrization of the model where within the first one a first order phase transition from a phase with broken chiral symmetry to a color superconducting phase for temperatures below the triple point at @xmath0 mev occurs whereas for the second one a second order phase transition for temperatures below @xmath1 mev is found . in the latter case , there is also a coexistence phase of broken chiral symmetry with color superconductivity , which is a new finding within this class of models .
possible consequences for the phenomenology of the qcd phase transition at high baryon densities are discussed .
rostock university + preprint no .
mpg - vt - ur 235/02 |
about ten years ago , a peculiar dynamical phenomenon was discovered in populations of identical phase oscillators : under nonlocal symmetric coupling , the coexistence of coherent ( synchronized ) and incoherent oscillators was observed @xcite .
this highly counterintuitive phenomenon was given the name chimera state after the greek mythological creature made up of different animals @xcite . since
then the study of chimera states has been the focus of extensive research in a wide number of models , from kuramoto phase oscillators @xcite to periodic and chaotic maps @xcite , as well as stuart - landau oscillators @xcite .
the first experimental evidence of chimera states was found in populations of coupled chemical oscillators as well as in optical coupled - map lattices realized by liquid - crystal light modulators @xcite .
recently , moreover , martens and coauthors showed that chimeras emerge naturally from a competition between two antagonistic synchronization patterns in a mechanical experiment involving two subpopulations of identical metronomes coupled in a hierarchical network @xcite . in the context of neuroscience
, a similar effort has been undertaken by several groups , since it is believed that chimera states might explain the phenomenon of unihemispheric sleep observed in many birds and dolphins which sleep with one eye open , meaning that one hemisphere of the brain is synchronouns whereas the other is asynchronous @xcite .
the purpose of this paper is to make a contribution in this direction , by identifying for the first time a variety of single and multi - chimera states in networks of non - locally coupled neurons represented by hindmarsh rose oscillators .
recently , multi - chimera states were discovered on a ring of nonlocally coupled fitzhugh - nagumo ( fhn ) oscillators @xcite .
the fhn model is a 2dimensional ( 2d ) simplification of the physiologically realistic hodgkin - huxley model @xcite and is therefore computationally a lot simpler to handle .
however , it fails to reproduce several important dynamical behaviors shown by real neurons , like rapid firing or regular and chaotic bursting .
this can be overcome by replacing the fhn with another well - known more realistic model for single neuron dynamics , the hindmarsh - rose ( hr ) model @xcite , which we will be used throughout this work both in its 2d and 3d versions . in section 2 of this paper
, we first treat the case of 2d - hr oscillators represented by two first order ordinary differential equations ( odes ) describing the interaction of a membrane potential and a single variable related to ionic currents across the membrane under periodic boundary conditions .
we review the dynamics in the 2d plane in terms of its fixed points and limit cycles , coupling each of the @xmath0 oscillators to @xmath1 nearest neighbors symmetrically on both sides , in the manner adopted in @xcite , through which chimeras were discovered in fhn oscillators .
we identify parameter values for which chimeras appear in this setting and note the variety of oscillating patterns that are possible due to the bistability features of the 2d model .
in particular , we identify a new `` mixed oscillatory state '' ( mos ) , in which the desynchronized neurons are uniformly distributed among those attracted by a stable stationary state .
furthermore , we also discover chimera like patterns in the more natural setting where only the membrane potential variables are coupled with @xmath2 of the same type .
next , we turn in section 3 to the more realistic 3d - hr model where a third variable is added representing a slowly varying current , through which the system can also exhibit bursting modes . here , we choose a different type of coupling where the two first variables are coupled symmetrically to @xmath2 of their own kind and observe only states of complete synchronization as well as mos in which desynchronized oscillators are interspersed among neurons that oscillate in synchrony . however , when coupling is allowed only among the first ( membrane potential ) variables chimera states are discovered in cases where spiking occurs within sufficiently long bursting intervals .
finally , the paper ends with our conclusions in section 4 .
following the particular type of setting proposed in @xcite we consider @xmath0 nonlocally coupled hindmarsh - rose oscillators , where the interconnections between neurons exist with @xmath3 nearest neighbors only on either side as follows : @xmath4 \label{eq:01 } \\
\dot y_k & = & 1 - 5x_k^2-y_k+ \frac{\sigma_y}{2r}\sum_{j = k - r}^{j = k+r } [ b_{yx}(x_j - x_k)+b_{yy}(y_j - y_k ) ] .
\label{eq:02 } \end{aligned}\ ] ] in the above equations @xmath5 is the membrane potential of the @xmath6-th neuron , @xmath7 represents various physical quantities related to electrical conductances of the relevant ion currents across the membrane , @xmath8 , @xmath9 and @xmath10 are constant parameters , and @xmath11 is the external stimulus current . each oscillator is coupled with its @xmath12 nearest neighbors on both sides with coupling strengths @xmath13 .
this induces nonlocality in the form of a ring topology established by considering periodic boundary conditions for our systems of odes . as in @xcite ,
our system contains not only direct @xmath14 and @xmath15 coupling , but also cross - coupling between variables @xmath16 and @xmath17 .
this feature is modeled by a rotational coupling matrix : @xmath18 depending on a coupling phase @xmath19 . in
what follows , we study the collective behavior of the above system and investigate , in particular , the existence of chimera states in relation to the values of all network parameters : @xmath0 , @xmath3 , @xmath19 , @xmath20 and @xmath21 .
more specifically , we consider two cases : ( @xmath22 ) direct and cross - coupling of equal strength in both variables ( @xmath23 ) and ( @xmath24 ) direct coupling in the @xmath16 variable only ( @xmath25 , @xmath26 ) . similarly to @xcite we shall use initial conditions randomly distributed on the unit circle ( @xmath27 ) . at @xmath28 of eqs .
( [ eq:01]),([eq:02 ] ) ( left ) and selected time series ( right ) for @xmath29 .
( a ) @xmath30 , ( b ) @xmath31 and ( c ) @xmath32 . @xmath33 and @xmath34.,scaledwidth=70.0% ] typical spatial patterns for case ( @xmath22 ) are shown on the left panel of fig .
[ fig : fig1 ] , where the @xmath16 variable is plotted over the index number @xmath6 at a time snapshot chosen after a sufficiently long simulation of the system . in this figure
the effect of different values of the phase @xmath19 is demonstrated while the number of oscillators @xmath0 and their nearest neighbors @xmath3 , as well as the coupling strengths @xmath23 are kept constant . for example , for @xmath30 ( fig .
[ fig : fig1](a ) ) an interesting novel type of dynamics is observed that we shall call `` mixed oscillatory state '' ( mos ) , whereby nearly half of the @xmath5 are attracted to a fixed point ( at this snapshot they are all at a value near @xmath35 ) , while the other half are oscillating interspersed among the stationary ones . from the respective time series ( fig . [
fig : fig1](a ) , right ) it is clear that the former correspond to spiking neurons whereas the latter to quiescent ones .
this interesting mos phenomenon is due to the fact that , in the standard parameter values we have chosen , the uncoupled hr oscillators are characterized by the property of _
bistability_. clearly , as shown in the phase portrait of fig .
[ fig : fig2 ] , each oscillator possesses three fixed points : the leftmost fixed point is a stable node corresponding to the resting state of the neuron while the other two correspond to a saddle point and an unstable node and are therefore repelling . for @xmath11 ( which is the case here )
a stable limit cycle also exists which attracts many of the neurons into oscillatory motion , rendering the system bistable and producing the dynamics observed in fig .
[ fig : fig1](a ) .
now , when a positive current @xmath36 is applied , the @xmath16-nullcline is lowered such that the saddle point and the stable node collide and finally vanish . in this case
the full system enters a stable limit cycle associated with typical spiking behaviour .
similar complex patterns including mos and chimeras have been observed in this regime as well . .
three fixed points coexist with a stable limit cycle.,scaledwidth=40.0% ] for @xmath31 , on the other hand , there is a `` shift '' in the dynamics of the individual neurons into the spiking regime , as seen in fig . [
fig : fig1](b ) ( right ) .
the corresponding spatial pattern has a wave - like form of period 2 .
then , for @xmath32 a classical chimera state with two incoherent domains is observed ( ( fig .
[ fig : fig1](c ) , left ) . diagonal coupling ( @xmath37 , @xmath38 ) is , therefore , identified as the necessary condition to achieve chimera states . by contrast , it is interesting to note that in nonlocally coupled fitz - hugh nagumo oscillators @xcite it has been shown both analytically and numerically that chimera states occur for _ off - diagonal _ coupling . by decreasing the range of coupling @xmath3 and increasing the system size @xmath0 , chimera states occur with multiple domains of incoherence and coherence for @xmath32 ( fig .
[ fig : fig3](c , d ) ) , and , accordingly , periodic spatial patterns with larger wave numbers arise for @xmath31 , as seen in fig .
[ fig : fig3](a , b ) .
this is in agreement with previous works reported in @xcite .
next we consider the case ( @xmath24 ) where the coupling term is restricted to the @xmath16 variable .
this case is important since incorporating the coupling in the voltage membrane ( @xmath16 ) alone is more realistic from a biophysiological point of view . in fig .
[ fig : fig4 ] spatial plots ( left ) and the corresponding @xmath39-plane ( right ) for increasing coupling strength are shown .
chimera states ( fig .
[ fig : fig4 ] ( b , c ) ) are observed as an intermediate pattern between desynchronization ( fig .
[ fig : fig4 ] ( a ) ) and complete synchronization ( fig .
[ fig : fig4 ] ( d ) ) . of eqs .
( [ eq:01]),([eq:02 ] ) at @xmath28 for @xmath23 .
@xmath40 ( top ) and @xmath41 ( bottom ) whereas @xmath31 ( left ) and @xmath32 ( right ) . in ( a ) and ( b ) @xmath42 , in ( c ) @xmath43 , and in ( d ) @xmath44.,scaledwidth=50.0% ] ( left ) and in the @xmath45-plane ( right ) of eqs . ( [ eq:01]),([eq:02 ] ) at @xmath28 for @xmath25 , @xmath33 and @xmath34 . (
a ) @xmath46 , ( b ) @xmath47 , ( c ) @xmath48 and ( d ) @xmath49.,scaledwidth=40.0% ]
in order to complete our study of the hindmarsh - rose model we shall consider , in this section , its three - dimensional version . the corresponding equations read : @xmath50 the extra variable @xmath51 represents a slowly varying current , which changes the applied current @xmath52 to @xmath53 and guarantees firing frequency adaptation ( governed by the parameter @xmath54 ) as well as the ability to produce typical bursting modes , which the 2d model can not reproduce .
parameter @xmath55 controls the transition between spiking and bursting , parameter @xmath56 determines the spiking frequency ( in the spiking regime ) and the number of spikes per bursting ( in the bursting regime ) , while @xmath57 sets the resting potential of the system .
the parameters of the fast @xmath58 system are set to the same values used in the two - dimensional version ( @xmath8 , @xmath10 ) and typical values are used for the parameters of the @xmath51-equation ( @xmath59 , @xmath60 , @xmath61 ) .
the 3d hindmarsh - rose model exhibits a rich variety of bifurcation scenarios in the @xmath62 parameter plane @xcite .
thus , we prepare all nodes in the spiking regime ( with corresponding parameter values @xmath9 and @xmath63 ) and , as in section 2 , we use initial conditions randomly distributed on the unit sphere ( @xmath64 ) . first let us consider direct coupling in both variables @xmath16 and @xmath17 and vary the value of the equal coupling strengths @xmath23 , while @xmath0 and @xmath3 are kept constant .
naturally , the interaction of @xmath0 spiking neurons will lead to a change in their dynamics , as we discuss in what follows . for low values of @xmath65
we observe the occurrence of a type of mos where nearly half of the neurons spike regularly in a synchronous fashion , while the rest are unsynchronized and spike in an irregular fashion .
this is illustrated in the respective time series in the right panel of fig .
[ fig : fig5](a ) at higher values of the coupling strength the the system is fully synchronized ( fig .
[ fig : fig5](b ) ) . of eqs .
( [ eq:03])-([eq:05 ] ) at @xmath66 ( left ) and selected time series ( right ) for : ( a ) @xmath67 and ( b ) @xmath68 . @xmath33 and @xmath34.,scaledwidth=50.0% ]
next we check the case of coupling in the @xmath16 variable alone ( @xmath69 , @xmath25 ) .
figure [ fig : fig6 ] displays characteristic synchronization patterns obtained when we increase @xmath20 ( left panel ) and selected time series ( right panel ) . at low values of the coupling strength
all neurons remain in the regular spiking regime and desynchronization alternates with complete synchronization as @xmath20 increases ( fig .
[ fig : fig6](a , b , c ) ) . for intermediate values of the coupling strength ,
chimera states with one incoherent domain are to be observed .
these are associated with a change in the dynamics of the individual neurons , which now produce irregular bursts ( fig .
[ fig : fig6](d ) ) .
the number of spikes in each burst increases at higher @xmath20 values and the system is again fully synchronized ( fig .
[ fig : fig6](e ) ) .
extensive simulations show that the chimera states disappear and reappear again by varying @xmath20 which is most likely due to the system s multistability and sensitive dependence on initial conditions . of eqs .
( [ eq:03])-([eq:05 ] ) at @xmath66 ( left ) and selected time series for @xmath25 , @xmath33 and @xmath34 . ( a )
@xmath70 , ( b ) @xmath71 , @xmath72 , ( d ) @xmath73 , and ( e ) @xmath74 . ,
in this paper , we have identified the occurrence of chimera states for various coupling schemes in networks of 2d and 3d hindmarsh - rose models .
this , together with recent reports on multiple chimera states in nonlocally coupled fitzhugh - nagumo oscillators , provide strong evidence that this counterintuitive phenomenon is very relevant as far as neuroscience applications are concerned .
chimera states are strongly related to the phenomenon of synchronization . during the last years
, synchronization phenomena have been intensely studied in the framework of complex systems @xcite . moreover , it is a well - established fact the key ingredient for the occurrence of chimera states is nonlocal coupling .
the human brain is an excellent example of a complex system where nonlocal connectivity is compatible with reality .
therefore , the study of chimera states in systems modelling neuron dynamics is both significant and relevant as far as applications are concerned .
moreover , the present work is also important from a theoretical point of view , since it verifies the existence of chimera states in coupled bistable elements , while , up to now , it was known to arise in oscillator models possessing a single attracting state of the limit cycle type .
finally , we have identified a novel type of mixed oscillatory state ( mos ) , in which desynchronized neurons are interspersed among those that are either stationary or oscillate in synchrony . as a continuation of this work , it is very interesting to see whether chimeras and mos states appear also in networks of _ populations _ of hindmarsh - rose oscillators , which are currently under investigation .
the authors acknowledge support by the european union ( european social fund esf ) and greek national funds through the operational program `` education and lifelong learning '' of the national strategic reference framework ( nsrf ) - research funding program : thales . investing in knowledge society through the european social fund .
funding was also provided by ninds r01 - 40596 .
hagerstrom , a. m. , thomas , e. , roy , r. , hvel , p. , omelchenko , i. & schll , e. [ 2012 ] `` chimeras in coupled - map lattices : experiments on a liquid crystal spatial light modulator system '' , _ nature physics _ * 8 * , 658 .
omelchenko , i. , omelchenko , o. e. , hvel , p. & schll , e. [ 2013 ] `` when nonlocal coupling between oscillators becomes stronger : patched synchrony or multichimera states '' , _ phys .
lett . _ * 110 * , 224101 .
rattenborg , n. c. , amlaner , c. j. & lim , s. l. [ 2000 ] `` behavioral , neurophysiological and evolutionary perspectives on unihemispheric sleep , '' _ neuroscience and biobehavioral reviews _ * 24 * , pp . 817-842 . | we have identified the occurrence of chimera states for various coupling schemes in networks of two - dimensional and three - dimensional hindmarsh - rose oscillators , which represent realistic models of neuronal ensembles .
this result , together with recent studies on multiple chimera states in nonlocally coupled fitzhugh - nagumo oscillators , provide strong evidence that the phenomenon of chimeras may indeed be relevant in neuroscience applications . moreover , our work verifies the existence of chimera states in coupled bistable elements , whereas to date chimeras were known to arise in models possessing a single stable limit cycle .
finally , we have identified an interesting class of mixed oscillatory states , in which desynchronized neurons are uniformly interspersed among the remaining ones that are either stationary or oscillate in synchronized motion . |
in the quest for better wireless connectivity and higher data rates , the cellular network is becoming heterogeneous , featuring multiple types of base stations ( bss ) with different cell size .
heterogeneity implies that the traditional strategies in cell planning , deployment and communication should be significantly revised @xcite .
since the number of bss becomes comparable to the number of devices @xcite and the deployment pattern of the bss is rather irregular , there are multiple bss from which a device can select one to associate with . the key issue in a wireless heterogeneous setting is the way in which a device selects an access point ( ap ) . the authors in @xcite and @xcite indicate that the ap selected for downlink ( dl ) , termed downlink ap ( dlap ) , is not necessarily the same as the uplink ap ( ulap ) .
the current cellular networks use a criterion applicable to the dl for association in both directions , i.e. a device selects the bs that offers maximal signal - to - interference - plus - noise ratio ( sinr ) in the dl and then uses the same bs for ul transmission .
when dlap@xmath0ulap , we say that the device has a _ decoupled access_. there are two main drivers for decoupled access : ( 1 ) the difference in signal power and interference in dl as compared to ul @xcite ; and ( 2 ) the difference in congestion between bss @xcite .
decoupled dl / ul access has been considered in @xcite , where the authors devise separate criteria for selection of dlap and ulap , respectively , and demonstrate the throughput benefits by using real - world data from planning tools of a mobile operator . another related work that considers different associations in ul and dl is@xcite , where coverage probability and throughput are analyzed for dynamic tdd networks enhanced with device - to - device ( d2d ) links .
this letter focuses on the analytical characterization of the decoupled access by using the framework of stochastic geometry @xcite .
we use the same association criteria as in @xcite .
we perform a joint analysis of the dl and ul association , using the same realization of the random process that describes spatial deployment of the bss and devices .
the analysis is performed for a two - tier cellular network , consisting of macro bss ( mbss ) and femto bss ( fbss ) .
this is used to obtain the central result of the paper , which is the set of association probabilities for different dl / ul configurations .
the analytical results are closely matching the simulations and provide interesting insights about the decoupled access in terms of e.g. fairness regarding the ul throughput .
combining novel results from this letter with already available results in the literature , we provide an analytical justification of the phenomenon of decoupled access compared to current dl - based association in heterogeneous networks .
the letter is organized as follows .
section ii describes the system model . in section iii , we derive the association probabilities and the average throughput .
section iv gives the numerical results and section v concludes the paper .
we model a two - tier heterogeneous cellular network . the locations of bss
are modeled with independent homogeneous poisson point processes ( ppps ) .
we use @xmath1 to denote the set of points obtained through a ppp with intensity @xmath2 , where @xmath3 for mbss , @xmath4 for fbss and @xmath5 for the devices .
similarly , we use @xmath6 with @xmath7 to denote the transmission power of the node @xmath8 . the variables @xmath9 denote the two - dimensional coordinate of mbs and fbs , respectively . the analysis is performed for a typical device located at the origin , which is the spatial point @xmath10 . by slivnyak
s theorem @xcite , the distribution of a point process in @xmath11 is unaffected by addition of a node at the origin . the power received by a typical device in dl from a bs located at @xmath12 , where @xmath13 is denoted by @xmath14 .
the power received by a bs from the typical device in ul is denoted by @xmath15 .
these powers are given by : @xmath16 where @xmath17 and @xmath18 are distances from the points @xmath19 and @xmath20 to the origin , respectively , and @xmath21 is the path loss exponent ( @xmath22 ) .
@xmath23 is independent exponentially distributed random variable with unit mean , representing rayleigh fading at the point @xmath24 .
each receiver in the system has a constant noise power of @xmath25 .
the dl sinr when the device is associated to @xmath8bs is : @xmath26 where @xmath27 and @xmath28 . with the notion of typical point located at the origin , ul sinr is calculated at the location of ulap
this involves calculation of distances between the interfering devices and ulap , which complicates the analysis because none of them is located at the origin .
the problem is solved by using the translation - invariance property of stationary point processes , by which the processes @xmath29 and @xmath30 have the same distribution for all @xmath31 @xcite .
thus , translation of the points for the same value of @xmath32 preserves the process properties .
we use this to shift the points for the distance between the typical device and ulap such that the ulap becomes located at the origin .
the interfering devices are modeled by thinning the ppp @xmath33 in order to take into account that only one device per bs acts as an interferer , using the same resource as the typical device @xcite . by thinning ,
we randomly select fraction of points from the original point process @xcite with probability @xmath34 .
the thinned process is denoted as @xmath35 with density @xmath36 .
the presence of a device in a voronoi cell of a bs forbids the presence of other devices and introduces dependence among the active devices .
however , this dependence is weak , as shown in @xcite , and it is justified to assume independent ppp for the active devices .
the ul sinr at @xmath8bs is defined as : @xmath37
the analysis is divided into two mutually related parts .
we first derive the association probabilities for dl / ul and afterward use them to evaluate the average throughput . in dl
, the device is associated to the bs from which it receives the highest average power . in
ul it is associated to bs to which it transmits with the highest average power .
the average power is obtained by averaging over the received signals given by ( [ dlul_signals ] ) with respect to the fading .
this is justified as the fading - induced variations can lead to ping - pong effects in the association process .
the average received signal powers in dl and ul are : @xmath38 = p_v \left\|x_v\right\|^{-\alpha } \label{dl_signals_ave } ; \mathbb{e}_h\left[\textrm{s}_{v , u}\right ] = p_d \left\|x_v\right\|^{-\alpha } \label{dlul_signals_ave}\end{aligned}\ ] ] ulap will always be the closest bs at distance @xmath39 , where @xmath40 .
the dl case is more complicated since @xmath6 is also variable .
let @xmath41 be the closest point to the origin from the set @xmath1 , with @xmath40 .
the device is associated to @xmath42 otherwise , the device is associated to fbs .
let @xmath43 .
the distribution of @xmath44 follows from the null probability of 2d ppp @xcite , the probability that there is no point in the circle with radius @xmath32 , i.e. @xmath45 . the pdf of @xmath44 is : @xmath46 for two - tier heterogeneous network , there are four possible combinations for choosing dlap and ulap : the probability that a device will be associated to mbs both in dl and ul is : @xmath48 assuming @xmath49 , it follows that @xmath50 .
therefore , the intersection of the events is the region defined by @xmath51 , denoted as region 1 on fig .
[ allregions ] .
= 46 dbm , @xmath52=20 dbm , @xmath21=4).,width=230 ] the association probability of case 1 is calculated as : @xmath53 the derivation of the remaining cases follows the same procedure and we thus only provide the final results .
case 2 defines decoupled access since dlap@xmath0ulap .
the association probability is defined as : @xmath54 the domain that satisfies both events is @xmath55 and is denoted as region 2 on fig . [ allregions ] .
the association probability for case 2 is equal to : @xmath56 the association probability for case 3 should satisfy the following conditions : @xmath57 the intersection ( [ case3 ] ) is an empty set and therefore the probability of dlap@xmath47fbs and ulap@xmath47mbs is @xmath58 .
the probability for associating to fbs in both dl and ul is defined as : @xmath59 since @xmath60 , the intersection of the events is @xmath61 , denoted as region 4 on fig .
[ allregions ] .
the association probability for case 4 is equal to : @xmath62 the average throughput for devices associated to @xmath8bs with @xmath40 in @xmath63l direction using @xmath64l association rules , with @xmath65 for dl and @xmath66 for ul , is calculated as : @xmath67 where @xmath68 is target sinr , @xmath69 is the probability that the instantaneous sinr is greater than @xmath68 and @xmath70 is the average number of associated devices on @xmath8bs and is equal to @xmath71 , with @xmath72 being the association probability for @xmath8bs using @xmath64l association rules . using the association probabilities derived in section [ sec : assocprob ] ,
@xmath72 is expressed as : @xmath73 in order to calculate the average throughput in a two - tier network , we first need to calculate the distribution of the distance to the serving bs , which depends on the association process . for dl association rules , given by ( [ dl_assocrule ] )
, the pdf of the distance to the serving bs is derived in @xcite and is given by : @xmath74 for ul association rules , given by ( [ ul_assocrule ] ) , the pdf of the distance to the serving bs is given by : @xmath75 the average throughput in the dl is calculated as : @xmath76 where @xmath77 and @xmath78 are expressed by the general formula given by ( [ throughputdef ] ) .
using the approach derived in @xcite , we derive the final expression for the average throughput in dl on @xmath8bs : @xmath79 where @xmath80 . the key point in the evaluation
is the following observation : if the device is associated to mbs located at @xmath32 the interfering mbss are at a distance greater than @xmath32 and the interfering fbss are at a distance greater than @xmath81 ; if the device is associated to fbs located at @xmath32 , the interfering fbss are at a distance greater that @xmath32 and the interfering mbss are at a distance greater than @xmath82 .
the average downlink throughput can be also calculated in a more elegant way by using the following : ( equivalent downlink model ) _ a two - tier heterogeneous model is equivalent to a novel homogeneous model with bss deployed by ppp @xmath83 with intensity @xmath84 .
then , the average throughput in dl can be calculated as : _
@xmath85 where @xmath86 .
the dl signal power at the typical point can be represented as @xmath87 , where z is a discrete random variable with two possible values , @xmath88 and @xmath52 , with probabilities @xmath89 and @xmath90 , respectively .
the points @xmath91 are from a ppp @xmath92 with density @xmath93 . by equivalence theorem @xcite ,
the spatial points @xmath94 form new ppp with density @xmath95 $ ] .
the average throughput in ul without decoupled access is calculated using dl association rules given by ( [ dl_assocrule ] ) : @xmath96 where @xmath97 and @xmath98 are evaluated as : @xmath99 it can be observed that the distribution to the serving bs and the association probabilities are from dl association rules .
the average throughput in ul with decoupled access is : @xmath100 where @xmath101 and @xmath102 are evaluated as : @xmath103 * remark 1 * ( equivalent uplink model ) _ a two - tier hetnet model with homogeneous devices is represented by an equivalent homogeneous model with bss deployed by ppp @xmath104 with intensity @xmath105 .
this is a consequence of the ul association rule , which is based on path - loss only .
then , the average throughput in ul can be elegantly calculated as : _
@xmath106 the throughput gain for case @xmath107 , @xmath108 , is defined as the ratio between the throughput achieved with and without decoupling and is denoted as @xmath109 .
the average throughput gain is calculated as : @xmath110 .
= 46 dbm ; @xmath52=20 dbm ; @xmath21=4).,width=302 ] the association probabilities for each of the cases are equal to the percentage of devices that will be associated with the particular case .
[ jap ] shows the association probabilities for different densities of fbss and it gives an important information about dl / ul decoupling .
the percentage of devices that choose decoupled access of case 2 ( dl through mbs and ul through fbs ) increases rapidly by increasing the density of fbss .
as the density of fbss increases further , the probability for decoupled access starts to decrease slowly at the expense of increased probability for case 4 .
there is a region of interest for @xmath111 with a high percentage @xmath112 of devices for which the decoupled access is optimal . as @xmath113
, the probability of decoupled access will go to zero .
[ ulthroughput ] shows the throughput gain for the devices associated to ( m / f)bss and the average gain . there is a difference between , on one side , the accurate simulation of the devices with ppp @xmath33 , and , on the other side , its approximation by simulation of independent ppp @xmath35 for the active devices only .
while the ul coverage probability with dl / ul decoupling is strictly superior , the congestion of the bss affects the throughput in a different manner
. basically , fbss have significantly small dl coverage and therefore associate very small number of devices compared to mbss , but each device gets higher throughput .
it is visible that for @xmath68=2 db the throughput achieved on mbss is 40 times higher with decoupling , while the throughput achieved on fbss is 5 times lower with decoupling .
the average throughput gain is always positive .
it can be concluded that by dl / ul decoupling the devices with low sinr ( located in regions 1 and 2 ) achieve significant improvement , at the expense of marginal decrease in the ul throughput of the devices in region 4 .
this suggests that decoupled access can be used as a tool towards achieving fairness among the accessing devices .
= 46dbm ; @xmath52=20dbm ; @xmath114=20dbm ; @xmath21=4 ; @xmath111=10@xmath115 ; @xmath25=@xmath116).,width=302 ]
this letter considers the problem of device association in a heterogeneous wireless environment .
the analysis is done using models based on stochastic geometry .
the main result is that , as the density of the femto bss ( fbss ) increases compared to the density of the macro bss ( mbss ) , a large fraction of devices chooses to receive from a mbs in the downlink ( dl ) and transmit to a fbs in the uplink ( ul ) .
this is the concept of _ decoupled access _ and challenges the common approach in which both dl and ul transmission are associated to the same bs .
it is shown that the decoupling of dl and ul can be used as a tool to improve the fairness in the ul throughput .
part of our future work refers to the architecture for decoupled access , which includes signaling and radio access protocols .
h. elsawy , e. hossain , and m. haenggi , `` stochastic geometry for modeling , analysis , and design of multi - tier and cognitive cellular wireless networks : a survey , '' _ ieee comm .
surveys & tutorials _ , vol .
15 , no . 3 , 2013 h
.- s jo , y. j. sang , p. xia and j. g. andrews , `` heterogeneous cellular networks with flexible cell selection : a comprehensive downlink sinr analysis , '' _ ieee trans . on wireless comm .
11 , no . 10 , 2012 | wireless cellular networks evolve towards a heterogeneous infrastructure , featuring multiple types of base stations ( bss ) , such as femto bss ( fbss ) and macro bss ( mbss ) .
a wireless device observes multiple points ( bss ) through which it can access the infrastructure and it may choose to receive the downlink ( dl ) traffic from one bs and send uplink ( ul ) traffic through another bs . such a situation is referred to as _ decoupled dl / ul access_. using the framework of stochastic geometry , we derive the association probability for dl / ul . in order to maximize the average received power , as the relative density of fbss initially increases , a large fraction of devices chooses decoupled access , i.e. receive from a mbs in dl and transmit through a fbs in ul .
we analyze the impact that this type of association has on the average throughput in the system .
smiljkovikj , popovski gavrilovska : decoupling of uplink and downlink transmissions heterogeneous networks , decoupled downlink / uplink , average throughput . |
recently , magnetic - field - induced superconductivity ( fisc ) has been observed in organic superconductors @xmath4 and @xmath3 @xcite , where bets is bis(ethylenedithio)tetraselenafulvalene . in these salts ,
localized spins on fe@xmath5 exhibit antiferromagnetic long - range order at ambient pressure at low temperatures .
the fisc in these compounds is considered to be due to the jaccarino - peter mechanism @xcite , where the localized spins are aligned uniformly at high fields .
et al . _ have observed in the compound @xmath3 that superconductivity coexists with the antiferromagnetic long - range order in a low - field region around the zero field @xcite .
they have fitted the experimental phase diagrams by fisher s theory @xcite based on the jaccarino - peter mechanism .
the resultant upper critical field of low - field superconductivity is much smaller than their experimental data .
they have suggested that the reason for the discrepancy is that the jaccarino - peter - fisher theory does not take into account the antiferromagnetic long - range order at low fields @xcite . in recent works ,
one of the authors has extended the jaccarino - peter mechanism to antiferromagnetic superconductors with canted spin structures in magnetic fields @xcite .
the canted spin structure generates the ferromagnetic moments that create exchange fields acting on the conduction electrons through kondo interactions .
if the kondo interactions are antiferromagnetic , the exchange fields partly cancel the zeeman energy . as a result
, the pauli paramagnetic pair - breaking effect can be largely reduced , and the upper critical field can exceed the pauli paramagnetic limit ( chandrasekhar and clongston limit ) @xcite .
this mechanism occurs even in the presence of the orbital pair - breaking effect @xcite .
we call this mechanism an extended jaccarino - peter mechanism in this paper .
since the canted antiferromagnetic phase occurs in the compound @xmath3 for @xmath6 @xcite , we apply the mechanism to this compound . in the compound @xmath3 ,
the fisc has been observed both for @xmath7 and @xmath8 @xcite .
the phase diagrams for @xmath6 and @xmath9 are rather different , and it is attributed to the anisotropy of the fermi surface and the kondo interactions between the localized spins and the conduction electrons .
we take into account the fermi surface anisotropy by effective masses .
the effective - mass model was introduced in ginzburg - landau equations by ginzburg @xcite .
hohenberg and werthamer @xcite pointed out that detailed structures of the fermi surface affect the upper critical field .
rieck and scharnberg @xcite and langmann @xcite obtained general equations for arbitrary fermi surfaces .
et al . _
@xcite and kim _ et al . _
@xcite calculated the upper critical fields of mixed _ d_-wave and _ s_-wave superconductors with effective - mass anisotropy .
recently , kita and arai @xcite have formulated an equation for the upper critical field , taking into account the fermi surface anisotropy and the gap anisotropy on the basis of the rieck and scharnberg theory @xcite .
they have performed the quantitative calculations of the upper critical fields for type - ii superconductors nb , nbse@xmath10 , and mgb@xmath10 using fermi surfaces obtained by first - principles calculations @xcite .
a theory of the upper critical field for layered superconductors has been proposed by lebed and yamaji @xcite , and developed by lebed and hayashi @xcite .
they have found that when the layer spacing is large the upper critical field exhibits a reentrant transition or an enhancement at low temperatures in the quantum region , due to an effect of dimensional crossover induced by the magnetic field @xcite . in the compounds @xmath3 , however , since the upper critical field of the low - field superconductivity did not exhibit either a reentrant transition or an enhancement in the experimental phase diagrams , the dimensional crossover does not seem to take place .
therefore , from a phenomenological consideration , we use the effective - mass model as an approximation instead of strict equations in refs .
@xcite and @xcite .
the effective - mass model is adequate in the ginzburg - landau region for layered superconductors . in this paper , first , we derive critical field equations for _ s_-wave , @xmath1-wave , and @xmath2-wave superconductors with effective - mass anisotropy in three directions , taking into account both orbital and paramagnetic pair - breaking effects . secondly , we take into account the extended jaccarino - peter mechanism .
lastly , we reproduce the phase diagrams of @xmath3 including both the fisc and low - field superconductivity .
we also examine the possibility of the fflo state .
the fflo state has extensively been studied @xcite since pioneering works by fulde and ferrell , and larkin and ovchinnikov @xcite .
the state is taken into account by an extension of the bcs mean - field theory to include the possibility of finite center - of - mass momenta @xmath11 s . in this study
, we adopt a model in which @xmath12 is assumed following gruenberg and gunther @xcite , since we consider the situation in which substantial orbital pair - breaking effect is present . in the organic compounds @xmath4 ,
the possibility of the fflo state in the fisc has been pointed out by uji _
et al . _
@xcite and balicas _ et al . _
@xcite , and also examined theoretically @xcite .
the shape of the phase boundary of the fisc is well reproduced by taking into account the fflo state @xcite .
_ have also argued that the fflo state may occur in @xmath13 from their experimental data @xcite .
recently , the quasi - two - dimensional heavy - fermion superconductor cecoin@xmath14 has been believed to exhibit the fflo state @xcite .
adachi and ikeda @xcite and won _ et al . _
@xcite have calculated the critical fields for @xmath1-wave pairing taking into account the fflo state in connection with cecoin@xmath14 .
this paper is constructed as follows . in the next section ,
we extend the theory of the critical field to the systems with anisotropic fermi surfaces .
s_-wave , @xmath1-wave , and @xmath2-wave pairing superconductors are examined . in sec .
[ sec : fisc ] , we take into account the extended jaccarino - peter mechanism @xcite . in sec . [
sec : res ] , we apply the present theory to the organic superconductor @xmath3 .
we compare experimental and theoretical phase diagrams .
the final section is devoted to the summary and a discussion .
in this section , we derive critical field equations for _ s_-wave , @xmath1-wave , and @xmath2-wave superconductors with an effective - mass anisotropy .
we consider the electron system with the energy dispersion @xmath15 and the pairing interaction @xmath16 where we have defined @xmath17 and introduced anisotropic effective masses @xmath18 , @xmath19 , and @xmath20 . in the magnetic field @xmath21 , the hamiltonian is written as @xmath22 where @xmath23 ^ 2 \psi_\sigma({\mathbf{r } } ) \ ; , \\ { \mathcal{h}}_m & = -\int { d^3}{\mathbf{r}}\ ; \mu_{\text{e}}\mathbf{h } \cdot \left [ \sum_{\sigma\sigma'}\psi_\sigma^\dag({\mathbf{r } } ) \boldsymbol{\sigma}_{\sigma\sigma ' } \psi_{\sigma'}({\mathbf{r } } )
\right ] \;.
\end{split}\ ] ] here , @xmath24 denotes the field operator which annihilates an electron of spin @xmath25 at a point @xmath26 .
we have defined the electronic magnetic moment @xmath27 with the bohr magneton @xmath28 and the _ g _ factor of the conduction electrons @xmath29 .
we use the units such that @xmath30 unless it is explicitly expressed .
we define the @xmath31 axis in the spin space along the direction of @xmath32 .
we should note that the @xmath31 axis does not necessarily coincide with the @xmath33 axis of the electron coordinate depending on the direction of the @xmath32 .
therefore , we have @xmath34 where @xmath35 with @xmath36 , and @xmath37 and @xmath38 for up and down spin states , respectively .
we apply the bcs mean - field approximation to the interaction hamiltonian ( [ eq : inth ] ) as @xmath39 \ ; , \end{split}\ ] ] defining the order parameter @xmath40 we define the center - of - mass coordinate @xmath41 and the relative coordinate @xmath42 , and redefine the gap function as @xmath43 .
by following the procedure of refs .
@xcite and @xcite , we obtain the linearized gap equation @xmath44 near the second - order phase - transition point .
we could examine both orbital and paramagnetic pair - breaking effects with eq .
( [ eq : gap1 ] ) . in the derivation of eq .
( [ eq : gap1 ] ) , it is assumed that the spatial variation of vector potential @xmath45 is sufficiently slow .
therefore , the order parameter becomes a slowly varying function of the center - of - mass coordinate .
we have defined @xmath46 the matsubara frequencies @xmath47 with integer @xmath48 , and the free - electron green s function @xmath49 we introduce @xmath50 with @xmath51 so that the dispersion relation becomes isotropic as @xmath52 in the @xmath53 space , where @xmath54 .
we also introduce @xmath55 with @xmath56 and @xmath57 , so that @xmath58 .
we also define the operator @xmath59 with @xmath60 where @xmath61 and @xmath62 .
we will calculate the explicit form of @xmath59 afterward .
we replace @xmath63 and @xmath64 in @xmath65 , @xmath66 , and @xmath67 with the fermi momentum @xmath68 since electrons with momenta @xmath53 and @xmath69 near the fermi surface mainly contribute to the gap equation .
therefore , we write @xmath65 and @xmath66 as @xmath70 and @xmath71 , respectively , with unit vectors @xmath72 and @xmath73 .
after simple algebra using eq .
( [ eq : freeg ] ) , we obtain @xmath74 t } \\ & \qquad\qquad \times \exp\left[\frac{t}{2i}\text{sgn}(\omega_n ) \tilde{\mathbf{v}}_{\text{f}}(\hat{{\tilde{\mathbf{p } } } } ' ) \cdot\tilde{{\boldsymbol{\pi } } } \right ] \delta_{-\sigma}({\tilde{\mathbf{r}}},\hat{{\tilde{\mathbf{p } } } } ' ) \ ; , \end{split}\ ] ] where we have defined @xmath75 and the density of states at the fermi energy @xmath76 @xcite . since eq .
( [ eq : gap2 ] ) does not depend on the spin value @xmath25 , we omit the spin suffix from now on . by redefining @xmath77 and summing up the matsubara frequencies , the gap equation ( [ eq : gap2 ] ) acquires the form @xmath78 \delta({\tilde{\mathbf{r}}},\hat{{\tilde{\mathbf{p } } } } ' ) \;. \end{split}\ ] ] it
is obvious that if we rewrite the integral variable @xmath79 as @xmath80 , the linearized gap equation ( [ eq : gap3 ] ) is similar to those obtained by many authors @xcite .
the differences between eq .
( [ eq : gap3 ] ) and the equations obtained so far for isotropic systems are that the vector potential @xmath45 is scaled as @xmath81 in @xmath82 , and that the pairing interaction @xmath70 is deformed in the @xmath79-space anisotropic superconductivity . the former scale transformation in @xmath45 has been studied in ginzburg - landau theory @xcite for _ s_-wave pairing .
for _ s_-wave pairing , since @xmath83 , eq . ( [ eq : gap3 ] ) is exactly reduced to the equations for the system with the isotropic fermi surface , except the scale of @xmath84 , which results in the scaling @xmath85 .
therefore , the only difference due to the fermi surface anisotropy is that the critical field in the @xmath86 direction is enhanced or reduced by the factor @xmath87 .
the latter deformation of @xmath70 occurs because @xmath88 for anisotropic superconductivity .
we take into account the deformation of the pairing interaction in the @xmath53 space .
these changes affect the mass anisotropy dependence of the critical field .
now , we derive a more explicit form of the upper critical field equation . we take the _ z _ axis in the direction where the effective mass is the largest .
the compounds with layered structures with large layer spacing are approximately described by the models with @xmath89 .
for example , in the application to the organic superconductor @xmath3 , we take the @xmath90 and @xmath91 axes along @xmath92 and @xmath93 axes of the compounds , respectively .
we study a quasi - two - dimensional superconductor under magnetic fields parallel to the layers ( @xmath94 or @xmath95 ) hereafter .
we derive the critical field equations only for @xmath94 , because those for @xmath95 are obtained by exchanging @xmath96 and @xmath97 .
we assume the uniform magnetic field @xmath32 in the @xmath98 direction parallel to the layers , which is expressed by a vector potential @xmath99 .
then , we obtain the explicit form of @xmath82 as @xmath100 from eqs .
( [ eq : pi ] ) and ( [ eq : tpi ] ) .
we define the differential operators @xmath101 which satisfy the bosonic commutation relations , where @xmath102 the factor @xmath103 is due to the effective - mass anisotropy .
we consider pairing interactions of the form @xmath104 where @xmath105 denotes the symmetry function of symmetry @xmath106 and @xmath107 , @xmath1 , @xmath2 and so on .
the symmetry function @xmath108 is normalized so that @xmath109 ^ 2\rangle = 1 $ ] , where the average on the fermi surface @xmath110 is defined by @xmath111 , where @xmath112 denotes the angle - dependent density of states on the fermi surface .
the gap function is proportional to the symmetry function @xmath108 and expanded by the abrikosov functions @xmath113 of the landau - level indexes @xmath114 , as @xmath115 the abrikosov functions @xmath113 are expressed as @xmath116 where @xmath117 denotes the solution of @xmath118 . in the gap function ( [ eq : gf1 ] ) , we have taken into account the possibility of nonzero @xmath119 for the fflo state , where @xmath120 .
the linearized gap equation ( [ eq : gap3 ] ) can be written as @xmath121 with @xmath122 where we have defined @xmath123 ^ 2 \frac{1-e^{2\tilde{p } \omega_{\text{d}}/{\tilde{v}_{\text{f } } } } } { \tilde{p}^2 \sinh(2 \pi t \tilde{p } / { \tilde{v}_{\text{f } } } ) } \ ; , \\ d_{nn'}^\alpha & = -\frac{t}{2{\tilde{v}_{\text{f } } } } \int { d^3}{\tilde{\mathbf{p}}}[\gamma_\alpha(\hat{{\mathbf{p}}})]^2 \frac{1-e^{2\tilde{p } \omega_{\text{d}}/{\tilde{v}_{\text{f } } } } } { \tilde{p}^2 \sinh(2 \pi t \tilde{p } / { \tilde{v}_{\text{f } } } ) } e^{i(n'-n)\tilde{\varphi } } \\ & \qquad\qquad\times \bigg [ \sum_{j=1}^\infty d_{nn'}^{(j)}(\tilde{p},{\tilde{\theta } } ) \left(\delta_{n'-n,2j } + \delta_{n'-n,-2j}\right ) \\ & \qquad\qquad\qquad\qquad + d_{nn'}^{(0)}(\tilde{p},{\tilde{\theta}})\delta_{nn ' } \bigg ] \end{split}\ ] ] and @xmath124 \\ & \times \sum_{k=0}^{\min(n , n ' ) } \left [ -\frac{\kappa_x}{2 } \tilde{p}^2\sin^2{\tilde{\theta}}\right]^{k+j } \\ & \qquad \times \frac{\sqrt{n!n'!}}{k!(k+2j)![\min(n , n')-k ] ! } \\ & \qquad \times \cos\left[\frac{2\tilde{p}}{{\tilde{v}_{\text{f } } } } ( h+\frac{\tilde{q}{\tilde{v}_{\text{f}}}}{2}\cos{\tilde{\theta}})\right ] -\delta_{nn'}\delta_{j,0 } \;.
\end{split}\ ] ] more explicit expressions of the matrix elements @xmath125 , @xmath126 , and @xmath127 are given below .
defining the zero - field transition temperature @xmath128 , the linearized gap equation ( [ eq : mgeq1 ] ) is written as @xmath129 the transition temperature and the critical field are given by the condition that eq .
( [ eq : mgeq1 ] ) has a nontrivial solution of @xmath130 for the first time when the magnetic field and temperature decrease , respectively , where @xmath119 is optimized . for _ s_-wave pairing
, we insert @xmath131 into eq .
( [ eq : mgeq2 ] ) .
it is easily verified by integrating over @xmath132 that @xmath133 for @xmath134 .
therefore , the matrix elements @xmath135 in the linearized gap equation ( [ eq : mgeqg ] ) are expressed as @xmath136 we consider @xmath1-wave pairing interaction given by eq .
( [ eq : gpi ] ) with @xmath137 , where @xmath138 .
the function @xmath139 is rewritten as @xmath140 \ ; , \end{split}\ ] ] where we have defined the polar coordinates ( @xmath141 ) taking the @xmath142 axis as the polar axis , so that @xmath143 is measured from the @xmath142 axis , and @xmath132 is measured from the @xmath144 axis .
therefore , in the linearized gap equation ( [ eq : mgeqg ] ) , the matrix elements @xmath126 are written as @xmath145 where @xmath146 is defined by @xmath147 with @xmath148 \left[\frac{c({\tilde{\theta}})-1}{c({\tilde{\theta}})+1 } \right]^j \\ & \times \bigg{\ { } \delta_{j,0}-2c({\tilde{\theta } } ) + 4j[c({\tilde{\theta}})]^2 \\ & \qquad + 2[c({\tilde{\theta}})]^3\bigg{\ } } \ ; , \end{split}\ ] ] and @xmath149 . the matrix equation ( [ eq : mgeqg ] ) is decoupled two sets of equations for @xmath150 of even and odd @xmath151 s .
next , we consider @xmath2-wave pairing interaction with @xmath152 the matrix elements @xmath127 take the form @xmath153 where @xmath154 is defined by @xmath155 with @xmath156 \\ & \times \bigg{\ { } 2c({\tilde{\theta}})\left\{1-[c({\tilde{\theta}})]^2\right\ } \delta_{j,0 } \\ & \qquad\quad -\frac{4j}{[c({\tilde{\theta}})]^{2j-1 } } \left[\frac{c({\tilde{\theta}})-1}{c({\tilde{\theta}})+1 } \right]^j \bigg{\ } } \;. \end{split}\ ] ] the order parameter components @xmath157 and @xmath158 couple only with @xmath159 and @xmath160 , respectively , where @xmath161 .
therefore the matrix equation ( [ eq : mgeqg ] ) for @xmath2-wave pairing is decoupled into two sets of equations for @xmath162 of even and odd @xmath151 s .
in this section , we review an extended jaccarino - peter mechanism in antiferromagnetic superconductors @xcite . we have formulated the upper critical field equation of the conduction electron system in the previous section .
now , we shall introduce a localized spin system that is coupled with the conduction electron system through the kondo interaction @xmath163 .
we consider a situation in which the antiferromagnetic long - range order has canted spin structures in magnetic fields .
then , ferromagnetic moments are induced in the localized spins and create exchange fields on the conduction electrons .
the exchange fields partly or completely cancel the magnetic fields in the zeeman energy terms of the conduction electrons when @xmath164 .
the exchange fields are not real magnetic fields in the sense that they only modify the zeeman energy terms but not the vector potential . on the other hand
, the ferromagnetic moments also create real internal magnetic fields . however , such internal magnetic fields are weak when the localized spins are spatially separated from the conduction electrons .
therefore , we neglect them in this study .
we examine the case in which the antiferromagnetic transition occurs at a temperature much higher than the superconducting transition temperature and thus the magnetic ordered state is rigid .
therefore , we neglect modification of the magnetic structure by occurrence of superconductivity . strictly speaking , the localized spin state is modified by the mobile electrons , and the state of the total system should be determined self - consistently @xcite . however , we consider a model in which such modification has already been included and the canted spin structure occurs as a result .
we modify the extended jaccarino - peter mechanism @xcite for the system with anisotropic kondo interactions , introducing kondo coupling constants @xmath165 between the localized spins and the conduction electrons of the @xmath166 component in spin space . when the magnetic field is oriented to the @xmath167 direction , where @xmath168 , we replace @xmath169 in eq .
( [ eq : allh ] ) with the effective hamiltonian @xmath170 where @xmath171 denotes the effective zeeman energy .
therefore , in the critical field equations derived in the previous section , the zeeman energy @xmath172 is replaced with @xmath171 . for @xmath173 , @xmath171
is given by @xmath174 with @xmath175 .
( [ eq : thlf ] ) , @xmath33 , @xmath176 , @xmath177 , and @xmath178 denote number of antiferromagnetic bonds for a given site , number of lattice sites that participate in the kondo interaction with the conduction electrons at @xmath179 , coupling constant of the exchange interaction of the localized spins , and the _ g _ factor of the localized spins , respectively .
here , @xmath180 denotes the critical field of the antiferromagnetic phase which is given by @xmath181 , where @xmath182 denotes the magnitude of the localized spins .
therefore , when @xmath183 , the pauli paramagnetic pair - breaking effect is reduced in antiferromagnetic superconductors . for @xmath184 ,
since the spins are aligned uniformly , the effective zeeman energy @xmath171 is given by @xmath185 because the second term is a negative constant when @xmath186 , the present mechanism is reduced to the conventional jaccarino - peter mechanism .
the zeeman energy is completely compensated at @xmath187 .
it is noted that fisc occurs only when @xmath188 @xcite .
theoretical phase diagram for _ s_-wave pairing in the magnetic fields parallel to the _ c _ axis .
here , sc stands for superconductivity .
the solid and dotted curves show boundaries of superconductivity in the presence and the absence of the fflo state , respectively .
the closed circles show the experimental data of the superconducting transition points for @xmath6 in @xmath3 by konoike _
et al_. ( ref .
the dot - dashed line shows the critical field of the antiferromagnetic phase.,width=245 ] theoretical phase diagram for _ s_-wave pairing in the magnetic fields parallel to the _ a _ axis .
the notations are the same as those in fig .
[ fig : sc ] .
since the canted spin structure does not occur for @xmath9 at low field , the upper critical field of low - field superconductivity has not been calculated.,width=245 ] in this section , we apply the present theory to the antiferromagnetic organic superconductor @xmath3 taking the _ x _ and _ y _ axes along the crystallographic _
c _ and _ a _ axes , respectively .
therefore , we write the effective masses as @xmath189 , @xmath190 , and @xmath191 , and the anisotropic kondo interactions as @xmath192 and @xmath193 . in @xmath3 , the canted spin structure was observed when @xmath6 , and we can apply the extended jaccarino - peter theory . however , for the magnetic field parallel to the magnetic easy axis , i.e. , @xmath9 , the canted spin structure has not been observed and the details of the metamagnetic transition have not been revealed .
therefore , we do not apply the present theory to the low - field region for @xmath9 .
it is convenient to introduce a constant @xmath194 with which @xmath195 and @xmath196 are expressed as @xmath197 ^ 2 \ ; , \\
\kappa_y & = a_m h \left ( \frac{m_a}{m_c } \right)^{1/4 } \left [ \frac{2\pi t_{\text{c}}^{(0)}}{{\tilde{v}_{\text{f } } } } \right]^2 \;. \end{split}\ ] ] we also define the strength ratio of the paramagnetic and orbital pair - breaking effects @xmath198 we explain the procedure to analyze the experimental phase diagram with the present theory .
our critical field equations contain five microscopic parameters : @xmath199 , @xmath200 , @xmath201 , @xmath202 , and @xmath203 .
first , we fit the experimental data of the critical field of the fisc for @xmath6 with the parameters @xmath199 , @xmath200 , @xmath201 , and @xmath202 . secondly ,
using the parameter values of @xmath199 , @xmath200 and @xmath201 determined in the first step , we fit the data of the fisc for @xmath9 with a single parameter @xmath203 .
it is noted that all five parameters are determined only from the curves of the fisc for @xmath6 and @xmath9 .
crudely speaking , the curve of the fisc can be characterized by three real numbers , that is , the magnetic field where the transition temperature is maximum , the maximum transition temperature , and the width of the magnetic fields at @xmath204 .
therefore , we obtain six real numbers from the two experimental data of the fisc , which sufficiently determine all five microscopic parameters .
lastly , using those five parameters determined in the first and second steps , we calculate the upper critical field of low - field superconductivity for @xmath6 , where the canted spin structure coexists . in this last step ,
we do not use any additional fitting parameter .
theoretical phase diagram for @xmath1-wave in the magnetic fields parallel to the _ c _ axis .
the notation is the same as those in fig .
[ fig : sc ] .
the inset shows the theoretical phase diagram below 0.8 k and between 2 and 2.6 t.,width=245 ] theoretical phase diagram for @xmath1-wave in the magnetic fields parallel to the _ a _ axis .
the notation is the same as those in fig .
[ fig : sc].,width=245 ] we carry out this procedure for _ s_-wave , @xmath1-wave , and @xmath2-wave pairings . in figs .
[ fig : sc][fig : ffq ] , we set @xmath205 k and @xmath206 t , which have been observed in the compounds @xmath3 .
for all pairing symmetries examined , the theoretical curves agree well with the experimental data , as shown in figs . [
fig : sc][fig : d2a ] .
in particular , the agreements at low field for @xmath6 are quantitative , because all parameters are determined only from the data of the fisc at high fields . in what follows , we explain the differences depending on the pairing symmetries .
the results for _ s_-wave pairing are shown in figs .
[ fig : sc ] and [ fig : sa ] .
we obtain the effective - mass ratio @xmath207 , which is reasonable in comparison with the value estimated from the shubnikov de haas ( sdh ) oscillations @xcite .
it is confirmed that the fflo state is not realized in the fisc .
in contrast , it is shown that low - field superconductivity exhibits the fflo state at low temperatures .
the results for @xmath1-wave pairing are shown in figs .
[ fig : d1c ] and [ fig : d1a ] .
we obtain @xmath208 , which is too small in comparison with the value estimated from the sdh oscillations @xcite .
it is confirmed that the fflo state is not realized in the fisc .
there are bends in the theoretical curves of the low fields , where the landau level indexes @xmath151 s of the optimum solution alternate .
the solution near @xmath204 is expressed by a linear combination of the abrikosov functions with odd @xmath151 s . in fig .
[ fig : d1c ] , it is found that there are two tricritical points in the phase diagram for @xmath6 .
one at @xmath209 k is the critical point of the normal state , the bcs state with even @xmath151 s , and the fflo state with even @xmath151 s , and another at @xmath210 k is that of the normal state , the fflo states with even @xmath151 s and odd @xmath151 s .
the results for @xmath2-wave pairing are shown in figs .
[ fig : d2c ] and [ fig : d2a ] .
we obtain @xmath211 , which is reasonable in comparison with the value estimated from the sdh oscillations @xcite . in this case , the area of the fflo state is very limited .
it is found that the fflo state is not realized in the fisc and low - field superconductivity near @xmath204 .
there is a bend in the theoretical curve , where the parity of the landau - level indexes @xmath151 s alternates .
the solution of the odd @xmath151 s gives the highest critical field near @xmath204 .
theoretical phase diagram for @xmath2-wave in the magnetic fields parallel to the _ c _ axis .
the notation is the same as those in fig .
[ fig : sc ] .
the inset shows the theoretical phase diagram below 0.8 k and between 2 and 2.6 t.,width=245 ] theoretical phase diagram for @xmath2-wave in the magnetic fields parallel to the _ a _ axis .
the notation is the same as those in fig .
[ fig : sc].,width=245 ] in fig .
[ fig : ffq ] , we show temperature dependences of @xmath212 along the upper critical field curves of low - field superconductivity for _ s_-wave , @xmath1-wave , and @xmath2-wave pairings when @xmath6 . for @xmath1-wave pairing , it is found that the @xmath119 jumps at a temperature where the landau - level indexes @xmath151 of the optimum solution alternate . for @xmath2-wave pairing , it is found that the @xmath119 vanishes for @xmath213 , where the fflo state with odd @xmath151 is unstable . in a brief summary of this section
, experimental results could be reproduced for all _ s_-wave , @xmath1-wave , and @xmath2-wave parings . for @xmath214-wave pairing ,
the upper critical fields of low - field superconductivity are smaller than the experimental data , but they are improved by taking into account the fflo state .
the in - plane mass anisotropies are estimated as @xmath215 , 0.468 , and 0.865 for _ s_-wave , @xmath1-wave , and @xmath2-wave pairings , respectively .
the results for _ s_-wave and @xmath2-wave pairings agree with the value obtained from the sdh oscillations @xcite , whereas the result for @xmath1-wave pairing largely deviates from it .
we found that the fflo state does not occur in the fisc , while it occurs in low - field superconductivity .
temperature dependence of the magnitudes of the total pair momentum along the upper critical field curves for ( a ) _ s_-wave , ( b ) @xmath1-wave , and ( c ) @xmath2-wave pairings in the magnetic fields parallel to the _ c _ axis .
the sets of parameters for ( a ) , ( b ) , and ( c ) are the same as in figs .
[ fig : sc ] , [ fig : d1c ] , and [ fig : d2c ] , respectively.,width=245 ]
we have studied the upper critical fields of antiferromagnetic _ s_-wave , @xmath1-wave , and @xmath2-wave superconductors with fermi surface anisotropy . we have derived the linearized gap equations when the magnetic field is applied parallel to layers , taking into account the effective - mass anisotropy , the fflo state , and the extended jaccarino - peter mechanism .
we have applied the theory to the organic superconductor @xmath3 , and obtained good agreement between the theoretical and experimental results .
while the agreements have been obtained for all cases of _ s_-wave , @xmath1-wave , and @xmath2-wave pairings , the resultant in - plane mass ratios @xmath216 are different from one another .
the mass ratio for @xmath1-wave pairing seems too small , although those for _ s_-wave and @xmath2-wave pairings are close to the value estimated from the sdh oscillations @xcite .
the mass anisotropy dependence of the pure orbital limits of the upper critical fields is shown in fig .
[ fig : mae ] . the upper critical fields for _ s_-wave , @xmath1-wave , and @xmath2-wave pairings become larger when the ratio @xmath217 decreases .
it is found that the upper critical field is the most influenced for @xmath2-wave pairing by a change in the ratio @xmath217 , while it is less influenced for @xmath1-wave pairing .
this is why the resultant ratio @xmath217 to fit the experimental data becomes much smaller for @xmath1-wave pairing than for @xmath2-wave pairing . the mass anisotropy dependence of the pure orbital limits for @xmath6 .
the open circles , open diamonds , and open triangles show the results for _ s_-wave , @xmath1-wave , and @xmath2-wave pairings , respectively .
the solid , dotted , and dashed curves are guides to the eye . here
, @xmath218 denotes the orbital limit @xmath219 of the isotropic case ( @xmath220).,width=245 ] this result is physically interpreted as follows .
for example , we consider the situation that @xmath221 .
the origin of the orbital pair - breaking effect is the lorentz force that acts on electrons .
when the magnetic fields are applied along the _ c _ axis , the lorentz force is the strongest for electrons near the @xmath222 axis , where the fermi velocity is the largest , in the two - dimensional momentum space . when the mass ratio @xmath217 decreases with the fixed electron number , i.e. , when the effective mass @xmath223 increases , the fermi velocity in the _ a _ direction decreases .
therefore , the lorentz force acting on the electrons on the fermi surface near the @xmath222 axis becomes stronger and superconductivity is more suppressed when @xmath217 decreases .
the order parameter of @xmath2-wave pairing has nodes on the @xmath222 axis , while that of @xmath1-wave pairing has the largest amplitude there .
therefore , @xmath1-wave pairing is less affected than @xmath2-wave pairing by a change in the ratio @xmath217 .
it has been found from fig .
[ fig : d2c ] that the fflo state does not occur for @xmath2-wave pairing in low - field superconductivity near @xmath204 .
this result is physically explained as follows .
since the effective density of states for @xmath106-wave pairing is proportional to @xmath224 ^ 2 $ ] in the gap equation , areas of the fermi surface near the peak of @xmath224 ^ 2 $ ] have a greater effect on the nesting condition for the fflo state , but those near the nodes of @xmath224 ^ 2 $ ] have less effect @xcite .
hence , in the absence of the orbital pair - breaking effect , the vector @xmath11 tends to points in the areas of the fermi surfaces where @xmath224 ^ 2 $ ] exhibits a peak . in this study , however , we adopt the model in which @xmath12 is assumed due to the orbital pair - breaking effect . therefore , for @xmath2-wave pairing , the fermi surface nesting is less effective when @xmath225 or @xmath226 , because the order parameter has nodes on the @xmath227 and @xmath228 axes .
in contrast , for @xmath1-wave pairing , because the order parameter has peaks on the @xmath227 and @xmath228 axes , the fflo state occurs in low - field superconductivity , when @xmath225 or @xmath226 .
theoretical predictions of the phase boundary of superconductivity for various values of @xmath229 .
the other parameters are common to fig .
[ fig : sc ] .
the thin dotted and thin two - dot - dashed curves show the pauli paramagnetic and orbital limits , respectively . in this figure ,
the fflo state is taken into account , but modifies the curves slightly at low temperatures . at the upper critical fields of the fisc for @xmath230 , the fflo state does not occur.,width=245 ] in the application to @xmath3
, it is found that the fflo state is not present in the fisc in contrast to @xmath4 , while it is present in low - field superconductivity . in @xmath4 ,
the phase boundary of the fisc can be well reproduced if the possibility of the fflo state is taken into account @xcite .
for the fflo state to occur , the orbital pair - breaking effect needs to be sufficiently suppressed . in @xmath4
, this condition seems to be satisfied even at high field due to low dimensionality .
in contrast , it was suggested that @xmath3 has a larger interlayer electron hopping energy than @xmath4 @xcite .
therefore , in @xmath3 , the orbital effect is substantial at high fields , as verified by comparison of the zero - field transition temperature and the maximum transition temperature of the fisc .
the upper critical fields of low - field superconductivity agree with the experimental data , as shown in figs .
[ fig : sc ] , [ fig : d1c ] , and [ fig : d2c ] .
further , for @xmath214-wave pairing , the agreement is improved if the fflo state is taken into account .
however , they are slightly smaller than the experimental data . this discrepancy may be removed by taking into account a mixing effect in the presence of the weak triplet pairing interaction , as in the compounds @xmath4 @xcite .
a slight discrepancy also occurs in the lower critical fields of the fisc , where the fflo state does not occur as explained above .
if the mixing effect occurs , this discrepancy may also be removed . for @xmath1-wave and @xmath2-wave pairings in low - field superconductivity
, we found an internal transition between the vortex states with different landau level indexes @xmath151 , which is analogous to internal transitions in a two - dimensional system in tilted magnetic fields @xcite .
both the fflo state and the vortex state with higher @xmath151 s originate from the pauli paramagnetic effect as discussed in ref .
@xcite , because in the absence of zeeman energy only the vortex state with lower @xmath151 s occurs . due to the transitions to the fflo state and
the vortex states with higher @xmath151 s , the upper critical fields of low field superconductivity exhibit downward convex curves for all pairing symmetries examined , which agrees with the experimental data . we have determined the microscopic parameters only from the curves of the fisc , and calculated the upper critical field of low - field superconductivity with those parameter values without any additional fitting parameters . from the accordance between the theoretical and experimental results on low - field superconductivity , the extended jaccarino - peter mechanism @xcite seems to be realized in @xmath3 . applying the extended jaccarino - peter mechanism to the same models with different parameter values
, we obtain some unusual phase diagrams as shown in fig .
[ fig : sj ] . when @xmath231 , the transition temperature curve has double peaks . when @xmath232 , superconductivity occupies large area in the phase diagram .
similar phase diagrams have been obtained in two- and three - dimensional systems @xcite . in this study
, we confirmed that the phase diagrams as obtained above can be realized in a more realistic model of organic superconductors .
if we control the value of @xmath163 by material design , such critical field curves would be observed . in the present theory ,
the effective mass @xmath233 is included in the parameter @xmath199 .
since we have obtained the value of @xmath199 , we can estimate a value of @xmath233 , if we have the value of @xmath234 appropriate for @xmath3 .
the upper critical field in low fields for @xmath9 could not be reproduced by the present theory .
it has been found that the magnetic easy axis of antiferromagnetic long - range order in low fields is along the _ a _ axis @xcite , and that the metamagnetic transition occurs at the antiferromagnetic phase boundary close to the upper critical field @xcite . in order to reproduce the upper critical field in low fields for @xmath9 , we need to reproduce the localized spin structure .
this remains for future study . in conclusion ,
the experimental phase diagrams of the antiferromagnetic superconductor @xmath3 are theoretically reproduced by the present theory . in particular ,
the low - field phase for @xmath6 is well reproduced by the model parameters that are determined from the critical fields of the fisc .
therefore , the extended jaccarino - peter mechanism seems to be realized in the present compound at low fields . due to this mechanism , some unusual phase diagrams may occur in compounds with appropriate energy parameters , in the future .
we would like to thank s. uji and t. konoike for discussions and for providing us with the experimental data .
we wish to acknowledge useful discussions with k. doi .
this work was partially supported by the ministry of education , science , sports and culture of japan , grant - in - aid for scientific research ( c ) , no . 16540320 , 2005 .
t. konoike , s. uji , t. terashima , m. nishimura , s. yasuzuka , k. enomoto , h. fujiwara , b. zhang , e. fujiwara , and h. kobayashi , phys .
b * 72 * , 094517 ( 2005 ) ; t. konoike , h. fujiwara , b. zhang , h. kobayashi , m. nishimura , s. yasuzuka , k. enomoto , and s. uji , physica c * 412 - 414 * , 107 ( 2004 ) .
l. balicas , j.s .
brooks , k. storr , d. graf , s. uji , h. shinagawa , e. ojima , h. fujiwara , h. kobayashi , a. kobayashi , and m. tokumoto , solid state commun .
* 116 * , 557 ( 2000 ) ; s. uji , h. shinagawa , y. terai , t. yakabe , t. terashima , l. balicas , j.s .
brooks , e. ojima , h. fujiwara , h. kobayashi , a. kobayashi , and m. tokumoto physica b * 298 * , 557 ( 2001 ) . | we derive critical field @xmath0 equations for antiferromagnetic _ s_-wave , @xmath1-wave , and @xmath2-wave superconductors with effective mass anisotropy in three dimensions , where we take into account ( i ) the jaccarino - peter mechanism of magnetic - field - induced superconductivity ( fisc ) at high fields , ( ii ) an extended jaccarino - peter mechanism that reduces the pauli paramagnetic pair - breaking effect at low fields where superconductivity and an antiferromagnetic long - range order with a canted spin structure coexist , and ( iii ) the fulde - ferrell - larkin - ovchinnikov ( fflo or loff ) state . as an example , experimental phase diagrams observed in organic superconductor @xmath3 are theoretically reproduced . in particular , the upper critical field of low - field superconductivity
is well reproduced without any additional fitting parameter other than those determined from the critical field curves of the fisc at high fields .
therefore , the extended jaccarino - peter mechanism seems to occur actually in the present compound .
it is predicted that the fflo state does not occur in the fisc at high fields in contrast to the compound @xmath4 , but it may occur in low - field superconductivity for _ s_-wave and @xmath1-wave pairings .
we also briefly discuss a possibility of compounds that exhibit unconventional behaviors of upper critical fields . |
observations of accretion powered pulsars began with the discovery of periodic x - ray pulsations from cen x-3 by _ uhuru _ ( giacconi et al .
1971 ; schreier et al .
. qualitative understanding of accretion powered pulsars was achieved in the 1970s ( pringle & rees 1972 ; davidson & ostriker 1973 ; lamb et al . 1973 ) .
ghosh and lamb presented an accretion disk theory to address the accretion powered pulsar observations in the 1970s in terms of a fastness parameter , material and magnetic torques in the case of a stable prograde accretion disk ( ghosh&lamb , 1979a , b ) . in the absence of a stable accretion disk ,
numerical simulations were used to probe the nature of accretion ( anzer et al . 1987 ; taam&fryxell 1988a,1988b,1989 ; blondin et al .
1990 ) .
observations of pulse frequency changes in accretion powered pulsars are direct signs of torques exerted on the pulsar .
these torques can originate either outside or inside the star ( lamb et al .
1978 ; baykal & gelman 1993 ) .
internal torques depend on the coupling between interior components , in particular the core superfluid , and the solid outer crust ( baykal et al .
external torques depend on the magnetic field strength of the neutron star and on the type of accretion flow to the neutron star .
if the neutron star accretes mass from an accretion disk , torques are produced either by the angular momentum transfer of the plasma to the magnetic field in the magnetospheric radius via interaction of the inner boundary of the disk and the magnetic field lines ( causing material torques ) or by the interaction of the disk and the magnetic field ( causing magnetic torques ) ( ghosh & lamb 1979a , b ) .
if the accretion results from roche lobe overflow of the companion , a persistent prograde keplerian accretion disk forms and the disk creates material and magnetic torques causing the neutron star to spin - up or spin - down .
for such a configuration , material torques can only give spin - up contribution to the net torque , while magnetic torques may give either spin - up or spin - down contribution .
if the companion does not fill its roche lobe , then the neutron star may still accrete mass from its companion s wind . from the hydrodynamic simulations
, it is seen that the stellar wind is disrupted in the vicinity of a compact x - ray source ( the neutron star for our case ) which causes plasma to lose its homogeneity .
the interaction of the incident flow with the shock fronts around the neutron star can produce retrograde and prograde temporary accretion disks ( anzer et al .
1987 , taam & fryxell 1988a , 1988b , 1989 ; blondin et al .
1990 ) .
the relations between x - ray luminosity , torque and specific angular momentum may lead to important clues about the accretion process .
if the neutron star accretes mass from a stable prograde accretion disk , we expect a positive correlation between x - ray flux and torque ( ghosh&lamb 1979a , b ) . for the case of continuous changes in accretion geometry , we can expect a correlation between specific angular momentum and torque which may be the sign of significant torque changes while the luminosity does not vary significantly ( taam & fryxell 1988a , 1988b , 1989 ; blondin et al . 1990 ) . in this paper ,
we use batse ( burst and transient source experiment ) 20 - 60 kev band x - ray flux and pulse frequency time series of three high mass systems ( vela x-1 , gx 301 - 2 , and oao 1657 - 415 ) .
this database is a part of the flux and pulse frequency database for accretion powered pulsars which was discussed before by bildsten et al .
( 1997 ) . using these time series
, we investigate the correlations of torque , x - ray luminosity and specific angular momentum .
detailed studies on torque and x - ray luminosity using the batse x - ray flux and pulse frequency data were presented before for gx 1 + 4 ( chakrabarty 1996,1997 ) and oao 1657 - 415 ( baykal 1997 ) .
baykal(1997 ) also discussed correlations of specific angular momentum with torque and x - ray luminosity for oao 1657 - 415 .
gx 1 + 4 , which was continuously spinning - up in the 1970 s , later exhibited a continuous spin - down trend with an anticorrelation of torque and x - ray luminosity , i.e the spin - down rate is increased with increasing x - ray luminosity ( chakrabarty 1996,1997 ) .
this spin - down episode was interpreted as evidence for a retrograde keplerian accretion disk ( nelson et al .
1997 ) which may originate from the slow wind of a red giant ( murray et al .
1998 ) . other explanations for this spin - down episode were the radially advective sub - keplerian disk ( yi et al .
1997 ) and warped disk ( van kerkwijk et al .
1998 ) models .
x - ray luminosity , torque and specific angular momentum correlations for oao 1657 - 415 were studied earlier ( baykal , 1997 ) . that work employed a flux and pulse frequency data string covering a @xmath030% shorter time interval compared to the content of the oao 1657 - 415 data studied in the present paper .
in that paper , correlations of pulse frequency derivative ( proportional to torque exerted on the neutron star ) , pulse frequency derivative over flux ( proportional to specific angular momentum of the accreted plasma ) and flux ( proportional to luminosity ) were discussed .
it was found that the most natural explanation of the observed x - ray flux and pulse frequency derivative fluctuations is the formation of temporary accretion disks in the case of stellar wind accretion .
the present paper extends the analysis on oao 1657 - 415 to cover a larger data string .
we also present the results of a similar analysis in two other pulsars , vela x-1 and gx 301 - 2 .
in the next section , database is introduced , and pulse frequency , pulse frequency derivative and flux time series are presented .
a discussion of the results and conclusions are given in section 3 .
batse is made up of eight detector modules located at the corners of cgro ( compton gamma ray observatory ) .
these detectors have enabled continuous all sky monitoring for both pulsed and unpulsed sources above 20 kev since 1991 .
batse daily monitors the pulse frequency and x - ray flux of three low mass binaries , five high mass binaries and seven previously known transients .
it has also discovered new transients ( bildsten et al . 1997 ) .
this paper is based on orbitally corrected batse 20 - 60 kev band x - ray flux and pulse frequency time series of vela x-1 , gx 301 - 2 and oao 1657 - 415 which are obtained from the ftp site `` ftp.cossc.gsfc.nasa.gov '' . in this paper , we assume that time variation of the 20 - 60 kev band flux represents the time variation of the bolometric x - ray flux . it should be noted that the flux time series , reported in fig . 1 , 4 , and 7 might not be representative of the time variation of the bolometric x - ray flux .
pulse frequency and flux time series are binned by considering that measurement errors dominate in short time lags or high frequencies in the pulse frequency derivative power spectrum ( bildsten et al . 1997 ) .
we choose our bin sizes to the extent that the measurement errors do not dominate on pulse frequency derivatives .
bin sizes are 45 days , 30 days and 16 days for vela x-1 , gx 301 - 2 and oao 1657 - 415 respectively .
pulse frequency derivatives are found by averaging left and right derivatives of pulse frequency values so that each pulse frequency derivative corresponds to a single flux and time value .
we present x - ray flux , pulse frequency and pulse frequency derivative time series in fig .
1,4 and 7 . in fig
. 2 , 5 and 8
, we present the plot of pulse frequency derivative and flux values corresponding to the same time value .
since pulse frequency derivative and x - ray flux are directly proportional to torque and x - ray luminosity , these figures show the relation between torque and x - ray luminosity .
pulse frequency derivative over flux is directly proportional to the specific angular momentum of the plasma ( @xmath1 where l is the specific angular momentum , @xmath2 is the frequency derivative and @xmath3 is the mass accretion rate ) .
fig . 3 , 6 , 9 which are plots of pulse frequency derivative over flux and pulse frequency derivative show the relation between specific angular momentum and torque .
pulse frequency derivative over flux time series are created by dividing each pulse frequency derivative value with the flux value corresponding to the same time .
283s pulsations from vela x-1 were discovered by _
sas-3 in 1975 ( mcclintock et al . 1976 ) .
it is the brightest persistent accretion powered pulsar in the 20 - 60 kev energy band ( bildsten et al .
. optical companion of vela x-1 is the b0.5 ib supergiant hd77581 ( vidal et al .
this system is an eclipsing binary with eccentricity of @xmath4 and period of 8.96days ( rappaport et al .
_ x - ray flux , pulse frequency and pulse frequency derivative values of vela x-1 cover the interval between 48371 mjd and 50580 mjd ( fig.1 ) . in fig.2 , pulse frequency derivative and corresponding x - ray flux data were plotted . no correlation between frequency derivative and flux is found . from fig.3 , it is seen that there is a correlation between pulse frequency derivative over flux and frequency derivative .
700s pulsations from gx 301 - 2 ( 4u 1223 - 62 ) were discovered by ariel 5 in 1975 ( white et al .
gx 301 - 2 was , on average , neither spinning up nor spinning down between 1975 and 1985 .
after 1985 , a spin - up episode began ( nagase 1989 ) , reaching the current pulsar spin period of @xmath5 .
gx 301 - 2 , being in a 41.5 day eccentric orbit ( @xmath6 ) , is an accreting pulsar with the supergiant companion wray 977 ( sato et al .
this source exhibited two rapid spin - up episodes at @xmath7 mjd and @xmath8 mjd .
these two spin - up episodes suggest the existence of a long - lived ( @xmath9days ) accretion disk ( koh et al .
1997 ) .
x - ray flux , pulse frequency and pulse frequency derivative values of gx 301 - 2 cover the interval between 48371 mjd and 50577 mjd ( fig.4 ) .
fig.5 is the plot of pulse frequency derivative and corresponding x - ray flux values .
no correlation between pulse frequency derivative and flux is found . from fig.6
, we see that there is a correlation between pulse frequency derivative over flux and pulse frequency derivative . in fig .
5 and 6 , there exist two points with pulse frequency derivative values greater than @xmath10 hz / day .
these points correspond to two rapid spin - up episodes which was interpreted as a sign of transient prograde accretion disk ( koh et al .
1997 ) .
1657 - 415 was first detected by the copernicus satellite ( poldan et al . 1978 ) .
38.22s pulsations from oao 1657 - 415 were found in 1978 from heao 1 observations ( white et al .
the optical companion of oao 1657 - 415 is probably a ob type star .
its binary orbit period with an eccentricity of @xmath11 was found to be 10.4 days from the eclipse due to its companion from timing observations of this source with the batse observations ( chakrabarty 1993 ) .
detailed studies on x - ray flux and pulse frequency derivative changes were performed earlier on a less extensive batse 20 - 60kev x - ray flux and pulse frequency data string ( baykal 1997 ) . in the current data string , x - ray flux and pulse frequency and pulse frequency derivative values of oao 1657 - 415 cover the interval between 48372 mjd and 50302 mjd ( fig.7 ) .
8 presents pulse frequency derivative and corresponding x - ray flux time series .
no correlation between pulse frequency derivative and flux is found .
there is a correlation between pulse frequency derivative over flux and pulse frequency derivative as seen from fig .
the torque on the neutron star can be expressed in terms of the specific angular momentum ( @xmath12 ) added to the neutron star by the accreted plasma and the mass accretion rate ( @xmath3 ) : this equation is a general expression which is valid for both accretion from a keplerian disk and accretion from the stellar wind . in the case of accretion from a keplerian disk ( ghosh & lamb 1979b ) we have , where @xmath17 is the fastness parameter and @xmath18 , the dimensionless torque , represents the ratio of the total ( magnetic plus material ) torque to the material torque .
@xmath19 is the radius of the inner edge of the disk , which can be written as @xmath20 where @xmath21 is the magnetic moment of the neutron star .
dimensionless torque can approximately be written as where @xmath23 is a critical value for the fastness parameter which defines the boundary between spin - up and net spin - down phases .
dimensionless torque may be positive or negative depending on the fastness parameter @xmath24 which is defined as where @xmath26 is the neutron star s spin angular velocity and @xmath27 is the angular velocity corresponding to the keplerian velocity at the magnetospheric radius .
for a slowly rotating neutron star for which @xmath28 ( ghosh&lamb 1979b ; wang 1995 ; li&wang 1996,1999 ) , we expect a spin - up torque and for a very fast rotating neutron star ( @xmath29 ) , we expect a spin - down torque .
we also expect to see positive correlation between torque and mass accretion rate if the disk is prograde . for a disk formed from roche lobe overflow of the companion
we expect the plasma to carry positive specific angular momentum , so a prograde disk should be formed .
the total torque exerted on the neutron star is proportional to the material torque for a given fastness parameter .
since the specific angular momentum weakly depends on mass accretion rate ( @xmath30 ) , the net torque becomes proportional to the mass accretion rate ( @xmath3 ) .
the bolometric x - ray luminosity is also correlated with the mass accretion rate ( @xmath31 where r is the radius of the neutron star ) .
thus , we expect a correlation between torque and x - ray luminosity for the sources accreting from prograde accretion disks ( torques are positive ) .
for the similar reasons , an anticorrelation between torque and x - ray luminosity is expected from a retrograde accretion disk ( torques are negative ) . however , for the pulsars we have considered , we see no correlation of pulse frequency derivative and flux(fig
. 2,5,8 ) .
moreover , there are several transitions from spin - up and spin - down .
these results suggest that we do not have stable accretion disks for these sources .
there is a model which does not exclude the possibility of a stable keplerian accretion disk(anzer&brner 1995 ) .
this model is proposed to explain the torque reversals in vela x-1 suggesting the existence of a stable accretion disk lying just outside the magnetosphere mass of which exhibits small variations .
it is found that change of the disk s mass in a random way can produce variations of torque in vela x-1 .
however , the authors have not identified a specific physical process which is responsible for such random mass fluctuations . where i is the moment of inertia of the neutron star , l is the specific angular momentum of the accreting matter , @xmath34 , @xmath35 , and @xmath36 are the variations of spin frequency derivative , mass accretion rate ( proportional to x - ray flux ) and specific angular momentum ( proportional to pulse frequency derivative over flux ) respectively . when the variations of torques for the three pulsars are concerned , we have considerable changes and transitions from negative values to positive values and vice versa . for the case of an accretion from the wind ,
numerical simulations show that the changes in the sign of specific angular momentum is possible ( anzer et al .
1987 ; taam & fryxell 1988a,1988b,1989 ; blondin et al .
1991 , murray et al .
so , we can observe transitions from spin - up to spin - down or vice versa even if there is not a significant change in mass accretion rate .
for wind accreting sources , continuos change in accretion geometry rules out the existence of a stable accretion disk .
thus , it is unlikely to see a correlation between torque and x - ray luminosity which is the case for our sources as well . for such sources
, a correlation between specific angular momentum and torque shows that there are considerable changes in torque while there are not very considerable changes in x - ray luminosity .
this indicates changes in accretion geometry .
the correlation between specific angular momentum and torque exists for all of the three sources ( fig . 3 , 6 and 9 ) .
there are recent developments which explain the torque reversals in accretion powered pulsars .
negative torques may come from a retrograde keplerian accretion disk ( nelson et al .
1997 ) which may , for instance , originate from a red giant ( murray et al . 1998 ) .
these spin - down torques may be the result of an advection dominated sub - keplerian disk for which the fastness parameter should be higher than that of a corresponding keplerian disk causing a net spin - down ( yi et al .
1997 ) or the warping of the disk so that the inner disk is tilted by more than 90 degrees ( van kerkwijk et al . 1998 ) .
torque and x - ray luminosity correlation is expected from these models which is not found for our sources . moreover , in these models timescales for torque reversals are either not certain or of the order of years .
so , these ideas about torque reversals are not supported by the behaviour of the three pulsars we have considered . our considerations about correlations between pulse frequency derivative , flux and specific angular momentum
give insights about the physics of the plasma flow in the vicinity of the neutron stars accreting from the winds of their companions .
we had similar conclusions for all three sources .
we found that it is unlikely for these sources to have stable prograde keplerian accretion disks since they show both spin - up and spin - down episodes and they do not show correlation between torque and luminosity .
correlation between specific angular momentum and torque for these sources may indicate the continuous change in accretion geometry .
this shows the possibility of temporary prograde and retrograde accretion disk formation
. it would be also possible to have stronger idea about the accretion geometry if we had measurements of the pulse frequency derivative with a time resolution of the order of hours , which is a typical time scale of accretion geometry changes for these systems ( taam & fryxell 1988a , 1988b , 1989 ; blondin et al .
for such a case , we would allow a better comparison of the changes in flux and specific angular momentum and it would be interesting to detect both high flux and low specific angular momentum points corresponding to the radial flow cases and to see low flux and high specific angular momentum points corresponding to the accretion from prograde ( high positive specific angular momentum ) or retrograde ( high negative specific angular momentum ) accretion disks .
our flux and pulse frequency time series are found using the x - ray flux and pulse frequency between 20 - 60 kev .
observatories with higher time resolution and capable of detecting photons from a wider energy band will be useful to observe the torque , x - ray luminosity and specific angular momentum changes for these pulsars .
an encouraging example is given by rxte ( rossi x - ray timing explorer ) observations of the accretion powered pulsar 4u 1907 + 09 .
dipping activity in the x - ray intensity was found which was interpreted as a consequence of inhomogeneity of the wind from the companion(int zand et al .
1997 ) , and more recently , it was shown from the rxte observations that 4u 1907 + 09 exhibited transient @xmath37s qpo oscillations during a flare which was superposed on long term spin - down rate .
this was interpreted as a sign of transient retrograde accretion disk ( int zand et al .
1998a , b ) .
rxte observations of vela x-1 , gx 301 - 2 , and oao 1657 - 415 may provide more understanding on the nature of accretion flow through tests on flux and frequency time series at higher resolution .
we acknowledge dr.ali alpar and dr.len balman for critical reading of the manuscript .
it is a pleasure to thank dr.matthew scott and dr.bob wilson for their help to our questions about the database .
we thank the compton gamma ray observatory team at heasarc for the archival data .
anzer , u. et al .
1987 , a&a , 176 , 235 anzer , u. , brner , g. 1995 , a&a , 299 , 62 baykal , a. et al .
1991 , a&a , 252 , 664 baykal , a. 1997 , a&a , 319,515 baykal , a. , gelman , h. 1993 , a&a , 267 , 119 bildsten , l. et al .
1997 , apjs , 113,367 blondin , j.m .
et al . 1990 , apj , 356 , 591 chakrabarty , d. 1993 , apj , 403 , l33 chakrabarty , d. 1996 , ph.d thesis submitted to california institute of technology chakrabarty , d. 1997 , apj , 481 , l101 davidson , k. , ostriker , j.p .
1973 , apj , 179 , 585 ghosh , p. , lamb , f.k .
1979a , apj , 232 , 259 ghosh , p. , lamb , f.k .
1979b , apj , 234 , 296 giacconi , r. et al .
1971 , apj , 167 , l67 int zand , j.j.m . ,
strohmayer , t.e . , baykal , a. 1997 , apj , 479 , l47 int zand , j.j.m . ,
baykal , a. , strohmayer , t.e . 1998a , apj , 496 , 386 int zand , j.j.m . ,
strohmayer , t.e . , baykal , a. 1998b , nuclear physics b ( proc.suppl .
) 69/1 - 3 , 224 koh , d.t .
1997 , apj , 479 , 913 lamb , f.k .
1973 , apj , 184,271 lamb , f.k .
1978 , apj , 224 , 969 li , x .- d . ,
wang , z .- r .
1996 , a&a , 307l , 5 li , x .- d .
, wang , z .- r .
1999 , astro - ph 9901083 mcclintock , j.e .
1976 , apj , 206 , l99 murray , j.r .
1998 , astro - ph 9810118 nagase , f. 1989 , pasj , 41 , 1 nelson , r.w .
1997 , apj , 488 , l117 polidan , r.s .
1978 , nature , 275 , 296 pringle , j.e .
, rees m.j . , 1972 ,
a&a , 21 , 1 rappaport , s. et al .
1976 , apj , 206 , l103 sato , n. et al .
1986 apj , 304 , 241 schreier , e. et al .
1972 , apj , 172 , l79 taam , r.e . ,
fryxell , b.a . , 1988a , apj , 327 , l73 taam , r.e . , fryxell , b.a . , 1988b , apj , 335 , 862 taam , r.e . , fryxell , b.a . , 1989 , apj , 339 , 297 van kerkwijk , m.h .
1998 , astro - ph 9802162 vidal , n.v .
1973 , apj , 265 , 1036 yi , i. et al .
1997 , apj , 481 , l51 wang , y .- m .
1995 , apj , 449 , l153 white , n.e .
1976 , apj , 209 , l119 white , n.e .
1979 , apj , 233 , l121 | using archival batse ( burst and transient source experiment ) 20 - 60 kev band x - ray flux and pulse frequency time series , we look for correlations between torque , luminosity and specific angular momentum for three high mass x - ray binary pulsars vela x-1 , gx 301 - 2 and oao 1657 - 415 .
our results show that there is no correlation between pulse frequency derivative and flux which may be an indication of the absence of stable prograde accretion disk . from the strong correlation of specific angular momentum and torque ,
we conclude that the accretion geometry changes continuously as suggested by the hydrodynamic simulations(blondin et al .
1990 ) . |
the formation of dense molecular condensations , and eventually stars , involves large changes in the physical properties of the atomic and molecular gas .
these changes also have very specific consequences on the chemical interactions of the gas and dust inside the forming cores / stars .
in particular as the density increases molecules in the gas phase collide with dust grains with greater frequency and , if the molecules stick with any reasonable efficiency , they will deplete from the gas phase .
thus , the density gradient that results from core condensation is accompanied by chemical gradients , with the inner parts of the core representing high density chemistry ( with freeze - out onto grains ) , and the outer parts representing the original low density molecular composition .
indeed , these molecular depletions have been suggested as the primary chemical indicator of the earliest stages of the star formation process ( mundy & mcmullin 1997 ; bergin & langer 1997 ) . such chemical structure should be particularly evident in dark clouds because the low temperatures ( @xmath0 k ) preclude thermal evaporation of the molecules frozen in the mantle .
the process of star formation is by nature a dynamic one and molecular emission is the primary method for obtaining information on kinematic motions inside molecular clouds .
extensive efforts have been placed towards using molecular tracers , typically cs and , to search for the presence of star - forming infall due to gravitational collapse ( see evans 1999 ; myers , evans , & ohashi 2000 and references therein ) .
the freeze - out of molecules onto grain surfaces significantly reduces the effectiveness of using molecular emission as a tracer of motions .
indeed the inside - out collapse models of shu ( 1977 ) predict that the highest infall speeds are found in the exact regions where molecules are expected to freeze - out . despite these arguments it is only recently that detections of gas phase freeze - out have become commonplace . until recently inherent difficulties in extracting molecular abundances from gas phase emission prevented definitive detections of gas - phase freeze - out .
previous studies assumed local thermodynamic equilibrium ( lte ) to estimate total column densities .
however , molecular emission from tracers with high dipole moments is far from lte .
in addition , secondary tracers such as co and its isotopic variants were previously used to estimate the total h@xmath1 column density , since it can not be observed directly .
it is now known that depletion of co in cold regions seriously hampers its utility as an estimator of the total hydrogen column density .
the recent advent of sensitive continuum and heterodyne arrays probing millimeter / sub - millimeter wavelengths , along with wide - field infrared imaging devices , has led to an explosion of clear detections of gas phase depletion / freeze - out ( bacmann et al .
2002 ; bergin et al .
2002 , tafalla et al .
2002 , hotzel et al .
2002 , caselli et al .
2002 , jessop & ward - thompson 2001 , kramer et al .
1999 , alves et al .
at this conference alone there are 8 contributions discussing gas - phase depletions ( carey et al . , di francesco et al . , feldman et al .
, savva et al . , kontinen et al . , lai et al .
, lee et al . , and peng et al .
, this volume ) .
observations of dust in emission or absorption provide direct knowledge of the dust distribution .
in addition , since the dust column density and mass is correlated with the h@xmath1 column density and mass ( hildebrand 1983 ; gordon 1995 ) , these observations provide the clearest information to date on the spatial distribution of h@xmath1 molecules .
furthermore , with some geometrical assumptions , the radial profiles of core density can be constructed ; these methods have greatly increased our knowledge of the physical conditions throughout the star formation process ( andr et al .
2000 , alves , lada , & lada 2001 ) . knowledge of the density and column density distribution of h@xmath1 also significantly aids the molecular observations in two ways . ( 1 ) by indirectly confirming the location of the h@xmath1 density and column density peak and , ( 2 ) by providing the density profile which helps to unravel the similar effects of density and abundance on excitation .
these advances have moved chemical analyses beyond surveys of objects that show evidence for complicated chemistry ( e.g. tmc-1 , l134n ) towards studies of more centrally concentrated objects that are closer to collapse and star formation . in this concise review
we outline how combined studies of dust emission / absorption with molecular emission have improved our ability to estimate molecular abundances and , in consequence , how our picture of dark cloud chemistry has been changed to one dominated by the effects of freeze - out . in the following ( 2 & 3 ) we briefly outline our current theoretical understanding of gas - grain interactions . furthermore we discuss key observational efforts that demonstrate the systematic nature of gas - phase freeze - out in condensed cores and how these observations can be qualitatively understood by a coupling of gas - grain chemical models to sophisticated radiative transfer models . in
4 we discuss how our improved understanding opens the door towards using astrochemistry as a new and potent tool to study the process of star formation .
the rate of deposition of a molecule in the gas phase onto a grain surface is @xmath2 where @xmath3 is the grain radius , @xmath4 is the mean thermal velocity , @xmath5 is the sticking coefficient , and @xmath6 the space density of grains .
if we use `` classical '' grains with @xmath7 1000 and @xmath6 = 1.3 @xmath8 10@xmath9n(@xmath10 ) @xmath11 ( spitzer 1978 ) , then the timescale for a molecule to freeze - out onto a grain surface is : @xmath12 where @xmath13 is the molecular weight and @xmath5 is the sticking coefficient .
co is a typical tracer of molecular cores which have typical densities of n(h@xmath1 ) @xmath14 @xmath11 ( evans 1999 ) . for a sticking coefficient of unity
the freeze - out timescale is @xmath15 year , well above the estimated time for thermal evaporation ( @xmath16 yr ) , but below estimated ages of @xmath17 myr ( palla and stahler 2000 ; hartmann 2001 ; lee and myers 1999 ) .
thus it is surprising that a gas - phase is observed at all , instead the near total freeze - out of the gas phase would be expected ( e.g. iglesias 1977 ) .
this points to the existence of some non - thermal mechanism to desorb or remove molecules from the grains and allow for the observed active gas phase chemistry .
various mechanisms have been identified , including spot heating due to cosmic - ray or x - ray impacts , chemical desorption , and others .
the reader is referred to williams ( 1993 ) for more detailed discussion .
the strength of the molecule bond to the grain surface is a key parameter that determines the efficiency of evaporation .
because low temperatures preclude the breaking of chemical bonds , molecules are not expected to be chemically bound to grain surfaces .
instead the approaching molecule has an induced dipole interaction with the grain surface or mantle and is bound through weaker van der waals - london interactions called physical adsorption ( kittel 1996 ) .
van der waals interactions are proportional to the product of the polarizabilities of the molecule and the nearby surface species ( @xmath18 ) . because of
this property some species are more tightly bound to the grain surface than others .
for example cs is expected to be more tightly bound to the grain surface than co or n@xmath1 .
in addition , theoretical analysis has suggested that n@xmath1 has a weaker bond to a water ice surface than co ( sadlej et al .
1995 ) .
models which have achieved the best success in reproducing observations are those that incorporate these assumptions regarding bond strengths , resulting in a desorption process that preferentially removes more weakly bound species ( bergin & langer 1997 ; charnley 1997 ; nejad & wagenblast 1999 ; aikawa et al .
2001 ; li et al .
2002 ) . as as sample in figure
1 we provide chemical abundances as a function of time for a cloud with a centrally concentrated density profile .
this result uses the model for b68 discussed by bergin et al .
( 2002 ) where the density profile is that for a bonnor - ebert sphere near - equilibrium that is kept constant with time as the chemistry evolves .
this is a simplification as more realistic models couple the physical and chemical evolution ( bergin & langer 1997 ; aikawa et al .
2001 ; li et al .
nonetheless it is illustrative of the observed effects .
figure 1 shows chemical abundances for species that are typically used to trace molecular clouds . at early times ( t = 2 @xmath19 yr ) the effects of freeze - out are evident , but the overall effect is small . at later times
( t = 7 @xmath19 yr ) strong gas - phase depletions are seen for nearly all species .
cs has the largest abundance decrease followed by other carbon - bearing species .
in contrast the nitrogen chemistry remains active at later times due to the presumed volatility of the n@xmath1 molecule ( see bergin & langer 1997 ) . one byproduct of co depletion is an increase in the dco@xmath20/hco@xmath20 ratio ( see aikawa et al . 2001 ; caselli et al 2002 ; caselli 2002 for further discussion ) . in the following section
we show that the pattern of differential depletions early cs depletion , followed by co , and finally n@xmath1
is observed in numerous sources .
similar effects would be observed in any model that incorporates these particular assumptions with regards to the gas - grain physics and evolves to a centrally concentrated state .
combined gas / dust investigations have resulted in the development of a new method to derive molecular abundances .
this method derives abundance profiles as a function of radius , as opposed to a line of sight average .
more specifically , the observed core density profile from the dust observations is used as an input to one - dimensional ( 1d ) radiative transfer codes ( monte - carlo is generally used ; e.g. bernes 1979 ; hogerheijde and van der tak 2000 ) .
the temperature profile is another parameter that is constrained by dust thermal emission or by multi - transitional studies of molecular emission .
variables are the abundance profile and velocity structure ( both bulk and turbulent ) . with these constraints , and iterating over the variables , the radiative transfer model predicts the radial profile of emission which can then be compared in detail to the observations .
this technique has been outlined by van dishoeck and hogerheijde ( 1999 ) and applications can be found in van der tak et al .
( 2000 ) , tafalla et al . (
2002 ) , bergin et al .
( 2002 ) , and lee et al .
( 2002 ) .
below we discuss some recent investigations that highlight : ( 1 ) the observed depletions are differential in nature and are systemic in the dense ism ( tafalla et al .
2002 ) and ( 2 ) current gas - grain chemical models are capable of qualitatively matching observations of a limited set of key species ( bergin et al .
2002 ) . in the top panels of figure 2 we show integrated molecular emission maps of the centrally concentrated molecular core l1517b for sample species with a map of the 1.2 mm continuum emission ( tafalla et al .
the notable feature in this figure is the lack of correlation between the 1.2 mm continuum a tracer of h@xmath1 with c@xmath21o and cs emission ; where the continuum peaks the molecular emission exhibit local minima .
in contrast , the n@xmath1h@xmath20 emission peak is coincident with the continuum emission maximum .
an analysis of these data using the physical description of the source given by the dust continuum emission is shown in the lower panels ( tafalla et al .
2002 ) . here
the radial profiles of emission ( solid squares ) are compared to model predictions .
models that assume constant abundance are shown as solid red lines ; these models predict excess emission at the core center for the co isotopes and cs . to reach agreement with observations the abundances of these species
must be sharply reduced in the densest regions of the core ( models shown as solid blue lines ) presumably through freeze - out .
constant abundance models provide a reasonable match for n@xmath1h@xmath20 , while requires a slight abundance increase towards the center .
tafalla et al . also compare the observations directly to the chemical models of bergin & langer ( 1997 ) , finding good agreement to models of fast - collapse .
most significantly similar results are found for all 5 cores examined in this work , as well as in sources observed by other studies ( kuiper , langer , & velusamy 1996 ; ohashi et al . 1999 ; lee et al .
2002 ; bergin et al .
this demonstrates the systematic nature of depletions they are observed in nearly all centrally concentrated low mass cores .
_ thus differential depletions in cold dark clouds must be a chemical signature of the earliest stages of star formation .
_ the dark cloud b68 is another example of a centrally concentrated object that exhibits emission features which are attributed to gas phase freeze - out . in this core
the h@xmath1 distribution is traced through a map of dust visual extinction ( a@xmath22 ) derived via near - infrared extinction mapping techniques ( e.g. alves , lada , and lada 2001 ) . in figure 3
we present the b68 a@xmath22 map along with a series of molecular emission maps . in this core
we see that the n@xmath1h@xmath20 emission peaks in a shell partially surrounding the peak of dust extinction . moreover , the n@xmath1h@xmath20 peaks inside the much larger c@xmath21o emission hole , which itself lies inside the cs emission depression .
analysis of these data differed from tafalla et al .
( 2002 ) in that a gas - grain chemical model is directly linked to a radiative transfer model predicting the profile of integration emission .
this allows for chemical considerations to determine how the abundance is structured with depth .
for example , photodissociation limits the abundance at cloud edges and the competition between depletion and desorption determines how the abundance declines with density .
figure 4 presents an analysis of the n@xmath1h@xmath20 and c@xmath21o emission ; in this case the open squares show the integrated emission as function of visual extinction .
this particular representation of the data is independent of cloud geometry , depending only on physical variables .
furthermore , if the emission is in lte and optically thin the integrated emission from a source with constant abundance would appear as a straight line in this diagram ( e.g. frerking et al .
1982 ; lada et al .
clearly the c@xmath21o - a@xmath22 relation deviates from a straight line at high extinctions .
such structure in this relation could be due to freeze - out or high opacities . to discriminate between these possibilities , observations of a lesser abundant isotope
are typically needed ( and in this case confirm freeze - out ) .
the dashed line is the `` best fit '' radial abundance profile predicted in the chemical model and the solid line is the resulting emission profile . for the observations
require significant freeze - out and also an abundance reduction at low extinction due to photodissociation . in the case of an abundance reduction at low extinction
is required , but only a small decrease in abundance at high a@xmath22 is needed to match observations . in figure 5 we present a more qualitative comparison of the structure predicted by the chemical models to the b68 data . here
the left - hand panel represents the peak contours predicted by a spherical radiative transfer model combined with the chemistry for transitions of cs , and .
the right - hand panel shows an identical representation of the observed emission in b68 .
although clear divergences from spherical symmetry are evident there exists striking agreement between model and observation .
the results presented here and elsewhere suggests that we are making progress in our understanding of the chemistry in these cold regions ( e.g. tafalla et al .
2002 ; aikawa et al . 2001 ; li et al . 2002
; lee et al .
the decrease in the abundance is also significant as this suggests that its parent molecule ( n@xmath1 ) is freezing onto grain surfaces .
if this result is observed beyond b68 it could have important implications for our ability to use molecules as the tracers of star formation dynamics .
the above results are excellent examples of the trends that are becoming apparent in chemical studies of low mass dark clouds .
these trends have suggested changes in the way molecular emission is used to probe star forming cores . 1 .
the use of c@xmath21o as a probe of the h@xmath1 column density in regions below the sublimation temperature of co ( @xmath23 k ) is no longer appropriate . whether this extends to warmer cores in giant molecular clouds
remains an open issue ( see mauersberger et al .
1992 ; chandler & carlstrom 1996 ; gibb & little 1998 ; savva et al . ,
this volume ) 2 . due to freeze - out the cs abundance can be sharply reduced when @xmath24 @xmath11 and is no longer a reliable tracer of motions in high density gas . however , this molecule is still a good probe of low density molecular gas prior to condensation and after star formation .
the emission from n@xmath1h@xmath25 and nh@xmath26 most closely resembles the maps of dust emission / extinction .
n@xmath1d@xmath20 , dco@xmath20 , and nh@xmath1d are also found to be well correlated with the dust ( caselli et al . 2002 ; shah and wooten 2001 ) .
these tracers are emerging as the best probes of the innermost regions of dense cores prior to the formation of a star .
once a star has been born the heating of the inner envelope returns some tracers to the gas phase .
this has not specifically been addressed in this paper but the reader is referred to recent work in this area including : hogerheijde et al .
( 1999 ) , maret et al .
( 2002 ) , jrgensen et al .
( 2002 ) , and references therein .
for a few key species we are now finding a convergence between observation and theory .
this opens the door towards using chemistry as a new tool to study the process of star formation .
below we discuss how this can be accomplished through both observational and theoretical efforts .
one new tool is the isolation of specific tracers that either do not freeze - out or freeze onto grain surfaces at late stages .
observations of these tracers ( e.g. ) over small or large scales can provide complementary kinematical information to continuum observations .
for instance , in regions that are currently forming stellar clusters , we can investigated the core - core velocity dispersion and dense core gas motions relative to the overall structure ( see walsh ; hogerheijde , this volume ) . on similar scales
combined studies of an `` early '' depleter ( e.g. cs ) and a high density tracer ( e.g. ) can also take advantage of the chemical evolution to detect cores prior to condensation ( cs ) , during condensed stages ( ) and after star formation ( see lee et al . 2002 ; lee , this volume ) .
the presence of differential depletions can certainly reduce the effectiveness of using molecular emission as a tracer of star formation kinematics , however , with careful choice of tracers one can use our expanded chemical knowledge to reconstruct the radial profile of kinematical motions .
indeed it is likely that _ the key to fully unraveling star formation dynamics lies in our improved understanding of the chemistry_. one example is in b68 where the emission lines of and are characterized by a systematic and well defined velocity gradient across the face of the cloud .
however , the magnitude and direction of the velocity gradient is different between the two tracers .
the chemical analysis presented in figure 4 shows that emission traces the innermost regions of the b68 core , while emission from predominantly traces the outer layers .
thus , if the gradient is due to rotation , then the combined chemical / dynamical analysis suggests that the core is differentially rotating ( lada et al .
2002 ) .
another example of the importance of chemistry to dynamics is shown in figure 8 .
the left - hand panel shows spectra of 3 molecules taken towards the same position in the l183 molecular core .
the double peaked structure observed in the optically thick cs and hco@xmath20 emission is now commonly used as a tracer of motions along the line of sight ( e.g. evans 1999 ) .
however , in l183 the cs profile is consistent with infalling motions , while hco@xmath20 is consistent with outflow .
since both of these transitions sample the same level of excitation this set of conflicting data must be due to chemistry .
in particular , the answer must lie in the differences in the abundance structure of each molecule , and changes in the velocity field , as a function of depth .
a qualitative example of how these tracers could sample different motions along the line of sight is shown in the right - hand panel of figure 8 .
this example uses the combined chemical / radiative transfer model constructed for b68 . in the plot
the x - axis is the radial position inside the cloud with a radius of 0.065 pc , and the y - axis shows the time evolution .
the lines delineate the @xmath27 surface of the line core emission .
all positions to the left of a given line for a particular transition have higher opacities . for example , for t
@xmath28 yr emission from is optically thin and traces the entire line of sight through the cloud . at later times
the abundance rises and the emission becomes thick and only probes a surface that is successively further from the core center .
when optically thick molecular transitions are used to probe motion via asymmetric self - reversals the emission preferentially probes the @xmath27 surface and hence the magnitude and direction of motions at that depth . _
this plot clearly demonstrates that as the chemistry evolves the various tracers are sensitive to different layers within the cloud .
_ for instance in this model cs ( j=2 - 1 ) and hco@xmath20 ( j=1 - 0 ) initially both probe the outer layers .
as the cloud evolves , cs differentially depletes and the @xmath27 surface retreats deeper into the cloud .
eventually , ( for t @xmath29 yr ) hco@xmath25 emission traces outer layers , while cs probes the inner layers .
this plot is not supplied as a template for the actual evolution of any particular core , rather it is shown as an example of how evolutionary changes in the chemical composition can affect observed emission , and more importantly the interpretation .
therefore , with observation of several species , and knowledge of the chemical abundance and physical profiles , another powerful new tool is the ability to search for structure in the radial velocity field , a key to understanding star formation .
our understanding of chemical theory in dark clouds also must evolve .
areas that clearly need further investigation include comparison of models to observations of similar cores ( l1544 , l1517b , b68 ) in species beyond the typical suite of tracers ( e.g. co , cs , ) .
for example lee et al .
( 2002 ) find that the abundance of ccs is depleted in the the l1512 and l1544 cores , but find evidence for a small abundance rise in the center .
this would be in agreement with the models of ruffle et al .
( 1997 ) that suggest a rebound in the abundances of complex molecules due to co freeze - out .
do other complex molecules provide similar results and can we further refine our understanding of gas phase pathways ?
this moves the testing of chemical models beyond comparison to abundances derived in template sources , such as tmc-1 and l134n , to observational comparisons with more centrally condensed objects that have well defined physical properties .
abundances derived in these sources still have uncertainties due to calibration , collision rates , and geometrical assumptions , but can generally be considered more reliable than simple line of sight averages . finally one challenge for future theory is to place the above observations in context of the non - detections of water vapor and molecular oxygen in cold cores ( snell et al .
2000 ; bergin & snell 2002 ) .
some efforts have begun to examine this question ( viti et al 2001 ; charnley et al 2001 ; roberts & herbst 2002 ) .
models that combine chemistry with dynamics have become increasingly more sophisticated .
the ultimate goal of these efforts is to use chemical observations to refine our understanding of star formation theory .
this goal is noteworthy and it remains to be seen whether the chemical unknowns regarding gas - grain interactions dominate ( e.g. sticking coefficients , binding energies , grain sub - structure , surface chemistry , exact desorption process ) .
combined models have recently begun to attempt to reproduce not only the pattern of depletions , but also the column densities observed in l1544 , a new template source ( aikawa et al .
2001 ; li et al .
this commendable step is needed to move theory beyond qualitative towards quantitative comparisons with observations .
however , in 3.2 we demonstrated that we now have the capability to place model predictions into radiative transfer models and predict direct observables .
this is important because column density estimates depend on assumptions regarding excitation that may not be reproduced in the model at the given time where model column densities best match observational estimates .
hence one could reproduce the column density and not the observed emission because of differences in the radial profile of physical conditions .
this task is time consuming , but sophisticated radiative transfer codes are now readily available ( see hogerheijde & van der tak 2002 ) .
the combination of chemical / dynamical models with radiative transfer truly offers the best opportunity to take chemical models and astrochemistry to the next level and potentially begin to test star formation theory .
the chemical processes that result in the formation of molecules play an important role in star and planetary formation through molecular contributions to the stability of the cloud against gravitational forces . given the widespread nature of gas - phase freeze - out it is becoming increasingly evident that knowledge of the chemistry is also required in order to fully characterize the star formation process .
the future of dark cloud chemical studies is promising as more complex and predictive models will be developed that can be compared to observations with increasing resolution and sensitivity .
this will lead to crucial tests of chemical theory but also presents a completely new approach to star formation studies , bringing astrochemistry into a wider realm as a potent tool .
we are grateful to i. bergin for proofing the manuscript and m. tafalla for providing comprehensive figures for this presentation .
we are also grateful for our collaboration with charlie lada , joo alves , and tracy huard which made the b68 analysis possible .
andr , p. , ward - thompson , d. , & barsony , m. 2000 , `` from prestellar cores to protostars : the initial conditions of star formation '' , protostars and planets iv , ( univ . of arizona press , tucson , 2000 ) ,
pp 59 van dishoeck , e. f. and hogerheijde , m. r. , `` models and observations of the chemistry near young stellar objects '' , the origin of stars and planetary systems , ( kluwer academic publishers , dordrecht , 1999 ) , edited by c. j. lada and n. d. kylafis , 97 | the past decade has led to significant improvements in our understanding of the physical structure of the molecular cores of cold dark clouds .
observational efforts , in combination with improved knowledge of cloud structure , now provide clear evidence that the chemistry of dark clouds is dominated by the depletion of gaseous species onto grain surfaces .
we outline the basis of these observational efforts and show how the abundance determinations have moved beyond single point analyses to the derivation of abundance profiles .
we discuss the basic physics of the interaction between molecules and grain surfaces and show that when physics is coupled into a chemical model there is excellent agreement , for a limited set of species , between theory and observations .
we discuss our improved understanding of cloud chemistry can be used as a new tool for studies of the formation of stars and planetary systems . |
the parton distribution functions ( pdf s ) of the nucleon are fundamental quantities associated with the strong interactions .
extensive experimental investigations have been undertaken during the past three decades to measure these distributions via deep - inelastic scattering ( dis ) of leptons from protons and light nuclei .
due to the intrinsically non - perturbative nature of the strong interactions in the low - momentum region , theoretical efforts to understand these distributions have had only limited success . with the ever increasing power of computers and significant developments in algorithms used to numerically simulate qcd on the lattice , it is hoped that properties of the pdf s can be determined from first principle @xcite at some point in the not so distant future ( for a review see ref .
it is the forward matrix elements of twist-2 operators that are computed numerically and these matrix elements are directly related to moments of the pdf s .
of course , only matrix elements computed with unquenched qcd with the physical values of the quark masses , @xmath1 , are to be directly compared with experimental data , but at this point in time such computations are not possible .
all present computations are performed with lattice quark masses , @xmath3 , that give a pion mass of @xmath4 , and most simulations are quenched . despite the fact that quenched computations require significantly less computer time , they can not , unfortunately , be connected to qcd in any way .
while solid progress is being made toward unquenched calculations @xcite of the moments of the pdf s , it is likely that partially - quenched @xcite computations will first provide a reliable connection between lattice computations and nature by allowing for calculations with smaller @xmath1 and thereby minimizing the impact of the @xmath1-extrapolation .
however , it is of interest , from a theoretical standpoint , to know the moments of the pdf s in quenched qcd ( qqcd ) . to extrapolate from @xmath3 down to @xmath1
, the @xmath1-dependence of the matrix elements is required @xcite , and recently chiral perturbation theory ( @xmath5pt ) has been used to determine the leading @xmath1-dependence in an expansion about the chiral limit in qcd @xcite and large-@xmath6 qcd @xcite .
@xmath5pt also allows for a connection to be made with existing convolution models of pdf s , and in addition , shows how to make them consistent with qcd @xcite . in this work
we determine the leading @xmath1-dependence of the matrix elements of isovector twist-2 operators about the chiral limit in qqcd using quenched chiral perturbation theory ( q@xmath5pt ) @xcite .
the lagrange density of qqcd is @xmath7_a^b\ q_b \ + \ \sum_{\tilde a,\tilde b=\tilde u , \tilde d,\tilde s } \overline{\tilde q}^{\tilde a } \ \left[\ i\dslash -m_{\tilde q}\ \right]_{\tilde a}^{\tilde b}\tilde q_{\tilde b } \nonumber\\ & = & \sum_{j , k = u , d , s,\tilde u , \tilde d,\tilde s } \overline{q}^j\ \left[\ i\dslash -m_{q}\
\right]_j^k\ q_k \ \ \ , \label{eq : qqcd}\end{aligned}\ ] ] where @xmath8 are the three light - quarks , @xmath9 , @xmath10 , and @xmath11 , and @xmath12 are three light bosonic quarks @xmath13 , @xmath14 , and @xmath15 .
the super - quark field , @xmath16 , is a six - component column vector with the three light - quarks , @xmath9 , @xmath10 , and @xmath11 , in the upper three entries and the three ghost - light - quarks , @xmath13 , @xmath14 , and @xmath15 , in the lower three entries .
the graded equal - time commutation relations for two fields is @xmath17 where @xmath18 are spin - indices and @xmath19 are flavor indices .
the objects @xmath20 correspond to the parity of the component of @xmath21 , with @xmath22 for @xmath23 and @xmath24 for @xmath25 .
the diagonal super mass - matrix , @xmath26 , has entries @xmath27 , i.e. @xmath28 , @xmath29 and @xmath30 , so that the contribution to the determinant in the path integral from the @xmath8 s and the @xmath12 s exactly cancel . in the absence of quark masses , the lagrange density in eq .
( [ eq : qqcd ] ) has a graded symmetry @xmath31 , where the left- and right - handed quark fields transform as @xmath32 and @xmath33 respectively .
however , the functional integral associated with this lagrange density does not converge unless the transformations on the left- and right - handed fields are related , @xmath34 , where @xmath35 denotes a superdeterminant @xcite , leaving the theory to have a symmetry @xmath36\times u(1)_v$ ] , where the `` @xmath37 '' denotes a semi - direct product as opposed to a direct product , `` @xmath38 '' .
it is assumed that this symmetry is spontaneously broken @xmath36\times u(1)_v\rightarrow su(3|3)_v\times u(1)_v$ ] so that an identification with qcd can be made .
the pseudo - goldstone bosons of qqcd form a @xmath39 matrix , @xmath40 , that can be written in block form @xmath41 where @xmath42 is the @xmath43 matrix of pseudo - goldstone bosons including the @xmath44 with quantum numbers of @xmath45 pairs , @xmath46 is a @xmath43 matrix of pseudo - goldstone bosons including the @xmath47 with quantum numbers of @xmath48 pairs , and @xmath5 is a @xmath43 matrix of pseudo - goldstone fermions with quantum numbers of @xmath49 pairs , @xmath50 as the object @xmath51 is invariant under @xmath36\times u(1)_v$ ] the most general lagrange density that describes low - momentum dynamics will contain arbitrary functions of @xmath52 @xcite . at lowest order in the chiral expansion , the lagrange density that describes the dynamics of the pseudo - goldstone bosons is , using the notation of ref .
@xcite , @xmath53 \ +
\ \lambda\ { \rm str } \left[\ m_q\sigma\ + \ m_q^\dagger\sigma^\dagger\ \right ] \ +
\ \alpha_\phi \ \partial^\mu\phi_0\partial_\mu\phi_0 \ -\
m_0 ^ 2\ \phi_0 ^ 2 \ \ \ , \label{eq : lagpi}\end{aligned}\ ] ] where the parameter @xmath54 is chosen to reproduce the meson masses , and @xmath55 is the exponential of the @xmath40 field , @xmath56 with this normalization , @xmath57 in qcd .
in addition , it is understood that the operators with coefficients @xmath58 and @xmath59 , the hairpin interactions , are inserted perturbatively .
expanding out the lagrange density in eq .
( [ eq : lagpi ] ) to quadratic order in the meson fields , one finds relations between the meson masses in the isospin limit , @xmath60 the lagrange density in eq .
( [ eq : lagpi ] ) has been used to compute several observables in the meson sector , such as @xmath61 , @xmath62 @xcite and the meson masses , to one - loop in perturbation theory @xcite ( for a review see ref .
@xcite ) .
the inclusion of the lowest - lying baryons , the octet of spin-@xmath63 baryons and the decuplet of spin-@xmath64 baryon resonances , is detailed in ref .
an interpolating field that has non - zero overlap with the baryon octet ( when the @xmath65 indices are restricted to @xmath66 ) is @xcite @xmath67 \epsilon_{abc } \left(c\gamma_5\right)_{\alpha\beta } \ \ \ , \label{eq : octinter}\end{aligned}\ ] ] where @xmath68 is the charge conjugation operator , @xmath69 are color indices and @xmath70 are spin indices . dropping the spin index , one finds that under the interchange of flavor indices @xcite , @xmath71 the object @xmath72 describes a * 70 * dimensional representation of @xmath73 @xcite .
it is convenient to decompose the irreducible representations of @xmath73 into irreducible representations @xmath74 @xcite .
the subscript denotes where the `` @xmath75 '' acts , either on the @xmath8 s or on the @xmath12 s .
the ground floor of the * 70 * -dimensional representation contains baryons that are comprised of three quarks , @xmath76 , and transforms as an @xmath77 of @xmath78 .
the octet baryons are embedded as @xmath79 where the indices are restricted to take the values @xmath80 only .
the octet baryon matrix is @xmath81 the first floor of the * 70 * -dimensional representation contains baryons that are composed of two quarks and one ghost - quark , @xmath82 , and transforms as an @xmath83 of @xmath78 . the tensor representation @xmath84 of the @xmath85 multiplet , where @xmath86 runs over the @xmath12 indices and @xmath87 run over the @xmath8 indices , has baryon assignment @xmath88 the right superscript denotes the third component of q - isospin , while the left subscript denotes the @xmath12 flavor . the tensor representation @xmath89 of the @xmath90 multiplet , where @xmath86 runs over the @xmath12 indices and @xmath91 run over the @xmath8 indices , has baryon assignment @xmath92 the @xmath84 and the @xmath89
are uniquely embedded into @xmath72 ( up to field redefinitions ) , constrained by the relations in eq .
( [ eq : bianchi ] ) : @xmath93 as we are only interested in one - loop contributions to observables with @xmath76-baryons in the asymptotic states , we do not explicitly show the second and third floors of the * 70*. an interpolating field that contains the spin-@xmath64 decuplet as the ground floor is @xcite @xmath94 \varepsilon_{abc } ( c\gamma^\mu)_{\beta\gamma } \ \ \ , \label{eq : tdef}\end{aligned}\ ] ] where the indices @xmath95 run from @xmath96 to @xmath97 . neglecting spin indices
, one finds that under the interchange of flavor indices @xcite @xmath98 @xmath99 describes a @xmath100 dimensional representation of @xmath73 , which has a ground floor transforming as @xmath101 under @xmath102 with @xmath103 where the indices are restricted to take the values @xmath80 , and where @xmath104 is the totally symmetric tensor containing the decuplet of baryon resonances , @xmath105 the first floor transforms as a @xmath85 under @xmath102 and has a tensor representation , @xmath106 , with baryon assignment @xmath107 the embedding of @xmath106 into @xmath99 is unique ( up to field redefinitions ) , constrained by the symmetry properties in eq .
( [ eq : ttrans ] ) : @xmath108 the free lagrange density for the @xmath72 and @xmath99 fields is @xcite , at leading order in the heavy baryon expansion @xcite , @xmath109 where @xmath110 , @xmath111 is the decuplet - octet mass splitting , @xmath112 is the baryon four - velocity , and @xmath113 .
the brackets , @xmath114 denote contraction of lorentz and flavor indices as defined in refs . @xcite .
the lagrange density describing the interactions of the baryons with the pseudo - goldstone bosons is @xcite @xmath115 \ + \
2\gamma^\prime \left(\overline{\cal t}^\nu s^\mu { \cal t}_\nu\right ) \ { \rm str}\left(a_\mu \right ) \ ,
\label{eq : ints}\end{aligned}\ ] ] where @xmath116 is the covariant spin - vector @xcite . restricting oneself to the @xmath76 sector , it is straightforward to show that in qcd @xmath117 where @xmath118 and @xmath119 are constants that multiply the @xmath120 invariants that are commonly used in qcd , but it should be stressed that the @xmath119 and @xmath118 in qqcd will not have the numerical values of those of qcd . in qqcd
there is an additional coupling that must be considered , @xmath121 , a hairpin interaction @xcite , that is usually not considered in the @xmath5pt description of low - energy qcd . in our calculation of the matrix elements of twist-2 operators
we will replace @xmath122 and @xmath123 with @xmath119 and @xmath118 , but we will keep @xmath121 explicit . in the above discussion , vector and axial - vector meson fields
have been introduced in direct analogy with qcd .
the covariant derivative acting on either the @xmath124 or @xmath125 fields has the form @xmath126 where the vector and axial - vector meson fields are @xmath127
in qcd , the nonsinglet twist-2 operators have the form , @xmath128 where the @xmath129 denotes symmetrization on all lorentz indices , and where @xmath130 are gell - mann matrices acting in flavor - space .
the @xmath131 transform as @xmath132 under @xmath133 chiral transformations @xcite .
of particular interest to us are the isovector operators where @xmath134 . in the isospin limit , there are no contributions to the matrix elements of @xmath135 from disconnected diagrams . in qqcd
the nonsinglet twist-2 operators have the form @xmath136 where the @xmath137 are super gell - mann matrices , and we are interested in the isovector matrix elements with @xmath138 ( for a discussion about the extension of electro - weak operators contributing to nonleptonic decays in qqcd see refs .
@xcite ) . for arbitrary @xmath139 and @xmath140
there will be contributions from disconnected ghost quark diagrams .
however , for @xmath141 there will be no contributions from disconnected diagrams involving quarks or ghost - quarks in the isospin limit . in this work we have chosen to examine the case where @xmath142 . ] .
the operators @xmath143 transform as @xmath144 under @xmath145 graded chiral transformations . at leading order in the chiral expansion
, matrix elements of the isovector operator @xmath143 between meson states are reproduced by operators of the form @xcite @xmath146 \nonumber\\ & & \ -\ { \rm traces } \ \ \ , \label{eq : mesontree}\end{aligned}\ ] ] where @xmath147 is the scale of chiral symmetry breaking .
operators involving more derivatives on the meson fields or more insertions of @xmath1 make contributions to the matrix elements that are higher order in the chiral expansion @xcite .
= 2.5 in 0.15 in .2 in in qcd , the one - loop diagrams shown in fig .
[ fig : pidiags ] with only the octet mesons in the loop give rise to the leading non - analytic terms in @xmath1 , of the form @xmath0 .
such diagrams have the potential to give analogous contributions to the matrix elements in qqcd .
however , explicit computation proves that such non - analytic contributions to the matrix element of @xmath143 resulting from the diagrams shown in fig .
[ fig : pidiags ] vanish , due to a cancellation between the `` @xmath42''-loops and `` @xmath5''-loops .
therefore , at one - loop order there are no nonanalytic contributions and the matrix element of @xmath143 in the pion is @xmath148 for n - odd , between an initial pion with isospin index @xmath122 and momentum @xmath149 , and a final pion with isospin index @xmath123 .
the matrix elements vanish for n - even and the ellipses denote terms that are analytic in @xmath1 , or are higher order in the chiral expansion . at leading order in the chiral expansion ,
the matrix elements of @xmath143 are described by @xmath150 where the @xmath151 brackets denote complete symmetrization of the enclosed indices , @xmath112 is the four - velocity of the heavy baryon , and @xmath152 in complete analogy with the work of refs .
the index contractions denoted by @xmath153 are the same as those in eq .
( [ eq : ints ] ) defined in refs . @xcite . in general , the coefficients @xmath154 and @xmath155 are not constrained by symmetries and must be determined from elsewhere .
however for @xmath156 they are fixed by the isospin charge of the baryons to be @xmath157 an expression analogous to eq .
( [ eq : baryonops ] ) exists in qcd , and can be obtained from eq .
( [ eq : baryonops ] ) by restricting the flavor - indices in each sum to take values @xmath158 only .
however , one must bear in mind that , in general , the values for @xmath154 and @xmath155 in qqcd will differ from those in qcd for @xmath159 . at
next - to - leading order ( nlo ) there are tree - level contributions from local counterterms involving one insertion of the quark mass matrix . in qcd
we write the operator set as @xmath160 , m_{q\xi+}\right]\right ] \ + \
b_2^{(n)}\ { \rm tr}\left[\ \overline{b } \ { \left[\lambda^3_{\xi+ } , b\right ] , m_{q\xi+ } \}\right ] \right.\nonumber\\ & & \left .
+ \ b_3^{(n)}\ { \rm tr}\left[\ \overline{b } \left[\ { \lambda^3_{\xi+ } , b\ } , m_{q\xi+}\right]\right ] \ + \
b_4^{(n)}\ { \rm tr}\left[\ \overline{b } \{\
{ \lambda^3_{\xi+ } , b\ } , m_{q\xi+}\}\right ] \right.\nonumber\\ & & \left . + \
b_5^{(n)}\ { \rm tr}\left[\ \overline{b } b \
\right]\ { \rm tr}\left[\ \lambda^3_{\xi+ } m_{q\xi+}\ \right ] \
+ \ b_6^{(n)}\ { \rm tr}\left[\ \overline{b } b \lambda^3_{\xi+ } \
\right]\ { \rm tr}\left[\ m_{q\xi+}\ \right ] \right.\nonumber\\ & & \left .
b_7^{(n)}\ { \rm tr}\left[\ \overline{b } \lambda^3_{\xi+ } b\ \right]\ { \rm tr}\left[\ m_{q\xi+}\ \right ] \
\right ) v_{\mu_1} ... v_{\mu_n}\ \ -\ { \rm traces } \ \ \ , \end{aligned}\ ] ] where @xmath161 and @xmath162
. the @xmath163 must be determined , however , isovector charge conservation ensures that @xmath164 .
the explicit renormalization scale dependence of the matrix element at each order in the chiral expansion requires that , in general , the @xmath163 have logarithmic dependence upon the renormalization scale , @xmath165 . in qqcd
the analogous operator set takes the form @xcite @xmath166\ v_{\mu_1 } v_{\mu_2} ... v_{\mu_n}\ \ -\ { \rm traces } \
, \label{eq : tcts}\end{aligned}\ ] ] where the @xmath163 of qqcd will , in general , differ from those of qcd .
isovector charge conservation forces @xmath164 . at tree - level
the matrix elements in an octet baryon `` @xmath167 '' is @xmath168 where the clebsch - gordan coefficients , @xmath169 and @xmath170 , are given in tables [ tab : tree ] , [ tab : mqqcd ] and [ tab : mqqqcd ] .
parametrically , but not numerically , the leading order matrix elements in qcd and qqcd are the same at tree - level . .
the leading order tree - level matrix elements of the isovector twist-2 operators between octet baryon states .
the constants @xmath171 and @xmath172 are defined in eq .
( [ eq : baryonops ] ) . [ cols="^,^,^,^",options="header " , ] further , the hairpin diagrams do not contribute to the matrix elements in the nucleon in the limit of exact isospin symmetry , as the possible intermediate states in @xmath99 are inaccessible . our motivation for undertaking this work is to construct a framework in which experimental data can be compared with the predictions of lattice qcd . as data exists only for the nucleon , it is natural to ask why we bothered with the other members of the baryon octet .
as the present values of @xmath173 are large , and in fact @xmath174 , we consider three light quarks as opposed to just two .
thus the quantities @xmath175 , @xmath155 and combinations of the @xmath163 must all be determined , in addition to the strong interaction couplings @xmath176 and @xmath177 in order to perform the quark mass extrapolation at one - loop order . for a one - loop analysis as we have presented here , the coefficients @xmath178 and @xmath155 can be determined at tree - level from calculations of matrix elements in the baryon decuplet , and it is consistent to use these values in the one - loop expression for the octet baryons .
calculation of the proton , @xmath179 and @xmath180 matrix elements will allow for a determination of @xmath171 , @xmath172 , and combinations of the @xmath163 , and hence the @xmath1-dependence of the nucleon matrix element . while the hairpin interactions , @xmath59 and @xmath58 , do not contribute directly to the matrix elements in the nucleon , they do contribute to the matrix elements in the other octet baryons , and so will indirectly influence the calculation of the matrix elements in the nucleon . to get a feel for the @xmath1-dependence of the matrix elements in the proton , we pick some values for the parameters describing the proton matrix element and vary the pion mass . for the qcd matrix element shown in fig .
[ fig : t2 ] , we choose @xmath181 , and @xmath182 , @xmath183 and @xmath184 for @xmath159 ( this does not eliminate contributions from the @xmath99 intermediate states , as they contribute in wavefunction renormalization ) . also , we use the tree - level qcd values for the axial coupling constants @xmath185 , @xmath186 @xcite and @xmath187 @xcite , and the measured value of the decuplet - octet mass splitting , @xmath188 .
for the qqcd matrix elements shown in fig .
[ fig : t2 ] , we choose three different parameter sets , set ( a ) , set ( b ) and set ( c ) , that reproduce the qcd matrix element at @xmath189 .
set ( a ) has @xmath181 , @xmath190 , @xmath183 and @xmath184 for the tree - level matrix elements , with @xmath185 , @xmath186 , @xmath187 , and @xmath188 .
set ( b ) has @xmath181 , @xmath191 , @xmath183 and @xmath184 for the tree - level matrix elements , with @xmath185 , @xmath186 , @xmath187 , and @xmath192 is @xmath193 @xcite . ] .
in contrast , set ( c ) has @xmath181 , @xmath194 , @xmath183 and @xmath184 for the tree - level matrix elements , with @xmath195 , @xmath196 , @xmath197 , and @xmath192 .
= 4.0 in 0.15 in .2 in
it is important to know the parton distribution functions of the nucleon in qqcd . in order to obtain these predictions from the lattice computations of the foreseeable future
, extrapolations in @xmath1 from the lattice values down to the physical values will still be required .
we have taken the first step toward establishing the theoretical @xmath1-dependence by computing the leading @xmath1-dependence of the forward matrix element of the isovector twist-2 operators in the octet baryons in qqcd using q@xmath5pt .
it is certainly true that higher order calculations are required , not only to establish the convergence of the expansion about the chiral limit , but also to allow an extrapolation to the relatively large values of the lattice quark masses . even in qcd ,
the convergence of @xmath75 chiral perturbation theory varies from observable to observable due to the large value of the kaon mass .
recently , it has been observed that baryon masses obtained from qqcd simulations by applying the qcd extrapolation agree reasonably well with qcd simulations of the masses @xcite .
it will be a pleasant surprise if this empirical feature is found to hold for other observables , and particularly the matrix elements of the twist-2 operators .
we look forward to a complete analysis of the quenched lattice data , including an extraction of the constants that describe these matrix elements and eventual predictions at the physical value of @xmath198 .
g. martinelli and c.t .
sachrajda , _ phys .
* b 196 * , 184 ( 1987 ) ; _ nucl .
phys . _ * b306 * , 865 ( 1988 ) ; m. gckeler _ et al .
_ , _ phys .
* d 53 * , 2317 ( 1996 ) ; m. gckeler _ et al .
_ , _ nucl .
suppl . _ * 53 * , 81 ( 1997 ) ; c. best _ et al .
_ , hep - ph/9706502 .
s. r. sharpe and n. shoresh , hep - lat/0108003 ; hep - lat/0011089 ; _ phys .
rev . _ * d62 * , 094503 ( 2000 ) ; _ nucl .
_ * 83 * , 968 ( 2000 ) ; s. r. sharpe , _ phys .
rev . _ * d56 * , 7052 ( 1997 ) ; c. w. bernard and m. f. l. golterman , _ phys .
rev . _ * d49 * , 486 ( 1994 ) .
a. w. thomas , w. melnitchouk and f. m. steffens , _ phys .
* 85 * , 2892 ( 2000 ) ; w. detmold , w. melnitchouk , j. w. negele , d. b. renner and a. w. thomas , hep - lat/0103006 ; w. detmold , d. b. leinweber , w. melnitchouk , a. w. thomas , and s. v. wright , nucl - th/0104043 . for a recent review see u .-
meiner , essay for the festschrift in honor of boris ioffe , in `` encyclopedia of analytic qcd '' , edited by m. shifman , to be published by world scientific .
hep - ph/0007092 . | we compute the leading quark mass dependence of the matrix elements of isovector twist-2 operators between octet baryon states in quenched qcd using quenched chiral perturbation theory .
there are contributions of the form @xmath0 and @xmath1 , in analogy with qcd , but there are also contributions of the form @xmath2 from hairpin interactions
. the nucleon does not receive such hairpin contributions .
# 1#2#3#4#1 * # 2 * , # 3 ( # 4 ) |
currently popular theories of structure formation in the universe postulate that structure grows through the amplification by gravitational forces of initially small density fluctuations ( see white 1996 for a review ) . depending on the characteristics of the spectrum of fluctuations , small objects can be the first to collapse ; they then merge to form progressively larger systems giving rise to the complex structure we observe today . in this hierarchical structure formation scenario our galaxy , as a typical galaxy , should also have been formed in part by merging and accretion of smaller galaxies , or ` building blocks ' .
these events should be imprinted in some of its present - day components , presumably as residual structure .
for example , when a galaxy is disrupted it leaves trails of stars along its orbit .
these could be superposed in a spheroidal component such as a stellar halo .
in fact , numerous observations suggest substructure in the halo of the galaxy ( see majewski in this volume ) . in this paper
, we attempt to describe what the signatures of different accretion events should be if indeed our galaxy formed as envisaged in current theories . should this merging history be observed in star counts , kinematic or abundance surveys of the galaxy ? how prominent or not would these substructures be ?
how well - mixed are the stars that made up these progenitors ?
what can we say about the properties of the accreted satellites from the observations we have today ?
to tackle the questions we just posed we carry out n - body simulations of accretion of satellite galaxies , where we represent the milky way by a fixed , rigid potential and the satellite by a collection of particles .
the self - gravity of the satellite is modelled by a monopole term as in white ( 1983 ) .
the galactic potential is represented by two components : a disk described by a miyamoto - nagai potential , @xmath0 where @xmath1 , @xmath2 , @xmath3 , and a dark halo with a logarithmic potential , @xmath4 with @xmath5 and @xmath6 .
the initial density distribution of the satellite is given by a plummer profile @xmath7 with @xmath8 , @xmath9 being the initial mass of the satellite and @xmath10 its scale length .
its one - dimensional internal velocity dispersion is @xmath11 .
we run several simulations which differ in their orbital parameters , which span a range in radial periods from 0.5 to 1.3 gyr , and have an apocentre to pericentric distance ratio of 5 to 10 ( fairly radial orbits ) .
we have also imposed that the orbits pass close to the solar circle to compare the results of the experiments with the known properties of the stellar halo of the milky way . in all cases
the satellite was represented by @xmath12 particles of equal mass .
we find that the satellites become completely unbound after , at most , three pericentric passages .
one process that will tend to erase any macroscopic correlation between the particles , making more difficult the detection of satellite debris , is phase - mixing . to quantify it we use the coarse - grained entropy defined as : @xmath13 = -\int \bar{f}\ln
\bar{f } d^3x d^3v,\ ] ] where @xmath14 is the coarse - grained distribution function , that is , the average of the actual distribution function @xmath15 over small cells in phase - space .
one of the interesting properties of @xmath14 is that it decreases as the system becomes phase - mixed .
therefore , @xmath16 $ ] is expected to increase with time , as shown for our simulations in figure 1 . in practice
, we replace the integral by a sum over cells , and @xmath14 by the fraction of particles in each cell .
the spatial properties of the debris can be studied by plotting isodensity surfaces .
these surfaces are indicative of how spread in its available configuration volume the system is , or equivalently how advanced the disruption is .
we find that for the region of parameter space probed , these volumes are almost completely filled after a hubble time . in terms of the spatial distribution of the particles on the plane of the sky ( see helmi , zhao and de zeeuw , this volume , their figure 1 ) we do not find any strong correlations , contrary to what johnston , hernquist and bolte ( 1996 ) find in their simulations of accretion in the outer halo
. we can understand this in terms of the short time scales and the strong flattening characteristic of the inner parts of our galaxy .
the maximum densities of such debris are three to four orders of magnitude lower than the initial density of their progenitors , roughly comparable to the local density of the stellar halo . in order to reflect what observers can do in surveys of the local halo , in figure 2
we plot the kinematical properties of stars inside a box of 3 kpc on a side in different locations along the orbit .
notice the strong correlations between the different components of the velocity vector inside any given box , and , in particular , the large velocity range in each component when close to the galactic centre .
this shows that the debris can appear kinematically hot .
this is the result of a combination of multiple streams within a given box ( clearly visible in figure 2 ) and strong gradients along each stream . at a given point along any particular stream
the dispersions are usually very small .
[ fig2 ] because the disruption of the satellite occurs very early in its history , it can be considered as an ensemble of test particles during most of its evolution .
one of the distinguishing properties of this ensemble is that it initially had a very high density in phase - space , and by virtue of liouville s theorem , this is true at all times . at later times , however , this is no longer reflected by a strong concentration in configuration space . since the behaviour of a dynamical system is particularly simple in action - angle variables : @xmath17 we can find the evolution of the distribution function in the following way .
we know the distribution function at the initial time @xmath18 . by a transformation of coordinates
we may write it as a function of action - angle variables at that initial time : @xmath19 . eqs .
( [ eq : evol ] ) then give the distribution function at a later time @xmath20 in action - angle variables . if we now transform locally from @xmath21 back to @xmath22 , we obtain the distribution function , and thereby the velocity dispersions and the density behaviour , in the region of interest . in action - angle variables the system expands along three directions and contracts along the remaining three .
this is directly associated with the equations of motion .
any initial dispersion in the angles can only increase ( eqs .
( [ eq : evol ] ) ) , so that the system becomes a very elongated ellipsoid in phase - space as time passes by .
the conservation of phase - space density ( liouville s theorem ) then forces the other directions to shrink .
this is reflected in the projection onto observable space .
it is possible to show ( see helmi & white 1998 for all the details ) that the velocity dispersions decrease on the average with time , and so the volume density also decreases with time as the system expands in the spatial directions .
for example in the spherical case the dispersions and the central density behave as @xmath23 the velocity dispersion in the direction transverse to the plane of motion ( @xmath24 ) is constant , except for periodic variations due to the orbital phase . in the axisymmetric case
, there no longer is a preferred orientation , so that also in the @xmath24-direction the velocity dispersion decreases , and therefore the density decreases as @xmath25 .
our results suggest that fossil structure from an accretion event which took place at any time during the the history of the milky way should be visible in velocity space .
a stream of stars is , most of the time , fairly cold .
however , in particular in the inner parts of the galaxy , it is quite likely to find more than one stream from any particular disrupted satellite in a small region in configuration space . in that case
, the apparent velocity dispersions can be much larger , although constrained by the initial dispersions in the integrals of motion .
majewski , munn & hawley ( 1994 ) reported the discovery of a moving group near the ngp ( for details see majewski , this volume ) with large velocity dispersions ( greater than 30 @xmath26 ) , and with a mean motion very different from that of the other stars in the field .
if we are to take these stars as satellite debris , we will have to invoke a multistream structure in order to explain their kinematics .
indeed , there is some evidence of substructure in their distribution of angular momenta .
this is to be expected if indeed there are multiple streams . with this in mind
, we can use our simulations and analytic results to estimate the mass of the progenitor using the simulations , and estimates for its initial size and velocity dispersions .
we find : @xmath27 , @xmath28 and @xmath29 .
helmi , a. & white , s.d.m .
1998 , in preparation helmi , a. , zhao , h.s . &
de zeeuw , p.t . 1998 , this volume johnston , k.v . , hernquist , l. & bolte , m. 1996 , , 465 , 278 majewski , s.r . ,
munn , j.a .
& hawley , s.l .
1994 , , 427 , l37 majewski , s.r .
1998 , this volume white , s.d.m . 1983 , , 274 , 53 white , s.d.m .
1996 , in cosmology and large scale structure , les houches , session lx .
amsterdam : elsevier , p. 349 | we study numerical simulations of satellite galaxy disruption in a potential resembling that of the milky way .
our goal is to assess whether a merger origin for the stellar halo would leave observable fossil structure in the phase - space distribution of nearby stars .
we show how mixing of disrupted satellites can be quantified using a coarse - grained entropy .
although after 10 billion years few obvious asymmetries remain in the distribution of particles in configuration space , strong correlations are still present in velocity space .
we briefly describe how we can understand these effects based on the conservation of fine - grained phase - space density in an action - angle formalism .
we also discuss the implications of our results on the known properties of the stellar halo . |
the signals received by the antennas obey the stationary stochastic process and then ergodic process .
the ergodic theory can be applied to the auto - correlation function for a spectrometer and the cross - correlation function for radio interferometer . under such conditions , weinreb ( 1963 )
developed the first digital spectrometer .
this digital spectrometer is called the xf correlator in which the correlation is calculated before fourier transform . meanwhile , chikada et al .
( 1987 ) developed the first the fx correlator of an another design , in which fourier transform is performed before cross multiplication .
although there is a difference of property between two basic designs , the obtained astronomical spectra of them were confirmed to be identical .
determining the number of correlation lags in the xf scheme or of fourier transform points in the fx scheme is essential for the realization of high - dispersion and wideband observation , because the frequency resolution is derived as @xmath0 where @xmath1 is the sampling period , @xmath2 is the number of correlation lags or fourier transform points , and the bandwidth of b is equal to @xmath3 .
the material size and cost of the correlator strongly depend on the sampling period , @xmath1 , and the number of correlation lags or fourier transform points , @xmath2 . the new xf architecture with the digital tunable filter bank that is designed with the finite impulse response ( fir ) has been proposed and developed for the next generation radio interferometers , the expanded very large array ( evla ) and the atacama large millimeter / submillimeter array ( alma ) ( @xcite , @xcite ) .
this is called the `` fxf correlator '' .
the architecture of the fxf scheme can make the material size smaller in comparison with that of the conventional xf scheme . since the digital filter allows a variety of observation modes [ scientific and observational availability
were shown in iguchi et al .
( 2004 ) ] , the fxf scheme will provide us with the most appropriate specifications which meet the scientific requirements .
this will lower the risk of over - engineering of the correlator .
the improved fx architecture with dft filterbank was developed by bunton ( 2000 ) . the use of polyphase filter banks allows arbitrary filter responses to be implemented in the fx scheme ( bunton 2003 ) .
this is called the `` polyphase fx correlator '' .
this scheme has a possibility to achieve the spectral leakage of about -120 db . in particular , this performance is significant to suppress the leakage from the spurious lines mixed in receiving , down - converting or digitizing . the ffx correlator is a new algorithm for correlation process in radio astronomy .
the ffx scheme consists of 2-stage fourier transform blocks , which perform the 1st - stage fourier transform as a digital filter , and the 2nd - stage fourier transform to achieve higher dispersion .
the first f of the ffx is the initial letter of the word `` filter '' . in this paper
, we present a new ffx architecture .
the principle of the ffx scheme in section 2 , the properties of the ffx scheme in section 3 , the algorithm verification and performance evaluation with the developed ffx correlator in sections 4 and 5 , and the summary of this paper in section 6 are presented .
this section shows the algorithm and the data flow diagram of the signal processing in the fourier transform of the ffx scheme ( see figure [ fig : ffx ] ) .
suppose that @xmath4 are the digital waveforms at the correlator input from the astronomical radio signals that are received by the telescope .
the inputs , @xmath4 , are real digital signals at sampling period of @xmath1 , and obey the zero - mean gaussian random variable .
the suffix @xmath5 is an integer for time .
fig1 ( 160mm,200mm)fig1.eps [ step 1 ] the correlator receives the time - domain digital sampling signals from the analog - to - digital converter ( adc ) , and accumulate them up to @xmath6 points .
[ step 2 ] the time - domain @xmath6-point data are transferred to the frequency - domain by using the @xmath6-point discrete complex fourier transform as follows : @xmath7 where @xmath8 is the spectrum after the 1st fourier transform , the suffix @xmath9 is an integer for frequency , and @xmath1 is equal to @xmath10 at the bandwidth of @xmath11 .
the @xmath12 is the minimum frequency resolution of the 1st fourier transform , which is equal to @xmath13 .
[ step 3 ] the extraction of the @xmath14 points from the frequency domain @xmath15-point data after the 1st fourier transform is conducted as if filter and frequency conversion are performed simultaneously : @xmath16 where @xmath17 is the minimum frequency channel in the extraction , and the suffix @xmath18 is an integer for frequency .
[ step 4 ] the @xmath14-point data after inverse fourier transform is written by @xmath19,\ ] ] where @xmath20 is the time - domain signal after inverse fourier transform , the suffix @xmath21 is an integer for time , and @xmath22 is the sampling period after filtering at the bandwidth of @xmath23 @xmath24 .
[ step 5 ] by repeating the procedure from step 1 to step 4 , the data are gathered up to @xmath25 points as follows ; @xmath26 where @xmath27 is @xmath28 , and @xmath29 is the number of repeating times of the procedure from step 1 to step 4 .
[ step 6 ] the time - domain @xmath25-point data after gathering are transferred to the frequency - domain by using the @xmath25-point discrete complex fourier transform as follows : @xmath30 where @xmath31 is the spectrum after the 2nd fourier transform , and the suffix @xmath32 is an integer for frequency .
the @xmath33 is the minimum frequency resolution after the 2nd fourier transform , which is equal to @xmath34 ( = @xmath35 ) .
.definition of functions . [ cols="<,<",options="header " , ] [ table : fxopmode ] llllll stage & bandwidth & spectral & spectral & velocity & correlation + & & points & resolution & resolution & + & & & & at 1 mm & + 1@xmath36 & 4096 mhz & 512 & 8 mhz & 8.0 km s@xmath37 & 2ac1cc + 2@xmath38 & 64 mhz & 2048 & 31.25 khz & 0.031 km s@xmath37 & 2ac1cc + + 1@xmath36 & 4096 mhz & 512 & 8 mhz & 8.0 km s@xmath37 & 2ac1cc + 2@xmath38 & 128 mhz & 4096 & 31.25 khz & 0.031 km s@xmath37 & 1ac + + 1@xmath36 & 2048 mhz & 512 & 4 mhz & 4.0 km s@xmath37 & 4ac2cc + 2@xmath38 & 32 mhz & 2048 & 15.625 khz & 0.016 km s@xmath37 & 4ac2cc + + 1@xmath36 & 2048 mhz & 512 & 4 mhz & 4.0 km s@xmath37 & 2ac1cc + 2@xmath38 & 64 mhz & 4096 & 15.625 khz & 0.016 km s@xmath37 & 2ac1cc + + 1@xmath36 & 2048 mhz & 512 & 4 mhz & 4.0 km s@xmath37 & 2ac1cc + 2@xmath38 & 128 mhz & 8192 & 15.625 khz & 0.016 km s@xmath37 & 1ac + + 1@xmath36 & 2048 mhz & 512 & 4 mhz & 4.0 km s@xmath37 & 2ac1cc + 2@xmath38 & 96 mhz & 6144 & 15.625 khz & 0.016 km s@xmath37 & 1ac + 2@xmath38 & 32 mhz & 2048 & 15.625 khz & 0.016 km s@xmath37 & 1ac + + + + [ table : ffxopmode ] @xmath39 correction is performed based on the baseline . @xmath39
correction coefficient is generated within fpga .
firstly , @xmath40 , the gradient of @xmath39 to time change , is evaluated and then the value of each frequency channel is calculated ( see figure [ fig : deltaw ] ) .
circuit diagram of the line graphs above is shown in figure [ fig : deltawcir ] .
for every 1 millisecond , the initial value ( init ) is provided by the monitor @xmath41 control computer , and read as a initial gradient [ @xmath42 .
in the border between segments , the previous value is multiplied by the gradient ( grad ) data to generate a gradient of a new segment [ @xmath40 ] . in the arbitrary time
@xmath44 , initial value ( the value of dc ) is set to 0 .
@xmath39 is set for every 128 channel in the full bandwidth of 8@xmath45 channel , which means the full bandwidth ( 8@xmath45 channel ) is corrected with 64 steps .
the initial value of @xmath40 can be set for every 1 millisecond , however , a given value is set within 128 channel .
the gradient of @xmath40 is the gradient variation of @xmath39 per one segment length for fft .
the monitor @xmath41 control computer specifies a set of the initial value and gradient approximately every 1 second .
( 80mm,120mm)fig17.eps ( 130mm,120mm)fig18.eps final correlation value is calculated by adding the output of 16 correlators .
the composition of the long - term accumulation / output board ( ltaob ) is shown in figure [ fig : lta ] .
when the correlation result is output , the data is converted into the format of ieee single - precision floating point . in the fx processing ,
the number of the output frequency channels is normally 4 @xmath45 ( = 4096 ) .
thus the output data size per one correlation is @xmath46 where @xmath47 is the number of single - precision floating bits , @xmath48 is the complex , and @xmath49 is the number of correlations ; auto - correlations of @xmath50 and @xmath51 , and a cross - correlation between them . assuming the minimum integration time is 0.1 second , the estimated output speed is approximately 7.7 mbps . in the ffx processing ,
the number of the output frequency channels is 512 + 2 @xmath45 , thus the data size per one correlation is 480 @xmath45 bits .
consequently , the estimated output speed with 0.1-second integration time is approximately 4.8 mbps .
correlation results are sent to the monitor @xmath41 control computer using tcp / ip protocol of 100 baset - ether .
the ffx correlator are divided into four f parts in principle ( not physically ) .
each f part in a fx mode and the first fft stage of ffx mode are operated at 2048 mhz , and each f part at the second fft stage of ffx mode are operated at 32 mhz .
also , the digital signals input from eib are distributed to arbitrary f parts by using the command ( `` fchsel '' ) from the monitor @xmath41 control computer . in fx mode and the first fft stage of ffx mode ,
the digital signals of 8192 msps are operated by combining two f parts (= 2@xmath522048 mhz ) , while the signals of 4096 msps are operated by using one parts ( = 1@xmath522048 mhz ) . at the second fft stage of ffx mode ,
the digital signals of 8192 msps are operated by combining two f parts ( = 2@xmath5232 mhz ) , while the signals of 4096 msps are operated using only one parts ( = 1@xmath5232 mhz ) . by using the command ( `` fchsel '' ) , the effective bandwidth after the second fft stage can be changed .
all of the operation modes available in this ffx correlator are listed in table [ table : fxopmode ] and [ table : ffxopmode ] .
the hardware of the fx correlator is shown in figure [ fig : farm ] .
the dts - r module consists of two electrical input interface boards ( eibs ) , two delay correction and data configuration boards ( dcdcbs ) , and one dts - r monitor @xmath41 control boards ( drmcbs ) .
the correlation module consists of eight correlation boards ( corbs ) , long - term accumulation / output board ( ltaob ) , and one correlation monitor @xmath41 control board(cormcb ) .
each module is connected to the independent power module .
the power consumption of the filter module is 400 w , while that of the output module is 600 w. the total ac power is 750 w at 1-phase 100 - 220 vac 50/60 hz ( 100 v @xmath53 or 220 v @xmath53 ) .
the total weight is 71.3 kg .
( 80mm,65mm)fig19.eps figure [ fig : measure ] shows a block diagram of the measurement setup of the frequency response of the ffx correlator . to investigate the frequency response , it is useful to use the cw signal , which can measure the folding effects by sweeping the frequency range of 0 to 4096 mhz .
the white noise is important to measure the frequency response , because the astronomical signals obey the gaussian random variable . to obtain the input signals that are approximated to the zero - mean gaussian probability ,
mixing of the cw signal with the white noise is necessary .
( 147mm,100mm)fig20.eps the frequency response of the ffx correlator when cw is included is written as @xmath54 \nonumber \\ & & \cdot h_{1}(f ) h_{2}^{\ast}(f ) \cdot |h_{\mathrm{d}}(f)|^2 , \label{eq : oncor}\end{aligned}\ ] ] where @xmath55 is the frequency response of the cw signal , @xmath56 is the frequency response of the white noise from the aste analog backend subsystem , @xmath57 and @xmath58 are the frequency response by different transmission paths , in which @xmath59 and @xmath60 ( see figure [ fig : measure ] ) , and @xmath61 is the frequency response of the ffx correlator , including the effects of requantization and the folding noise after downsampling . the bandpass calibration is essential for estimating the cw power accurately , because the bandpass response becomes a time - variable due to outdoor air temperature .
the frequency response without the cw signal is written as @xmath62 the adcs work as 1-bit performance ( @xcite ) . in that case
, it is important to adjust the power as precisely as possible so as to avoid the high - order spurious effects .
the frequency responses of @xmath63 and @xmath64 depend on the relative power of the cw signal to the white noise , and also the threshold levels in quantization . to correct these effects , we need to calibrate the bandpass by sensitively adjusting the continuum floor level of @xmath65 to that of @xmath66 in the data analysis .
these values are @xmath67 and @xmath68 . from equations ( [ eq : oncor ] ) and
( [ eq : coroff ] ) , the frequency response in a ffx mode is written as @xmath69 \cdot h_{1}(f ) h_{2}^{\ast}(f ) \cdot |h^{\mathrm{f}}_{\mathrm{d}}(f)|^2 , \label{eq : ffxon}\end{aligned}\ ] ] while the frequency response without the cw signal can be written as @xmath70 and then the cw frequency response including the response of the measurement system is derived as @xmath71 similarly , the frequency response in a fx mode is written as @xmath72 \cdot h_{1}(f ) h_{2}^{\ast}(f ) \cdot |h^{\mathrm{n}}_{\mathrm{d}}(f)|^2 , \label{eq : fxon}\end{aligned}\ ] ] while the frequency response without the cw signal can be written as @xmath73 and then the cw frequency response including the response of the measurement system is derived as @xmath74 from equations ( [ eq : ffxpwr ] ) and ( [ eq : fxpwr ] ) , we can derive the frequency response in a ffx mode from the correlated spectra obtained in ffx and fx modes as @xmath75 since the frequency response of @xmath76 in the fx mode is well known ( @xcite ) , the frequency response in a ffx mode can be derived finally .
figure [ fig : response ] shows the frequency response of the ffx correlator at a bandwidth of 64 mhz in the range of 2048 to 2112 mhz , and that the lower limits of about -33 db to -40 db are successfully measured with this method .
the measurement frequency resolution was 31.25 khz .
the measurement results show that the effective bandwidth is about 59.28 mhz , which was obtained by the passband responses of about @xmath77 db at 2046.40625 and 2106.28125 mhz , @xmath78 db at 2045.3125 and 2107.28125 mhz , and by a stopband response with the first sidelobe of about @xmath79 db . the measurement results are well consistent with the theoretical curve in the passband , both bandedges ( sharpness ) , and first sidelobe levels . in the stopband response except these responses
, however , it is shown that there are differences between theoretical curve and measured results .
this problem is probably due to the non - linear response of 1-bit adc and the precision of the data reduction process in this measurement method .
the cross - modulation distortion is strongly generated in digitizing the cw signals at 1 bit .
this character complicates the data reduction method , and will reduce the measurement precision .
if the adcs with 3 bits or more are feasible , this problem will be relaxed .
finally , the measurement results show that the theory of the ffx scheme can be confirmed , and the development of the ffx correlator was successfully realized .
there are two basic designs of a digital correlator : the xf - type in which the cross - correlation is calculated before fourier transformation , and the fx - type in which fourier transformation is performed before cross multiplication . to improve the fx - type correlator , we established a new algorithm for correlation process , that is called the ffx scheme .
the ffx scheme demonstrates that the realization of a stopband response with first and second sidelobes of @xmath80 db and higher - order sidelobes of @xmath81 db is technically feasible .
the ffx scheme consists of 2-stage fourier transform blocks , which perform the 1st - stage fourier transform as a digital filter , and the 2nd - stage fourier transform to achieve higher dispersion .
the ffx scheme provides flexibility in the setting of bandwidth within the sampling frequency .
the input data rate of the developed ffx correlator is about 48 giga bit per second ( gbps ) with 3-bit quantization at the sampling frequency of 8192 or 4096 msps , which is 8192 msps x 3 bits x 2 if or 4096 msps x 3 bits x 4 ifs .
we have successfully evaluated the feasibilities of the ffx correlator hardware . also , this hardware will be installed and operated as a new spectrometer for aste .
we successfully developed the ffx correlator , measured its performances , and demonstrated the capability of a wide - frequency coverage and high - frequency resolution of the correlation systems .
our development and measurement results will also be useful and helpful in designing and developing the next generation correlator .
the authors would like to acknowledge yoshihiro chikada for his helpful technical discussions .
the author would like to express gratitude to brent carlson who provided constructive comments and suggestions on this paper .
this research was partially supported by the ministry of education , culture , sports , science and technology , grant - in - aid for young scientists ( b ) , 17740114 , 2005 .
bunton , j , 2000 , alma memo 342 , ( charlottesville : nrao ) bunton , j , 2003 , alma memo 447 , ( charlottesville : nrao ) carlson , b. 2001 , nrc - evla memo 014 chikada , y. , ishiguro , m. , hirabayashi , h. , morimoto , m. , morita , k. , kanzawa , t. , iwashita , h. , nakazima , k. , et al .
1987 , proc .
ieee , 75 , 1203 escoffier , r. p. , comoretto , g. webber , j. c. , baudry , a. , broadwell , c. m. , greenberg , j. h. , r. r. treacy , r. r. , cais , p. , et al .
2007 , a@xmath41a 462 , 801 ezawa , h. , kawabe , r. , kohno , k. , yamamoto , s. 2004 , spie 5489 , 763 iguchi , s. , okuramu , s. k. , okiura , m. , momose , m. , chikada , y. ursi general assembly ( j6:recent scientific developments ) , maastricht , 2002 .
iguchi , s. , kurayama , t. , kawaguchi , n. , @xmath41 kawakami , k. 2005 , pasj 57 , 259 narayanan , d. , groppi , c. e. , kulesa , c. a. , & walker , c. k. 2005 , , 630 , 269 okuda , t. , iguchi , s. 2008 , pasj 60 , 315 okumura , s. k. , chikada , y. , momose , m. , iguchi , s. 2001 , alma memo no.350 ( charlottesville : nrao ) rabiner , l. r. , schafer , r. w. 1971 , ieee trans .
audio electroacoust . , au-19 , 200 thompson , a. r. , moran , j. m. , @xmath41 swenson , g. w.jr .
2001 , interferometry and synthesis in radio astronomy , 2nd ed . ,
( new york : john wiley @xmath41 sons ) , 289 weinreb , s. 1963 , digital spectral analysis technique and its application to radio astronomy , ( r. l. e. , mit , cambridge , mass . ) , tech . rep . | we established a new algorithm for correlation process in radio astronomy .
this scheme consists of the 1st - stage fourier transform as a filter and the 2nd - stage fourier transform for spectroscopy .
the `` ffx '' correlator stands for filter and fx architecture , since the 1st - stage fourier transform is performed as a digital filter , and the 2nd - stage fourier transform is performed as a conventional fx scheme .
we developed the ffx correlator hardware not only for the verification of the ffx scheme algorithm but also for the application to the atacama submillimeter telescope experiment ( aste ) telescope toward high - dispersion and wideband radio observation at submillimeter wavelengths .
in this paper , we present the principle of the ffx correlator and its properties , as well as the evaluation results with the production version . |
random walks are widely used to model natural processes in physics , chemistry , and biology @xcite . in particular , first - passage and persistence statistics @xcite of multiple random walks underlie reaction - diffusion processes @xcite , spin systems @xcite , and polymer dynamics @xcite .
first - passage processes involving multiple random walks are equivalent to diffusion in a restricted region of space .
for example , the probability that @xmath0 ordinary random walks do not meet is equivalent to the probability that a `` compound '' walk in @xmath0 dimensions remains confined to the region .
this probability decays as in the long - time limit @xcite . when there are only two or three particles , the compound walk is , in many cases , confined to a wedge , formed by two intersecting planes .
moreover , the well - known properties of diffusion inside an absorbing wedge @xcite explain the long - time kinetics @xcite .
in general , however , the absorbing boundary is defined by multiple intersecting planes in a high - dimensional space . apart from a few special cases , diffusion subject to such complicated boundaries conditions remains an open problem @xcite .
our goal is to use cones in high dimensions to approximate the absorbing boundaries that underlie such first - passage processes . in this study
, we obtain analytic results for the survival probability of a diffusing particle inside an absorbing cone in arbitrary dimension . in a follow - up study @xcite
, we demonstrate that cones provide useful approximations to first - passage characteristics of multiple random walks @xcite . .
the initial location of the particle is parametrized by the radial distance @xmath1 and the polar angle @xmath2.,scaledwidth=25.0% ] we consider a single particle that diffuses inside an unbounded cone with opening angle @xmath3 in spatial dimension @xmath4 ( figure [ fig - cone ] ) .
the central quantity in our study is the probability @xmath5 that the particle does not reach the cone boundary up to time @xmath6 .
regardless of the starting position , this survival probability decays algebraically , @xmath7 , in the long - time limit .
first , we find the exponent @xmath8 analytically by solving the laplace equation inside the cone . in dimensions two and four
, this exponent is an explicit function of the opening angle @xmath3 , and in particular , @xmath9 when @xmath10 . in general dimension
, we find @xmath8 as a root of a transcendental equation involving the associated legendre functions .
second , we derive scaling properties of the exponent .
interestingly , the exponent @xmath8 becomes a function of a single scaling variable in the large-@xmath4 limit .
we obtain the scaling function as a root of the transcendental equation @xmath11 involving the parabolic cylinder function @xmath12 .
the exponent @xmath8 is of order one only in a small region around @xmath13 .
the width of this region shrinks as @xmath14 in the infinite dimension limit .
the exponent diverges algebraically , @xmath15 as @xmath16 , and it is exponentially small , @xmath17 when @xmath18 .
thus , in the large-@xmath4 limit , the exponent @xmath8 is huge if the opening angle is acute , and conversely , it is tiny if the opening angle is obtuse .
strikingly , if we fix the opening angle @xmath3 and take the limit @xmath19 , there are three distinct possibilities , @xmath20 of course , a cone with opening angle @xmath13 is simply a plane , and hence , @xmath21 for all @xmath4 .
third , we study the limiting cases of very thin and very wide cones .
the exponent diverges algebraically , , when the cone is extremely thin .
when the cone is extremely wide , the exponent is exponentially small , .
the rest of this paper is organized as follows . in section
ii , we write the diffusion equation that governs the survival probability , and show that finding the leading asymptotic behavior of the survival probability requires a solution to the laplace equation @xcite .
we present the solutions to this laplace equation in two and four dimensions in section iii , and for an arbitrary dimension in section iv .
the bulk of the paper deals with asymptotic analysis for very large dimensions .
in particular , we derive scaling properties of the exponent and obtain the limiting behaviors of the scaling function ( section v ) .
asymptotic results for extremely thin and extremely wide cones are detailed in sections vi and vii , respectively .
we also obtain the first - passage time ( section viii ) and conclude with a discussion in section ix .
consider a particle undergoing brownian motion @xcite inside an unbounded cone in spatial dimension @xmath4 .
the opening angle @xmath3 , that is , the angle between the cone axis and its surface , fully specifies the cone ( figure [ fig - cone ] ) .
the range of opening angles is @xmath22 , and for @xmath13 , the cone surface is planar .
moreover , the exterior of the cone is itself a cone with opening angle @xmath23 . in two dimensions ,
the cone is a wedge , and in three dimensions , the cone is an ordinary circular cone . at time @xmath24
, the particle is released from a certain location inside the cone .
our goal is to determine the probability that the particle does not reach the cone surface up to time @xmath6 . by symmetry , this survival probability , @xmath25 , depends on the initial distance to the apex @xmath1 , and the initial angle with the cone axis @xmath2 . using a spherical coordinate system where the origin is located at the cone apex and the @xmath26-axis is along the cone axis , the pair of parameters @xmath27 are simply the radial and the polar angle coordinates of the initial location ( figure [ fig - cone ] ) .
the survival probability fully quantifies the first - passage process .
for example , the probability that the particle first reaches the cone surface during the time interval equals @xmath28\times dt$ ] . in general
, the survival probability satisfies the diffusion equation @xcite @xmath29 where @xmath30 is the diffusion constant .
the initial condition is @xmath31 , and the boundary condition is @xmath32 .
we are primarily interested in the large - time kinetics . based on the behavior in two and three dimensions @xcite
, we expect that the survival probability decays algebraically , @xmath33 as @xmath34 .
the exponent @xmath35 depends on the spatial dimension @xmath4 and the opening angle @xmath3 .
the dependence on the initial location enters only through the amplitude @xmath36 .
we now substitute the leading asymptotic behavior into the diffusion equation . since the time derivative becomes negligible in the long - time limit , the amplitude @xmath36 satisfies laplace s equation , @xmath37 subject to the boundary conditions .
the survival probability is finite and positive everywhere inside the cone , and consequently , the amplitude must be finite and positive , @xmath38 for all @xmath39 .
in addition , , to avoid a cusp along the cone axis .
we use dimensional analysis @xcite to solve eq . .
the survival probability is dimensionless , and hence , implies @xmath40=t^\beta$ ] where @xmath41 denotes dimension of time .
the amplitude @xmath36 depends on three variables : the radial coordinate @xmath1 with dimension of length , @xmath42=l$ ] , the diffusion coefficient @xmath30 with @xmath43=l^2/t$ ] , and the dimensionless angle @xmath2 . as the quantity @xmath44 is the only combination of the variables @xmath1 and @xmath30 with dimension @xmath45
, we seek a solution in the form @xmath46 the angular function @xmath47 must be finite and positive everywhere inside the cone , @xmath48 for @xmath39 , and it vanishes on the cone surface , @xmath49 .
we now write the laplacian operator in explicitly by using @xmath50 next , we substitute into the laplace equation and find that the angular function @xmath51 satisfies the eigenvalue equation @xcite @xmath52 + 2\beta(2\beta+d-2)\psi=0.\ ] ] the boundary conditions are @xmath53 and @xmath49 . from equations and ,
the leading asymptotic behavior is @xmath54 as @xmath34 . in particular
, the survival probability grows algebraically with distance , @xmath55 .
we note that the problem of finding the leading asymptotic behavior of the survival probability reduces to an electrostatic problem @xcite as the amplitude satisfies laplace s equation .
our goal is to find the exponent @xmath8 and the angular function @xmath47 .
we expect that the exponent is a monotonically decreasing function of the opening angle @xmath3 .
also , in all dimensions , @xmath21 because a cone with opening angle @xmath13 is a half - space and consequently , the first - passage probability is identical to that of a particle in the vicinity of a trap in one dimension @xcite .
we first discuss two special cases where explicit solutions are possible . in dimension two , the second order differential equation reads , and the two independent solutions are simply @xmath56 and @xmath57 .
the boundary condition @xmath53 excludes the latter and therefore , @xmath58 henceforth , the subscript indicates the dimension .
since the linear equation specifies the angular function only up to an overall constant , we set the prefactor to one throughout this paper .
the boundary condition and the requirement that the angular function must be positive everywhere inside the cone together imply @xmath59 . therefore , the exponent is @xcite @xmath60 as expected , the exponent is a monotonically decreasing function of @xmath3 , and @xmath61 .
the exponent is minimal , yet finite , @xmath62 , for an absorbing needle @xcite .
thus , in dimension two , a diffusing particle reaches a needle with certainty . in dimension four
, the transformation reduces the eigenvalue equation to @xmath63 .
now , the two independent solutions are @xmath64 $ ] and @xmath65 $ ] , but the latter is forbidden because the function @xmath66 must be finite
. therefore , @xmath67}{\sin\theta}.\ ] ] the boundary condition @xmath49 and the requirement @xmath68 for @xmath39 together give @xmath69 .
hence , the exponent is an explicit function of the opening angle in dimension four as well , @xmath70 the exponent vanishes , @xmath71 as @xmath72 , so it is no longer guaranteed that a diffusing particle reaches a needle .
in general dimension , we transform the eigenvalue equation using the variable . in terms of this variable ,
the function @xmath73 satisfies @xmath74 the boundary condition is @xmath75 .
we now use a second transformation , @xmath76 substituting this form into shows that the auxiliary function @xmath77 satisfies the associated legendre equation @xcite @xmath78 \psi=0.\end{aligned}\ ] ] the two independent solutions of this equation are @xmath79 and @xmath80 , the associated legendre functions of degree @xmath81 and order @xmath82 . in even dimensions ,
the first solution is not physical because it implies a divergence of @xmath66 as @xmath83 @xcite . in odd dimensions ,
the second solution is excluded for the same reason .
therefore @xcite , @xmath84 the boundary condition @xmath49 relates the exponent @xmath8 and the opening angle @xmath3 , @xmath85 in general dimension , the exponent @xmath8 is the smallest root of the transcendental equation involving the associated legendre functions .
we must always choose the smallest root because @xmath68 for all @xmath39 .
versus @xmath86 at different dimensions .
this exponent is given explicitly by and in dimensions @xmath87 and @xmath10 , respectively .
in general , the exponent is a root of the transcendental equation .,scaledwidth=45.0% ] we can verify that the exponent @xmath8 decreases monotonically with the opening angle @xmath3 , and that @xmath88 ( figure [ fig - beta ] ) . in all dimensions , the exponent diverges for extremely thin cones , @xmath89 , and when @xmath90 , the exponent vanishes when @xmath72 . versus the normalized polar angle @xmath91 at different dimensions .
the function @xmath47 is normalized such that @xmath92 .
figure 3a shows the behavior for an acute opening angle , @xmath93 , and figure 3b shows the behavior for an obtuse opening angle , @xmath94.,title="fig:",scaledwidth=40.0% ] versus the normalized polar angle @xmath91 at different dimensions .
the function @xmath47 is normalized such that @xmath92 .
figure 3a shows the behavior for an acute opening angle , @xmath93 , and figure 3b shows the behavior for an obtuse opening angle , @xmath94.,title="fig:",scaledwidth=40.0% ] let us fix the opening angle and increase the dimension .
there are two possibilities : ( i ) if @xmath95 , the first - passage process speeds up with increasing dimension because @xmath8 increases ( figure [ fig - beta ] ) and @xmath47 declines ( figure [ fig - psi]a ) ; ( ii ) if @xmath96 , the first - passage process slows down because @xmath8 shrinks ( figure [ fig - beta ] ) and @xmath47 grows ( figure [ fig - psi]b ) .
ultimately , in the infinite - dimension limit , first passage becomes instantaneous for acute angles but infinitely long for obtuse angles . in dimension three , @xmath97 , and
the solution reduces to the legendre function of index @xmath98 , namely , @xcite .
also , the exponent @xmath8 is the smallest root of the transcendental equation @xcite @xmath99 for half - integer values of the exponent , @xmath100 , the angular function @xmath101 is a polynomial of degree @xmath102 in @xmath103 .
this follows directly from equation .
for example , @xmath104 in three dimensions , the polynomials coincide with the legendre polynomials : @xmath105 when . using @xmath49
, we find the opening angles for which the exponent is a half - integer : @xmath106
the special values listed in suggest that the survival exponent has the scaling form @xmath107 in the limit @xmath19 . from equation , we deduce , , and .
to show the scaling behavior in general , we introduce the variable @xmath108 .
performing this scaling transformation on the laplace equation shows that the angular function depends on a single scaling variable , @xmath109 , in the @xmath19 limit .
the scaling function @xmath110 satisfies @xmath111 the boundary condition is @xmath112 , and additionally , @xmath113 for all @xmath114 . , specified by equation , versus the scaling variable @xmath115.,scaledwidth=45.0% ] using the transformation @xmath116 , we recast into the parabolic cylinder equation @xcite @xmath117 the two independent solutions of this equation are @xmath118 and @xmath119 where @xmath120 is the parabolic cylinder function of index @xmath121 .
the second solution is not physical : the asymptotic behavior as @xmath122 @xcite implies a divergent survival probability in the limit @xmath123 .
therefore , @xmath124 as always , we set the overall proportionality constant to one .
the boundary condition @xmath112 relates the scaling function @xmath125 and the scaling variable @xmath126 , defined in , @xmath127 the proper solution is the largest root of the parabolic cylinder function . for half - integer values of the exponent ,
the parabolic cylinder function is related to a hermite polynomial , @xmath128 , and hence , is equivalent to @xmath129 , where the largest root is the appropriate one .
therefore , the half - integer values of the scaling function @xmath125 occur at zeroes of hermite polynomials .
in addition to the aforementioned values , we also quote @xmath130 and @xmath131 .
the scaling behavior is valid in the limits and @xmath19 with the scaling variable @xmath115 or alternatively @xmath132 kept finite .
hence , the scaling function in quantifies the shape of @xmath8 in a `` scaling window '' @xcite centered on @xmath13 .
the size of this window shrinks as @xmath14 at large dimensions . as shown in figure [ fig - scaling ]
, @xmath8 vanishes as , and it diverges as . surprisingly ,
if we fix the opening angle @xmath3 and then take the @xmath19 limit , there are three distinct possibilities , as stated in : the exponent vanishes if @xmath96 , it always equals @xmath133 when @xmath13 , and it diverges if @xmath95 . , specified by , is plotted versus the scaling variable @xmath115 for @xmath134 and @xmath135 . also shown is the scaling function @xmath125.,scaledwidth=45.0% ] we also note that the approach to the scaling behavior is not uniform .
the scaling exponent @xmath8 converges rapidly to the scaling function for negative @xmath126 , but the convergence is quite slow for positive @xmath126 ( figure [ fig - convergence ] ) . to find the asymptotic behavior when @xmath136 , we use the fact that the largest root @xmath137 of the parabolic cylinder function of index @xmath121 , @xmath138 , is located at @xcite @xmath139 when @xmath140 . here
, @xmath141 is the first root of the airy function . substituting @xmath142 into this expression
, we find @xmath143 as @xmath144 .
therefore , the leading behavior is @xmath145 .
furthermore , the first correction to this asymptotic form is @xmath146 thus , the exponent @xmath8 diverges algebraically when .
we use perturbation analysis to find how @xmath8 vanishes in the complementary limit @xmath136 .
equation shows that the angular function becomes constant when @xmath147 .
thus , @xmath148 . substituting this form into , the correction function @xmath149 obeys @xmath150 the boundary condition @xmath151 follows from the small-@xmath8 , small-@xmath26 , behavior of the scaling solution @xmath152 here , we used the values and @xcite , as well as the identities @xmath153 and @xmath154 . using the integrating factor ,
we simplify eq . to @xmath155 .
integration of this equation subject to the boundary condition gives @xmath156 the second integral approaches a constant , and consequently , , when .
we now impose the boundary condition , and find that the exponent is exponentially small , @xmath157 in the limit @xmath158 . in summary ,
the scaling function @xmath125 has the following extremal behaviors @xmath159 at large dimensions , these limiting behaviors apply only inside the scaling window , that is , when @xmath160 is of the order @xmath14 . in the next two sections , we perform asymptotic analysis for the limiting cases of extremely thin cones ( @xmath83 ) and extremely wide cones ( @xmath161 .
the explicit expressions and show that the exponent is inversely proportional to the opening angle , @xmath162 , when the cone is extremely thin .
we anticipate that this divergence is generic .
when the cone is very thin , we have and equation simplifies , @xmath163 this equation holds as long as @xmath164 .
in writing this equation we tacitly assumed a divergent @xmath8 .
we now introduce the scaling variable @xmath165 and transform equation as follows , @xmath166 next , we seek a solution in the form @xmath167 , where as in , @xmath168 .
performing this second transformation gives the bessel equation @xcite @xmath169 this equation has two independent solutions : the bessel functions @xmath170 and @xmath171 .
the boundary condition @xmath53 excludes the latter and hence , @xmath172 the other boundary condition , @xmath49 , gives an implicit relation @xmath173 between the exponent @xmath8 and the opening angle @xmath3 .
we see that the exponent is inversely proportional to the opening angle in all dimensions : @xmath174 as @xmath89 . here , @xmath175 is the first positive zero of the bessel function , @xmath176 .
table i lists the coefficients @xmath177 for @xmath178 .
.the coefficient @xmath177 in for @xmath179 .
equations and yield @xmath180 and @xmath181 , respectively . [ cols="^,<",options="header " , ] in the marginal case @xmath182 , the coefficient in vanishes , @xmath183 . in this case
, the exponent vanishes gently , @xmath184 as follows from @xcite .
this behavior is consistent with the near-@xmath185 behavior of , @xmath186 .
as the opening angle approaches its maximal value , the cone surface turns into an infinitely long , yet infinitesimally thin needle . in two dimensions ,
a diffusing particle is bound to reach such a needle . yet , in dimensions three and higher , the particle may or may not reach the needle .
moreover , the exponentially small exponent shows that the first - passage process becomes extremely slow at high dimensions .
we now briefly discuss the first - passage time .
let @xmath187 be the average duration of the first - passage process , namely , the average time it takes a particle released at @xmath27 to reach the cone surface for the first time ( figure [ fig - cone ] ) .
this quantity obeys the poisson equation @xcite @xmath188 and the boundary condition @xmath189 . as in
, we use dimensional analysis and write where @xmath190 is a dimensionless function of the angle . from eq . , we get @xmath191 + 2d u = -1,\ ] ] while the boundary condition becomes @xmath192 .
the linear equation has the particular solution @xmath193 , and thus , we seek a solution in the form .
the function @xmath194 obeys the homogeneous linear differential equation @xmath195 + 2du = 0,\ ] ] subject to the boundary condition @xmath196 .
this equation is a special case of , and using , we immediately find where @xmath197 is set by the boundary condition .
finally , we find that the first - passage time has the compact form @xmath198 this first - passage time is finite if and only if @xmath199 .
in summary , we studied first - passage kinetics for a particle diffusing in a cone . in all dimensions , the probability that the particle does not reach the cone boundary decays algebraically with time .
we found the exponent underlying this power - law behavior as a root of a transcendental equation involving associated legendre functions .
we also obtained scaling and extremal properties of the exponent .
our results generalize the known properties of first passage in two- and three - dimensional cones @xcite to arbitrary dimensions . moreover , the statistical physics perspective , where scaling plays a central role , extends rigorous studies of diffusion in cones in probability theory @xcite and potential theory @xcite .
scaling implies that the behavior in a narrow window becomes universal : cones with different combinations of @xmath3 and @xmath4 , have the same exponent @xmath8 , as long as the scaling variable @xmath115 is the same .
the exponent is exponentially small when @xmath158 , and it grows algebraically when @xmath136 .
as the dimension increases , the first - passage process speeds up if the opening angle is acute but slows down if the opening angle is obtuse .
this behavior is reflected by the asymptotic behavior for fixed @xmath2 in the large-@xmath4 limit .
we merely quote the results of the corresponding asymptotic analysis , @xmath200 the exponent is proportional to the dimension at acute angles , and it decays exponentially with dimension at obtuse angles .
the asymptotic behavior is consistent with the limiting behavior of the scaling function as well as the asymptotic results and .
kinetics of first - passage in a cone provide useful information about first - passage problems involving multiple random walks .
for example , let us consider the random - walk problem mentioned in the introduction .
the probability that the trajectories of @xmath0 ordinary random walks do not meet up to time @xmath6 is equivalent to the probability that a compound random walk in @xmath0 dimensions remains inside the region . since there are @xmath201 permutations of the positions , the total solid angle inside this `` allowed region '' accounts for a fraction @xmath202 of space .
we replace the allowed region with an @xmath0-dimensional cone that has the same solid angle , @xmath203 . using the stirling formula ,
, the opening angle is @xmath204 .
the survival probability decays algebraically , @xmath205 , and using , the exponent is @xmath206 .
this cone approximation correctly gives the asymptotic @xmath0-dependence as the exact exponent is @xmath207 @xcite .
first passage in the exterior of this narrow cone gives an approximation for the probability that the order of @xmath0 random walks does not turn into the mirror image of the initial state .
if the initial ordering of the random walks is , then @xmath5 is the probability that the particles do not reach the configuration up to time @xmath6 .
since the angle of the cone is now @xmath208 , the asymptotic behavior gives @xmath205 with the tiny exponent @xmath209 in a follow - up study , we use cones to understand other random walk problems @xcite . with a few exceptions such as paraboloids @xcite ,
the study of first passage in elementary geometries is still in its infancy .
we considered first passage in basic circular cones .
a natural extension is to generalized cones @xcite , defined as domains with the property that all rays emanating from the apex do not intersect the cone boundary .
understanding first - passage properties in such generalized cones ultimately requires a solution to the laplace - dirichlet boundary value problem . | we study statistics of first passage inside a cone in arbitrary spatial dimension .
the probability that a diffusing particle avoids the cone boundary decays algebraically with time .
the decay exponent depends on two variables : the opening angle of the cone and the spatial dimension . in four dimensions
, we find an explicit expression for the exponent , and in general , we obtain it as a root of a transcendental equation involving associated legendre functions . at large dimensions ,
the decay exponent depends on a single scaling variable , while roots of the parabolic cylinder function specify the scaling function .
consequently , the exponent is of order one only if the cone surface is very close to a plane .
we also perform asymptotic analysis for extremely thin and extremely wide cones . |
in this paper we consider asymptotically anti de sitter ( ads ) type metrics on @xmath9-dimensional manifolds with boundary @xmath3 , @xmath10 .
we recall the actual definition of ads space below , but for our purposes the most important feature is the asymptotic of the metric on these spaces , so we start by making a bold general definition . thus , an asymptotically ads type space is a manifold with boundary @xmath3 such that @xmath11 is equipped with a pseudo - riemannian metric @xmath12 of signature @xmath13 which near the boundary @xmath14 of @xmath3 is of the form @xmath15 @xmath16 a smooth symmetric 2-cotensor on @xmath3 such that with respect to some product decomposition of @xmath3 near @xmath14 , @xmath17 , @xmath18 is a section of @xmath19 ( rather than merely would have variable order , and the restrictions on @xmath6 that arise here , @xmath20 , would have to be modified . ]
@xmath21 ) and is a lorentzian metric on @xmath14 ( with signature @xmath22 ) .
note that @xmath14 is time - like with respect to the conformal metric @xmath23 i.e. the dual metric @xmath24 of @xmath2 is negative definite on @xmath25 , i.e. on @xmath26 , in contrast with the asymptotically de sitter - like setting studied in @xcite when the boundary is space - like .
moreover , @xmath14 is _ not _ assumed to be compact ; indeed , under the assumption ( tf ) below , which is useful for global well - posedness of the wave equation , it never is .
let the wave operator @xmath27 be the laplace - beltrami operator associated to this metric , and let @xmath28 be the klein - gordon operator , @xmath29 .
the convention with the positive sign for the ` spectral parameter ' @xmath6 preserves the sign of @xmath6 relative to the @xmath30 component of the metric in both the riemannian conformally compact and the lorentzian de sitter - like cases , and hence is convenient when describing the asymptotics .
we remark that if @xmath31 then up to a change of the ( overall ) sign of the metric , these spaces are asymptotically de sitter , hence the results of @xcite apply .
however , some of the results are different even then , since in the two settings the role of the time variable is reversed , so the formulation of the results differs as the role of ` initial ' and ` boundary ' conditions changes .
these asymptotically ads - metrics are also analogues of the riemannian ` conformally compact ' , or asymptotically hyperbolic , metrics , introduced by mazzeo and melrose @xcite in this form , which are of the form @xmath32 with @xmath33 smooth riemannian on @xmath3 , and @xmath18 is a section of @xmath19 .
these have been studied extensively , in part due to the connection to ads metrics ( so some phenomena might be expected to be similar for ads and asymptotically hyperbolic metrics ) and their riemannian signature , which makes the analysis of related pde easier .
we point out that hyperbolic space actually solves the riemannian version of einstein s equations , while de sitter and anti - de sitter space satisfy the actual hyperbolic einstein equations .
we refer to the works of fefferman and graham @xcite , graham and lee @xcite and anderson @xcite among others for analysis on conformally compact spaces .
we also refer to the works of witten @xcite , graham and witten @xcite and graham and zworski @xcite , and further references in these works , for results in the riemannian setting which are of physical relevance .
there is also a large body of literature on asymptotically de sitter spaces . among others ,
anderson and chruciel studied the geometry of asymptotically de sitter spaces @xcite , while in @xcite the asymptotics of solutions of the klein - gordon equation were obtained , and in @xcite the forward fundamental solution was constructed as a fourier integral operator .
it should be pointed out that the de sitter - schwarzschild metric in fact has many similar features with asymptotically de sitter spaces ( in an appropriate sense , it simply has two de sitter - like ends ) .
a weaker version of the asymptotics in this case is contained in the part of works of dafermos and rodnianski @xcite ( they also study a non - linear problem ) , and local energy decay was studied by bony and hfner @xcite , in part based on the stationary resonance analysis of s barreto and zworski @xcite ; stronger asymptotics ( exponential decay to constants ) was shown in a series of papers with antnio s barreto and richard melrose @xcite . for the universal cover of ads space itself ,
the klein - gordon equation was studied by breitenlohner and freedman @xcite , who showed its solvability for @xmath20 , @xmath34 , and uniqueness for @xmath35 , in our normalization .
analogues of these results were extended to the dirac equation by bachelot @xcite ; and on exact ads space there is an explicit solution due to yagdjian and galstian @xcite . finally , for a class of perturbations of the universal cover of ads , which still possess a suitable killing vector field , holzegel @xcite recently showed well - posedness for @xmath20 by imposing a boundary condition , see ( * ? ? ?
* definition 3.1 ) .
he also obtained certain estimates on the derivatives of the solution , as well as pointwise bounds .
below we consider solutions of @xmath36 , or indeed @xmath37 with @xmath38 given . before describing our results , first we recall a formulation of the conformal problem , namely @xmath39 , so @xmath2 is lorentzian smooth on @xmath3 , and @xmath14 is time - like at the end of the introduction we give a full summary of basic results in the ` compact ' and ` conformally compact ' riemannian and lorentzian settings , with space - like as well as time - like boundaries in the latter case .
let @xmath40 adding @xmath6 to the operator makes no difference in this case ( unlike for @xmath41 ) .
suppose that @xmath42 is a space - like hypersurface in @xmath3 intersecting @xmath14 ( automatically transversally ) .
then the cauchy problem for the dirichlet boundary condition , @xmath43 @xmath38 , @xmath44 , @xmath45 given , @xmath46 a vector field transversal to @xmath42 , is locally well - posed ( in appropriate function spaces ) near @xmath42 . moreover , under a global condition on the generalized broken bicharacteristic ( or ) flow and @xmath42 , which we recall below in definition [ def : gen - br - bich ] , the equation is globally well - posed .
namely , the global geometric assumption is that @xmath47 where @xmath48 is the bundle projection . in the above formulation of the problem
, we would assume that @xmath42 is a level set , @xmath49 note that locally this is always true in view of the lorentzian nature of the metric and the conditions on @xmath14 and @xmath42 . as is often the case in the presence of boundaries , see e.g. ( * ? ? ?
* theorem 24.1.1 ) and the subsequent remark , it is convenient to consider the special case of the cauchy problem with vanishing initial data and @xmath38 supported to one side of @xmath42 , say in @xmath50 ; one can phrase this as solving @xmath51 this forward cauchy problem is globally well - posed for @xmath52 , @xmath53 , and the analogous statement also holds for the backward cauchy problem .
here we use hrmander s notation @xmath54 , see ( * ? ? ?
* appendix b ) , to avoid confusion with the ` zero sobolev spaces ' @xmath55 , which we recall momentarily .
in addition , ( without any global assumptions ) singularities of solutions , as measured by the b - wave front set , @xmath56 , relative to either @xmath57 or @xmath58 , propagate along as was shown by melrose , sjstrand and taylor @xcite , see also @xcite in the analytic setting . here recall that in @xmath11 , bicharacteristics are integral curves of the hamilton vector field @xmath59 ( on @xmath60 ) of the principal symbol @xmath61 inside the characteristic set , @xmath62 we also recall that the notion of a @xmath63 and an analytic is somewhat different due to the behavior at diffractive points , with the analytic definition being more permissive ( i.e. weaker ) . throughout this paper
we use the analytic definition , which we now recall .
first , we need the notion of the compressed characteristic set , @xmath64 of @xmath65 .
this can be obtained by replacing , in @xmath66 , @xmath67 by its quotient @xmath68 , where @xmath25 is the conormal bundle of @xmath14 in @xmath3 .
one denotes then by @xmath64 the image @xmath69 of @xmath70 in this quotient .
one can give a topology to @xmath64 , making a set @xmath71 open if and only if @xmath72 is open in @xmath70 .
this notion of the compressed characteristic set is rather intuitive , since working with the quotient encodes the law of reflection : points with the same tangential but different normal momentum at @xmath14 are identified , which , when combined with the conservation of kinetic energy ( i.e. working on the characteristic set ) gives the standard law of reflection .
however , it is very useful to introduce another ( equivalent ) definition already at this point since it arises from structures which we also need .
the alternative point of view ( which is what one needs in the proofs ) is that the analysis of solutions of the wave equation takes place on the b - cotangent bundle , @xmath73 ( ` b ' stands for boundary ) , introduced by melrose .
we refer to @xcite for a very detailed description , @xcite for a concise discussion .
invariantly one can define @xmath73 as follows .
first , let @xmath74 be the set of all @xmath63 vector fields on @xmath3 tangent to the boundary .
if @xmath75 are local coordinates on @xmath3 , with @xmath5 defining @xmath14 , elements of @xmath74 have the form @xmath76 with @xmath77 and @xmath78 smooth .
it follows immediately that @xmath74 is the set of all smooth sections of a vector bundle , @xmath79 : @xmath80 , @xmath81 , give local coordinates in terms of .
then @xmath73 is defined as the dual bundle of @xmath79 .
thus , points in the b - cotangent bundle , @xmath73 , of @xmath3 are of the form @xmath82 so @xmath83 give coordinates on @xmath73 .
there is a natural map @xmath84 induced by the corresponding map between sections @xmath85 thus @xmath86 i.e. @xmath87 , @xmath88 . over the interior of @xmath3 we can identify @xmath89 with @xmath90 , but
this identification @xmath91 becomes singular ( no longer a diffeomorphism ) at @xmath14 .
we denote the image of @xmath70 under @xmath91 by @xmath92 called the compressed characteristic set .
thus , @xmath64 is a subset of the vector bundle @xmath73 , hence is equipped with a topology which is equivalent to the one define by the quotient , see ( * ? ? ?
* section 5 ) . the definition of _ analytic _ then becomes : [ def : gen - br - bich ] _ generalized broken bicharacteristics , _ or , are continuous maps @xmath93 , where @xmath94 is an interval , satisfying that for all @xmath95 real valued , @xmath96 since the map @xmath97 is a derivation , @xmath98 at @xmath70 , so bicharacteristics are merely reparameterized if @xmath99 is replaced by a conformal multiple . in particular , if @xmath41 is the klein - gordon operator , @xmath100 , for an asymptotically ads - metric @xmath12 , the bicharacteristics over @xmath11 are , up to reparameterization , those of @xmath2 .
we make this into our definition of . the compressed characteristic set @xmath64 of @xmath41
is that of @xmath101 .
generalized broken bicharacteristics , or , of @xmath41 are _ in the analytic sense _ of the smooth lorentzian metric @xmath2 .
we now give a formulation for the global problem . for this purpose
we need to recall one more class of differential operators in addition to @xmath74 ( which is the set of @xmath63 vector fields _ tangent to the boundary _ ) .
namely , we denote the set of @xmath63 vector fields _ vanishing at the boundary _ by @xmath102 . in local
coordinates @xmath103 , these have the form @xmath104 with @xmath105 ; cf . .
again , @xmath102 is the set of all @xmath63 sections of a vector bundle , @xmath106 , which over @xmath11 can be naturally identified with @xmath107 ; we refer to @xcite for a detailed discussion of 0-geometry and analysis , and to @xcite for a summary .
we then let @xmath108 , resp .
@xmath109 , be the set of differential operators generated by @xmath74 , resp .
@xmath102 , i.e they are locally finite sums of products of these vector fields with @xmath110-coefficients . in particular ,
@xmath111 which explains the relevance of @xmath109 .
this can be seen easily from @xmath12 being in fact a non - degenerate smooth symmetric bilinear form on @xmath106 ; the conformal factor @xmath112 compensates for the vanishing factors of @xmath5 in , so in fact this is _ exactly _ the same statement as @xmath2 being lorentzian on @xmath113 .
let @xmath114 denote the zero - sobolev space relative to @xmath115 so if @xmath116 is an integer then @xmath117 negative values of @xmath118 give sobolev spaces by dualization .
for our problem , we need a space of ` very nice ' functions corresponding to @xmath108 .
we obtain this by replacing @xmath110 with the space of conormal functions to the boundary relative to a fixed space of functions , in this case @xmath114 , i.e. functions @xmath119 such that @xmath120 for every @xmath121 ( of any order ) .
the finite order regularity version of this is @xmath122 , which is given for @xmath123 integer by @xmath124 while for @xmath125 integer , @xmath126 if @xmath127 , @xmath128 , @xmath129 .
thus , @xmath130 is the dual space of @xmath122 , relative to @xmath131 .
note that in @xmath11 , there is no distinction between @xmath74 , @xmath102 , or indeed simply @xmath132 ( smooth vector fields on @xmath3 ) , so over compact subsets @xmath133 of @xmath11 , @xmath122 is the same as @xmath134 . on the other hand , at @xmath135 , @xmath122 distinguishes precisely between regularity relative to @xmath102 and @xmath74 .
although the finite speed of propagation means that the wave equation has a local character in @xmath3 , and thus compactness of the slices @xmath49 is immaterial , it is convenient to assume @xmath136 even as stated , the propagation of singularities results ( which form the heart of the paper ) do not assume this , and the assumption is made elsewhere merely to make the formulation and proof of the energy estimates and existence slightly simpler , in that one does not have to localize in spatial slices this way .
suppose @xmath20 .
suppose @xmath137 we want to find @xmath138 such that @xmath139 we show that this is locally well - posed near @xmath42 .
moreover , under the previous global assumption on , this problem is globally well - posed : ( see theorem [ thm : well - posed ] . ) assume that ( tf ) and ( pt ) hold .
suppose @xmath20 . the forward dirichlet problem , , has a unique global solution @xmath138 , and for all compact @xmath140 there exists a compact @xmath141 and a constant @xmath142 such that for all @xmath38 as in , the solution @xmath143 satisfies @xmath144 in fact
, one can be quite explicit about @xmath145 in view of ( pt ) , since @xmath146}$ ] can be estimated by @xmath147 , @xmath94 open containing @xmath148 $ ] .
we also prove microlocal elliptic regularity and describe the propagation of singularities of solutions , as measured by @xmath56 relative to @xmath149 .
we define this notion in definition [ def : wfb ] and discuss it there in more detail .
however , we recall the definition of the standard wave front set @xmath150 on manifolds without boundary @xmath3 that immediately generalizes to the b - wave front set @xmath56 . thus , one says that @xmath151 is _ not _ in the wave front set of a distribution @xmath143 if there exists @xmath152 such @xmath153 is invertible and @xmath154 for all @xmath155 this is equivalent to @xmath156 by the sobolev embedding theorem . here
@xmath157 can be replaced by @xmath158 instead , with @xmath159 arbitrary .
moreover , @xmath160 can also be defined analogously : we require @xmath161 for @xmath162 elliptic at @xmath163 .
thus , @xmath164 means that @xmath143 is ` microlocally @xmath63 at @xmath163 ' , while @xmath165 means that @xmath143 is ` microlocally @xmath166 at @xmath163 ' . in order to microlocalize @xmath122 , we need pseudodifferential operators , here extending @xmath108 ( as that is how we measure regularity ) .
these are the b - pseudodifferential operators @xmath167 introduced by melrose , their principal symbol @xmath168 is a homogeneous degree @xmath159 function on @xmath169 ; we again refer to @xcite .
then we say that @xmath170 is _ not _ in @xmath171 if there exists @xmath172 with @xmath173 invertible and such that @xmath174 is @xmath175-conormal to the boundary .
one also defines @xmath176 : @xmath177 if there exists @xmath167 with @xmath173 invertible and such that @xmath178 .
one can also extend these definitions to @xmath125 . with this definition
we have the following theorem : ( see proposition [ prop : elliptic ] and theorem [ thm : prop - sing ] . )
suppose that @xmath179 , @xmath20 , @xmath180 or @xmath181 .
suppose @xmath182 for some @xmath183 .
then @xmath184 moreover , @xmath185 is a union of maximally extended generalized broken bicharacteristics of the conformal metric @xmath2 in @xmath186 in particular , if @xmath36 then @xmath187 is a union of maximally extended generalized broken bicharacteristics of @xmath2 . as a consequence of the preceding theorem
, we obtain the following more general , and precise , well - posedness result .
( see theorem [ thm : well - posed - precise ] . )
assume that ( tf ) and ( pt ) hold .
suppose that @xmath179 , @xmath20 , @xmath180 , @xmath188 .
suppose @xmath189 . then has a unique solution in @xmath190 , which in fact lies in @xmath191 , and for all compact @xmath140 there exists a compact @xmath141 and a constant @xmath142 such that @xmath192 while we prove this result using the propagation of singularities , thus a relatively sophisticated theorem , it could also be derived without full microlocalization , i.e. without localizing the propagation of energy in phase space . we also generalize propagation of singularities to the case @xmath193 ( @xmath194 arbitrary ) , in which case we prove one sided propagation depending on the sign of @xmath195 .
namely , if @xmath196 , resp .
@xmath197 , @xmath185 is a union of _ maximally forward , resp .
backward _ , extended generalized broken bicharacteristics of the conformal metric @xmath2 .
there is no difference between the case @xmath198 and @xmath199 , resp .
@xmath193 , at the elliptic set , i.e. the statement @xmath184 holds even if @xmath193 .
we refer to proposition [ prop : elliptic ] and theorem [ thm : prop - sing - im ] for details .
these results indicate already that for @xmath193 there are many interesting questions to answer , and in particular that one can not think of @xmath6 as ` small ' ; this will be the focus of future work . in particular , if @xmath38 is conormal relative to @xmath200 then @xmath201 .
let @xmath202 denote the branch square root function on @xmath203 $ ] chosen so that takes positive values on @xmath204 .
the simplest conormal functions are those in @xmath110 that vanish to infinite order ( i.e. with all derivatives ) at the boundary ; the set of these is denoted by @xmath205 .
if we assume @xmath206 then @xmath207 as we show in proposition [ prop : asymp ] . since the indicial roots of @xmath100 are @xmath208 this explains the interpretation of this problem as a ` dirichlet problem ' , much like it was done in the riemannian conformally compact case by mazzeo and melrose @xcite : asymptotics corresponding to the growing indicial root , @xmath209 , @xmath210 , is ruled out . for @xmath20 , one can then easily solve the problem with inhomogeneous ` dirichlet ' boundary condition , i.e. given @xmath211 and @xmath206 , both supported in @xmath212 , @xmath213 if @xmath214 is not an integer .
if @xmath215 is an integer , the same conclusion holds if we replace @xmath210 by @xmath216 ; see theorem [ thm : inhomog - dirichlet ] .
the operator @xmath217 is the analogue of the dirichlet - to - neumann map , or the scattering operator . in the de sitter
setting the setup is somewhat different as both pieces of scattering data are specified either at past or future infinity , see @xcite .
nonetheless , one expects that the result of ( * ? ? ?
* section 7 ) , that the scattering operator is a fourier integral operator associated to the relation can be extended to the present setting , at least if the boundary is totally geodesic with respect to the conformal metric @xmath2 , and the metric is even with respect to the boundary in an appropriate sense . indeed , in an ongoing project , baskin and the author are extending baskin s construction of the forward fundamental solution on asymptotically de sitter spaces , @xcite , to the even totally geodesic asymptotically ads setting . in addition , it is an interesting question what the ` best ' problem to pose is when @xmath193 ; the results of this paper suggest that the global problem ( rather than local , cauchy data versions ) is the best behaved .
one virtue of the parametrix construction is that we expect to be able answer lorentzian analogues of questions related to the work of mazzeo and melrose @xcite , which would bring the lorentzian world of ads spaces significantly closer ( in terms of results ) to the riemannian world of conformally compact spaces .
we singled out the totally geodesic condition and evenness since they hold on actual ads space , which we now discuss .
we now recall the structure of the actual ads space to justify our terminology .
consider @xmath218 with the pseudo - riemannian metric of signature @xmath219 given by @xmath220 with @xmath221 denoting coordinates on @xmath218 , and the hyperboloid @xmath222 inside it .
note that @xmath223 on the hyperboloid , so we can ( diffeomorphically ) introduce polar coordinates in these two variables , i.e. we let @xmath224 , @xmath225 , @xmath226 .
then the hyperboloid is of the form @xmath227 inside @xmath228 .
as @xmath229 , @xmath230 , @xmath231 and @xmath232 are linearly independent at the hyperboloid , @xmath233 give local coordinates on it , and indeed these are global in the sense that the hyperboloid @xmath11 is identified with @xmath234 via these . a straightforward calculation
shows that the metric on @xmath218 restricts to give a lorentzian metric @xmath12 on the hyperboloid . indeed , away from @xmath235
, we obtain a convenient form of the metric by using polar coordinates @xmath236 in @xmath237 , so @xmath238 : @xmath239 where @xmath240 is the standard round metric ; a similar description is easily obtained near @xmath235 by using the standard euclidean variables .
we can compactify the hyperboloid by compactifying @xmath237 to a ball @xmath241 via inverse polar coordinates @xmath242 , @xmath243 , @xmath244 thus , the interior of @xmath241 is identified with @xmath237 , and the boundary @xmath245 of @xmath241 is added at @xmath246 to compactify @xmath237 .
we let @xmath247 be this compactification of @xmath11 ; a collar neighborhood of @xmath8 is identified with @xmath248 in this collar neighborhood the lorentzian metric takes the form @xmath249 which is of the desired form , and the conformal metric is @xmath250 with respect to which the boundary , @xmath251 , is indeed time - like .
note that the induced metric on the boundary is @xmath252 , up to a conformal multiple .
as already remarked , @xmath2 has the special feature that @xmath14 is totally geodesic , unlike e.g. the case of @xmath253 equipped with a product lorentzian metric , with @xmath254 carrying the standard euclidean metric . for global results ,
it is useful to work on the universal cover @xmath255 of @xmath3 , where @xmath256 is the universal cover of @xmath257 ; we use @xmath258 to emphasize the time - like nature of this coordinate .
the local geometry is unchanged , but now @xmath258 provides a global parameter along generalized broken bicharacteristics , and satisfies the assumptions ( tf ) and ( pt ) for our theorems .
we use this opportunity to summarize the results , already referred to earlier , for analysis on conformally compact riemannian or lorentzian spaces , including a comparison with the conformally related problem , i.e. for @xmath259 or @xmath101 .
we assume dirichlet boundary condition ( dbc ) when relevant for the sake of definiteness , and global hyperbolicity for the hyperbolic equations , and do not state the function spaces or optimal forms of regularity results .
1 . riemannian : @xmath260 , with dbc is well - posed for @xmath261 ; moreover , if @xmath206 , then @xmath262 .
( also works outside a discrete set of poles @xmath6 in @xmath263 . )
lorentzian , @xmath264 is spacelike , @xmath38 supported in @xmath50 , @xmath29 : @xmath265 , @xmath143 supported in @xmath50 , is well - posed .
if @xmath206 , the solution is @xmath63 up to @xmath266 .
3 . lorentzian , @xmath8 is timelike , @xmath38 supported in @xmath50 , @xmath29 : @xmath265 , with dbc at @xmath14 , @xmath143 supported in @xmath50 , is well - posed . if @xmath206 , the solution is @xmath63 up to @xmath266 .
we now go through the original problems .
let @xmath267 be as in . 1
. asymptotically hyperbolic , @xmath268 : there is a unique solution of @xmath269 , @xmath206 , such that @xmath270 , @xmath271 .
( analogue of dbc ; mazzeo and melrose @xcite . )
( indeed , @xmath272 , and this can be extended to @xmath273 , apart from finitely many poles in @xmath274 $ ] , and analytically continued further . ) 2 .
asymptotically de sitter , @xmath29 : for @xmath38 supported in @xmath50 , there is a unique solution of @xmath275 supported in @xmath50 .
moreover , for @xmath206 , @xmath276 , @xmath277 and @xmath278 is specified , provided that @xmath279 .
( see @xcite . )
asymptotically anti de sitter , @xmath280 : for @xmath206 supported in @xmath50 , there is a unique solution of @xmath275 such that @xmath270 , @xmath271 and @xmath281 .
the structure of this paper is the following . in section
[ sec : poincare ] we prove a poincar inequality that we use to allow the sharp range @xmath20 for @xmath6 real .
then in section [ sec : energy ] we recall the structure of energy estimates on manifolds without boundary as these are then adapted to our ` zero geometry ' in section [ sec : diff0-diffb ] . in section [ sec : diff0-psib ] we introduce microlocal tools to study operators such as @xmath41 , namely the zero - differential - b - pseudodifferential calculus , @xmath282 . in section [ sec : gbb ] the structure of is recalled . in section [ sec : elliptic ] we study the dirichlet form and prove microlocal elliptic regularity .
finally , in section [ sec : prop - sing ] , we prove the propagation of singularities for @xmath41 .
i am very grateful to dean baskin , rafe mazzeo and richard melrose for helpful discussions .
i would also like to thank the careful referee whose comments helped to improve the exposition significantly and also led to the removal of some very confusing typos .
let @xmath16 be a conformally compact riemannian metric , i.e. a positive definite inner product on @xmath106 , hence by duality on @xmath283 ; we denote the latter by @xmath284 .
we denote the corresponding space of @xmath285 sections of @xmath283 by @xmath286 . while the inner product on @xmath287 depends on the choice of @xmath16 , the corresponding norms are independent of @xmath16 , at least over compact subsets @xmath133 of @xmath3 .
we first prove a hardy - type inequality : [ lemma : local - sharp - poincare ] suppose @xmath288 is real with @xmath289 , and let @xmath290 be given by @xmath291 .
given any compact subset @xmath133 of @xmath3 and @xmath292 , there exists @xmath293 such that if @xmath294 is supported in @xmath133 then for @xmath295 supported in @xmath296 , @xmath297 recall here that @xmath205 denotes elements of @xmath110 that vanish at @xmath135 to infinite order , and the subscript @xmath298 on @xmath299 below indicates that in addition the support of the function under consideration is compact . for any @xmath290 real , and @xmath300 , @xmath301
, we have , using @xmath302 , @xmath303u , u\rangle = \langle \chi u , v^ * u\rangle-\langle vu,\chi u\rangle\\ & \qquad=-\langle \chi u , vu\rangle-\langle vu,\chi u\rangle -\langle \chi u,({\operatorname{div}}v)u\rangle . \end{split}\ ] ] now , if @xmath291 , @xmath288 transversal to @xmath8 , and if we write @xmath304 , @xmath305 a smooth non - degenerate density then in local coordinates @xmath306 such that @xmath307 , @xmath308 , @xmath309 where the subscript @xmath2 in @xmath310 denotes that the divergence is with respect to @xmath2 .
thus , assuming @xmath288 with @xmath289 , @xmath311 let @xmath312 be such that @xmath313 in @xmath314 .
thus , if @xmath315 , @xmath316 near @xmath317 , @xmath318 , @xmath319 is supported in @xmath314 , @xmath320 , then @xmath321 hence @xmath322 @xmath323 and thus given any @xmath324 there is @xmath293 such that for @xmath143 supported in @xmath133 , @xmath325 namely we take @xmath326 such that @xmath327 , choose @xmath316 on @xmath328 $ ] , supported in @xmath329 .
this completes the proof of the lemma .
the basic poincar estimate is : [ prop : nonlocal - poincare - no - weight ] suppose @xmath140 compact , @xmath330 , @xmath71 open with @xmath331 , @xmath71 arcwise connected to @xmath8 , @xmath332 compact .
there exists @xmath142 such that for @xmath138 one has @xmath333 where the norms are relative to the metric @xmath16 .
it suffices to prove the estimate for @xmath294 , for then the proposition follows by the density of @xmath205 in @xmath149 and the continuity of both sides in the @xmath149 topology .
let @xmath334 be as in lemma [ lemma : local - sharp - poincare ] , and let @xmath335 identically @xmath336 on a neighborhood of @xmath337 , supported in @xmath71 , and let @xmath293 be as in the lemma with @xmath133 replaced by @xmath145 .
we pull back @xmath338 to a function @xmath339 defined on a neighborhood of @xmath14 by the @xmath340-flow ; thus , @xmath341 . by decreasing @xmath342
if needed , we may assume that @xmath339 is defined and is @xmath63 in @xmath296 , and @xmath343 .
now , let @xmath295 identically @xmath336 where @xmath344 , supported where @xmath345 , and let @xmath346 be identically @xmath336 where @xmath345 , supported in @xmath296 ; thus @xmath347 . then
, by lemma [ lemma : local - sharp - poincare ] applied to @xmath348 , @xmath349 the full proposition follows by the standard poincar estimate and arcwise connectedness of @xmath133 to @xmath14 ( hence to @xmath344 ) , since one can estimate @xmath350 in @xmath285 in terms of @xmath351 in @xmath285 and @xmath352 .
we can get a more precise estimate of the constants if we restrict to a neighborhood of a space - like hypersurface @xmath42 ; it is convenient to state the result under our global assumptions . _ thus , ( tf ) and ( pt ) are assumed to hold from here on in this section . _
[ prop : poincare - no - weight ] suppose @xmath288 is real with @xmath289 , @xmath353 near @xmath14 and let @xmath290 be given by @xmath291 .
let @xmath94 be a compact interval .
let @xmath354 , @xmath355 .
then there exist @xmath356 , @xmath293 and @xmath357 such that the following holds . for @xmath358 , @xmath359 and for @xmath138
one has @xmath360)\}}\\ & \leq c^{-1}\|v u\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0+{\epsilon}],\ x(p)\leq x_0\ } ) } + \gamma\|du\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0+{\epsilon}]\})}\\ & \qquad\qquad+c'\|u\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0]\ } ) } , \end{split}\ ] ] where the norms are relative to the metric @xmath16 .
we proceed as in the proof of proposition [ prop : nonlocal - poincare - no - weight ] , using that the @xmath258-preimage of the enlargement of the interval by distance @xmath361 points is still compact by ( pt ) ; we always use @xmath362 correspondingly .
we simply let @xmath363 , where @xmath364 is the characteristic function of @xmath365 $ ] . thus @xmath366 vanishes near @xmath14
; at the cost of possibly decreasing @xmath342 we may assume that it vanishes in @xmath296 . by , with @xmath367 , @xmath368 on @xmath369 , supported in @xmath370 , @xmath371 thus , it remains to give a bound for @xmath372)\}}$ ] .
let @xmath42 be the space - like hypersurface in @xmath3 given by @xmath49 , @xmath358 .
now let @xmath373 be transversal to @xmath42 .
the standard poincar estimate ( whose weighted version we prove below in lemma [ lemma : weighted - poincare ] ) obtained by integrating from @xmath374 yields that for @xmath294 with @xmath375 , @xmath376\})}\leq c'({\epsilon}+\delta)^{1/2}\|w u\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0+{\epsilon}]\})},\ ] ] with @xmath377 as @xmath378 . applying this with @xmath143 supported where @xmath379 @xmath380\})}\leq c''({\epsilon}+\delta)^{1/2}\|xw u\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0+{\epsilon}]\})},\ ] ] with @xmath381 as @xmath378 .
as we want @xmath359 , we choose @xmath356 such that @xmath382 let @xmath383)$ ] be identically @xmath336 on @xmath384 , and be supported in @xmath385 .
applying to @xmath386 , @xmath387\})}\\ & \qquad\leq c''({\epsilon}+\delta)^{1/2}\|xw u\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0+{\epsilon}]\})}\\ & \qquad\qquad+c''({\epsilon}+\delta)^{1/2}\|x\chi'(t ) ( wt ) u\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0]\})}. \end{split}\ ] ] in particular , this can be applied with @xmath143 replaced by @xmath388 .
this completes the proof .
we also need a weighted version of this result .
we first recall a poincar inequality with weights .
[ lemma : weighted - poincare ] let @xmath389 .
suppose that @xmath373 real , @xmath390 , @xmath391 , and @xmath392 for @xmath50 , @xmath393 .
then there exists @xmath142 such that for @xmath138 with @xmath50 on @xmath394 , @xmath395 we compute , using @xmath396 , @xmath397u , u\rangle = \langle \chi u , w^ * u\rangle-\langle wu,\chi u\rangle\\ & \qquad=-\langle \chi u , wu\rangle-\langle wu,\chi u\rangle -\langle \chi u,({\operatorname{div}}w)u\rangle , \end{split}\ ] ] so @xmath398 dividing through by @xmath399 and rearranging yields @xmath400 hence the claim follows .
our poincar inequality ( which could also be named hardy , in view of the relationship of to the hardy inequality ) is then : [ prop : poincare ] suppose @xmath288 is real with @xmath289 , @xmath353 near @xmath14 , and let @xmath290 be given by @xmath291 .
let @xmath94 be a compact interval .
let @xmath354 .
then there exist @xmath356 , @xmath293 , @xmath357 , @xmath401 such that the following holds .
suppose @xmath358 , @xmath402 .
let @xmath403 , @xmath404 and @xmath405 on @xmath365 $ ] , @xmath319 supported in @xmath406 $ ] , @xmath407 . for @xmath138 one
has @xmath408)\}}\\ & \leq c^{-1}\| |\chi'|^{1/2 } vu\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0+{\epsilon } ] , \ x(p)\leq x_0\})}\\ & \qquad\qquad + c'\gamma\| \chi^{1/2 } du\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0+{\epsilon}]\})}\\ & \qquad\qquad+c'\|u\|_{l^2_0(\{p:\ t(p)\in[t_0-\delta , t_0]\ } ) } , \end{split}\ ] ] where the norms are relative to the metric @xmath16 .
let @xmath42 be the space - like hypersurface in @xmath3 given by @xmath49 , @xmath358 .
we apply lemma [ lemma : weighted - poincare ] with @xmath373 transversal to @xmath42 as follows .
one has from applied with @xmath339 replaced by @xmath409 that @xmath410 we now use lemma [ lemma : weighted - poincare ] with @xmath411 replaced by @xmath412 , @xmath413 on @xmath414 , @xmath415 , to estimate @xmath416 .
we choose @xmath417 so that in addition @xmath418 ; this can be done by pulling back a function @xmath419 from @xmath42 under the @xmath420-flow .
we may also assume that @xmath417 is supported where @xmath421 in view of @xmath422 on @xmath414 ( we might need to shorten the time interval we consider , i.e. @xmath356 , to accomplish this ) .
thus , @xmath423 , and hence @xmath424 as @xmath421 on @xmath425 , one can estimate @xmath426 in terms of @xmath427 ( even though @xmath16 is a riemannian 0-metric ! ) , giving the desired result .
we recall energy estimates on manifolds without boundary in a form that will be particularly convenient in the next sections .
thus , we work on @xmath11 , equipped with a lorentz metric @xmath12 , and dual metric @xmath428 ; let @xmath27 be the dalembertian , so @xmath429 .
we consider a ` twisted commutator ' with a vector field @xmath430 , where @xmath431 is a real vector field , typically of the form @xmath432 , @xmath411 a cutoff function .
thus , we compute @xmath433 the point being that the use of @xmath434 eliminates zeroth order terms and hence is useful when we work not merely modulo lower order terms . note that @xmath435 is a second order , real , self - adjoint operator , so if its principal symbol agrees with that of @xmath436 for some real self - adjoint bundle endomorphism @xmath437 , then in fact both operators are the same as the difference is @xmath317th order and vanishes on constants .
correspondingly , there are no @xmath317th order terms to estimate , which is useful as the latter tend to involve higher derivatives of @xmath411 , which in turn tend to be large relative to @xmath438 .
the principal symbol in turn is easy to calculate , for the operator is @xmath439,\ ] ] whose principal symbol is @xmath440 in fact , it is easy to perform this calculation explicitly in local coordinates @xmath306 and dual coordinates @xmath441 .
let @xmath442 , so @xmath443 .
we write the components of the metric tensors as @xmath444 and @xmath445 , and @xmath446 when this does not cause confusion .
we also write @xmath447 .
_ in the remainder of this section only , we adopt the standard summation convention .
_ then @xmath448 so @xmath449 ( the first two terms of @xmath450 are the same after summation , but it is convenient to keep them separate ) hence @xmath451 relabelling the indices , we deduce that @xmath452 with the first and fourth terms combining into @xmath453 , so @xmath454 where @xmath455 are the matrix entries of @xmath437 relative to the basis @xmath456 of the fibers of the cotangent bundle .
we now want to expand @xmath457 using @xmath432 , and separate the terms with @xmath411 derivatives , with the idea being that we choose the derivative of @xmath411 large enough relative to @xmath411 to dominate the other terms .
thus , @xmath458 and multiplying the first term on the right hand side by @xmath459 ( and summing over @xmath460 ) gives @xmath461 which is twice the sesquilinear stress - energy tensor associated to the wave @xmath143 .
this is well - known to be positive definite in @xmath462 , i.e. for covectors @xmath463 , @xmath464 vanishing if and only if @xmath465 , when @xmath420 and @xmath438 are both forward time - like for smooth lorentz metrics , see e.g. ( * ? ? ?
* section 2.7 ) or ( * ? ? ?
* lemma 24.1.2 ) . in the present setting ,
the metric is degenerate at the boundary , but the analogous result still holds , as we show below .
if we replace the wave operator by the klein - gordon operator @xmath466 , @xmath29 , we obtain an additional term @xmath467 in @xmath468 as compared to . with @xmath430 , @xmath432 , as above
, this contributes @xmath469 in terms containing derivatives of @xmath411 to @xmath470 . in particular , @xmath471 @xmath472 , @xmath473 .
now suppose that @xmath420 and @xmath438 are either both time like ( either forward or backward ; this merely changes an overall sign ) .
the point of is that one controls the left hand side if one controls @xmath474 ( in the extreme case , when @xmath475 , it simply vanishes ) , and one can regard all terms on the right hand side after @xmath476 as terms one can control by a small multiple of the positive definite quantity @xmath477 due to the poincar inequality if one arranges that @xmath478 is large relative to @xmath411 , and thus one can control @xmath477 in terms of @xmath474 .
in fact , one does not expect that @xmath438 will be non - degenerate time - like everywhere : then one decomposes the energy terms into a region @xmath479 where one has the desired definiteness , and a region @xmath480 where this need not hold , and then one can estimate @xmath481 in @xmath479 in terms of its behavior in @xmath480 and @xmath474 : thus one propagates energy estimates ( from @xmath480 to @xmath479 ) , provided one controls @xmath474 .
of course , if @xmath143 is supported in @xmath479 , then one automatically controls @xmath143 in @xmath480 , so we are back to the setting that @xmath143 is controlled by @xmath474 .
this easily gives uniqueness of solutions , and a standard functional analytic argument by duality gives solvability .
it turns out that in the asymptotically ads case one can proceed similarly , except that the term @xmath482 is not negligible any more at @xmath8 , and neither is @xmath483 .
in fact , the @xmath194 term is the ` same size ' as the stress energy tensor at @xmath8 , hence the need for an upper bound for it , while the @xmath195 term is even larger , hence the need for the assumption @xmath198 because although @xmath411 is not differentiated ( hence in some sense ` small ' ) , @xmath420 is a vector field that is too large compared to the vector fields the stress energy tensor can estimate at @xmath8 : it is a b - vector field , rather than a 0-vector field : we explain these concepts now .
we start by recalling that @xmath74 is the lie algebra of @xmath63 vector fields on @xmath3 tangent to @xmath8 , while @xmath102 is the lie algebra of @xmath63 vector fields vanishing at @xmath8 .
thus , @xmath102 is a lie subalgebra of @xmath74 .
note also that both @xmath102 and @xmath74 are @xmath110-modules under multiplication from the left , and they act on @xmath484 , in the case of @xmath102 in addition mapping @xmath110 into @xmath485 . the lie subalgebra property can be strengthened as follows .
[ lemma : vz - ideal ] @xmath102 is an ideal in @xmath74 .
suppose @xmath486 , @xmath373 .
then , as @xmath46 vanishes at @xmath8 , there exists @xmath487 such that @xmath488 .
thus , @xmath489=[xv',w]=[x , w]v'+x[v',w].\ ] ] now , as @xmath420 is tangent to @xmath14 , @xmath490=-wx\in x{{\mathcal{c}}^\infty}(x)$ ] , and as @xmath491 , @xmath492\in{\mathcal{v}}(x)$ ] , so @xmath493\in x{\mathcal{v}}(x)={\mathcal{v}_0}(x)$ ] . as usual
, @xmath109 is the algebra generated by @xmath102 , while @xmath108 is the algebra generated by @xmath74 .
we combine these in the following definition , originally introduced in @xcite ( indeed , even weights @xmath494 were allowed there ) .
let @xmath495 be the ( complex ) vector space of operators on @xmath205 of the form @xmath496 where the sum is locally finite , and let @xmath497 we recall that this space is closed under composition , and that commutators have one lower order in the 0-sense than products , see ( * ? ? ?
* lemma 4.5 ) : [ lemma : diff - filtered - ring ] @xmath498 is a filtered ring under composition with @xmath499 moreover , composition is commutative to leading order in @xmath500 , i.e. for @xmath501 as above , with @xmath502 , @xmath503\in{\mathrm{diff}_0}^{k+k'-1}{\mathrm{diff}_{{\mathrm b}}}^{m+m'}(x).\ ] ] here we need an improved property regarding commutators with @xmath108 ( which would a priori only gain in the 0-sense by the preceding lemma ) .
it is this lemma that necessitates the lack of weights on the @xmath108-commutant .
[ lemma : diffb - commutant ] for @xmath504 , @xmath505 , @xmath506 , @xmath503\in{\mathrm{diff}_0}^k{\mathrm{diff}_{{\mathrm b}}}^{s+m-1}(x).\ ] ] we first note that only the leading terms in terms of @xmath507 order in both commutants matter for the conclusion , for otherwise the composition result , lemma [ lemma : diff - filtered - ring ] , gives the desired conclusion .
we again write elements of @xmath498 as locally finite sums of products of vector fields and functions , and then , using lemma [ lemma : diff - filtered - ring ] and expanding the commutators , we are reduced to checking that 1 .
@xmath486 , @xmath373 , @xmath508=-[v , w]\in{\mathrm{diff}_0}^1(x)$ ] , which follows from lemma [ lemma : vz - ideal ] , 2 . and for @xmath373 , @xmath509 , @xmath510=wf\in{{\mathcal{c}}^\infty}(x)={\mathrm{diff}_{{\mathrm b}}}^0(x)$ ] . in both cases
thus , the commutator drops b - order by 1 as compared to the product , completing the proof of the lemma .
we also remark the following : [ lemma : b - to-0-conversion ] for each non - negative integer @xmath511 with @xmath512 , @xmath513 this result is an immediate consequence of @xmath514 .
integer ordered sobolev spaces , @xmath122 were defined in the introduction .
it is immediate from our definitions that for @xmath515 , @xmath516 is continuous .
a particular consequence of lemma [ lemma : diffb - commutant ] is that if @xmath290 , @xmath517 , the @xmath518\in{\mathrm{diff}_0}^m(x)$ ] .
we also note that for @xmath519 , @xmath520 , @xmath431 real , we have @xmath521 , where the adjoint is taken with respect to the @xmath522 inner product .
namely : [ lemma : divergence ] suppose @xmath519 , @xmath520 , @xmath431 real .
then @xmath521 , and with @xmath523 @xmath524 with the metric density given by @xmath525 , @xmath526 .
combining these results we deduce : suppose @xmath519 , @xmath520 , @xmath431 real .
then @xmath527 where @xmath528 and in the basis @xmath529 , @xmath530 we write @xmath531\in{\mathrm{diff}_0}^2(x),\ ] ] and compute the principal symbol , which we check agrees with that of @xmath532 .
one way of achieving this is to do the computation over @xmath11 ; by continuity if the symbols agree here , they agree on @xmath283 .
but over the interior this is the standard computation leading to ; in coordinates @xmath306 , with dual coordinates @xmath441 , writing @xmath533 , @xmath534 , both sides have principal symbol @xmath535 now both sides of are elements of @xmath536 , are formally self - adjoint , real , and have the same principal symbol .
thus , their difference is a first order , self - adjoint and real operator ; it follows that its principal symbol vanishes , so in fact this difference is zeroth order .
since it annihilates constants ( as both sides do ) , it actually vanishes .
we particularly care about the terms in which the coefficients @xmath537 are differentiated , with the idea being that we write @xmath432 , and choose the derivative of @xmath411 large enough relative to @xmath411 to dominate the other terms .
thus , as in , @xmath538 and multiplying the first term on the right hand side by @xmath459 ( and summing over @xmath460 ) gives @xmath539 which is twice the sesquilinear stress - energy tensor @xmath540 associated to the wave @xmath143 .
as we mentioned before , this is positive definite when @xmath420 and @xmath438 are both forward time - like for smooth lorentz metrics . in the present
setting , the metric is degenerate at the boundary , but the analogous result still holds since @xmath541 so the lorentzian non - degenerate nature of @xmath24 proves the ( uniform ) positive definiteness in @xmath542 , considered as an element of @xmath543 , hence in @xmath462 , regarded as an element of @xmath544 .
indeed , we recall the quick proof here since we need to improve on this statement to get an optimal result below . thus , we wish to show that for @xmath545 , @xmath546 , @xmath463 and @xmath420 forward time - like , @xmath547 is positive definite as a quadratic form in @xmath548 . since replacing @xmath420 by a positive multiple
does not change the positive definiteness , we may assume , as we do below , that @xmath549 . then we may choose local coordinates @xmath550 such that @xmath551 and @xmath552 , thus @xmath553 .
then @xmath554 being forward time - like means that @xmath555 and @xmath556 .
thus , @xmath557 with the last inequality strict if @xmath558 , and the preceding one ( by the strict forward time - like character of @xmath463 ) strict if @xmath559 and @xmath560 .
it is then immediate that at least one of these inequalities is strict unless @xmath561 , which is the claimed positive definiteness .
we claim that we can make a stronger statement if @xmath562 and @xmath563 and @xmath564 ( thus @xmath565 is necessarily space - like , i.e. @xmath566 ) : @xmath567 is positive definite in @xmath548 .
indeed , in this case ( again assuming @xmath568 ) we can choose coordinates as above such that @xmath551 , @xmath565 is a multiple of @xmath569 , namely @xmath570 , @xmath571 . to achieve this
, we complete @xmath572 and @xmath573 ( which are orthogonal by assumption ) to a @xmath2 normalized orthogonal basis @xmath574 of @xmath575 , and then choose coordinates such that the coordinate vector fields are given by the @xmath576 at @xmath163 .
then @xmath463 forward time - like means that @xmath555 and @xmath556 , and @xmath563 means that @xmath577 .
thus , with @xmath578 , @xmath579 on the right hand side the term in the large parentheses is the same kind of expression as in , with the terms with @xmath580 dropped , thus is positive definite in @xmath581 , and for @xmath578 , the first term is positive definite in @xmath582 , so the left hand side is indeed positive definite as claimed . rewriting this in terms of @xmath428 in our setting
, we obtain that for @xmath578 @xmath583 is positive definite in @xmath462 , considered an element of @xmath584 , when @xmath585 , and hence is positive definite sufficiently close to @xmath8 . stating the result as a lemma : [ lemma : energy - normal - deriv ] suppose @xmath586 , @xmath587 , @xmath588 and @xmath563 and @xmath564 .
then @xmath589 is positive definite in @xmath590 . at this point
we modify the choice of our time function @xmath258 so that we can construct @xmath565 and @xmath420 satisfying the requirements of the lemma .
assume ( tf ) and ( pt ) . given @xmath591 and
a compact interval @xmath94 there exists a function @xmath592 such that @xmath593 for @xmath594 , @xmath595 is time - like in the same component of the time - like cone as @xmath596 , and @xmath597 at @xmath246 .
let @xmath598 , identically @xmath336 near @xmath317 , @xmath599 , @xmath600 , supported in @xmath601 $ ] , and for @xmath602 to be specified let @xmath603 note that on the support of @xmath604 , @xmath605 , so if @xmath606 is sufficiently small , @xmath607 , and bounded away from @xmath317 , there in view of ( pt ) and as @xmath607 at @xmath14 . at @xmath246 @xmath608 so @xmath597 .
as already noted , on the support of @xmath604 , @xmath605 , so for @xmath594 , @xmath94 compact , in view of ( pt ) , @xmath609 .
next , @xmath610 where @xmath611 now , @xmath612 which is @xmath613 if @xmath614 , i.e. @xmath615 $ ] .
but @xmath616 , so @xmath617 indeed , i.e. @xmath618 is timelike . since @xmath619 is still time - like for @xmath620 , @xmath618 is in the same component of time - like covectors as @xmath596 , i.e. is forward oriented .
next , observe that with @xmath621 , @xmath622 so over compact sets @xmath623 can be made arbitrarily small by first choosing @xmath624 sufficiently small and then @xmath356 sufficiently small .
thus , @xmath625 is forward time - like as well . reducing @xmath356 further if needed , completes the proof .
this lemma can easily be made global .
[ lemma : t - global - construction ] assume ( tf ) and ( pt ) .
given @xmath591 there exists a function @xmath592 such that @xmath593 for @xmath626 , @xmath595 is time - like in the same component of the time - like cone as @xmath596 , and @xmath597 at @xmath246 .
in particular , @xmath627 also satisfies ( tf ) and ( pt ) .
we proceed as above , but let @xmath628 we then have two additional terms , @xmath629 and @xmath630 in @xmath595 .
note that on the support of both terms @xmath631 , while @xmath632 is uniformly bounded .
thus , if @xmath633 , @xmath634 , @xmath635 , the factors in front of @xmath596 in both terms is bounded in absolute value by @xmath636 .
now for any @xmath118 there are @xmath637 , which we may assume are in @xmath638 and are decreasing with @xmath118 , such that on @xmath639 $ ] , @xmath627 so defined , satisfies all the requirements if @xmath640 , @xmath641 on @xmath94 and @xmath635 , @xmath634 .
but now in view of the bounds on @xmath642 and @xmath643 it is straightforward to write down @xmath644 and @xmath645 with the desired properties , e.g. by approximating the piecewise linear function which takes the value @xmath642 at @xmath646 , @xmath647 , to get @xmath644 , and similarly with @xmath648 , finishing the proof . _ from this point on , within this section , we assume that ( tf ) and ( pt ) hold . from now on we simply replace @xmath258 by @xmath627 . _ we let @xmath649 , @xmath650 .
thus , _ at _ @xmath246 , @xmath651 we extend @xmath652 to a vector field @xmath565 such that @xmath653 , i.e. @xmath565 is tangent to the level surfaces of @xmath258 .
then we have on all of @xmath3 , @xmath654 and @xmath655 on a neighborhood of @xmath14 , with uniform upper and lower bounds ( bounding away from @xmath317 ) for both and on compact subsets of @xmath3 . using lemma [ lemma : energy - normal - deriv ] and the equations just above
, we thus deduce that for @xmath656 , @xmath578 , @xmath657 , identically @xmath336 near @xmath14 , supported sufficiently close to @xmath14 , @xmath520 , @xmath432 , @xmath658 with @xmath659 , @xmath660 and @xmath661 is positive definite , all independent of @xmath411 . here
@xmath417 is used since @xmath662 is only positive definite near @xmath14 .
fix @xmath663 .
let @xmath664 for @xmath665 , @xmath666 for @xmath667 , @xmath668 identically 1 on @xmath669 , vanishing on @xmath670 $ ] , thus , @xmath671 for @xmath672 .
now consider @xmath673 so @xmath674\ ] ] and @xmath675\rightarrow \tilde\chi'=-\digamma^{-1 } \chi_0'(-\digamma^{-1}(s - t_1)),\ ] ] so @xmath675\rightarrow \tilde\chi=-\digamma^{-1 } ( s - t_1)^2\tilde\chi',\ ] ] so for @xmath676 sufficiently large , this is bounded by a small multiple of @xmath677 , namely @xmath678\rightarrow \tilde\chi=-\gamma\tilde\chi ' , \
\gamma=(t_1-t_0)^2\digamma^{-1}.\ ] ] in particular , for sufficiently large @xmath679 , @xmath680 on @xmath681 $ ] .
in addition , by and , for @xmath199 , and @xmath682 sufficiently close to @xmath336 @xmath683 while @xmath684 and @xmath685 however , @xmath686 is too large to be controlled by the stress energy tensor since @xmath420 is a b - vector field , but not a 0-vector field . thus , in order to control the @xmath195 term for @xmath687 $ ] , we need to assume that @xmath198 .
then , writing @xmath688 , and choosing @xmath676 sufficiently large to absorb the first term on the right hand side of , @xmath689 for sufficiently small @xmath624 and sufficiently large @xmath676 we absorb all but the first parenthesized term on the right hand side into the left hand side by the positive definiteness of @xmath661 and the poincar inequality , proposition [ prop : poincare ] , to conclude that for @xmath143 supported in @xmath681 $ ] , @xmath690 in view of the poincar inequality we conclude : [ lemma : energy - est ] suppose @xmath20 , @xmath663 , @xmath411 as above . for @xmath294
supported in @xmath681 $ ] one has @xmath691 [ rem : uniform - length ] note that if @xmath94 is compact then there is @xmath692 such that for @xmath358 we can take any @xmath693 $ ] , i.e. the time interval over which we can make the estimate is uniform over such compact intervals @xmath94 .
this lemma gives local in time uniqueness immediately , hence iterative application of the lemma , together with remark [ rem : uniform - length ] , yields : [ cor : unique ] suppose @xmath20 . for @xmath694 supported in @xmath695 , there is at most one @xmath696 such that @xmath697 and @xmath37 . via the standard functional analytic argument
, we deduce from : [ lemma : weak - local - exist ] suppose @xmath20 , @xmath94 a compact interval .
there is @xmath698 such that for @xmath358 , and for @xmath699 supported in @xmath695 , there exists @xmath700 , @xmath697 and @xmath37 in @xmath701 . for any subspace @xmath702 of @xmath703 let @xmath704}$ ] consist of elements of @xmath702 restricted to @xmath705 $ ] , @xmath706}$ ] consist of elements of @xmath702 supported in @xmath705 $ ] . in particular ,
an element of @xmath707}$ ] vanishes to infinite order at @xmath708 .
thus , the dot over @xmath63 denotes the infinite order vanishing at @xmath8 , while the @xmath709 denotes the infinite order vanishing at the time boundaries we artificially imposed .
we assume that @xmath38 is supported in @xmath710 .
we use lemma [ lemma : energy - est ] , with the role of @xmath711 and @xmath712 reversed ( backward in time propagation ) , and our requirement on @xmath713 is that it is sufficiently small so that the backward version of the lemma is valid with @xmath714 .
( this can be done uniformly over @xmath94 by remark [ rem : uniform - length ] . )
let @xmath715 and @xmath712 be such that @xmath716 . applying the estimate , using @xmath717 , with @xmath143 replaced by @xmath718}$ ] with @xmath712 in the role of @xmath711 there ( backward estimate ) , @xmath719 in the role of @xmath711 ,
we obtain : @xmath720 } } \leq c\|p^*\phi\|_{h^{-1,1}_{0,{{\mathrm b}}}(x)|_{[\tau_0,t_1 ] } } , \
\phi\in{\dot { \mathcal{c}}^\infty}_{{\mathrm{comp}}}(x)^\bullet_{[\tau_0,t_1]}.\ ] ] it is also useful to rephrase this as @xmath721 } } \leq c\|p^*\phi\|_{h^{-1,1}_{0,{{\mathrm b}}}(x)|_{[\tau_0,t_1 ] } } , \
\phi\in{\dot { \mathcal{c}}^\infty}_{{\mathrm{comp}}}(x)^\bullet_{[\tau_0,t_1]},\ ] ] when @xmath722 . by
, @xmath723}\to{\dot { \mathcal{c}}^\infty}_{\mathrm{comp}}(x)^\bullet_{[t_0,t_1]}$ ] is injective .
define @xmath724}}p^*\to { \dot { \mathcal{c}}^\infty}_{{\mathrm{comp}}}(x)^\bullet_{[t_0,t_1]}\ ] ] by @xmath725 being the unique @xmath726}$ ] such that @xmath727 .
now consider the conjugate linear functional on @xmath728 } } p^*$ ] given by @xmath729 in view of , and the support condition on @xmath38 ( namely the support is in @xmath710 ) and @xmath730 ( the support is in @xmath731 ) as an element of @xmath732 , while @xmath725 as an element of @xmath733}$ ] , so these can be naturally paired , with the pairing bounded in the appropriate norms .
we then write these norms as @xmath734}$ ] and @xmath735}$ ] .
] , @xmath736 } } \,\|(p^*)^{-1}\psi\|_{h^1_0(x)|_{[t_0+\delta_0,t_1]}}\\ & \leq c\|f\|_{h^{-1}_0(x)|_{[t_0+\delta_0,t_1 ] } } \|\psi\|_{h^{-1,1}_{0,{{\mathrm b}}}(x)|_{[t_0,t_1 ] } } , \end{split}\ ] ] so @xmath737 is a continuous conjugate linear functional if we equip @xmath738 } } p^*$ ] with the @xmath739}$ ] norm .
if we did not care about the solution vanishing in @xmath740 , we could simply use hahn - banach to extend this to a continuous conjugate linear functional @xmath143 on @xmath741}$ ] , which can thus be identified with an element of @xmath742}$ ] .
this would give @xmath743 @xmath718}$ ] , so @xmath37 .
we do want the vanishing of @xmath143 in @xmath744 , i.e. when applied to @xmath339 supported in this region . as a first step in this direction , let @xmath745 , and note that if @xmath746 } } p^*\ ] ] then @xmath747 directly by , namely the right hand side vanishes by the support condition on @xmath38 .
correspondingly , the conjugate linear map @xmath748 is well - defined on the algebraic sum @xmath749 } } p^*\ ] ] by @xmath750 } } p^*.\ ] ] we claim that the functional @xmath748 is actually continuous when is equipped with the @xmath739}$ ] norm . but this follows from @xmath751 } } \|\psi\|_{h^{-1,1}_{0,{{\mathrm b}}}(x)|_{[t_0+\delta'_0,t_1]}}\ ] ] together with @xmath752}}\leq \|\phi+\psi\|_{h^{-1,1}_{0,{{\mathrm b}}}(x)|_{[t_0,t_1]}}\ ] ] since @xmath339 vanishes on @xmath753 $ ] .
correspondingly , by the hahn - banach theorem , we can extend @xmath748 to a continuous conjugate linear map @xmath754}\to{\mathbb{c}},\ ] ] which can thus by identified with an element of @xmath742}$ ] .
this gives @xmath743 @xmath718}$ ] supported in @xmath755 , so @xmath37 , and in addition @xmath756},\ ] ] so @xmath757 in particular , extending @xmath143 to vanish on @xmath758 , which is compatible with the existing definition in view of , we have a distribution solving the pde , defined on @xmath759 , with the desired support condition . in particular , using a cutoff function @xmath411 which is identically @xmath336 for @xmath760 $ ] , is supported for @xmath761 $ ] , @xmath762 , @xmath763 vanishes for @xmath764 as well as @xmath765 , and @xmath37 on @xmath766 , thus completing the proof .
[ prop : weak - exist ] suppose @xmath20
. for @xmath699 supported in @xmath695 , there exists @xmath700 , @xmath697 and @xmath37 .
we subdivide the time line into intervals @xmath767 $ ] , each of which is sufficiently short so that energy estimates hold even on @xmath768 $ ] ; this can be done in view of the uniform estimates on the length of such intervals over compact subsets . using a partition of unity
, we may assume that @xmath38 is supported in @xmath769 $ ] , and need to construct a global solution of @xmath37 with @xmath143 supported in @xmath770 .
first we obtain @xmath771 as above solving the pde on @xmath772 $ ] ( i.e. @xmath773 is supported in @xmath774 ) and supported in @xmath775 $ ] .
let @xmath776 , this is thus supported in @xmath777 $ ] .
we next solve @xmath778 on @xmath779 $ ] with a result supported in @xmath780 $ ] .
then @xmath781 is supported in @xmath782 $ ] , etc .
proceeding inductively , and noting that the resulting sum is locally finite , we obtain the solution on all of @xmath3 .
well - posedness of the solution will follow once we show that for solutions @xmath783 of @xmath37 , @xmath784 supported in @xmath695 , we in fact have @xmath785 ; indeed , this is a consequence of the propagation of singularities .
we state this as a theorem now , recalling the standing assumptions as well : [ thm : well - posed ] assume that ( tf ) and ( pt ) hold .
suppose @xmath20 . for @xmath694 supported in @xmath695 , there exists a unique @xmath696 such that @xmath697 and @xmath37 .
moreover , for @xmath140 compact there is @xmath141 compact , depending on @xmath133 and @xmath711 only , such that @xmath786 while we used @xmath627 of lemma [ lemma : t - global - construction ] instead of @xmath258 throughout , the conclusion of this theorem is invariant under this change ( since @xmath591 is arbitrary in lemma [ lemma : t - global - construction ] ) , and thus is actually valid for the original @xmath258 as well .
uniqueness and follow from corollary [ cor : unique ] and the estimate preceding it . by proposition [ prop : weak - exist ] , this problem has a solution @xmath700 with the desired support property . by the propagation of singularities ,
theorem [ thm : prop - sing ] , @xmath138 since @xmath143 vanishes for @xmath787 .
in order to microlocalize , we need to replace @xmath108 by @xmath788 and @xmath789 .
we refer to @xcite for a thorough discussion and ( * ? ? ? * section 2 ) for a concise introduction to these operator algebras including all the facts that are required here . in particular , the distinction between @xmath788 and @xmath789 is the same as between @xmath790 and @xmath791 of _ classical _ , or _
one step polyhomogeneneous _ ,
standard , pseudodifferential operators , i.e. elements of the former ( @xmath788 , resp .
@xmath790 ) are ( locally ) quantizations of symbols with a full one - step polyhomogeneous asymptotic expansion ( also called _ classical _ symbols ) , while those of the latter ( @xmath789 , resp .
@xmath791 ) are ( locally ) quantizations of symbols which merely satisfy symbolic estimates .
while the former are convenient since they have homogeneous principal symbols , the latter are more useful when one must use approximations ( e.g. by smoothing operators ) , as is often the case below . before proceeding
, we recall that points in the b - cotangent bundle , @xmath73 , of @xmath3 are of the form @xmath792 thus , @xmath83 give coordinates on @xmath73 .
if @xmath793 are the standard coordinates on @xmath66 induced by local coordinates on @xmath3 , i.e. one - forms are written as @xmath794 , then the map @xmath84 is given by @xmath795 . to be a bit more concrete ( but again we refer to @xcite and ( * ? ? ?
* section 2 ) for more detail ) , we can define a large subspace ( which in fact is sufficient for our purposes here ) of @xmath796 and @xmath797 locally by explicit _ quantization maps _ ; these can be combined to a global quantization map by a partition of unity as usual .
thus , we have @xmath798 , which restrict to @xmath799 , @xmath800 denoting classical symbols .
namely , over a local coordinate chart @xmath565 with coordinates @xmath103 , @xmath801 , and with @xmath77 supported in @xmath802 , @xmath803 compact , we may take @xmath804 understood as an oscillatory integral , where @xmath805 is identically @xmath336 near @xmath317 , and the integral in @xmath806 is over @xmath263 .
note that @xmath339 is irrelevant as far as the behavior of schwartz kernels near the diagonal is concerned ( it is identically @xmath336 there ) ; it simply localizes to a neighborhood of the diagonal .
somewhat inaccurately , one may write @xmath807 as @xmath808 , so @xmath77 is symbolic in b - vector fields ; a more accurate way of reflecting this is to change variables , writing @xmath87 , @xmath88 , so @xmath809 with this explicit quantization , the _ principal symbol _ @xmath168 of @xmath810 is the class @xmath811 $ ] of @xmath77 in @xmath812 .
if @xmath77 is classical , this class can be further identified with a homogeneous , of degree @xmath159 , symbol , i.e. an element of @xmath813 .
on the other hand , the _ operator wave front set _ @xmath814 of @xmath810 can be defined by saying that @xmath815 is _ not _ in @xmath814 if @xmath99 has a conic neighborhood @xmath816 in @xmath169 such that @xmath817 is rapidly decreasing ( i.e. is an order @xmath818 symbol ) in @xmath816 .
thus , on the complement of @xmath814 , @xmath661 is microlocally order @xmath818 .
a somewhat better definition of @xmath789 and @xmath788 is directly in terms of the schwartz kernels .
the schwartz kernels are well - behaved on the b - double space @xmath819 $ ] created by blowing up the corner @xmath820 in the product space @xmath821 ; in particular they are smooth away from the diagonal and vanish to infinite order off the front face . in these terms @xmath339 above localizes to a neighborhood of the diagonal which only intersects the boundary of @xmath822 in the front face of the blow - up .
the equivalence of the two descriptions can be read off directly from , which shows that the schwartz kernel is a right b - density valued ( this is the factor @xmath823 in ) conormal distribution to @xmath824 , @xmath825 , i.e. the lift of the diagonal to @xmath822 .
the space @xmath789 forms a filtered algebra , so for @xmath826 , @xmath827 , one has @xmath828 . in addition , the commutator satisfies @xmath829\in{\psi_{\mathrm{bc}}}^{m+m'-1}(x)$ ] , i.e. it is one order lower than the product , but there is no gain of decay at @xmath8 .
however , we also recall the following crucial lemma from ( * ? ? ?
* section 2 ) : [ lemma : xd_x - improvement ] for @xmath826 , resp .
@xmath167 , one has @xmath830\in x{\psi_{\mathrm{bc}}}^m(x)$ ] , resp . @xmath830\in x{\psi_{{\mathrm b}}}^m(x)$ ] .
the lemma is an immediate consequence of @xmath831 having a commutative normal operator ; see @xcite for a detailed discussion and ( * ? ? ?
* section 2 ) for a brief explanation .
_ for simplicity of notation we state the results below through lemma [ lemma : b - to-0-conversion - ps ] for @xmath788 ; they work equally well if one replaces @xmath788 by @xmath789 throughout . _ the analogue of lemma [ lemma : diffb - commutant ] with @xmath108 replaced by @xmath788 still holds , without the awkward restriction on positivity of b - orders ( which is simply due to the lack of non - trivial negative order differential operators ) .
let @xmath832 be the ( complex ) vector space of operators on @xmath205 of the form @xmath833 where the sum is locally finite , and let @xmath834 we define @xmath835 similarly , by replacing @xmath788 by @xmath789 throughout the definition .
the ring structure ( even with a weight @xmath494 ) of @xmath282 was proved in ( * ? ? ?
* corollary 4.4 and lemma 4.5 ) , which we recall here .
we add to the statements of ( * ? ? ?
* corollary 4.4 and lemma 4.5 ) that @xmath282 is also closed under adjoints with respect to any weighted non - degenerate b - density , in particular with respect to a non - degenerate 0-density such as @xmath836 , for both @xmath109 and @xmath788 are closed under these adjoints and @xmath837 .
[ lemma : psib - filtered - ring ] @xmath282 is a filtered * -ring under composition ( and adjoints ) with @xmath838 and @xmath839 where the adjoint is taken with respect to a ( i.e. _ any fixed _ ) non - degenerate 0-density .
moreover , composition is commutative to leading order in @xmath500 , i.e. for @xmath501 as above , @xmath502 , @xmath503\in{\mathrm{diff}_0}^{k+k'-1}{\psi_{{\mathrm b}}}^{m+m'}(x).\ ] ] just like for differential operators , we again have a lemma that improves the b - order ( rather than merely the 0-order ) of the commutator provided one of the commutants is in @xmath788 .
again , it is crucial here that there are no weights on @xmath788 .
[ lemma : b - comm - improve ] for @xmath840 , @xmath841 , @xmath829\in{\mathrm{diff}_0}^k{\psi_{{\mathrm b}}}^{s+m-1}(x)$ ] .
expanding elements of @xmath842 as finite sums of products of vector fields and functions , and using that @xmath788 is commutative to leading order , we need to consider commutators @xmath843 $ ] , @xmath509 , @xmath840 and show that this is in @xmath844 , which is automatic as @xmath845 , as well as @xmath846 $ ] , @xmath486 , @xmath840 , and show that this is in @xmath847 , i.e. @xmath848=\sum w_jb_j+c_j,\
b_j , c_j\in{\psi_{{\mathrm b}}}^{s-1}(x),\ w_j\in{\mathcal{v}_0}(x).\ ] ] but @xmath488 , @xmath487 , and @xmath849=\sum_j w'_j b'_j+c'_j,\ w'_j\in{\mathcal{v}}(x),\ b'_j , c'_j\in{\psi_{{\mathrm b}}}^{s-1}(x),\ ] ] see ( * ? ? ?
* lemma 2.2 ) , while @xmath850x^{-1}\in { \psi_{{\mathrm b}}}^{s-1}(x)$ ] , so @xmath848=[x , a]v'+x[v',a]=b''(xv')+\sum_j ( xw'_j ) b'_j+xc'_j,\ ] ] which is of the desired form once the first term is rearranged using lemma [ lemma : psib - filtered - ring ] , i.e. explicitly @xmath851 $ ] , with the last term being an element of @xmath844 .
we also have an analogue of lemma [ lemma : b - to-0-conversion ] .
[ lemma : b - to-0-conversion - ps ] for any @xmath852 integer , @xmath853 it suffices to show that @xmath854 , for the rest follows by induction .
also , we may localize and assume that @xmath661 is supported in a coordinate patch ; note that @xmath855 since @xmath856 . thus , suppose @xmath167 . then there exist @xmath857 , @xmath858 , and @xmath859 such that @xmath860 indeed , one simply needs to use the ellipticity of @xmath861 to achieve this by constructing a parametrix @xmath862 to it , and writing @xmath863 , @xmath864 .
as @xmath865 , the conclusion follows . as a consequence of our results
thus far , we deduce that @xmath866 is bounded on @xmath867 , already stated in ( * ? ? ?
* lemma 4.7 ) .
[ prop : h10-bd ] suppose @xmath868 .
any @xmath869 with compact support defines a bounded operator on @xmath867 , with operator norm bounded by a seminorm of @xmath661 in @xmath870 . for @xmath123
this is a special case of ( * ? ? ?
* lemma 4.7 ) , though the fact that the operator norm is bounded by a seminorm of @xmath661 in @xmath870 was not explicitly stated there though follows from the proof ; @xmath125 follows by duality .
for the convenience of the reader we recall the proof in the case we actually use in this paper , namely @xmath871 ( then @xmath872 follows by duality ) .
any @xmath661 as in the statement of the proposition is bounded on @xmath157 with the stated properties .
thus , we need to show that if @xmath486 , then @xmath873 . but @xmath874 $ ] , @xmath846\in{\mathrm{diff}_0}^1{\psi_{{\mathrm b}}}^{-1}(x)\subset{\psi_{{\mathrm b}}}^0(x)$ ] , hence @xmath875 and @xmath846:l^2(x)\to l^2(x)$ ] , with the claimed norm behavior .
if @xmath163 is a homogeneous function on @xmath876 , then we again consider the hamilton vector field @xmath877 associated to it on @xmath60 .
a change of coordinates calculation shows that in the b - canonical coordinates given above @xmath878 so @xmath877 extends to a @xmath63 vector field on @xmath876 which is tangent to @xmath879 .
if @xmath880 , @xmath881 , then @xmath882\in{\psi_{{\mathrm b}}}^{m+m'-1}(x)$ ] has principal symbol @xmath883)=\frac{1}{\imath}{\mathsf{h}}_q p.\ ] ] using proposition [ prop : h10-bd ] we can define a meaningful @xmath56 relative to @xmath200 .
first we recall the definition of the corresponding global function space from ( * ? ? ?
* section 4 ) : for @xmath116 the b - sobolev spaces relative to @xmath884 are given by .
on the other hand , we are working on a non - compact space , so we must consider local spaces and spaces of compactly supported functions as in ( * ? ? ?
* section 3 ) .
note also that we reversed the index convention ( which index comes first ) relative to @xcite , to match the notation for the wave front sets . ]
@xmath885 these can be normed by taking any properly supported elliptic @xmath886 and letting @xmath887 although the norm depends on the choice of @xmath661 , for @xmath143 supported in a fixed compact set , different choices give equivalent norms , see
* section 4 ) for details in the 0-setting ( where supports are not an issue ) , and ( * ? ? ?
* section 3 ) for an analysis involving supports .
we also let @xmath888 be the subspace of @xmath889 consisting of @xmath890 such that for any @xmath891 , @xmath892 . here
it is also useful to have sobolev spaces with a negative amount of b - regularity , in a manner completely analogous to ( * ? ? ?
* definition 3.15 ) : [ def : h1m - neg ] let @xmath893 be an integer , @xmath894 , and @xmath895 be elliptic on @xmath896 with proper support .
we let @xmath897 be the space of all @xmath898 of the form @xmath899 with @xmath900 .
we let @xmath901 we also let @xmath888 be the space of all @xmath898 such that @xmath892 for all @xmath891 .
as discussed for analogous spaces in @xcite following definition 3.15 there , this definition is independent of the particular @xmath661 chosen , and different @xmath661 give equivalent norms for distributions @xmath143 supported in a fixed compact set @xmath133 .
moreover , we have suppose @xmath902 , @xmath183 .
any @xmath903 with compact support defines a bounded operator on @xmath904 , with operator norm bounded by a seminorm of @xmath457 in @xmath870 .
suppose @xmath116 first .
then for an @xmath886 as in the definition above , @xmath905 the first term on the right hand side is bounded in the desired manner due to proposition [ prop : h10-bd ] .
letting @xmath906 be a properly supported parametrix for @xmath661 so @xmath907 , @xmath864 , we have @xmath908 , with @xmath909 , @xmath910 , so @xmath911 by proposition [ prop : h10-bd ] , with @xmath437 bounded by a seminorm of @xmath457 .
this completes the proof if @xmath116 . for @xmath894 ,
let @xmath895 be as in the definition .
if @xmath899 , and @xmath912 is a parametrix for @xmath661 so @xmath913 , @xmath914 , hence @xmath915 now , @xmath916 so @xmath917 indeed , and choosing @xmath918 , @xmath919 so that @xmath920 shows the desired continuity , as well as that the operator norm of @xmath457 is bounded by a @xmath870-seminorm .
now we define the wave front set relative to @xmath889 .
we also allow negative a priori b - regularity relative to this space .
[ def : wfb ] suppose @xmath921 , @xmath902 , @xmath183 .
then @xmath170 is _ not _ in @xmath922 if there is an @xmath172 such that @xmath173 is invertible and @xmath923 for all @xmath121 , i.e. if @xmath924 .
moreover , @xmath170 is _ not _ in @xmath925 if there is an @xmath167 such that @xmath173 is invertible and @xmath926 .
proposition [ prop : h10-bd ] implies that @xmath789 acts microlocally , i.e. preserves @xmath56 ; see ( * ? ? ?
* section 3 ) for a similar argument . in particular , the proofs for both the qualitative and quantitative version of microlocality go through without any significant changes ; one simply replaces the use of ( * ? ? ?
* lemma 3.2 ) by proposition [ prop : h10-bd ] .
* lemma 3.9)[lemma : wfb - mic ] suppose that @xmath927 , @xmath928 .
then @xmath929 . as in (
* section 3 ) , the wave front set microlocalizes the ` b - singular support relative to @xmath889 ' , meaning : ( cf .
* lemma 3.10 ) ) [ lemma : microloc - to - loc ] suppose @xmath930 , @xmath931 .
if @xmath932 , then in a neighborhood of @xmath99 , @xmath143 lies in @xmath933 , i.e. there is @xmath891 with @xmath934 near @xmath99 such that @xmath935 .
* corollary 3.11 ) ) [ cor : wf - to - h1 ] if @xmath930 and @xmath936 , then @xmath937 . in particular , if @xmath930 and @xmath936 for all @xmath159 , then @xmath938 , i.e. @xmath143 is conormal in the sense that @xmath926 for all @xmath939 ( or indeed @xmath940 ) . finally , we have the following quantitative bound for which we recall the definition of the wave front set of bounded subsets of @xmath941 : ( cf .
* definition 3.12 ) ) suppose that @xmath942 is a bounded subset of @xmath941 , and @xmath943 .
we say that @xmath944 if there is some @xmath940 which is elliptic at @xmath163 such that @xmath945 is a bounded subset of @xmath946 .
* lemma 3.13 , lemma 3.18 ) ) [ lemma : wfb - mic - q ] suppose that @xmath947 is compact , and @xmath565 a neighborhood of @xmath133 in @xmath896 .
let @xmath948 compact , and @xmath949 be a neighborhood of @xmath950 in @xmath3 with compact closure .
let @xmath951 be elliptic on @xmath133 with @xmath952 , with schwartz kernel supported in @xmath953 .
let @xmath942 be a bounded subset of @xmath941 with @xmath954 and schwartz kernel supported in @xmath953 .
then for any @xmath955 there is a constant @xmath142 such that for @xmath956 , @xmath957 with @xmath958 , @xmath959 we can use this lemma to obtain uniform bounds for pairings .
we call a subset @xmath942 of @xmath960 bounded if its elements are locally finite linear combinations of a fixed , locally finite , collection of elements of @xmath961 with coefficients that lie in a bounded subset of @xmath962 .
[ cor : wfb - mic - q ] suppose that @xmath947 is compact , and @xmath565 a neighborhood of @xmath133 in @xmath896 .
let @xmath948 compact , and @xmath949 be a neighborhood of @xmath950 in @xmath3 with compact closure .
let @xmath951 be elliptic on @xmath133 with @xmath952 , with schwartz kernel supported in @xmath953 .
let @xmath942 be a bounded subset of @xmath963 with @xmath954 and schwartz kernel supported in @xmath953 .
then there is a constant @xmath142 such that for @xmath956 , @xmath964 with @xmath965 , @xmath966 using lemma [ lemma : psib - filtered - ring ] we can write @xmath457 as @xmath967 , where @xmath968 , @xmath969 ( which we take to be elliptic on @xmath133 , but such that @xmath970 is elliptic on @xmath971 ) , @xmath972 lies in a bounded subset @xmath973 of @xmath974 and the sum is finite
. then @xmath975 where in the last step we used lemma [ lemma : wfb - mic - q ] .
it is useful to note that infinite order b - regularity relative to @xmath131 and @xmath200 are the same .
[ lemma : wfb - infty - independence ] for @xmath138 , @xmath976 the complements of the two sides are the set of points @xmath943 for which there exist @xmath172 ( with compactly supported schwartz kernel , as one may assume ) such that @xmath173 is invertible and @xmath977 , resp . @xmath978 . since @xmath979
, @xmath980 follows immediately . for the converse ,
if @xmath978 for all @xmath981 , then in particular @xmath982 shows that @xmath983 for @xmath984 and @xmath981 , so @xmath977 , i.e. @xmath985 , completing the proof .
we finally recall that @xmath986 , i.e. that @xmath143 is conormal relative to @xmath987 , means that @xmath988 for all @xmath981 , so in particular @xmath989 .
thus , @xmath990 in view of @xmath991 .
we recall here the structure of the compressed characteristic set and from ( * ? ? ?
* section 1-section 2 ) . in that paper @xmath3 is a manifold with corners , and @xmath118 is the codimension of the highest codimension corner in the local coordinate chart .
thus , for application to the present paper , the reader should take @xmath992 when referring to ( * ? ? ?
* section 1-section 2 ) .
it is often convenient to work on the cosphere bundle , here @xmath896 , which is equivalent to working on conic subsets of @xmath169 . in a region
where , say , @xmath993 @xmath142 fixed , we can take @xmath994 as ( projective ) local coordinates on @xmath169 , hence @xmath995 as local coordinates on the image of this region under the quotient map in @xmath896 ; cf .
* equation ( 1.4 ) ) .
first , we choose local coordinates more carefully . in arbitrary
local coordinates @xmath996 on a neighborhood @xmath997 of a point on @xmath135 , so that @xmath14 is given by @xmath246 inside @xmath998 , any symmetric bilinear form on @xmath66 can be written as @xmath999 with @xmath1000 smooth . in view of ,
using @xmath5 given there and coordinates @xmath1001 on @xmath14 pulled by to a collar neighborhood of @xmath14 by the product structure , we have in addition @xmath1002 for all @xmath1003 , and @xmath1004 is lorentzian for all @xmath1003 .
below we write covectors as @xmath1005 thus , @xmath1006 and hence the metric function , @xmath1007 is @xmath1008 since @xmath1009 , @xmath14 is indeed time - like in the sense that the restriction of the dual metric @xmath24 to @xmath25 is negative definite , for locally the conormal bundle @xmath25 is given by @xmath1010 we write @xmath1011 for the metric function on the boundary .
also , from , @xmath1012 where @xmath46 is a @xmath63 vector field in @xmath1013 , and @xmath548 is a @xmath63 function on @xmath1014 .
it is sometimes convenient to improve the form of @xmath457 near a particular point @xmath1015 , around which the coordinate system is centered .
namely , as @xmath457 is lorentzian , we can further arrange , by adjusting the @xmath1001 coordinates , @xmath1016 we now recall from the introduction that @xmath84 is the natural map corresponding to the identification of a section of @xmath66 as a section of @xmath73 , and in local coordinates @xmath91 is given by @xmath1017 moreover , the image of the characteristic set @xmath1018 , given by @xmath1019 under @xmath91 is the compressed characteristic set , @xmath1020 note that gives that @xmath1021 in particular , in view of , @xmath1022 lies in the region , at least after we possibly shrink @xmath997 ( recall that @xmath1013 ) , as we assume from now .
we also remark that , using , @xmath1023 and correspondingly @xmath1024 as we already noted , @xmath1025 can not vanish on @xmath1026 , so @xmath1027 in order to better understand the generalized broken bicharacteristics for @xmath1028 , we divide @xmath64 into two subsets .
we thus define the _ glancing set _ @xmath1029 as the set of points in @xmath64 whose preimage under @xmath1030 consists of a single point , and define the _ hyperbolic set _ @xmath1031 as its complement in @xmath64 .
thus , @xmath1032 as @xmath91 is a diffeomorphism on @xmath89 , while @xmath1033 lies in @xmath1029 if and only if on @xmath1034 , @xmath1035 . more explicitly , with the notation of
, @xmath1036 thus , @xmath1029 corresponds to generalized broken bicharacteristics which are tangent to @xmath14 in view of the vanishing of @xmath1037 at @xmath1038 ( recall that the @xmath1039 component of @xmath59 is @xmath1040 ) , while @xmath1031 corresponds to generalized broken bicharacteristics which are normal to @xmath14 .
note that if @xmath14 is one - dimensional ( hence @xmath3 is 2-dimensional ) , then @xmath1041 necessarily implies @xmath1042 , so in fact @xmath1043 , hence there are no glancing rays .
we next make the role of @xmath1029 and @xmath1031 more explicit , which explains the relevant phenomena better .
a characterization of , which is equivalent to definition [ def : gen - br - bich ] , is ( see the discussion in ( * ? ? ?
* section 1 ) after the statement of definition 1.1.)[lemma : gen - br - bichar ] a continuous map @xmath93 , where @xmath1044 is an interval , is a ( in the analytic sense that we use here ) if and only if it satisfies the following requirements : 1 .
if @xmath1045 then for all @xmath95 , @xmath1046 2 . if @xmath1047 then there exists @xmath356 such that @xmath1048 the idea of
the proof of this lemma is that at @xmath1029 , the requirement in ( i ) is equivalent to definition [ def : gen - br - bich ] since @xmath1049 contains a single point . on
the other hand , at @xmath1031 , the requirement in ( ii ) follows from definition [ def : gen - br - bich ] applied to the functions @xmath1050 , using , to conclude that @xmath1051 is strictly decreasing at @xmath1031 along .
since on @xmath1052 , one has @xmath1053 , for a @xmath1054 through @xmath1055 , on a punctured neighborhood of @xmath1056 , @xmath1057 , so @xmath1058 ( as @xmath1059 ) .
for the converse direction at @xmath1031 we refer to @xcite ; see ( * ? ? ?
* section 1 ) for details .
we first note the form of @xmath1028 with commutator calculations in mind . note that rather than thinking of the tangential terms , @xmath1060 , as ` too degenerate ' , we think of @xmath831 as ` too singular ' in that it causes the failure of @xmath1028 to lie in @xmath1061 .
this makes the calculations rather analogous to the conformal case , and also it facilitates the use of the symbolic machinery for b-ps.d.os .
[ prop : box - form ] on a collar neighborhood of @xmath14 , @xmath1028 has the form @xmath1062 with @xmath1063 where @xmath1064 is the dalembertian of the conformal metric on the boundary ( extended to a neighborhood of @xmath14 using the collar structure ) . writing the coordinates as @xmath550 ,
the operator @xmath1065 is given by @xmath1066 with adjoints taken with respect to @xmath1067 . with @xmath1068 , @xmath1069 for @xmath230 , this can be rewritten as @xmath1070 as @xmath1071 , we may take @xmath1072 and conclude that @xmath1073 . as @xmath1074 , taking @xmath1075 and @xmath1076 , @xmath1077 follow .
finally , @xmath1078 and modulo @xmath1079 , we can pull out the factors of @xmath5 and restrict @xmath1080 to @xmath14 , so @xmath1081 differs from @xmath1082 by an element of @xmath1079 , completing the proof .
we next state the lemma regarding dirichlet form which is of fundamental use in both the elliptic and hyperbolic / glancing estimates . below the main assumption is that @xmath179 , with @xmath1065 as in .
we first recall the notation for local norms : [ rem : localize ] since @xmath3 is non - compact and our results are microlocal , we may always fix a compact set @xmath948 and assume that all ps.d.os have schwartz kernel supported in @xmath953 .
we also let @xmath949 be a neighborhood of @xmath950 in @xmath3 such that @xmath949 has compact closure , and use the @xmath1083 norm in place of the @xmath200 norm to accommodate @xmath138 .
( we may instead take @xmath1084 identically @xmath336 in a neighborhood of @xmath950 , and use @xmath1085 . )
below we use the notation @xmath1086 for @xmath1087 to avoid having to specify @xmath949 .
we also use @xmath1088 for @xmath1089 .
* lemma 4.2 ) ) [ lemma : dirichlet - form ] suppose that @xmath947 is compact , @xmath1090 is open , @xmath803 .
suppose that @xmath1091\}$ ] is a bounded family of ps.d.os in @xmath1092 with @xmath1093 , and with @xmath1094 for @xmath1095 $ ] .
then there are @xmath1096 , @xmath1097 with @xmath1098 and @xmath389 such that for @xmath1099 $ ] , @xmath182 ( here @xmath1100 ) with @xmath1101 , @xmath1102 , we have @xmath1103 the point of this lemma is @xmath428 is @xmath1104 order lower ( @xmath1105 vs. @xmath1106 ) than the _ family _ @xmath1107 .
we will later take a limit , @xmath1108 , which gives control of the dirichlet form evaluated on @xmath1109 , @xmath1110 , in terms of lower order information .
the role of @xmath1111 , @xmath1112 , is to regularize such an argument , i.e. to make sure various terms in a formal computation , in which one uses @xmath1113 directly , actually make sense . the main difference with ( * ? ? ?
* lemma 4.2 ) is that @xmath6 is _ not negligible_. then for @xmath1099 $ ] , @xmath1114 , so @xmath1115 here the right hand side is the pairing of @xmath1116 with @xmath200 .
writing @xmath1117 $ ] , the right hand side can be estimated by @xmath1118 the second term is bounded by @xmath1119 . _
( recall that the ` local ' norms were defined in remark [ rem : localize ] . ) _ the first term is straightforward to estimate .
let @xmath1120 be elliptic with @xmath1121 a parametrix , so @xmath1122 then @xmath1123 since @xmath1124 is uniformly bounded in @xmath1125 , and @xmath1126 is uniformly bounded in @xmath1127 , @xmath1128 is uniformly bounded , with a bound like using cauchy - schwartz and lemma [ lemma : wfb - mic - q ] .
indeed , by lemma [ lemma : wfb - mic - q ] , choosing any @xmath1096 which is elliptic on @xmath133 , there is a constant @xmath1129 such that @xmath1130 similarly , by lemma [ lemma : wfb - mic - q ] and its analogue for @xmath1131 , choosing any @xmath1097 which is elliptic on @xmath133 , there is a constant @xmath1132 such that @xmath1133 . combining these gives , with @xmath1134 , @xmath1135 as desired .
a similar argument , using that @xmath1111 is uniformly bounded in @xmath1125 ( in fact in @xmath1136 ) , and @xmath1137 is uniformly bounded in @xmath1127 ( in fact in @xmath1138 ) , shows that @xmath1139 is uniformly bounded .
now we turn to the second term in , whose uniform boundedness is a direct consequence of lemma [ lemma : b - comm - improve ] and corollary [ cor : wfb - mic - q ] .
indeed , by lemma [ lemma : b - comm - improve ] , @xmath1140 $ ] is a bounded family in @xmath1141 , hence @xmath1142 $ ] is a bounded family in @xmath1143
. then one can apply corollary [ cor : wfb - mic - q ] to conclude that @xmath1144u , u\rangle\leq c'(\|u\|^2_{h^{1,k}_{0,{{\mathrm b}},{\mathrm{loc}}}(x)}+\|gu\|^2_{h^1(x)}),\ ] ] proving the lemma .
a more precise version , in terms of requirements on @xmath474 , is the following . here , as in section [ sec : poincare ] , we fix a positive definite inner product on the fibers of @xmath283 ( i.e. a riemannian 0-metric ) to compute @xmath1145 ; as @xmath1146 has support in a compact set below , the choice of the inner product is irrelevant .
[ lemma : dirichlet - form-2 ] ( cf .
* lemma 4.4 ) ) suppose that @xmath947 is compact , @xmath1090 is open , @xmath803 .
suppose that @xmath1091\}$ ] is a bounded family of ps.d.os in @xmath1092 with @xmath1093 , and with @xmath1094 for @xmath1095 $ ] .
then there are @xmath1096 , @xmath1147 with @xmath1098 and @xmath389 such that for @xmath356 , @xmath1099 $ ] , @xmath182 ( @xmath1100 ) with @xmath1101 , @xmath1148 , we have @xmath1149 the point of this lemma is that on the one hand the new term @xmath1150 can be absorbed in the left hand side in the elliptic region , hence is negligible , on the other hand , there is a gain in the order of @xmath1151 ( @xmath1106 , versus @xmath1152 in the previous lemma ) .
we only need to modify the previous proof slightly .
thus , we need to estimate the term @xmath1153 in differently , namely @xmath1154 now the lemma follows by using lemma [ lemma : wfb - mic - q ] and the remark following it , namely choosing any @xmath1147 which is elliptic on @xmath133 , there is a constant @xmath1132 such that @xmath1155 , then using the poincar inequality to estimate @xmath1156 by @xmath1157 , and finishing the proof exactly as for lemma [ lemma : dirichlet - form ] .
we next state microlocal elliptic regularity . note that for this result the restrictions on @xmath29 are weak ( only a half - line is disallowed ) , but on the other hand , a solution @xmath143 satisfying our hypotheses may not exist for values of @xmath6 when @xmath1158 .
( microlocal elliptic regularity.)[prop : elliptic ] suppose that @xmath466 , @xmath1159 and @xmath180 or @xmath181 .
suppose @xmath1160 for some @xmath1100 .
then @xmath1161 we first prove a slightly weaker result in which @xmath1162 is replaced by @xmath1163 we rely on lemma [ lemma : dirichlet - form ] .
we then prove the original statement using lemma [ lemma : dirichlet - form-2 ] .
suppose that @xmath1164 .
we may assume iteratively that @xmath1165 ; we need to prove then that @xmath1166 provided @xmath1167 ( note that the inductive hypothesis holds for @xmath1168 since @xmath182 ) .
we use local coordinates @xmath103 as in section [ sec : gbb ] , centered so that @xmath1169 , arranging that holds , and further group the variables as @xmath1170 , and hence the b - dual variables @xmath1171 .
we denote the euclidean norm by @xmath1172 .
let @xmath1173 be such that @xmath1174 and have @xmath814 in a small conic neighborhood @xmath565 of @xmath163 so that for a suitable @xmath142 or @xmath356 , in @xmath565 1 .
@xmath1175 if @xmath1176 , 2 .
@xmath1177 for all @xmath1178 , and @xmath1179 , if @xmath1180 and @xmath1181 .
let @xmath1182 for @xmath1112 , such that @xmath1183\}$ ] is a bounded family in @xmath866 , and @xmath1184 as @xmath1108 in @xmath1185 , @xmath1186 , e.g. the symbol of @xmath1187 could be taken as @xmath1188 .
let @xmath1189 .
let @xmath77 be the symbol of @xmath661 , and let @xmath1111 have symbol @xmath1190 , @xmath1112 , so @xmath1191 for @xmath1112 , and @xmath1111 is uniformly bounded in @xmath1136 , @xmath1192 in @xmath1193 . by lemma [ lemma : dirichlet - form ] , @xmath1194 is uniformly bounded for @xmath1099 $ ] , so @xmath1195 are uniformly bounded .
if @xmath193 , then taking the imaginary part at once shows that @xmath1196 is in fact uniformly bounded . on the other hand , whether @xmath198 or not , @xmath1197 using that @xmath1198 , we see that if @xmath1111 is supported in @xmath1199 , @xmath1200 for all @xmath1178 , then for some @xmath142 ( independent of @xmath1111 ) , @xmath1201 with analogous estimates and @xmath1202 if @xmath1203 , @xmath1204 if @xmath1205 , @xmath1206 if @xmath1207 . ] for @xmath1208 and for @xmath1209 .
thus , there exists @xmath1210 and @xmath591 such that if @xmath1211 and @xmath661 is supported in @xmath1212 and @xmath1213 then @xmath1214 now we distinguish the cases @xmath1180 and @xmath1176 . if @xmath1180 , we choose @xmath1215 , @xmath1211 , so that @xmath1216 on a neighborhood of @xmath814 , which is possible in view of ( ii ) at the beginning of the proof
. then the second integral on the left hand side of can be written as @xmath1217 , with the symbol of @xmath457 given by @xmath1218 ( which is @xmath1219 ) , modulo a term @xmath1220 but @xmath1221 is uniformly bounded in @xmath1222 , so this expression is uniformly bounded as @xmath1108 by corollary [ cor : wfb - mic - q ] .
we thus deduce that @xmath1223 is uniformly bounded as @xmath1108 .
if @xmath1176 , and @xmath661 is supported in @xmath1212 , @xmath1224 on the other hand , near @xmath1225 , for @xmath624 sufficiently small , @xmath1226 with the symbol of @xmath457 given by @xmath1227 ( which does not vanish on @xmath565 for @xmath624 small ) , while @xmath1228 , so the second term on the right hand side is uniformly bounded as @xmath1108 just as above .
we thus deduce in this case that @xmath1229 is uniformly bounded as @xmath1108 .
if @xmath193 then we already saw that @xmath1230 is uniformly bounded , so we deduce that @xmath1231 if @xmath198 , but @xmath20 , then the poincar inequality allows us to reach the same conclusion , since on the one hand in case ( ii ) @xmath1232 resp . in case ( i ) @xmath1233
is uniformly bounded , on the other hand by proposition [ prop : poincare - no - weight ] , for @xmath624 sufficiently small there exists @xmath1234 such that @xmath1235 correspondingly there are sequences @xmath1236 , @xmath1237 , @xmath1238 , weakly convergent in @xmath157 , and such that @xmath1239 , as @xmath1240 . since they converge to @xmath174 , @xmath1241 , @xmath1242 , respectively , in @xmath703 , we deduce that the weak limits are @xmath174 , @xmath1241 , @xmath1242 , which therefore lie in @xmath157 .
consequently , that @xmath1166 , hence proving the proposition with @xmath1162 replaced by @xmath1163 . to obtain the optimal result , we note that due to lemma [ lemma : dirichlet - form-2 ] we still have , for any @xmath356 , that @xmath1243 is uniformly bounded above for @xmath1099 $ ] . by arguing just as above , with @xmath457 as above , for sufficiently small @xmath356
, the right hand side gives an upper bound for @xmath1244 which is thus uniformly bounded as @xmath1108 .
the proof is then finished exactly as above .
we remark that the analogous argument works for the conformally compact elliptic problem , i.e. on asymptotically hyperbolic spaces , to give that for @xmath1159 , local solutions of @xmath1245 are actually conormal to @xmath14 provided they lie in @xmath200 locally , or indeed in @xmath1246 .
in this section we prove propagation of singularities for @xmath41 by positive commutator estimates .
we do so by first performing a general commutator calculation in proposition [ prop : commutator ] , then using it to prove rough propagation estimates first at hyperbolic , then at glancing points , in propositions [ prop : normal - prop ] , resp .
[ prop : tgt - prop ] . an argument
originally due to melrose and sjstrand @xcite then proves the main theorems , theorems [ thm : prop - sing ] and [ thm : prop - sing - im ] .
finally we discuss consequences of these results .
we first describe the form of commutators of @xmath41 with @xmath788 .
we state this as an analogue of ( * ? ? ?
* proposition 3.10 ) , and later in the section we follow the structure of @xcite as well . given proposition [ prop : commutator ] below , the proof of propagation of singularities proceeds with the same commutant construction as in @xcite , see also @xcite .
although it is in a setting that is more complicated in some ways , since it deals with the equation on differentials forms , we follow the structure of @xcite as it was written in a more systematic way than @xcite .
recall from the introduction that @xmath1051 is the b - dual variable of @xmath5 , @xmath1247 .
[ prop : commutator ] suppose @xmath1091\}$ ] is a family of operators @xmath1248 uniformly bounded in @xmath1125 , of the form @xmath1249 , @xmath172 , @xmath1250 , @xmath1251 . then @xmath1252=(xd_x)^ * c_r^\sharp(xd_x)+(xd_x)^ * xc_r'+xc_r''(xd_x ) + x^2c_r^\flat,\ ] ] with @xmath1253;{\psi_{\mathrm{bc}}}^{2s}(x ) ) , \ c'_r , c''_r\in l^\infty((0,1];{\psi_{\mathrm{bc}}}^{2s+1}(x ) ) , \
c_r^\flat\in { \psi_{\mathrm{bc}}}^{2s+2}(x),\ ] ] and @xmath1254 with @xmath1255 uniformly bounded in @xmath1256 respectively , @xmath1257 smooth and homogeneous of degree @xmath1258 respectively on @xmath169 , @xmath1259 and @xmath1260 annihilate @xmath1051 and @xmath1261 we start by observing that , in proposition [ prop : box - form ] , @xmath1028 is decomposed into a sum of products of weighted b - operators , so analogously expanding the commutator , all calculations can be done in @xmath1262 for various values of @xmath511 . in particular , keeping in mind lemma [ lemma : xd_x - improvement ] ( which gives the additional order of decay ) , @xmath1263,\imath[a_r^*a_r,(xd_x)^ * ] \in l^\infty((0,1]_r , x{\psi_{{\mathrm b}}}^{2s+1}(x)),\ ] ] with principal symbol @xmath1264 . by this observation ,
all commutators with factors of @xmath831 or @xmath1265 in can be absorbed into the ` next term ' of , so @xmath1266 \alpha ( xd_x)$ ] is absorbed into @xmath1267 , @xmath1268 $ ] is absorbed into @xmath1269 , @xmath1266 m'$ ] and @xmath1270 $ ] are absorbed into @xmath1271 .
the principal symbols of these terms are of the desired form , i.e. after factoring out @xmath1272 , they are the result of a vector field applied to @xmath77 plus a multiple of @xmath77 , and this vector field is @xmath1273 in the case of the first two terms ( thus annihilates @xmath1051 ) , @xmath1274 in the case of the last two terms , which in view of @xmath1275 , shows that it actually does not affect @xmath1276 . next , @xmath1277(xd_x)$ ] can be absorbed into ( in fact taken equal to ) @xmath1278 with principal symbol of @xmath1279 given by @xmath1280 in local coordinates , thus again is of the desired form since the @xmath1281 term has a vanishing factor of @xmath5 preceding it . since @xmath1282,[a_r^*a_r , m'']$ ]
are uniformly bounded in @xmath1283 , the corresponding commutators can be absorbed into @xmath1269 , resp .
@xmath1267 , without affecting the principal symbols of @xmath1284 and @xmath1285 at @xmath14 , and possessing the desired form .
next , @xmath1286 , @xmath1287 , so @xmath1288 $ ] is uniformly bounded in @xmath1289 , and thus can be absorbed into @xmath1290 without affecting its principal symbol at @xmath14 and possessing the desired form
. finally , @xmath1291 \in x^2{\psi_{{\mathrm b}}}^{2s+2}(x)$ ] has principal symbol @xmath1292 , thus can be absorbed into @xmath1290 yielding the stated principal symbol at @xmath14 .
[ prop : normal - prop](normal , or hyperbolic , propagation . )
suppose that @xmath179 , @xmath1293 .
let @xmath1294 , and let @xmath1295 be the function defined in the local coordinates discussed above , and suppose that @xmath1160 for some @xmath1100 , @xmath1296 , @xmath1297 . if @xmath1298 and there exists a conic neighborhood @xmath565 of @xmath1299 in @xmath169 such that @xmath1300 then @xmath1301 . in fact ,
if the wave front set assumptions are relaxed to @xmath1302 ( @xmath1297 ) and the existence of a conic neighborhood @xmath565 of @xmath1299 in @xmath169 such that @xmath1303 then we can still conclude that @xmath1304 .
[ rem : normal - remark ] as follows immediately from the proof given below , in and , one can replace @xmath1305 by @xmath1306 , i.e. one has the conclusion for either direction ( backward or forward ) of propagation , _ provided one also switches the sign of @xmath195 , when it is non - zero , i.e. the assumption should be @xmath1307_. in particular , if @xmath198 , one obtains propagation estimates both along increasing and along decreasing @xmath1308 .
note that @xmath1308 is _ increasing _ along the of @xmath101 by .
thus , the hypothesis region , @xmath1309 , on the left hand side of , is _ backwards _ from @xmath1299 , so this proposition , roughly speaking , propagates regularity _
moreover , every neighborhood @xmath565 of @xmath1310 in @xmath64 contains an open set of the form @xmath1311 see ( * ? ? ?
* equation ( 5.1 ) ) .
note also that implies the same statement with @xmath565 replaced by any smaller neighborhood of @xmath1299 ; in particular , for the set , provided that @xmath648 is sufficiently small .
we can also assume by the same observation that @xmath1312 .
furthermore , we can also arrange that @xmath1313 on @xmath565 since @xmath1314 .
we write @xmath1315 for the rehomogenized version of @xmath16 , which is thus homogeneous of degree zero and bounded below by a positive constant on @xmath565 .
this proposition is the analogue of proposition 6.2 in @xcite , and as the argument is similar , we mainly emphasize the differences .
these enter by virtue of @xmath6 not being negligible and the use of the poincar inequality . in @xcite , one uses a commutant @xmath172 and weights @xmath1316 , @xmath1317 , uniformly bounded in @xmath1125 , @xmath1249 , in order to obtain the propagation of @xmath1318 with the notation of that paper , whose analogue is @xmath1318 here ( the difference is the space relative to which one obtains b - regularity : @xmath1319 in the previous paper , the zero - sobolev space @xmath200 here ) .
one can use _ exactly the same _ commutant as in @xcite
. then proposition [ prop : commutator ] lets one calculate @xmath1320 $ ] to obtain a completely analogous expression to equation ( 6.18 ) of @xcite in the hyperbolic case .
we also refer the reader to @xcite because , although it studies a more delicate problem , namely natural boundary conditions ( which are not scalar ) , the main ingredient of the proof , the commutator calculation , is written up exactly as above in proposition [ prop : commutator ] , see ( * ? ? ?
* proposition 3.10 ) and the way it is used subsequently in propositions 5.1 there . as in (
* proof of proposition 5.1 ) , we first construct a commutant by defining its scalar principal symbol , @xmath77 . this completely follows the scalar case , see ( * ? ? ? * proof of proposition 6.2 ) .
next we show how to obtain the desired estimate .
so , as in ( * ? ? ?
* proof of proposition 6.2 ) , let @xmath1321 with @xmath1322 denoting the euclidean norm . for @xmath356 , @xmath624 , with other restrictions to be imposed later on , let @xmath1323 let @xmath1324 be equal to @xmath317 on @xmath670 $ ] and @xmath1325 for @xmath1326 .
thus , @xmath1327 for @xmath1328 .
let @xmath668 be @xmath317 on @xmath670 $ ] , @xmath336 on @xmath669 , with @xmath1329 satisfying @xmath1330 .
finally , let @xmath1331 be supported in @xmath1332 $ ] , identically @xmath336 on @xmath1333 $ ] , where @xmath1334 is such that @xmath1335 in @xmath1336 .
thus , @xmath1337 is a cutoff in @xmath1338 , with its support properties ensuring that @xmath1339 is supported in @xmath1340 $ ] hence outside @xmath64 it should be thought of as a factor that microlocalizes near the characteristic set but effectively commutes with @xmath41 ( since we already have the microlocal elliptic result ) .
then , for @xmath1341 large , to be determined , let @xmath1342 so @xmath77 is a homogeneous degree zero @xmath63 function on a conic neighborhood of @xmath1299 in @xmath169 . indeed , as we see momentarily , for any @xmath356 , @xmath77 has compact support inside this neighborhood ( regarded as a subset of @xmath896 , i.e. quotienting out by the @xmath1343-action ) for @xmath648 sufficiently small , so in fact it is globally well - defined .
in fact , on @xmath1344 we have @xmath1345 and @xmath1346 .
since @xmath1347 , the first of these inequalities implies that @xmath1348 , so on @xmath1344 @xmath1349 hence , @xmath1350 in view of and , this shows that given any @xmath1351 there exists @xmath591 such that for any @xmath1352 and @xmath1353 , @xmath77 is supported in @xmath565 . the role that
@xmath1354 large plays ( in the definition of @xmath77 ) is that it increases the size of the first derivatives of @xmath77 relative to the size of @xmath77 , hence it allows us to give a bound for @xmath77 in terms of a small multiple of its derivative along the hamilton vector field , much like the stress energy tensor was used to bound other terms , by making @xmath478 large relative to @xmath411 , in the ( non - microlocal ) energy estimate . now let @xmath1355 with @xmath1356 , supported in the coordinate chart .
also let @xmath1187 be scalar , have symbol @xmath1357 so @xmath1358 for @xmath1112 and it is uniformly bounded in @xmath1125 .
then , for @xmath1112 , @xmath1359 u , u\rangle+\langle \imath(p - p^*)a^*_ra_r u , u\rangle\\ & = \langle \imath [ a^*_r a_r , p ] u , u\rangle-2{\operatorname{im}}\lambda\|a_r u\|^2 . \end{split}\ ] ] we can compute this using proposition [ prop : commutator ] .
we arrange the terms of the proposition so that the terms in which a vector field differentiates @xmath1360 are included in @xmath1361 , the terms in which a vector fields differentiates @xmath1362 are included in @xmath1363 .
thus , we have @xmath1364 with @xmath1365 where @xmath1366 , @xmath1367 , @xmath1368 and @xmath1369 as well as @xmath1370 and @xmath1371 are all smooth functions on @xmath876 , homogeneous of degree 0 ( independent of @xmath1372 and @xmath648 ) , and @xmath1373 is the rehomogenized version of @xmath16 .
moreover , @xmath1374 arise from when @xmath1375 is differentiated in @xmath1376 , and thus vanish when @xmath1377 , while @xmath1370 and @xmath1371 arise when @xmath1308 is differentiated in @xmath1376 , and comprise all such terms with the exception of those arising from the @xmath1378 component of @xmath1276 ( which gives @xmath1379 on the last line above ) hence are the sums of functions vanishing at @xmath246 ( corresponding to us only specifying the restrictions of the vector fields in at @xmath14 ) and functions vanishing at @xmath1380 ( when @xmath1381 in @xmath1382 is differentiated ) . in this formula
we think of @xmath1383 is positive near @xmath1299 .
compared to this , the terms with @xmath1384 are negligible , for they can all be bounded by latexmath:[\[c\digamma^{-1}(\digamma^{-1}\delta^{-1}w_r^2 a the vanishing condition on the @xmath1388 ensures that , on @xmath1344 , @xmath1389 is not too large , i.e. there exists @xmath1390 such that if @xmath1391 , the terms with @xmath1388 can be treated as error terms . on the other hand , we have @xmath1392 now , on @xmath64 , @xmath1393 ( for @xmath1394 with @xmath565 sufficiently small ) . thus we can write @xmath1395 with @xmath1396 supported away from @xmath64 and @xmath1397 satisfying @xmath1398 we can also obtain a similar decomposition for @xmath1399 . *
@xmath1366 , @xmath1367 , @xmath1368 and @xmath1369 are all smooth functions on @xmath876 , homogeneous of degree 0 , satisfying ( and are independent of @xmath1401 ) , * @xmath1402 , @xmath1403 , @xmath1404 and @xmath1371 are all smooth functions on @xmath876 , homogeneous of degree 0 , with @xmath1395 , @xmath1405 satisfying ( and are independent of @xmath1401 ) , while @xmath1406 is supported away from @xmath64 , * and @xmath1407 , @xmath1408 , @xmath1409 and @xmath1410 are all smooth functions on @xmath876 , homogeneous of degree 0 , uniformly bounded in @xmath1411 .
let @xmath1412 and let @xmath1413 with principal symbol @xmath1414 .
then let @xmath1415 where @xmath1416 is identically @xmath336 on @xmath565 considered as a subset of @xmath896 ; recall from remark [ rem : normal - remark ] that @xmath1417 is bounded below by a positive quantity here . if @xmath1418 with principal symbol @xmath1419 then we deduce from - that terms are included in @xmath1420 , while the @xmath1421 terms are included in @xmath1422 , and similarly for the other analogous terms in @xmath1367 , @xmath1368 , @xmath1369 .
moreover , in view of lemma [ lemma : b - comm - improve ] , we can freely rearrange factors , e.g. writing @xmath1423 as @xmath1424 , if we wish , with the exception of commuting powers of @xmath5 with @xmath831 or @xmath1265 since we need to regard the latter as elements of @xmath1425 rather than @xmath1426 .
indeed , the difference between rearrangements has lower b - order than the product , in this case being in @xmath1427 , which in view of lemma [ lemma : b - to-0-conversion - ps ] , at the cost of dropping powers of @xmath5 , can be translated into a gain in 0-order , @xmath1428 , with the result that these terms can be moved to the ` error term ' , @xmath1429 . ]
@xmath1430 with @xmath1431 with @xmath1432 ) \cap u$ ] , @xmath1433 , and with @xmath1434 , @xmath1435 , @xmath1436 , @xmath1437 , @xmath1438 this is almost completely analogous to ( * ? ? ?
* equation ( 6.18 ) ) with the understanding that each term of ( * ? ? ?
* equation ( 6.18 ) ) inside the parentheses attains an additional factor of @xmath1439 ( corresponding to @xmath1028 being in @xmath536 rather than @xmath1440 ) which we partially include in @xmath831 ( vs. @xmath1441 ) .
the only difference is the presence of the @xmath1442 term which however is treated like the @xmath1443 term for @xmath679 sufficiently large , hence the rest of the proof proceeds very similarly to that paper .
we go through this argument to show the role that @xmath6 and the poincar inequality play , and in particular how the restrictions on @xmath6 arise .
having calculated the commutator , we proceed to estimate the ` error terms ' @xmath1444 , @xmath1445 , @xmath1446 and @xmath1420 as operators .
we start with @xmath1444 . by the standard square root construction to prove the boundedness of ps.d.os on @xmath285 , see e.g. the discussion after ( * ? ? ?
* remark 2.1 ) , there exists @xmath1447 such that @xmath1448 for all @xmath1449 .
here @xmath1450 is the @xmath157-norm , as usual .
thus , we can estimate , for any @xmath355 , @xmath1451 now we turn to @xmath1445 .
let @xmath1452 be elliptic ( which we use to shift the orders of ps.d.os at our convenience ) , with symbol @xmath1381 on @xmath1344 , @xmath1453 a parametrix , so @xmath1454 , @xmath914 .
then there exists @xmath1455 such that @xmath1456 for all @xmath1457 with @xmath1458 , and similarly , there exists @xmath1459 such that @xmath1460 finally , there exists @xmath1461 such that @xmath1462 for all @xmath1457 with @xmath1458 .
thus , @xmath1463 and , writing @xmath1464 in the right factor , and taking the adjoint of @xmath1465 , @xmath1466 with @xmath1467 .
now , by , @xmath1468u , u\rangle & = \|cx\tilde b_r u\|^2 + \langle r^\flat x\tilde b_r u , x\tilde b_r u\rangle\\ & \qquad+\langle \tilde r '' xd_x \tilde b_r u , x\tilde b_r u\rangle + \langle x\tilde b_r u,(\tilde r')^*xd_x \tilde b_r u\rangle\\ & \qquad+\langle r^\sharp xd_x \tilde b_r u , xd_x\tilde b_r u\rangle\\ & \qquad+\langle r''_ru , u\rangle + \langle ( e_r+e'_r)u , u\rangle \end{split}\ ] ] on the other hand , this commutator can be expressed as in , so @xmath1469 so the sign of the first two terms agree if @xmath197 , and the @xmath195 term vanishes if @xmath6 is real .
assume for the moment that @xmath1470 this is certainly the case in our setup if @xmath1471 , but this assumption is a little stronger than @xmath1472 , which is what we need to assume for the second paragraph in the statement of the proposition .
we deal with the weakened hypothesis @xmath1472 at the end of the proof . returning to
, the utility of the commutator calculation is that we have good information about @xmath474 ( this is where we use that we have a microlocal solution of the pde ! ) .
namely , we estimate the left hand side as @xmath1473 since @xmath1474 is uniformly bounded in @xmath1475 , @xmath1476 is uniformly bounded in @xmath1127 , both with @xmath1477 in @xmath565 , with @xmath1478 , resp .
@xmath1479 disjoint from them , we deduce ( using lemma [ lemma : wfb - mic - q ] and its @xmath1480 analogue ) that @xmath1481 is uniformly bounded . similarly , taking into account that @xmath1482 is uniformly bounded in @xmath946 , we see that @xmath1483 is also uniformly bounded , so @xmath1484 is uniformly bounded for @xmath1099 $ ] .
thus , @xmath1485 all terms but the ones involving @xmath1486 or @xmath1054 ( not @xmath1487 ) remain bounded as @xmath1108 .
the @xmath1486 and @xmath1054 terms can be estimated by writing @xmath1488 for some @xmath1489 , and using lemma [ lemma : dirichlet - form ] and the poincar lemma where necessary .
namely , we use either @xmath193 or @xmath20 to control @xmath1490 and @xmath1491 in @xmath157 in terms of @xmath1492 where @xmath1493 ; this is possible by factoring @xmath1494 ( which is elliptic on @xmath1495 ) out of @xmath1496 modulo an error @xmath1497 bounded in @xmath1136 , which in turn can be incorporated into the ` error ' given by the right hand side of lemma [ lemma : dirichlet - form ] .
thus , there exists @xmath1498 , @xmath1096 , @xmath1097 as in lemma [ lemma : dirichlet - form ] such that @xmath1499 we further estimate @xmath1500 in terms of @xmath1501 and @xmath1502 using that @xmath437 is elliptic on @xmath1503 and lemma [ lemma : wfb - mic - q ] .
we conclude , using @xmath1298 , taking @xmath1372 sufficiently large , then @xmath1504 sufficiently small , and finally @xmath679 sufficiently large , that there exist @xmath355 , @xmath356 , @xmath591 and @xmath1505 , @xmath1506 such that for @xmath1353 , @xmath1507 letting @xmath1108 now keeps the right hand side bounded , proving that @xmath1500 is uniformly bounded as @xmath1108 , hence @xmath1508 ( cf . the proof of proposition [ prop : elliptic ] ) . in view of lemma [ lemma : dirichlet - form ] and
the poincar inequality ( as in the proof of proposition [ prop : elliptic ] ) this proves that @xmath1304 , and hence proves the first statement of the proposition .
in fact , recalling that we needed @xmath1509 for the uniform boundedness in , this proves a slightly weaker version of the second statement of the proposition with @xmath1510 replaced by @xmath1478 . for the more precise statement
we modify this is the only term in that needs modification to prove the optimal statement .
let @xmath1511 be elliptic , @xmath1512 a parametrix , @xmath1513 .
then , similarly to , we have for any @xmath355 , @xmath1514 the last term on the right hand side can be estimated as before .
as @xmath1515 is bounded in @xmath1516 with @xmath1477 disjoint from @xmath565 , we see that @xmath1517 is uniformly bounded .
moreover , @xmath1518 can be estimated , using lemma [ lemma : dirichlet - form ] and the poincar inequality , by @xmath1519 modulo terms that are uniformly bounded as @xmath1108 .
the principal symbol of @xmath1520 is @xmath1521 , with @xmath1522 , where @xmath319 stands for @xmath1523 , etc .
, so we can write : @xmath1524 where we used that @xmath1525 when @xmath1526 , while @xmath77 , @xmath1527 vanish otherwise . correspondingly , using that @xmath1528 is @xmath63 , homogeneous degree zero , near the support of @xmath77 in @xmath169 , we can write @xmath1529 , @xmath1530 , @xmath1531 .
thus , modulo terms that are bounded as @xmath1108 , @xmath1532 ( hence @xmath1518 ) can be estimated from above by @xmath1533 .
therefore , modulo terms that are bounded as @xmath1108 , for @xmath355 sufficiently small , @xmath1534 can be absorbed into @xmath1535 . as the treatment of the other terms on the right hand side of requires no change ,
we deduce as above that @xmath1508 , which ( in view of lemma [ lemma : dirichlet - form ] ) proves that @xmath1304 , completing the proof of the iterative step .
we need to make one more remark to prove the proposition for @xmath1536 , namely we need to show that the neighborhoods of @xmath1299 which are disjoint from @xmath1318 do not shrink uncontrollably to @xmath1537 as @xmath1538 .
this argument parallels to last paragraph of the proof of ( * ? ? ?
* proposition 24.5.1 ) .
in fact , note that above we have proved that the elliptic set of @xmath1539 is disjoint from @xmath1318 . in the next step , when we are proving @xmath1540 , we decrease @xmath624 slightly ( by an arbitrary small amount ) , thus decreasing the support of @xmath1541 in , to make sure that @xmath1542 is a subset of the elliptic set of the union of @xmath1543 with the region @xmath1544 , and hence that @xmath1545 .
each iterative step thus shrinks the elliptic set of @xmath1543 by an arbitrarily small amount , which allows us to conclude that @xmath1299 has a neighborhood @xmath1546 such that @xmath1547 for all @xmath1106 .
this proves that @xmath1548 , and indeed that @xmath1549 , for if @xmath167 with @xmath1550 then @xmath1551 by lemma [ lemma : wfb - mic ] and corollary [ cor : wf - to - h1 ] . before turning to tangential propagation
we need a technical lemma , roughly stating that when applied to solutions of @xmath475 , @xmath1552 , microlocally near @xmath1029 , @xmath831 and @xmath1553 are not merely bounded by @xmath1554 , but it is small compared to it , provided that @xmath1159 .
this result is the analogue of ( * ? ? ?
* lemma 7.1 ) , and is proved as there , with the only difference being that the term @xmath1555 can not be dropped , but it is treated just as in proposition [ prop : elliptic ] above . below
a @xmath648-neighborhood refers to a @xmath648-neighborhood with respect to the metric associated to any riemannian metric on the manifold @xmath73 , and we identify @xmath896 as the unit ball bundle with respect to some fiber metric on @xmath73 .
* lemma 7.1).)[lemma : dt - dx ] suppose that @xmath179 , @xmath1556 suppose @xmath182 , and suppose that we are given @xmath947 compact satisfying @xmath1557 then there exist @xmath591 and @xmath389 with the following property .
let @xmath1211 , @xmath1090 open in a @xmath648-neighborhood of @xmath133 , and @xmath1091\}$ ] be a bounded family of ps.d.os in @xmath1092 with @xmath1558 , and with @xmath1094 for @xmath1095 $ ] .
as @xmath133 is compact , this is essentially a local result . in particular
, we may assume that @xmath133 is a subset of @xmath73 over a suitable local coordinate patch .
moreover , we may assume that @xmath591 is sufficiently small so that @xmath1494 is elliptic on @xmath565 . by lemma [ lemma : dirichlet - form ] applied with @xmath133 replaced by @xmath1563 in the hypothesis ( note that the latter is compact ) , we already know that @xmath1564 for some @xmath1565 and for some @xmath428 , @xmath1151 as in the statement of the lemma .
freezing the coefficients at @xmath14 , as in the proof of proposition [ prop : elliptic ] , see ( * ? ? ?
* lemma 7.1 ) for details , we deduce that @xmath1566 now , one can show that @xmath1567 _ precisely _ as in the proof of ( * ? ? ?
* lemma 7.1 ) .
equations - imply with the left hand side replaced by @xmath1568 .
if @xmath193 , taking the imaginary part of @xmath1569 gives the desired bound for @xmath1570 , hence taking the real part gives the desired bound for @xmath1571 as well . if @xmath198 but @xmath20 , we finish the proof using the poincar inequality , cf . the proof of proposition [ prop : elliptic ] .
[ prop : tgt - prop](tangential , or glancing , propagation . )
suppose that @xmath179 , @xmath1293 .
let @xmath997 be a coordinate chart in @xmath3 , @xmath565 open with @xmath1572 .
let @xmath1160 for some @xmath1100 , and let @xmath1573 be the coordinate projection @xmath1574 given @xmath1575 compact with @xmath1576 there exist constants @xmath389 , @xmath591 such that the following holds . if @xmath1298 , @xmath1577 , @xmath1578 , @xmath1579 considered as a constant vector field in local coordinates , and for some @xmath1580 , @xmath1581 and for all @xmath1582 @xmath1583 then @xmath1301 .
[ rem : tgt - rem ] just like proposition [ prop : normal - prop ] , this proposition gives regularity propagation in the _ forward _ direction along @xmath1586 , i.e. in order to conclude regularity at @xmath1299 , one needs to know regularity in the _ backward _
@xmath1586-direction from @xmath1299 .
one can again change the direction of propagation , i.e. replace @xmath648 by @xmath1587 in @xmath1588 , provided one also changes the sign of @xmath195 to @xmath1307 .
in particular , if @xmath198 , one obtains propagation estimates in both the forward and backward directions .
again , the proof follows a proof in @xcite closely , in this case proposition 7.3 , as corrected at a point in @xcite , so we merely point out the main steps .
again , one uses a commutant @xmath172 and weights @xmath1316 , @xmath1317 , uniformly bounded in @xmath1125 , @xmath1249 , in order to obtain the propagation of @xmath1318 with the notation of that paper , whose analogue is @xmath1318 here ( the difference is the space relative to which one obtains b - regularity : @xmath1319 in the previous paper , the zero - sobolev space @xmath200 here ) .
one can use _ exactly the same _ commutants as in @xcite ( with a small correction given in @xcite )
. then proposition [ prop : commutator ] lets one calculate @xmath1320 $ ] to obtain a completely analogous expression to the formulae below equation ( 7.16 ) of @xcite , as corrected in @xcite ) .
the rest of the argument is completely analogous as well .
again , we refer the reader to @xcite because the commutator calculation is written up exactly as above in proposition [ prop : commutator ] , see ( * ? ? ?
* proposition 3.10 ) and it is used subsequently in 6.1 there the same way it needs to be used here any modifications are analogous to those in proposition [ prop : normal - prop ] and arise due to the non - negligible nature of @xmath6 . again , we first construct the symbol @xmath77 of our commutator following ( * ? ? ? * proof of proposition 7.3 ) as corrected in @xcite .
note that ( with @xmath1589 ) @xmath1590 and let @xmath1591 so @xmath420 is homogeneous of degree zero ( with respect to the @xmath1343-action on the fibers of @xmath1592 ) .
we use @xmath1593 now to measure propagation , since @xmath1594 at @xmath1299 by , so @xmath1595 is @xmath1596 at @xmath1299 . note that @xmath1597 is thus increasing along of @xmath2 .
first , we require @xmath1598 note that @xmath1599 at @xmath1299 for @xmath1600 there , but @xmath1601 , so @xmath1602 next , since @xmath1603 , @xmath1604 , hence @xmath1605 . with a slight abuse of notation ,
we also regard @xmath1299 as a point in @xmath1606 recall that @xmath1607 .
we can also regard @xmath420 as a vector field on @xmath1606 in view of its homogeneity . as @xmath420 does not vanish as a vector in @xmath1608 in view of @xmath1609 , @xmath1597 being homogeneous degree zero , hence a function on @xmath1606 , the kernel of @xmath420 in @xmath1610 has dimension @xmath1611 . thus there
exist @xmath1612 , @xmath1613 be homogeneous degree zero functions on @xmath1614 ( hence functions on @xmath1606 ) such that @xmath1615 by dimensional considerations , @xmath1616 , @xmath1617 , together with @xmath1618 span the cotangent space of @xmath1606 at @xmath1299 , i.e. of the quotient of @xmath1614 by the @xmath1343-action , so the @xmath1612 , together with @xmath1597 , can be used as local coordinates on a chart @xmath1619 near @xmath1299 .
we also let @xmath1620 be a neighborhood of @xmath1299 in @xmath896 such that @xmath1612 , together with @xmath1597 , @xmath5 and @xmath1621 are local coordinates on @xmath1620 ; this holds if @xmath1622 is identified with a subset of @xmath1623 and @xmath1620 is a product neighborhood of this in @xmath896 in terms of the coordinates .
note that as @xmath1624 on @xmath1625 , for points @xmath163 in @xmath64 , one can ensure that @xmath1621 is small by ensuring that @xmath1626 is close to @xmath1299 and @xmath1627 is small ; cf . the discussion around and after .
by reducing @xmath1620 if needed ( this keeps all previously discussed properties ) , we may also assume that it is disjoint from @xmath1584 . hence , @xmath1628 with @xmath1629 smooth , @xmath1630 , @xmath1613 .
then we extend @xmath1612 to a function on @xmath169 ( using the coordinates @xmath83 ) , and conclude that @xmath1631 with @xmath1632 smooth . similarly , @xmath1633 with @xmath1634 smooth .
let @xmath1635 finally , we let @xmath1636 and define @xmath77 by @xmath1637 with @xmath1638 and @xmath1362 as in the case of the normal propagation estimate , stated after .
we always assume @xmath362 , so on @xmath1344 we have @xmath1639 since @xmath1347 , the first of these inequalities implies that @xmath1640 , so on @xmath1344 @xmath1641 hence , @xmath1642 thus , @xmath1344 lies in @xmath1620 for @xmath624 sufficiently small .
moreover , on @xmath1643 , @xmath1644,\ \omega^{1/2 } \leq 2{\epsilon}\delta,\ ] ] so this region lies in after @xmath1372 and @xmath648 are both replaced by appropriate constant multiples , namely the present @xmath648 should be replaced by @xmath1645 .
we proceed as in the case of hyperbolic points , letting @xmath1355 with @xmath1356 , supported in the coordinate chart .
also let @xmath1187 be scalar , have symbol @xmath1357 so @xmath1358 for @xmath1112 and it is uniformly bounded in @xmath1125 .
then , for @xmath1112 , @xmath1646 u , u\rangle+\langle \imath(p - p^*)a^*_ra_r u , u\rangle\\ & = \langle \imath [ a^*_r a_r , p ] u , u\rangle-2{\operatorname{im}}\lambda\|a_r u\|^2 .
\end{split}\ ] ] and we compute the commutator here using proposition [ prop : commutator ] .
we arrange the terms of the proposition so that the terms in which a vector field differentiates @xmath1360 are included in @xmath1361 , the terms in which a vector fields differentiates @xmath1362 are included in @xmath1363 .
thus , we have @xmath1647 with @xmath1648 where @xmath1366 , @xmath1367 , @xmath1368 and @xmath1369 as well as @xmath1370 and @xmath1371 are all smooth functions on @xmath876 , homogeneous of degree 0 ( independent of @xmath1372 and @xmath648 ) .
moreover , @xmath1374 arise from when @xmath1375 is differentiated in @xmath1649 , while @xmath1370 and @xmath1371 arise when @xmath1597 is differentiated in @xmath1649 , and comprise all such terms with the exception of part of that arising from the @xmath1650 component of @xmath1276 ( which gives the @xmath1651 on the last line above , modulo a term included in @xmath1371 and which vanishes at @xmath1377 ) .
in addition , as @xmath1652 for any function @xmath417 , the terms @xmath1653 , @xmath1654 , have vanishing factors of @xmath1655 , resp .
@xmath5 , with the structure of the remaining factor dictated by the form of @xmath1656 , resp . @xmath1657 .
thus , using to compute @xmath1369 , to compute @xmath1371 , we have @xmath1658 with @xmath1659 , etc . ,
we deduce that @xmath1660 while @xmath1661 @xmath1662 , and @xmath1663 we remark that although thus far we worked with a single @xmath1664 , the same construction works with @xmath1299 in a neighborhood @xmath1665 of a fixed @xmath1666 , with a _ uniform _ constant @xmath437 . in view of the compactness of @xmath133 ,
this suffices ( by the rest of the argument we present below ) to give the uniform estimate of the proposition .
since - are exactly the same ( with slightly different notation ) as ( * ? ? ?
* equations ( 6.16)-(6.18 ) ) , the rest of the proof is analogous , except that ( * ? ? ?
* lemma 4.6 ) is replaced by lemma [ lemma : dt - dx ] here .
thus , for a small constant @xmath1667 to be determined , which we may assume to be less than @xmath437 , we demand below that the expressions on the right hand sides of are bounded by @xmath1668 , those on the right hand sides of are bounded by @xmath1669 , while those on the right hand sides of are bounded by @xmath1670 .
this demand is due to the appearance of two , resp .
one , resp .
zero , factors of @xmath831 in for the terms whose principal symbols are affected by these , taking into account that in view of lemma [ lemma : dt - dx ] we can estimate @xmath1671 by @xmath1672 if @xmath1146 is microlocalized to a @xmath1443-neighborhood of @xmath1029 , which is the case for us with @xmath1673 in terms of support properties of @xmath77 .
thus , recalling that @xmath1667 is to be determined , we require that @xmath1674 and @xmath1675 see ( * ? ? ?
* proposition 6.1 ) for motivation .
then with @xmath1676 satisfying and , hence @xmath1677 , - give that @xmath1678 while @xmath1679 @xmath1662 , and @xmath1680 as desired .
one deduces that @xmath1681 with @xmath1431 with @xmath1682 ) \cap\omega^{-1}([0,4\delta^2{\epsilon}^2))\subset \tilde{\mathcal{u}}\ ] ] ( cf . ) , @xmath1433 , and with @xmath1434 , @xmath1435 , @xmath1436 , @xmath1437 , @xmath1683 these are analogues of the result of the second displayed equation after ( * ? ? ?
* equation ( 7.16 ) ) , as corrected in @xcite , with the small ( at this point arbitrary ) constant @xmath1670 replacing some constants given there in terms of @xmath1372 and @xmath648 ; see ( * ? ? ? * equation ( 6.25 ) ) for estimates stated in exactly the same form in the form - valued setting .
the rest of the argument thus proceeds as in ( * ? ? ?
* proof of proposition 7.3 ) , taking into account @xcite , and using lemma [ lemma : dt - dx ] in place of ( * ? ? ?
* lemma 7.1 ) . since for @xmath6 real , @xmath20 , both forward and backward propagation
is covered by these two results , see remarks [ rem : normal - remark ] and [ rem : tgt - rem ] , we deduce our main result on the propagation of singularities : [ thm : prop - sing ] suppose that @xmath466 , @xmath20 , @xmath180 or @xmath181 .
suppose @xmath182 for some @xmath1100 .
then @xmath185 is a union of maximally extended generalized broken bicharacteristics of the conformal metric @xmath2 in @xmath186 the proof proceeds as in ( * ? ? ?
* proof of theorem 8.1 ) , since propositions [ prop : normal - prop ] and [ prop : tgt - prop ] are complete analogues of ( * ? ? ?
* proposition 6.2 ) and ( * ? ? ?
* proposition 7.3 ) . given the results of the preceding sections of @xcite , the argument of ( * ? ? ?
* proof of theorem 8.1 ) is itself only a slight modification of an argument originally due to melrose and sjstrand @xcite , as presented by lebeau @xcite ( although we do not need lebeau s treatment of corners here ) . for the convenience of the reader
we give a very sketchy version of the proof .
to start with , propagation of singularities has already been proved in @xmath11 ; this is the theorem of duistermaat and hrmander @xcite .
now , the theorem can easily be localized the global version follows by a zorn s lemma argument , see ( * ? ? ?
* proof of theorem 8.1 ) for details . indeed , in view of the duistermaat - hrmander result it suffices to show that if @xmath1684 then @xmath1685\to\dot\sigma,\ { \epsilon}_0>0,\\ & \qquad\qquad \gamma(0)=q_0 , \ \gamma(s)\in{\mathrm{wf}_{{\mathrm b}}}^{1,m}(u)\setminus{\mathrm{wf}_{{\mathrm b}}}^{-1,m+1}(pu),\ s\in[-{\epsilon}_0,0 ] , \end{split}\ ] ] for the existence of a on @xmath1686 $ ] can be demonstrated similarly by replacing the forward propagation estimates by backward ones , and , directly from definition [ def : gen - br - bich ] , piecing together the two s gives one defined on @xmath1687 $ ] .
note that by microlocal elliptic regularity , proposition [ prop : elliptic ] , implies that @xmath1688 .
now suppose @xmath1689 .
we use the notation of proposition [ prop : normal - prop ] .
then @xmath1054 in is constructed by taking a sequence @xmath1690 , @xmath1691 with @xmath1692 and @xmath1693\to \dot\sigma$ ] with @xmath1694 and with @xmath1695 for @xmath1696 $ ] .
once this is done , by compactness of with image in a compact set , see ( * ? ? ?
* proposition 5.5 ) and lebeau s paper ( * ? ? ?
* proposition 6 ) , one can extract a uniformly convergent subsequence , converging to some @xmath1054 , giving .
now , the @xmath1697 arise directly from proposition [ prop : normal - prop ] , by shrinking @xmath565 ( via shrinking @xmath648 in ) , namely under our assumption on @xmath1299 , for each such @xmath565 there must exist a @xmath1698 in @xmath1699 .
the @xmath1700 then arise from the theorem of duistermaat and hrmander , using that @xmath1701 implies that the backward from @xmath1697 can not meet @xmath14 for some time @xmath1702 , uniform in @xmath9 this is essentially due to @xmath1308 being strictly increasing along microlocally , and @xmath1308 vanishing at @xmath1703 , so as long as @xmath1308 is negative , the can not hit the boundary .
see ( * ? ? ?
* proof of theorem 8.1 ) for more details .
finally , suppose @xmath1704 , which is the more technical case .
this part of the argument is present in essentially the same form in the paper of melrose and sjstrand @xcite ; lebeau s paper ( * ? ? ?
* proposition vii.1 ) gives a very nice presentation , see ( * ? ? ?
* proof of theorem 8.1 ) for an overview with more details .
the rough idea for constructing the @xmath1054 for is to define approximations to it using proposition [ prop : tgt - prop ] .
first , recall that in proposition [ prop : tgt - prop ] , applied at @xmath1299 , @xmath1586 is the coordinate projection ( push forward ) of @xmath59 , evaluated at @xmath1049 , to @xmath1614 .
thus , one should think of the point @xmath1705 in @xmath1614 as an @xmath1706 approximation of where a backward should be after ` time ' ( i.e. parameter value ) @xmath648 .
this is used as follows : given @xmath624 , proposition [ prop : tgt - prop ] gives the existence of a point @xmath1707 in @xmath1708 which is , roughly speaking , @xmath1706 from @xmath1709 , with @xmath1710 being @xmath1706 as well .
then , from @xmath1707 , one can repeat this procedure ( replacing @xmath1299 by @xmath1707 in proposition [ prop : tgt - prop ] ) there are some technical issues corresponding to @xmath1707 being in the boundary or not , and also whether in the former case the backward hits the boundary in time @xmath648 .
taking @xmath1711 , this gives @xmath1712 points @xmath1713 corresponding to the dyadic points on the parameter interval @xmath1714 $ ] .
it is helpful to consider this as analogous to a discrete approximation of solving an ode without the presence of the boundary by taking steps of size @xmath1715 . defining @xmath1716 for only these dyadic values , one
can then get a subsequence @xmath1717 which converges , as @xmath1240 , at @xmath1718 for all @xmath1719 and @xmath1720 integers .
( note that @xmath1721 is defined for these values of @xmath1106 for @xmath118 sufficiently large ! )
one then checks as in lebeau s proof that the result is the restriction of a to dyadic parameter values .
again , we refer to lebeau s paper ( * ? ? ?
* proposition vii.1 ) and ( * ? ? ?
* proof of theorem 8.1 ) for more detail .
[ thm : prop - sing - im ] suppose that @xmath466 , @xmath196 , resp .
@xmath197 , and @xmath180 or @xmath181 .
suppose @xmath182 for some @xmath1100 .
then @xmath185 is a union of maximally _ forward extended , resp .
backward extended _
generalized broken bicharacteristics of the conformal metric @xmath2 in @xmath186 the proof proceeds again as for theorem [ thm : prop - sing ] , but now propositions [ prop : normal - prop ] and [ prop : tgt - prop ] only allow propagation in one direction . thus , if @xmath197 , they allow one to conclude that if a point in @xmath1722 is in @xmath1708 , then there is another point in @xmath1708 which is roughly along a _ backward _ segment emanating from it .
then an actual backward can be constructed as in the works of melrose and sjstrand @xcite , and lebeau @xcite .
this proposition reiterates the importance of the constraint on @xmath6 in that @xmath1730 for @xmath1731 ; for @xmath1732 , the growth or decay relative to @xmath149 does not distinguish between the two approximate solutions @xmath1733 , @xmath277 . for the first part of the lemma , by lemma [ lemma : wfb - infty - independence ] and the subsequent remark , under our assumptions we have @xmath1734 . by , @xmath1735 this is , up to a change in overall the sign of the second summand , @xmath1736 the same as the analogous expression in the de sitter setting , see the first line of the proof of lemma 4.13 of @xcite .
thus , the proof of that lemma goes through without changes the reader needs to keep in mind that @xmath1737 excludes one of the indicial roots from appearing in the argument of that lemma .
( in the de sitter setting , in lemma 4.13 of @xcite , there was no a priori weight ( relative to which one has conormality ) specified . )
the proof of lemma 4.13 of @xcite shows that there exists @xmath1739 , supported in @xmath50 , such that @xmath1740 , @xmath1741 as in the statement of the theorem , and @xmath1742 .
now let @xmath1743 be the solution of @xmath1744 supported in @xmath212 , whose existence follows from theorem [ thm : well - posed ] , and which is of the form @xmath1745 by theorem [ thm : prop - sing ] and proposition [ prop : asymp ] .
then @xmath1746 solves the pde as stated .
uniqueness follows from the basic well - posedness theorem , theorem [ thm : well - posed ] .
[ thm : well - posed - precise ] assume ( tf ) and ( pt ) .
suppose @xmath189 for some @xmath180 , and let @xmath188 . then has a unique solution in @xmath190 , which in fact lies in @xmath191 , and for all compact @xmath140 there exists a compact @xmath141 and a constant @xmath142 such that @xmath192 if @xmath123 , then by theorem [ thm : well - posed ] , has a unique solution in @xmath1747 , and by propagation of singularities it lies in @xmath191 , with the desired estimate .
moreover , again by the propagation of singularities , any solution of in @xmath190 lies in @xmath191 , so the solution is indeed unique even in @xmath190 . if @xmath125 , uniqueness and the stability estimate follow as above . to see existence ,
let @xmath1748 , and let @xmath1749 such that @xmath1750 and @xmath1751 .
this can be achieved by taking @xmath1752 with properly supported schwartz kernel ( of sufficiently small support ) such that @xmath1753\}$ ] is a bounded family in @xmath870 , converging to @xmath1553 in @xmath1754 for @xmath356 , then with @xmath1755 , @xmath1756 , we have the desired properties . by theorem [ thm : well - posed ] , with @xmath38 replaced by @xmath1757 has a unique solution @xmath1758 .
moreover , by the propagation of singularities , one has a uniform estimate @xmath1759 with @xmath437 independent of @xmath1760 . in view of the convergence of the @xmath1757 in @xmath1761
, we deduce the convergence of the @xmath1762 in @xmath1763 to some @xmath1764 , hence ( by uniqueness ) we deduce the existence of @xmath1765 solving @xmath37 with support in @xmath1766 .
however , as @xmath1767 , uniqueness shows the vanishing of @xmath143 on @xmath1768 , proving the theorem .
a. vasy .
propagation of singularities for the wave equation on manifolds with corners . in _ sminaire : quations aux drives partielles , 20042005 _ , smin .
partielles , pages exp .
xiii , 21 .
cole polytech . ,
palaiseau , 2005 . | in this paper we describe the behavior of solutions of the klein - gordon equation , @xmath0 , on lorentzian manifolds @xmath1 which are anti - de sitter - like ( ads - like ) at infinity . such manifolds are lorentzian analogues of the so - called riemannian conformally compact ( or asymptotically hyperbolic ) spaces , in the sense that the metric is conformal to a smooth lorentzian metric @xmath2 on @xmath3 , where @xmath3 has a non - trivial boundary , in the sense that @xmath4 , with @xmath5 a boundary defining function .
the boundary is conformally time - like for these spaces , unlike asymptotically de sitter spaces studied in @xcite , which are similar but with the boundary being conformally space - like .
here we show local well - posedness for the klein - gordon equation , and also global well - posedness under global assumptions on the ( null)bicharacteristic flow , for @xmath6 below the breitenlohner - freedman bound , @xmath7 .
these have been known under additional assumptions , @xcite .
further , we describe the propagation of singularities of solutions and obtain the asymptotic behavior ( at @xmath8 ) of regular solutions .
we also define the scattering operator , which in this case is an analogue of the hyperbolic dirichlet - to - neumann map .
thus , it is shown that below the breitenlohner - freedman bound , the klein - gordon equation behaves much like it would for the conformally related metric , @xmath2 , with dirichlet boundary conditions , for which propagation of singularities was shown by melrose , sjstrand and taylor @xcite , though the precise form of the asymptotics is different .
[ section ] [ lemma]proposition [ lemma]theorem [ lemma]corollary [ lemma]result [ lemma]remark [ lemma]definition |
first we need the concept of a reusable secure carrier @xcite , . a bell state like @xmath0 shared between alice(a ) and
bob(b ) can be used as a reusable secure carrier between two parties as follows .
alice entangles a qubit @xmath1 by the action of a cnot gate @xmath2 ( acting on the qubit @xmath3 and controlled by @xmath4 ) , which produces a state like @xmath5 at the destination bob disentangles the qubit by a cnot operation @xmath6 , leaving the carrier in its original state for reusing . during the transmission the qubit has been disguised in a highly mixed state .
+ any of the bell states [ bells ] |^_ab = ( |00|11)_ab,|^_ab = ( |01|10)_abcan be used as a carrier .
+ for three parties @xcite , a carrier shared between alice(a ) , bob(b ) and charlie(c ) can be a ghz state like @xmath7 or an even parity state like @xmath8 throughout @xcite , the comment @xcite and the present reply the subscripts @xmath9 and @xmath10 are used for the quibts shared by , or the local operators acted by , alice , bob and charlie respectively , while the subscripts @xmath3 and @xmath11 are used for the qubits sent to bob and charlie respectively . +
it was shown in @xcite that by suitable local operations , alice can send a qubit @xmath12 to bob and charlie , by entangling it to the above carriers ( hence hiding it from eavesdroppers ) . in order to share the secret between bob and charlie ,
half of the bits ( the bits in the odd rounds ) were sent to bob and charlie , as states of the form @xmath13 which they could read without the help of each other and the other half ( the bits in the even rounds ) were sent to them in the form @xmath14 which they could use to decipher the value of @xmath12 only by their cooperation .
note that @xmath15 . in order to be able to send both types of states in disguised form
, alice needs to use two types of carriers , namely the @xmath16 carrier for the states @xmath17 and the @xmath18 carrier for the states @xmath19 .
the interesting point is that the two types of carriers are transformed to each other at the end of every round by the local action of hadamard gates by the three parties , due to the following easily verified property an important property which requires careful attention is that the carrier alternates between the above two forms regardless of the value of the qubit @xmath12 which has been sent to bob and charlie by alice .
+ in @xcite the authors show that in the second round where a qubit say @xmath21 has been encoded as @xmath22 and entangled to the carrier @xmath18 , bob ( assuming that he has access to the channel between alice and charlie ) can intercept the qubit 2 sent to charlie ( assuming that he has access to the channel used between alice and charlie ) and perform a suitable unitary operation @xmath23 , on the state of the carrier and the two bits @xmath3 and @xmath11 , to split the carrier @xmath18 to two simple carriers of the type [ bells ] .
this process is shown schematically in figure ( [ split ] ) .
let us denote by @xmath24 the qubit sent by alice in the second round .
bob keeps this qubit for himself and denotes it hereafter by @xmath25 , since it is now in possession of bob and plays a role as part of his new carriers . + it is important to note that the pattern of entanglement splitting depends on the value of this qubit @xmath24 as follows ( equation 3 of the comment ) : as it stands in @xcite , this does not harm the cheating strategy of bob , since as mentioned before any of the bell states can be used as a carrier between two parties .
he then uses the above two pairs of entangled states for retrieving the qubits sent by alice on his own and sending counterfeit qubits to charlie in a clever way so that to avoid detection after public announcement of subsequence of the bits . +
what is crucial in this attack is that bob acts by hadamard gates on his qubits @xmath27 and @xmath25 along with alice and charlie who are doing the same thing at the end of each round . in this way
he almost maintains the pattern of the new carriers , which he has created in the second round , between himself and the other two parties .
+ the reason for `` almost '' is that the hadamard operations act as follows ( equation 4 of the comment ) : at first sight one may argue that alice and charlie who are no longer entangled after bob s trick , can detect their new disentangled situation ( i.e. by testing a bell inequality ) and hence detect bob s cheating .
however this test requires statistical analysis which requires many measurements . in each measurement the carrier collapses and will not be usable anymore .
being in conflict with the whole idea of reusable carrier , we do not follow this line of argument .
instead we modify the protocol in a way which prevents bob s from entanglement splitting .
+ to this end we note that the operator @xmath31 is not the only operator which transforms the carriers @xmath16 and @xmath18 into each other .
consider a unitary operator of the form @xmath32 where @xmath33 is an arbitrary parameter @xmath34 .
for @xmath35 this is the usual hadamard operator .
a simple calculation shows that a generalization of ( [ three ] ) is possible in the following form @xmath37 provided that @xmath38 therefore in the modified protocol alice , bob and charlie act alternatively by the operators @xmath39 , @xmath40 , and @xmath41 , and their inverses , on the qubits in their possession .
the angles @xmath42 and @xmath43 can be announced publicly at the beginning of the protocol .
we now show that after entanglement splitting , bob can not retain his pattern of carriers by any operator @xmath44 which he acts on his qubits @xmath25 and @xmath27 .
we need the following + * proposition * : + * a : * the only operator @xmath45 which in conjunction with @xmath46 leaves invariant the state @xmath29 is the operator @xmath47 .
+ * b : * the only operator @xmath45 which in conjunction with @xmath46 transforms the state @xmath48 into @xmath49 is the operator @xmath50 , where @xmath51 means transpose . + * proof : * the proof is simply straightforward calculations .
we highlight the basic steps .
consider part * a*. we want an operator @xmath44 such that @xmath52 where we use @xmath53 as an abbreviations of @xmath54 and so forth .
acting on both sides by @xmath55 we obtain @xmath56 we now rearrange both sides to the convenient form @xmath57 and effect the operators @xmath58 and @xmath59 on the right hand side by using ( [ ha ] ) . after comparing both sides in the basis @xmath60 we arrived at the stated assertion , namely that @xmath47 . +
similar reasoning proves part @xmath27 .
+ we now come to our main conclusion .
bob , being among the original legitimate parties knows the values of the angles , @xmath61 .
however in order to scape detection he has to apply either the operator @xmath62 or @xmath63 at the end of each round .
however his choice depends on the value of the second bit which he does not know . without this knowledge
he can not retain the pattern of fraud carriers which he has constructed between him and the other two parties .
this then introduces errors in half of the bits sent by alice and received by him and charlie , which in subsequent public announcement of substrings of bits reveals his cheating .
incidentally we note that the equality @xmath64 holds only for @xmath65 , that is for the ordinary hadamard gate .
s. bagherinezhad , v. karimipour , phys .
rev . * a * , 67 , 044302 , ( 2003 ) .
jian - zhong du et al . , entanglement split : comment on `` quantum secret sharing based on reusable greenberger - horne - zeilinger states as secure carriers [ phys . rev .
a * 67 * , 044302 ( 2003 ) ] '' , quant - ph/0605088 .
y. zhang , c. li and g. gao , phys .
rev . * a * , 64 , 024302 , ( 2001 ) . | in a recent comment , it has been shown that in a quantum secret sharing protocol proposed in [ s. bagherinezhad , v. karimipour , phys . rev . *
a * , 67 , 044302 , ( 2003 ) ] , one of the receivers can cheat by splitting the entanglement of the carrier and intercepting the secret , without being detected . in this reply
we show that a simple modification of the protocol prevents the receivers from this kind of cheating .
* reply to `` comment on quantum secret sharing based on reusable greenbergr - horne - zeilinger states as secure carriers '' * + v. karimipour + department of physics , sharif university of technology , + p.o .
box 11365 - 9161 , + tehran , iran pacs numbers : 03.67.dd , 03.65.ud . to set up the context and the notations
, it is appropriate to first review briefly the protocol itself @xcite and the basic feature of the attack or cheating suggested in @xcite . |
it is nowadays well established that entanglement represents a fundamental resource for quantum information tasks @xcite .
however , being a purely quantum feature it is fragile with respect to enviromental contamination .
notwithstanding that , the possibility to achieve entangled states as stationary ones in open quantum systems has been put forward in many different contexts ( for what concern qubits systems see e.g. refs.@xcite ) .
the subject has attracted a lot of attention up until a recent striking experiment on long living entanglement @xcite .
the works on this topic can be considered as falling into two main categories : one where all qubits are plunged in the same environment @xcite and the other where each qubit is plunged in its own environment @xcite . in the former case
the environment can provide an indirect interaction between otherwise decoupled qubits and thus a means to entangle them . in the latter case a direct interaction between qubits
is needed to create entanglement , and usually to maintain it one has to also exploit other mechanisms ( state resetting , driving , etc . ) .
here we consider a hybrid situation as depicted in fig.1 .
it represents a sort of spin-@xmath0 chain dimerized by environments . in practice each environment induces a chain link between contiguous qubits .
hence , we can expect that a simple dissipative dynamics in such a configuration is able to establish entanglement along the chain without the need to exploit any other mechanism .
actually , we will show for the case of three qubits the possibility of achieving stationary entanglement for each qubits pair .
the amount of entanglement results strongly dependent on the initial ( separable ) state .
also the dependance from the chain boundary conditions ( open or closed ) will be analyzed as well as a left - right asymmetry in qubit - environment interaction .
the layout of the paper is the following : in section ii we introduce the model relying on physical motivations and we discuss the general dynamical properties ; in section iii we restrict our attention to the three qubits case and investigate the entanglement dynamics in the open boundary condition ; in section iv we analyze the same system but with closed boundary conditions .
concluding remarks are presented in section v.
the model of fig.1 can be motivated by physically considering two - level atoms inside cavities connected by fibers @xcite .
in such a scheme each atom - qubit can be thought as exchanging energy with the optical modes supported by the fiber . in turn
this latter can modeled as an environment .
thus each qubit dissipates energy through two environments ( one on the left and the other on the right ) .
it happens that two contiguous qubits dissipates energy into the same environment
. then this environment mediates the interaction between the contiguous qubits .
more specifically , let us consider at the @xmath1th site of a chain a qubit described by ladder operators @xmath2 satisfying the usual spin-@xmath0 algebra @xmath3=\hat{\sigma}_i^z$ ] .
let us also consider at the @xmath1th site of a chain radiation modes described by ladder operators @xmath4 satisfying the usual bosonic algebra @xmath5=\delta_{j , j^\prime}$ ] .
then , the interaction hamiltonian reads @xmath6 by considering the @xmath7 as environment s operators for the @xmath1th qubit , we can use standard techniques @xcite to arrive at the following master equation @xmath8,\end{aligned}\ ] ] where @xmath9 denotes the anti - commutator and we have assumed unit decay rate .
since we are interested on the steady state we have to notice that , given a master equation written in the standard linbladian form , @xmath10,\ ] ] the uniqueness of the stationary solution is guaranteed if the only operators commuting with every lindblad operator @xmath11 are multiples of identity @xcite . in the case of eq . the @xmath12s commute with lindblad operators .
hence the steady state may not be unique , that is it may depend on the initial conditions . due to that we need to study the full dynamics of the system .
we restrict our attention to a chain of three sites .
we first consider open boundary conditions .
then , the dynamics will be described by a master equation that can be easily derived from eq.([dyn ] ) @xmath13\nonumber\\ & + & ( 1-\gamma)\left[2(\hat{\sigma}_2 ^ -+\hat{\sigma}_{3}^-)\hat{\rho } ( \hat{\sigma}_{2}^{+}+\hat{\sigma}_{3}^{+})\right.\nonumber\\ & & \left.-\{(\hat{\sigma}_{2}^{+}+\hat{\sigma}_{3}^{+})(\hat{\sigma}_{2}^-+\hat{\sigma}_{3}^-),\hat{\rho}\}\right].\end{aligned}\ ] ] here we have considered the possibility for each qubit of having an asymmetric decay rate on the left and right environments .
this has been accounted for by the real factors @xmath14 and @xmath15 with the assumption @xmath16 .
clearly the symmetric situation is recovered when @xmath17 . by arranging the density matrix ( expressed in the computational basis @xmath18 ) as a vector @xmath19 ( e.g. writing @xmath20 )
, the master equation can be rewritten as a linear set of differential equations @xmath21 where @xmath22 is a @xmath23 matrix of constant coefficients given by @xmath24 where @xmath25 with @xmath26 the @xmath27 dimensional identity matrix and @xmath28 , @xmath29 .
then , the set of differential equations can be converted into a set of algebraic equations via the laplace transform , @xmath30 , i.e. @xmath31 decoupling these equations one finds that the laplace transforms of the density matrix elements are rational functions of polynomials and the inverse laplace transformation can be performed analytically .
the results are not explicitly reported because the expressions are too much cumbersome .
having the density matrix of the system , we can study the entanglement dynamics for each qubit pair of the system . to quantify the amount of entanglement between each of the qubits we use the concurrence @xcite .
we recall that to find the concurrence of a bipartite system described by the density matrix @xmath32 , the following steps should be done : * find the complex conjugate of the density matrix in the computational basis and denote it by @xmath33 .
* define @xmath34 , where @xmath35 . *
find the square root of the eigenvalues of @xmath36 and sort them in decreasing order : @xmath37 . *
the concurrence is given by @xmath38 figure [ 3e ] shows the evolution of entanglement between each qubit pair for @xmath39 initial state . as it can be seen in this figure , when all the qubits are initially in excited state it takes longer time for the first and the third qubits to become entangled compared to the time needed to generate entanglement between the first and second or second and third qubits . as a consequence for not nearest neighborhood qubits we have a _ sudden birth of entanglement _ , i.e. it suddenly becomes non zero at times greater than zero ( it does not smoothly increases starting from initial time ) @xcite .
initial state . dashed line :
@xmath40 , solid line : @xmath41 and dotted line : @xmath42 . ]
if we start with different initial states , the entanglement behaves differently in time .
figure [ 2e]-left shows the value of concurrence in time for @xmath43 as initial state . like the previous case , it shows that the time needed by the first and third qubits to become entangled is longer than the time needed by nearest neighborhood qubits .
moreover , we still have the entanglement sudden birth phenomenon , but this time it manifests not only for distant qubits ( first and third ) but also for the first and the second qubits . the interesting point in this figure
is that the entanglement generation between the nearest neighbors is quicker if they are initially prepared in @xmath44 rather than @xmath45 state .
the other possibility with two number of excitations in the initial state is @xmath46 , for which the time evolution of entanglement in shown in figure [ 2e]-right .
this time the entanglement sudden birth phenomenon only manifests for distant qubits ( first and third ) .
( top ) and @xmath47 ( bottom ) initial states . dashed line : @xmath40 , solid line : @xmath41 and dotted line : @xmath42.,title="fig : " ] ( top ) and @xmath47 ( bottom ) initial states . dashed line : @xmath40 , solid line : @xmath41 and dotted line : @xmath42.,title="fig : " ] ( top ) and @xmath48 ( bottom ) initial states . dashed line : @xmath40 , solid line : @xmath41 and dotted line @xmath42.,title="fig : " ] ( top ) and @xmath48 ( bottom ) initial states . dashed line : @xmath40 , solid line : @xmath41 and dotted line @xmath42.,title="fig : " ] a big difference appears if the number of excitation of the initial state , reduces to one .
figures [ 1e]-left and - right show entanglement evolution for the initial states @xmath49 and @xmath50 , respectively . as it can be seen in this figure
, entanglement is generated between each qubits pair from the beginning , no matter how far they are from each other , i.e. we no more have entanglement sudden birth phenomenon .
finally , in the case of @xmath51 initial state there is no entanglement at any time because this state represent a fixed point of the liouvillian superoperator at right hand side of the master equation , or in other words @xmath52 with @xmath53 .
taking the limit @xmath54 in the density matrix elements , we arrive at the following general form for the steady state @xmath55 where @xmath56 is determined by the initial state of the system .
in particular we have the following correspondence : @xmath57 to find the amount of entanglement in each qubits pair at the steady state , we first write the reduced density matrices @xmath58 @xmath59 for different initial states . ]
then , it is easy to show that the concurrence becomes @xmath60 figure ( [ mathematicasteady ] ) shows the stationary entanglement vs @xmath14 for different initial state . for initial states like @xmath61 the entanglement does not depend on @xmath14 . for @xmath62 initial state , by increasing the difference in the dissipation rates more entanglement will be induced in the system . on the contrary for @xmath47 the maximum value of entanglement
is achieved when the dissipation rates into the two environments are the same . as it is expected the entanglement for initial states
@xmath63 are equal at @xmath64 . in these cases ,
the maximum value can be attained when there is one excitation in the initial state .
we now study the chain of three sites with closed boundary condition . in this case
the master equation reads @xmath65\nonumber\\ & + & \mu\left[2(\hat{\sigma}_2 ^ -+\hat{\sigma}_{3}^-)\hat{\rho } ( \hat{\sigma}_{2}^{+}+\hat{\sigma}_{3}^{+})\right.\nonumber\\ & & \left.-\{(\hat{\sigma}_{2}^{+}+\hat{\sigma}_{3}^{+})(\hat{\sigma}_{2}^-+\hat{\sigma}_{3}^-),\hat{\rho}\}\right]\nonumber\\ & + & \nu\left[2(\hat{\sigma}_3 ^ -+\hat{\sigma}_{1}^-)\hat{\rho } ( \hat{\sigma}_{3}^{+}+\hat{\sigma}_{1}^{+})\right.\nonumber\\ & & \left.-\{(\hat{\sigma}_{3}^{+}+\hat{\sigma}_{1}^{+})(\hat{\sigma}_{3}^-+\hat{\sigma}_{1}^-),\hat{\rho}\}\right ] , \label{mecc}\end{aligned}\ ] ] where the possibility for each qubit of having an asymmetric decay rate on the left and right environments is accounted for by the real factors @xmath14 , @xmath66 and @xmath67 ( @xmath68 ) .
clearly the symmetric situation is recovered when @xmath69 ( @xmath70 ) . again arranging the density matrix ( expressed in the computational basis @xmath18 ) as a vector @xmath19 ( e.g. writing @xmath20 ) ,
the master equation can be rewritten as a linear set of differential equations @xmath71 where @xmath22 is a @xmath23 matrix of constant coefficients given by @xmath72 where @xmath73 in this case it is possible to see that @xmath74 admits a unique solution @xmath75 , i.e. the only possible solution is @xmath76 , for all values of @xmath77 .
hence no entanglement survive at stationary conditions .
in this work we have considered a spin-@xmath0 chain dimerized by environments . by means of dissipative mechanism , each environment induces a chain link between contiguous qubits .
then we have studied the possibility of having long living entanglement without resorting to any other mechanism .
in particular for the case of three qubits chain with open boundary condition we have classified the amount of stationary entanglement accordingly to some initial ( separable ) states .
here we have also shown the appearance of the entanglement sudden birth .
on the contrary , for the case of three qubits chain with closed boundary condition we have proved the impossibility of stationary entanglement .
this fact can be interpreted as entanglement frustration phenomenon @xcite , induced in this context by the imposed periodic boundary conditions .
the proposed scheme lends itself to to be extended to @xmath78 sites where one can evaluate how entanglement scales as function of the distance between the two qubits .
moreover , in the limit of large @xmath79 , it could also results useful for studying possible connections between quantum phase transitions and reservoirs properties .
finally , the discussed method of generating entanglement seems economical and offers interesting perspectives for the generation of the so called graph states ( useful for measurement based quantum compuatation ) @xcite , when one considers a network topologies more complicate than a simple chain .
m. b. plenio and s. f. huelga , phys .
lett . * 88 * , 197901 ( 2002 ) ; s. clark , a. peng , m. gu and s. parkins , phys .
lett . * 91 * , 177901 ( 2003 ) ; s. mancini and j. wang , eur .
j. d * 32 * , 257 ( 2005 ) ; l. hartmann , w. dur , h.j .
briegel , phys .
a * 74 * , 052304 ( 2006 ) ; d. angelakis , s. bose and s. mancini , europhys
. lett . * 85 * , 20007 ( 2007 ) . | we investigate the possibility of chaining qubits by letting pairs of nearest neighbours qubits dissipating into common environments .
we then study entanglement dynamics within the chain and show that steady state entanglement can be achieved . |
the so - called `` ridge '' , found in hadronic collisions at rhic and lhc energies , @xcite , has been subject to years of intense theoretical and experimental investigation and can be regarded as a crucial observable in the study of dense qcd .
the ridge is a 2-particle correlation focused in the azimuthal difference @xmath2 and elongated in pseudo - rapidity @xmath3 .
this correlation seems to be specific to high multiplicity hadronic processes ( more central high @xmath4 heavy ion , pa , da @xcite , and very high multiplicity pp collisions @xcite ) , suggesting it to be related to the appearance of a `` medium '' .
it is further possible to relate ridge physics to bulk anisotropies particle production @xcite : the two - particle correlation function , including the ridge , can be successfully decomposed in fourier components ( commonly known as @xmath5 ) w.r.t .
the reaction plane @xmath6 @xmath7 it is trivial to see that a single ridge is mostly generated by @xmath8 components .
a `` double ridge '' , observed in aa and pa / da collisions , is dominated by `` elliptic flow '' @xmath9 and higher even components .
since elliptic flow was key to the widely - publicized announcement that matter in heavy ion collisions behaves as a low viscosity liquid @xcite , a proper understanding of ridge phenomenology is crucial to our understanding of the properties of the matter created in heavy ion collisions .
@xmath5s are thought to reflect the structure of fourier components of the initial transverse energy density @xmath10 .
thus , @xmath10 @xmath11 gets somehow converted into momentum anisotropies @xmath5 .
further use of the ridge as a medium probe , however , is somewhat impeded by uncertainty of the mechanism which converts initial state shape anisotropy into a momentum correlation : it is far from clear if the ridges in @xmath12 have the same origin , or a single origin , and how exactly the ridge in each system is generated .
many models have appeared explaining the ridge .
while experimental data @xcite seems to suggest that the ridge is generated by `` soft '' phenomena rather than `` jetty physics '' , two broad classes of models remain : they are distinguished by the ridge s relationship to collective flow . in the first class of models it is hydrodynamic flow itself that creates the correlation in azimuthal angle @xcite .
the local structures ( be they hotspots , qcd strings or glasma flux tubes ) generate a `` long '' correlation in rapidity which is then azimuthally collimated by transverse flow `` pushing out '' the hotspot .
thus , the ridge is made by an interplay of transverse flow ( the zeroth fourier component of the flow expansion ) and local hotspots , which can be represented by higher harmonics in eq .
the deviation from isotropy , ie the width of the @xmath2 correlation , should therefore depend directly on the global collective flow of the event and its anisotropies ( be they average features , which are there for even coefficients , or event - by - event fluctuations , which can generate odd coefficients ) . in the second class of models ,
the ridge is generated by local partonic dynamics .
as shown in @xcite , a system extended in both transverse space and rapidity , approximately boost - invariant , and of high enough partonic density that more than one parton - parton interaction per unit of transverse space per unit of rapidity will naturally yield ridges : in a spatially extended system , partonic interactions at _ finite impact parameter _ will naturally generate an angular correlation , centered around the impact parameter vector as well as color - coherence regions @xcite .
because of boost - invariance , this angular correlation with be wide in rapidity , since a parton - parton interaction at one rapidity is more likely at a hotspot , and , given that hotspots are elongated in rapidity , a _
second _ parton - parton interaction is more likely .
the model used in @xcite has been calculated in a generic pqcd model , but qualitatively similar dynamics emerges in cgc - based calculations such as @xcite as well as string dynamics @xcite .
the scaling behind @xmath13 is then the local gluon density , given , in @xcite , by powers of the saturation scale @xmath14 and the qcd coupling constant @xmath15 .
the recent appearance of a `` strong '' @xmath16 and @xmath17 double ridge @xcite , comparable to that of heavy ion collisions , has provided us with further puzzles : hydrodynamics is generally thought to be a good description of the system when the knudsen number @xmath18 , i.e. the mean free path is much less than the system size .
given the empirical multiplicity scaling with @xmath4 @xcite , the ideal gas equation of state and the bjorken model @xcite with initial time @xmath19 , entropy density @xmath20 and overlap area @xmath21 @xmath22 in terms of the unknown @xmath23 , and assuming the initial radius of a @xmath24 collision is set by the strong coupling scale @xmath25 @xmath26 assuming @xmath27 and @xmath28 . for larger systems
the scaling will be @xmath29 where the extra energy power in @xmath30 is due to the faster dependence of multiplicity per participant in @xmath30 collisions @xcite .
as can be seen , by comparing ridges at the same @xmath31 for different systems , one expects about the same knudsen number for @xmath24 vs @xmath0 collisions , but @xmath30 will have a lower knudsen number than an equal multiplicity @xmath0 bin : while the former will be denser but smaller than the other , the greater transverse size of the @xmath30 system beats the smaller density in determining @xmath32 , and the extra energy dependence will suppress the knudsen number in @xmath30 further . therefore ,
unless one assumes a breakdown in @xmath33 scaling @xcite , binning by multiplicity is _ not _ the same as binning in knudsen number .
harmonics should follow a scaling of the type @xcite @xmath34 where @xmath19 is the hydrodynamic initialization time , @xmath35 is the freeze - out entropy density , @xmath36 the initial entropy density , @xmath37 the speed of sound and @xmath38 the lifetime .
@xmath38 needs to be determined via a hydrodynamic code but is a non - trivial function of both @xmath39 and the size @xmath40 even in the ideal limit .
it is driven by rarefaction wave dynamics for small ( @xmath41 ) systems ( such as pa , pp ) and by expansion dynamics for large ones ( @xmath30 ) .
@xmath42 hence , @xmath38 should further break the knudsen number scaling .
while , as discussed in @xcite , the initial eccentricity in small systems is too dominated by initial - state uncertainties ( the `` transverse shape of the nucleon '' ) to allow for a quantitative comparison , the scaling patterns of the ridge structure as one goes from pp to pa to aa are puzzling .
if hydrodynamics is to work , eq .
[ v2kn ] would imply @xmath43fm , putting the likely value of @xmath44 below the `` universal '' @xmath45 limit . in this case , however , it is remarkable that pp vs aa systems are so different : from pa to pp , the knudsen number should be different by no more than 30 - 50@xmath46 , since the transverse size is comparable and the temperatures scale as @xmath47 .
these considerations arise just from the knudsen number , and hold independently of the initial geometry .
the latter , for large nuclei , is thought to be dominated by the number of participants . for smaller systems such as pa and @xmath24 , however , the largely unexplored sub - nucleon fluctuations are crucial in determining the initial @xmath48
. one can not quantitatively compare @xmath49 for @xmath30 and @xmath50 using eq .
[ v2kn ] since the average geometry in these systems is very different . over all events , however , @xmath51 by symmetry for _ all _ systems , so event - by - event @xmath52 is determined only by fluctuations , @xmath53 at present nuclear scale @xmath54 is thought to be well - approximated by a glauber calculation , while partonic scale @xmath55 is much more model - dependent @xcite .
we know , however , that in all models @xmath56 for @xmath30 and @xmath57 for pa , pp .
initial conditions , therefore , give the same hierarchy as the knudsen number , a `` large '' @xmath30 vs a `` small '' @xmath50 .
experimental data has a diametrically opposite behavior : the correlation strength of pa w.r.t .
aa collisions are _ nearly - identical _ , especially in @xmath58 where data points at relevant centralities overlap @xcite . for @xmath9 , where aa and pa collisions are geometrically different , a @xmath59 scaling seems to hold @xcite .
however , when going from aa to @xmath0 ( lhc ) and @xmath1 ( rhic ) , the knudsen number certainly rises by @xmath60 or so if transverse size is counted as @xmath40 . at present , only one experiment @xcite reported seeing a ridge in @xmath24 collisions , and only in an experiment - specific sample of very high multiplicity events .
this ridge is of markedly smaller amplitude w.r.t aa , and might not exhibit the accompanying double ridge ( which for pa was seen in all centrality bins and can be used to estimate @xmath9 and @xmath58 ) .
while no harmonic decomposition was attempted to date , @xmath9 and @xmath58 in pp collisions could be compatible with zero , albeit with a large error bar .
this large systematic uncertainty is primarily due to the large background , in @xmath24 collisions , of off - center of mass energy ( different bjorken @xmath61 ) parton - parton collisions producing a strong away - side `` fake ridge '' @xcite . unlike in pa collisions ,
the admixture of such events is highly correlated with multiplicity , making their subtraction difficult .
hence , a centrality - based quantitative comparison is to date problematic .
even a qualitative comparison between pp and pa ridges @xcite however , shows it is difficult to see how the knudsen and density scaling described in the previous paragraph can be made compatible with experimental data .
i should note that , as pointed out earlier , this is a more general issue than the pa ridge .
it seems ( see the references in @xcite that when different energies , system sizes and rapidities are put together , @xmath9 is described by the following empirical formula @xmath62 where @xmath63 and @xmath64 are universal functions ( with very weak to no dependence on @xmath65 and so on ) and the residual dependence on @xmath66 is through @xmath67 rather than @xmath68 . while a detailed study of this scaling within hydrodynamics is yet to be performed , the compatibility of such a system with a fluid whose speed of sound @xmath37 and @xmath44 depend non - trivially on temperature is not so obvious .
expanding the cooper - frye formula @xcite ( discussed later in eq .
[ cf ] ) in the 2nd fourier harmonic of the flow and freeze - out hypersurface @xmath69\ ] ] which , when integrated , yields eq .
[ v2kn ] ( @xmath38 , defined after eq .
[ v2kn ] , also determines @xmath70 ) .
hence , size , density and @xmath71 dependence are _ not _ expected to factorize they do in eq . [ v2scaling ] , but maintain a non - trivial dependence in which density and size ( and hence geometry , @xmath33 , system size and rapidity ) mix in a non - trivial way .
@xmath72 if all @xmath73 parameters are non - negligible and uncorrelated , no projection of the matrix in eq .
[ matrix ] to a lower dimension is possible . the only way to reduce the dimensions of eq .
[ matrix ] is to assume a negligible knudsen number .
but , as discussed before , this predicts similar correlations between p - p and pa and does not eliminate scaling violations associated with @xmath38 .
hence , the scaling difficulty of the pa ridge is an `` in your face '' ( due to pa collisions s small size ) illustration of the larger difficoulty of hydrodynamics to describe the _ scaling _ of harmonic flow with energy , system size and rapidity @xcite .
in contrast , initial state effects depend only on the geometry _ and only one intensive parameter _
( such as the `` saturation scale '' @xmath14 , or more generally the transverse gluon density @xcite ) .
thus , such models might be more amenable to simple scaling @xcite .
the long - established difference between in - vacuum and in - medium partonic wavefunctions @xcite also makes the similarity of pa and aa w.r.t .
pp more natural in such an initial state model .
the systematic uncertainties over the initial geometry in small systems , however , make disentangling such initial state effects from collectivity remnants just from particle harmonics in a model - independent way non - trivial . from a scaling point of view , therefore , hydrodynamics looks disfavored but the entanglement of geometric and `` intensive '' quantities might be too high for a `` clean '' to be made by scaling arguments _
alone_. this puts the onus on exploiting the main difference between the two models of ridge production outlined in the previous section - in one case azimuthal collimation is done by `` global '' collective flow , in the other by `` local '' parton - parton interactions , to devise a qualitative experimental observable which can clarify the situation . in the next section
i will argue that pid two - particle correlations is a good candidate for such an analysis .
in @xmath30 collisions , collective flow has historically been ascertained by looking at the difference between @xmath74 spectra .
when an expanding hydrodynamic fluid emits particles , energy - momentum and entropy conservation constrain the particle distribution to be of the cooper - frye form @xmath75 where the emission function is a `` boosted '' thermal exponential @xcite . @xmath76 + { \mathcal{o } \left ( kn \times \left [ \frac{p_t}{t } \right]^{3/2 - 2 } \right ) } \ ] ] the form of the viscous correction is currently controversial @xcite , but it is subleading in knudsen number and important at @xmath77 .
the @xmath78 limit is the usual boosted exponential @xmath79 \right\rangle}_r\ ] ] where @xmath80 is the temperature and @xmath70 is the transverse flow at freeze - out ( the radial component of the flow 4-velocity @xmath81 ) and @xmath82 is the 3d surface specifying the spacetimes of particle emission .
@xmath83 have to be calculated within a hydrodynamic code . in this work ,
i quote the results in @xcite ( a similar analysis was done in @xcite ) calculated with the code in @xcite , which was also used successfully to describe small systems @xcite in the past .
a qualitative signature of eq . [ cf ] is a mass - scaling of spectra and flow correlations . for a particle of mass @xmath84 ,
the exponent at mid - rapidity ( no longitudinal momentum ) becomes @xmath85 consequently , each particle still maintains an exponential shape , but its `` effective temperature '' @xmath86 is related to the true freeze - out temperature @xmath80 by @xmath87 the @xmath84-dependence is steepening in both the massive and the massless particle limit , with the spectrum becoming flatter as @xmath84 increases ( naturally , since more momentum is transmitted to the particle by flow and less by thermal motion ) .
thus , mass ordering arises naturally if angular correlations are generated by flow , as assumed in the first class of scenarios discussed in the introduction .
if , however , the second class is physically correct , and the angular correlations are generated by initial state high density effects , mass ordering can also arise from final state fragmentation effects . as noted in @xcite , a non - negligible mass - ordering
is generated within pythia with no need for transverse flow , by the fact that more massive particles are typically produced by more `` transverse '' strings . for small ( p - p ) systems ,
mass ordering using this effect is _ greater _ than the mass ordering from a hypothetical hydrodynamic phase .
while @xcite focused on @xmath88 strings , while the processes considered in @xcite are purely gluonic , such effects should also be present in fragmentation functions : given a distribution of gluons @xmath89 fragmenting as in elementary collisions , the final hadron distribution will be given by @xmath90 different @xmath91 will generate different hadron spectra .
it is therefore not surprising that models such as hijing reproduce qualitatively , although not quantitatively @xcite the @xmath67 scaling in pa collisions . since our understanding of gluon fragmentation is far from precise , therefore , the mass hierarchy , _ by itself _ , can not be taken as evidence of hydrodynamic flow , although an extension of @xmath67 analysis for particles of similar masses but different partonic wavefunctions , such as the proton and the @xmath92 meson , might help in clarifying whether the scaling variable in @xmath67 is mass or partonic structure .
2-particle correlations are however qualitatively different from 1-particle correlations in this respect . if the ridge is generated by transverse flow focusing hotspots , its strength should be correlated , and its width anti - correlated , with the mass of the particle , because of the greater sensitivity to flow of higher - mass particles .
this leads to higher , slimmer correlations for more massive particles .
mathematically , @xcite accounts for this effect by modeling 2-particle angular correlations through the local flow field .
up to an emission volume @xmath93 , and ignoring quantum ( hbt ) effects @xmath94 with @xmath95 specified in eq .
[ emission ] and the mixed event background given by the square of eq .
it is clear from the form of eq .
[ emission ] that correlations will get an additional boost for higher - mass particles , since the latter are more sensitive to flow .
this is why hydrodynamics predicts consistently higher @xmath9 form more massive particles @xcite . as an illustration ,
@xcite calculated the 2-particle distribution function , binned by particle species , with the model @xcite .
the result is shown in fig .
[ figcorr ] for @xmath96 where @xmath97 is the overall number of events , and indeed follows expectations .
experimental data seems to behave in a similar way @xcite .
in contrast , in initial - state based models @xcite the effective theory of high - density qcd will give a 2-parton correlation . in the absence of subsequent evolution , however , fragmentation will happen `` as in vacuum '' , or independently for each gluon .
gluon - gluon correlations will be suppressed by factors of @xmath98 , assumed as small in such approaches .
the hadron - hadron ( @xmath99 ) correlation function will therefore be of this form @xmath100 where @xmath101 is a function incorporating the predictions from dense parton dynamics for the ridge - like structure .
the mixed event background is given by the square of the distribution in eq .
[ fragment ] .
even qualitatively , eq . [ fragment2part ] is a very different equation from eq .
[ cf2part ] , since fragmentation is parton - specific and independent of and partonic correlations .
thus , the fragmentation function @xmath102 is independent of rapidity and angle .
hence , for the azimuthal correlation function all dependence on @xmath103 factorizes into a function of @xmath104 only . provided normalization is done by a species and @xmath105bin specific factor so one point on the curve matches for all species , any residual fragmentation effects should factor out
. this should _ not _ be the case if particle production happens via cooper - frye freezeout , eqn [ cf2part ] . to test the phenomenological consequences of this ,
i describe @xmath106 by an empirical function in terms of @xmath2 and @xmath107 ( the narrow peak in @xmath108 due to jet fragmentation is disregarded ) . @xmath109
e @xmath110 are adjustable parameters and proceed to convolute it with the cteq @xcite gluon fragmentation functions according to eq .
[ fragment2part ] . the result is also shown in fig .
[ figmar ] , for both a normalization of the type @xmath111 and assuming zero yield at minumum ( zyam ) separately for all particle species , @xmath112 ( @xmath113 at minumum by definition ) . as can be seen ,
once an overall normalization factor is common for all particles , the correlation function shape is independent of the particle species .
this is clearly not the case for the plot in fig .
[ figcorr ] . changing the form of the empirical function eq .
[ empirical ] without the hadronization implied in eq .
[ fragment2part ] will not alter these basic conclusions , although of course the shape of the angular correlation and its normalization will vary . physically , the species - independence of the shape of the 2-particle correlation function is a trivial consequence of the form of eq .
[ fragment2part ] : since fragmentation into a particular species are determined independently of the correlation , the fragmentation function factorizes into `` the average energy of the gluon that fragments into a hadron of that momentum'',which retains no memory of the shape of the angular distribution . by how much can this feature be altered by altering fragmentation dynamics ?
collinear fragmentation assumed in eq .
[ fragment2part ] is usually appropriate for @xmath114 . for lower momenta ,
fragmentation is certainly not collinear .
however , if the typical parton momentum @xmath115 fragmentation should remain parton specific since correlations between oppositely moving partons are higher order in @xmath15 .
one way to incorporate this is to introduce species - specific deviations from collinear fragmentation , i.e. , @xmath116 .
the @xmath117 dependence might introduce the observed species - specific pattern in the 2-particle correlation function if ( as is reasonable ) fragmentation into more massive particles gives a larger `` side - kick '' to the fragmented hadron . assuming a normalization such that @xmath118 and a gluon distribution correlated by @xmath9 only @xmath119 one can see that , when one integrates @xmath120 out of eq .
[ fragment2part ] one gets @xmath121 note that no correlation exists between @xmath122 and @xmath123 , integrating over one of these angles for a large ensemble of both events and fragmentations will average the other to zero .
hence , fourier - decomposing eq .
[ vn1part ] one can rotate @xmath124 or @xmath125 away into an irrelevant phase . using the convolution theorem , one will get @xmath126 where @xmath127 , the fourier - transform of the product of fragmentation function , and _ must be independent of geometry and @xmath128-coefficient _ ( in the collinear limit @xmath129 , in case fragmentation smearing dominates @xmath130 ) .
in contrast , @xmath131 must depend only on geometry and not of particle species
. ratios of @xmath5s for different system sizes and particle species , a la @xcite , therefore , could be used to see if fragmentation of correlated gluons occurs independently for each gluon or is correlated .
if relation eq.[dtilde ] holds , for two systems @xmath132 ( be they @xmath133 or a particular centrality of@xmath30 ) at the same @xmath4 and @xmath71 @xmath134 where the last relation is motivated by the nearly identical @xmath58 in pa / da vs aa collisions @xcite .
no such relations are apparent in the hydrodynamic model .
in fact the last two ratios in eq . [ tests ] should exhibit a systematic upward trend for larger systems , as transverse flow is larger in such systems ( as the system size increases , @xmath135 grows faster than @xmath136 ) .
this is something experiment can readily check .
if such scalings are not apparent , initial - state correlations might still work if the multi - gluon ladders commonly used to justify the initial - state @xmath137 factorization approach @xcite also influence fragmentation , leading to some fragmentation function sensitive to multi - gluon correlations @xmath138 @xmath139 such a fragmentation would be make the experimental distinction between an initial - state model and a hydrodynamic model much more involved . the good description , by the ip - glasma flux - tube model with parton - hadron duality , of the negative - binomial multiplicity distributions @xcite places some bounds of the importance of such effects within an initial state - based model .
nevertheless , these ideas have been explored in the literature before @xcite , although they have not been explored quantitatively in the context of 2-particle correlations . if multi - gluon correlations persist until and including fragmentation , however , it is difficult to see the qualitative difference between this system and a dense ensemble of sequentially interacting gluons , which should universally approach hydrodynamic evolution @xcite . in this limit ,
`` high density gluons '' and `` hydrodynamics '' become indistinguishable .
further tests of the observables described here can be experimentally performed via 2-particle correlations of the @xmath92-meson , which has the approximately same mass as the baryon but a very different partonic structure : a similar pattern to the proton would point to mass scaling , typical of hydrodynamics and equation [ cf2part ] .
a very different pattern might indicate non - hydrodynamic fragmentation , possibly correlated fragmentation of the type of eq .
[ corrfrag ] if the tests outlined in eq . [ tests ] are not passed . along the same direction ,
2-particle correlations with identified _ heavy quark mesons _ in pa and da collisions might be instrumental in assessing the relative importance of flow w.r.t .
initial state correlations : while correlations due to high initial gluon density should be equally strong for light and heavy quarks according to formulae similar to eq [ fragment2part ] ( updated with the heavy quark creation diagrams @xcite ) , in a hydrodynamic system the `` effective knudsen number '' for a heavier particle is @xmath140@xcite , so a hydrodynamic correlation will be parametrically weaker
. it will also be illuminating to see whether @xmath141 in pa go to the same @xmath142 gev as in aa @xcite . for pa the most commonly used tomographic variable , @xmath143 from @xmath144 gev onward @xcite .
this would naively suggests the `` medium '' in pa is transparent to particles at 3 - 4 gev . yet
@xmath68 in pa was measured to be non - negligible for these momenta @xcite .
this suggests that the influence of `` the medium '' on 1 and 2 particle correlations is dramatically different .
this observation , as well as giving us an explanation for the inability of most jet energy loss models to describe both @xmath145 and @xmath9 together ( unless a non - trivial opacity evolution with density is assumed @xcite ) , raises the question of how far in @xmath71 does @xmath141 reach in `` small systems '' such as pa and da collisions . to answer these questions , one must determine if @xmath146 gev in the `` soft '' regime determined by hydrodynamics or the `` hard '' regime determined by fragmentation .
knudsen number scaling suggests , as shown in the appendix , that for a `` fluid '' with low absolute opacity , after a @xmath147 , one expects hydrodynamic @xmath141 to be @xmath148 .
tomographic @xmath5 for these @xmath71s is ruled out by @xmath149 .
a sizable @xmath9 for @xmath150 collisions at @xmath151 gev might suggest that either @xmath44 is _ really _ low ( but then again , what about pp collisions ? ) , or initial state is somehow impacting 2-particle correlations even at high momenta . in conclusion ,
i have discussed the possible physical origin of the observed `` ridge - like '' 2-particle correlations observed in pa and aa collisions .
i can not draw any firm conclusions : on the one hand , the scaling of the observed ridge as the system size increases from pp to pa / da and aa is difficult to see in a hydrodynamic model . on the other , the successful fit by hydrodynamic calculations @xcite of the pid ridge , together with the failure , even on a qualitative level , of existing in - vacuum fragmentation ansatze to reproduce the pattern observed in experiment , suggests that the ridge is very much a _ final state _ phenomenon : the mass ordering , provided fragmentation happens through usual vacuum fragmentation , is non - trivial to reproduce when the 2-gluon correlation function is tuned to reproduce charged dihadron correlations .
i eagerly hope that further tests for this mass ordering , involving @xmath92 mesons , heavy quarks and high @xmath71 charged particles , can clarify these issues .
combining the standard formula @xmath152 with the dimensional estimate @xmath153 and the knudsen number defined in eq .
[ knudsen ] i get that , for @xmath154 the number of `` hard '' scatterings is @xmath155 . for a fourier component in the scattering difference in azimuthal angles to vanish , a it is enough that the azimuthal difference in number of scatterings be small . for this one requires @xmath156 for the @xmath128th component .
a radiative dominated freezeout will bring the square root to a 3/2 root @xcite .
`` tomographic '' soft scatterings and radiative corrections could remain , as they do not contribute to the transport properties of the bulk per se .
however , @xmath157 after a @xmath158 gev seems to place strict limits on their significance .
g.t . acknowledges the financial support received from the helmholtz international center for fair within the framework of the loewe program ( landesoffensive zur entwicklung wissenschaftlich - konomischer exzellenz ) launched by the state of hesse .
gt also acknowledges support from doe under grant no .
de - fg02 - 93er40764 .
i would like to thank wojciech broniowski , piotr bozek , miklos gyulassy , mike lisa and jurgen schukraft for discussions .
p. bozek , w. broniowski and g. torrieri , phys .
lett . * 111 * , 172303 ( 2013 ) j. d. bjorken , s. j. brodsky and a. scharff goldhaber , phys . lett .
b * 726 * , 344 ( 2013 ) s. chatrchyan _ et al .
_ [ cms collaboration ] , phys .
b * 724 * , 213 ( 2013 ) k. aamodt _ et al .
_ [ alice collaboration ] , phys .
b * 708 * , 249 ( 2012 ) s. mohapatra [ atlas collaboration ] , j. phys .
ser . * 389 * , 012011 ( 2012 ) b. i. abelev _ et al . _ [ star collaboration ] , phys .
c * 80 * , 064912 ( 2009 ) a. adare _ et al .
_ [ phenix collaboration ] , phys .
c * 78 * , 014901 ( 2008 ) b. alver _ et al .
_ [ phobos collaboration ] , j. phys .
g * 35 * , 104080 ( 2008 ) v. khachatryan _ et al .
_ [ cms collaboration ] , jhep * 1009 * , 091 ( 2010 ) a. adare _ et al . _
[ phenix collaboration ] , phys .
lett . * 111 * , 212301 ( 2013 ) j. takahashi , b. m. tavares , w. l. qian , r. andrade , f. grassi , y. hama , t. kodama and n. xu , phys .
lett . * 103 * , 242301 ( 2009 ) m. gyulassy and l. mclerran , nucl .
a * 750 * , 30 ( 2005 ) ; j. l. nagle , nucl .
a * 830 * , 147c ( 2009 ) s. a. voloshin , phys .
b * 632 * , 490 ( 2006 ) j. d. bjorken , phys .
d * 27 * , 140 ( 1983 ) .
[ alice collaboration ] , phys . rev .
* 105 * , 252301 ( 2010 ) k. dusling and r. venugopalan , phys .
d * 87 * , 094034 ( 2013 ) a. dumitru , k. dusling , f. gelis , j. jalilian - marian , t. lappi and r. venugopalan , phys .
b * 697 * , 21 ( 2011 ) k. dusling , f. gelis , t. lappi and r. venugopalan , nucl .
a * 836 * , 159 ( 2010 ) r. a. lacey , y. gu , x. gong , d. reynolds , n. n. ajitanand , j. m. alexander , a. mwai and a. taranenko , arxiv:1301.0165 .
a. bzdak , b. schenke , p. tribedy and r. venugopalan , phys .
c * 87 * , 064906 ( 2013 ) g. -y .
qin and b. mller , arxiv:1306.3439 [ nucl - th ] . c. ecoleman - smith and b. mller , phys .
d * 89 * , 025019 ( 2014 ) the author thanks jurgen schukraft for clarifying this issue in a private discussion l. mclerran , m. praszalowicz and b. schenke , nucl .
a * 916 * , 210 ( 2013 ) l. mclerran and m. praszalowicz , acta phys .
b * 42 * , 99 ( 2011 ) d. higinbotham , g. amiller , o. hen and k. rith , cern courier * 53n4 * , 24 ( 2013 ) f. cooper and g. frye , phys .
d * 10 * , 186 ( 1974 )
. y. cheng , l. p. csernai , v. k. magas , b. r. schlei and d. strottman , phys .
c * 81 * , 064910 ( 2010 ) g. basar and d. teaney , arxiv:1312.6770 [ nucl - th ] . g. c. rossi and g. veneziano , nucl .
b * 123 * , 507 ( 1977 ) .
s. e. vance and m. gyulassy , phys .
* 83 * , 1735 ( 1999 ) i. bouras , a.
el , o. fochler , c. greiner , e. molnar , h. niemi and z. xu , acta phys .
b * 40 * , 973 ( 2009 ) .
k. tuchin , phys .
b * 593 * , 66 ( 2004 ) | i review the models suggested , to date , as an explanation for the so called `` ridge '' phenomenon , an elongation in rapidity of 2-particle correlations seen at rhic and lhc energies .
i argue that these models can be divided into two phenomenologically distinct classes : `` hotspot+flow '' driven correlations , where initial state correlations created by structures local in configuration space are collimated by transverse flow , and models where the azimuthal correlation is created through local partonic interactions in a high gluon density initial state .
i argue that the measurement of a strong double ridge in @xmath0 and @xmath1 collisions allows a good opportunity to understand the ridge s origin because it allows to see if a common knudsen number scaling , expected if the ridge has a hydrodynamic origin , can be used to understand all data .
i show that current data presents evidence that this scaling is lacking , presenting a challenge to the hydrodynamic models .
on the other hand , particle - identified correlations are a particularly promising way of testing the assumption , which distinguishes the two models , of whether the correlation is formed initially in the partonic phase , or as a final state effect .
assuming fragmentation occurs `` as in vacuum '' can be used to predict scaling trends which are generally broken by models , such as hydrodynamics , where the ridge is created as a final state effect . while evidence is again not fully conclusive , data does seem to follow a scaling compatible with hydrodynamics @xcite .
i close by discussing experimental observables capable of clarifying the situation |
the dressing of a nucleon with its meson cloud is deeply and consistently connected to the spontaneous chiral symmetry breaking exhibited by quantum chromodynamics ( qcd ) . in the specific case of deep - inelastic - scattering ( dis ) ,
the presence of a meson cloud has relevant consequences as first discussed by feynman and sullivan @xcite,@xcite .
one had to wait till the discovery of the gottfried sum rule violation @xcite , to realize that the cloud contribution can quantitatively account for the excess of @xmath9 anti - quarks over @xmath10 anti - quarks @xcite .
the inclusion of a strange component into the meson cloud has been first discussed by signal and thomas @xcite showing that the strange cloud component exhibits peculiar characteristics since all the @xmath11 antiquarks come from the kaon , whereas all the @xmath12 quarks come from the hyperon . in particular it was demonstrated that the meson cloud contribution to the anti - strange distribution is softer than the contribution to the strange component .
some shortcomings were present in the approach proposed in refs.@xcite and a better formulation of the meson cloud model was proposed within a time ordered perturbation theory in the infinite momentum frame @xcite . using time ordered approach one has , in fact , the advantage that the struck hadrons remain on - mass - shell avoiding ambiguities and allowing the use of experimental structure functions as input .
the question of a possible quark - antiquark asymmetry in the strange sea received new interest in the early 2000 s as a result of the experimental evidence of the nutev collaboration @xcite for a significantly smaller value of the weak mixing angle .
the measure has to be corrected for several charge symmetry violation ( csv ) effects @xcite - @xcite and a possible significant contribution of the second moment of the strange asymmetry @xmath13 . as matter of fact a quark - antiquark asymmetry
@xmath14 would explain the discrepancy between the nutev value and the accepted value of the weak mixing angle @xmath15 .
however , at present , even the sign of the strange asymmetry is not known and the phenomenological values range in the interval @xmath16 .
in particular @xmath17 from ref.@xcite ; @xmath18 from ref.@xcite ; @xmath19 from ref.@xcite ; @xmath20 from refs.@xcite .
the strange - antistrange asymmetry manifests not only the non - perturbative component due to strange meson cloud , but it exhibits also a dependence on the hard - scattering scale , @xmath21 , at which the nucleon is probed .
perturbative qcd alone definitely predicts a nonvanishing , @xmath21-dependent value of the strange asymmetry .
in fact the non - singlet evolution of the parton densities at three loops level ( next - to - next - to - leading order : nnlo ) generates a strange asymmetry of the order @xmath22 at @xmath23 even if the strange asymmetry is vanishing at the starting scale , an effect which occurs in both qcd and qed , and it is a genuine quantum phenomenon @xcite . as a consequence the three - loops level is the correct and needed order for discussing parton distributions when strange sea is included , in particular strange asymmetry @xcite , @xcite .
the present paper will make use of an original ( matlab ) nnlo evolution code built by using the unpolarized splitting functions of refs.@xcite and making reference to the pegasus ( fortran ) code by andreas vogt @xcite ( see sec .
[ sec : evolution ] and the appendix ) .
the matlab code is a descendant of the nlo ( fortran ) code written fifteen years ago in collaboration with vento , mair and zambarda @xcite .
various versions of that code have been regularly used by the author and by others both for unpolarized @xcite and polarized @xcite parton distributions , as well as the evolution of their more complex generalization ( the generalized parton distributions , gpds ) @xcite . the perturbative approach to qcd
is able to connect observables at different resolution scales , but the full knowledge of the consequences of the gluon and quark dynamics require the input of unknown non - perturbative matrix elements to provide absolute values for the observables at any scale . in the present paper
i will apply a radiative parton model procedure which , starting from low resolution scale @xmath24 , has been able to reproduce and predict @xcite important features of the experimental deep inelastic structure functions at high momentum transfer ( see also @xcite and @xcite ) .
the procedure assumes the existence of a scale where the short range ( perturbative ) part of the interaction is suppressed ( and therefore the gluons and the sea are suppressed ) and the long range part of the interaction produces a proton composed mainly by three valence quarks @xcite .
jaffe and ross @xcite proposed to ascribe the quark model calculations of matrix elements to that hadronic scale @xmath24 . in this way the quark models , summarizing a great deal of hadronic properties , may substitute the low energy parametrization while evolution to larger @xmath21 is dictated by perturbative qcd . in the following
i will use a relativistic quark model ( formulated within a light - front dynamics ) to describe the three quark bound states .
the inclusion of the non perturbative sea contribution is obtained within a fluctuation expansion of meson - baryon states in the infinite momentum frame .
non - strange ( @xmath0 , @xmath1 , @xmath2 , together with @xmath5 , @xmath6 ) as well a strange ( @xmath3 and @xmath4 , together with @xmath7 and @xmath8 ) meson - baryon fluctuations are considered following the rich literature @xcite . the light - front quark model with a meson cloud
is revisited in sec.[sec : model ] ( the bare nucleon model is discussed in sec.[subsec : bare ] , while sec.[subsec : cloud ] is devoted to investigate the role of the nucleon cloud within different scenarios ( sec.[subsubsec : scenarios ] ) ; sec.[sec : evolution ] is dedicated to the nnlo evolution procedure and its features ( expanded also in the appendix ) , results are presented and discussed in sec.[sec : results ] .
some conclusions are drawn in sec.[sec : conclusions ] . in the appendix the relevant formulae for the adopted nnlo evolution procedure .
in the light - front quark model the intrinsic momenta of the constituent quarks ( @xmath25 ) can be obtained from the corresponding momenta ( @xmath26 ) in a generic reference frame , through a light - front boost @xmath27 such that the wigner rotations reduce to identities . with the specific choice @xmath28 one has @xmath29 and @xmath30 .
the nucleon state is characterized by isospin ( and its third component ) , parity , light - front ( non - interacting ) angular momentum operators @xmath31 and projection @xmath32 , where the unitary vector @xmath33 defines the spin quantization axis .
the nucleon state factorizes into @xmath34 , where @xmath35 is the total light - front nucleon momentum @xmath36 . @xmath37 and the subscript @xmath38 indicates the perpendicular projection with respect to the @xmath39 axis . in order to retrieve the usual composition rules , the intrinsic light - front angular momentum eigenstate @xmath40 must be obtained from the _ canonical _ angular momentum eigenstates @xmath41 by means of a unitary transformation that is a direct product of generalized melosh rotations @xcite .
finally the intrinsic part of the nucleon state , @xmath42 , is eigenstate of the mass operator @xmath43 where the interaction term @xmath44 must be independent of the total momentum @xmath45 and invariant under rotations @xcite . in the following
i will discuss results of a confining three quark mass operator @xmath46 where @xmath47 is the radius of the hypersphere in six dimensions and @xmath48 and @xmath49 are the intrinsic jacobi coordinates @xmath50 and @xmath51 .
the intrinsic nucleon state is antisymmetric in the color degree of freedom and symmetric with respect the orbital , spin and flavor coordinates .
in particular ( disregarding the color part ) , one has @xmath52}^{(0,0)}(\omega ) \ , { \left [ \chi_{\rm ms}\ , \phi_{\rm ms } + \chi_{\rm ma } \ , \phi_{\rm ma } \right ] \over \sqrt 2}\ , , \label{n00-wf}\ ] ] where @xmath53 is the hyper - radial wave function solution of eq.([eq : massopxi ] ) , @xmath54}^{(l , m)}(\omega)$ ] the hyperspherical harmonics defined on the hypersphere of unitary radius , and @xmath55 and @xmath56 the flavor and spin wave functions of mixed @xmath57 symmetry . in order to preserve relativistic covariance @xmath56
has to be formulated by means of the appropriate melosh transformation of the @xmath58-th quark spin wave function .
the mass equation ( [ eq : massopxi ] ) is solved numerically by expanding the hyper - radial wave function @xmath53 on a truncated set of hyper - harmonic oscillator basis states @xcite . making use of variational principles the ho
constant has been determined and convergence has been reached considering a basis of 17 components .
the parameters of the interaction have been determined phenomenologically in order to reproduce the basic features of the ( non - strange ) baryonic spectrum up to @xmath59 gev , namely the position of the roper resonance and the average value of the @xmath60 states .
the well known problem of the energy location of the roper resonance is solved , in the present case , by the use of the `` coulomb - like '' potential @xmath61 as already discussed in ref.@xcite for the nonrelativistic formulation and in ref.@xcite within the light - front dynamics .
one obtains @xmath62 and @xmath63 @xmath64 ( to be compared with the corresponding non - relativistic fit @xmath65 and @xmath66 @xmath64 ) @xcite . as a result a huge amount of high momentum components is generated in solving the mass equation and they play an important role on the high-@xmath67 behavior of the parton distributions .
such effect emerges naturally if one evaluate the momentum distribution @xmath68 as it is shown in fig.[fig : nk_vr ] .
mass equation ( [ eq : massopxi ] ) does not contain hyperfine interaction terms which would split the nucleon and the @xmath6 masses .
such @xmath69-breaking term must be considered in a more detailed analysis of the nucleon and baryon spectrum and the nucleon wave function ( [ n00-wf ] ) would include a larger number of @xmath70 configurations ( e.g. ref@xcite ) .
the consequences and limitations due to the choice ( [ eq : massopxi ] ) will become more clear when i will discuss the large-@xmath67 behavior of the valence partons in sec .
[ sec : valencegluons ] .
for the moment let me recall that i am mostly interested in the sea distribution , located at lower-@xmath67 . of eq.([eq
: nk_vr ] ) as function of @xmath71 . the results of a full covariant light - front calculation for the mass operator of eq.([eq : massopxi ] ) ( full curve ) are compared with the non - relativistic approximation ( dashed curve ) .
the normalization is such that @xmath72 . ]
following previous work ( e.g. refs.@xcite ) the parton distribution in the ( bare ) relativistic light - front quark model takes the form @xcite : @xmath73}(\{\vec{k}_i;\lambda_i,\tau_i\})\vert^2\ , , \label{eq : lfq - val}\end{aligned}\ ] ] where @xmath74 is the quark light - cone momentum , and @xmath75 is the free mass for the three - quark system .
@xmath76}(\{\vec{k}_i;\lambda_i,\tau_i\})$ ] is the canonical wave function of the nucleon in the instant form obtained by solving the eigenvalue equation for the mass operator ( [ eq : massopxi ] ) in momentum space . the distributions ( [ eq : lfq - val ] ) automatically fulfills the support condition and satisfies the ( particle ) baryon number and momentum sum rules at the hadronic scale @xmath77 where the valence contribution dominates the twist - two response : @xmath78 with @xmath79 being the number of valence quarks of flavor @xmath80 ; in addition @xmath81 = 1\,,\end{aligned}\ ] ] and the valence partons exhaust the momentum sum rule at the scale of the bare nucleon .
let me now introduce the meson - cloud model to incorporate @xmath82 pairs into the valence - quark picture of the parton distributions described in the previous section , and dressing the bare nucleon to a physical nucleon .
the physical nucleon state is built expanding it ( in the infinite momentum frame ( imf ) and in the one - meson approximation ) in a series involving bare nucleons and two - particle , meson - baryon states .
its wave function can , therefore , be written as a sum of meson - baryon fock states @xmath83 @xmath84 is the probability amplitude of the fock state containing a virtual baryon ( @xmath85 ) with longitudinal momentum fraction @xmath86 , transverse momentum @xmath87 , and helicity @xmath88 , and a meson ( @xmath89 ) with longitudinal momentum fraction @xmath90 , transverse momentum @xmath91 , and helicity @xmath92 .
@xmath93 is the renormalization constant and is equal to the probability to find the bare nucleon in the physical nucleon . [
cols="^,^,^,^,^,^ " , ] [ tableiv ] recently the results shown in fig.[fig : hermescomparison2013 ] have been critically reviewed @xcite arguing that the sole analysis of the sum of @xmath94 multiplicities , as done in the hermes paper @xcite , might not be sufficient to draw solid conclusion about the ( rather elusive ) strange sector .
a simultaneous analysis of the _ difference _ of @xmath94 multiplicities should be performed ( see also ref .
complementary information come from the differential measurements of the inclusive @xmath95 and @xmath93 boson cross section at the lhc , recently performed by the atlas collaboration @xcite using @xmath96 collision data recorded in 2010 @xcite .
because of the weak couplings of the quark involved , complementary information to @xmath97 is provided constraining the total light sea @xmath98 .
the parton distribution analysis of ref .
@xcite is performed at nnlo ( and directly comparable with the present approach ) using atlas data jointly with inclusive deep inelastic scattering data from hera .
the atlas analysis is compared with the present theoretical results in fig.[fig : xsigmaatlas ] where the total light atlas sea data @xmath99 for a fitted fraction of strangeness of about unity ( squares ) are shown in the region @xmath100 and compared with the nnlo evolutions to the scale of the data , i.e. @xmath101 gev@xmath102 . predictions for both scenarios are shown to appreciate the role of the unperturbed strange sea .
in particular it is evident that the experimental data at low-@xmath67 ( @xmath103 ) show a shoulder not reproduced by the present approach while the data at larger @xmath67 ( @xmath104 ) are well reproduced in both scenarios .
the large perturbative production of strange sea due ( in the present approach ) to the large evolution scale difference seems to be at the origin of the steep behavior of the total sea below @xmath105 . in fig.[fig : asymmetrytotq2_4 ] the results for the asymmetry @xmath106 at @xmath107 gev@xmath102 are shown . as already discussed
such a quantity is extremely sensitive to the choice of the quark wave functions and meson cloud content .
the results shown ( within scenario * b * ) refer to the choice of eq.([eq : sk ] ) ( see sec .
[ subsubsec : strangeness ] for a discussion ) .
the figure emphasizes also the contribution due to the presence of an asymmetric strange component in the meson cloud and the role of perturbative evolution at different order .
in particular the asymmetry vanishes at lo and nlo if one neglects asymmetric strange non - perturbative components ( namely the @xmath3 and @xmath4 fluctuations ) . at nnlo
a large perturbative component is present both for scenario * a * and * b * ( cfr . also ref.@xcite ) . as a matter of fact the perturbative part gives a significantly large contribution to the asymmetry at nnlo softening the dependence of the present results on the cloud model and stressing the role of nnlo effects .
+ fig.[fig : asymmetrynp_p_q2_16 ] emphasizes such a conclusion showing the effects of non - perturbative and perturbative components at the scale @xmath108 gev@xmath102 .
disentangle , at such large resolution scale , perturbative and non - perturbative contributions to the strange asymmetry is assured by a _
complete _ parton distribution evolution at nnlo as in the present approach .
an approximate attempt has been proposed in ref.@xcite where the analysis is performed evolving in a separate way the non - perturbative strange asymmetry due to the meson cloud and valence distribution ( related to the perturbative component of the asymmetry , in the @xmath67-space ) . neglecting the interference effects between the two responses , in a small and delicate quantity like the strange asymmetry , can introduce uncontrolled uncertainties . in fig.[fig : asymmetrynp_p_q2_16 ] a comparison with the results by feng _
@xcite is explicitly performed at the scale @xmath109 gev@xmath102 chosen by those authors .
table [ tableiv ] confirms that the asymmetry is not well constrained at nlo and a nnlo investigation is mandatory .
an observation which helps in drawing some conclusion also on the role of strange asymmetry in the analysis of the nutev experiment .
charge symmetry violations play an important role in the interpretation of the experiment , however the sources of uncertainty and the needed corrections are of more general origin including electrodynamic isospin violations and nuclear effects on structure functions . taking into account all the corrections coherently ( for a summary see ref .
@xcite ) no more anomaly in the nutev experiment can be invoked even for an accepted ( mean ) value of zero strange asymmetry . in ref .
@xcite , for instance , i have demonstrated that the interpretation of the experiment is consistent with a value of @xmath110 and it is not necessary to invoke large strange asymmetry to explain the nutev result .
gev@xmath102 ( scale of the h1 collaboration experiments @xcite ) within scenario * b*. continuous lines lines represent the results of the evolution at nnlo .
the dot - dashed lines the results at nlo , the dotted curves the lo results . ]
gev@xmath102 ( scale of the h1 collaboration experiments @xcite ) .
the dashed lines show the result obtained without including strange components in the meson - cloud at ( non - perturbative level , scenario * a * ) .
the continuous lines represent the numerical results obtained when the non - perturbative strange sea component is introduced explicitly at the model scale ( scenario * b * ) .
triangles represent the results of the dynamical fit to the experimental data of ref .
] i have emphasized , till now , the effects of non - perturbative and perturbative origin on the specific flavor sea components , in particular strangeness .
the results show strong nnlo effects with respect to nlo , as if the nnlo expansion had no perturbative meaning .
however one has to notice that i discussed the largest effects and therefore also effects which appears _ for the first time _ at nnlo .
for those observables ( the strange asymmetry is the prototype ) the perturbative expansion _ starts _ at nnlo . in this
section results of genuine higher - order will be investigated , in particular the effects on the non - singlet valence and gluon distributions .
both these observables are strongly influenced by perturbative expansion .
the valence partons share almost the total momentum at low - resolution scale , a choice discussed in sec .
[ subsubsec : scenarios ] and [ subsec : q0 ] .
the introduction of the effects due to the meson cloud lowers the total valence momentum from 1 to @xmath111 within scenario * a * and to @xmath112 within scenario * b * ( cfr .
( [ eq : q0scenarios ] ) of sec .
[ subsec : q0 ] ) . during the evolution
gluon radiation enhances the gluon and sea components and lowers the valence contribution keeping the total number of valence quarks and total momentum fixed .
a typical example is shown in fig.[fig : nnlo_nlo_lopartons_q26464c ] where the initial partons of scenario * b * are evolved at nnlo , nlo and lo from low resolution scale to @xmath113 , a scale of the h1 collaboration experiments @xcite .
the huge amount of sea and gluon components is stressed by the factor 100 needed to make the figure illustrative .
also the convergence of the perturbative expansion shows up clearly despite the large range chosen at low-@xmath67 . in fig.[fig
: nnlo_strange_partons_q26464d ] the nnlo evolution at the h1 scale @xmath114 gev@xmath102 is shown for both scenarios * a * and * b * making clear the role of the non - perturbative meson cloud in calculating the valence , total sea and gluon distributions .
let me stress that the results represent absolute predictions starting from a quark model wave function on the light - front and meson fluctuations .
the model is fixed and also the low resolution scale , without adjustable parameters .
the comparison with the dynamical fits of ref .
@xcite ( evolved at the same scale ) gives a quantitative comparison with the experimental results .
the investigation of valence partons implies the description of the distributions at large-@xmath67 ( @xmath115 ) .
this region manifests specific aspects : higher - order twist effects could become important as well as the re - summation of large @xmath116 contribution to the evolution near the @xmath117 threshold @xcite .
the model i am presenting could be the framework to investigate also such effects and work in this direction is in progress @xcite .
a more standard aspect , of the large-@xmath67 region has been mentioned at the end of sect.[subsubsec : lfqm ] : su(6 ) breaking effects should show up in that region .
in particular the ratio of the neutron to proton structure functions ( @xmath118 ) is predicted to be @xmath119 in model where the @xmath120 distribution dominates with respect the @xmath121 distribution at large-@xmath67 . in @xmath69-symmetric models where the symmetry implies @xmath122 ,
the ratio is simply constant and , in the region where the valence quark contribution dominates , equal to @xmath123 .
i am using a @xmath69-symmetric model and the expected result ( @xmath123 ) is in disagreement with the experiments which confirm a lower limiting value ( for a critical review on the extraction of the @xmath124 ratio at large-@xmath67 , see , for example , ref.@xcite ) .
while all familiar parton distributions vanish at @xmath125 , ratios of two of them need not and , under evolution , the value of such a ratio is ( practically ) invariant and can reflect significant non - perturbative features of qcd @xcite .
since fundamental aspects of non - perturbative qcd are encoded in the ratios at large-@xmath67 , it seems that a direct relation between the @xmath69-breaking effects at low - energy and the parton distributions can be established in a simple way .
actually the situation is more intricate as it has been demonstrated calculating the @xmath118 ratio both within non - relativistic @xcite and relativistic quark - parton models @xcite as input scale .
to summarize the situation i will refer to the paper by pasquini , traini and boffi @xcite dedicated explicitly to that topic .
it has been shown that the effects due to the @xmath69-breaking needed at low - energy to reproduce the mass spectrum ( or other nucleon properties ) can not reproduce the high-@xmath67 behavior of the @xmath118 ratio @xcite .
apparently such a conclusion contradicts other contributions ( e.g. the well known paper by close @xcite ) .
the reason is rather simple : the calculation by close ( as well as other similar investigation and references therein ) does not satisfy the pauli principle for the symmetry of the three - quark wave - function , an observation due to nathan isgur @xcite .
the approach proposed in ref.@xcite has the advantage to fully satisfy the pauli principle of the quark and parton distributions opening the possibility of a direct connection between the effects incorporated within low - energy quark - models and the parton dynamics .
the result is mainly negative : the @xmath69-breaking manifested by the large-@xmath67 behavior can not be reduced to the @xmath69-breaking mechanism at low - energy .
the approach i am proposing in the present manuscript seems to be promising also for investigating in more detail the large-@xmath67 region once higher order effects @xcite and re - summation contribution have been integrated .
as already mentioned , work , in this direction is in progress @xcite .
the radiative parton model has been demonstrated to have a predictive power since the pioneer work of glck , reya and vogt @xcite where the small @xmath67-behavior of parton distributions was predicted and later checked experimentally .
the meaning of such experimental behavior at low-@xmath67 is rather simple : the structure functions is entirely due to qcd dynamics at @xmath126 and the parton distributions at @xmath127 gev@xmath102 are generated radiatively from input distributions at an optimally determined low - scale @xmath128 gev@xmath102 .
the phenomenological view @xcite that there exist a scale where the short - range ( perturbative ) part of the interaction is suppressed ( and therefore the gluons and the sea ) and that the long - range part of the interaction reveals the nucleon as a composite system of ( mainly ) three quarks , receives some specific support . following this suggestive hypothesis
the paper investigates the actual results of parton distributions generated when the low - input scale is rigidly fixed by a detailed model of the nucleon which includes quarks and meson - cloud .
the light - front hamiltonian dynamics fixes the three - quark wave functions and the meson cloud is introduced by means of high - order fock component in time - ordered perturbation theory in the infinite momentum frame .
non - strange as well as strange mesons - baryon fluctuations are considered following the rich literature ( @xmath0 , @xmath1 , @xmath2 , @xmath3 and @xmath4 , together with @xmath5 , @xmath6 , @xmath7 and @xmath8 ) and the effects on strange and non - strange parton distributions investigated in detail showing the large effects due to ( non - strange ) sea asymmetries and the delicate balance of strange asymmetry .
the total strange component is compared with the available experiments .
the intense comparison of the results of the present approach with more conventional fits of parton distributions shows the quality of a study fully based on the quark and meson dynamics at low resolution scale .
the dynamical origin of some features of the parton distributions have been directly related to properties of the nucleon wave functions and meson - baryon fluctuations . in this way
the presence of a strange asymmetry can be correlated to both perturbative ( nnlo is mandatory ! ) and non - perturbative effects ( the non -perturbative strange sea can not be neglected ! ) . despite the lack of gluons at low resolution scale (
non perturbative gluons have been neglected in favor of a non - ambiguous determination of the sea contribution and the related input scale ) the large amount of gluons at small-@xmath67 ( @xmath129 ) is reasonably reproduced together with the samll-@xmath67 behavior of the total sea . again
, nnlo approximation produces results better in agreement with data fits .
the model is complete and flexible enough ( both in the non - perturbative and perturbative sector ) to investigate in detail the role of the individual mesons in the virtual cloud . in this way , for instance , the role of both @xmath3 and @xmath4 strange mesons have been investigated with particular emphasis on the total strange sea and asymmetry .
the completeness of the approach is also seen by the possibility of integrating subsidiary elements in the study of delicate quantites like charge - symmetry - violation effects both of strong and electrodynamics ( qed - radiative corrections ) origin
. the topic has been shortly mentioned , but a specific contribution has been recently devoted to a detailed analysis ( cfr .
ref.@xcite ) within the same approach .
it seems to me that one of the actual potentiality of the radiative dynamical approach to parton distributions is just opening the door to the connection of low - energy model and the large amount of high - energy data .
the discussion devoted to the choice of the ( initial ) low - resolution scale in sec .
[ sec : evolution ] ( in particular the sec . [ freezing ] ) on the freezing of the qcd running coupling @xcite , has to be further elaborated , but can be a real interesting key to bridge different worlds .
i thank jean - paul blaizot for the warm hospitality at the institute de physique theorique , saclay , in february 2013 , where part of this work has been done .
i am member of the interdisciplinary laboratory for computational science ( lisc ) , a joint venture of trento university and fbk ( bruno kessler foundation ) .
the mellin @xmath5-th moments of the parton distributions @xmath130 ( @xmath131 ) , @xmath132 evolve according to the coupled flavor-@xmath133 evolution equation @xmath134\,\ , , \label{mellin - s - evol}\ ] ] where has been introduced to define the functions @xmath135 which have been used in our previous work @xcite . ] @xmath136 with @xmath137 and @xmath138 the @xmath139 @xmath140 and @xmath141 are called the lo , nlo and nnlo ( 3-loop ) splitting functions @xcite . in the flavor-@xmath142-@xmath133 ( ns )
sector eq.([mellin - s - evol ] ) is uncoupled and reads : @xmath143\,\ , , \label{mellin - ns - n}\ ] ] with ( see footnote ) @xmath144 and @xmath145 refers to the ns splitting functions @xmath146 and @xmath147 ( i.e. @xmath148 ) .
they are the evolving functions for the ns combinations of parton distributions @xmath149 , @xmath150 , where @xmath151 and @xmath152 .
the nnlo splitting functions @xmath153 can be found in ref.@xcite . for the evolution i used the basis @xcite @xmath154 with @xmath155 . after performing the evolution individual quark and anti - quark distributions
ca be recovered by means of the relations @xmath156 where @xmath157 , as well as the corresponding equation for the differences @xmath158 .
since no matrices are involved in the ns evolution , let us start the discussion with the evolution of that sector .
combining equations ( [ alpha_s ] ) , ( [ mellin - ns - n ] ) and ( [ pns - expansion ] ) one can express the evolution of the ns - moments @xmath160 and the solution can be written @xmath161 where @xmath162 .
the previous integrals have a closed form , and for @xmath163 ( lo ) , @xmath164 ( nlo ) and @xmath165 ( nnlo ) become : @xmath166^{\lambda_0(n ) } \ , ; \label{ns - nlo}\\ \nonumber \\ \nonumber \\
\left.{\langle q_{\rm ns}(q^2)\rangle_n \over \langle q_{\rm ns}(q_0 ^ 2)\rangle_n}\right|_{\rm nnlo } & = & \exp\left\{-\int^{a_s}_{a_{s0}}da_s\,{a_s p^{(0)}_{\rm ns}(n)+ a^2_s p^{(1)}_{\rm ns}(n ) + a^3_s p^{(2)}_{\rm ns}(n ) \over \beta_0 a_s^{2 } + \beta_1 a_s^3 + \beta_2 a_s^4 } \right\ } = \nonumber \\ \nonumber \\ & = & \left({a_s \over a_{s0}}\right)^{-{p^{(0)}_{\rm ns}(n)\over \beta_0 } } \ , \left[{1-{1 \over x_1 } a_s \over 1- { 1 \over x_1 } a_{s0}}\right]^{\lambda_1(n)}\,\cdot \,\left[{1-{1 \over x_2 } a_s \over 1- { 1 \over x_2 } a_{s0}}\right]^{\lambda_2(n)}\,;\label{ns - nnlo}\end{aligned}\ ] ] where @xmath167 and @xmath168\ , ; \nonumber
\\ \lambda_2(n ) & = & -{1 \over \beta_2 } { 1 \over x_1-x_2 } \left[- p_{\rm ns}^{(1)}(n ) + { \beta_1 \over \beta_0 } p_{\rm ns}^{(0)}(n ) - x_2 \left(p_{\rm ns}^{(2)}(n ) - { \beta_2 \over \beta_0 } p_{\rm ns}^{(0)}(n)\right)\right]\,.\label{lambda1lambda2}\end{aligned}\ ] ] eq.([ns - lo ] ) is the well known lo evolution expression for ns components and it keeps a symmetric form for evolutions from @xmath169 and back .
eq.([ns - nlo ] ) is its analog at nlo , but it is less known in the literature .
the reason is rather simple : it contains higher order corrections in @xmath170 and several authors prefer to expand eq.([ns - nlo ] ) keeping lower order terms ( in @xmath170 and @xmath171 ) around the lo solution .
developing eq.([ns - nnlo ] ) , one gets @xmath172 \ , ; \label{ns - nlo - approx}\ ] ] an expansion which has to pay the price of a non symmetric form in @xmath21 and @xmath24 , loosing the possibility of a backward evolution into the region of validity of the same eq.([ns - nlo ] ) . in our previous works we tried to keep a symmetric form to establish a backward evolution needed to fix the actual scale of some nucleon model , and
we have developed the following nlo expansion @xcite @xmath173 \ , ; \label{ns - nlo - approx - ours}\ ] ] which can also be derived from eq.([ns - nlo ] ) expanding the numerator and the denominator independently . , see
eqs.([mellin - s - evol]),([mellin - ns - n ] ) . ] the nnlo ns evolution of eq.([ns - nnlo ] ) , found application after the explicit calculation of all splitting functions at nnlo @xcite . similarly to the nlo expression ( [ ns - nlo ] ) , also the nnlo expression ( [ ns - nnlo ] ) contains higher order corrections and some author prefers to further expand it around the lo order result ( [ ns - lo ] ) ( e.g. @xcite and references therein ) .
the expansion would lead to ( let me use a simplified notation dropping the @xmath5 and @xmath21 dependence where obvious ) : @xmath174\,a_s^2\ ; \over 1 - \left({\lambda_1 \over x_1 } + { \lambda_2 \over x_2 } \right)\,a_{s0 } + { 1 \over 2 } \left[\left({\lambda_1 \over x_1 } + { \lambda_2 \over x_2 } \right)^2 -\left({\lambda_1 \over x_1 ^ 2 } + { \lambda_2 \over x_2 ^ 2 } \right ) \right]\,a_{s0}^2 } \right\}= \nonumber \\ \label{ns - nnlo - ours } \\ & = & \left({a_s \over a_{s0}}\right)^{-{p^{(0)}_{\rm ns}(n)\over \beta_0 } } \left\ { { 1 - r_1 \,a_s\ ; + { 1\over 2 } \left(r_1 ^ 2 - r_2\right)\,a_s^2\ ; \over 1 - r_1 \,a_{s0 } + { 1\over 2 } \left(r_1 ^ 2 - r_2\right)\,a_{s0}^2 } \right\}= \nonumber \\ \label{ns - nnlo - ours1}\\ & \approx & \left({a_s \over a_{s0}}\right)^{-{p^{(0)}_{\rm ns}(n)\over \beta_0 } } \left\{1 - \left({\lambda_1 \over x_1 } + { \lambda_2 \over x_2 } \right ) \left(a_s - a_{s0}\right)+ { 1\over 2 } \left[\left({\lambda_1 \over x_1 } + { \lambda_2 \over x_2 } \right)^2 + \right.\right .
\nonumber \\ & - & \left .
\left({\lambda_1 \over x_1 ^ 2 } + { \lambda_2 \over x_2 ^ 2 } \right ) \right ] \left(a_s^2 - a_{s0}^2\right ) - \left({\lambda_1 \over x_1 } + { \lambda_2 \over x_2 } \right)^2\,a_{s0 } \left(a_s - a_{s0}\right)\right\ } = \label{ns - nnlo - approx1}\\ & = & \left({a_s \over a_{s0}}\right)^{-{p^{(0)}_{\rm
ns}(n ) \over \beta_0 } } \left[1 - ( a_s - a_{s0 } ) r_1 + { 1 \over 2 } ( a_s^2 - a_{s0}^2 ) \left(r_1 ^ 2 - r_2\right ) - a_{s0 } ( a_s - a_{s0 } ) r_1 ^ 2 \right]\ , , \label{ns - nnlo - jdr}\end{aligned}\ ] ] with @xmath175 .
explicitly : @xmath176\label{r1}\\ r_2 & = & { 1 \over \beta_0 } \left\{p_{\rm ns}^{(2 ) } -{\beta_1 \over \beta_0 } p_{\rm ns}^{(1 ) } + \left[\left({\beta_1 \over \beta_0}\right)^2 - { \beta_2 \over \beta_0}\right ] p_{\rm ns}^{(0)}\right\}\label{r2}\end{aligned}\ ] ] and from ( [ lambda1lambda2 ] ) ( [ r1 ] ) and ( [ r2 ] ) @xmath177 which check the validity of the equality ( [ ns - nnlo - ours ] ) ( [ ns - nnlo - ours1 ] ) .
summarizing : : * the basic ns evolution at nnlo is given by eq.([ns - nlo ] ) and its symmetric ( [ ns - nnlo - ours1 ] ) or not symmetric ( [ ns - nnlo - jdr ] ) expansions ; * the coupling constant @xmath178 is given , at nnlo by the solution of the transcendental equation ( [ alpha_snnlo ] ) .
the approxximation ( [ alpha_snnloapprox ] ) is not appropriate within a radiative approach like ours , because the starting point of the evolution is rather low ; * the use of a symmetric evolution form is specifically suitable for fixing the low - resolution scale allowing for a forward and back evolution of valence partons .
the singlet splitting - function matrices @xmath180 of different orders @xmath181 do not commute .
therefore the solution of the evolution equation ( [ mellin - s - evol ] ) can not be written in a closed exponential form beyond lo and one is left with a series expansion around the lowest order solution . differently from the @xmath182 evolution
, one can not introduce symmetrized version of the evolution at nnlo ( for an attempt at nlo see ref.@xcite ) .
the procedure is lengthy , but standard and i refer to the paper by a. vogt @xcite for a clear discussion .
the nnlo ( truncated ) evolution can be written @xmath183\ , \langle \vec q_{\rm s}(q_0 ^ 2)\rangle_n\ , , \label{singlet - n - evolution}\ ] ] the lo and nlo approximations are obtained from eq.([singlet - n - evolution ] ) by respectively retaining only the first , or the first second and third terms ( linear in @xmath170 and @xmath171 ) in the square bracket @xmath184 \,\langle \vec q_{\rm s}(q_0 ^ 2)\rangle_n\,.\end{aligned}\ ] ] with @xmath185 and @xmath186 .
the projection matrices @xmath187\ ] ] and @xmath188 denote the smaller ( larger ) eigenvalue of @xmath189 : @xmath190\;\;\;\;{\rm i.e.}\;\;\;\ ; \hat r_0 = \lambda_- \hat e_- + \lambda_+ \hat e_+\,.\ ] ] the @xmath191 matrices are defined by @xmath192 with @xmath193 for the actual calculations : * all the matrix manipulations have been performed numerically ; * the nnlo splitting @xmath194 as well as the standard lo @xmath195 and nlo @xmath196 are from ref.@xcite ; * the bjorken-@xmath67 distributions @xmath197 and @xmath198 are obtained numerically inverting the mellin moments in the complex-@xmath5 plane by means of a contour integral around the singularities of @xmath199 and @xmath200 ( see for example @xcite .
zeller ( nutev collaboration ) , phys .
88 * , 091802 ( 2002 ) ; [ erratum - ibid .
* 90 * , 239902 ( 2003 ) ] ; g.p .
zeller ( nutev collaboration ) , phys .
d * 65 * , 111103 ( 2002 ) ; [ erratum - ibid .
d * 67 * , 119902 ( 2003 ) ] .
for a recent revision the reader could refer to the recent workshop _ flavor structure of the nucleon sea _
, ect@xmath201 : trento july 1 - 5 , 2013 ; website : http://www.ectstar.eu/node/95 and link to the talks : http://www.npl.illinois.edu/ect/program.php .
w. melnitchouck , j. speth and a.w .
thomas , phys .
d * 59 * , 014033 ( 1999 ) ; j. speth and a.w .
thomas , adv .
* 24 * , 83 ( 1997 ) ; j.t . londergan and a.w .
thomas , prog .
phys * 41 , 49 ( 1998 ) * in ref.@xcite different spin dependent forces which break @xmath69 are used , making the conclusions even more stringent . in particular models with cromo - magnetic interaction of one - gluon - exchange type @xcite and the goldstone - boson - exchange motivated by breaking of chiral symmetry and by effects due to instanton @xcite . | a detailed study of nucleon parton distribution functions is performed within a radiative next - to - next - to leading order ( nnlo ) parton model whose low - scale input is rigidly described by wave functions which include quarks and dressing meson - cloud . the light - front hamiltonian dynamics fixes the three - quark wave functions and the meson cloud is introduced by means of high - order fock components in time - ordered perturbation theory in the infinite momentum frame .
non - strange as well as strange meson - baryon fluctuations are considered ( @xmath0 , @xmath1 , @xmath2 , @xmath3 and @xmath4 , together with @xmath5 , @xmath6 , @xmath7 and @xmath8 ) and the effects on strange and non - strange parton distributions investigated showing the large effects due to ( non - strange ) sea asymmetries and the delicate balance of the strange asymmetry .
the non - strange and the strange components , the valence as well as the gluon distributions are compared with available experimental data confirming the need of both non - perturbative degrees of freedom and perturbative ( nnlo ) radiative effects . |
kauffman networks are disordered dynamical systems proposed by kauffman in 1969 as a model for genetic regulatory systems @xcite .
they attracted the interest of physicists in the 80 s @xcite , due to their analogy with the disordered systems studied in statistical mechanics , such as the mean field spin glass @xcite .
a dynamical phase transition was found and studied in the framework of mean field theory . in this and in the next paper
@xcite we deal with some structural properties of the networks that determine their attractors . in the present paper
we introduce the relevant elements , a notion that was suggested by flyvbjerg @xcite and flyvbjerg and kjaer @xcite , and we study their probability distribution . in the next one
we describe how the relevant elements are subdivided into asymptotically non communicating , independent modules .
the modular organization of random boolean networks was already suggested by kauffman @xcite , and it was used by flyvbjerg and kjaer to study analytically the attractors in @xmath4 networks .
we shall show that it is possible to describe the phase transition in random boolean networks in terms of the scaling of the number of relevant elements with system size , or in terms of a percolation transition in the set of the relevant elements .
the interest of this approach is that some consequences about the statistical properties of the attractors can be directly drawn . in @xcite
we computed the properties of the attractors in the framework of the annealed approximation , introduced by derrida and pomeau @xcite , but we observed that the results of this approximation are reliable only when the system is chaotic enough , becoming exact for a random map .
the study of the relevant elements is complementary to this approach , and we sketch the lines of a new approximation scheme that works better in the frozen phase and on the critical line .
this region in parameter space is the most interesting one , since , according to kauffman , it reproduces some features of real cells , and is also the less understood , since neither approximate computations nor simulations @xcite give a precise picture of the properties of the attractors for systems of large size . in next section
we define the model , discussing some old results together with open problems . in section 3
we define the relevant elements and in section 4 we give an approximate argument predicting the scaling of their number with system size in the different phases of the model . in the following section we present our numerical results , starting from the magnetization and the stable elements ( section 5.1 ) and then discussing the distribution of the relevant elements and its connection with the properties of the attractors , respectively in the chaotic phase ( section 5.2 ) and on the critical line ( section 5.3 ) .
the discussion of the results is postponed to our following paper @xcite , concerning the modular organization of the relevant elements on the critical line .
kauffman model is defined as follows .
we consider a set of @xmath1 elements @xmath5 and we associate to each of them a binary variable , @xmath6 . in the biological interpretation proposed by kauffman each element of the network
represents one gene and the binary variable @xmath7 represents its state of activation .
each element is under the control of @xmath8 elements , in the sense that its state at time @xmath9 is determined by the states at time @xmath10 of the @xmath8 control genes , @xmath11 and by a response function of @xmath8 binary variables , @xmath12 , that specifies how the element @xmath13 responds to the signals coming from its control variables .
the control elements are chosen in @xmath14 with uniform probability .
the response functions are also extracted at random , and it s believed that the properties of the model do not depend on the details of their distribution @xcite .
the rule most generally used in teh literature is the following : for each of the possible inputs @xmath15 we extract independently the value of @xmath16 , and we call @xmath17 the probability that @xmath16 is equal to 0 . the dynamics of the system obey the equation _
i(t+1)=f_i(_j_1(i),_j_k(i ) ) .
this evolution law is deterministic , but the system is disordered because the control rules ( elements and functions ) are chosen at random from the beginning and kept fixed : thus we deal with a statistical ensemble of deterministic dynamical systems , and we are interested in the statistical properties of systems of large size . for finite @xmath1 ,
every trajectory becomes periodic after a long enough transient time , and the configuration space is partitioned into the attraction basins of the different periodic orbits .
we are interested in the probability distributions of the number , the length and the size of the attraction basin of the periodic orbits , as well as in that of transient times . in the biological metaphor ,
given a set of rules ( a genome ) an attractor represents a possible cellular type , its length represents the duration of the cellular cycle , and the number of attractors represents the number of cells that can be formed with a given genome .
it was observed already in the first simulations that two dynamical regimes are present , and that the line separating them has properties reminiscent of those of real cells @xcite . in the so - called chaotic phase ( large connectivity , @xmath17 close to @xmath18 ) the average length of the cycles increases exponentially with system size .
the limit case of the chaotic phase , @xmath19 , was already known as random map in the mathematical literature , and was studied in detail by derrida and flyvbjerg @xcite , who pointed out interesting analogies between this system and the mean field spin glass @xcite concerning the distribution of the weights of the attraction basins . in the frozen phase ,
on the other hand , the typical length of the cycles does not increase with @xmath1 .
the limit case of this phase , @xmath4 , was analytically studied to some extent by flyvbjerg and kjaer @xcite , who introduced in that context the concept of relevant elements ( though without using this name ) . the first description of this dynamical phase transition in terms of an order parameter was given by derrida and pomeau @xcite .
they studied the evolution of the hamming distance between configurations in the kauffman networks approximating it with a markovian stochastic process .
such approximation ( the so - called annealed approximation ) was then shown to be exact in the infinite size limit , concerning the average value of the distance @xcite . below a critical line in parameter space
the average distance goes to zero in the infinite size limit ( _ frozen phase _ ) and above it the distance goes to a finite value ( _ chaotic phase _ ) .
the position of the phase transition depends only on the parameter @xmath20 , representing the probability that the responses to two different signals are different one has @xmath21 , so its value is comprised between zero and 1/2 , but for @xmath4 @xmath20 can be taken as an independent parameter in [ 0,1 ] ] , and is given by the equation @xmath22 . the properties of the attractors can be easily computed from the knowledge of the whole stationary distribution of the distance , and this can also be obtained within the annealed approximation @xcite , but the validity of this approximation in this more general case is not guaranteed .
comparison with simulations shows that the agreement is satisfactory in the chaotic phase , while the approximation fails on the critical line .
in the chaotic phase it is possible to compute the value of the exponent of the typical length of a cycle , @xmath23 , in good agreement with numerical results , but the distribution of cycle lengths is much broader than it is expected .
the annealed approximation predicts also that the distribution of the weights of the attraction basins is universal in the whole chaotic phase , and equal to the one obtained by derrida and flyvbjerg in the case of the random map @xcite .
the corrections to this prediction appear small , if any , even for @xmath24 .
finally , the number of different cycles in a network is expected to be linear in @xmath1 , but it is very hard to test numerically this prediction .
the annealed approximation makes also predictions about the critical line of the model @xcite .
it predicts that the properties of the attractors are universal on the critical line @xmath25 ( with the exceptions of the points @xmath26 and @xmath27 , which are not transition points ) .
in particular , the typical length of the cycles should increase as @xmath0 all along the critical line .
numerical results are not clear under this respect @xcite : it seems that the rescaled cycle length @xmath28 has a limit distribution if @xmath29 is small ( roughly , smaller than 2 ) but for larger values the distribution becomes broader and broader as @xmath1 increases @xcite , so that it is possible to define an effective length scale increasing much faster with system size ( as a stretched exponential ) .
the distribution of the number of cycles has exactly the same characteristics .
these results cast doubts on the validity of the biological analogy proposed by kauffman , that relies very much on the fact that in critical networks the typical number of cycles scales as @xmath0 , reminiscent of the fact that the number of cell types of multicellular organisms very far apart in the filogenetic tree scales as the square root of the number of the genes , and that in critical networks the typical length of the cycles increases as a power law of system size , also consistently with the behavior of cell cycles time .
thus it is interesting to understand how these distributions look like in the limit of very large systems .
another reason of interest of the present approach is that it allows to understand the limits of the annealed approximation . in our interpretation
the annealed approximation is valid as far as the system loses memory of the details of its evolution .
this , of course , does not happen if in a realization of a random network some structural properties that are able to influence its asymptotic dynamics emerge .
thus the approach presented here is complementary to the one used in @xcite .
let us start recalling the definition of the stable elements @xcite .
these are elements that evolve to a constant state , independent of the initial configuration . flyvbjerg defined them and computed their fraction @xmath30 using the annealed approximation , which becomes exact in the infinite size limit .
we now recall briefly , for future convenience , the main steps of this calculation .
let us suppose that an element is controlled by @xmath31 stable elements and @xmath13 stable ones
. then it will be stable if the control function does not depend on the unstable arguments when the stable arguments assume their fixed values .
otherwise it will be unstable .
when all the @xmath13 unstable elements are different ( this can always be taken to be the case if @xmath8 is finite and @xmath1 grows ) , the probability @xmath32 to choose a constant control function of @xmath13 binary variables is given by @xmath33 , with @xmath34 . in the framework of the annealed approximation ,
extracting at random connections and response functions at each time step , we get the following equation for the fraction of variables that are stable at time @xmath10 : [ stable ] s(t+1)=(s(t))= _ i=0^k ki s(t)^k - i(1-s(t))^i p_i .
this equation can be shown to be exact in the infinite size limit .
the fixed point of this map ( which can be interpreted as a self - consistency equation for the fraction of stable variables ) has only the trivial solution @xmath35 in the frozen phase , in other words all the elements are stable except eventually a number increasing less than linearly with @xmath1 . in the chaotic phase
this solution becomes unstable and another solution less than 1 appears .
this happens when @xmath36 .
since @xmath37 ( it is just the probability that the response to two different signals are different ) this condition is equivalent to the condition obtained from the study of the hamming distance .
the existence of the stable variables is due to the finite connectivity of the network ( @xmath38 goes to zero very fast when @xmath8 increases ) .
these variables do not take part in the asymptotic dynamics . among the remaining unstable variables ,
some are irrelevant for the dynamics , either because they do not send signals to any other variable , or because they send signals , but the response functions are independent of this signal when the stable variables have attained their fixed values .
the remaining variables , that are unstable and control some unstable variable , are what we call the relevant variables .
they are the only ones that can influence the long time behavior of the system . to be more clear we now describe the algorithm that we used to identify the relevant variables . as a first step , we have to identify the stable variables .
these are the variables that assume the same constant state in every limit cycle , and identifying them is computationally very hard , but very simple in principle .
we then eliminate from the system the stable variables , reducing the response functions to functions of the unstable variables alone . some of the connections left are still irrelevant , and we have to eliminate them ( a connection between the elements @xmath13 and @xmath39 is irrelevant if the reduced response function @xmath40 does not depend on the argument @xmath41 for all the configurations of the remaining @xmath42 control variables ) . at this point
we iterate a procedure to eliminate the irrelevant variables . at each iteration
we eliminate the variables that do not send any signal to anyone of the variables that are left , until we remain with a set that can not be further reduced .
this is the set of the relevant variables .
measuring the number of relevant variables is computationally a very hard task . in order to identify the stable variables , in fact
, we should find all the cycles in the network , and , to be rigorous , we should simulate a number of trajectories of the same order of the number of configurations in the system .
of course this is not feasible and we run only 200 ( in some case 300 ) randomly chosen trajectories in every network .
thus we overestimate the number of stable elements .
nevertheless , the number of stable elements changes very little when we simulate more initial conditions and we think that the error that we make is not very large .
however , for every network we simulate some hundreds of trajectories and every trajectories has to be followed until the closing time .
this grows exponentially with system size in the chaotic phase . on the critical line
the typical closing time increase roughly as a power law of system size , but the distribution becomes broader and broader and the average closing time is more and more dominated by rare events .
the average depends thus on the number of samples generated and on the cutoff of the closing time , _
i.e. _ the maximum time that we are disposed to wait to look for a cycle . to reduce the bias determined by the cutoff
, we had to run simulations lasting a time which increases roughly as a stretched exponential of system size on the critical line .
thus it is not possible to simulate systems of more than about one hundred elements in the chaotic phase and one thousands of elements on the critical line .
the mean field analysis @xcite shows that the fraction of relevant variables vanishes in the frozen phase and on the critical line , but does not tell how the number of relevant variables scales with @xmath1 as @xmath1 grows .
in order to clarify this point , we have to go beyond the mean field picture . in the special case of @xmath4 , belonging to the frozen phase for every @xmath43
, there are detailed analytical results about the distribution of the relevant variables @xcite .
we propose here a rough argument that generalizes those results to the whole frozen phase and predicts that the typical number of relevant elements scales as @xmath0 on the critical line . though this argument is based on some approximations which we can not control , its results coincide for @xmath4 with the exact results by flyvbjerg and kjaer . let us suppose that we add a new element to a system with @xmath1 elements , @xmath44 of which are relevant , while @xmath45 are stable and @xmath46 are indifferent , _ i.e. _ neither stable nor relevant .
the probability that the new element is relevant can be computed as a function of @xmath44 and @xmath45 , within some approximations that we are going to discuss in a while .
this probability is equal to the fraction of relevant elements in the system with @xmath47 elements , given that the relevant elements are @xmath44 and the stable ones are @xmath45 in the system with @xmath1 elements .
we can then average over @xmath44 and @xmath45 in order to get an equation connecting @xmath48 to the moments of the distribution of @xmath44 in the system with @xmath1 elements . since in the frozen phase and on the critical line
@xmath49 vanishes , it will be enough to consider the first two moments of the distribution , and the resulting equation can be solved asymptotically in @xmath1 .
the weakness of this approach lies on the assumptions that allow us to express the probability that the new element is relevant as a function of @xmath44 and @xmath45 , as it will become soon clear .
we compute now this probability . to this aim
, we need two steps : 1 . as a first step
, we have to extract the @xmath8 control elements and the response function of the new element . as a consequence ,
the new element can be stable , unstable or , if it receives an input from itself and this input is relevant in the sense discussed above , relevant .
the evaluation of the stability is perfectly equivalent to the mean field argument , but this stability is only temporary because it can be altered by the second step described below .
thus we call a new element that is stable ( unstable ) after the first step a _ temporarily _ stable ( unstable ) element .
then we have to send to the old system the signal of the new element . for each of the @xmath50 old control connections we have a probability @xmath51 that the connection is broken and the old control element is substituted by the new element .
this step perturbs the elements that control the new element and modifies its temporary stability .
we have no chance to take this into account , unless we use some drastic approximations . in the second step
, three situations can occur . 1 .
if the new element was relevant in the first step , the new step can not modify this condition .
2 . if the new element was unstable , it can not become stable through the feedback of its signal .
so it will be relevant or indifferent , depending on whether it sends an input to at least one relevant element or not .
3 . if the new element was stable , its signal can destabilize some of the elements that control it and thus it can become relevant through a feedback mechanism , very hard to investigate analytically . to compute the probability of case 3 , we should know the organization of the network in the very detail and not only the number of relevant and stable elements .
we propose to bypass this difficulty considering a different event : we will consider the new element relevant if it receives a signal from a previously relevant element or from itself .
this is the simplest way to get a closed equation for the average number of relevant elements . in this way we make two errors of opposite sign
: on one hand we overestimate the probability that a temporarily unstable element becomes relevant , on the other one we underestimate the probability that the new element is temporarily unstable and we neglect the probability that a temporarily stable element becomes relevant through a feedback loop . we think that this method captures at least the qualitative behavior of the number of relevant elements .
we have then to compare the estimate given by this approximation to the simulations , because the approximation is not under control .
we present this argument because its results agree with both the numerical results and with the analytical calculations for @xmath4 and because we believe that it is possible to improve this method and to keep the approximation under control .
since we are interested in the frozen phase , where the fraction of unstable elements vanishes in the infinite size limit , we can neglect the eventuality that the new element is controlled by more than two elements that were relevant in the old system .
the results are consistent with this assumption . with these approximations
we obtain the following equation for the probability that the new element is relevant : r_n+1=_n=0^n \{r_n = n } [k(n+1n+1 ) (1-n+1n+1)^k-1.[rilev ] + + ._2k2(n+1n+1)^2 (1-n+1n+1)^k-2 ] , where @xmath52 represents the probability that a boolean function of two arguments is not constant and in terms of @xmath20 is given by _ 2=1-p^4-(1-p)^4=(2 - 2 ) . in the frozen phase it is sufficient to consider that the new element receives only one signal from the previously relevant elements .
so , posing @xmath53 , the equation for the new fraction of relevant elements , @xmath54 , is r_n+1cr_n+cn .
the first term represents a new element that receives a relevant signal from one of the previously relevant elements , the second term represents a new element that receives its own relevant signal .
thus the average number of relevant elements is independent on @xmath1 and its asymptotic value is r_n = c1-c.[froz ] this number diverges on the critical line @xmath55 . in this case
, we have to consider also the eventuality that the new element receives a signal from two of the previously relevant elements .
expanding to the second order in @xmath56 , and using the fact that @xmath57 , we get the equation r_n+1r_n - (k-14k ) r^2_n+1n , whence , in the asymptotic regime where the variations of @xmath58 are of order @xmath59 , we finally get r^2_n(k-14k)1n .
this means that the scale of the number of relevant elements grows , on the chaotic phase , as @xmath0 .
we stress here that these computations are valid because of the finite connectivity of the system .
if we perform the limit @xmath19 on the above result , we get that the scale of the number of relevant elements grow as @xmath60 .
if , instead , we apply the limit @xmath19 prior to the limit @xmath61 we get the trivial critical point @xmath62 , where all the elements are stable after one time step , while for every other @xmath20 value all the elements are relevant .
thus , the two limits do not commute .
in fact , the equation ( [ stable ] ) for the fraction of stable variables and all the computations performed in this section are valid only if we can neglect that the same element is chosen more than once to control a given element , _
i.e. _ for @xmath63 . the result ( [ froz ] ) coincides for @xmath4 with the analytical computation by flyvbjerg and kjaer @xcite , thus suggesting that the distribution of relevant elements is independent on @xmath1 in the whole frozen phase , and depends on the two parameters @xmath8 and @xmath20 only through their product .
this picture agrees with the results of the annealed approximation , which predicts that the distribution of the number of different elements in two asymptotic configurations is independent on @xmath1 and depends only on the product of the parameters @xmath8 and @xmath20 in the frozen phase @xcite .
our simulations confirm that on the critical line the number of relevant elements scales as @xmath0 ( see figure [ fig_ril4 ] ) .
also the annealed approximation is consistent with this result , since it predicts that the number of elements whose state is different in two asymptotic configurations has to be rescaled with @xmath0 on the critical line @xcite . on the other hand
the number of unstable elements grows much faster with @xmath1 ( numerically it is found that it goes as @xmath64 , see below ) but this discrepancy is only apparent , since the asymptotic hamming distance is related more to the number of relevant elements than to this quantity .
for later convenience ( see our next paper ) it is also interesting to compute the effective connectivity , defined as the average value of the relevant connections between relevant elements .
let us compute it by imposing the condition that the network has @xmath44 relevant elements .
the effective connectivity is equal to the average number of connections between the new element and the other relevant elements of the older system , with the condition that the new element is relevant . from equation ( [ rilev ] ) we have , at the leading order in @xmath65 : [ ceff ] k_eff(r)= & & c(1-)^k-1 + 2_2k2()^2(1-)^k-2c(1-)^k-1+_2k2()^2(1-)^k-2 + & & 1+a(k , ) .
this equation shows that the effective connectivity minus 1 goes to zero as @xmath65 in the frozen phase ( where @xmath66 ) and on the critical line ( where @xmath67 ) . for a fixed system size ,
the effective connectivity increases linearly with the number of relevant elements .
the distribution of this variable , shown in figure [ fig_magneti ] for @xmath24 and @xmath70 , has many peaks , corresponding to simple rational values .
this perhaps reflects the fact that the relevant elements are divided into asymptotically independent modules , so that a cycle can be decomposed into several independent shorter cycles .
this subject will be further discussed in our second paper .
our results have to be compared to the analytical work by derrida and flyvbjerg @xcite .
they defined the magnetization of element @xmath13 at time @xmath10 on a given network as the activity of the element at time @xmath10 averaged over many initial configurations and could compute analytically its stationary distribution , in the limit @xmath61 , using the annealed approximation , that can be shown to be exact for this purpose .
the picture they got is different from ours , in particular we see peaks much higher than theirs .
for instance the peak at @xmath71 , which gives information on the size of the stable core of the network , is about 10 times larger then expected , and the first moments of the magnetization , that can be computed analytically , are larger than the predicted values .
thus we performed other simulations that strongly suggest that these discrepancies are finite size effects , and we present an argument that explains their origin . in order to investigate larger systems we had to change the definition of the magnetization .
the definition ( [ mag ] ) is numerically cumbersome , since the measure takes place only after that a cycle has been found , and this means , for chaotic systems , that we have to wait a time exponentially increasing with @xmath1 .
thus we neglected this condition and we measured the magnetization of the variable @xmath13 at time @xmath10 as the average activity of the variable with respect to different initial conditions ( this definition coincides with the one used by derrida and flyvbjerg ) .
for very large @xmath10 , when all trajectories have reached a limit cycle , this quantity tends to the asymptotic value m_i=_w_m_i^,[magna]where @xmath72 is the weight of the basin of cycle @xmath69 and @xmath68 is defined in ( [ mag ] ) .
we observed that @xmath73 reaches a stationary value ( within some precision ) much earlier than the typical time at which the trajectories reach their limit cycles . at first sight
surprisingly , the time after which @xmath73 reaches its stationary distribution does decrease with system size instead of increasing ( see figure [ fig_score ] ) .
we measured the second and fourth moment of the magnetization in a system with @xmath24 and @xmath74 , and we found a large positive correction to the infinite size values computed by derrida and flyvbjerg @xcite .
the values found coincide within the statistical error with those obtained from equation ( [ mag ] ) for a system of small size for which we did an explicit comparison .
these values can be fitted to the sum of the infinite size value , that we got from @xcite , plus an exponentially decreasing term .
the exponent of the best fit turned out to be the same for both the moments that we measured : we found @xmath75 , and @xmath76 .
the measure of the magnetization allows to identify the stable elements as the elements with @xmath77 equal either to 0 or to 1 .
the two definitions of the magnetization gave roughly the same number of stable elements in the cases where we could compare the results , but with the second method we could consider much larger systems ( we recall that the difference between the two methods is that in the first case a cycle has been reached while in the second one the system is still in some transient configuration ) .
the second method was used only to study finite size effects , since it does not allow to identify the relevant elements ( see below ) .
both the methods overestimate the number of stable elements , since it could happen that an element appearing stable in our sample of trajectories ( some hundreds ) oscillates in a cycle that is not reached by any of them .
we checked that the results do not change qualitatively if we consider a larger number of trajectories .
the fraction of stable nodes measured in simulations with @xmath24 and @xmath1 ranging from 50 to 200 have been compared to the prediction of the mean field theory by flyvbjerg .
the networks with @xmath78 have a stable core about 10 times larger than the mean field value ( in this case we measured the magnetization using both the above definitions , while for larger systems only equation ( [ magna ] ) was used ) .
the corrections to the mean field value , that is exact in the infinite size limit , appear to decay exponentially with a rate identical , within statistical errors , to the decay rate of the corrections to the moments of the magnetization : we found @xmath79 .
for every size of the systems which we simulated the stable core is then much larger than it would be in an infinite system . on this ground
, we may expect very important finite size effects concerning the dynamical properties of the system .
summarizing , the distribution of the magnetization for finite systems has the following characteristic : 1 ) the asymptotic value is reached after a time that _ decreases _ with system size ; 2 ) the corrections to the infinite size values are very large ; and 3 ) these corrections decrease exponentially with system size .
these apparently strange finite size effects have a simple interpretation : they arise as a consequence of the periodic dynamic of the random networks .
the mean field values of the magnetization and of the stable core are computed within the annealed approximation without taking into account the fact that the asymptotic dynamic is periodic . as we proposed in @xcite , the existence of limit cycles
must be taken into account in the framework of the annealed approximation in this way : if at time @xmath10 all the configurations generated are different ( _ i.e. _ the trajectory is still open ) we treat the quantities of interest ( distance , magnetization or stable core ) as a markovian stochastic process ; if one configuration has been found twice ( the trajectory is closed ) we impose the condition that all quantities are periodic .
thus the master equation for the distribution for the number of stable variables is , in the framework of the annealed approximation : & & \ { s(t+1)=s,o(t+1)s(t)=s , o(t)}= n s((s))^s(1-(s))^n - s(1-_n(s , t ) ) + & & \ { s(t+1)=s,s(t)=s , o(t)}= ns((s))^s(1-(s))^n - s_n(s , t ) + & & \ { s(t+1)=s,s(t)=s , } = _ ss , [ stable2 ] where @xmath80 is the number of stable elements , @xmath30 , @xmath81 stands for the condition that the trajectory is open at time @xmath10 ( no configuration has been visited twice ) , @xmath82 stands for the condition complementary to @xmath81 ( the trajectory has closed on a previously visited configuration ) and @xmath83 is the probability that a trajectory open at time @xmath10 and with @xmath45 stable elements at time @xmath9 closes at that time .
finally , @xmath84 is given by equation ( [ stable ] ) .
we do nt know how to compute @xmath83 , but it is clear that this is an increasing function of @xmath45 for fixed @xmath10 : the more elements are stable , the more it is likely that the trajectory closes .
the infinite size value of the stable core is given by equation ( [ stable ] ) , that represents the evolution of the most probable number of stable variables .
it is clear that the corrections to this value are positive , and that they go to zero as soon as the closing time becomes much larger than the time necessary for the stable core to reach its stationary value in an infinite system ( where all trajectories are still open ) .
thus we expect that these corrections vanish as a power law of the typical length of the cycles : in the chaotic phase this means that the finite size corrections due to this effect vanish exponentially with system size , as we observed simulating systems with @xmath24 and @xmath74 .
lastly , this argument implies that the time after which the distribution of the stable elements becomes stationary is shorter in an infinite system than in a small system , where the evolution of @xmath80 is coupled to the closure of the periodic orbits .
thus the correction of the annealed approximation to take into account the existence of periodic attractors can account for all the features of the finite size effects that we observed . after having identified the stable elements we detect the relevant elements using the algorithm described in the second section and we study how this quantity influences the dynamical properties of the network .
the main results are that the average cycle length grows almost exponentially with the number of relevant variables in some range of this variable and the average weight of the attraction basins has apparently a non monotonic behavior versus the number of relevant variables .
this qualitative features are the same both in the chaotic phase and on the critical line , but the ranges of @xmath44 in which these things happen are quite different in the two cases .
we start discussing the situation in the chaotic phase .
the simulations were done generating at random @xmath85 sample networks and running 200 trajectories on each of them .
the parameters considered in this section are @xmath24 , @xmath74 and system size @xmath1 ranging from 30 to 60 elements .
figure [ fig_ril ] shows the density of the distribution of the fraction @xmath49 of relevant variables , @xmath86 . the density relative to the most probable value increases with system size , and it appears that @xmath49 tends to be delta - distributed in the infinite size limit , as it is expected on the ground of the annealed master equation ( [ stable2 ] ) .
we observe an excess of networks with very few relevant elements ( _ i.e. _ very many stable elements ) , consistently with the finite size effects discussed in last section .
this excess seems to disappear in the infinite size limit .
then we show the average length of the cycles in networks with @xmath44 relevant elements ( figure [ perril ] ) .
this quantity increases almost exponentially with @xmath44 when @xmath56 is large , while its behavior is different for small @xmath54 .
the crossover takes place at about @xmath87 .
thus the number of relevant elements turns out to have a very important influence on the typical length of the cycles we have also measured the conditional distribution of the length of the cycles in networks with @xmath44 relevant elements .
when @xmath44 is close to @xmath1 the distribution decays as a stretched exponential with an exponent smaller than one , very close to the one found in the unconditioned distribution .
thus the deviation of the unconditioned distribution from the prediction of the annealed approximation , that predicts a much narrower distribution , is not a consequence of the existence of the relevant elements .
y_2=_w_^2 , where @xmath72 is the attraction basin of cycle @xmath89 .
we used the method proposed by derrida and flyvbjerg @xcite , that is based on the fact that @xmath88 is equal to the probability that two trajectories chosen at random end up on the same attractor . from our data
( not shown ) it appears that @xmath88 has a non monotonic behavior as a function of @xmath54 : for very small @xmath54 it decreases from the value 1 , corresponding to @xmath90 , reaches a minimum and rapidly increases . at large @xmath54 , @xmath88 does not seem to be correlated with @xmath54 ( at least within the statistical error , that is rather large ) .
we will see in the next paper that the decreasing behavior at small @xmath54 can be interpreted as an effect of the modular organization of kauffman networks .
we simulated systems with @xmath91 and the critical value @xmath92 .
systems size ranges from 120 to 1600 . concerning the statistical properties of the attractors , these networks have a behavior very similar to that of the more studied @xmath93 , @xmath74 networks @xcite . in these networks
, the number of relevant elements appears to scale as @xmath0 , in agreement with the argument presented in section 4 .
the number of unstable elements , on the other hand , appears to scale as @xmath64 .
this implies that the probability to extract at random an element which is relevant , scaling as @xmath94 , is approximately proportional to the square of the probability to extract at random an element which is relevant ( @xmath95 ) .
these scaling laws can be observed both looking at the average quantities and looking at the whole distribution .
the average number of unstable variables is found to follow the power law @xmath96 , with @xmath97 .
we then define the rescaled variable @xmath98 , and we compare its probability density for various system sizes . as it can be seen in figure [ fig_unst ] the different curves superimpose within the statistical errors .
this suggests that @xmath99 has a well defined probability density in the infinite size limit , although our data are rather noisy to state this point without doubts .
we can distinguish in the distribution three different ranges with different characteristics : at vanishingly small values of @xmath99 ( ranging from @xmath100 up to @xmath101 ) the density decreases very fast . at intermediate values , roughly up to @xmath102 , it looks to decrease approximately as a power law with a small exponent ( the best fit exponents that we found range from 0.25 to 0.40 , showing some tendency to increase with system size ) .
asymptotically , for large @xmath99 , the best fit is a stretched exponential , @xmath103 , with an exponent compatible with @xmath104 for all the systems that with studied with @xmath1 larger than 240 .
the number of relevant variables was studied in a similar way .
its average value increases as a power law of @xmath1 , @xmath105 , with @xmath106 .
the rescaled variable @xmath107 looks to have a well defined distribution in the infinite size limit , as it is shown in figure [ fig_ril4 ] , where the probability density of @xmath108 is plotted for system sizes ranging from 120 to 1600 . for large @xmath109 the density of the distribution
is well fitted by a stretched exponential , @xmath110 , with the exponent @xmath111 compatible with the value @xmath112 for system size larger than 240 .
the average length of the cycles increases exponentially as a function of the number of relevant elements , for @xmath54 large , and more slowly for @xmath54 small , just as it happens in the chaotic phase .
figure [ fig_perril4 ] shows on a logarithmic scale the behavior of the average length of the cycles as a function of the rescaled number of relevant elements , @xmath107 , for different system sizes at the critical point @xmath91 , @xmath92 .
the average weight of the attraction basins , @xmath88 , has a non monotonic behavior as a function of the number of relevant elements , as it happens in the chaotic phase .
the value of @xmath113 is one for @xmath114 , then decreases to a minimum value and increases very slowly , as it is shown in figure [ fig_yril4 ] , where @xmath115 is plotted against @xmath107 , for @xmath91 , @xmath92 and different system sizes .
nevertheless , there are two important differences with respect to the chaotic phase : first , the range where @xmath113 is a decreasing function is much wider on the critical line than in the chaotic phase ; then , on the critical line the curves corresponding to a smaller @xmath1 value are lower , while in the chaotic phase the contrary holds . as a consequence ,
if we average @xmath113 over @xmath44 on the critical line , we get a quantity vanishing in the infinite size limit @xcite , while the average weight of the attraction basins is finite and very close to the random map value in the chaotic phase @xcite .
this difference and the non monotonic @xmath44 behavior of @xmath113 have a clear interpretation in the framework of the modular organization of kauffman networks @xcite .
we thus postpone to that paper the discussion of our results .
kauffman ( 1969 ) , metabolic stability and epigenesis in random constructed genetic nets , _ j. theor .
_ * 22 * , 437 - 467 .
kauffman ( 1993 ) , _ origins of order : self - organization and selection in evolution _ , oxford university press derrida b. and y. pomeau ( 1986 ) , random networks of automata : a simple annealed approximation , _ biophys . lett . _ * 1*(*2 * ) , 45 - 49 b. derrida and g. weisbuch ( 1986 ) , _ j. physique _ * 47 * 1297 .
hilorst and m. nijmajer ( 1987 ) , _ j. phys .
paris , * 48 * _ , 185 .
b. derrida and h. flyvbjerg ( 1986 ) , multivalley structure in kauffman s model : analogy with spin glasses , _ j.phys.a : math.gen._ * 19 * , l1003-l1008 h. flyvbjerg ( 1988 ) , an order parameter for networks of automata , _
a : math.gen . _ * 21 * l955-l960 h. flyvbjerg , n.j . kjaer ( 1988 ) , exact solution of kauffman model with connectivity one , _ j.phys . a : math.gen . _ * 21*(*7 * ) , 1695 - 1718 m. mezard , g. parisi , m.a .
virasoro ( 1987 ) , _ spin glass theory and beyond _ , singapore : world scientific u. bastolla and g. parisi ( 1997 ) , the modular structure of kauffman networks , _ physica d _ u. bastolla and g. parisi ( 1996 ) , closing probabilities in the kauffman model : an annealed computation , _ physica d _ 98 , 1 .
u. bastolla and g. parisi ( 1996 ) , the critical line of kauffman networks , _ journal of theoretical biology _ , in press .
b. derrida , h. flyvbjerg ( 1986 ) , the random map model : a disordered model with deterministic dynamics , _ journal de physique _ * 48 * , 971 - 978 a. bhattacharijya and s. liang ( 1996 ) , power - law distribution in the kauffman net , _ physica d _ b. derrida and h. flyvbjerg ( 1987 ) , distribution of local magnetizations in random networks of automata , _ j.phys.a : math.gen._ * 20 * , l1107-l1112 . | this is the first of two papers about the structure of kauffman networks . in this paper
we define the relevant elements of random networks of automata , following previous work by flyvbjerg @xcite and flyvbjerg and kjaer @xcite , and we study numerically their probability distribution in the chaotic phase and on the critical line of the model .
a simple approximate argument predicts that their number scales as @xmath0 on the critical line , while it is linear with @xmath1 in the chaotic phase and independent on system size in the frozen phase .
this argument is confirmed by numerical results .
the study of the relevant elements gives useful information about the properties of the attractors in critical networks , where the pictures coming from either approximate computation methods or from simulations are not very clear .
48 pt -1.5 truecm -1.2 truecm 16.5 cm 23 cm @xmath2dipartimento di fisica , universit `` la sapienza '' , p.le aldo moro 2 , i-00185 roma italy @xmath3hlrz , forschungszentrum jlich , d-52425 jlich germany keywords : disordered systems , genetic regulatory networks , random boolean networks , cellular automata |
one of the most distinctive features of the supergiant cd - type galaxies found at the centers of rich environments is the very obvious presence of huge populations of old - halo globular clusters ( gcs ) , often numbering above 10,000 gcs in a single system and spanning radii upwards of 100 kpc ( e.g. * ? ? ? * ) .
these `` high specific frequency '' ( `` high-@xmath9 '' ) globular cluster systems ( gcss ) provide a platform for statistical studies of halo cluster properties that can not be carried out in any other environment . despite their rarity
, they are therefore prime targets for both observations and interpretive modelling on the formation and evolution of gcs and their host galaxies .
the gcss in cd galaxies ( most of which are also `` brightest cluster galaxies '' or bcgs ) have been used to identify a new correlation between gc luminosity ( or mass ) and color ( or metallicity ) . while the cluster metallicity distribution function ( mdf ) usually has a bimodal shape ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* among many others ) , recent photometric work suggests that the more metal - poor gcs ( those in the `` blue sequence '' ) have higher heavy - element abundance at progressively higher luminosity . this _ mass - metallicity relation _
( mmr ) was first discovered in eight giant bcgs by ( * ? ? ?
* hereafter h06 ) and in the virgo giants m87 and ngc 4649 @xcite .
the same trend can be seen for the gcs around the giant sa galaxy ngc 4594 @xcite , and @xcite find evidence for a similar mmr in their extensive sample of virgo cluster survey galaxies by combining their gc data into groups by host galaxy luminosity . as h06 discuss ,
a clear trace of this trend had already been found for the fornax cd ngc 1399 by @xcite and @xcite , who noted that the bright end of the gc population appeared broadly unimodal rather than bimodal . as we now realize , this effect comes about essentially because the blue and red sequences merge at the top end .
intriguingly , the more metal - rich `` red '' sequence is not seen to exhibit a mmr of its own in any galaxy so far .
that is , for the more metal - rich gcs , metallicity is independent of mass .
the basic interpretation proposed by h06 is that the blue - sequence clusters formed primarily in dwarf - sized pregalactic clouds while the first stage of hierarchical merging was beginning . in these sites ,
more self - enrichment could occur within the more massive dwarfs where more of the enriched gas from supernova ejecta could be held back .
these same higher - mass dwarfs would also , on average , produce the most massive gcs @xcite . by contrast , the more metal - rich gcs on the red sequence formed later , perhaps in major mergers , and at a stage when all the material could be held in the much deeper potential well of the final big galaxy . in that stage
, enrichment would have been more independent of gc mass .
@xcite discuss a similar view .
@xcite use semi - analytic galaxy formation models to test a different approach , namely the contribution of stripped dwarf nuclei to the mmr .
they find that a mmr along the blue sequence appears naturally with a standard model prescription for star formation and enrichment rates in semi - analytic models
. an integral part of their interpretation is that many of the massive blue gcs are actually the dense nuclei of stripped dwarfs ( de , n or ucd objects ) .
still more intriguingly , at least one clear anomaly is already known to exist : although the mmr along the blue sequence has now been found unequivocally in many galaxies , it does not seem to occur in the well studied virgo giant ngc 4472 , where both the red and blue gc sequences run vertically .
this result has been confirmed in several studies and in a variety of photometric bandpasses @xcite .
@xcite find that increasing the efficiency of supernova feedback for heating in dwarfs can produce this effect , leaving the blue gc sequence without a mmr .
but it is not yet clear what underlying physical or environmental conditions would cause this to occur . in short , the existence of the mmr is beginning to provide rich new insights into the early formation era of large galaxies .
ngc 3311 , the central cd galaxy in the hydra cluster a1060 , has long been an attractive target for gcs studies because it is relatively nearby ( @xmath10 mpc for a hydra redshift @xmath11 km s@xmath12 and @xmath13 km s@xmath12 mpc@xmath12 ) and has an enormously populous gcs comparable with the largest known , such as m87 in virgo or ngc 4874 in coma . its ranking among the high-@xmath9 systems was established with the early photometric survey by @xcite .
further ground - based imaging in the washington system @xcite indicated that its gcs was moderately metal - rich with a hint of bimodality , and that its spatial distribution was similar to the shape of the cd envelope light .
hubble space telescope imaging with the wfpc2 camera @xcite in the frequently used @xmath14 color index was employed to argue that the mdf for its clusters was more like the `` standard '' pattern in giant ellipticals that is , a probable bimodal structure with metal - poor and metal - rich sequences at their usual mean metallicities .
the most recently published study @xcite combined the _
wfpc2 _ @xmath15 data with _
nicmos _ imaging in f160w ( approximately the @xmath16band ) to gain a first rough estimate of the metallicity and age distributions through models in the @xmath17 two - color diagram .
none of these earlier studies have given a satisfactory picture of the system .
the hst - based studies @xcite are hampered by the small fields of view of wfpc2 and nicmos , yielding very incomplete spatial coverage of this big cd galaxy , and thus relatively small gc samples to work with .
by contrast , the initial study of @xcite in the pre - ccd photographic era covered a wide field but penetrated to only quite shallow photometric limits by today s standards .
relatively shallow limits and photometric calibration problems also affected the first ccd - based imaging @xcite . to study this
unusually attractive gcs further , we need to draw from a new photometric study that combines depth with wide field .
as part of a new program ( gemini gs-2006a - q-24 ) intended to obtain deep multicolor imaging of globular cluster systems around cd galaxies , we obtained ( @xmath18 ) images of ngc 3311 and the central hydra cluster field using the gmos imager on gemini south , which has a @xmath19 field of view ( fov ) and a scale of 0.146@xmath20 after @xmath21 binning .
the raw observations were taken on the nights of 2006 february 8 and march 23 , under dark skies and photometric conditions , with an average seeing of @xmath22 . in each waveband , 13 individual exposures were taken , which we then reregistered and median - combined to create the final , deep images in ( @xmath23 ) .
the total integration time in each filter was 3900s .
the preprocessing of the raw exposures was completed with the gemini package in iraf . in the @xmath24 frames ,
a small amount of night - sky fringing was present , which we successfully removed with a calibration fringe frame from the gemini archives .
after subtracting the fringe frame and then median combining all 13 @xmath24-band frames , the resulting fringe signal contributes only 0.3 percent to the background fluctuations .
for the photometric calibration , we used exposures of landolt standard stars @xcite taken adjacent to our program - field images during photometric conditions .
we transformed the landolt standard - star indices in @xmath25 to ( @xmath18 ) using the conversion equations of @xcite .
since our standard star exposures covered only a small range in airmass , we used the ngc 3311 images taken within just the same airmass range , and taken during the same night within one hour of the standard - star exposures .
this calibration was suitable only for establishing a mean zero - point value for our ngc 3311 frames , and not for a more complete evaluation of the coefficients in airmass or color .
we therefore regard the calibration as only a preliminary one , to be refined when a larger network of standards can be obtained .
ccccccccc @xmath26 & e2/cd + @xmath27 ( j2000 ) & @xmath28 + @xmath29 ( j2000 ) & @xmath30 + @xmath31 ( helio ) & 3593 km s@xmath12
+ @xmath32 & 0.26 + @xmath33 & @xmath34 + @xmath35 & 0.158 + @xmath36 & 11.04 + @xmath37 & @xmath38 + the final measurement of our combined , deep ( @xmath18 ) images was carried out with the latest implementation of the standalone _ daophot _ code , _ daophot4 _ @xcite .
we used an extensive series of artificial - star tests with the _ addstar _ component of _ daophot _ to evaluate the internal photometric detection completeness as a function of magnitude : stars were added to each the @xmath39 and @xmath24 images in groups of 500 over a wide range of magnitudes , the images were then remeasured in the same way as the original frames , and the fraction @xmath40 of stars recovered in each magnitude bin was then determined .
the results are shown in figure [ completeness ] .
the limits of our data , defined as the magnitudes at which @xmath40 drops to 0.5 , are @xmath41 and @xmath42 .
the trend of @xmath40 with magnitude is well matched by a pritchet interpolation function @xcite , @xmath43.\ ] ] which has two free parameters : the limiting magnitude @xmath44 , and a parameter @xmath27 giving the steepness of the dropoff .
for these images we find @xmath45 and @xmath46 .
as will be seen below , the limiting magnitudes @xmath47 and @xmath48 are deep enough to be comparable with the gc luminosity function turnover point of ngc 3311 . within the bright bulge of ngc 3311 , the photometry depth diminishes due to the increased local sky noise at small radii .
however , for radii larger than @xmath49 from ngc 3311 , which include the vast majority of our data , we find no change in the completeness limits @xmath50 .
the random uncertainties of the photometry , as determined from _ daophot / allstar _ and from the same artificial - star tests , are well approximated by the interpolation curves @xmath51 they show little variation from place to place within the gmos field , indicating that crowding is unimportant , and that background light from ngc 3311 and ngc 3309 has little influence except near their very centers .
on the basis of these error curves , we would expect the gc sequences to start being noticeably ` broadened ' in color , purely because of random uncertainties in the photometry , for levels @xmath52 .
lastly , the artificial - star tests show that any _ systematic _ error ( net bias ) in the photometry is less than 0.02 mag in each filter for @xmath53 . throughout the following discussion
we adopt a distance modulus @xmath54 for ngc 3311 and foreground reddening @xmath55 , as we did in @xcite .
after carrying out the photometry in @xmath39 and @xmath24 with _ daophot _ we used the additional code sextractor @xcite to help reject obvious nonstellar objects from the detection lists ( at @xmath56 seeing and at the distance of the hydra cluster , globular clusters appear indistinguishable from the starlike psf . under hst - type resolution of @xmath57 ,
the profiles of the very biggest gcs can , however , be resolved ; see wehner & harris 2007 ) .
detections from the two wavebands were then correlated , and only those objects found in both were kept for the final analysis .
the final color - magnitude measurements for 8108 starlike objects across the entire gmos field are shown in figure [ cmd ] .
the globular cluster population is readily visible as the dominant concentration of points in the color range @xmath58 .
fainter than @xmath59 the field contamination and photometric scatter increase markedly , but brighter than this , the gc population is well defined .
evidence of the standard `` blue '' and `` red '' sequences can already be seen and will be discussed more quantitatively below through objective bimodal fits .
on the red - sequence side ( approximately in the range @xmath60 ) a distinctive bright extension to high luminosity can be seen that , intriguingly , does not exist for the blue gc sequence .
this bright - end feature extends up into the gc mass range @xmath61 that overlaps with the regime of the `` ultra - compact dwarfs '' ( ucds ) .
we have discussed these objects in our previous paper @xcite . to date , not much other photometry of globular cluster systems in the @xmath62 indices has been published that we can make comparisons with . @xcite present @xmath63 data taken with the gmos camera on gemini north for the virgo giant ngc 4649 .
the bimodal fit they obtain to its gc color distribution has mean values @xmath64 for the blue and red sequences which are @xmath65 mag _ redder _ than our calibrated and dereddened colors for ngc 3311 .
we would expect _ a priori _ , however , that the colors of the two sequences would be quite similar from one galaxy to the next ( e.g. * ? ? ?
* ; * ? ? ?
* ) with only a modest trend due to the parent galaxy luminosity .
for the present , we do not regard this difference as significant , because ( as emphasized above ) we regard our zeropoint calibration as a preliminary one , which still needs a separate series of observations especially to establish the photometric color terms .
notably , forbes et al . also regard their own zeropoint calibration as preliminary .
what is perhaps more important at this stage , and equally worth noting , is that the mean color _ difference _
@xmath66 between the red and blue sequences that we find for ngc 3311 ( see the discussion below ) is identical with the separation btween those two sequences measured for ngc 4649 by @xcite .
in sections 4.1 4.2 we use our gmos data to discuss the spatial distribution of the gcs around both ngc 3311 and 3309 , including the radial and azimuthal components .
a continuing theme in this analysis is the relative contribution of the two galaxies to the gc population we see on our images . in section 4.3 , we discuss the color ( metallicity ) distribution of the gcs by radius around ngc 3311 , and in sections 4.4 4.5 we discuss the gc luminosity function and specific frequencies .
the core of the hydra cluster contains several other large galaxies in addition to the central supergiant that was our primary target . of these others
, the closest in projection and by far the most luminous is the giant elliptical ngc 3309 , just @xmath6 away from the center of ngc 3311 .
its integrated magnitude ( @xmath67 ; see below ) places it 0.9 magnitude fainter than ngc 3311 ( @xmath68 ) , though the true total luminosity of ngc 3311 is estimated to be about 1.3 mag brighter than that of ngc 3309 because of its extended cd envelope .
nevertheless , we might expect that ngc 3309 would contribute significant numbers of gcs of its own to the population of objects in the field . by using starcounts around each of the two galaxies on the sides opposite to the other galaxy , @xcite and
@xcite both argued that ngc 3309 was likely to be contributing no more than a few percent of the gcs in the hydra core field , with ngc 3311 acting as the dominant contributor .
even though ngc 3311 is clearly a high-@xmath9 elliptical like many cds , such claims might mean that ngc 3309 would have unusually low @xmath9 for a giant elliptical in a rich cluster . a better and more quantitative estimate of the relative populations , as well as the spatial structures of the two gcss , is possible from our deeper data .
to quantify our analysis , let us assume that the projected number density of gcs on the sky from each of the two galaxies can be modelled by a circularly symmetric function @xmath69 giving the number of gcs per unit projected area ( we justify the assumption of symmetry below ) .
then , the total number density of counted objects at a point @xmath70 anywhere in the field is given by @xmath71 where @xmath72 is the distance of the given point @xmath70 from the center of ngc 3311 , @xmath73 is its distance from the center of ngc 3309 , and @xmath74 are the two radial functions specifying the gcs around each galaxy .
finally , @xmath75 is the uniform background density of whatever contaminating objects fall within our selected sample .
our goal is to deduce , from a fit to the observed gc distribution in the field , the best - estimate parameters for @xmath76 and @xmath77 _
simultaneously_. to perform the fit , we adopt two possible models for the distribution : \(a ) a standard hubble profile , @xmath78 which is a power law with an added core profile of core radius @xmath79 .
\(b ) a generalized sersic profile , @xmath80\ ] ] where @xmath81 @xcite to guarantee that @xmath82 will be the half - light radius , and for typical e galaxies , the exponent @xmath83 is found observationally to fall in the range @xmath84 .
the case @xmath85 recovers the standard de vaucouleurs law . to carry out the actual numerical fit , we divide up the observed area into @xmath86 boxes , each of area @xmath87
; we use the model parameters to predict how many objects @xmath88 should be found in each box at position @xmath89 ; and then calculate the residuals ( observed @xmath90 predicted ) in each box . a standard goodness of fit based on poisson statistics
can then be formed , @xmath91 the outline of the model is shown in figure [ grid ] for the @xmath92 grid that we used for the final reductions .
an advantage of this procedure is that it uses the information everywhere in the frame , not just the counts within some arbitrary radius centered on ngc 3311 . in practice , we ignored any boxes located at the centers of the two galaxies and at one very bright foreground star , as well as the boxes under the shadow of the camera guide probe ( lower right part of fig . [ grid ] ) and at the vignetted corners of the field . to reduce field contamination from the outset
( see fig .
[ cmd ] ) , we used the sample of 4393 objects within the restricted range @xmath93 , expecting that this sample would be dominated by gcs .
when we fit eq .
[ fiteqn ] to the data , in principle we have a total of six free parameters . for the hubble profile model ,
these six are @xmath94 , and the relative amplitude @xmath95 . for the sersic model , they are @xmath96 and @xmath97 .
once a particular ratio @xmath95 is chosen , the actual values of both @xmath98 and @xmath99 individually are automatically set by the condition that the _ total _ predicted population of objects in the field must precisely equal the observed total .
the same applies to @xmath100 and @xmath101 for the sersic model . in practice ,
the background density @xmath75 is extremely hard to determine by the fitting procedure alone , because the ngc 3311 gcs is so extensive that it dominates the starcounts even near the edges of our gmos field .
an external constraint on @xmath75 is therefore desirable . lacking any remote `` control field '' to rely on
, we use the distribution of stars in the cmd itself ( fig .
[ cmd ] ) to estimate the background .
the background contamination will be due to a combination of foreground milky way stars and faint , small remote galaxies .
we can estimate the number and color distribution of possible field stars using the galactic stellar population synthesis models of @xcite . using a @xmath19 field centered on ngc 3311 ,
the models find only 13 stars , suggesting contamination from galactic foreground stars is small at these latitudes .
using a larger field of view , such as a square degree , results in over 2 thousand stars with a color distribution leaning blueward of the blue peak of the gcs distribution .
our main source of contamination , as indicated by the initial fov search , is most likely distant background galaxies . without a control field
, we can not specifically quantify these background galaxies .
if we assume that their distribution of colors is roughly uniform over the interval @xmath102 , we can use the numbers of objects with colors in the ranges @xmath103 and @xmath104 to predict how many background objects are in our selected midrange of @xmath105 .
for @xmath106 we count 860 objects total that are bluer or redder than the gc color range .
normalizing to the same color interval , we then estimate that 570 of the 4393 objects within @xmath105 ( or about 13% ) are background contamination , equivalent to @xmath107 arcmin@xmath108 . in the numerical fitting procedure we adopted this background throughout and treated only the 5 remaining parameters in each model as free . for the hubble profile model ,
we obtain a best - fit set of parameters @xmath109 and @xmath110 for ngc 3311 ; and @xmath111 and @xmath112 for ngc 3309 .
the ratio @xmath113 is @xmath114 . for this set of parameters
the minimum reduced @xmath115 value was 1.44 .
we experimented with different grid sizes and found that the results were not sensitive within broad limits to the fineness @xmath116 of the chosen grid .
given the simple nature of the model , and other factors not accounted for such as the necessarily patchy sampling across the grid , the internally uncertain adopted level of the background , and possible clumpiness ( clustering ) in the background population , the fit is successful at accomplishing our main goal to evaluate the relative contributions of the two giant galaxies .
a very slightly better overall fit ( minimum @xmath117 ) is obtained with the generalized sersic model . for the ngc 3311 gcs
, we find @xmath118 and @xmath119 , while for ngc 3309 @xmath120 and @xmath121 .
the ratio of the two amplitudes is @xmath122 . for comparison , the scale sizes of the underlying halo _ light _ of each galaxy
are significantly smaller : @xcite find @xmath123 kpc for ngc 3309 and @xmath124 25 kpc for ngc 3311 from an integrated - light study in the gunn @xmath125 filter .
these effective radii are not strictly comparable to our solutions for the gcss , because they assume a classic de vaucouleurs profile ( @xmath85 in the sersic model ) ; but they confirm that the gcs is spatially more extended , as has usually been found for giant galaxies @xcite .
one very evident result of this numerical exercise is that ngc 3309 contributes only a minor part of the gc population in our field ( 2 to 3 percent ; see section 4.5 below ) . as a consequence ,
its gcs shape parameters are quite weakly constrained . assuming that the detected objects in the field belong _
entirely _ to ngc 3311 and ignoring ngc 3309 completely ( that is , setting @xmath126 in the numerical fitting ) yields a minimum @xmath127 negligibly worse than the two - galaxy fit .
we quantify the relative importance of the two galaxies a bit further below .
a rough comparison of the two model fits is shown in figure [ radial ] , where we simply plot the number density of objects in each grid box as a function of distance from the center of ngc 3311 .
the best - fit hubble and sersic models _ for ngc 3311 alone _ are shown overplotted , verifying the first - order description of the entire population in the field as belonging to the central cd . the same grid plot , but this time centered on ngc 3309 , gives a radically different impression , similar to picking just a random point somewhere away from ngc 3311 .
that is , no significant concentration towards ngc 3309 is evident
. a significant improvement in this estimate of the two gcs spatial structures will be possible if a better external measurement of the background can be obtained .
the ngc 3311 system is so extensive , however , that photometry at projected radii @xmath128 ( 150 kpc ) will be necessary for a control - field location .
@xcite also attempted to measure the ellipticity of the ngc 3311 gcs and to compare it with the shape of its cd envelope .
we can , in principle , make a better test with our much larger measured population . in this case
, for our test sample we take all 3544 measured objects within radii @xmath129 of ngc 3311 ( the outer circle is the largest complete one fitting within the gmos frame boundary ) and within the photometric limits @xmath93 .
this sample is subdivided into @xmath130 sectors and plotted against azimuthal angle @xmath131 of the sector , as shown in figure [ azimuth ] .
the two sectors containing the center of ngc 3309 and the guide probe shadow are shown as open symbols in fig .
[ azimuth ] and are ignored .
it is evident that there is no strong trace of ellipticity in the residuals about the mean level of @xmath132 objects per sector after background subtraction .
although there are sector - to - sector differences larger than the internal statistical uncertainties , they can not easily be made to match an approximate double - cosine curve that would be the mark of a modestly elliptical distribution @xcite .
the integrated halo light of ngc 3311 has a small but definite ellipticity of @xmath133 and major axis at @xmath134 @xcite .
if the formation of the gcs is closely associated with the galaxy halo , we should expect no large differences in their intrinsic shapes and orientations .
contrarily , a dramatic difference in the gcs shape or orientation relative to the halo light would be a signal of a more fundamental difference in the gcs formation era or its internal dynamical evolution .
we have carried out a solution for @xmath135 and @xmath136 with our gcs data , using the formalism of @xcite .
they show that for an intrinsically elliptical distribution of points sampled in _
annuli , the observed number density is @xmath137^{-\alpha / 2}\ ] ] where the radial falloff is modelled as a simple power law of exponent @xmath27 . here
, @xmath138 denotes the position angle of the major axis at which @xmath139 reaches its maximum value . for ngc 3311
, we have @xmath140 ( for the hubble profile fit described above , we found @xmath141 , but if we remove the core - radius parameter @xmath79 , we require a flatter overall power law to match the whole profile adequately ) .
we then vary the assumed @xmath135 and @xmath136 till the residuals about this model are minimized .
the best - fit solution we find is for @xmath142 and @xmath143 , shown as the solid curve in fig .
[ azimuth ] .
the corresponding major axis line is thus @xmath144 .
( in fig . [ grid ] , the major axis would run vertically . )
the best - fit solution is weakly determined with a shallow minimum at @xmath145 , and indeed solutions with @xmath146 are scarcely worse than this one .
our deduced ellipticity is lower than for the integrated light , and our solution for @xmath136 falls @xmath147 eastward of the integrated halo light , though its uncertainty is large enough that there is no formal disagreement . in short ,
we appear to be confronted with no difficulties by assuming first - order circular symmetry . for comparison , we show one of the solutions by @xcite as the dashed line in fig .
[ azimuth ] ; this clearly does not match our solution very well .
@xcite performed solutions for two different magnitude ranges ( @xmath148 , @xmath149 ) and for the more slightly more restricted radial range @xmath150 . for @xmath148 , they found @xmath151 , @xmath152 , while for @xmath149 they found @xmath153 , @xmath154 ( this latter one is shown in our fig .
[ azimuth ] ) .
they comment that `` the uncertainty in @xmath136 is at least @xmath155 '' . both of the mclaughlin et al .
solutions differ from ours ; we find a distinctly smaller ellipticity and a position angle between their two solutions ( although , again , it is not clear that any genuine formal disagreement exists ) . to attempt to trace down causes for the evident differences
, we rebinned our data to magnitude ranges corresponding more closely to theirs , and also restricted to exactly the same radial range as they used . for these numerical trials
we also included objects at all colors , as they did .
for our objects brighter than @xmath156 , we find @xmath157 and @xmath158 . for a slightly deeper cut down to @xmath159 , we find @xmath160 and @xmath161 .
interestingly , the best - fit position angles of our two subsamples bracket our best solution from the large , ` best ' dataset ( @xmath106 and color - selected ) , but restricting the sample much more severely to the brightest objects tends to lead to a larger @xmath135 , as found by mclaughlin et al . without being able to offer definitive reasons for the disagreement between these two studies , we suggest that the earlier gcs result was likely to have been compromised by the combination of small - number statistics , and especially by much larger relative field contamination ( since the dominant source of field contamination is faint , small background galaxies which have clumpy space distributions , they are quite able to bias our solutions for the spatial distribution if they contribute a high fraction of the total sample ) . in the present data , we have been able to reject non - gc objects more completely by color and morphology than in the older study .
the most intriguing feature of the azimuthal distribution is the residual difference between the gc system ( @xmath162 , @xmath163 ) and the halo _ light _
( @xmath164 , @xmath165 ) .
unfortunately , we are not yet in a position to place heavy significance on this , simply because the solution for the best - fit position angle is extremely uncertain if the ellipticity is very small .
if the difference is verified with deeper and cleaner samples , it will , perhaps , be a tantalizing hint that the gc accretion or destruction histories differed in second - order ways from the field stars . with the knowledge that the vast majority of gcs in this field belong to ngc 3311
, we can also investigate their color distribution as a function of galactocentric distance .
in many gcss within large ellipticals , a color gradient can be found in the sense that the mean color of the gcs ( averaging together both the red and blue ones ) conventionally decreases with increasing @xmath166 .
however , beginning especially with the key work of @xcite on ngc 4472 , it became clear that ( in geisler et al.s concise wording ) `` most of this gradient appears to be due to the varying radial concentrations of the two populations '' . that is , the metal - poor , blue gc population is less centrally concentrated than the metal - richer population and thus contributes a higher proportion of clusters at larger radius , generating an overall declining metallicity gradient in the gc population as a whole .
notably , each of the two metallicity components shows little or no color ( metallicity ) gradient of its own .
very much the same conclusion had already become clear long ago ( albeit with a much smaller sample of gcs than can be found in giant ellipticals ) for the milky way gcs @xcite . in figure [ radcolor ]
we show the color indices for the objects in the ngc 3311 field ( in order to minimize field contamination , we restrict the sample to the 1360 objects brighter than @xmath167 ) .
a broad range in color is present at all radii , but no strong radial gradient in color is evident for either the red or blue subpopulations , a similar result to other galaxies . in order to determine whether there was a trend in color versus radius , we binned the clusters in radius and performed a fit using rmix ( a fitting code discussed in detail in section 6 ) .
the locations of the red and blue peaks are shown as diamonds in figure [ radcolor ] .
a comparison between the cumulative distributions of the red and blue clusters ( shown in the lower panel of the figure ) indicates that any difference has weak significance , at least within the regions our data are sampling .
unfortunately , the field size we sample reaches a maximum @xmath168 kpc , which is not large enough to test either the true spatial extent of the system or what happens to the color distribution near its outskirts ( recall , for example , that the m87 gcs extends detectably outward to at least @xmath169 kpc ; see @xcite ) .
deep imaging from galaxies of all types now shows that the luminosity distribution of their gcs follows a roughly universal shape that has most often been approximated by a gaussian distribution in number of gcs per unit magnitude ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* among many others ) . for giant galaxies ,
the turnover or peak frequency of the gclf is consistently near @xmath170 @xcite while the intrinsic dispersion is in the range @xmath171 ( see particularly * ? ? ?
* for an extensive analysis of the correlation of the turnover and dispersion with host galaxy luminosity , although it is worth noting that the only cd - type galaxy included in their data is m87 ) .
ideally , a determination of the gclf for our target ngc 3311 would best be done by taking the lf of an adjacent `` control field '' , observed under identical conditions , and subtracting it from the lf of the objects in our @xmath172wide central field . lacking this ,
a second - best alternative is to use the outer parts of our gmos field as a `` background '' .
in fact , much of the outer zone is still populated by gcs , and so by defining it as background we will partially subtract some of the gc population as well as the true background . however , as long as no large changes in the gclf with radius exist , the residual gclf we obtain this way will be systematically correct even if a bit oversubtracted .
an inevitable limitation of using the outer parts of the image field to define the background level is that the background estimate @xmath75 is much more uncertain numerically than if we had a large , remote control field to work with .
we divide our field into two regions : an inner annulus of inner radius @xmath173 , outer radius @xmath72 , and area @xmath174 ; and an outer zone of inner radius @xmath72 and outer boundary set by the edge of the frame , and area @xmath175 .
we set @xmath173 to be the innermost radius at which the photometry is still highly complete to the magnitude limit of interest ( @xmath176 ) .
the only significant free parameter we have to define these two zones is the mid - radius @xmath72 .
the residual gc population within @xmath174 after background subtraction will clearly be @xmath177 where , by hypothesis , the background @xmath75 is constant across the field and subtracts cleanly out after the area normalization .
we would like to choose the midpoint @xmath72 in such a way as to minimize the relative uncertainty in @xmath178 .
if @xmath72 is too large , then @xmath179 will be large , but the area normalization factor @xmath180 will blow up and the background correction becomes extremely uncertain .
but if @xmath72 is too small , then so is @xmath181 , and again the relative uncertainty in the residual population becomes large .
to fix @xmath72 , we adopted the radial profile parameters described above for the gc system ( the best - fit hubble profile plus the background level @xmath107 arcmin@xmath108 ) .
the inner radius @xmath173 is set at @xmath182 .
we then calculated the resulting relative uncertainty @xmath178 as a function of @xmath72 , where we let @xmath72 vary from just slightly larger than @xmath173 out to almost the borders of the field .
the optimum @xmath72 turns out to be roughly halfway from the center of ngc 3311 to the edge of the field , given the geometry of this system , but the relative uncertainty goes through a very shallow minimum near this point and the precise choice is not critical . for the following discussion ,
we use @xmath183 .
the `` background '' region is defined to be everything beyond that radius , though not including the vignetted regions in the outskirts . in each zone
we count all objects within the broad color range @xmath184 , and also correct the raw numbers of objects in each bin for photometric incompleteness @xmath185 with the completeness function parameters derived above .
the results are in figure [ gclf ] , shown in @xmath186mag bins . for @xmath187
, the net completeness @xmath40 falls below 0.5 and the lf becomes considerably more uncertain ( note the rapid growth of the internal error bars ) .
this level is also very near the _ predicted _ magnitude of the gclf turnover point for a normal giant or supergiant elliptical galaxy . for @xmath188 and @xmath189
, we expect @xmath190 or @xmath191 adding the distance modulus @xmath54 . to convert @xmath192 to @xmath24 , we use the mean difference @xmath193 determined directly within our field from 12 of the brightest gcs ( see * ? ? ?
finally adding the foreground extinction @xmath194 , we have @xmath195 .
a gaussian curve with that peak magnitude and with @xmath196 mag is shown superimposed on the observed gclf in fig .
encouragingly , the raw lf of all objects within the inner zone ( shown as the open circles ) has the same peak to within @xmath65 mag even without background subtraction . the gaussian curve shown in fig .
[ gclf ] is not intended to represent a rigorous derivation of the observed turnover point or dispersion ; we use it only to show that the standard values provide an entirely plausible description of the data for ngc 3311 .
that is , the gclf in this cd galaxy gives no indication of being unusual .
our photometry does not reach far enough past the turnover to permit any kind of useful test of more sophisticated numerical models , such as the asymmetric schechter - like function explored in depth by @xcite from the virgo cluster survey .
however , in a broader sense our data clearly show that imaging with ground - based 8m - class telescopes is capable of penetrating deep into the globular cluster populations of galaxies in the 50-mpc distance range and even beyond .
recent interest has followed the possibility that the gclfs may differ between the redder and bluer metallicity groups of clusters . if the underlying
_ mass _ distributions of both types of clusters are similar , then the gclf turnover magnitude should differ between photometric bandpasses by typically @xmath197 mag purely because of differential bolometric correction @xcite
. larger differences between the red and blue gclfs ( or ones in the opposite sense expected from bolometric corrections ) may be indicators of genuine differences in their mass distributions . in an hst - based survey of several nearby ellipticals
, @xcite found a mean difference @xmath198 mag in the sense that the blue gclf peak was brighter , though large and uncertain differences show up from one system to another .
they conclude that it is unclear whether any true underlying mass differences are required to explain this mean difference .
ngc 3311 has large numbers of both red and blue clusters , and using the same numerical technique as described above , we determined their gclfs separately .
`` blue '' , metal - poor clusters are defined to lie within the color range @xmath199 and `` red '' , metal - rich clusters within @xmath200 .
the results are shown in figure [ gclf_color ] , where we display only the residual lfs and not the raw numbers before background subtraction .
the blue population falls close to the same roughly - gaussian curve describing the system as a whole . for the red population ,
the small excess of very luminous clusters @xmath201 is evident , but the most striking difference is in the range @xmath202 , where a noticeable dip is seen just around the nominal turnover point .
we are , however , reluctant to conclude that this effect is real given the uncertainties in the background correction and the determination of residuals . a standard kolmogorov - smirnov two - sample test shows that the two lfs differ over the range @xmath203 at more than 99% confidence , but the strongest statement we feel is justified from the present data
is that the bright - end slope for the red clusters is somewhat flatter than for the blue ones .
a better control field , and perhaps even better selection of gcs versus field contamination , will be needed to carry this discussion any further . finally , we use the radial distributions we have derived for ngc 3311 and ngc 3309 to estimate the total cluster populations around each one , and thus their specific frequencies .
we use the best - fit sersic profile parameters given above , and integrate them outward to large enough radius @xmath204 that the gc population totals converge . for ngc 3309 , the particular choice of @xmath204 is quite unimportant , since its profile is steep and negligible numbers of gcs are added for @xmath205 . however , the much more extended profile for ngc 3311 requires us to integrate outward to @xmath206 or about 140 kpc , beyond which the sersic model profile adds negligible numbers of gcs .
this large radius is , however , quite consistent with the existing surface photometry @xcite which shows that ngc 3311 extends detectably to @xmath207 kpc . with these parameters in mind ,
we find total gc populations brighter than @xmath208 equal to @xmath209 and @xmath210 .
evaluating the actual population totals requires correcting for the magnitude limit and the fraction of gcs fainter than that limit . assuming a gaussian - like gclf
as discussed in the previous section , with a turnover point at @xmath211 and standard deviation @xmath212 , then @xmath208 is @xmath213 mag or @xmath214 brighter than the turnover point .
this part of the gclf includes a fraction @xmath215 of the total population .
we therefore estimate the total gc population for ngc 3311 as @xmath216 , where the main uncertainty is the necessary extrapolation to large radius . for ngc 3309 we find @xmath217 ; here , the relatively large internal error is dominated instead by the large uncertainty in the ratio @xmath122 for the two - function fit described above . in this case
our calculation may be a slight underestimate , not because of what may be present at larger radius , but because of the uncertain _ inward _ extrapolation into the central @xmath218 .
its gcs core radius is both smaller and considerably more uncertain ( by a factor of two ; see above ) , and our profile integration may not have accounted for a significant part of the total .
the hst photometry to be discussed in the next section verifies that more gcs are present in its core .
their specific frequencies can be determined once we know their total luminosities .
@xcite measure integrated magnitudes @xmath68 for ngc 3311 and 11.94 for ngc 3309 ( we have reconstructed these numbers from their table 5 and their adopted distance and reddening ) . with our adopted @xmath219 ,
their luminosities are @xmath220 and @xmath221 .
the ( global ) specific frequency of the ngc 3311 gcs is then @xmath222 , and for ngc 3309 @xmath223 .
it should be noted , however , that these total magnitudes are extrapolations to large radius of the @xmath224 profiles that were found to fit the inner regions , and for ngc 3311 at least a part of the excess cd envelope light would be missed .
vasterberg et al .
estimate that it should be made @xmath225 mag brighter , which would give @xmath226 .
if this adjustment is applied , its specific frequency would be lowered to @xmath227 .
however , it is not clear that the luminosity should be increased without also making a similar adjustment to the total cluster population , since we might have missed some of the gcs by using only the extrapolation from the inner profile . for the present time , and pending a more quantitative measurement of the cd envelope light , we use @xmath228 , while @xmath229 should be considered an extreme lower limit . by contrast , ngc 3309 has a remarkably low deduced cluster population , even with the uncertainties about the core population .
giant e galaxies in rich cluster environments with @xmath230 are rare , but not unprecedented .
for example , @xcite found @xmath231 for one of the coma cluster giants ngc 4881 .
it is tempting to imagine that an object like ngc 3309 might have had much of its original gcs removed by tidal stripping in passages through the cluster core .
regardless of its history , it is clear that today , ngc 3311 completely dominates the gcs population in the core hydra field , in equal measure because of its higher total luminosity and its higher intrinsic @xmath9 .
the wide ranges of @xmath9 that we find among these giant and structurally simple e galaxies are still not fully understood and may require combinations of formation conditions , later dynamical evolution in the potential field of their cluster , and accounting for the presence or absence of hot halo gas ( e.g. * ? ? ?
* ; * ? ? ?
* ) . as a final test of the relative gc populations of the two galaxies
, we ran another solution for eq .
[ fiteqn ] where we imposed the _ assumption _ that the total number of gcs in each galaxy was in proportion to the luminosity of the galaxy ; that is , they have the same @xmath9 .
this requirement converts to @xmath232 .
the best - fit solution gave @xmath233 , worse than our optimum solutions described above ( @xmath234 ) though not dramatically so .
the implication is simply that the parameters for the ngc 3309 system are poorly constrained within broad limits , a statement that is already reflected in the large uncertainty on its deduced best - fit cluster population .
ngc 3311 clearly holds one of the very largest globular cluster systems in the local universe , rivalling that of the classic high@xmath235 prototype m87 which has @xmath236 and @xmath237 @xcite .
the ngc 3311 gcs is spatially very extended , marked by a huge core radius @xmath238 kpc or effective radius @xmath239 kpc also comparable to that of the gcs in m87 @xcite .
to gain additional insight into the relative properties of the gcss around ngc 3309 and 3311 , we employed imaging in @xmath15 from the hst wfpc2 camera that was available in the hst archive .
two wfpc2 fields were taken as part of go program 6554 ( pi brodie ) in cycle 6 , each with total exposures of 3700 sec in @xmath240 and 3800 sec in @xmath241 . in one field ,
the pc1 chip was centered on ngc 3311 , and in the second field , the pc1 chip was centered on ngc 3309 .
photometry of the ngc 3311-centered field was published in @xcite , but to our knowledge , no measurements from the ngc 3309 imaging have been published .
we have carried out photometry on both of these wfpc2 fields with _
daophot _ in the same manner as described above
. a typical globular cluster with a half - light radius of @xmath242 pc will , at the distance of hydra , have an intrinsic fwhm @xmath243 , making it entirely feasible to perform psf - fitting photometry with the wfpc2 resolution of @xmath244 . for
each of the four wfpc2 ccds , we registered and co - added the individual exposures downloaded from the hst archive , and performed the usual sequence of _ daophot _ find / phot / psf / allstar .
detected objects on the @xmath245 and @xmath192 frames that matched locations within 2 pixels , and with allstar goodness - of - fit @xmath246 , were kept for the final data list .
calibrations followed the normal prescriptions in @xcite , based on @xmath247 aperture photometry of the brightest objects .
the data for the four ccds were then combined to yield the results shown in figures [ wfpc2_xy ] and [ wfpc2_2cmd ] . the photometric limit near @xmath248 or @xmath249 is about half a magnitude shallower than for our gmos data .
our psf - based photometry for ngc 3311 has generated a slightly tighter cmd than the work of @xcite ( who used only aperture photometry ) , but no major differences .
though the @xmath14 color index is notoriously insensitive to metallicity ( with an intrinsic difference of only 0.2 mag between the red and blue sequences ) , a clear trace of bimodality can be seen ( also see * ? ? ?
notably , on the red side at @xmath250 , the upward extension of the gc sequence into the high - mass ucd - like range that is more evident in our more comprehensive gmos photometry @xcite can also be seen here . in general ,
both cmds in fig .
[ wfpc2_2cmd ] resemble each other closely . in the @xmath251 plots of fig .
[ wfpc2_xy ] , we see the `` ngc 3309 problem '' once again . in the so - called ngc 3309 field ,
a population of gcs is present on and near the pc1 chip centered on 3309 ; but the presence of the ngc 3311 gcs is extensive everywhere over the field ( note particularly the lower left region of the @xmath251 plot where the increase in gc number density toward the center of ngc 3311 is particularly obvious ) .
even though the exposures were intended for study of the ngc 3309 system , the majority of detected objects in it actually belong to ngc 3311 . from our quantitative double - profile solutions in the previous section , it can easily be shown that for any location further than @xmath252 from the center of ngc 3309 , the gcs of ngc 3309 contributes less than 10% of the gc counts . this transition boundary
beyond which the ngc 3309 system becomes unimportant is indicated by the dashed circle in fig .
[ wfpc2_xy ] .
the cmd for this inner zone is shown in figure [ wfpc2_cmd_3309 ] , where we have included the @xmath253 detected objects within an even smaller radius of @xmath254 . in this diagram , which gives us our best look at the gc population of ngc 3309 alone , the high - luminosity end of the cmd is not as well populated as in the ngc 3311 gcs ( perhaps a simple result of small - number statistics ) , but otherwise there are no striking differences to note . in figure [ wfpc2_cmd ] , we have combined the photometry from both wfpc2 fields , excluding only the points within the @xmath252 zone around ngc 3309 .
this cmd , dominated by ngc 3311 , will be analyzed below simultaneously with the gmos data , for bimodality characteristics of the metallicity distribution .
a primary goal of our study was to investigate the multimodal structure of the gc color and metallicity distribution , and in particular to find out whether or not the blue gc sequence would reveal the same mass - metallicity relation as in other large galaxies .
having measured several other properties of the system , including the fact that the gc population in our target field is dominated by ngc 3311 , we now discuss the color distribution .
the histogram of colors @xmath63 over any selected magnitude range drawn from fig .
[ cmd ] indicates immediately that the color distribution is broad and not well fit , for example , by a single gaussian - type curve ( also found by * ? ? ? * from their wfpc2 photometry ) .
a better description of the system is by conventional `` blue '' and `` red '' sequences whose color distributions partially overlap .
the bright - end extension of the red sequence in particular reaches upward into the ucd luminosity range @xcite .
we therefore experimented with bimodal fits in a variety of ways , to help define the mean colors and dispersions of these two components . to carry out quantitative fits to the color distribution , we employed the statistical code rmix @xcite , a library written with the statistics programming language , r. this package allows the user to fit data with various functional shapes and to set constraints on them at a range of possible levels .
rmix does not restrict the fitting to unimodal or bimodal forms ; the number of components adopted to fit the data is user - defined , as are the functional forms of the curve ( e.g. gaussian , poisson , binomial ) and the starting values for the modes and dispersions of each component ( e.g. the peak and standard deviation , in the gaussian case ) . in addition , the dispersions can be constrained to be equal ( homoscedastic ) or allowed to differ between components ( heteroscedastic ) , as appropriate .
the numerical values of the dispersions can also be set to previously determined numbers , or can be determined by the fit , in any combination between the components .
the kmm software @xcite is a well known code frequently used in the literature for bimodal fits to gcs color distributions ( see , e.g. * ? ? ?
* ; * ? ? ?
* among many such examples ) .
we have experimented with both kmm and rmix in this study .
both give quite similar results for a given dataset , but we find that the rmix package offers more flexibility in its choice of fitting functions as well as better flexibility in managing user - defined constraints on the fitting parameters .
it is also more robust , successfully converging on multimodal fits in small - sample cases where kmm would not run .
we recommend rmix as an attractive and statistically rigorous alternative .
figure [ mdf_4panel ] shows some of our results for four different magnitude subsets of our ( @xmath39,@xmath24 ) data . for each magnitude bin shown , we fit a pair of gaussian curves to the @xmath63 histogram .
reasonable initial guesses for the peaks and dispersions were provided , and the component means and dispersions were determined by the fit itself with the condition that the dispersions of the two gaussians were required to be the same .
the results are summarized in table [ rmixfits ] . here
, the first column gives the magnitude range ; column 2 the deduced proportions of the blue and red components ; columns 3 and 4 the mean colors of the blue and red sequences ; column 5 the standard deviation of each sequence ; and the last column the goodness of fit .
bimodal fits are strongly preferred over unimodal ones for the two brighter bins . in the two faintest bins
the overlap between the red and blue components increases because of the combination of field contamination and larger photometric uncertainty , and the distinction between bimodal and unimodal fits becomes less clear .
cccccc @xmath255 & ( 0.48,0.52 ) & @xmath256 & @xmath257 & @xmath258 & 1.38 + @xmath259 & ( 0.64,0.36 ) & @xmath260 & @xmath261 & @xmath262 & 2.34 + @xmath263 & ( 0.62,0.38 ) & @xmath264 & @xmath265 & @xmath266 & 1.35 + @xmath267 & ( 0.72,0.28 ) & @xmath268 & @xmath269 & @xmath270 & 2.11 + in fig .
[ mdf_4panel ] , one can see that the peak of the bluer component shifts closer to the red component at brighter magnitudes , suggesting the presence of a mass - metallicity trend for at least one of the modes . to explore this further
, we separated the gc population into blue and red halves by taking the intervals @xmath271 ( blue ) and @xmath272 ( red ) , and then calculating the mean color of each mode in half - magnitude bins .
the combined results can be seen in figure [ mmr ] .
linear least - squares fits over the entire gc luminosity range @xmath273 , weighted by @xmath274 where the uncertainty of each mean color is @xmath275 , give @xmath276 for the blue sequence with a correlation coefficient of the fit equal to 0.995 , indicating that a mmr is present at high significance . for the red sequence , @xmath277 , indicating no significant trend with magnitude . in fig .
[ mmr ] we show it simply as an average @xmath278 .
field contamination at magnitude levels @xmath279 is unimportant , so at these relatively bright levels in the color - magnitude diagram the existence of an mmr seems secure .
fainter than this , however , contamination becomes progressively worse and is likely to affect these simple linear fits to uncertain degrees .
thus for the fainter parts of the sequences ( roughly , @xmath280 ) we can not test with any confidence whether or not the blue sequence really does continue to get `` bluer '' as it goes down . in @xcite , we claimed that the blue sequence became approximately vertical ( that is , no mmr ) for luminosity levels @xmath281 .
that claim was based on bimodal fits to the color distributions in finely spaced magnitude bins for the eight bcgs studied there ; however , in most of those galaxies the issue of faint - end field contamination is also a concern just as it is here . for ngc 4594 ,
@xcite suggest that the blue - sequence mmr continues downward roughly linearly to fainter luminosities , though their two faintest blue - sequence bins are not distinguishably different in mean color , and their conclusion of the same mmr at all luminosities is in part forced by their assumption of a linear fit . to quote them directly , `` indeed , whether the faintest bins follow the trends established by the brighter points is indeterminate . '' for our ngc 3311 data , removing the two faintest bins , where we expect the contamination to be most significant , as well as the top bin , which may be affected by the presence of ultra compact dwarfs @xcite , yields a similar fit of @xmath282 . as such , the resulting slope does not appear to be driven by contamination in the faintest magnitude bins .
lastly , @xcite use the combined gcs data for 79 virgo galaxies to test for color trends by combining their target galaxies into four groups by luminosity . in their data ,
field contamination is much less of a concern because individual gcs were carefully selected by a series of morphological and photometric criteria .
for the highest - luminosity galaxies ( m87 , m49 , and m60 ; see their figures 1 and 2 ) it can reasonably be claimed that the blue - sequence mmr continues roughly linearly over the entire luminosity range @xmath283 . for the other groups containing a total of 76 galaxies from giants to dwarfs
, the blue gc sequence clearly shows a color trend at bright levels but could reasonably be claimed to become nearly vertical for @xmath284 , corresponding roughly to @xmath285 . at the present stage in this subject
, we can not yet say on the basis of the empirical evidence whether or not the blue - sequence mmr persists at all luminosity levels .
more complete data allowing for accurate removal of field contamination will be needed to trace it out at the crucial faint levels .
all studies to date agree , however , that a mmr exists for the blue clusters brighter than about @xmath286 , and that this seems to be a nearly universal phenomenon . comparing our mmr for ngc 3311 with those found
previously for other large galaxies requires transformations between photometric systems , since h06 used @xmath287 , whereas @xcite and @xcite used @xmath288 .
ideally , we would like to convert @xmath63 to either one of these other indices , but transformations based directly on observations of standard globular clusters ( the milky way clusters , for example ) do not yet exist . for the purposes of a preliminary comparison , we choose instead to transform the observed blue - sequence slope @xmath289 into a metallicity scale @xmath290[fe / h ] through the use of models .
representative calculations of integrated colors for ssp systems are given by @xcite and @xcite .
for the range with [ m / h ] @xmath291 , these two sets of models agree well with each other and give @xmath292 } = 0.18 \pm 0.03 $ ] for the metallicity regime we are concerned with , i.e. [ m / h ] @xmath293 .
then , converting color index to metallicity , and magnitude to luminosity and hence mass , we find that the mmr for the blue sequence in ngc 3311 corresponds to a heavy - element abundance scaling with cluster mass @xmath294 . as we did in @xcite
, we assume that @xmath295 does not change systematically with luminosity in order to convert @xmath296 to @xmath297 and thus derive the scaling . by comparison
, h06 found a mean scaling @xmath298 for their sample of eight giant galaxies .
@xcite found @xmath299 for their virgo galaxy sample .
our results for ngc 3311 are entirely consistent with the other studies within measurement uncertainties . as an additional check of the @xmath63 analysis
, we perform the same kinds of bimodal gaussian fits to the wfpc2 data in @xmath14 . here
, the smaller size of the @xmath14 sample , and its lower sensitivity to metallicity , prevent us from tracing out mean color trends with magnitude with any confidence , so we use a color histogram defined from only a single magnitude range @xmath300 over which the field contamination and photometric scatter are minimized .
as before , we exclude the region within @xmath252 of ngc 3309 . the histogram of 1730 such objects is shown in figure [ wfpc2_hist ] .
we ran rmix under similar assumptions as described above ( two gaussians , with means , dispersions , and proportions determined by the fit ) .
for the blue , metal - poor component we find @xmath301 , @xmath302 , and proportion of the total population @xmath303 .
for the red , metal - rich component , @xmath304 , @xmath305 , and @xmath306 , with an overall @xmath307 .
the relative proportions of the two components are quite uncertain again , a direct result of the insensitivity of @xmath308 to metallicity and thus the heavy overlap between the two components but agree nominally with the near - equal proportions we found from the @xmath63 solutions ( fig . [ mdf_4panel ] ) . for comparison purposes
, we also tried a fit of a single gaussian curve to the entire data : this yielded a best - fitting mean @xmath309 and dispersion @xmath310 , with goodness - of - fit @xmath311 .
the bimodal fit performs better at matching the data . although @xmath14 is less sensitive to metallicity , it has the advantage over the sdss indices that its conversion to [ fe / h ] is more well determined . with @xmath312
, we have @xmath313 and @xmath314 . adopting a conversion @xmath315 [ fe / h ]
@xmath316 @xcite based on calibration from the milky way and m31 clusters then gives [ fe / h ] @xmath317 for the blue sequence and @xmath318 for the red sequence .
these mean metallicities should be taken only as rough estimates , since the red - sequence mode in particular requires an extension of the conversion equations into a high - metallicity regime that the calibrating gcs in m31 and the milky way do not adequately cover .
however , the mean @xmath319 values themselves are close to what would be predicted from the established correlations with host galaxy luminosity @xcite , namely @xmath320 for @xmath321 these relations predict @xmath322 and 1.19 for the blue and red modes respectively , in good agreement with our observed mean colors .
supergiant ellipticals such as this one are the most logical places to search for gcs of super - solar metallicity , although higher quality spectra than currently exist will be required to establish their abundances accurately .
we have presented an extensive new photometric study of the large globular cluster system around ngc 3311 , the central cd galaxy in the hydra cluster .
the relatively wide field and deep limiting magnitude of our data from the gemini / gmos camera has allowed us to analyze the gcs with a sample of several thousand globular clusters ; our data reach to @xmath323 , just beyond the turnover point of the gc luminosity function . among our findings are these : * the classic metal - poor blue gc sequence and the metal - richer red sequence are both present at their expected colors , and with nearly equal populations .
we show the color distributions both from our gmos @xmath63 photometry and from hst wfpc2 archival data in @xmath14 .
* bimodal fits to the complete color distribution , subdivided by magnitude , confirm that the blue sequence has the same correlation of progressively increasing metallicity with gc mass that was previously found in many other large galaxies ; the correlation we find corresponds to a scaling of gc metallicity with mass of @xmath4 .
* by contrast , the red sequence shows no change of mean metallicity with mass , though it does extend upward to much higher than normal luminosity into the ucd - like range , strengthening the potential connections between massive gcs and ucds . * confirming previous but much more preliminary investigations , we find that the gcs in ngc 3311 falls definitively into the `` high specific frequency '' category inhabited almost uniquely by supergiant ellipticals ( cd s or bcg s ) at the centers of rich clusters .
ngc 3311 has an estimated total population of about 16,000 clusters and a specific frequency @xmath228 .
* at the core of the hydra cluster , another giant elliptical ngc 3309 is sitting projected just @xmath6 from the cd ngc 3311 .
we use our database to solve simultaneously for the spatial structure and total gc populations of both galaxies at once .
the results show that ngc 3309 contributes only a few percent of the gc population over the entire field , and has an unusually ( though not unprecedentedly ) low value @xmath8 .
the central cd is completely dominant in the central hydra region . *
the gc luminosity function , which we measure down to the `` turnover '' point at @xmath5 , also has a normal structure relative to other giant ellipticals .
additional observations that could make clear improvements to the understanding of the ngc 3311 system would include imaging with a still wider field ( to establish a true `` control '' field for background measurement ) , and multicolor photometry with still more sensitive color indices to improve the definition of the metallicity distribution function .
the resources held by this enormous and fascinating gc system are far from being exhausted .
the authors would like to thank the anonymous referee for their helpful comments in improving this paper .
this work is based on observations obtained at the gemini observatory , which is operated by the association of universities for research in astronomy , inc .
, under a cooperative agreement with the nsf on behalf of the gemini partnership : the national science foundation ( united states ) , the science and technology facilities council ( united kingdom ) , the national research council ( canada ) , conicyt ( chile ) , the australian research council ( australia ) , cnpq ( brazil ) and conicet ( argentina ) .
emhw , weh , and kaw thank the natural sciences and engineering research council of canada for financial support .
ashman , k.m . , bird , c.m . , & zepf , s.e .
1994 , , 108 , 2348 ashman , k.m . , conti , a. , & zepf , s.e .
1995 , , 110 , 1164 baum , w.a . , hammergren , m. , groth , e.j .
, ajhar , e.a . ,
lauer , t.r .
, oneil , e.j.jr . ,
lynds , c.r . ,
faber , s.m . ,
grillmair , c.j . ,
holtzman , j.a .
, & light , r.m .
1995 , , 110 , 2537 barmby , p. , huchra , j.p . ,
brodie , j.p . ,
forbes , d.a . ,
schroder , l.l . , & grillmair , c.j .
2000 , , 119 , 727 bertin , e . , & arnouts , s. 1996 , , 117 , 393 brodie , j.p . , larsen , s.s . , & kissler - patig , m . 2000 , , 543 , l19 dirsch , b. , richtler , r. , geisler , d. , forte , j.c .
, bassino , l.p .
, & gieren , w.p .
2003 , , 125 , 1908 disney , m.j . , &
wall , j.v .
1977 , , 179 , 235 ferrarese , l. , ct , p. , jordn , a. , peng , e.w . ,
blakeslee , j.p . ,
piatek , s. , mei , s. , merritt , d. , milosavljevic , m. , tonry , j.l . , & west , m.j .
2006 , , 164 , 334 fleming , d.e.b . ,
harris , w.e . ,
pritchet , c.j . , & hanes , d.a .
1995 , , 109 , 1044 forbes , d.a .
, faifer , f.r . ,
forte , j.c . , bridges , t. , beasley , m.a . , gebhardt , k. , hanes , d.a . , sharples , r. , & zepf , s.e
2004 , , 355 , 608 fukugita , m. , ichikawa , t. , gunn , j.e . , doi , m. , shimasaku , k. , & schneider , d.p .
1996 , , 111 , 1748 .
geisler , d. , lee , m.g .
, & kim , e. 1996 , , 111 , 1529 girardi , l. , grebel , e.k . ,
odenkirchen , m. , & chiosi , c. 2004 , , 422 , 205 harris , w.e . 2001 , in star clusters , saas - fee advanced course 28 , edited by l.labhardt & b.binggeli ( berlin : springer - verlag ) harris , w.e . , harris , g.l.h . ,
layden , a.c . , & stetson , p.b .
2007 , , 134 , 43 harris , w.e . , kavelaars , j.j . ,
hanes , d.a . , hesser , j. e. , & pritchet , c.j . 2000 , , 533 , 137 harris , w.e . ,
& pudritz , r.e .
1994 , , 429 , 177 harris , w.e . ,
smith , m.g . , &
myra , e.s .
1983 , , 272 , 456 harris , w.e . ,
whitmore , b.c .
, karakla , d. , okon , w. , baum , w.a .
, hanes , d.a . , & kavelaars , j.j .
2006 , , 636 , 90 hempel , m . ,
geisler , d . , hoard , d.w . , & harris , w.e .
2005 , , 439 , 59 holtzman , j.a . , burrows , c.j . ,
casertano , s. , hester , j. , trauger , j.t . ,
watson , a.m. , & worthey , g. 1995 , , 107 , 1065 jordn , a. , mclaughlin , d.e .. , ct , p. , ferrarese , l. , peng , e.w . , mei , s. , villegas , d. , merritt , d. , tonry , j.l . , & west , m.j .
2007 , , 171 , 101 kundu , a. , & whitmore , b.c .
2001a , , 121 , 2950 kundu , a. , & whitmore , b.c .
2001a , , 122 , 1251 landolt , a.u .
1992 , , 104 , 340 larsen , s.s . ,
brodie , j.p .
, huchra , j.p . , forbes , d.a . ,
& grillmair , c.j .
2001 , , 121 , 2974 lee , m. g. , kim , e. & geisler , d. 1998 , , 115 , 947 macdonald , p.d.m .
2007 , documentation and code at http://www.math.mcmaster.ca/peter/mix/mix.html , department of mathematics and statistics , mcmaster university maraston , c. 2005 , , 362 , 799 mclachlan , g.j . , & basford , k.e . 1987 , mixture models : inference and applications to clustering .
marcel dekker , new york mclaughlin , d.e .
1999 , , 117 , 2398 mclaughlin , d.e . ,
harris , w.e . , &
hanes , d.a .
1994 , , 422 , 486 mclaughlin , d.e .
, secker , j. , harris , w.e . , & geisler , d. 1995 , , 109 , 1033 mieske , s. , jordn , a. , ct , p. , kissler - patig , m. , peng , e.w . ,
ferrarese , l. , blakeslee , j.p . ,
mei , s. , merritt , d. , tonry , j.l . , & west , m.j .
2006 , , 653 , 193 ostrov , p.g . , forte , j.c . , & geisler , d. 1998 , , 116 , 2854 puzia , t. h. , kissler - patig , m. , brodie , j. p. , & huchra , j. p. 1999
, , 118 , 2734 rhode , k.l . , & zepf , s.e .
2001 , , 121 , 210 richter , o .- g . ,
materne , j. , & huchtmeier , w , k , 1982 , , 111 , 193 robin , a. c. , reyl , c. , derrire , s. , & picaud , s. 2003 , , 409 , 523 rothberg , b. , harris , w.e . , whitmore , b.c . ,
somerville , r. , & cockcroft , r. 2007 , , submitted secker , j . ,
geisler , d . ,
mclaughlin , d.e .
, & harris , w.e .
1995 , , 109 , 1019 spitler , l.r . ,
larsen , s.s .
, strader , j. , brodie , j.p . ,
forbes , d.a . , & beasley , m.a .
2006 , , 132 , 1593 strader , j. , brodie , j.p . , & forbes , d.a .
2004 , , 127 , 3431 strader , j. , brodie , j.p . ,
spitler , l. , & beasley , m.a .
2006 , , 132 , 2333 tamura , n. , sharples , r.m . ,
arimoto , n. , onodera , m. , ohta , k. , & yamada , y. 2006 , , 373 , 588 vasterberg , a.r . , jorsater , s. , & lindblad , p.o .
1991 , , 247 , 335 wehner , e.h.m , & harris , w.e .
2007 , , in press ( astro - ph / arxiv:0708:1514 ) zepf , s.e . , & ashman , k.m .
1993 , , 264 , 611 zinn , r. 1985 , , 293 , 424 | we present extensive new photometry in @xmath0 of the large globular cluster ( gc ) system around ngc 3311 , the central cd galaxy in the hydra cluster .
our gmos data cover a @xmath1 field of view and reach a limiting magnitude @xmath2 , about 0.5 magnitude fainter than the turnover point of the gc luminosity function .
we find that ngc 3311 has a huge population of @xmath3 gcs , closely similar to the prototypical `` high specific frequency '' virgo giant m87 .
the color - magnitude distribution shows that the metal - poor `` blue '' gc sequence and the metal - richer `` red '' sequence are both present , with nearly equal numbers of clusters .
bimodal fits to the color distributions confirm that the blue sequence shows the same trend of progressively increasing metallicity with gc mass that has previously been found in many other large galaxies ; the correlation we find corresponds to a scaling of gc metallicity with mass of @xmath4 .
by contrast , the red sequence shows no change of mean metallicity with mass , but it shows an upward extension to much higher than normal luminosity into the ucd - like range , strengthening the potential connections between massive gcs and ucds
. the gc luminosity function , which we measure down to the turnover point at @xmath5 , also has a normal form like those in other giant ellipticals . within the hydra field , another giant elliptical ngc
3309 is sitting just @xmath6 from the cd ngc 3311 .
we use our data to solve simultaneously for the spatial structure and total gc populations of both galaxies at once .
their specific frequencies are @xmath7 and @xmath8 .
ngc 3311 is completely dominant and entirely comparable with other cd - type systems such as m87 in virgo . |
this work has its roots in the challenge of extending differential equation models for diffusion or wave propagation from domains in euclidean space to infinite graphs intended to resemble biological transport systems such as the arteries of the human circulatory system .
such biological systems can include enormous numbers of branching segments .
short time transport across the network is essential , so treelike structures with small numbers of large edges and vast collections of microscopic edges are typical . faced with such complex heterogeneous structures , one hopes that appropriate infinite graph models will suggest useful structural features and robustly posed problems . building on an earlier quantum graph analysis of such problems
@xcite , this work uses infinite graph and operator theoretic methods to treat a class of continuous time markov chains . recall that continuous time markov chains use a system of constant coefficient differential equations @xmath0 to describe the evolution of probability densities @xmath1 on a finite or countably infinite set of states .
an associated graph may be constructed by connecting states ( vertices ) @xmath2 and @xmath3 with an edge if @xmath4 . in the finite state case
the solution of is simply @xmath5 .
when the set of states is infinite the formal description of the operator @xmath6 may not be adequate to determine the semigroup @xmath7 , an issue known in probability as the problem of explosions .
infinite graph models inspired by biological transport systems will typically face the explosion problem . by imposing restrictions on both the form of the markov chain generator @xmath6 and the structure of the associated graph viewed as a metric space
, this work provides a resolution in terms of reflecting and absorbing behavior at a graph boundary
. it will be advantageous to use the dirichlet form theory @xcite . to that end , consider a graph @xmath8 whose edges @xmath9 $ ] are equipped with positive weights @xmath10 which are interpreted as edge length . with @xmath11 , a symmetric bilinear form for functions on the vertex set is defined by @xmath12 each vertex is also given a positive weight @xmath13 .
formal semigroup generators @xmath14 are defined by @xmath15 @xmath10 is often interpreted as electrical resistance . the electrical network analogy is treated at length in @xcite . the recent work @xcite treats electrical currents in a context similar to this paper , while @xcite treats related topological questions .
an analysis of function theory on infinite trees motivated by modeling the human lungs is in @xcite . with the domain of functions with finite support
, @xmath14 is a symmetric operator on @xmath16 .
in contrast to this paper , other recent work @xcite has stressed cases when this symmetric operator is essentially selfadjoint , and so behavior at the graph boundary is not an issue .
the vertex set @xmath17 of an edge weighted locally finite graph @xmath8 can be equipped with a metric @xmath18 obtained by minimizing the sum of the edge lengths of paths from @xmath19 to @xmath20 . by completing this metric space
we obtain a metric space @xmath21 in which one can discuss features like the graph boundary and compactness .
if @xmath8 is a tree , then distinct points of @xmath21 can be separated by deleting a suitable edge .
generalizing this idea , our graphs will be required to have weakly connected completions , with the property that , for any two distinct points , any path joining them must include an edge from a finite set .
this generalization identifies a rich class of edge weighted graphs with useful topological and function theoretic properties .
the properties of weakly connected graph completions are developed in the second section .
in addition to trees , arbitrary graphs with finite volume have weakly connected completions .
this class is also preserved if we add suitably constrained edge sequences to a graph .
weakly connected completions are totally disconnected metric spaces .
when also compact , these spaces are topologically stable with respect to decrease of the metric .
the weakly connected class will be characterized using the separation of points property for an algebra of eventually flat functions .
the third section treats the bilinear forms , vertex weights , and associated operators .
the choice of vertex weights typically used for discrete time markov chains are contrasted with weights making @xmath14 resemble a discretized second derivative .
the bilinear form is used to construct several sobolev style hilbert spaces on @xmath8 whose elements extend continuously to @xmath22 .
two main problems are treated in the fourth section .
the first , a version of the dirichlet problem , asks for conditions under which continuous functions on @xmath23 have a unique harmonic extension to @xmath8 .
an example shows that a lack of compactness can lead to a negative result .
using assumptions of compactness and weak connectivity , a general positive result is established .
the second problem is the resolution of the explosion problem in terms of reflecting and absorbing boundary conditions .
the semigroups generated by the operators defined using these boundary conditions are positivity preserving contractions on @xmath24 . despite the connections with probability ,
this work will not explicitly use probabilistic techniques or interpretations .
we simply mention the classic work @xcite , and the recent works @xcite as pointers to the enormous literature related to analysis of infinite state markov chains .
@xmath8 will denote a simple graph with a countable vertex set @xmath17 and a countable edge set @xmath25 .
each vertex will have at least one and at most finitely many incident edges .
vertices of degree @xmath26 are _ boundary vertices _ ; the rest are _
interior vertices_. @xmath8 is assumed to have _ edge weights _ ( _ resistances _ ) .
that is , there is a function @xmath27 , denoted by @xmath28 when @xmath29 \in \edgeset $ ] .
general references on graphs are @xcite .
edge weights , considered the length of the edges , are commonly identified with electrical network resistance @xcite or @xcite , and then _ edge conductance _ is the reciprocal @xmath30 if @xmath31 , and @xmath32 otherwise . a _ finite path _ ( sometimes called a walk ) in @xmath8 connecting vertices @xmath19 and @xmath20 is a finite sequence of vertices @xmath33 such that @xmath34 \in \edgeset $ ] for @xmath35 .
@xmath8 is connected there is a finite path from @xmath19 to @xmath20 for all @xmath36 .
define a metric on ( the vertices of ) @xmath8 by @xmath37 the infimum taken over all finite paths @xmath38 joining @xmath19 and @xmath20 . if there is no finite path from @xmath19 to @xmath20 then @xmath39 .
@xmath21 , with the extended metric @xmath40 , will denote the metric space completion @xcite of @xmath8 . extending the combinatorial notion of path
, a _ path _ in @xmath21 will be a sequence @xmath41 with @xmath42 , @xmath34 \in \edgeset$ ] , where the index set may be finite ( finite path ) , the positive integers ( a ray ) , or the integers ( a double ray ) .
the role of continuous paths in @xmath21 is played by paths going from @xmath43 to @xmath44 , which in the double ray case requires @xmath45 and @xmath46 .
the ray case is similar .
a path for which all vertices are distinct is a _
simple path_. if @xmath8 is connected then there is a path joining any pair of points @xmath47 . modifying ideas from @xcite ,
say that @xmath21 is _ weakly connected _ if for every pair of distinct points @xmath48 there is a finite set @xmath49 of edges in @xmath8 such that every path from @xmath19 to @xmath20 contains an edge from @xmath49 .
one may extend @xmath8 to a metric graph @xmath50 by identifying the combinatorial edge @xmath29 $ ] with an interval of length @xmath51 . with the usual metric on @xmath50
, its vertex set will be isomorphic to @xmath8 . by this device some of the results of @xcite ,
which should be consulted for more details , carry over to the present context .
the next result is a simple example .
if @xmath52 is a tree then @xmath53 is weakly connected .
the volume of a graph is defined as the sum of its edge lengths , @xmath54 \in \edgeset } \rest(v_1,v_2).\ ] ] finite volume graphs also have weakly connected completions @xcite . if @xmath55 then @xmath21 is weakly connected .
the main case considers distinct points @xmath56 and @xmath57 in @xmath58 .
remove a finite set of edges from @xmath59 so that the remaining edgeset @xmath60 satisfies @xmath61 \in \edgeset _ 1 } \rest(v_1,v_2 ) <
\frac{d(x , y)}{2}.\ ] ] proceeding with a proof by contradiction , suppose @xmath62 is a path from @xmath56 to @xmath57 using only edges in @xmath60 .
then for @xmath63 sufficiently large @xmath64 , but there is a simple path from @xmath65 to @xmath66 using only edges from @xmath60 , so @xmath67 .
the next result gives conditions allowing edges to be added to a weakly connected graph without disturbing that property .
[ addedge ] suppose the graph @xmath68 , with vertex set @xmath17 , has a weakly connected completion @xmath69 .
using the same vertex set @xmath17 , enlarge @xmath68 to a graph @xmath70 by adding a sequence @xmath60 of edges @xmath71 whose lengths @xmath72 satisfy @xmath73 .
assume there is a positive constant @xmath74 such that @xmath75 then @xmath76 is weakly connected .
first note that @xmath77 , since the set of cauchy sequences of vertices has not changed . it will suffice to consider distinct points @xmath19 and @xmath20 in @xmath78 which are joined by a path in @xmath76 .
let @xmath79 be a finite set of edges in @xmath68 such that every path in @xmath69 from @xmath19 to @xmath20 contains an edge @xmath80 $ ] from @xmath79 .
pick @xmath81 such that @xmath82 for all edges @xmath80 \in w_0 $ ] .
find @xmath83 so that the lengths @xmath84 of edges @xmath85 satisfy @xmath86 for @xmath87 .
let @xmath88 be the set of edges @xmath89 in @xmath70 .
suppose there is a path @xmath38 in @xmath76 joining @xmath19 to @xmath20 , but not containing any edge from @xmath88 .
let @xmath90 be the set of edges @xmath91 which are in @xmath38 .
@xmath90 is not empty since @xmath38 contains at least one edge not in @xmath68 .
since @xmath38 contains no edge from @xmath88 , the edges @xmath92 have length @xmath93 .
if an edge @xmath94 \in \gamma $ ] has @xmath95 , then @xmath96 can be replaced by a finite path @xmath97 in @xmath68 with length at most @xmath98 , and containing no edge from @xmath79 . thus there is at least one @xmath94 \in \gamma $ ] with @xmath99 .
but the inequalities @xmath100 while @xmath99 contradict the hypotheses , so no such path from @xmath19 to @xmath20 exists , and @xmath76 is weakly connected .
the conclusion of theorem [ addedge ] may be false if the edge lengths @xmath84 have a positive lower bound .
start with @xmath68 which is be weakly connected and connected .
take distinct points @xmath101 .
suppose @xmath38 is a path from @xmath56 to @xmath57 .
take sequences of vertices @xmath102 from @xmath38 with @xmath103 and @xmath104 such that @xmath105 , and @xmath106 $ ] is not an edge in @xmath68 , form @xmath70 by adding edges @xmath106 $ ] to @xmath68 , with @xmath107 .
since @xmath108 for all vertices @xmath109 , the vertex sets for @xmath68 and @xmath70 are isometric .
the edge lengths @xmath110 have a positive lower bound , and @xmath76 is not weakly connected , [ disconnect ] assume @xmath21 is weakly connected .
if @xmath111 and @xmath112 are disjoint compact subsets of @xmath21 , then there is a finite set @xmath49 of edges in @xmath8 such that every path from @xmath111 to @xmath112 contains an edge in @xmath49 . since @xmath21 is weakly connected , if @xmath113 and @xmath114 there is a finite set @xmath115 of edges in @xmath8 such that every path from @xmath19 to @xmath20 contains an edge in @xmath49 .
take @xmath81 such that @xmath82 for all edges @xmath80 \in w$ ] . if @xmath116 , the open @xmath117 ball centered at @xmath19 , and @xmath118 , then every path from @xmath119 to @xmath120 contains an edge in @xmath115 .
the collection @xmath121 is an open cover of the compact set @xmath122 , so there is a finite subcover @xmath123 for @xmath124
. take @xmath125 .
suppose @xmath49 is a nonempty finite set of edges in @xmath59 .
for @xmath126 , let @xmath127 be the set of points @xmath128 which can be connected to @xmath56 by a path containing no edge of @xmath49 .
[ cloplem ] for all @xmath126 the set @xmath127 is both open and closed in @xmath21 .
take @xmath81 such that @xmath82 for all edges @xmath80 \in w$ ] .
suppose @xmath57 and @xmath129 are vertices with @xmath130 , so there is a path of length smaller than @xmath117 from @xmath57 to @xmath129 .
if there is a path from @xmath57 to @xmath56 containing no edge from @xmath49 , then by concatenating these paths there is a path from @xmath129 to @xmath56 containing no edge from @xmath49 .
this shows that @xmath127 and the complement of @xmath127 are both open . [ totally ]
a weakly connected @xmath21 is totally disconnected .
suppose @xmath131 and @xmath132 are distinct points in @xmath22 , with @xmath49 being a finite set of edges in @xmath8 such that every path from @xmath131 to @xmath132 contains an edge from @xmath49 .
the set @xmath133 is both open and closed . since @xmath133 and @xmath134 are disjoint , @xmath135 , the complement of @xmath133 in @xmath21 .
thus @xmath131 and @xmath132 lie in different connected components .
if @xmath21 is totally disconnected and compact , it has a rich collection of clopen sets , that is sets which are both open and closed .
in fact @xcite or @xcite for any @xmath126 and any @xmath81 there is a clopen set @xmath111 such that @xmath136 . in particular any compact subset of @xmath21
can then be approximated by a clopen set .
changing the metric on @xmath8 may change the completion @xmath21 and the functions on @xmath8 that extend continuously to graphbar . given @xmath8 and two weight functions @xmath137 and @xmath138 ,
let @xmath69 and @xmath76 denote the completions of @xmath139 with respect to the associated metrics @xmath140 and @xmath141 .
say that the weight function @xmath142 is smaller than the weight function @xmath143 if @xmath144 for every edge @xmath29
\in \edgeset $ ] .
if @xmath138 is smaller than @xmath137 , the metric @xmath141 on @xmath8 extends to a pseudometric @xcite on @xmath69 ; that is , there may be distinct points @xmath145 with @xmath146 .
the next result establishes a stability property for weakly connected graph completions [ changemet ] suppose @xmath147 is weakly connected .
if @xmath138 is smaller than @xmath137 then the pseudometric on @xmath69 induced by @xmath138 is a metric .
if in addition @xmath148 is compact , then @xmath148 and @xmath149 are homeomorphic .
the usual construction @xcite identifies the completion of a metric space with equivalence classes of cauchy sequences , two sequences @xmath150 and @xmath151 being equivalent when @xmath152 .
since @xmath8 is locally finite , it is only possible to have distinct points @xmath145 with @xmath146 if @xmath153 .
suppose @xmath56 and @xmath57 are distinct points of @xmath154 , and that @xmath49 is a finite set of edges such that any path from @xmath56 to @xmath57 contains an edge from @xmath49 .
let @xmath150 and @xmath151 be sequences in @xmath155 with @xmath156 and @xmath104 . by lemma [ cloplem ]
there is an @xmath83 such that @xmath157 implies any path in @xmath8 from @xmath158 to @xmath159 must contain an edge from @xmath49 .
let @xmath160 for all edges @xmath29 \in w$ ] .
then @xmath161 for all @xmath157 , so the sequences @xmath150 and @xmath151 are still inequivalent cauchy sequences with respect to @xmath141 , showing that @xmath141 is a metric on @xmath69 .
let @xmath162 denote the identity map from @xmath148 to @xmath149 , which is distance reducing , so continuous . if @xmath148 is compact , then so is @xmath149 .
following @xcite , suppose @xmath163 is closed .
then @xmath164 is compact , so closed , and @xmath165 is continuous .
there is an algebra of functions with pointwise addition and multiplication well matched to the weakly connected graph completions .
define the eventually flat functions @xmath166 to be the real algebra of functions @xmath167 such that the set of edges @xmath29 $ ] in @xmath59 with @xmath168 is finite .
all functions @xmath169 extend continuously to @xmath21 .
let @xmath49 be the finite set of edges @xmath29 $ ] with @xmath168 .
suppose @xmath126 , @xmath170 , and @xmath156 . if @xmath171 then @xmath172 for @xmath63 sufficiently large .
suppose @xmath173 .
by lemma [ cloplem ] the set @xmath174 is open , so there is an @xmath83 such that @xmath157 implies @xmath175 and @xmath176 .
take @xmath177 . making use of this lemma , functions @xmath169 are extended continuously to functions on @xmath21 .
if @xmath8 is connected , then any @xmath169 has finite range . for @xmath178
, @xmath179 is a clopen set in @xmath21 .
suppose @xmath169 is not constant , and suppose @xmath180 .
find a path @xmath181 such that @xmath182 for @xmath183 and @xmath184 for some @xmath185 adjacent to @xmath186 . since the set of such vertices @xmath186 is finite , @xmath187 has one of a finite set of values . for @xmath178
, @xmath179 is a closed set , and its complement in @xmath21 is the union of a finite collection of closed sets .
the next result shows that @xmath166 separates points of @xmath21 if and only if @xmath21 is weakly connected .
[ separate ] suppose @xmath21 is weakly connected .
if @xmath56 and @xmath57 are distinct points of @xmath21 , there is a function @xmath169 whose range is @xmath188 such that @xmath189 for @xmath129 in an open neighborhood @xmath111 of @xmath56 , and @xmath190 for @xmath129 in an open neighborhood @xmath112 of @xmath57 .
conversely , if @xmath166 separates points of @xmath21 , then @xmath21 is weakly connected .
let @xmath49 be a finite set of edges in @xmath8 such that every path from @xmath56 to @xmath57 contains an edge from @xmath49 .
by lemma [ cloplem ] the set @xmath127 is both open and closed in @xmath21 , as is @xmath191 define @xmath189 for @xmath192 and @xmath190 for @xmath193 . for every vertex @xmath20 and adjacent vertex @xmath185 we have @xmath194 unless @xmath195 \in w$ ] . since @xmath49 is finite , @xmath169 . in the other direction ,
suppose @xmath166 separates points of @xmath21 .
let @xmath56 and @xmath57 be distinct points in @xmath21 , and suppose @xmath169 with @xmath196 .
let @xmath49 be the finite set of edges @xmath29 $ ] such that @xmath168 .
if @xmath38 is any path starting at @xmath56 which contains no edge from @xmath49 , then @xmath197 must be constant along @xmath38 .
that is , every path from @xmath56 to @xmath57 must contain an edge from @xmath49 , so @xmath21 is weakly connected .
[ sepsets ] suppose @xmath22 is weakly connected .
if @xmath198 and @xmath199 are nonempty disjoint compact subsets of @xmath21 , then there is a function @xmath200 such that @xmath201 , @xmath202 first fix @xmath203 . using theorem [ separate ] , find a finite cover @xmath204 of @xmath198 by open sets with corresponding functions @xmath205 which satisfy @xmath206 for all @xmath129 in some open neighborhood @xmath207 of @xmath57 , while @xmath208 for all @xmath129 in @xmath209 .
define @xmath210 .
find a finite collection @xmath211 whose corresponding open sets @xmath212 cover @xmath199 .
the function @xmath213 has the desired properties . the combination of theorem [ separate ] and the stone - weierstrass theorem yields the next result . [ stone ]
if @xmath21 is weakly connected and compact , then @xmath166 is uniformly dense in the continuous functions on @xmath21 .
vertex weights and the corresponding measure are now added to the edge weighted graph @xmath8 .
a vertex weight function @xmath214 provides the borel measure @xmath215 on @xmath8 , which may be extended @xcite to @xmath21 by defining the measure of @xmath216 to be @xmath32 .
the real hilbert space @xmath16 will consist of functions @xmath217 with @xmath218 , and inner product @xmath219 .
the set of functions @xmath217 which are @xmath32 at all but finitely many vertices is denoted by @xmath220 . also introduce @xmath221 .
the next proposition collects basic facts about the symmetric bilinear forms induced by the edge conductances @xmath222 .
closely related results using the smaller domain @xmath220 are in @xcite and @xcite .
[ bform ] suppose @xmath223 satisfies @xmath224 , with @xmath225 if and only if @xmath29 \in \edgeset$ ] .
given a vertex weight @xmath226 , the symmetric bilinear form @xmath227 has a nonnegative quadratic form @xmath228 , and satisfies @xmath229 where @xmath230 the nonnegativity of the quadratic form is immediate from the definition .
note that for any @xmath200 there are only finitely many vertices @xmath231 for which @xmath232 is nonzero if @xmath19 is adjacent to @xmath20 . to identify the operator @xmath14 , start with @xmath233 @xmath234
suppose a graph edge @xmath235 has vertices @xmath236 and @xmath237 .
the second sum over @xmath238 in can be viewed as a sum over edges , with each edge contributing the terms @xmath239 and @xmath240 . using this observation to change the order of summation gives @xmath241
@xmath242 @xmath243 employing this identity in gives @xmath244.\ ] ] with respect to the standard basis consisting of functions @xmath245 with @xmath246 and @xmath247 for @xmath248 , the operators @xmath14 have the matrix representation @xmath249 if @xmath20 is fixed , then summing on @xmath185 gives @xmath250 so @xmath251 is a @xmath6-matrix in the sense of markov chains @xcite . in the @xmath6 - matrix formulation
the matrix entries represent transition rates , so decreasing the vertex measure @xmath226 increases the rates .
consistent with the boundary value themes of this work an interesting choice is to take the vertex weight @xmath252 to be half the sum of the lengths of the incident edges , @xmath253 this choice makes the vertex measure consistent with the previously defined graph volume , @xmath254 if @xmath255 , then @xmath256 will include all functions in @xmath257 .
with respect to this measure @xmath258 this operator resembles the symmetric second difference operator from numerical analysis . like the classical laplace operator in euclidean space , @xmath259 exhibits quadratic scaling behavior .
suppose @xmath260 have equal numbers of incident edges , and there is a bijection of vertex neighborhoods with @xmath261 for @xmath262 , and @xmath263 .
if @xmath217 satisfies @xmath264 and @xmath265 , then @xmath266 the vertex weight @xmath267 is typically distinct from @xmath268 , a choice which appears in the study of discrete time markov chains @xcite , @xcite , @xcite with transition probabilities @xmath269 for @xmath270 .
the next result considers continuous extension of functions to @xmath21 when the quadratic form is finite .
[ formcont ] suppose @xmath8 is connected .
using the metric of , functions @xmath271 with @xmath272 are uniformly continuous on @xmath8 , and so @xmath273 extends uniquely to a continuous function on @xmath21 .
if @xmath274 and @xmath275 is any finite simple path from @xmath20 to @xmath185 , then the cauchy - schwarz inequality gives @xmath276 \frac{c^{1/2}(v_{k+1},v_k)}{c^{1/2}(v_{k+1},v_k ) } \big |^2\ ] ] @xmath277 \sum_k r(v_{k+1},v_k)\ ] ] @xmath278 there is a simple path with @xmath279 , so latexmath:[\[\label{formest } uniformly continuous on @xmath8 .
by @xcite @xmath273 extends continuously to @xmath21 .
the bilinear form may be used to define a sobolev style hilbert space @xmath281 with inner product @xmath282 let @xmath283 be the closure of @xmath220 in @xmath284 .
[ h1lem ] if @xmath8 is connected with finite diameter , then there is a constant @xmath74 such that @xmath285 so a cauchy sequence in @xmath281 is a uniform cauchy sequence .
the functions @xmath273 in the unit ball of @xmath286 are uniformly equicontinuous @xcite . fixing a vertex @xmath287 ,
gives @xmath288 which is .
the uniform equicontinuity follows from .
[ zerobnd ] suppose @xmath8 is connected and has finite diameter . if @xmath289 , then @xmath273 has a unique continuous extension to @xmath22 which is zero at all points @xmath173 .
any function @xmath289 is the limit in @xmath281 of a sequence @xmath290 from @xmath220 . the functions @xmath273 and @xmath290 have unique continuous extensions to @xmath22 by theorem [ formcont ] .
the extended functions @xmath290 satisfying @xmath291 for all @xmath173 . by lemma [ h1lem ]
the sequence @xmath290 converges to @xmath273 uniformly on @xmath8 , so the extensions @xmath290 converge uniformly to the extension @xmath273 on @xmath22 .
thus @xmath292 for all @xmath173 .
the symmetry and nonnegativity of @xmath294 are given by .
since @xmath294 is densely defined , @xmath296 is the operator whose graph is the set of pairs @xmath300 such that @xmath301 for all @xmath302 .
suppose @xmath303 then for any @xmath304 in the domain of @xmath296 , @xmath305h(w)\ ] ] @xmath306 proposition [ lapdef ] provides a basic laplace operator , the friedrich s extension @xcite of @xmath294 , whose domain is a subset of @xmath283 the closure of @xmath220 in @xmath284 .
let @xmath307 denote the friedrich s extension of @xmath293 .
several features of @xmath307 are implied by the condition @xmath308 .
the friedrich s extension @xmath307 of the nonnegative symmetric operator @xmath294 has the same lower bound , so it suffices to consider functions @xmath302 . since @xmath312 there must be some vertex @xmath20 where @xmath313 .
since @xmath273 has finite support , there is another vertex @xmath19 with @xmath314 .
an application of gives @xmath315 ^ 2 \le 4b(f , f)d(u , v).\ ] ] suppose @xmath8 is connected , @xmath21 is compact , and @xmath316 is finite . let @xmath317 be a symmetric extension of @xmath294 in @xmath16 whose associated quadratic form is @xmath318 then the friedrich s extension @xmath319 of @xmath317 has compact resolvent .
the resolvent of @xmath319 maps a bounded set in @xmath16 into a bounded set in @xmath281 .
suppose @xmath290 is a bounded sequence in @xmath16 , with @xmath320 .
by lemma [ h1lem ] and the arzela - ascoli theorem @xcite the sequence @xmath321 has a uniformly convergent subsequence , which converges in @xmath16 .
when considering harmonic functions , it will be convenient to treat vertices of degree one ( boundary vertices ) as part of the boundary of @xmath21 .
for instance , this convention allows finite graphs with degree one vertices to have nonconstant harmonic functions .
let @xmath322 denote the set of interior vertices , that is vertices with degree at least @xmath323 .
say that a function @xmath271 is _ harmonic on @xmath8 _ if @xmath324 for all vertices @xmath325 .
since the value @xmath326 of a harmonic function is the weighted average of the values at the adjacent vertices , @xmath273 has a minimum or maximum at some @xmath325 of a connected graph if and only if @xmath273 is constant .
the set of harmonic functions is independent of the vertex weight @xmath327 , but harmonic functions may be described as those for which @xmath328 if @xmath32 were an eigenvalue of @xmath307 , then there would be a corresponding eigenfunction @xmath273 which is positive somewhere . by theorem [ zerobnd ] , @xmath273 would extend continuously to @xmath22 , with @xmath292 for all @xmath173 . since @xmath21 is compact
, @xmath273 has a positive maximum at some @xmath171 .
since holds at all vertices , @xmath273 is constant , but then @xmath329 implies @xmath330 .
define @xmath23 to be the complement of @xmath322 in @xmath21 .
the dirichlet problem asks whether every continuous function @xmath331 has a continuous extension @xmath332 which is harmonic on @xmath8 .
the following example shows that the dirichlet problem is not always solvable .
as shown in figure a , construct a graph @xmath8 beginning with a sequence of vertices @xmath66 and with edges @xmath333 , n = 0,1,2, .
take @xmath334 .
for each of the vertices @xmath66 introduce @xmath335 additional vertices @xmath336 , and edges @xmath337 of length @xmath26 . in this example
, @xmath338 consists of the boundary vertices @xmath336 together with the boundary point at @xmath339 .
all of these points are isolated in @xmath338 .
[ harm1 ] suppose @xmath8 is connected , while @xmath22 is weakly connected and compact .
every continuous function @xmath343 has a unique extension to a continuous function @xmath344 that is harmonic on @xmath8 .
the proof breaks into two main parts . for the first part ,
assume that @xmath8 has no boundary vertices , so @xmath340 is defined on @xmath216 .
the second part of the proof will extend this partial result to the full theorem .
the uniqueness is a standard consequence of the maximum principle since the difference of two solutions would be @xmath32 on @xmath216 . by the tietze extension theorem @xcite
, @xmath340 may be extended to a continuous function on @xmath21 . by theorem
[ stone ] , for any @xmath345 there is a @xmath346 with @xmath347 pick vertex weights @xmath226 with @xmath348 . since @xmath346 , the function @xmath349 has the value @xmath32 except at a finite set of vertices .
the compactness of @xmath21 means that @xmath8 has finite diameter . by proposition [ lbnd ]
the operator @xmath350 has a strictly positive lower bound , so the equation @xmath351 has a solution .
the function @xmath352 is harmonic on @xmath155 , continuous on @xmath21 , and satisfies @xmath353 since @xmath354 converges uniformly on @xmath216 , the maximum principle implies it is a uniformly cauchy sequence on @xmath21 , so there is a continuous limit @xmath273 on @xmath21 . at each @xmath231 @xmath355 @xmath356
so @xmath273 is harmonic on @xmath8 .
this completes the first main part of the proof .
the second main step involves extending the theorem to include boundary vertices . for each edge
@xmath357 $ ] with boundary vertex @xmath358 and edge length @xmath359 , delete the vertex @xmath358 and edge @xmath235 from @xmath8 .
for @xmath360 , add a sequence of new vertices @xmath66 and edges @xmath361 $ ] . assume the edges @xmath71 have length @xmath362 with @xmath363 .
the resulting graph @xmath70 will have a completion @xmath76 containing a point @xmath364 .
suppose a continuous function @xmath365 is given , and is extended to the harmonic function @xmath366 on @xmath70 .
with @xmath367 we have @xmath368\ ] ] @xmath369 or @xmath370 and so @xmath366 is a linear function of the distance along the path from @xmath287 to @xmath185 . at @xmath287
the fact that @xmath366 is harmonic means @xmath371 the linearity of @xmath366 along the path means that @xmath372 now define the function @xmath373 by @xmath374 for @xmath375 , and @xmath376 .
the function @xmath273 is harmonic on @xmath8 , and extends to the continuous function @xmath377 where @xmath378 for @xmath173 , while @xmath379 for each boundary vertex @xmath358 . in this section
absorbing and reflecting boundary conditions are used to construct distinct nonnegative self adjoint extensions of @xmath293 .
the constructed operators extend to semigroup generators which are positivity preserving contractions on @xmath24 .
given a closed set @xmath380 , let @xmath381 denote the subalgebra of @xmath166 vanishing on @xmath198 . define the domain @xmath382 let @xmath383 denote the operator with domain @xmath384 acting on @xmath295 by @xmath385 . by proposition [ bform ]
the operator @xmath386 is nonnegative and symmetric , with quadratic form @xmath387 .
let @xmath388 denote the friedrich s extension of @xmath386 , and note that the domain of @xmath388 is a subset of @xmath281 .
a slight modification of the proof of theorem [ zerobnd ] shows that every function @xmath389 in the domain of @xmath390 extends continuously to @xmath21 with @xmath391 for @xmath392 .
[ distinct ] suppose @xmath21 is connected , weakly connected , and compact .
assume that for @xmath397 the sets @xmath398 are compact , and that @xmath399 for some @xmath400 .
if @xmath401 , then @xmath402 and @xmath403 have distinct domains . reversing the roles of @xmath404 and @xmath405 if necessary
, we may assume @xmath406 .
let @xmath407 denote the complement of @xmath405 .
find @xmath408 , and an @xmath409 with @xmath410 such that @xmath411 .
the sets @xmath412 and @xmath413 are disjoint and compact .
corollary [ sepsets ] shows there is a function @xmath200 with @xmath414 , while @xmath415 for all @xmath416 .
since @xmath417 , the function @xmath273 is in the domain of @xmath418 , but @xmath273 is not in the domain of @xmath419 . since the operators @xmath420 are nonnegative self adjoint on @xmath16 , the operators of the semigroup @xmath421 are @xmath295 contractions for @xmath422 . for applications to probability
this is not sufficient .
let @xmath423 denote the domain of @xmath424 .
the method of @xcite will show that the quadratic forms @xmath425 associated to @xmath426 are dirichlet forms .
there are two conditions to check .
the first condition is that @xmath427 implies @xmath428 and @xmath429 .
since the form is @xmath430 the first condition holds for @xmath431 .
if @xmath432 then there is a sequence @xmath433 with @xmath434 it follows that @xmath435 and @xmath429 .
arguing by analogy , one expects the operator @xmath445 to exhibit absorption at @xmath198 and reflection at @xmath446 .
this analogy may be tested by estimating the rate of decay of a probability density function initially supported near the reflecting boundary .
suppose @xmath21 is compact , connected and weakly connected .
using theorem [ totally ] and the subsequent remarks , let @xmath112 be a clopen neighborhood of @xmath198 , with @xmath447 .
the edge boundary of @xmath111 will be the set @xmath448 of edges @xmath9 \in \graph $ ] with @xmath449 and @xmath114 .
the edge boundary @xmath450 is a finite set since by theorem [ disconnect ] there is a finite set of edges @xmath49 such that any path from @xmath111 to @xmath112 contains an edge from @xmath49 , and @xmath451 .
let @xmath452 be the indicator function of @xmath111 , with @xmath453 for @xmath454 and @xmath455 for @xmath456 .
the function @xmath452 is in @xmath166 . if @xmath457 then @xmath458 is in the domain of @xmath459 .
since the semigroup is positivity preserving and a contraction on @xmath469 , @xmath470 that is , the decay rate for @xmath471 is controlled by what happens at @xmath448 , without regard to the rest of the boundary of @xmath111 . | formal laplace operators are analyzed for a large class of resistance networks with vertex weights .
the graphs are completed with respect to the minimal resistance path metric . compactness and a novel connectivity hypothesis for the completed graphs play an essential role .
a version of the dirichlet problem is solved .
self adjoint laplace operators and the probability semigroups they generate are constructed using reflecting and absorbing conditions on subsets of the graph boundary .
_ mathematics subject classification .
_ primary 34b45 _ keywords . _
boundary value problems on networks , resistance networks , dirichlet forms , markov chain explosions . |
turbulence is a strong nonequilibrium phenomenon , exhibiting unpredictable behavior of the velocity field , which can result from the multiple degrees of freedom and nonlinearity of fluid systems @xcite .
this situation complicates our understanding of turbulence , making it one of the unresolved problems in modern physics .
the studies on turbulence in quantum fluids have possibility to shed new light on our understanding of turbulence . in quantum turbulence ( qt ) for quantum fluids such as superfluid helium and one - component atomic bose - einstein condensate ( bec ) @xcite , many quantized vortices with quantized circulation are nucleated , forming tangles .
this situation is drastically different from classical turbulence ( ct ) realized in classical fluids because the circulation of velocity field is not quantized in this system .
therefore , the element of qt is more obvious than that in ct , and qt is believed to be useful for the understanding of ct .
recently , a new trend begins to appear in turbulence study in atomic becs , which is turbulence in multi - component becs @xcite .
atomic becs have various characteristic features , one of which is a realization of multicomponent becs @xcite . in this system
, there exists not only a velocity field but also a ( quasi ) spin field . owing to the spin degrees of freedom , various topological excitations such as monopole , skyrmion , knot , domain wall , and vortex
therefore , in this system , one expects novel turbulence , in which both the velocity and spin fields are much disturbed and various topological excitations are generated .
thus , this kind of turbulence should give us new observations for turbulence not found in conventional systems .
previously , we studied turbulence in a spin-1 ferromagnetic spinor bec , which is a typical multicomponent bec @xcite . in this turbulence , the spin field
is disturbed , so that we call it spin turbulence ( st ) . in our previous study
, we focused on the spectrum of the spin - dependent interaction energy corresponding to the spin correlation , finding the characteristic @xmath1 power law .
however , this observation sees only one side of st . in st ,
the velocity and spin fields interact with each other , so that a coupled turbulence with two fields can be realized , lending the possibility of showing a property of the velocity field not seen in conventional ct and qt . studying the both sides of turbulence completes the story . in this paper
, we treat this problem , focusing on the spectrum of the superflow kinetic energy in a spin-1 ferromagnetic spinor bec .
the spectrum of the kinetic energy is theoretically and numerically found to show a kolmogorov spectrum through the interaction between the spin and velocity fields .
the kolmogorov spectrum refers to the @xmath0 power law in the kinetic energy spectrum , which is known to appear in ct and qt @xcite .
this spectrum is considered to be related to the vortex dynamics , so that it is significant to confirm it for understanding the turbulence addressed here .
furthermore , when the @xmath0 power law is sustained , the spectrum of the spin - dependent interaction energy exhibits a @xmath1 power law simultaneously .
therefore , we obtain the coupled turbulence with the disturbed spin and superfluid velocity fields sustaining the two power laws , anew calling it spin - superflow turbulence ( sst ) instead of st .
we consider a bec of spin-1 bosonic atoms with mass @xmath2 at zero temperature without trapping and magnetic fields .
this system is well described by the macroscopic wave functions @xmath3 ( @xmath4 ) with magnetic quantum number @xmath5 , which obey the spin-1 spinor gorss - pitaevskii ( gp ) equations @xcite given by @xmath6 in this paper , roman indices that appear twice are to be summed over @xmath7 , 0 , and 1 , and greek indices are to be summed over @xmath8 , @xmath9 , and @xmath10 .
the parameters @xmath11 and @xmath12 are the coefficients of the spin - independent and spin - dependent interactions , which are expressed by @xmath13 and @xmath14 , respectively
. here , @xmath15 and @xmath16 are the @xmath17-wave scattering lengths corresponding to the total spin-@xmath18 and spin-2 channels .
the total density @xmath19 and the spin density vector @xmath20 ( @xmath21 ) are given by @xmath22 and @xmath23 , where @xmath24 are the spin-1 matrices .
the sign of the coefficient @xmath12 drastically changes the spin dynamics . in this paper
, we consider the ferromagnetic interaction @xmath25 . in this paper , we focus on the energy spectra for the superflow kinetic and the spin - dependent interaction energy .
the kinetic energy of superfluid velocity @xmath26 per unit mass is given by @xmath27 where @xmath28 and @xmath29 is a total particle number .
the superfluid velocity @xmath30 is given by @xmath31 by using the fourier series @xmath32 , we define the spectrum for the kinetic energy per unit mass : @xmath33 where @xmath34 and @xmath35 are given by @xmath36 and @xmath37 , respectively , with system size @xmath38 and spatial dimension @xmath39 .
similarly , the spectrum of the spin - dependent interaction energy per unit mass is defined by @xmath40 where @xmath41 is defined by @xmath42 } $ ] with the fourier transformation @xmath43 } = \int \cdot \hspace{0.5mm}e^{-i\bm{k}\cdot \bm{r } } d\bm{r}/v$ ] and @xmath44 .
we discuss the possibility for a kolmogorov spectrum in sst with hydrodynamic equations obtained from eq .
( [ gp ] ) .
these hydrodynamic equations are derived in @xcite , being composed of the equations of the total density @xmath19 , the superfluid velocity @xmath26 , the spin vector @xmath45 , and the nematic tensor @xmath46 with @xmath47/2 $ ] .
we apply the following three approximations to these hydrodynamic equations : ( i ) the macroscopic wave functions are expressed by the fully magnetized state , ( ii ) the total density is almost uniform ( @xmath48 ) , and ( iii ) the magnitude of velocity @xmath26 is much smaller than the density sound velocity @xmath49 .
these similar approximations are discussed in our previous papers @xcite .
the approximation ( i ) leads to the relation between the spin vector and nematic tensor @xcite : @xmath50 thus , by eliminating the nematic tensor in the hydrodynamic equations of @xcite , we obtain the following equations @xmath51 @xmath52 @xmath53 with the approximations ( ii ) and ( iii ) , eqs .
( [ 1 ] ) - ( [ 3 ] ) become @xmath54 , \label{spin_motion}\end{aligned}\ ] ] @xmath55 in eq .
( [ velocity_motion ] ) , the inertial term @xmath56 is smaller than the other terms @xcite , but we retain this term for the following explanation .
we apply a kolmogorov - type dimensional scaling analysis @xcite to eqs .
( [ spin_motion ] ) and ( [ velocity_motion ] ) , obtaining the @xmath0 power law .
we consider the scale transformation @xmath57 and @xmath58 .
then , if @xmath59 and @xmath60 are transformed to @xmath61 and @xmath62 , eqs .
( [ spin_motion ] ) and ( [ velocity_motion ] ) are invariant .
thus , the velocity field satisfies @xmath63 with a nondimensional coefficient @xmath64
. then the spatial and temporal dependence of @xmath65 is same as that of @xmath26 , which leads to @xmath66 with @xmath67 .
thus , in sst , the spectrum of the superflow kinetic energy can be determined by the kinetic energy flux @xmath68 and the coefficient @xmath69 , which , by using a kolmogorov - type dimensional analysis , leads to @xmath70 the coefficient @xmath71 is nondimensional , which corresponds to the kolmogorov constant in ct .
therefore , the spectrum of the kinetic energy of sst can obey the kolmogorov spectrum . applying a similar analysis to the spin field
, we obtain the relation @xmath72 with a dimensional coefficient @xmath73 , which leads to the @xmath1 power law given by @xmath74 with the spin - dependent interaction energy flux @xmath75 and a dimensional coefficient @xmath76 .
this was discussed in the previous study @xcite .
we note that this @xmath0 power law in sst is much different from that in ct . in three ( two)-dimensional ct
, there is the direct ( inverse ) energy cascade , where the @xmath0 power law is generated by the inertial term @xmath56 in the navier - stokes equation @xcite . on the contrary , in sst , the spatial gradient of the spin vector in eq .
( [ velocity_motion ] ) leads to the @xmath0 power law because , in eq .
( [ velocity_motion ] ) , the order estimation @xcite finds that the inertial term is smaller than the nonlinear spin term in the range @xmath77 with the density coherence length @xmath78 .
this suggests that the mechanism responsible for the @xmath0 power law in sst should be different from that in ct . finally , the assumptions used in the derivation of the @xmath0 and @xmath1 power laws are discussed .
we use five assumpotions , which are ( i ) the macroscopic wave functions are expressed by the fully magnetized state , ( ii ) the total density is almost uniform ( @xmath48 ) , ( iii ) the magnitude of velocity @xmath26 is much smaller than the density sound velocity @xmath79 , ( iv ) the spin and velocity fields are scale invariant , and ( v ) the energy flux is independent of the wave number .
it is difficult to theoretically confim the validity of these assumptions .
then , we consider that the validity may be indirectly confirmed if the @xmath0 and @xmath1 power laws based on these assumptions appear in the numerical result . in the following ,
we show our numerical method and results to confirm these theoretical considerations .
[ b ] components and the rotations for the velocity and spin fields in sst at @xmath80 . ( a ) and ( b ) show the @xmath8 components for the velocity @xmath26 and spin @xmath81 fields , and ( c ) and ( d ) show the @xmath10 component of their rotation .
the system size is @xmath82.,width=321 ] [ t ] in our numerical calculation , we introduce a phenomenological small - scale energy dissipation term into eq . ( [ gp ] ) for the following reasons . in three - dimensional ct
, an energy cascade from low to high wave number is assumed , and the energy in the high - wave - number region is considered to dissipate @xcite . unless this kind of dissipation takes place , energy accumulates in the high - wave - number region , which can break the power law in the spectrum . based on this description of the energy cascade in three - dimensional ct , in the previous study of qt in the one - component becs @xcite
, a phenomenological small - scale dissipation was added to the one - component gp equation given by @xmath83 @xmath84},\end{aligned}\ ] ] where the macroscopic wave function and the interaction coefficient for the one - component bec are denoted by @xmath85 and @xmath86 , respectively .
@xmath87 is the fourier component @xmath88 } $ ] of this wave function .
the function @xmath89 is defined by @xmath90 with the step function @xmath91 and the strength of dissipation @xmath92 , which dissipates the energy in the high - wave - number region larger than the wave number @xmath93 corresponding to the coherence length in the one - component gp equation . also , this dissipation reduces the particle number , so that , in the study of ref .
@xcite , the chemical potential is adjusted for it s conservation .
thus the chemical potential @xmath94 has a time dependence .
we introduce a similar phenomenological dissipation term into the spin-1 spinor gp equations ( [ gp ] ) . in a spin-1 spinor bec
, there are two characteristic lengths : the density coherence length @xmath95 and the spin coherence length @xmath96 , which are defined by @xmath97 and @xmath98 . in the usual experiments
, @xmath99 is larger than unity , which leads to the condition @xmath100 .
the size of the spin structure such as the spin domain wall and the spin vortex is on the order of @xmath101 . with reference to qt in the one - component bec
, we expect that the energy dissipation occurs for wave numbers greater than the wave number @xmath102 corresponding to the spin coherence length @xmath103 . therefore , we add a phenomenological small - scale energy dissipation to eq .
( [ gp ] ) , which is given by @xmath104 @xmath105 } , \label{dgpk2}\end{aligned}\ ] ] where @xmath106 and @xmath107 is defined by @xmath108}$ ] and @xmath109 . in the previous study ,
the chemical potential has a time dependence in order to conserve the total particle number .
however , in our calculation , we do not adjust the chemical potential @xmath110 , because the total particle number hardly decreases in sst .
we describe the parameters and the initial state in our numerical calculation of eqs .
( [ dgpk1 ] ) and ( [ dgpk2 ] ) .
all our numerical results are obtained in a two - dimensional system , whose size is @xmath82 . to generate sst ,
we prepare the dynamically unstable state as the initial state , which is the counterflow state @xcite .
this state is given by @xmath111 with a relative velocity @xmath112 . in our numerical calculation
, we set @xmath113 . the dependence of numerical result on @xmath112 is discussed in sec .
vi a. the usual experimental ratio @xmath114 for @xmath115 is about @xmath116 @xcite , but , in our numerical calculation , we set @xmath117 with a positive @xmath11 and negative @xmath12 to grow the dynamical instability quickly .
we use the strength of the dissipation @xmath118 and the chemical potential @xmath119 . in the initial state , we add a small white noise contribution to cause the counterflow instability .
figures [ fig1](a ) and 1(b ) show the distribution of the @xmath8 components of the velocity and spin fields in sst at @xmath120 with @xmath121 . through the counterflow instability
, these two fields are disturbed .
the @xmath10 components of the rotation for these two fields are shown in figs . [ fig1](c ) and 1(d ) , where the rotations are also disturbed @xcite . in a spin-1 spinor bec , the superfluid velocity is related to the spin field through the mermin - ho relation , so that the vortical field can be continuous , which is much different from the one - component bec @xcite . actually , as seen in fig . [
fig1](c ) , the vortical field @xmath122_{z } $ ] has a smooth spatial dependence .
the time development of the spectrum of the spin - dependent interaction and the superflow kinetic energy is shown in fig .
[ fig2 ] . in the early stage of the instability , as shown in fig .
[ fig2](a ) , these spectra have a peak corresponding to the most unstable wave number @xmath123 for the counterflow instability , which was theoretically obtained by using the bogoliubov - de gennes equations for this initial state @xcite .
this wave number @xmath124 is the energy injection scale .
as time progresses , the energy is transferred from low to high wave number , as seen in fig . [ fig2](b ) .
after time @xmath125 , the spectra of the velocity and the spin - dependent interaction energy exhibit the @xmath0 and @xmath1 power laws in figs .
[ fig2](c ) and ( d ) . in fig .
[ fig2 ] ( d ) , the range @xmath126 of these scaling laws is not so wide , but the scaling behavior is consistent with our consideration in sec .
therefore , we conclude that our numerical result confirms the -5/3 and @xmath1 power laws
. we comment on the scaling range .
this narrow scaling range may come from how to excite to the initial state . as shown in fig .
[ fig2](a ) , the energy is injected at the wave number @xmath127 , which roughly determines the lower limit of the scaling range .
therefore , if we use another method with lower @xmath124 for generating sst , the scaling range may be wider . in our calculation
, the @xmath0 power law applies in the range @xmath128 . for @xmath129 , we confirm that the spectrum of the kinetic energy begins to deviate from the @xmath0 power law as shown in fig .
3 ; this is caused by the shortage of energy in the low - wave - number region and the energy accumulation near @xmath102 .
on the other hand , the @xmath1 power law in the spectrum of spin - dependent interaction energy sustains even for @xmath129 .
as time sufficiently passes , the spectrum in low wave number region grows , and this spectrum has a configuration similar to that in our previous study @xcite .
this growth is discussed in sec .
b. finally , we comment on what happens if the energy dissipation is absent .
we perform numerical calculation without the dissipation , which shows two following behaviors : ( i ) the spectrum of superflow kinetic energy still shows the @xmath0 power law , but the time period sustaining this power law becomes shorter , and ( ii ) the fluctuation of the spectrum is larger .
these similar behaviors were observed in the previous study for qt of one - component bec too @xcite .
[ t ] ( a ) 800 and ( b ) 1500 . the dotted lines in the upper and lower graphs are proportional to @xmath130 and @xmath131 , respectively.,width=321 ] we now consider what structure of velocity field leads to the kolmogorov @xmath0 power law in sst . in qt ,
the vortical velocity field seems to be important for the kolmogorov spectrum .
then , to investigate the vortical flow of @xmath132 in sst , we decompose the vector @xmath133 into incompressible @xmath134 and compressible @xmath135 parts @xcite .
the helmholtz theorem leads to @xmath136 , where the relations @xmath137 and @xmath138 are satisfied .
thus , the superflow kinetic energy per unit mass is expressed by @xmath139 with @xmath140 ( @xmath141 ) . using the fourier series @xmath142 ( @xmath141 )
, we can define the spectra for each kinetic energy per unit mass as @xmath143 we calculate the time dependence of @xmath144 in fig .
[ fig4](a ) ; one can see that the incompressible superflow kinetic energy is much larger than the compressible one .
this can be caused by ( i ) the condition @xmath145 under which the total density is hard to disturb , and ( ii ) the dissipation , which prevents the total density modulation from accumulating . figure [ fig4](b ) shows the spectrum @xmath146 of the compressible kinetic energy at @xmath147 , which deviates from the @xmath0 power law in comparison with @xmath148 in fig . [
fig2](d ) .
therefore , the vortical structure of @xmath133 is significant for the kolmogorov spectrum in sst , which is similar to the situation of qt .
[ t ] ( @xmath149 ) and ( b ) spectrum for compressible superflow kinetic energy at @xmath150 . the dotted line in ( b ) is proportional to @xmath131.,width=321 ] we numerically confirm that the @xmath0 power law in sst is different from that in ct . in sec .
iii , we point out that the @xmath0 power law in sst originates from the nonlinear spin term of eq .
( [ velocity_motion ] ) .
the magnitude of the inertial and nonlinear spin terms is dependent on the scale , so that we perform the fourier transform for these two terms , comparing the magnitude of the fourier components in order to confirm the argument of sec .
specifically , we numerically calculate the following quantities : @xmath151 @xmath152 where @xmath153 and @xmath154 are defined by @xmath155 , \end{aligned}\ ] ] @xmath156.\end{aligned}\ ] ] figure 5 shows the wave number dependence of @xmath157 and @xmath158 .
one sees that the nonlinear spin term @xmath159 is larger than the inertial term @xmath160 , and these ratios are about @xmath161 in the scaling range @xmath126 of fig . 2 where the @xmath0 power law appears .
thus we can consider that the kolmogorov spectrum in sst is generated by the nonlinear spin term .
[ t ] ( @xmath162 ) at @xmath163 ( a ) 500 and ( b ) 700 .
, width=321 ]
we now discuss the dependence of the spectra of the superflow kinetic and spin - dependent interaction energy on @xmath112 .
when the relative velocity @xmath112 is large , the period sustaining the kolmogorov spectrum in the superflow kinetic energy is numerically confirmed to become short .
this can be caused by the following two reasons .
one is ( i ) modulation of total density , and the other is ( ii ) high energy injection wave number . as for ( i ) , when the relative velocity is large , the total density is easy to fluctuate , which leads to the growth of the compressible velocity field .
actually , we numerically confirm the increase of the compressible kinetic energy . in our derivation of the power law ,
the uniformity of the total density is assumed , so that this increase shortens the time period sustaining the kolmogorov spectrum . as for ( ii ) , the most unstable wave number becomes high when the relative velocity is large @xcite , which leads to the high injection wave number @xmath124 . as discussed in sec . v. a
, this generates the narrow scaling range , which disturbs the appearance of the power law . because of the above reasons , in our calculation , we set @xmath113 . on the other hand , in the spectrum of spin - dependent interaction energy , the @xmath1 power law can appear even under the condition of large relative velocity .
this was confirmed in our previous study @xcite . at present , we do not understood why this spectrum can exhibit the @xmath1 power law independently of the relative velocity . in this paper
, we perform the two - dimensional numerical calculation for sst , so that the possibility of inverse cascade in sst is discussed . as briefly described in sec .
iii , in three - dimensional ct , the kinetic energy is transported from low wave numbers to high ones , which is called direct energy cascade . however , in two - dimensional ct , the energy inverse cascade occurs , where the kinetic energy is transported from high wave numbers from low ones .
this inverse cascade is caused by the conservation of ensthropy as well as that of kinetic energy in two - dimensional fluid system @xcite . at present , we consider that the inverse energy cascade for superflow kinetic energy does not occur in the two - dimensional sst because the enstrophy in ferromagnetic spin-1 spinor bec is not conserved . actually , our numerical calculation does not exhibit the sign of the energy inverse cascade in the superflow kinetic energy .
on the other hand , as for the spin - dependent interaction energy , the inverse energy cascade may occur because our previous study and fig .
[ fig3 ] seem to exhibit the growth of spectrum of spin - dependent interaction energy in low wave number region as the time sufficiently passes .
this growth may be caused by the ferromagnetic interaction since it tends to align the spin density vector and make the spin domain .
however , as shown in fig .
[ fig2 ] , the spectrum apparently seems to exhibit the sign of direct energy cascade .
thus , at the moment , we does not sufficiently understand the direction of energy cascade for the spin - dependent interaction energy .
we theoretically and numerically studied sst in a spin-1 ferromagnetic spinor bec at zero temperature by using the spin-1 spinor gp equations , finding that both the @xmath0 and @xmath1 power laws appear in the spectrum of the superflow kinetic and the spin - dependent interaction energy .
first , we discussed the possibility for the kolmogorov spectrum in sst , pointing out that this spectrum can be generated by the nonlinear spin term .
second , we showed the numerical results of the spin-1 spinor gp equation with the phenomenological small - scale energy dissipation , whereby sst in the two - dimensional system was obtained by the counterflow instability .
our numerical results indicated that both the @xmath0 and @xmath1 power laws appeared .
furthermore , we estimated the magnitude of the inertial term and the nonlinear spin term of eq .
( [ velocity_motion ] ) in the wave number space , numerically confirming that the kolmogorov spectrum in sst can be generated by the latter term .
k. f. was supported by a grant - in - aid for jsps fellows grant number 262524 .
m. t. was supported by jsps kakenhi grant number 26400366 and mext kakenhi `` fluctuation @xmath164 structure '' grant number 26103526 .
99 p. a. davidson , _ turbulence : an introduction for scientists and engineers _ ( oxford university press , oxford , 2004 ) .
u. frisch , _ turbulence _
( cambridge university press , cambridge , 1995 ) . , edited by w. p. halperin and m. tsubota ( elsevier , amsterdam , 2008 ) , vol .
w. f. vinen , j. low temp . phys . * 161 * , 419 ( 2010 ) . l. skrbek and k. r. sreenivasan , phys .
fluids * 24 * , 011301 ( 2012 ) . c. nore , m. abid , and m. e. brachet , phys .
* 78 * , 3896 ( 1997 ) ; phys .
fluids * 9 * , 2644 ( 1997 ) .
n. g. parker and c. s. adams , phys .
rev . lett . * 95 * , 145301 ( 2005 ) .
m. kobayashi and m. tsubota , phys .
lett . * 94 * , 065302 ( 2005 ) ; j. phys .
. jpn . * 74 * , 3248 ( 2005 ) .
m. kobayashi and m. tsubota , phys .
a * 76 * , 045603 ( 2007 ) .
horng , c .- h .
hsueh , s .- w .
su , y .- m .
kao , and s .- c .
gou , phys .
a * 80 * , 023618 ( 2009 ) .
e. a. l. henn , j. a. seman , g. roati , k. m. f. magalhaes , and v. s. bagnato , phys .
* 103 * , 045301 ( 2009 ) .
a. c. white , c. f. barenghi , n. p. proukakis , a. j. you d , and d. h. wacks , phys .
lett . * 104 * , 075301 ( 2010 ) .
b. nowak , j. schole , d. sexty , and t. gasenzer , phys .
a * 85 * , 043627 ( 2012 ) .
m. t. reeves , b. p. anderson , and a. s. bradley , phys .
a * 86 * , 053621 ( 2012 ) .
k. e. wilson , e. c. samson , z. l. newman , t. w. neely , and b. p. anderson , annu .
* 1 * , 261 ( 2013 ) .
h. takeuchi , s. ishino , and m. tsubota , phys .
lett . * 105 * , 205301 ( 2010 ) .
k. fujimoto and m. tsubota , phys .
a * 85 * , 033642 ( 2012 ) .
k. fujimoto and m. tsubota , phys .
a * 88 * , 063628 ( 2013 ) .
m. karl , b. nowak , and t. gasenzer , sci .
rep . * 3 * , 2394 ( 2013 ) .
m. karl , b. nowak , and t. gasenzer , phys .
a * 88 * , 063615 ( 2013 ) .
d. kobyakov , a. bezett , e. lundh , m. marklund , and v. bychkov , phys .
rev . a * 89 * , 013631 ( 2014 ) .
b. villase@xmath165or , r. zamora - zamora , d. bernal , and v. romero - roch@xmath166n , phys .
rev . a * 89 * , 033611 ( 2014 ) .
k. kasamatsu , m. tsubota , and m. ueda , int . j. mod
b * 19 * , 1835 ( 2005 ) .
y. kawaguchi and m. ueda , phys . rep . *
520 * , 253 ( 2013 ) .
d. m. stamper - kurn and m. ueda , rev .
* 85 * , 1191 ( 2013 )
. t. ohmi and k. machida , j. phys .
. jpn . * 67 * , 1822 ( 1998 ) .
ho , phys . rev .
lett . * 81 * , 742 ( 1998 ) .
e. yukawa and m. ueda , phys .
a * 86 * , 063614 ( 2012 ) . as shown in fig .
1 , the velocity is smaller than @xmath79 , whose order is about @xmath167 .
the magnitude @xmath168 of the spin vector is of the order of unity because of the ferromagnetic interaction .
then , by using the nondimensional form of eq .
( [ velocity_motion ] ) , we can estimate the order of each term .
m. ottaviani and j. a. krommes , phys .
* 69 * , 2923 ( 1992 ) .
t. watanabe , h. fujisaka , and t. iwayama , phys .
e * 55 * , 5575 ( 1997 ) .
in contrast to a one - component bec , the superfluid velocity and the vorticity in fig .
( [ fig1 ] ) do not diverge .
this is because the vortex core is filled with other components .
however , if the instability is much stronger , the velocity may diverge since the other components may escape from the vortex core .
e. g. m. van kempen , s. j. j. m. f. kokkelmans , d. j. heinzen , b. j. verhaar , phys .
lett . * 88 * , 093201 ( 2002 ) . | spin - superflow turbulence ( sst ) in spin-1 ferromagnetic spinor bose - einstein condensates is theoretically and numerically studied by using the spin-1 spinor gross - pitaevskii ( gp ) equations .
sst is turbulence in which the disturbed spin and superfluid velocity fields are coupled . applying the kolmogorov - type dimensional scaling analysis to the hydrodynamic equations of spin and velocity fields
, we theoretically find that the @xmath0 and @xmath1 power laws can appear in spectra of the superflow kinetic and the spin - dependent interaction energy , respectively .
our numerical calculation of the gp equations with a phenomenological small - scale energy dissipation confirms sst with the coexistence of disturbed spin and superfluid velocity field with two power laws . |
the multiple sequence alignment ( msa ) problem constitutes one of the fundamental research areas in bioinformatics . while at first sight it may seem a simple extension of the two - string alignment problem _ two strings good , four strings better _
, for biologists , the multiple alignment of proteins or dna is crucial in deducing their common properties @xcite . quoting arthur lensk @xcite : _ one or two homologous sequences whisper ... a full multiple alignment shouts out loud_. in general , the sequences consist of a linear array of symbols from an alphabet of @xmath3-letters ( @xmath4 for dna and @xmath5 for proteins ) . given @xmath0 sequences to determine a good multiple sequence alignment is a relative task .
usually one defines a score function that depends on the distances between the letters of the alphabet , and assumes that the better alignment is the one that minimizes this score function .
it is a common use to define the msa score in terms of the scores of the pairwise global alignments of the sequences ( sum of pairs score)@xcite . given two sequences @xmath6 and @xmath7 let @xmath8 be a cost of the mutation of @xmath9 into @xmath10 and @xmath11 the cost of inserting or deleting of the letter . extending @xmath8 so that @xmath12 and @xmath13 and considering that a null ( - ) symbol isolated from others ( - ) pays an extra cost @xmath14 @xcite we may define the score of a pairwise alignment @xmath15 for sequences @xmath16 and @xmath17 of size @xmath18 as : @xmath19 where @xmath20 is the number of isolated ( - ) .
then , the score for the multiple alignment is given by : @xmath21 the multiple sequence alignment has at least three important applications in biology : classification of protein families and superfamilies , the identification and representation of conserved sequences features of both dna or proteins that correlate structure and/or function and the deduction of the evolutionary history of the sequences studied @xcite .
unfortunately the problem is known to be _ np - complete _ and no complete algorithm exist to solve real or random instances .
therefore , many heuristic algorithms have been proposed to solve this problem .
the algorithm of carrillo - lipman @xcite ( which is complete ) , is a dynamic programming algorithm able to find the multiple alignment of 3 sequences , and with some heuristic added , to find the alignments , in reasonable time , of up to 6 sequences @xcite .
however , its computational cost scales very fast with the number of sequences and is of little utility for more ambitious tasks . in the first 90 s
the problem was approached using ideas coming from physics , j. kim and collaborators @xcite and m. ishikawa and collaborators @xcite used different version of the simulated annealing @xcite technique with some success , but their algorithms were unable to change the number of gaps in the alignment .
this means that once they started with a given initial configuration ( usually taken from some heuristics ) , any motion of segments in the sequences conserved the number of gaps . to extend these programs allowing the number of gaps to change
will cause the appearance of global moves in the algorithm that are very expensive from the computational point of view .
probably the must successful attempt to solve this problem has been the clustal project @xcite , a progressive algorithm that first organizes the sequences according to their distances and then aligns the sequences in a progressive way , starting with the most related ones .
moreover , it uses a lot of biological information , some motifs of residues rarely accept gaps , sub - sequences of residues associated with structural sub - units are preferred to stay together during the alignment , etc .
these features , and a platform easy to use and integrated with other standard bioinformatic tools , have made clustal the favorite multiple sequence alignment program for biologists and people doing bionformatics in general @xcite .
however , it also has important drawbacks .
once the first @xmath3 sequences are aligned , the inclusion of a new sequence would not change the previous alignment , the gap penalties are the same independently on how many sequences have been already aligned or their properties , and being a progressive method the global minimum obtained is strongly biased by those sequences which are more similar@xcite .
another , recent and also successful approach uses the concepts of hidden markov models @xcite . while some of the previous drawback associated to clustal disappear , because for example , the sequences do not need to be organized a priori , one most start assuming a known model of protein ( or dna ) organization , which is usually obtained after training the program in a subset of sequences . then
, one must be aware that the results usually depend on the training set , specially if it is not too large .
moreover , if we are dealing with sequences of unknown family , or difficult to be characterized this approach does not guarantee good alignments .
therefore we decided to propose a new simulated annealing ( sa ) algorithm that avoids the main difficulty of the previous attempts @xcite .
our algorithm allows for the insertion and deletion of gaps in a dynamic way using only local moves .
it makes use of the mathematical mapping between the multiple sequence alignment and a directed polymer in a random media ( dprm ) pointed out some years ago by hwa et al @xcite .
in such a way , it should be also possible to extrapolate all the computer facilities and techniques , developed in the field of polymers to this biological problem .
the rest of the paper is organized in the following way . in the next section
we make a short review of the theoretical foundations of our algorithm .
then in section [ sec : alg ] we explain the implementation details to discuss the results in section [ sec : res ] .
finally the conclusions are presented including and outlook for future improvements of this program .
usually , multiple sequence alignments are studied and visualized writing one sequence on the top of the other , miming a table , ( see figure [ fig : usual_alignment ] ) and all the probabilistic algorithms devised so far use the simplicity of this representation to generate the moves . instead of that , we will use the well known fact @xcite that the alignment of @xmath0 sequences may be represented in a @xmath0 dimensional lattice ( see figure [ fig : square ] for @xmath22 ) .
the cells of the @xmath0-dimensional lattice are labeled by the @xmath0 indexes @xmath23 .
the bonds encode the adjacency of letters : a diagonal bond in a @xmath0 dimensional space represents the @xmath0-pairing @xmath24 .
the insertion of gaps are represented by bonds without components in the sequences where the gaps were inserted .
for example , a @xmath0-pairing @xmath25 is represented by a bond whose projection on the third sequence is zero , and the @xmath0-pairing @xmath26 is represented by a bond whose projection on the second and third sequences are zero .
then , any alignment maps onto a lattice path that is directed along the diagonal of the @xmath0-dimensional hypercube .
this lattice path may be interpreted as a directed polymer and the random media in the problem is provided by the structure of the sequences to be aligned and by the distance between the residues in the different sequences .
this mapping was already fruitfully used by hwa and co.@xcite to prove that the similarities between two sequences can be detected only if their amount exceeds a threshold value and for proposing a dynamic way to determine the optimal parameters for a good alignment of two sequences . here , our main focus will be to optimize the directed polymer ( lattice path ) under the constraints imposed by the sequences and their interactions in dimensions larger than 2 . to use a simulated annealing algorithm we extend the usual representation of computer science of determining the ground state of the problem to a finite temperature description .
then , a finite - temperature alignment is a probability distribution @xmath27 over all possible conformations @xmath28 of the polymer and where @xmath29 is given by equation [ eq : cost_function ] , and @xmath30 is the partition function of the alignment @xcite .
the temperature ( @xmath31 ) controls the relative weight of alignments with different scores ( different conformations of the polymer ) while @xmath8 the length of the polymer and the frequency of the gaps . in physical terms
, @xmath32 defines and ensemble at temperature @xmath31 with line tension @xmath11 and chemical potential @xmath8.@xcite
the simulated annealing ( sa ) was introduced many years ago by kirkpatrick et al @xcite to find a global minimum of a function in combinatorial optimization problems .
sa is a probabilistic approach , that in general needs a state space ( the different configurations of the directed path ) and a cost or energy function ( [ eq : cost_function ] ) to be minimize ( eq . [ eq : cost_function ] ) .
simulated annealing generates new alignments from a current alignment by applying transition rules of acceptance . the criteria for acceptance
are the following : * if @xmath33 , accept the new alignment * if @xmath34 accept it with probability @xmath35 the parameter @xmath36 controls the probability to accept a new configuration .
initially , one starts at low values of @xmath36 ( high temperatures ) and then increases it applying an annealing schedule .
if the temperature is lowered slowly enough , it can be proved that the system reaches a global minimum @xcite .
unfortunately it will require infinitely computational time and one usually selects the best scheduling that is possible to afford with the computational facilities at hand .
then sa is run over many initial conditions , and one assumes that the output of minimum energy is ( or is close ) to the global minimum . in the multiple sequence alignment
this is also the case , but differently to what happens in other combinatorial optimization problems , here the average solution , i.e. that obtained after averaging over all the local minima s may be interesting by itself .
in fact , researchers are often interested not in the particular details of the alignment , but in its robust properties , and comparing all the outputs of the sa is a way to get this information . from the technical point of view
, once a cost function is defined , one needs to select the moves to be associated to the transition rates . our description of the multiple sequence alignment problem as a directed polymer in a random media allows us to define three types of moves , insertion , deletion and motion of gaps .
all these moves are represented in figure ( [ fig : local_moves ] ) in a two - dimensional grid .
the extension to @xmath0-dimensional systems is straightforward . in this way
we get an algorithm that allows for the creation of gaps , which means a search space larger than the usually proved by similar methods . at the same time the algorithm is quadratic in the number of sequences .
in fact , the computational cost of any move is limited by the square of the number of sequences to be aligned . in this work
, we did not follow any heuristic strategy of optimization .
our intention was to prove the potentiality of this strategy and we kept things as simple as possible .
for example , if we start too far from the global minimum , the selection of local moves alone will make the algorithm to converge very slowly to it .
this drawback may be overcome using very different initial conditions or trying , every some time steps , global moves that change radically the conformation of the polymer .
we did not take care of this . during the simulation
the three moves were chosen with probability @xmath37 .
the only biological information inserted was given by the cost matrix used to align the protein sequences .
we avoid the use of important and well know biological information , fixed residues , phylogenetic tree of the sequences , etc , and the program was designed without the use programming optimization tricks .
all the results presented in this section reflects the alignments of proteins from the kinase family , but qualitatively similar results were obtained for the gpcrs ( g protein - coupled receptors ) and crp ( camp receptor protein ) families .
the simulations started with @xmath38 and every @xmath39 montecarlo steps @xmath36 was increased by a multiplicative factor of 1.01 until @xmath40 .
we took care and in all cases the system reached the equilibrium .
the different initial conditions were chosen inserting gaps randomly in all the sequences but the larger one , such that considering these gaps at @xmath41 , all the sequences were of equal length . to define the distance between the letters of the alphabet we use the pam matrix @xcite . in figure
[ fig : relax3d ] are shown the approach to equilibrium of the multiple sequence alignment of 3 sequences averaged over 100 initial conditions for @xmath42 and @xmath43 . comparing with the result of the carrillo - lipman algorithm it is evident that for @xmath44 the algorithm is very close to the global minimum .
however , one should take care that differently to what is usually obtained in other algorithms for the multiple sequence alignment problem , these figures reflects averages values of the multiple sequence alignment . in fact , one is often interested , rather than on the average , in the minimum of all the alignments . in figure
[ fig : hist0 - 3d-123 ] we represent an histogram in energy for the alignment , over 1000 initial conditions for different values of @xmath39 , of the same three proteins of the kinase family presented in figure [ fig : relax3d ] .
note again that for @xmath45 we obtain exactly the carillo - lipman result with probability larger than 0 .
moreover , looking to the structure of the histogram for @xmath45 , one can also conjecture that if the average multiple sequence alignment were calculated only with those alignments concentrated in the peak of lower energies , the result presented in figure [ fig : relax3d ] will be closer to the one of the carrillo - lipman algorithm .
another symptom suggesting that the average over the realization must be taken with care comes from the analysis of figure [ fig : hist-3d-1 - 9 ] .
there we present again histograms for @xmath45 but using three different samples of the kinase family .
note that while sample 1 and sample 3 are very well behaved and the results compare very well with the carillo - lipman method , the situation for sample 2 is different . to get good results in this case , it is clearly necessary to go beyond 1000 montecarlo steps . in the same spirit of figure [ fig : relax3d ] ,
figures [ fig : relax9d ] and [ fig : relax18d ] reflect results suggesting that also for higher dimensions , if @xmath39 is large enough the algorithm should produce good alignments .
in fact , also in these cases the energy decreases linearly with @xmath39 . if the number of sequences is higher , the correlations between the sequences increases , and the algorithm should find better results .
this fact may be clearly seen in figure [ fig : equilibritation - e - time ] where the equilibration time of the algorithm ( in m.c.s ) is shown as a function of the number of sequences .
the equilibration time , measured as the time necessary to reduce the energy by a factor @xmath46 , decreases linearly with the number of sequences to be aligned . of course the results may change if very different sequences are aligned , but in this case all other known algorithms fail to predict good alignments .
then , we may say that for the most common cases , of correlated sequences , we present an algorithm whose convergence time decreases with @xmath0 , and whose moves , only increase quadratically with @xmath0 . with these results at hand , we follow to study the structure of the space of the solutions as a function of the number of aligned sequences .
we define a distance ( @xmath47 ) between two alignments @xmath48 , and @xmath49 in the following way .
given two solutions , @xmath50 and @xmath51 ( where the index @xmath52 stands for the sequence and @xmath53 for the position of the symbol in the sequence ) we , aligned one by one of the @xmath0 sequences of each solution using a dynamic programming algorithm reminiscent of the needleman - wunsh algorithm with the following score function : @xmath54 and express @xmath55 as : @xmath56 in this way identical alignments will be at distance @xmath57 from each other , and the insertion of gaps to obtain good alignments is penalized , such that the original alignments are altered minimally during the calculation of @xmath55 .
we calculated @xmath47 with @xmath58 and @xmath59 and the results were the same ( apart a constant shift in @xmath47 ) .
below we present figures for @xmath58 .
we study how different are , from the solution of minimum score , the other solutions obtained aligning @xmath0 sequences . for every value of @xmath0 we use 1000 initial conditions and a @xmath45 .
note that the sequences used were always the same .
this mean , we first aligned three sequences from the kinase family .
then , to align 4 sequences we just add a new one to the previous three and started the alignment from scratch
. the procedure was repeated for every new value of @xmath0 .
the results appear in figures [ fig : clustering - result - down ] and [ fig : clustering - result - up ] where @xmath60 is plotted as a function of @xmath47 . from the figures it becomes evident that the space of solution strongly depends on the number of sequences aligned .
for example , while for three sequences the distance between the alignments is correlated with the difference in score , for more than 4 sequences it is not true anymore .
moreover , while for @xmath61 the distance between the solutions decreases , it increases for @xmath62 and remains constant for @xmath63 .
surely , these results reflect the correlation between the proteins aligned .
for example , we may speculate that for @xmath64 , any new protein added contributed to find more similar alignments , i.e. added relevant information to the system .
however , for @xmath62 the sequences contributed with new and uncorrelated information that produce more distant alignments and for @xmath63 to add new proteins that belong to the kinase family will not change the relevant characteristics of the alignment .
these results are relevant , considering the limitations of progressive algorithms to change the previous alignment when new sequences are added .
figures [ fig : clustering - result - down ] and [ fig : clustering - result - up ] , clearly suggest that the inclusion of one single sequence may dramatically change the character of the solutions .
in this work we presented a new probabilistic algorithm , to perform the multiple alignment of proteins .
the algorithm is based on the mapping between the dprm and the multiple sequence alignment problem . at variance with other probabilistic algorithms
our algorithm permits the variation of the number of gaps in the alignment without the necessity of expensive global moves .
it is proved that for small number of sequences it reproduces the results of a complete algorithm .
moreover , we show that for practical purposes the equilibration time is almost independent on the number of sequences aligned @xmath0 and in the worst case , it scale linearly with @xmath0 . finally , we study the space of solutions for different number of aligned sequences , and find a very rich structure that indicates the importance of just one sequence in the multiple alignment .
of course the algorithm is still far from being competitive with other approaches like hmm and clustal .
we are already working in implementing a similar work but using parallel tempering instead of simulated annealing .
it is know that parallel tempering is more useful that s.a . when dealing with very hard problems , like spin glasses .
moreover , it is very suitable to parallelization . also a direction of current work is the introduction of biological information relevant for the alignment .
this may impose important constraints in the possible alignments , that may in turn strongly reduce the space of possible solutions . and
last , but not least , important programing optimization are necessary to make competitive this program from the computing time point of view .
d. gunsfield , _ algorithms on strings , trees and sequences _ ( 1999 ) cambridge university press .
t. j. p. hubbard , a. m. lensk and a. tramontano , nature structural biology * 4 * , 313 ( 1996 ) m.s .
waterman , _ introduction to computational biology _ ( 200 ) , chapm & hall .
p. pevzner , _ computational molecular biology _ ( 2000 ) mit press . h.
carrillo and d. lipman , siam j. appl .
math , * 48 * , 1073 , ( 1988 ) j. kim , s. pramanik and m. j. chung , cabios * 10 * 419 ( 1994 ) m. ishikawa , t. toya , m. hoshida , k. nitta , a. ogiwara and m. kanehisa , cabios * 9 * 267 , ( 1993 ) s. kirkpatrick , c.d .
gelatt and m. p. vecchi , science * 220 * , 671 ( 1983 ) j. d. thompson , d. g. higgings and t. j. gibson , nucleid acids res . * 22 * , 4673 ( 1994 ) we are not aware of any statistical study proving this , but we believe that this statment is well accepted within the colleagues we contacted .
r. durbin , s. r. eddy , a. krogh and g. mitchison , _ biological sequence analysis _
( 1998 ) cambridge university press .
t. hwa and michael lassig , phys .
* 76 * , 2591 ( 1996 ) m. q. zhang and t. g. marr , j. theor
. biol . * 174 * , 119 ( 1995 ) m. kschischo and m. l@xmath65ssig , preprint s. geman , s. geman , ieee transictions on pattern analysis and machine inteligence , * 6*,721 ( 1984 ) s.b . needleman and c. d. wunsch , j. mol
48 * , 444 ( 1981 ) | we proposed a probabilistic algorithm to solve the multiple sequence alignment problem .
the algorithm is a simulated annealing ( sa ) that exploits the representation of the multiple alignment between @xmath0 sequences as a directed polymer in @xmath0 dimensions . within this representation
we can easily track the evolution in the configuration space of the alignment through local moves of low computational cost . at variance with other probabilistic algorithms
proposed to solve this problem , our approach allows for the creation and deletion of gaps without extra computational cost .
the algorithm was tested aligning proteins from the kinases family . when @xmath1 the results are consistent with those obtained using a complete algorithm .
for @xmath2 where the complete algorithm fails , we show that our algorithm still converges to reasonable alignments .
moreover , we study the space of solutions obtained and show that depending on the number of sequences aligned the solutions are organized in different ways , suggesting a possible source of errors for progressive algorithms . |
= 9.5 t , @xmath0 = 4.2 k , and @xmath5 = -4 v. continuous lines correspond to the edges of the qpc .
the black bar represents 1 @xmath6 m .
( b ) @xmath7-dependence of @xmath8-profiles over the region marked with a dashed line in ( a ) , with @xmath5 = -6 v. using eq .
( [ rosenow ] ) for the two consecutive fringes highlighted with the white dashed lines in ( b ) , we calculate in ( c ) the diameter of the qhi as the tip - qhi distance @xmath9 is varied . ,
width=10 ] our sample is a qpc etched in an ingaas / inalas heterostructure holding a 2des 25 nm below the surface .
the qpc lithographic width is 300 nm .
all the experiments were performed at temperature between 4.2 k and 100 mk , in a dilution refrigerator . here , the perpendicular magnetic field @xmath10 t , which corresponds to a ll filling factor @xmath11 in the 2des .
the sgm experiment is schematically depicted in fig .
[ fig : fig1 ] .
it consists in scanning a metallic atomic force microscope tip , polarized at voltage @xmath5 , along a plane parallel to the 2des at a tip-2des distance of 50 nm while recording a map of the device resistance @xmath8 @xcite .
the qpc resistance is defined as @xmath12 , where @xmath13 and @xmath14 are the voltage and the current through the device , respectively .
the 2des being on a quantized hall plateau , whenever some current tunnels between opposite edge channels , @xmath8 deviates from the zero value expected in qh systems at very low @xmath0 @xcite . in our case , the sgm resistance map recorded at @xmath7 = 9.5 t ,
@xmath5 = -4 v and @xmath0 = 4.2 k and presented in fig . [
fig : fig2](a ) reveals concentric fringes superimposed on a slowly varying background .
the origin of the background , related to reflection of es at the qpc , is discussed in the supplementary information .
the fringe pattern can easily be understood in the presence of a qhi surrounding a potential hill , close to the saddle point of the qpc and tunnel - coupled to the propagating es ( fig .
[ fig : fig1 ] ) . indeed , approaching the polarized tip gradually changes the potential of the qhi , and hence its area @xmath15 , defined as the surface enclosed by the `` looping '' es . the enclosed magnetic flux @xmath16 varies accordingly and the tip generates iso-@xmath16 lines when circling around the qhi . since adding one flux quantum @xmath17 corresponds to trapping one electron per populated ll in the island , cb oscillations are generated whenever @xmath7 or @xmath15 are varied @xcite , thereby producing ab - like oscillations @xcite .
isoresistance lines visible on fig .
[ fig : fig2](a ) are , therefore , iso-@xmath16 lines that are crossed as the tip - island distance is varied @xcite .
consequently , the center of concentric fringes in fig .
[ fig : fig2](a ) indicates the position of the active qhi , which connects opposite propagating edge channels through tunnel junctions ( fig .
[ fig : fig1 ] ) . in the framework of this model , the area of the qhi can be determined thanks to the @xmath7-dependence of ab - like oscillations @xcite : @xmath18 where @xmath19 is the number of completely filled ll in the bulk ( here @xmath20 ) .
the combined effect of moving the tip along the dashed line in fig .
[ fig : fig2](a ) and changing @xmath7 is illustrated in fig .
[ fig : fig2](b ) for @xmath5 = -6 v. along the @xmath7-axis , ab - like oscillations are highlighted with the white dashed lines .
the negatively polarized tip approaching the qhi raises its potential , which increases its area @xmath15 , and hence reduces the magnetic field that separates two resistance peaks @xmath21 .
this is illustrated in fig .
[ fig : fig2](c ) , where we assume that the qhi has a surface equivalent to that of a disk with diameter @xmath22 obtained from eq .
( [ rosenow ] ) : @xmath22 is found to increase from @xmath2365 nm to @xmath2395 nm as the tip - island distance @xmath9 decreases from 1300 nm to 300 nm , respectively .
noteworthy , as expected for coulomb dominated transport in a qh interferometer , increasing @xmath7 is equivalent to applying a more negative @xmath5 , yielding a positive @xmath24 for isoresistance stripes @xcite . since approaching the negatively charged tip
has the same effect as decreasing @xmath5 , fig .
[ fig : fig2](b ) seems consistent with the coulomb dominated transport .
vs @xmath0 obtained from sgm maps with @xmath5 = -1 v ( circles ) and from data in ref .
@xcite ( squares ) .
the dashed line corresponds to a @xmath25 dependence .
the gray region corresponds to an exponential dependence exp@xmath26 with @xmath27 , consistent with magnetoresistance data and edge state velocity estimate along ref .
the solid line corresponds to @xmath28 16.2 k , consistent with the spectroscopy data ( see text ) . , width=377 ] but , surprisingly , the temperature dependence of fringes amplitude ( @xmath29 , measured on sgm maps ) , shown on fig .
[ fig : fig3 ] , reveals a peculiar behaviour : it clearly does not follow the @xmath25 dependence expected in the quantum regime of cb @xcite ( data from ref .
@xcite are presented for comparison in fig .
[ fig : fig3 ] ) . instead , @xmath29
deceases very slowly from 100 mk to 4.2 k. indeed , for coherent transport through a fabry - prot geometry , thermal smearing of interference gives rise to a temperature dependence @xmath30exp@xmath26 in the low temperature regime .
in contrast , for transport processes involving a weakly coupled coulomb island , this form for @xmath31 is expected only for temperatures larger than the charging energy @xcite . in the fabry
prot situation , @xmath32 is linked to the excited states level spacing @xmath33 according to the relation @xmath34 where @xmath35 is the local edge state velocity , related to the gradient of the confining potential . from experimental data measured in a gaas qh fabry - prot interferometer @xcite , one can infer that , in our sample , @xmath36 m / s . given this range for @xmath35 , and @xmath37 80 nm ( from fig .
[ fig : fig2](c ) , taking into account that the @xmath0-dependence data were measured at @xmath38 630 nm ) , we obtain the range of @xmath0-dependence represented as a gray region in fig .
[ fig : fig3 ] , which reproduces quite well the behaviour observed experimentally .
the corresponding range of @xmath39 k is consistent with the low temperature limit and hence with a fabry - prot behaviour .
earlier experiments already evidenced such an exponential decay with temperature , but only in mach - zehnder and ballistic devices , which are known to be coherent @xcite .
as a function of the dc component of @xmath5 and @xmath40 at @xmath41 t and @xmath0=100 mk .
voltage modulation of @xmath5 was set to 50 mv .
( b ) 2d fit of @xmath42 using eq .
( [ checkerboard ] ) .
( c - d ) transresistance vs @xmath40 taken along the red ( c ) and blue ( d ) dashed lines in ( a - b ) .
the circles correspond to the experimental data and the continuous lines to the fit . , width=10 ] however , our main observation confirming the preserved electron phase coherence emerges from the analysis of non - linear transport through the qpc .
scanning gate spectroscopy is realized by positioning the tip right above the qhi , and sweeping both @xmath5 and the dc current @xmath14 through the qpc .
the voltage across our tunnel device , i.e. between propagating edge states , is the hall voltage : @xmath43 @xcite , where @xmath44 is the ( integer ) number of transmitted es at the qpc ( fig .
s1 - supplementary material ) .
the measurement configuration is indeed identical to that used to perform conventional electrical spectroscopy on isolated quantum dots .
if the qhi were weakly tunnel coupled to the propagating edge states , one would expect to observe a `` coulomb diamond '' pattern @xcite .
[ fig : fig4](a ) shows @xmath42 as a function of the dc component of @xmath5 and @xmath40 . instead of coulomb diamonds , the spectroscopy displays a checkerboard pattern of maxima and minima , indicating that both @xmath5 and @xmath40 tune the interference of transiting electrons .
each bias independently adds a phase shift between interfering paths , so that the transresistance is modulated by a product of cosines and an exponential term accounting for a voltage - dependent dephasing induced by electrons injected at an energy @xmath45 @xcite : @xmath46 where @xmath47 is the zero - bias visibility of the oscillations , @xmath48 is the oscillation period induced by @xmath5 , @xmath49 is a constant phase factor , @xmath50 is the oscillation period along the @xmath40 axis , and @xmath51 is directly related to the @xmath40-dependent dephasing rate : @xmath52 @xcite .
@xmath53 varies from 1 to 2 according to ref .
@xcite and was set to 1 as we could not discriminate from fitting the data . as shown in figs .
[ fig : fig4](b - d ) , we obtain an excellent fit of the data in fig .
[ fig : fig4](a ) using eq .
( [ checkerboard ] ) with with a transist time @xmath54 s , and a parameter @xmath51 = 0.2 in the range found in ref .
note that in such a small qhi , @xmath55 turns out to be smaller by at least one order of magnitude than the intrinsic @xmath56 in the same 2des @xcite .
this renders coherent resonant tunnelling through the whole qhi device possible .
to interpret @xmath48 obtained from the fit , one first notes that @xmath8 evolves very similarly when changing either @xmath5 or @xmath7 in the vicinity of @xmath57 9 t ( fig . s2 - supplementary material ) .
therefore , one can convert @xmath48 into an equivalent @xmath21 , through a lever arm @xmath58 0.108 t / v .
hence , @xmath59 v corresponds to @xmath21 = 50 mt for the ab - like oscillations . in that range of @xmath5 , @xmath60 ( fig .
this means that @xmath61 nm , consistent with data in fig .
[ fig : fig2](c ) since @xmath22 is at a maximum when the tip is above the qhi ( @xmath62 ) .
moreover , given the value of @xmath63 = @xmath64 s found in fitting the spectroscopy data , one obtains @xmath65 m / s , within the range of values that was expected from data in ref .
@xcite , and in agreement with the exponential temperature dependence in fig .
[ fig : fig3 ] .
we therefore have a fully consistent picture that explains all magnetoresistance , temperature dependence and spectroscopy data , and shows that tunnelling across the qhi is indeed coherent .
one fundamental question remains : why do we observe two distinct behaviours of transport through apparently similar qh devices , coulomb blockaded transport in our previous work @xcite , and coherent transport in this one ?
the qualitative difference can not be explained by the fact that @xmath22 is smaller than previously examined qhis . @xmath32 and @xmath66 would be reduced proportionally , but not enough to explain the observed @xmath0-dependence and spectroscopy . on the other hand , signs of coherent transport through cb quantum dots were only obtained for symmetric tunnel junctions @xcite that allow resonant tunnelling instead of sequential tunnelling . in that framework , one might thus ascribe the loss of electron coherence in other qhis to an asymmetry of tunnel junctions .
however , a difference in the transmission coefficients @xmath67 of the tunnel barriers may point towards an alternative explanation . in the coherent regime
, we find a rather strong coupling between the qhi and propagating es ( @xmath68 ) , which contrasts with the coulomb blockade regime where @xmath69 @xcite .
a similar trend is observed in transport experiments at @xmath7 = 0 t on carbon nanotubes @xcite : phase coherence is maintained when electrons tunnel through barriers with a large transmission coefficient , so that interference effects can be observed . between edge states .
, width=10 ] up to this point , our analysis is based on the presence of a qhi near the qpc , connected to propagating es on both sides through tunnelling paths ( fig .
[ fig : fig1 ] ) .
however , one could imagine the presence of additional tunnelling paths between propagating es , in the vicinity of the qpc saddle point . the resulting model is presented in fig .
[ fig : fig5](a - b ) .
while the `` green '' paths occur naturally when propagating es and the qhi are close enough , the `` red '' paths may originate from potential anharmonicities ( _ i.e. _ non parabolicity ) on both sides of the saddle point , similar to the fast potential variations suggested in ref .
@xcite . in this model
, transport depends in principle on the various tunnelling probabilities , denoted @xmath70 and @xmath71 in fig .
[ fig : fig5 ] .
however , the presence of the qhi should always induce oscillations in the magnetoresistance and spectroscopy of the qpc , either because it is enclosed in an interferometer , created by the `` red '' paths and propagating es , when @xmath72 , or because tunnelling occurs directly through it ( @xmath73 ) as discussed above ( fig .
[ fig : fig1 ] ) .
therefore , whichever @xmath70 or @xmath71 dominates , transport is still controlled by the flux trapped in the qhi and hence its coulomb charging , so that the analysis developed above to extract parameters from the magnetoresistance and spectroscopy are still valid . in that case ,
i.e. @xmath74 , the amplitude of the fringes leads us to @xmath75 ( for details see supplementary information ) . in summary , we report first evidence for preserved electron phase coherence in tunnelling across a quantum point contact in the quantum hall regime .
we propose a framework that explains all magnetoresistance , temperature dependence and spectroscopy data .
this scenario relies on the presence of a potential hill that generates a quantum hall island near the saddle point of the qpc .
our data therefore provide new signatures of coherent tunnelling in an ultra - small qh device .
* device fabrication and 2des parameters . *
our device is fabricated from a ingaas / inalas heterostructure grown by molecular beam epitaxy where a 2des is confined 25 nm below the surface ( the layer sequence of this heterostructure is detailed in @xcite ) .
the qpc was patterned using e - beam lithography followed by wet etching .
the qpc resistance @xmath8 is measured in a four - probe configuration : a low - frequency ( typically 10 to 20 hz ) oscillating current @xmath14 is driven between contacts 1 and 2 on fig . [
fig : fig1 ] , and @xmath13 is measured between contacts 3 and 4 using a lock - in technique , with @xmath13 across the qpc always less than @xmath76 .
next to the qpc , we patterned a hall bar where we measured a low-@xmath0 electron density and mobility of @xmath77 and @xmath78 , respectively .
* sgm and sgs techniques . *
all the experiments are carried out inside a @xmath79he/@xmath80he dilution refrigerator where a home - made cryogenic atomic force microscope ( afm ) was integrated @xcite .
the afm is based on a quartz tuning fork to which a commercial metallized si cantilever ( model csc17 from mikromasch ) is glued by means of a conductive silver epoxy .
we image the sample topography by imposing a feedback loop on the shift in the tuning fork resonant frequency and using standard dynamic afm mode of operation . after locating the qpc we perform sgm .
it consists of scanning the tip along a plane parallel to the 2des at constant distance of 25 nm from the surface , i.e. 50 nm from the 2des , with a bias voltage @xmath5 applied to the tip and recording simultaneously the device resistance @xmath8 . at the end of a set of sgm experiments , we image the topography of the qpc to ensure that , during that period , the position of the qpc did not change .
the sgs is performed by positioning the afm tip at a fixed position in the vicinity of the qhi and by adding a dc current @xmath14 to the lock - in ac signal between contacts 1 and 2 ( fig .
[ fig : fig1 ] ) . the voltage between edge states @xmath40
is obtained by multiplying the dc current @xmath14 by @xmath81 .
the transresistance @xmath42 is measured with a second lock - in using an ac modulation of @xmath5 .
* acknowledgements : * + the authors are grateful to m. treffkorn and t. hyart for helpful discussions .
f.m . and b.h
are postdoctoral and associate researchers with the belgian frs - fnrs , respectively .
this work has been supported by frfc grants no .
2.4.546.08.f and 2.4503.12 , fnrs grant no .
1.5.044.07.f , by the fsr and arc program `` stresstronics '' , by belspo ( interuniversity attraction pole iap-6/42 ) , and by the pnano 2007 program of the anr ( micatec project ) .
v.b . acknowledges the award of a chaire dexcellence by the nanoscience foundation in grenoble .
* author contributions : * + f. m. , b. h. and s. f. performed the low - temperature sgm experiment ; f. m. , b. h. , v. b. and b. r. analysed the experimental data ; l. d. and x. w. grew the ingaas heterostructure ; b. h. processed the sample ; b. h. , s. f. and f. m. built the low temperature scanning gate microscope ; b. h. , f. m. , s. f. , h. s. , s. h. , m. p. and v. b. contributed to the conception of the experiment ; f.m . , b. h. and v. b. wrote the paper and all authors discussed the results and commented on the manuscript . * additional information : * + supplementary information accompanies this paper .
+ competing financial interests : the authors declare no competing financial interests . 10 beenakker , c. w. j. & van houten , h. quantum transport in semiconductor nanostructures . _ solid state physics _
* 44 * , 1 - 228 ( 1991 ) .
fischer , j. & loss , d. dealing with decoherence .
_ science _ * 324 * , 1277 - 1278 ( 2009 ) .
roulleau , p. _ et al .
_ , direct measurement of the coherence length of edge states in the integer quantum hall regime .
lett . _ * 100 * , 126802 ( 2008 ) .
huckestein , b. , scaling theory of the integer quantum hall effect .
phys . _ * 67 * , 357 - 396 ( 1995 ) .
ilani , s. _ et al . _ the microscopic nature of localization in the quantum hall effect .
_ nature _ * 427 * , 328 - 332 ( 2004 ) .
altimiras , c. _ et al . _
non - equilibrium edge - channel spectroscopy in the integer quantum hall regime .
_ nature phys .
_ * 6 * , 34 - 39 ( 2009 ) .
sim , h. -s . ,
kataoka , m. & ford , c. j. b. electron interactions in an antidot in the integer quantum hall regime .
_ * 456 * , 127 - 165 ( 2008 ) .
goldman , v. j. & su , b. resonant tunnelling in the quantum hall regime : measurement of fractional charge .
_ science _ * 267 * , 1010 - 1012 ( 1995 ) .
maasilta , i. j. & goldman , v. j. energetics of quantum antidot states in the quantum hall regime .
b _ * 57 * , r4273-r4276 ( 1998 ) . topinka , m. a. _ et al . _
imaging coherent electron flow from a quantum point contact .
_ science _ * 289 * , 2323 - 2326 ( 2000 ) .
crook , r. , smith , c. g. , simmons , m. y. & ritchie , d. a. imaging cyclotron orbits and scattering sites in a high - mobility two - dimensional electron gas .
b _ * 62 * , 5174 - 5178 ( 2000 ) .
pioda , a. _ et al .
_ spatially resolved manipulation of single electrons in quantum dots using a scanned probe .
lett . _ * 93 * , 216801 ( 2004 ) .
hackens , b. _ et al . _ imaging and controlling electron transport inside a quantum ring .
_ nature phys . _
* 2 * , 826 - 830 ( 2006 ) .
martins , f. _ et al .
_ imaging electron wave functions inside open quantum rings .
. lett . _ * 99 * , 136807 ( 2007 ) .
pala , m. g. _ et al .
_ local density of states in mesoscopic samples from scanning gate microscopy .
b _ * 77 * , 125310 ( 2008 ) .
pala , m. g. _ et al .
_ scanning gate microscopy of quantum rings : effects of an external magnetic field and of charged defects .
_ nanotechnology _ * 20 * , 264021 ( 2009 ) .
paradiso , n. _ et al .
_ spatially resolved analysis of edge - channel equilibration in quantum hall circuits .
b _ * 83 * , 155305 ( 2011 ) .
hackens , b. _ et al .
_ imaging coulomb islands in a quantum hall interferometer .
_ nature comm . _ * 1 * , 39 ( 2010 ) .
taylor , r. p. _ et al .
_ aharonov - bohm oscillations in the coulomb blockade regime .
lett . _ * 69 * , 1989 - 1992 ( 1992 ) .
kataoka , m. _ et al .
_ detection of coulomb charging around an antidot in the quantum hall regime .
lett . _ * 83 * , 160 - 163 ( 1999 ) .
rosenow , b. & halperin , b. i. influence of interactions on flux and back - gate period of quantum hall interferometers .
lett . _ * 98 * , 106801 ( 2007 ) .
zhang , y. _ et al . _ distinct signatures for coulomb blockade and aharonov - bohm interference in electronic fabry - prot interferometers .
b _ * 79 * , 241304 ( 2009 ) .
van loosdrecht , p. h. m. _ et al .
_ aharonov - bohm effect in a singly connected point contact .
b _ * 38 * , 10162 - 10165 ( 1988 ) .
jain , j.k .
& kivelson , s. model tunneling problems in a high magnetic - field . _
b _ * 37 * , 4111 - 4117 ( 1988 ) .
ji , y. _ et al .
_ an electronic mach - zehnder interferometer .
_ nature _ * 422 * , 415 - 418 ( 2003 ) .
neder , i. , heiblum , m. , levinson , y. , mahalu , d. & umansky , v. unexpected behavior in a two - path electron interferometer .
_ * 96 * , 16804 ( 2006 ) .
sivan , u. , imry , y. & hartzstein , c. aharonov - bohm and quantum hall effects in singly connected quantum dots .
b _ * 39 * , 1242 - 1245 ( 1989 ) .
van wees , b. j. _ et al .
_ observation of zero - dimensional states in a one - dimensional electron interferometer .
_ * 62 * , 2523 - 2526 ( 1989 ) .
mcclure , d. t. _ et al .
_ edge - state velocity and coherence in a quantum hall fabry - prot interferometer .
_ * 103 * , 206806 ( 2009 ) .
yamauchi , y. , _ et al .
_ universality of bias- and temperature - induced dephasing in ballistic electronic interferometers .
b _ * 79 * , 161306(r ) ( 2009 ) .
ofek , n. _ et al .
_ role of interactions in an electronic fabry - prot interferometer operating in the quantum hall effect regime . _ proc .
sci . _ * 107 * , 5276 - 5281 ( 2010 ) .
halperin , b.i .
, stern , a. , neder , i. & rosenow , b. theory of the fabry - prot quantum hall interferometer .
b _ * 83 * , 155440 ( 2011 ) .
bttiker , m. absence of backscattering in the quantum hall effect in multiprobe conductors .
_ phys . rev .
b _ * 38 * , 9375 - 9389 ( 1988 ) .
yacoby , a. , heiblum , m. , mahalu , d. & shtrikman , h. coherence and phase sensitive measurements in a quantum dot .
* 74 * , 4047 - 4050 ( 1995 ) .
kouwenhoven , l. p. _ et al .
_ electron transport in quantum dots . in mesoscopic electron transport .
sohn , l. l. , kouwenhoven , l. p. & schon , g. , ) _ series e : applied sciences ( kluwer academic , dordrecht ) _ , * 345 * , 105 - 214 ( 1997 ) .
roulleau , p. _ et al .
_ d. finite bias visibility of the electronic mach - zehnder interferometer .
b _ * 76 * , 161309(r ) ( 2007 ) .
litvin , l. v. , helzel , a. , tranitz , h. p. , wegscheider , w. & strunk , c. edge - channel interference controlled by landau level filling .
b _ * 78 * , 075303 ( 2008 ) .
martins , f. _ et al . _ scanning gate spectroscopy of transport across a quantum hall nano - island _ new j. phys . _ * 15 * , 013049 ( 2013 ) .
van der wiel , w. g. _ et al .
_ electromagnetic aharonov - bohm effect in a two - dimensional electron gas ring .
b _ * 67 * , 033307 ( 2003 ) .
hackens , b. _ et al . _ dwell - time - limited coherence in open quantum dots .
lett . _ * 94 * , 146802 ( 2005 ) .
van houten , h. , beenakker , c. w. j. & staring , a. a. m. coulomb blockade oscillations in semiconductor nanostructures , ( in single charge tunneling , edited by grabert , h. & devoret , m. h. ) _ nato asi series _ * b294 * ( plenum , new york , 1992 ) .
biercuk , m. j. , ilani , s. , marcus , c. m. & mceuen , p. l. electrical transport in single - wall carbon nanotubes carbon nanotubes , ( in advanced topics in the synthesis , structure , properties and applications , edited by jorio , a. dresselhaus , g. & dresselhaus , m. s. ) * 111 * , 455 - 493 ( springer , berlin , 2008 ) .
s[fig : sup_1 ] allows inferring the origin of the broad background in sgm images . by sweeping @xmath5 below 0 v , the sgm maps measured at 100 mk shown in figs .
s[fig : sup_1](a - c ) reveal concentric fringes marking the presence of a qhi near the saddle point of the qpc .
the diameter of circling fringes increases with decreasing @xmath5 , consistent with the observations in fig .
s [ fig : sup_1 ] .
importantly , the sgm pattern around the qpc exhibits a strong variation that adds to the concentric fringes on figs .
s[fig : sup_1](b - c ) .
the origin of the stronger contrast is found by positioning the tip near the saddle point of the qpc and continuously decreasing @xmath5 ( fig .
s[fig : sup_1](d ) ) .
the main trend shows step - like increases of the device resistance which can be understood by invoking es reflections at the constriction .
decreasing @xmath5 raises the energy of the saddle point and decreases the local filling factor @xmath82 near the constriction . every time @xmath82 passes a half integer value , an es is totally reflected and the device resistance shifts to the next plateau given by @xmath83 @xcite ( brown dashed lines in fig .
s[fig : sup_1](d ) ) , where @xmath44 is the ( integer ) number of transmitted es at the constriction .
the presence of oscillations superimposed on the first plateau at @xmath44 = 5 , similar to those around @xmath5 = 0 v , indicate that the qhi is active even when one es is reflected .
here we summarize the details concerning the determination of the coefficients of transmission across the tunnel barriers defining the qhi in the coherent regime ( we assume here that all barriers have equal transmission coefficients ) . in this work
it was found that the peak - to - peak amplitudes in the coherent regime ( @xmath84 ) were within the following intervals : @xmath85 @xmath86 @xmath87 @xmath88 . assuming that @xmath89 @xcite where @xmath90 is the total transmission through the qhi ( i.e. taking into account the two barriers defining the qhi ) , we conclude that @xmath90 is the interval : @xmath91 . in fig . s [ fig : sup_3 ] we draw the two models considered in the main article . in the following subsections
we deduce the coefficients of transmission of the tunnel barriers for the two different situations presented in fig .
s [ fig : sup_3](a ) and ( b ) .
we first consider the situation represented in fig .
s [ fig : sup_3](a ) where a qhi is located at the centre of a qpc . in this case
we assume that coherence is maintained during the multiple reflections .
for an off - resonance condition and assuming @xmath92 , @xmath67 is given by : @xmath93 @xcite , which implies that @xmath94 .
in the case of an interferometer formed around the saddle point of a qpc , as illustrated in fig .
s [ fig : sup_3](b ) , we compute the reflection coefficient @xmath95 , taking into account interferences between different semiclassical paths for electrons : 1 ) a direct path along the edge state , which does not include transmission through the tunnel barriers , and 2 ) paths including multiple transmissions through the tunnel barriers . the reflection coefficient is then given by @xcite : where @xmath97 is the phase difference accumulated along the two types of trajectories and @xmath98 and @xmath99 are the transmissions at each side of the saddle point . assuming that @xmath100
, we obtain : @xmath75 .
mk , @xmath101 t and @xmath102 = 0 , -2 and -4 v , respectively .
the top black bar represents 1 @xmath6 m .
( d ) @xmath8 vs @xmath102 with the tip positioned near the saddle point of the qpc .
the brown dashed lines indicate the resistance expected for @xmath44 . , width=120 ] and @xmath104 are transmission coefficients of the tunnel barriers between es and the qhi ..
( b ) alternative model for the situation at the qpc : two tunnelling paths ( dotted lines ) on both sides of the saddle point connect counterpropagating edge states and form a closed loop . note that in both cases , only one edge state is represented , for the sake of clarity , width=120 ] | the unique properties of quantum hall devices arise from the ideal one - dimensional edge states that form in a two - dimensional electron system at high magnetic field .
tunnelling between edge states across a quantum point contact ( qpc ) has already revealed rich physics , like fractionally charged excitations , or chiral luttinger liquid .
thanks to scanning gate microscopy , we show that a single qpc can turn into an interferometer for specific potential landscapes .
spectroscopy , magnetic field and temperature dependences of electron transport reveal a quantitatively consistent interferometric behavior of the studied qpc . to explain this unexpected behavior
, we put forward a new model which relies on the presence of a quantum hall island at the centre of the constriction as well as on different tunnelling paths surrounding the island , thereby creating a new type of interferometer .
this work sets the ground for new device concepts based on coherent tunnelling .
+ electron phase coherence is the cornerstone of quantum devices and computation @xcite . in that perspective , quantum hall ( qh ) devices are particularly attractive in view of their large coherence times @xcite .
quantum hall edge states ( es ) formed by landau levels ( ll ) crossing the fermi energy near sample borders are ideal one - dimensional ( 1d ) systems in which scattering vanishes exponentially at low temperature @xmath0 @xcite .
edge state loops surrounding potential hills or wells , referred to as localized states or quantum hall islands ( qhis ) , then form unique zero - dimensional ( 0d ) systems @xcite .
the last few years witnessed great progresses in the transport spectroscopy of model qh localized states created by patterning quantum dots @xcite or antidots @xcite in a two - dimensional electron system ( 2des ) . in parallel
, new tools were developed to probe the microscopic structure of confined electron systems in the qh regime .
in particular , scanning gate microscopy @xcite ( sgm ) makes use of a movable metallic tip , which is voltage - biased , to finely tune the electrons confining potential in its vicinity .
this way , the geometry of propagating edge states and localized states can be modified at will @xcite . very recently , sgm allowed us to locate active qhis in a qh interferometer @xcite .
importantly , it appeared that qhis do not only form around antidots , but potential inhomogeneities also induce qhis in the arms or near the constrictions connecting a quantum ring to source and drain reservoirs @xcite .
therefore , lateral confinement , e.g. in quantum point contacts ( qpcs ) , offers the possibility to connect a qhi to es through tunnel junctions , and thus form a new class of 1d-0d-1d qh devices ( fig .
[ fig : fig1 ] ) . in this case , the 0d island is characterized by a weak coupling ( @xmath1 ) and a large charging energy ( @xmath2 ) ( @xmath3 is the island capacitance ) , which induce coulomb blockade ( cb ) @xcite . in such devices ,
aharonov - bohm ( ab ) like oscillations of the resistance can be explained by coulomb coupling between fully occupied lls and confined states in the qhi @xcite .
it was also suggested that ab oscillations reported on a qpc @xcite could be attributed to tunnelling paths around the saddle point @xcite .
in contrast , transport through qh devices , but in the strong coupling limit ( @xmath4 ) , revealed coherent effects analog to those observed in optical mach - zehnder @xcite or fabry - prot @xcite interferometers . here , we examine an unexplored regime of transport across a qpc where qh edge states are weakly coupled , but phase coherence is preserved .
the sgm tip is used as a nanogate to tune the potential landscape and hence edge states pattern and coupling . at first sight ,
one expects that transport should be driven by tunnelling , and possibly by coulomb blockade if a quantum hall island were mediating transport between edge states ( fig .
[ fig : fig1 ] ) @xcite .
indeed , sgm and magnetoresistance data corroborate with coulomb blockade across a qhi located near the saddle point of the qpc .
however , temperature dependence and scanning gate spectroscopy show clear signatures of quantum interferences .
since , up to now , such interferences were exclusively observed in open qh devices , this observation sets the stage for a new electron transport scenario .
we propose a new model that provides a quantitative interpretation of the data . |
a degree @xmath7 polynomial - like mapping is a degree @xmath7 proper holomorphic map @xmath8 , where @xmath9 and @xmath10 are topological disks and @xmath9 is compactly contained in @xmath11 .
this definition captures the behaviour of a polynomial in a neighbourhood of its filled julia set .
the filled julia set is defined in the polynomial - like case as the set of points which do not escape the domain .
the external class of a polynomial - like map is the ( conjugacy classes of ) the map which encodes the dynamics of the polynomial - like map outside the filled julia set . the external class of a degree
@xmath7 polynomial - like map is a degree @xmath7 real - analytic orientation preserving and strictly expanding self - covering of the unit circle : the expansivity of such a circle map implies that all the periodic points are repelling , and in particular not parabolic . in order to avoid this restriction , in @xcite
we introduce an object , which we call _ parabolic - like mapping _ , to describe the parabolic case .
a parabolic - like mapping is thus similar to a polynomial - like mapping , but with a parabolic external class ; that is to say , the external map has a parabolic fixed point .
a parabolic - like map can be seen as the union of two different dynamical parts : a polynomial - like part and a parabolic one , which are connected by a dividing arc .
[ definitionparlikemap ] * ( parabolic - like maps ) * a parabolic - like map of degree @xmath7 is a 4-tuple ( @xmath12 ) where * @xmath9 and @xmath11 are open subsets of @xmath13 , with @xmath14 and @xmath15 isomorphic to a disc , and @xmath9 not contained in @xmath11 , * @xmath16 is a proper holomorphic map of degree @xmath7 with a parabolic fixed point at @xmath17 of multiplier 1 , * @xmath18 \rightarrow \overline { u}$ ] is an arc with @xmath19 , forward invariant under @xmath20 , @xmath21 on @xmath22 $ ] and on @xmath23 $ ] , and such that @xmath24 @xmath25 ) \subseteq u \setminus u ' , \,\,\,\,\,\,\gamma(\pm 1 ) \in \partial u.\ ] ] it resides in repelling petal(s ) of @xmath26 and it divides @xmath9 and @xmath11 into @xmath27 and @xmath28 respectively , such that @xmath29 ( and @xmath30 ) , @xmath31 is an isomorphism and @xmath32 contains at least one attracting fixed petal of @xmath26 .
we call the arc @xmath33 a _ dividing arc_. in @xcite we extend the theory of polynomial - like maps to parabolic - like maps , and we straighten degree @xmath0 parabolic - like maps to members of the family of quadratic rational maps with a parabolic fixed point of multiplier @xmath34 at infinity and critical points at @xmath34 and @xmath35 , which is @xmath36\,|\,p_a(z)=z+\frac{1}{z}+a\}.\ ] ] more precisely , we prove the following : every degree @xmath0 parabolic - like mapping @xmath37 is hybrid equivalent to a member of the family @xmath2 .
moreover , if @xmath38 is connected , this member is unique .
note that @xmath39= \ { p_a,\,\,p_{-a}\},$ ] since the involution @xmath40 conjugates @xmath41 and @xmath42 , interchanging the roles of the critical points .
we refer to a member of the family @xmath2 as one of the representatives of its class . the family @xmath2 is typically parametrized by @xmath43 , which is the multiplier of the free fixed point @xmath44 of @xmath45
. the connectedness locus of @xmath2 is called @xmath6 . if @xmath46 is a family of degree @xmath0 parabolic - like maps with parameter space @xmath47 , calling @xmath48 the connectedness locus of @xmath49 , by the uniqueness of the straightening we can define a map @xmath50 @xmath51 which associates to each @xmath52 the multiplier of the fixed point @xmath44 of the member @xmath39 $ ] hybrid equivalent to @xmath53 . in this paper
we will prove that if the family @xmath49 is _ analytic _ and _ nice _
( see def .
[ def ] and [ nf ] ) , _ the map @xmath54 extends to a map defined on the whole of @xmath55 ( see [ external ] ) , whose restriction to @xmath48 , under suitable conditions ( see def .
[ properfam ] ) is a ramified covering of @xmath4 ( see thm .
[ bigthm])_. the reason why the map @xmath54 covers @xmath4 , instead of the whole of @xmath6 , resides in the definition of analytic family of parabolic - like mappings , and it will be explained in section [ anfam ] .
[ m ] [ m1 ] as an application , we will show that the connectedness locus of the family @xmath56 ( see fig.[m ] ) presents @xmath0 baby @xmath6 ( see fig .
+ the results in this paper were developed during the author s ph.d .
so the author would like to thank her former advisor , carsten lunde petersen , for suggesting the idea of parabolic - like mapping and for his help , roskilde university and universit paul sabatier for their hospitality , and roskilde university , the anr-08-jcjc-0002 founded by the agence nationale de la recherche and the marie curie rtn 035651-cody for their financial support during her ph.d .
in this section we define an analytic family of parabolic - like maps and its connectedness locus , nice families of parabolic - like maps , and we give an example of nice analytic family of parabolic - like maps .
then we give a review of the straightening theorem , an overview of this paper and we state the main result .
[ def ] let @xmath47 , @xmath57 and let @xmath58 be a family of degree @xmath7 parabolic - like mappings .
set @xmath59 , @xmath60 , @xmath61 , and @xmath62 .
then @xmath49 is a degree @xmath7 analytic family of parabolic - like maps if the following conditions are satisfied : 1 .
@xmath63 , @xmath64 , @xmath65 and @xmath66 are domains in @xmath67 ; 2 .
the map @xmath68 is holomorphic in @xmath69 .
all the parabolic - like maps in the family have the same number of attracting petals in the filled julia set . for all
@xmath70 let us call @xmath71 the parabolic - fixed point of @xmath72 , and let us set @xmath73 , and @xmath74 .
define @xmath75 an analytic family of parabolic - like mappings is _ nice _ if there exists a holomorphic motion of the dividing arcs @xmath76 and there exists a holomorphic motion of the ranges @xmath77 which is a piecewise @xmath21-diffeomorphism with no cusps in @xmath78 ( for every fixed @xmath52 ) , and @xmath79 .
a nice family is basically endowed by definition with a holomorphic motion of a fundamental annulus ( see section [ holmot ] ) .
we did not require analytic families to have these properties , because the concept of parabolic - like map is local .
on the other hand , since all the maps in an analytic family of parabolic - like maps have the same number of attracting petals in its filled julia set , it follows from the holomorphic parameter dependence of fatou coordinates ( see appendix in @xcite ) , that in many cases there is a holomorphic motion of the dividing arcs ( however , in individual cases further detail might be required according to circumstances ) . moreover , since the concept of parabolic - like map is local , in many cases it is not difficult to construct a holomorphic motion of the ranges for an analytic family of parabolic - like mappings .
the definition of analytic family of parabolic - like maps is valid for any degree .
however , since in this paper we are interested in proving that , under suitable conditions , the map @xmath54 defined in the introduction is a ramified covering between @xmath48 and @xmath80 , in the remainder we will restrict our attention to degree @xmath0 nice analytic families of parabolic - like maps .
all the maps of an analytic family of parabolic - like maps have the same number of attracting petals in their filled julia set , and each ( maximal ) attracting petal requires a critical point in its boundary . hence
, if @xmath49 is a degree @xmath0 analytic family of parabolic - like maps , either for each @xmath70 the map @xmath72 has no attracting petals in @xmath81 , or for each @xmath70 the map @xmath72 has exactly one attracting petal in @xmath81 .
consider now the family @xmath2 .
the @xmath82-part of a parabolic - like mapping requires ( at least ) one attracting petal , and for all the members of the family @xmath2 with @xmath83 the parabolic fixed point has parabolic multiplicity @xmath34 .
so a parabolic - like restriction of @xmath45 , with @xmath84 has no attracing petals in the filled julia set . on the other hand
, @xmath85 has a parabolic fixed point of parabolic multiplicity @xmath0 and the julia set of @xmath86 is the common boundary of the immediate parabolic basins , so a parabolic - like restriction of @xmath86 has exactly one attracting petal in its filled julia set .
so , if all the members an analytic family of degree @xmath0 parabolic - like mappings @xmath49 have exactly one attracting petal in the filled julia set , they are all hybrid conjugate to the map @xmath85 , and @xmath87 ( but this case is not really interesting ) . on the other hand ,
if all the members of @xmath49 have no petals in the filled julia set , there is no @xmath70 such that @xmath72 is hybrid conjugate to the map @xmath85 , and finally the image of @xmath48 under the map @xmath54 is not the whole of @xmath6 , but it belongs to @xmath4 .
this is the case we are interested in .
consider the family of cubic polynomials @xmath88 the maps belonging to this family have a parabolic fixed point at @xmath89 of multiplier @xmath34 , and critical points at @xmath90 and @xmath91 .
call @xmath92 the connectedness locus for this family .
let @xmath93 denote the bttcher coordinates for @xmath94 tangent to the identity at infinity , call @xmath95 the co - critical point of @xmath96 and let @xmath97 be the conformal rappresentation of @xmath98 given by @xmath99 define @xmath100 as the open set bounded by the external rays of angle @xmath101 and @xmath102 ( see @xcite ) . in this section
we are going to prove that the family @xmath103 yields to a nice family of parabolic - like mappings .
let us construct a parabolic - like restriction for every member of the family @xmath105 .
call @xmath106 the immediate basin of attraction of the parabolic fixed point @xmath89
. then @xmath107 belongs to @xmath106 , while @xmath96 does not belong to @xmath106 .
let @xmath108 be the riemann map normalized by setting @xmath109 and @xmath110 , and let @xmath111 be its inverse . by the carathodory theorem
the map @xmath112 extends continuously to @xmath113 .
note that @xmath114 .
let @xmath115 be a @xmath116 periodic point in the first quadrant , such that the hyperbolic geodesic @xmath117 connecting @xmath115 and @xmath118 separates the critical value @xmath119 from the parabolic fixed point @xmath120 .
let @xmath121 be the jordan domain bounded by @xmath122 , union the arcs up to potential level @xmath34 of the external rays landing at @xmath123 and @xmath124 , together with the arc of the level @xmath34 equipotential connecting this two rays around @xmath96 ( see fig .
[ cubicbcnnew ] ) .
let @xmath125 be the connected component of @xmath126 containing @xmath127 and the dividing arcs @xmath128 be the fixed external rays landing at the parabolic fixed point @xmath127 and parametrized by potential .
then ( @xmath129 ) is a parabolic - like map of degree @xmath0 ( see fig .
[ cubicbcnnew ] ) . for every @xmath104 the parabolic fixed point @xmath127 of @xmath94 has parabolic multiplicity @xmath34 , and ( @xmath129 ) is a parabolic - like map with no attracing petals in its filled julia set . by the construction we gave , it follows easily that @xmath105 restricts to an analytic family of parabolic - like mappings . since external rays move holomorphically , to prove that this analytic family of parabolic - like maps is nice it suffices to show that the boundaries of @xmath121 move holomorphically with the parameter ( by construction the motion defines a piecewise @xmath21-diffeomorphisms with no cusps in @xmath78 ) .
let us start by proving that the basin of attraction @xmath106 of @xmath127 depends holomorphically on the parameter .
call @xmath132 the maximal attracting petal in @xmath106 , and let @xmath133 be fatou coordenates for @xmath94 normalized by sending the critical point @xmath107 to @xmath34 .
since the family @xmath134 depends holomorphically on @xmath131 , @xmath135 depends holomorphically on @xmath131 and the extended fatou coordenates to the whole parabolic basin @xmath136 depend holomorphically on @xmath131 . on the other hand ,
let @xmath137 be extended fatou coordinates for the map @xmath116 , normalized by sending the critical point to @xmath34 . since the riemann map @xmath138 is a holomorphic conjugacy between @xmath94 and @xmath116 , @xmath139 are fatou coordinates for @xmath94 . since @xmath140 , we have that @xmath141 .
hence the riemann map @xmath138 depends holomorphically on @xmath131 .
so ( fixing a base point @xmath142 ) the dynamical holomorphic motion @xmath143 ( holomorphic in @xmath78 ) induces a dynamical holomorphic motion @xmath144 ( holomorphic in @xmath78 ) , which extends by the @xmath52-lemma to a dynamical holomorphic motion of @xmath145 .
since @xmath146 moves holomorphically , the points @xmath147 and @xmath148 and the arc @xmath149 defined in [ dynamicalconstruction ] depend holomorphically on @xmath131 .
since equipotentials and external rays move holomorphically , for every @xmath104 the set @xmath150 moves holomorphically .
hence the family @xmath130 restricts to a degree @xmath0 nice analytic family of parabolic - like maps .
in @xcite we proved that a degree @xmath0 parabolic - like map is hybrid conjugate to a member of the family @xmath2 by changing its external class into the class of @xmath151 ( see theorem 6.3 in @xcite ) and showing that a parabolic - like map is holomorphically conjugate to a member of the family @xmath2 if and only if its external class is given by the class of @xmath116 ( see proposition 6.2 in @xcite ) .
we defined a ( quasiconformal ) conjugacy between two parabolic - like maps @xmath152 and @xmath153 to be a ( quasiconformal ) homeomorphism between ( appropriate ) restrictions of @xmath154 and @xmath155 which conjugates dynamics on @xmath156 ( see def .
@xmath157 in @xcite ) .
let us review how we changed the external class of a degree @xmath0 parabolic - like map @xmath20 into the class of @xmath116 . as first step
, we constructed a homeomorphism @xmath158 , quasiconformal everywhere but at the parabolic fixed point , between a fundamental annulus @xmath159 of @xmath20 and a fundamental annulus @xmath160 of @xmath116 .
then we defined on @xmath161 an almost complex structure @xmath162 by pulling back the standard structure by @xmath158 . in order to obtain on @xmath154 a bounded and invariant ( under a map coinciding with @xmath20 on @xmath163 ) almost complex structure @xmath164 we replaced @xmath20 with @xmath116 on @xmath82 , and spread @xmath162 by the dynamics of this new map @xmath165 ( and kept the standard structure on @xmath38 ) . finally , by integrating @xmath164 we obtained a parabolic - like map hybrid conjugate to @xmath20 and with external map @xmath116 . in this paper
we want to perform this surgery for nice analytic families of degree @xmath0 parabolic - like maps , and prove that the map @xmath166 induced by the family of hybrid conjugacies extends to a continuous map @xmath167 which under suitable conditions restricts to a branched covering of @xmath168 .
we will start by defining a family of quasiconformal maps , depending holomorphically on the parameter , between a fundametal annulus of @xmath116 and fundamental annuli @xmath169 of @xmath170 . in analogy with the polynomial - like setting
we will call this family a _
holomorphic tubing_. in order to construct a holomorphic tubing , fixed a @xmath171 , we will start by constructing a quasiconformal homeomorphism @xmath158 between @xmath172 and @xmath173 ( see section [ psidiff ] ) and a dynamical holomorphic motion @xmath174 ( see section [ holmot ] ) .
hence we will obtain a holomorphic tubing by composing the inverse of @xmath158 with the holomorphic motion ( see section [ a ] ) . by tubing , we will extend the map @xmath54 to the whole of @xmath55 ( see section [ external ] ) .
we will prove that the map @xmath54 is continuous ( see section [ con ] ) , holomorphic on the interior of @xmath48 ( see section [ anal ] ) and with discrete fibers ( see section [ d ] ) . finally , we will prove that , on compact subsets of @xmath55 , the map @xmath54 is a degree @xmath175 branched covering ( see section [ final ] ) . by defining proper families of parabolic - like maps we wil
give the condition under which , for each neighborhood @xmath11 of @xmath34 , @xmath176 is a compact subset of @xmath55 ( see section [ p ] ) .
this implies the following result : [ bigthm ] given a proper family of parabolic - like maps @xmath177 , the map @xmath178 is a degree @xmath175 branched covering .
more precisely , for every neighborhood @xmath11 of @xmath34 in @xmath13 ( with @xmath179 ) there exists a neighborhood @xmath180 of @xmath181 in @xmath182 such that the map @xmath183 is a degree @xmath175 branched covering .
in this section we will first , fixed a @xmath171 , construct a fundamental annulus for @xmath116 and one for @xmath184 , and recall the quasiconformal @xmath21-diffeomorphism @xmath158 between these fundamental annuli .
then we will construct a holomorphic motion of the ranges of the nice analytic family of parabolic - like maps , and by this holomorphic motion derive fundamental annuli for @xmath72 from the fundamental annulus of @xmath184 .
finally we will obtain a holomorphic tubing by composing the inverse of @xmath158 with the holomorphic motion .
the map @xmath185 is an external map of every member of the family @xmath2 ( see prop . 4.2 in @xcite ) .
let @xmath186 ( where @xmath187 for an @xmath188 , and @xmath189 ) be a degree @xmath0 covering extension ( this is , an extension such that @xmath186 is a degree @xmath0 covering and there exists a dividing arc which devides @xmath190 and @xmath190 into @xmath191 and @xmath191 respectively , such that @xmath192 is a topological quadrilateral ; see def . 5.2 in @xcite ) .
choose @xmath171 .
let @xmath193 be an external map of @xmath184 , @xmath26 be its parabolic fixed point and define @xmath194 , @xmath195 ( where @xmath196 is an external equivalence between @xmath184 and @xmath193 ) .
let @xmath197 and @xmath198 be repelling petals for the parabolic fixed point @xmath26 which intersect the unit circle and @xmath199 be fatou coordinates for @xmath193 with axis tangent to the unit circle at the parabolic fixed point @xmath26 .
let @xmath200 and @xmath201 be repelling petals which intersect the unit circle for the parabolic fixed point @xmath120 of @xmath116 , and let @xmath202 be fatou coordinates for @xmath116 with axis tangent to the unit circle at @xmath34
. define @xmath203 and @xmath204 .
define @xmath205 and @xmath206 .
we call _ fundamental annulus for @xmath116 _ the topological annulus @xmath207 .
let @xmath209 be a homeomorphism coinciding with @xmath210 on @xmath211 , quasiconformal on @xmath212 ( where @xmath213 is the parabolic fixed point of @xmath184 ) and real - analytic diffeomorphism on @xmath214 ( see claim @xmath215 in the proof of thm .
@xmath216 in @xcite ) .
define @xmath217 , @xmath218 , and @xmath219 .
consider @xmath220 define @xmath221 , @xmath222 , and the _ fundamental annulus _ @xmath223 .
let @xmath224 be a quasiconformal map which coincides with @xmath210 on @xmath211 ( see the proof claim @xmath225 in thm .
@xmath216 in @xcite ) .
define a map @xmath226 as follows : @xmath227 the map @xmath158 is a homeomorphism , quasiconformal on @xmath228 , so the map @xmath229 is a homeomorphism , quasiconformal on @xmath230 .
define for all @xmath70 the set @xmath231 .
then the set @xmath232 is a topological annulus .
define the map @xmath233 as follows : @xmath234 since @xmath235 and @xmath236 are holomorphic motions with disjoint images on @xmath237 , and @xmath238 is a degree @xmath7 covering , @xmath239 is a holomorphic motion with basepoint @xmath240 . since @xmath57 ,
by the slodkowski s theorem we can extend @xmath239 to a holomorphic motion @xmath241 .
in particular we obtain a holomorphic motion of @xmath242 . for every @xmath70 ,
define @xmath243 , and @xmath244 .
define for every @xmath70 the map @xmath245 as follows : @xmath246 and the set @xmath247 .
finally , define for all @xmath70 the set @xmath248 .
then the set @xmath249 is a topological annulus , and we call it _ fundamental annulus of @xmath72_. the holomorphic motion @xmath241 restricts to a holomorphic motion @xmath250 which respects the dynamics
. define @xmath251 .
we call the map @xmath252 a * holomorphic tubing*. a holomorphic tubing is not a holomorphic motion , since @xmath253 , but nevertheless it is quasiconformal in @xmath78 for every fixed @xmath70 and holomorphic in @xmath52 for every fixed @xmath254 . let us now straighten the members of the family @xmath255 to members of the family @xmath2 . for every @xmath70 define on @xmath256 the beltrami form @xmath257 as follows : @xmath258 for every @xmath52 the map @xmath259 is quasiconformal , hence @xmath260 on every compact subset of @xmath55 . on @xmath261
the beltrami form @xmath262 is obtained by spreading @xmath263 by the dynamics of @xmath72 , which is holomorphic , while on @xmath264 the beltrami form @xmath262 is constant for all @xmath265 ( by construction of the map @xmath266 ) .
thus @xmath267 , which is bounded .
by the measurable riemann mapping theorem ( see @xcite ) for every @xmath70 there exists a quasiconformal map @xmath268 such that @xmath269 .
finally , for every @xmath70 the map @xmath270 is a parabolic - like map hybrid conjugate to @xmath72 and holomorphically conjugate to a member of the family @xmath2 ( see prop . 6.2 in @xcite ) .
[ locboundil ] note that for every @xmath70 , the dilatation of the integrating map @xmath271 is equal to the dilatation of the holomorphic tubing @xmath259 .
so the family of integrating maps @xmath272 has locally bounded dilatation .
let us lift the tubing @xmath259 .
define @xmath273 , @xmath274 , @xmath275 and @xmath276 .
hence @xmath277 and @xmath278 are degree @xmath0 covering maps , and we can lift the tubing @xmath259 to @xmath279 ( such that @xmath280 on @xmath281 ) .
define recursively @xmath282 , @xmath283 , @xmath284 and @xmath285 .
hence @xmath286 and @xmath287 are degree @xmath0 covering maps , and we can lift the tubing to @xmath288 ( such that @xmath289 on @xmath290 ) . in the case
@xmath81 is connected , we can lift the tubing @xmath259 to the whole of @xmath291 .
if @xmath81 is not connected , the maximum domain we can lift the tubing @xmath259 to is @xmath292 , such that @xmath293 contains the critical value of @xmath72 . note that the extension is still quasiconformal in @xmath78 .
consider the map @xmath294 ( defined in section [ intro ] ) which associates to each @xmath295 the multiplier of the fixed point of the map @xmath45 hybrid equivalent to @xmath53 . in this section , we will first extend the map @xmath54 to the whole of @xmath55 ( see section [ external ] ) , then prove that the map @xmath296 is continuous at the boundary of @xmath48 ( see prop .
[ con ] ) and that it depends analytically on @xmath52 for @xmath297 ( see prop .
finally , we will prove that the map @xmath54 has discrete fibers ( see prop .
[ discretness ] ) let @xmath252 be a holomorphic tubing for the nice analytic family of parabolic - like maps @xmath49 .
call @xmath298 the critical point of @xmath72 and let @xmath265 be such that @xmath299 , @xmath300 .
lift the holomorphic tubing @xmath259 to @xmath301 ( see section [ li ] ) .
we can therefore extend the map @xmath54 by setting : @xmath302 @xmath303 where @xmath304 is the canonical isomorphism between the complement of @xmath6 and the complement of the unit disk ( see @xcite ) . since the tubing @xmath259 has locally bounded dilatation , the map @xmath305 is quasiregular on @xmath306 for any open @xmath307 .
an indifferent periodic point @xmath308 for @xmath309 , is called _ persistent _ if for each neighborhood @xmath310 of @xmath308 there exists a neighborhood @xmath311 of @xmath240 such that , for every @xmath312 the map @xmath53 has in @xmath310 an indifferent periodic point @xmath313 of the same period and multiplier ( see @xcite ) .
let @xmath1 be an family of parabolic - like mappings . for all @xmath70 ,
the parabolic fixed point is persistent . since all the other indifferent periodic points are non persistent , in the remainder we will call them _ indifferent periodic points _ without further notation .
[ neutparametervalues ] the indifferent parameter values for a family of parabolic - like mappings belong to @xmath314 .
the proof follows the proof of prop . 11 in @xcite . since for all @xmath315 the critical point @xmath298 of @xmath72 belongs to @xmath316 , the map @xmath72 is hyperbolic .
assume that for @xmath317 the map @xmath318 has an indifferent periodic point @xmath319 of period @xmath320 , and assume first @xmath321
. by the implicit function theorem there exist @xmath322 neighborhoods of @xmath240 and @xmath319 respectively , with @xmath323 , where the indifferent cycle and the critical point move holomorphically with the parameter , and where the multiplier map @xmath324 is a holomorphic non constant map .
set @xmath325 . by taking a restrictions if necessary
, we can assume @xmath240 is the only parameter in @xmath326 for which @xmath72 has in @xmath327 an indifferent periodic point .
let @xmath328 be a sequence in @xmath326 converging to @xmath240 , such that for all @xmath265 , @xmath329 .
hence for all @xmath265 , there exists a @xmath330 such that @xmath331 we can assume @xmath332 independent of @xmath52 by choosing a subsequence . let us define for all @xmath333 the sequence @xmath334 since @xmath335 is a family of analytic maps bounded on any compact subset of @xmath326 , it is a normal family .
let @xmath336 be a subsequence converging to some function @xmath337 .
then @xmath338 for all @xmath265 , hence @xmath339 and for all @xmath333 , @xmath340 .
but in @xmath326 there are parameters
@xmath341 for which @xmath342 is a repelling periodic point , and thus it can not attract the sequence @xmath343 . in the case
@xmath344 , let @xmath345 be a neighborhood of @xmath346 , let @xmath347 , @xmath348 be a branched covering of @xmath345 branched at @xmath127 for some neighborhood @xmath349 of @xmath127 , and repeat the previous argument .
in this section we prove that the map @xmath296 is continuous on the boundary of @xmath48 . [ 7 ] suppose @xmath350 , with @xmath351 . if the maps @xmath352 and @xmath353 are quasiconformally conjugate , then @xmath354 .
let @xmath355 and @xmath356 be parabolic - like restrictions of @xmath352 and @xmath353 respectively ( for the construction of a parabolic - like restriction of members of the family @xmath2 see the proof of prop .
4.2 in @xcite ) , and let @xmath357 be a quasiconformal conjugacy between them . if @xmath358 is of measure zero ( where @xmath359 , and @xmath360 is the parabolic basin of attraction of infinity , see section 1 in @xcite ) , then @xmath361 is a hybrid conjugacy and the result follows from prop . 6.5 in @xcite
let @xmath358 be not of measure zero .
define on @xmath362 the following beltrami form : @xmath363 since @xmath361 is quasiconformal , @xmath364 .
therefore for @xmath365 we can define on @xmath362 the family of beltrami form @xmath366 , and @xmath367 .
the family @xmath368 depends holomorphically on @xmath369 .
let @xmath370 be the family of integrating maps fixing @xmath371 , @xmath34 and @xmath127 .
hence the family @xmath372 depends holomorphically on @xmath369 , @xmath373 and @xmath374 .
the family of holomorphic maps @xmath375 has the form @xmath376 ( since it is a family of quadratic rational maps with a parabolic fixed point at @xmath377 with preimage at @xmath89 and a critical point at @xmath378 ) and it depends holomorphically on @xmath369 .
therefore the map @xmath379 is holomorphic , hence it is either an open or constant , and @xmath380 . if @xmath381 is open , there exists a neighborhood @xmath326 of @xmath127 such that @xmath382 .
hence the map @xmath381 is constant , so for all @xmath369 , @xmath383 . in particular @xmath384 , and @xmath385 .
finally the map @xmath386 is a quasiconformal conjugacy between @xmath352 and @xmath353 with @xmath387 on @xmath358 .
so @xmath352 and @xmath353 are hybrid equivalent , and the result follows by prop . 6.5 in @xcite .
[ conti ] the map @xmath296 is continuous at any point @xmath388 , and moreover @xmath389 . in order to prove continuity at any point @xmath388
, we have to show that for every sequence @xmath390 converging to @xmath391 there exists a subsequence @xmath392 such that @xmath393 converges to @xmath394 .
let us start by proving that @xmath395 .
let @xmath396 be a sequence of indifferent parameters converging to @xmath240 .
hence there exists a sequence @xmath397 such that , for each @xmath398 , @xmath399 is hybrid conjugate to @xmath400 by some quasiconformal map @xmath401 .
the sequence @xmath401 is a sequence of quasiconformal maps with locally bounded dilatation ( see remark [ locboundil ] ) , hence it is precompact in the topology of uniform convergence on compact subsets of @xmath402 ( see @xcite ) .
therefore there exists a subsequence @xmath403 which converges to some quasiconformal limit map @xmath404 , which conjugates @xmath184 to some @xmath405 , so @xmath406 . for all @xmath398 the map @xmath399
has an indifferent periodic point , hence @xmath400 has an indifferent periodic point , thus @xmath407 and finally @xmath408 .
since the map @xmath184 is hybrid conjugate to @xmath409 and quasiconformally conjugate to @xmath405 , and @xmath410 , by prop.[7 ] @xmath411 .
let now @xmath390 be a sequence converging to @xmath391 .
since the sequence @xmath412 is precompact , there exists a subsequence @xmath413 such that @xmath414 converges to some limit map @xmath415 which conjugates @xmath184 to @xmath416 so @xmath417 . finally , since @xmath395 and @xmath184 is quasiconformally conjugate to both @xmath418 and @xmath409 , by prop.[7 ] @xmath419 .
the proof of the analycity of the map @xmath54 on the interior of @xmath420 ( see [ an ] ) follows the proof lyubich gave in the setting of polynomial - like mappings ( see @xcite ) .
we will prove that the map @xmath54 is holomorphic on hyperbolic components first , and then on queer components . to prove that @xmath54 is holomorphic on queer components , we first need the following proposition . [ invariantline ] let @xmath421 be a queer component of @xmath422 .
then for every @xmath423 , @xmath424 admits an invariant beltrami form with positive support .
in particular , @xmath425 . choose @xmath426 and set @xmath427 .
let us start by proving that there exists a dynamical holomorphic motion @xmath428 with base point @xmath429 , holomorphic in @xmath78 .
let @xmath430 be an attracting petal of @xmath86 containing the critical value @xmath431 , and let @xmath432 be the incoming fatou coordinates of @xmath86 normalized by @xmath433 . for @xmath423 , let @xmath434 be an attracting petal of @xmath45 and let @xmath435 be the incoming fatou coordinates of @xmath45 with @xmath436 .
the map @xmath437 is a conformal conjugacy between @xmath86 and @xmath45 . defining @xmath438 as the connected component of @xmath439 containing @xmath430 , and @xmath440 as the connected component of @xmath441 containing @xmath434
, we can lift the map @xmath442 to @xmath443 . since @xmath444 and @xmath445 are connected ( where @xmath444 and @xmath445 are the complements of the basin of attraction of the parabolic fixed point for @xmath86 and @xmath45 respectively ) , by interated lifting we can extend @xmath446 to @xmath447 .
the map @xmath448 is a holomorphic conjugacy between @xmath86 and @xmath45 , and since the family @xmath2 is a family of holomorphic maps depending holomorphic on the parameter and so fatou coordenates depend holomorphically on the parameter , the family @xmath449 depends holomorphically on the parameter .
hence @xmath450 is a dynamical holomorphic motion with base point @xmath429 , holomorphic in @xmath78 .
since for every @xmath423 all the periodic points of @xmath45 ( but the parabolic fixed point ) are repelling , @xmath451 is nowhere dense .
hence @xmath452 and thus by the @xmath52-lemma we can extend @xmath448 to @xmath453 .
note that @xmath454 still conjugates dynamics , and it is conformal on @xmath455 .
define @xmath456 . by construction , @xmath457 on @xmath455 .
on the other hand , if @xmath458 on @xmath362 , by the weyl s lemma for every @xmath423 the map @xmath454 is holomorphic , hence for all @xmath423 the maps @xmath45 are conformally equivalent .
therefore @xmath459 on @xmath444 , and thus @xmath460 on @xmath444 .
in particular , this implies @xmath461 .
[ an ] the map @xmath3 depends analytically on @xmath52 for @xmath297 .
let us start by proving that , for every hyperbolic component @xmath462 , there exists a hyperbolic component @xmath463 such that @xmath464 is a holomorphic map . by the implicit function theorem and prop .
[ neutparametervalues ] , all the parameter values in @xmath465 are hyperbolic . hence for @xmath466 , @xmath72 has an attracting cycle , thus @xmath467 has an attracting cycle and @xmath463 . since @xmath271 is conformal on @xmath468 , calling @xmath469 the multiplier maps on @xmath470 respectively , @xmath471 .
hence on @xmath465 we can write the map @xmath54 as @xmath472 .
since @xmath72 is holomorphic in both @xmath52 and @xmath78 , and by the implicit function theorem the attracting cycle moves holomorphically on @xmath465 , the map @xmath473 is holomorphic .
for the same reason , @xmath474 is holomorphic as well .
since @xmath474 has degree @xmath34 ( see @xcite ) , it is conformal and then @xmath54 is holomorphic . let now @xmath475 be a queer component in @xmath420 , and let @xmath476 . since @xmath477 we can lift the holomorphic motion @xmath478 constructed in [ holmot ] to @xmath479 ( as we did for the holomorphic tubing , see [ li ] ) .
since for all @xmath480 , @xmath81 is nowhere dense , by the @xmath52-lemma we can extend @xmath481 to a dynamical holomorphic motion @xmath482 .
let @xmath483 be the member of the family @xmath2 hybrid conjugate to @xmath184 , let @xmath484 be a hybrid conjugacy between them and set @xmath485 .
note that , for all @xmath480 , the map @xmath486 is a quasiconformal conjugacy between @xmath483 and @xmath72 .
define on @xmath487 the following family of beltrami forms : @xmath488 the family @xmath489 is a family of @xmath483-invariant beltrami forms depending holomorphically on @xmath52 . by prop .
[ invariantline ] , for every @xmath480 , on @xmath490 @xmath491 .
let @xmath492 be the family of integrating maps fixing @xmath493 and @xmath371 , then the family @xmath494has the form @xmath495 , where @xmath496 depends holomorphically on the parameter . finally , for every @xmath480 , the map @xmath497 is a hybrid conjugacy between @xmath72 and @xmath498 , hence @xmath498 is the straightening of @xmath72 and the map @xmath499 is holomorphic .
set @xmath500 .
[ discretness ] for every @xmath501 , @xmath502 is discrete .
this follows the proof of lemma @xmath503 in @xcite .
let us assume there exists a @xmath504 such that there exists a sequence @xmath505 and @xmath506 .
the map @xmath54 is quasiregular on @xmath507 and holomorphic on @xmath420 , hence @xmath508 ( or @xmath509 in a queer component of @xmath510 for which @xmath54 is constant , and then we can replace @xmath509 with a boundary point ) .
note that , for every @xmath265 , the maps @xmath511 are hybrid equivalent to @xmath512 by some hybrid equivalence @xmath513 .
let us assume that for all @xmath52 in a neighborhood of @xmath514 , @xmath515 ( in other case , take a nice analytic family of parabolic - like restrictions for which the assumption holds ) .
for every @xmath52 , call @xmath71 the parabolic fixed point of @xmath53 , @xmath298 its critical point and @xmath516 its critical value .
consider @xmath514 as the base point of a holomorphic motion @xmath517 , extend it by the slodkowski s theorem and then restrict it to a holomorphic motion @xmath518 .
define for every @xmath265 the map @xmath519 as follows : @xmath520 where the maps @xmath521 are as in [ holmot ] and the sets @xmath522 are constructed in [ li ] .
the proof of prop . 6.4 in @xcite shows that , for every @xmath265 , the map @xmath523 is continuous and hence quasiconformal .
therefore , for every @xmath265 , @xmath523 restricts to a hybrid equivalence between @xmath512 and @xmath511 .
consider on @xmath487 the family of beltrami forms @xmath524 .
note that trivially @xmath523 integrates @xmath525 , and for some subsets @xmath526 of @xmath527 and @xmath528 of @xmath529 , @xmath530 .
on the other hand , define on @xmath487 the family of beltrami forms @xmath257 as follows : @xmath531 where for every @xmath52 the map @xmath532 which defines the sets @xmath533 and spreads @xmath257 is defined as follows : @xmath534 note that @xmath535 and @xmath532 coincide on @xmath536 , hence for every @xmath537 .
therefore , for all @xmath265 , @xmath538 .
the family @xmath539 depends holomorphically on @xmath52 , because @xmath481 is a holomorphic motion , on @xmath540 it is constant and on @xmath541 it is spread by the dynamics of @xmath512 ( which does not depends on @xmath52 ) .
let @xmath542 be the holomorphic family of integrating maps mapping @xmath543 to @xmath71 , @xmath544 to @xmath298 and @xmath545 to @xmath516 .
the family @xmath546 is a holomorphic family of parabolic - like mappings hybrid equivalent to @xmath512 . for all @xmath265 , @xmath547 ( since they are solutions of the same family of beltrami form and they coincide on @xmath548 points ) , and therefore on the @xmath526 , @xmath549 .
choose a neighborhood @xmath550 of @xmath514 in @xmath55 and an open set @xmath327 in @xmath551 such that the maps @xmath552 and @xmath72 are well - defined in @xmath553 .
since @xmath554 on @xmath555 , and @xmath556 converges to @xmath514 , @xmath557 on @xmath558 .
this is impossible , because in @xmath550 there are @xmath52 for which @xmath72 has disconnected julia set ( since @xmath559 ) , while for all @xmath52 , @xmath552 has connected julia set ( since it is equivalent to @xmath512 ) .
in this chapter , we will first prove that for any closed and connected subset @xmath561 of @xmath500 , if @xmath562 is compact , then @xmath499 is a proper map of degree @xmath7 ( prop .
[ th ] ) , and then that @xmath563 is a @xmath7-fold branched covering ( prop . [ cov ] ) . finally , we will prove that , under certain conditions ( def .
[ properfam ] ) , for every neighborhood @xmath11 of the root of @xmath6 ( without specifications , we consider a neighborhood open ) , the set @xmath176 is compact in @xmath55 .
this implies theorem [ bigthm ] .
[ th ] let @xmath561 be a closed and connected subset of @xmath567 , and @xmath562 .
if @xmath475 is compact , then there exist neighborhoods @xmath180 of @xmath561 in @xmath567 and @xmath568 of @xmath475 in @xmath55 such that @xmath569 is a proper map of degree @xmath7 , where , for every @xmath570 , @xmath571 .
the proof follows the analogous one in @xcite .
let @xmath572 be a closed neighborhood of @xmath475 in @xmath55 with @xmath573 ( it exists since @xmath475 compact ) .
hence @xmath574 , @xmath575 is proper and @xmath576 . call @xmath180 the connected component of @xmath577 which contains @xmath561 , and set @xmath578 .
then @xmath579 , hence the map @xmath580 is proper . since @xmath54 is continuous and @xmath180
is connected , @xmath568 is the union of connected components .
set @xmath581 .
the restriction @xmath582 is then a proper map between connected sets , thus it has a degree @xmath583 , and since @xmath54 has discrete fibers , for every @xmath584 ( see @xcite pg .
136 ) : @xmath585 ( note that @xmath586 ) .
therefore @xmath569 has a degree @xmath587 and in particular for all @xmath588 the map @xmath580 is continuous , and by the previous proposition it is a proper surjective map of degree @xmath7 .
let @xmath589 , and let @xmath590 be a neighborhood of @xmath591 in @xmath180 such that for all @xmath592 and @xmath593 , @xmath594 is a regular point ( such a @xmath590 exists since the fiber of @xmath54 are finite ) .
hence @xmath595 , with the @xmath596 disjoint and @xmath597 .
if @xmath598 , @xmath591 is a regular point , and for all @xmath599 , @xmath600 is a homeomorphism . if @xmath601 , we want to show that for every @xmath602 there exists a neighborhood @xmath603 of @xmath604 such that @xmath605 , with the @xmath606 disjoint and @xmath607 homeomorphism
this is clear because all the points in @xmath608 different from the preimage of @xmath591 are regular points .
as we saw in [ anfam ] , the range @xmath567 of the map @xmath54 is not the whole of @xmath13 but a proper subset , because there is no @xmath70 for which @xmath72 is hybrid equivalent to @xmath609 .
hence @xmath610 .
however , we could hope that @xmath611 is the only point of @xmath6 missing from @xmath567 , or in other words , that as @xmath612 @xmath613 or @xmath614 .
indeed this is the case under appropriate conditions ( e.g. the following one ) .
assume @xmath475 is not compact in @xmath55 .
then there exists a sequence @xmath622 such that @xmath623 as @xmath624 . on the other hand , for all @xmath265 , @xmath625 .
let @xmath626 be a subsequence converging to some parameter @xmath627 .
since @xmath561 is compact , the limit point @xmath627 belongs to @xmath628 .
this is a contradiction , because @xmath49 is a proper family of parabolic - like mappings .
therefore @xmath475 is compact in @xmath55 .
hence if @xmath49 is a proper family of parabolic - like mappings , @xmath619 a neighborhood of @xmath611 , @xmath620 , and @xmath621 , by prop .
[ cov ] there exist neighborhoods @xmath568 of @xmath475 in @xmath55 and @xmath180 of @xmath561 in @xmath567 such that the restriction @xmath569 is a @xmath629-fold branched covering .
next proposition tells us how to compute the degree @xmath629 of the covering .
let @xmath49 be a proper family of parabolic - like mappings , @xmath619 a neighborhood of @xmath611 ( with @xmath630 ) , @xmath620 , and @xmath180 a neighborhood of @xmath561 given by prop .
[ th ] . then calling @xmath298 the critical point of @xmath72 and @xmath631 , the degree @xmath629 of the branched
covering @xmath632 is equal to the number of times @xmath633 turns around @xmath127 as @xmath52 describes @xmath634 .
let us remember that for every @xmath172 the map @xmath635 has two critical points : @xmath120 and @xmath378 .
after a change of coordenates we can assume @xmath120 is the first critical point attracted by @xmath371 . hence for all @xmath636 , @xmath378 is the critical point of the parabolic - like restriction of @xmath45 ( see the proof of prop .
4.2 in @xcite ) .
the proof follows the analogous one in @xcite .
let @xmath298 be the critical point of @xmath72 .
choose @xmath240 such that @xmath637 .
let @xmath409 be the member of the family @xmath2 hybrid equivalent to @xmath184 .
therefore @xmath638 , hence @xmath639 .
this means that the multiplicity of @xmath240 as zero of the map @xmath640 is the multiplicity of @xmath240 as zero of the map @xmath641 .
this last one is @xmath642 .
let @xmath1 be a proper family of parabolic - like mappings .
for every @xmath70 , @xmath72 is the restriction of some map @xmath643 .
consequently , @xmath55 is the restriction of the parameter plane of the maps @xmath643 , call it @xmath644 .
call @xmath645 the connectedness locus of @xmath643 , hence @xmath646 let @xmath49 be a proper family of parabolic - like mappings .
if the map @xmath647 is a homeomorphism , and @xmath648 , then @xmath54 extends to a homeomorphism @xmath649 for a unique @xmath650 . more generally ,
if the the map @xmath647 is a degree @xmath629 branched covering , and @xmath648 , then map @xmath54 extends continuously to @xmath651 for exactly @xmath652 points in @xmath653 .
let @xmath49 be a proper family of parabolic - like mappings for which the map @xmath647 is a degree @xmath629 covering and @xmath648 . since @xmath49 is a proper family , as @xmath618 , @xmath654 .
we will prove that for every @xmath655 , @xmath656 , and that @xmath657 is a discrete set . then by continuity , @xmath658 .
the original family @xmath643 has a persistent parabolic fixed point of multiplier @xmath34 and it depends holomorphically on @xmath52 . take a succession @xmath659 such that @xmath660 , and call @xmath661 the limit of the @xmath662 in @xmath644 . since for every @xmath332
, @xmath662 is a hyperbolic parameter , the limit @xmath661 is a hyperbolic or indifferent parameter .
so , if @xmath663 , @xmath664 presents a degree @xmath0 parabolic - like restriction and @xmath665 .
since @xmath666 , @xmath667 .
let us prove now that the set @xmath657 is discrete .
note that this is the set of parameters for which the parabolic fixed point @xmath71 of @xmath643 has parabolic multiplicity @xmath668 , where @xmath265 is the multiplicity of @xmath71 for @xmath669 .
then , in a neighborhood of @xmath71 we can consider @xmath643 as @xmath670 with @xmath671 and @xmath672 holomorphic in @xmath52 .
hence the set @xmath673 for which @xmath674 is a discrete set .
[ [ the - parameter - plane - of - the - family - c_azzaz2z3-presents - baby - m_1 ] ] the parameter plane of the family @xmath675 presents baby-@xmath6 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
let us show that the family of parabolic - like mappings @xmath130 is proper .
call @xmath676 the connectedness locus of @xmath105 .
the finite boundaries of @xmath55 are the external rays of angle @xmath101 and @xmath102 , which can not intersect the connectedness locus @xmath676 in other point than the landing point , if they land .
since these rays land at @xmath677 ( see @xcite ) , for @xmath678 either @xmath679 , hence @xmath613 , or @xmath680 , hence @xmath614 .
finally , by the relation @xmath681 between external rays in dynamical and parameter plane , the degree of @xmath682 is @xmath34 .
therefore @xmath92 presents a baby @xmath6 . by symmetry
, we can repeat the construction for the family @xmath683 , where @xmath684 is the open set bounded by the external rays of angle @xmath685 and @xmath686 .
hence the connectedness locus of the family @xmath687 presents two baby @xmath6 , namely in the connected component bounded by the external rays of angle @xmath101 and @xmath102 , and in the connected component bounded by the external rays of angle @xmath685 and @xmath686 ( see fig .
[ m ] and [ m1 ] in the introduction ) .
13 k. astala , _ elliptic partial differential equations and quasiconformal mappings in the plane _ , princeton univ . press , ( 2008 ) .
l. ahlfors , lectures on quasiconformal mappings , second edition .
ams university lecture series , vol .
b. branner & j. h. hubbard , the iteration of cubic polynomials ii : patterns and parapatterns , _
acta math._,157 ( @xmath688 ) , no
. 1 - 2 , 2348 . a. douady & c. j. earle , conformally natural extension of homeomorphisms of the circle , _
acta math._,169 ( @xmath689 ) , no . 3 - 4 , 229 - 325 .
a. douady & j. h. hubbard , on the dynamics of polynomial - like mappings , _ ann .
sci . cole norm .
sup._,(4 ) , vol.@xmath690 ( @xmath691 ) , 287 - 343 .
o. forster , _ lectures on riemann surfaces _ , springer , ( 1981 ) .
a. hatcher , _ algebraic topology _ , cambridge univ
. press , ( 2002 ) .
l. hrmander , _ the analysis of linear partial differential operators i _ , springer - verlag , ( 1983 ) .
j. hubbard , _ teichmller theory , volume 1 : teichmller theory _ , matrix editions , ( 2006 ) .
l. lomonaco , parabolic - like maps , arxiv:1111.7150 .
l. lomonaco , _ parabolic - like maps _ , imfufa tekst , ( 2013 )
. m. lyubich , _ conformal geometry and dynamics of quadratic polynomials _ , www.math.sunysb.edu/ mlyubich / book.pdf .
j. milnor , _ dynamics in one complex variable _ , annals of mathematics studies , ( 2006 ) .
j. milnor , on rational maps with two critical points , _ experimental mathematics _ , ( 4 ) , vol
. 9 ( 2000 ) , 481 - 522 .
r. ma , p. sad & d. sullivan , on the dynamics of rational maps , _ ann .
cole norm .
sup._,(4 ) , vol.@xmath692 ( @xmath693 ) , 193 - 217 .
d. sullivan , quasiconformal homeomorphisms and dynamics iii , _ ann .
cole norm .
sup._,(4 ) , vol.@xmath692 ( @xmath693 ) , 193 - 217 . s. nakane , capture components for cubic polynomials with parabolic fixed points , _ academic reports fac . eng .
tokio polytech .
univ._,(1 ) , vol.@xmath694 ( @xmath695 ) , 33 - 41 . c. petersen & l. tan , branner - hubbard motions and attracting dynamics , _ dynamics on the riemann sphere _ , ( 45 - 70 ) , eur . math .
( @xmath696 ) .
m. shishikura , bifurcation of parabolic fixed points , _ the mandelbrot set , theme and variations , _ ( 325 - 363 ) , _ london math .
lecture note ser .
, 274 _ cambridge univ . press , ( 2000 ) . | in this paper we study families of degree @xmath0 parabolic - like mappings @xmath1 ( as we defined in @xcite ) .
we prove that the hybrid conjugacies between a nice analytic family of degree @xmath0 parabolic - like mappings and members of the family @xmath2 induce a continuous map @xmath3 , which under suitable conditions restricts to a ramified covering from the connectedness locus of @xmath1 to the connectedness locus @xmath4 of @xmath2 . as an application
, we prove that the connectedness locus of the family @xmath5 presents baby @xmath6 . |
modulation level classification ( mlc ) is a process which detects the transmitter s digital modulation level from a received signal , using a priori knowledge of the modulation class and signal characteristics needed for downconversion and sampling . among many modulation classification methods @xcite ,
a cumulant ( cm ) based classification @xcite is one of the most widespread for its ability to identify both the modulation class and level .
however , differentiating among cumulants of the same modulation class , but with different levels , i.e. 16qam vs. 64qam , requires a large number of samples .
a recently proposed method @xcite based on a goodness - of - fit ( gof ) test using kolmogorov - smirnov ( ks ) statistic has been suggested as an alternative to the cm - based level classification which require lower number of samples to achieve accurate mlc . in this letter
, we propose a novel mlc method based on distribution distance functions , namely kuiper ( k ) @xcite ( * ? ? ?
3.1 ) and ks distances , which is a significant simplification of methods based on gof .
we show that using a classifier based only on k - distance achieves better classification than the ks - based gof classifier . at the same time
, our method requires only @xmath0 additions in contrast to @xmath1 additions for the ks - based gof test , where @xmath2 is the number of distinct modulation levels , @xmath3 is the sample size and @xmath4 is the number of test points used by our method .
following @xcite , we assume a sequence of @xmath3 discrete , complex , i.i.d . and sampled baseband symbols , @xmath5 $ ] , drawn from a modulation order @xmath6 , transmitted over awgn channel , perturbed by uniformly distributed phase jitter and attenuated by an unknown factor @xmath7 . therefore , the received signal is given as @xmath8 $ ] , where @xmath9 , @xmath10 and @xmath11 .
the task of the modulation classifier is to find @xmath12 , from which @xmath13 was drawn , given @xmath14 .
without loss of generality , we consider unit power constellations and define snr as @xmath15 .
the proposed method modifies mlc technique based on gof testing using the ks statistic @xcite .
since the ks statistic , which computes the minimum distance between theoretical and empirical cumulative distribution function ( ecdf ) , requires _ all _ cdf points , we postulate that similarly accurate classification can be obtained by evaluating this distance using a smaller set of points in the cdf .
let @xmath16=f(\mathbf{r})$ ] where @xmath17 is the chosen feature map and @xmath18 is the number extracted features .
possible feature maps include @xmath19 ( magnitude , @xmath20 ) or the concatenation of @xmath21 and @xmath22 ( quadrature , @xmath23 ) .
the theoretical cdf of @xmath24 given @xmath12 and @xmath25 , @xmath26 , is assumed to be known a priori ( methods of obtaining these distributions , both empirically and theoretically , are presented in ( * ? ? ?
iii - a ) ) . the @xmath2 cdfs , one for each modulation level , define a set of test points @xmath27 with the distribution distances given by @xmath28 for @xmath29 , @xmath30 , and @xmath31 , corresponding to the maximum positive and negative deviations , respectively .
note the symmetry in the test points such that @xmath32 .
thus , there are @xmath33 test points for a @xmath2 order classification .
the ecdf , given as @xmath34 is evaluated at the test points to form @xmath35 , @xmath36 . here
, @xmath37 equals to one if the input is true , and zero otherwise . by evaluating @xmath38 only at the test points in ( [ eq : testpoints ] ) , we get @xmath39 which are then used to find an estimate of the maximum positive and negative deviations @xmath40 of the ecdf to the true cdfs .
the operation of finding the ecdf at the given testpoints ( [ eq : dij ] ) can be implemented using a simple thresholding and counting operation and does not require samples to be sorted as in @xcite . the metrics in ( [ eq : dj ] ) are used to find the final distribution distance metrics @xmath41 which are the reduced complexity versions of the ks distance ( rcks ) and the k distance ( rck ) , respectively .
finally , we use the metrics in ( [ eq : metrics ] ) as substitutes to the true distance - based classifiers with the following rule : choose @xmath42 such that @xmath43 in the remainder of the letter , we define @xmath44 and @xmath45 , where @xmath46 .
let @xmath47 $ ] denote the set of test points , @xmath48 , sorted in ascending order . for notational consistency
, we also define the following points , @xmath49 and @xmath50 . given that these points are distinct , they partition @xmath51 into @xmath52 regions
. an individual sample , @xmath53 , can be in region @xmath54 , such that @xmath55 , with a given probability , determined by @xmath26 . assuming @xmath53 are independent of each other
, we can conclude that given @xmath51 , the number of samples that fall into each of the @xmath52 regions , @xmath56 $ ] , is jointly distributed according to a multinomial pmf given as @xmath57 where @xmath58 $ ] , and @xmath59 is the probability of an individual sample being in region @xmath54 . given that @xmath51 is drawn from @xmath12 , @xmath60 , for @xmath61 .
now , with particular @xmath62 , the ecdf at all the test points is @xmath63,\quad f_n(t_l ) = { \frac{1}{n } } \sum\limits_{i=1}^l n_i.\ ] ] therefore , we can analytically find the probability of classification to each of the @xmath2 classes as @xmath64 for the rck classifier .
a similar expression can be applied to rcks , replacing @xmath65 with @xmath66 in ( [ eq : probabilityofclassification ] ) .
given that the theoretical cdfs change with snr , we store distinct cdfs for @xmath67 snr values for each modulation level ( impact of the selection of @xmath67 on the accuracy is discussed further in section [ sec : detection_samples ] . )
further , we store @xmath68 theoretical cdfs of length @xmath69 each . for the non - reduced complexity classifiers that require sorting samples ,
we use a sorting algorithm whose complexity is @xmath70 . from table [
table : complexity ] , we see that for @xmath71 rck / rcks tests use less addition operations than k / ks - based methods @xcite and cm - based classification @xcite . for @xmath72 ,
the rck method is more computationally efficient when implemented in asic / fpga , and is comparable to cm in complexity when implemented on a cpu .
in addition , the processing time would be shorter for an asic / fpga implementation , which is an important requirement for cognitive radio applications .
furthermore , their memory requirements are also smaller since @xmath69 has to be large for a smooth cdf .
it is worth mentioning that the authors in @xcite used the theoretical cdf , but used @xmath69 as the number of samples to generate the cdf in their complexity figures .
the same observation favoring the proposed rck / rcks methods holds for the magnitude - based ( mag ) classifiers ( * ? ? ?
* sec iii - a ) .
.number of operations and memory usage [ cols="<,<,<,<",options="header " , ] [ table : complexity ]
as an example , we assume that the classification task is to distinguish between m - qam , where @xmath73 . for comparison we also present classification result based on maximum likelihood estimation ( ml ) .
in the first set of experiments we evaluate the performance of the proposed classification method for different values of snr .
the results are presented in fig .
[ fig : vary_snr ] .
we assume fixed sample size of @xmath74 , in contrast to ( * ? ? ?
1 ) to evaluate classification accuracy for a smaller sample size . we confirm that even for small sample size , as shown in ( * ? ? ?
1 ) , cm has unsatisfying classification accuracy at high snr . in
( 10,17)db region rck clearly outperforms all detection techniques , while as snr exceeds @xmath7517db all classification methods ( except cm ) converge to one . in low snr region , ( 0,10)db , ks , rcks , rck perform equally well , with cm having comparable performance .
the same observation holds for larger sample sizes , not shown here due to space constraints .
note that the analytical performance metric developed in section [ sec : analysis ] for rck and rcks matches perfectly with the simulations .
for the remaining results , we set @xmath76db , unless otherwise stated .
= 50 ; ( an . )
analytical result using ( [ eq : probabilityofclassification ] ) . ] in the second set of experiments , we evaluate the performance of the proposed classification method as a function of sample size @xmath3 .
the result is presented in fig .
[ fig : vary_n ] . as observed in fig .
[ fig : vary_snr ] , also here cm has the worst classification accuracy , e.g. 5% below upper bound at @xmath77 .
the rck method performs best at small sample sizes , @xmath78 . with @xmath79 ,
the accuracy of rck and ks is equal .
classification based on rcks method consistently falls slightly below rck and ks methods .
in general , rcks , rck and ks converge to one at the same rate .
db . ] in the third set of experiments we evaluate the performance of the proposed classification method as a function of snr mismatch and phase jitter .
the result is presented in fig .
[ fig : mismatch ] . in case of snr mismatch ,
[ fig : vary_gamma ] , our results show the same trends as in ( * ? ? ? * fig .
4 ) ; that is , all classification methods are relatively immune to snr mismatch , i.e. the difference between actual and maximum snr mismatch is less than 10% in the considered range of snr values .
this justifies the selection of the limited set of snr values @xmath67 for complexity evaluation used in section [ sec : complexity ] .
as expected , ml shows very high sensitivity to snr mismatch . note
again the perfect match of analytical result presented in section [ sec : analysis ] with the simulations . in the case of phase jitter caused by imperfect downconversion , we present results in fig .
[ fig : vary_phase ] for @xmath80db as in @xcite , in contrast to @xmath76db used earlier , for comparison purposes .
we observe that our method using the magnitude feature , rck / rcks ( mag ) , as well as the cm method , are invariant to phase jitter .
rck and rcks perform almost equally well , while cm is worse than the other three methods by @xmath7510% .
as expected , the ml performs better than all other methods .
quadrature - based classifiers , as expected , are highly sensitive to phase jitter .
note that in the small phase jitter , @xmath81 , quadrature - based classifiers perform better than others , since the sample size is twice as large as in the former case .
in this letter we presented a novel , computationally efficient method for modulation level classification based on distribution distance functions .
specifically , we proposed to use a metric based on kolmogorov - smirnov and kuiper distances which exploits the distance properties between cdfs corresponding to different modulation levels .
the proposed method results in faster mlc than the cumulant - based method , by reducing the number of samples needed .
it also results in lower computational complexity than the ks - gof method , by eliminating the need for a sorting operation and using only a limited set of test points , instead of the entire cdf .
o. a. dobre , a. abdi , y. bar - ness , and w. su , `` survey of automatic modulation classification techniques : classical approaches and new trends , '' _ iet communications _ , vol . 1 , no . 2 , pp .
137156 , apr .
2007 .
g. a. p. cirrone , s. donadio , s. guatelli , a. mantero , b. masciliano , s. parlati , m. g. pia , a. pfeiffer , a. ribon , and p. viarengo , `` a goodness of fit statistical toolkit , '' _ ieee trans .
_ , vol .
51 , no . 5 ,
pp . 20562063 , oct . | we present a novel modulation level classification ( mlc ) method based on probability distribution distance functions .
the proposed method uses modified kuiper and kolmogorov - smirnov distances to achieve low computational complexity and outperforms the state of the art methods based on cumulants and goodness - of - fit tests .
we derive the theoretical performance of the proposed mlc method and verify it via simulations .
the best classification accuracy , under awgn with snr mismatch and phase jitter , is achieved with the proposed mlc method using kuiper distances . |
neutrino oscillation physics has entered the precision era .
present data determines all three leptonic mixing angles and the absolute value of the two mass - squared differences with few percent precision @xcite .
crucial goals of future oscillation experiments are ( @xmath1 ) the determination of the neutrino mass ordering and the cp - violating phase @xmath2 , and ( @xmath3 ) establishing the robustness of three - flavour oscillations with respect to physics beyond the standard model ( sm ) . in the present work
we show that those two items are intimately related .
we consider the hypothesis that additional interactions affect the neutrino sector , beyond the sm weak interaction @xcite , see @xcite for recent reviews .
we will show that , for a certain choice of these non - standard interactions ( nsi ) , the determination of the neutrino mass ordering
one of the main goals of upcoming oscillation experiments @xcite becomes impossible , due to an exact degeneracy in the evolution equation governing neutrino oscillations in matter .
the paper is structured as follows . in sec .
[ sec : nsi ] we introduce the nsi framework and the notation used in the rest of the paper .
section [ sec : deg ] shows the origin of the degeneracy and how it can be realized in both vacuum and matter regimes . in sec .
[ sec : osc ] we explain how the degeneracy affects neutrino oscillation data , while in sec .
[ sec : scattering ] we explore the possible combination with neutrino scattering data to try to remove the degeneracy .
finally , our conclusions are summarized in sec .
[ sec : conclusions ] .
three - flavour neutrino evolution in an arbitrary matter potential is described by the schroedinger equation @xmath4 where @xmath5 is a vector of the flavour amplitudes , @xmath6 , and @xmath7 .
the hamiltonian describing evolution in vacuum is @xmath8 with @xmath9 , where @xmath10 stands for the neutrino mass - squared difference , and @xmath11 is the neutrino energy . from neutrino oscillation data , we know that @xmath12
. the neutrino mass ordering is parametrized by the sign of the larger mass - squared difference , with normal ordering ( no ) corresponding to @xmath13 and inverted ordering ( io ) to @xmath14 .
the sign of @xmath15 by convention is chosen positive .
the standard parametrization for the leptonic mixing matrix is @xmath16 , where @xmath17 ( @xmath18 ) denotes a real ( complex ) rotation in the @xmath19 sector , with mixing angle @xmath20 .
here we find it convenient to use an equivalent parametrization , where we put the complex phase @xmath2 in the 12 rotation , such that @xmath21 . after subtracting a term proportional to the unit matrix , the vacuum hamiltonian becomes @xmath22 with the 12 block given by @xmath23 let us consider now the presence of neutral - current ( nc ) nsi in the form of dimension-6 four - fermion operators , which may contribute to the effective potential in matter in @xmath24 .
we follow the notation of @xcite , for a recent review see _ e.g.
_ @xcite .
nsi are described by the lagrangian @xmath25 where , @xmath26 , and @xmath27 denotes a fermion present in the background medium .
the parameter @xmath28 parametrizes the strength of the new interaction with respect to the fermi constant @xmath29 .
hermiticity requires that @xmath30 .
note that we restrict to vector interactions , since we are interested in the contribution to the effective matter potential .
in generic models of new physics nsi parameters are expected to be small .
however , examples of viable gauge models leading to @xmath31 can be found in @xcite ( see also @xcite for a discussion of nsi models ) .
the matter part of the hamiltonian is then obtained as @xmath32 \epsilon_{\alpha\beta } & = \sum_{f = e , u , d } y_f(x ) \epsilon_{\alpha\beta}^f \ , , \label{eq : eps}\end{aligned}\ ] ] with @xmath33 , @xmath34 being the density of fermion @xmath27 along the neutrino path .
this implies that the effective nsi parameters @xmath35 may depend on @xmath36 .
the `` 1 '' in the @xmath37 entry in eq
. corresponds to the standard matter potential @xcite . for neutral matter ,
the densities of electrons and protons are equal .
thus , the relative densities of up and down quarks are @xmath38 where @xmath39 is the relative neutron density along the neutrino path . below we will use the notation @xmath40 and @xmath41 to indicate when the @xmath35 refer to the specific matter composition of the earth or the sun , respectively .
let us consider first the vacuum part of the hamiltonian , @xmath42 defined in eqs . and .
it is easy to show that the transformation @xmath43 implies that @xmath44 . inserting this into eq . and
taking the complex conjugate we recover exactly the same evolution equation , when we take into account that complex conjugation of the amplitudes ( @xmath45 ) is irrelevant , as only moduli of flavour amplitudes are observable .
is a consequence of cpt invariance .
it has been noted in the context of nsi in @xcite and applied in some limiting cases , see also @xcite . ]
this proves that the transformation leaves the three - flavour evolution in vacuum invariant .
note that this transformation corresponds to a complete inversion of the neutrino mass spectrum .
the transformation @xmath46 exchanges no and io , while changing the octant of @xmath0 exchanges the amount of @xmath47 present in @xmath48 and @xmath49 . we denote the effect of the transformation as `` flipping '' the mass spectrum .
the corresponding degeneracy is known in limiting cases , for instance , the so - called mass ordering degeneracy in the context of long - baseline experiments @xcite .
it is manifest also in the exact expression for the three - flavour @xmath47 survival - probability @xmath50 in vacuum , relevant for medium - baseline reactor experiments @xcite .
it is clear that for a non - zero standard matter effect , eq . with @xmath51 ,
the transformation no longer leaves the evolution invariant , since @xmath24 remains constant .
the matter effect in the 13-sector is the basis of the mass ordering determination in long - baseline accelerator @xcite and atmospheric neutrino @xcite experiments .
moreover , the observation of the msw @xcite matter resonance in the sun requires that @xmath52 , which forbids the transformation in the second line of eq . .
this allows , in principle , for the determination of the mass ordering via a precise measurement of @xmath50 in vacuum @xcite , as intended for instance by the juno collaboration @xcite .
however , if in addition to the transformation , it is also possible to transform @xmath53 , then the full hamiltonian including matter would transform as @xmath54 , leaving the evolution equation invariant . this can be achieved in presence of nsi , supplementing the transformation with @xcite @xmath55 the transformation of @xmath56 is crucial to change the sign of the @xmath37 element of @xmath24 including the standard matter effect .
note that eq .
depends on the parametrization used for @xmath42 in eq . .
if the standard parametrization with @xmath57 was used instead , then we would obtain @xmath58 , @xmath59 , and all other @xmath35 transforming as in eq . .
since in general the @xmath35 are dependent on the neutron density , the degeneracy can be broken in principle by comparing experiments in matter with different neutron densities , or in configurations where the neutron density changes significantly along the neutrino path .
however , one can choose couplings such that nsi with neutrons are zero and take place only with protons and/or electrons , by choosing @xmath60 proportional to the quark electric charge , _
i.e. _ , @xmath61 . in this situation
the @xmath35 are always independent of @xmath36 , the degeneracy is complete and can not be broken by any combination of neutrino oscillation experiments .
let us illustrate the degeneracy by the following example : assume that there are no nsi in nature .
then we can fit data from any neutrino oscillation experiment either with standard oscillations and the correct spectrum , or equally well with a flipped spectrum and @xmath62 .
for @xmath63 we obtain @xmath62 independent of the neutron density , and hence the degeneracy will be perfect , irrespective of the matter environment .
a manifestations of this result is the so - called lma - dark solution for solar neutrinos @xcite , which corresponds to a fit to solar neutrino data with @xmath64 ( `` dark octant '' ) and values of @xmath65 . in the sun
the neutron fraction @xmath66 drops from about 1/2 in the centre to about 1/6 at the border of the solar core . from eqs . and follows , that for @xmath65 we obtain @xmath67 , close to the value needed for the generalized mass ordering degeneracy . in @xcite
a recent analysis of solar neutrino data has been performed , assuming either nsi with up or down quarks . in this case
@xmath66 does not drop out of @xmath35 defined in eq . , and
hence the condition @xmath68 can not be fulfilled along the whole neutrino path in the sun .
therefore , the degeneracy is not perfect . in @xcite
the @xmath69 of the lma - dark solution is nearly zero for nsi on up quarks and @xmath70 for down quarks .
while the sign of @xmath71 is irrelevant for solar neutrino phenomenology , it has been realised in @xcite , that the @xmath72 ambiguity introduced by the lma - dark solution leads to a mass ordering ambiguity in the planned reactor experiment juno @xcite .
this is a manifestation of the generalized degeneracy discussed above . as another example
, we will now demonstrate the impact of the generalized degeneracy on the sensitivity of the long - baseline deep underground neutrino experiment ( dune ) @xcite to the mass ordering . in absence of nsi
, the dune experiment would be able to reject the wrong hypothesis for the mass ordering with a significance above @xmath73 regardless of the true value of @xmath2 @xcite .
we calculate expected data for no , @xmath74 , @xmath75 , and no nsi .
the simulation is performed using globes @xcite ; the simulation details are the same as in ref .
. then these artificial data are fitted by allowing for the simultaneous presence of @xmath56 and @xmath76 , while all other nsi parameters are set to zero , for simplicity .
results are shown by the shaded regions in fig .
[ fig : vplot ] .
the lower panel confirms the perfect degeneracy of the flipped mass spectrum at @xmath77 and @xmath78 , with @xmath79 with respect to the true best fit point . in both panels of fig .
[ fig : vplot ] we observe also a strong correlation between @xmath80 and @xmath81 , see @xcite .
therefore , while the degeneracy is exact for @xmath82 , it is recovered to a good accuracy for nonzero values of @xmath81 as long as @xmath83 .
the importance of @xmath80 and @xmath81 for the mass ordering determination in dune has been pointed out recently in @xcite . and @xmath81 .
in the upper panel we fit with the correct mass spectrum , while in the lower panel we adopt io and exchange @xmath72 .
the shaded regions correspond to dune alone , whereas the contour curves include the constraints from global oscillation data @xcite and from the charm experiment @xcite on the nc cross section .
we marginalize over @xmath84 , @xmath2 , @xmath85 and the complex phase of @xmath81 . ]
since the generalized degeneracy is exact and holds for any oscillation experiment , the only way to break it are non - oscillation experiments .
indeed , operators of the type in eq .
contribute to the neutral current ( nc ) neutrino scattering cross section .
unfortunately , data on electron neutrino nc scattering is scarce .
a relevant constraint on the parameters of interest to us comes from the historical charm experiment @xcite , which has measured the quantity @xmath86 , where @xmath87 is ratio of the electron neutrino plus antineutrino nc cross sections to the corresponding charged current ones . in presence of nsi
we have @xmath88 , where @xcite @xmath89 \,,\ ] ] with @xmath90 , and @xmath91 are the sm nc couplings .
we have included only the vector - like nsi parameters .
note that the charm constraint is somewhat model dependent , since it would not apply if the mediator particle responsible for the nsi is much lighter than the momentum transfer in charm ( typically of several tens of gev ) @xcite .
( 2 d.o.f . ) from charm @xcite ( assuming nsi from a heavy mediator ) in the plane of @xmath92 and @xmath93 , for @xmath94 .
the cross corresponds to the point of exact degeneracy , eq . .
the diagonal lines indicate the parameters for which @xmath68 in earth and solar matter . ] assuming that the charm bound applies , we follow ref .
@xcite and we show in fig .
[ fig : charm ] the allowed region for @xmath92 and @xmath93 from the charm data .
the point from eq .
, corresponding to perfect degeneracy for any matter profile , is indicated by the cross in the figure .
we observe that it is excluded by charm data : for this point we predict @xmath95 , which disagrees with the charm experimental value at @xmath96 .
the diagonal lines in the figure indicate the parameters for which @xmath68 in earth and solar matter .
we use that in earth matter , @xmath97 , and for the sun we show the spread induced by @xmath98 . for neutrino trajectories in the earth ,
the generalized degeneracy holds along the line indicated in the plot .
however , since the degeneracy for the sun appears for slightly different values of @xmath99 there is the potential to break it by the combination . indeed , for the lma - dark solution for nsi either on up or down
quarks we have @xmath100 @xcite . from fig .
[ fig : charm ] we see that @xmath101 is in strong disagreement with charm , while @xmath102 is within the @xmath103 region .
let us therefore adopt the hypothesis of nsi with up quarks only and check whether solar neutrino and charm data could break the mass ordering degeneracy for dune .
the contour curves in fig .
[ fig : vplot ] show the combined analysis , where we include global oscillation data ( including the solar neutrino `` sno - poly '' analysis ) from @xcite , assuming that @xmath92 and @xmath104 are approximately uncorrelated .
if we restrict @xmath78 , the degeneracy is broken , since @xmath105 ( as required by solar data ) implies @xmath106 , which can be excluded at high confidence level by dune for @xmath107 . however ,
if we allow for non - zero @xmath81 , we see that a large region with the flipped mass spectrum remains below the @xmath108 level around @xmath106 and @xmath109 .
hence , we conclude that including present constraints from oscillation and scattering data , the degeneracy will severely affect the mass ordering sensitivity of dune .
let us note that if more nsi parameters are allowed to vary , the fit with the flipped spectrum may even improve further .
we have demonstrated that the so - called lma - dark solution is a manifestation of an exact degeneracy at the level of the neutrino evolution equation .
this degeneracy makes it impossible to determine the neutrino mass ordering by neutrino oscillation experiments .
it requires nsi comparable in strength to weak interactions .
the only way to break the degeneracy is via non - oscillation experiments . we have shown that taking into account current data on the @xmath47 nc cross section excludes nsi needed for the exact generalized degeneracy ( subject to some model dependence ) ; however , the degeneracy remains to be present at an approximate level , still destroying the mass ordering sensitivity of planned experiments . in order to break
the degeneracy at high confidence level , improved data on @xmath47 nc interactions is mandatory .
these may be provided , for instance , by coherent neutrino
nucleus interaction experiments @xcite
. _ acknowledgments . _
we thank stefan vogl and david v. forero for useful discussions .
fermilab is operated by fermi research alliance , llc under contract no . with the united states department of energy .
this project has received funding from the european union s horizon 2020 research and innovation programme under the marie sklodowska - curie grant agreement no 674896 .
note that flavour evolution in oscillation experiments is only sensitive to differences of the diagonal elements of the hamiltonian .
therefore , eq . should be replaced by the more general expression : @xmath110 indeed , the degeneracy can also be realized for zero @xmath56 , shifting the @xmath111-term to the @xmath112 , @xmath113 entries in the potential . in the above analysis including charm data
we have implicitly assumed @xmath114 .
the assumption @xmath115 is motivated by strong limits from nutev @xcite , which should apply under the assumptions of heavy mediator particles as relevant for charm .
since oscillation experiments constrain @xmath116 @xcite , the combination of oscillation and nutev data implies @xmath117 as well . we have performed a detailed study of the generalized mass ordering degeneracy using a combination of oscillation plus scattering data in ref .
@xcite . in general , in order to exclude the degeneracy we need data from scattering experiments on both , electron - neutrino as well as muon - neutrino nc scattering . | we consider the impact of neutral - current ( nc ) non - standard neutrino interactions ( nsi ) on the determination of the neutrino mass ordering .
we show that in presence of nsi there is an exact degeneracy which makes it impossible to determine the neutrino mass ordering and the octant of the solar mixing angle @xmath0 at oscillation experiments .
the degeneracy holds at the probability level and for arbitrary matter density profiles , and hence , solar , atmospheric , reactor , and accelerator neutrino experiments are affected simultaneously .
the degeneracy requires order - one corrections from nsi to the nc neutrino
quark interaction and can be tested in neutrino
nucleus nc scattering experiments .
fermilab - pub-16 - 115-t |
let @xmath5 be a sequence of independent @xmath6 random matrices and @xmath7 in a celebrated paper furstenbergkesten60 , furstenberg and kesten proved that @xmath8 converges provided that @xmath9 later , oseledec in @xcite proved convergence for other singular values of @xmath10 and cohen and newman in @xcite studied the behavior of the limit in the situation when @xmath1 approaches infinity .
this paper investigates the question of how the speed of convergence depends on the dimension of matrices @xmath1 .
consider a dynamical system ( a gas , an economy , an ecosystem , etc . ) .
its evolution can be described by a mapping @xmath11 where @xmath12 is a vector that describes the state of the system at time @xmath13 we can often model the mapping as a multiplication by a random matrix @xmath14 stability and other long - run properties of the system depend on the growth in the norm of the product @xmath15 which we can measure by calculating the quantity @xmath16 the sub - multiplicativity property of the norm ( @xmath17 ) ensures that @xmath18 converges to @xmath19 where @xmath20 is an arbitrary vector .
intuitively , this means that it is not important what was the starting vector of the system . after some time , all products grow at the same rate independently of the initial state .
it is of interest to investigate whether this erasure of memory about the initial state occurs slower in more complex systems , that is , in systems , which are described by matrices of larger size .
of course , when we compare long - run properties of systems , we should only look at the systems that are comparable in the short run , that is , the system that have comparable one - step behavior .
roughly , the difference between one - step growth of a specially - chosen and a random vector can be measured by the ratio of @xmath21 to @xmath22 where @xmath1 is the dimension of the matrix @xmath23 indeed , @xmath21 is the square of the maximal possible increase in the length of the state vector , and @xmath24 is the average of the squared singular values of @xmath25 hence it can be considered as a measure of the increase in the length of a random state vector . hence , if we want systems to be comparable in the short run , then we should restrict this ratio by a constant that does not depend on the dimension of the system .
( otherwise , some directions may become more and more unusual as the dimension of the system grows . )
we will call this property uniform boundedness of singular values .
we also want to look at sufficiently symmetric systems , that is , systems without preferential directions .
we codify this by requiring that matrices @xmath26 are rotationally invariant , that is , the distribution of matrix elements does not depend on the choice of basis .
the main result of this paper is that under these assumptions the speed with which the memory of the initial state is erased does not decrease as the dimension of the system grows . intuitively , the asymptotic behavior of @xmath27 depends on three factors .
first of all , for a fixed vector @xmath28@xmath29for a certain sequence of vectors @xmath30 and this averaging is likely to concentrate the distribution of @xmath31 this factor does not depend on the dimension @xmath1 . on the other hand , we are interested in the convergence of the supremum of @xmath32 over all @xmath33 and to ensure the convergence of this supremum we have to make sure that variables @xmath32 are all close to the limit @xmath34 for a sufficiently dense set of vectors @xmath35 the number of elements in such a set is likely to grow exponentially in @xmath36 and this might make the convergence of @xmath37 slower for large @xmath38 the third factor appears because for every fixed vector @xmath28 the norm @xmath39 becomes concentrated around some particular value as @xmath40 this factor is likely to speed up the convergence of @xmath41 and therefore of @xmath42 we will show in this paper that the third factor dominates and the speed of convergence of @xmath37 is not slowed down by the growth in the dimension @xmath1 .
previously , the speed of convergence in the furstenberg
kesten theorem was investigated in @xcite and guivarchraugi85 .
they proved a central limit theorem for @xmath43 and studied large deviations of @xmath8 for a large class of random matrices .
however , the results in these papers do not provide effectively computable bounds on the rate of convergence in limit theorems , and , as a consequence , do not help us to investigate how the speed of convergence changes as the dimension of matrices grows .
one of the contributions of this paper is deriving more explicit bounds on the speed of convergence in limit theorems .
let us describe the problem in a more formal fashion .
consider independent identically - distributed @xmath1-by-@xmath1 matrices @xmath44 we are interested in the behavior of the norm of the product @xmath45 and we will make the following assumptions about matrices @xmath46 .
first of all , we assume that random matrices @xmath46 are rotationally invariant ; that is , the distribution of their entries does not depend on the choice of coordinates .
formally , we use the following definition .
a random matrix @xmath47 is rotationally invariant if for every integer @xmath48 for every collection of vectors @xmath49 , @xmath50 and for every orthogonal matrix @xmath51 the joint distributions of random vectors @xmath52 and @xmath53 are the same .
[ aa ] matrices @xmath54 are rotationally invariant .
we also impose an assumption needed for the validity of the furstenberg kesten theorem .
[ ab ] for all @xmath36 @xmath55 exists .
second , we restrict our study to two important cases .
the first one is the case of ( real ) gaussian matrices @xmath46 , that is , independent random @xmath1-by-@xmath1 matrices with independent entries distributed according to the gaussian distribution with zero mean and variance @xmath56 that is , as @xmath57 the second case is that of independent rotationally invariant @xmath1-by-@xmath1 matrices @xmath46 that satisfy the following assumptions .
let @xmath58 be the eigenvalues of @xmath59 ( i.e. , squared singular values of @xmath46 ) , and let @xmath60(we will sometimes omit superscripts to lighten the notation . )
[ ac ] with probability 1 , @xmath61 where the constant @xmath62 does not depend on @xmath38 in other form , assumption [ ac ] says that @xmath63 with probability 1 .
[ ad ] @xmath64 $ ] exists and bounded by a constant which does not depend on @xmath38 one example of a matrix family that satisfies these assumptions is hermitian matrices @xmath46 which are generated in the following way .
sample @xmath1 independent values from a distribution supported on @xmath65 , $ ] where @xmath66 and construct a diagonal matrix @xmath67 by putting these values on the main diagonal .
then take a haar - distributed random orthogonal matrix @xmath68 and define @xmath46 as @xmath69 a sequence of these matrices ( with independent @xmath70 ) will satisfy all the assumptions . the main result is as follows .
[ theorem_uniform_convergence_norm_products]let @xmath71 be independent , identically distributed @xmath6 matrices , which satisfy assumptions [ aa ] and [ ab ] and which are either gaussian with independent entries @xmath72 or satisfy assumptions [ ac ] and [ ad ] .
let @xmath73 and let @xmath74 be an arbitrary unit vector .
then @xmath37 converges in probability to @xmath75 and the convergence is uniform in @xmath38 that is , for each @xmath76 there exists an @xmath77 such that for all @xmath78 and all @xmath79 @xmath80 the assumptions of the theorem are sufficient but not necessary .
the assumption that @xmath81 is used in the proof of proposition [ proposition_ld_inequality_unbounded_rv ] below , where it is used to estimate the probability of large deviations of @xmath82 and to show that the rate in the corresponding exponential inequality is proportional to @xmath38 it is likely that this assumption can be somewhat relaxed by requiring instead that @xmath83 .
one particular implication of the assumption @xmath81 is that the bound on singular values does not depend on the dimension of the matrix . in order to understand this assumption better ,
consider the following example .
let @xmath84where @xmath85 is a haar - distributed row @xmath1-dimensional vector , and @xmath86 is a haar - distributed column @xmath1-dimensional vector .
( vectors @xmath85 and @xmath87 are assumed to be independent . )
then the squared singular values of @xmath5 are all zero except one , which equals @xmath38 hence , @xmath88 and @xmath89 we can conclude that assumptions [ aa ] , [ ab ] and [ ad ] are satisfied , and assumption [ ac ] is not satisfied . next , consider @xmath90 where @xmath74 is an arbitrary vector .
it is easy to see that @xmath91 is distributed as the first coordinate of a haar - distributed vector @xmath92 in other words @xmath91 is distributed as @xmath93where @xmath94 are independent standard gaussian variables .
then it is clear that @xmath95 next , let us compute @xmath96 note that @xmath97and @xmath98 hence , @xmath99where @xmath91 are independent and distributed as @xmath100 above .
hence , @xmath91 converges in distribution to @xmath101 as @xmath40 it is clear that @xmath102 in probability as @xmath103 therefore , for large @xmath36 @xmath104this bias term can not be made small uniformly in @xmath1 by an increase in @xmath105 this means that the claim of theorem theorem_uniform_convergence_norm_products fails in this case .
later , in section [ section_necessity ] , we will prove a necessary condition for the uniform convergence by using the basic idea of this example . in order to understand the role of the rotational invariance assumption ,
consider the following example .
let @xmath5 be independent , identically distributed , diagonal matrices .
the diagonal elements of a matrix @xmath5 are independent bernoulli variables that take values @xmath106 and @xmath62 .
that is , a diagonal element takes the value @xmath107 with probability @xmath108 and the value @xmath109 with probability @xmath110 assume that @xmath111 it is easy to see that the norm of @xmath112 is given by the following expression:@xmath113where @xmath114 and @xmath115 are independent random variables with the binomial distribution @xmath116 taking the logarithm and dividing by @xmath117 we get @xmath118where @xmath119 note that as @xmath120 grows , each @xmath121 approaches the gaussian distribution @xmath122 if @xmath1 is fixed , then @xmath123 however , if @xmath1 grows simultaneously with @xmath117 then the limit of @xmath37 may be nonexistent , or may depend on the speed of growth in @xmath1 relative to the speed of growth in @xmath105 hence , the conclusion of theorem theorem_uniform_convergence_norm_products is invalid in this case .
it is an interesting problem whether the assumption of rotational invariance can be relaxed so that the result in theorem theorem_uniform_convergence_norm_products holds for a larger class of matrices , for example , for matrices with i.i.d .
non - gaussian entries ( i.e. , wigner matrices )
. however , this problem appears to be hard since at this moment very little is known about effective bounds on the rate of convergence in the furstenberg
kesten theorem .
let me now explain two results which will be used as tools in the proof of theorem [ theorem_uniform_convergence_norm_products ] .
the proofs of these results will be given in later sections .
our main tool is the following proposition .
[ proposition_ld_inequality_unbounded_rv ] suppose that all @xmath5 are gaussian with independent entries @xmath124 .
then for all sufficiently small @xmath125 , all @xmath126 and all @xmath127 @xmath128 suppose that i.i.d .
@xmath1-by-@xmath1 matrices @xmath5 are rotationally invariant and satisfy assumption [ ac ] with constant @xmath129 let @xmath130then for all @xmath131 , all @xmath132 and all @xmath127 @xmath133 in its essence , proposition [ proposition_ld_inequality_unbounded_rv ] is a large deviation result which quantifies the speed of convergence of @xmath134 for a fixed vector @xmath35 its main point is that the rate in this large deviation estimate is proportional to the dimension @xmath1 .
the proof of this proposition will be given in section [ section_large_deviation_bound ] .
the other tool is as follows .
let a set of points on the unit sphere in @xmath135 be called an @xmath136-net if the sphere is covered by spherical caps with centers at these points and angular radius @xmath137 [ proposition_norm_and_sup]let @xmath138 be an arbitrary @xmath1-by-@xmath1 matrix .
suppose that the endpoints of vectors @xmath30 form an @xmath139-net of the unit sphere in @xmath135 . then for all sufficiently small @xmath139 , @xmath140 this proposition allows us to control the matrix norm @xmath141 by the norms of vectors @xmath142 where @xmath30 runs through a finite set of values .
proof of proposition [ proposition_norm_and_sup ] let @xmath30 be a vector in the net which is closest to a unit vector @xmath35 then@xmath143taking the supremum over @xmath28 we obtain that @xmath144hence , @xmath145and the claim of the proposition follows .
this proposition is useful in conjunction with the following result about the size of sphere coverings . by lemma 2.6 on page 7 of milmanschechtman01 , for @xmath139 smaller than a certain constant
, there exists an @xmath139-net with cardinality @xmath146 now let us prove theorem [ theorem_uniform_convergence_norm_products ] by using propositions [ proposition_ld_inequality_unbounded_rv ] and proposition_norm_and_sup .
proof of theorem [ theorem_uniform_convergence_norm_products ] we focus on the case when assumptions [ ac ] and [ ad ] hold .
the proof for the case of gaussian matrices goes along a similar route and it is simpler .
first of all , note that is enough to prove that ( [ inequality_theorem ] ) holds for all sufficiently large @xmath36 that is , for all @xmath147 indeed , for each @xmath148 we can apply results in @xcite and find that inequality ( [ inequality_theorem ] ) holds if @xmath149 hence , inequality ( inequality_theorem ) holds for all @xmath150 provided that @xmath151 we will choose an appropriate @xmath152 later .
we are going to prove that for all sufficiently large @xmath1 and @xmath120 [ i.e. , @xmath153 and all @xmath154 , it is true that @xmath155 let vectors @xmath156 @xmath157 form a @xmath158-net on the unit sphere .
then , by using propositions proposition_norm_and_sup and [ proposition_ld_inequality_unbounded_rv ] , the union bound and the estimate on the number of elements in the net we obtain @xmath159 where @xmath160 is a certain constant .
clearly , we can choose @xmath161 in such a way that for all @xmath162 it is true that @xmath163for some @xmath164 and then choose @xmath165 such that for all @xmath166 it is true that@xmath167 this choice of @xmath168 and @xmath169 is sufficient to ensure that ( [ inequality_proof_3 ] ) holds .
next , let @xmath170 since variance of @xmath171 is bounded above by a finite constant which does not depend on @xmath1 ( assumption [ ad ] ) , therefore we can find such @xmath172 that for all @xmath173 it is true that @xmath174 for all @xmath38 it follows that for all @xmath175 and @xmath176 it is true that @xmath177 note that by the furstenberg
kesten theorem , @xmath178 for all @xmath179 this implies that for all @xmath180 there exists such @xmath117 that both inequalities ( [ inequality_proof_4 ] ) and ( [ inequality_proof_5 ] ) hold .
this implies that for all such @xmath36 and for all @xmath181 the following inequality holds : @xmath182 otherwise , the sum of the events in ( [ inequality_proof_4 ] ) and ( inequality_proof_5 ) would cover all probability space and hence the sum of probabilities in ( [ inequality_proof_4 ] ) and ( [ inequality_proof_5 ] ) , @xmath183 would have to be greater then 1 .
this contradicts to the assumption that @xmath184 . inequalities ( [ inequality_proof_3 ] ) , ( [ inequality_proof_2 ] ) and ( inequality_proof_6 ) imply that @xmath185for all @xmath186 and @xmath187 where @xmath188 and @xmath189 are sufficiently large functions of @xmath190 .
it remains to complete the proof by proving proposition proposition_ld_inequality_unbounded_rv .
we will do this in the next section .
the rest of the paper consists of section [ section_large_deviation_bound ] , which is devoted to the proof of proposition proposition_ld_inequality_unbounded_rv , section [ section_necessity ] , which gives a necessary condition for uniform convergence in furstenberg
kesten theorem , and section [ section_conclusion ] , which concludes .
everywhere in this section , we assume that random matrices @xmath5 are independent , identically distributed and rotationally invariant , and that @xmath191 let us consider the following random variables : @xmath192it is known ( e.g. , @xcite ) that the random variables @xmath193 are independent and identically distributed .
their distribution coincides with the distribution of @xmath194 where @xmath74 is an arbitrary unit vector . in this section
, we consider an important case when each matrix @xmath5 has independent gaussian entries distributed according to @xmath195 . in this case
, @xmath196 is distributed in the same way as the random variable @xmath197where @xmath198 are independent standard gaussian variables . in order to prove proposition [ proposition_ld_inequality_unbounded_rv ] for this case ,
it is enough to show that the following result holds .
let @xmath193 be independent copies of the variable @xmath199 .\ ] ] if @xmath200 then there exists a function @xmath201 such that for all @xmath202 and all @xmath117 the following inequality holds:@xmath203 first of all , let us compute @xmath204where @xmath205 is a real number . by explicit calculation ,
@xmath206where @xmath207 is the gamma function .
this formula is valid for @xmath208 let @xmath209 where @xmath210 then using the stirling formula for large @xmath1 , we can write @xmath211\\[-8pt ] & \sim & \frac{1}{\sqrt{1 + 2\alpha } } n^{\alpha n}\exp \biggl\ { \biggl [ \biggl ( \frac{1}{2}+\alpha \biggr ) \log ( 1 + 2\alpha ) -\alpha \biggr ] n\biggr\ } .\nonumber\end{aligned}\ ] ] note that for all @xmath212 @xmath213 and for all @xmath214 @xmath215 with equalities only for @xmath216 if @xmath217 we set @xmath218 and @xmath219 and use the fact that for all sufficiently large @xmath36 the asymptotic term in ( stirling_expansion ) dominates all other terms . hence , we obtain the estimate @xmath220 if @xmath221 , then we can take @xmath222 and @xmath223 and we obtain @xmath224 by standard arguments , we can translate these inequalities into statements about probabilities of large deviations .
if @xmath225 then@xmath226 in this section , we are going to prove the second part of proposition proposition_ld_inequality_unbounded_rv . since @xmath5 are i.i.d . and rotationally invariant
, therefore the distribution of @xmath227 coincides with the distribution of @xmath228 and equals the distribution of the random variable @xmath229 here @xmath230 are components of the random vector @xmath231 which is uniformly distributed on the unit sphere and which is independent of @xmath232 let us start with considering large deviations of @xmath233 .
let @xmath234 suppose that with probability 1 , @xmath235 for all @xmath236 then for all @xmath217 @xmath237 \leq \exp \biggl\ { -\frac{nt^{2}}{4b ( b+t ) } \biggr\ } .\ ] ] let @xmath238 denote @xmath239 and let us estimate @xmath240 we will estimate the conditional probability @xmath241 which we denote as @xmath242 for simplicity .
note that @xmath243 where @xmath244 and @xmath245 let us use von neumann s formulas from ( @xcite , pages 373375 ) for the uncentered moments of the random variable @xmath238 .
namely , let @xmath246and let @xmath247then , von neumann s result is that @xmath248 using this result , we write @xmath249 next , note that @xmath250 and that @xmath251 this implies that @xmath252 \biggr\ } \\ & & \qquad = e^{-zt}\exp \biggl\ { \frac{n}{4}\frac { ( { 2bz / n } ) ^{2}}{1-{2bz / n}}\biggr\ } \\ & & \qquad = e^{-zt}\exp \biggl\ { \frac{b^{2}z^{2}}{n-2bz}\biggr\ } .\end{aligned}\ ] ] let @xmath253then @xmath254 altogether , we get @xmath255the proof of the inequality for @xmath256 is similar .
suppose that with probability 1 , @xmath81 for all @xmath236 then for all @xmath217 @xmath257\nonumber \\ & & \qquad\leq \exp \biggl\ { -\frac{nt^{2}}{4b ( b+t ) } \biggr\ } .\end{aligned}\ ] ] let @xmath258 for each @xmath259 and @xmath260 then : @xmath261 @xmath262 @xmath263 where @xmath264 let @xmath238 denote @xmath265 then @xmath266 & \geq & \pr \biggl\ { \log x\geq \log \overline{s}+\frac{t}{\overline{s}}\biggr\ } .\end{aligned}\ ] ] this and ( [ inequality_probability_of_deviation0 ] ) prove the first inequality .
the second inequality is proved similarly , and the third one is a consequence of the first two inequalities .
[ lemma_bound_laplace_transform1]suppose that @xmath47 is a random variable such that @xmath267where @xmath268 .
let @xmath269 then @xmath270 consider the case when @xmath271 first , let us estimate @xmath272 where @xmath273 is the distribution measure of @xmath274 let @xmath275 then , by integrating by parts and using the inequalities @xmath276 and @xmath79 we get @xmath277 & \leq & 2e^{-1/ [ 4c ( c+1 ) ] } e^{z/\sqrt{n}}+2z\int_{1/\sqrt{n}}^{\infty } e^{zt}\exp \biggl\ { -\frac{nt^{2}}{4c ( c+t ) } \biggr\ } \ , dt.\end{aligned}\ ] ] in order to estimate the integral in the last line , we divide it into two pieces , @xmath278 and @xmath279 then @xmath280 next , for the second piece , we have @xmath281 where we used the assumption that @xmath282 hence , combining the previous inequalities and using the assumption that @xmath283 again , we get @xmath284 } e^{z/\sqrt{n}}+2z\sqrt{\frac{8\pi c^{2}}{n}}\exp \biggl ( \frac{2c^{2}}{n}z^{2}\biggr ) \\ & & { } + 2\exp \biggl ( -\frac{n}{16}\biggr ) .\end{aligned}\ ] ] in addition , @xmath285 combining all the parts , we get @xmath286 } \bigr ) e^{z/\sqrt{n } } \\ & & { } + 2\exp \biggl ( -\frac{n}{16}\biggr ) , \end{aligned}\ ] ] from which the claim of the lemma follows for @xmath287 . the case when @xmath288 is similar .
let @xmath289 and let @xmath290 where @xmath264 then @xmath270 this follows directly from lemma lemma_bound_laplace_transform1 and inequality ( inequality_probability_of_deviation1 ) .
proof of the second part of proposition proposition_ld_inequality_unbounded_rv note that @xmath291,\ ] ] where @xmath292 are components of independent haar - distributed @xmath1-vectors @xmath293 let @xmath294we aim to estimate @xmath295 as usual , @xmath296where @xmath297 note that by assumption [ ab ] , @xmath298 hence , our previous lemmas are applicable .
we set @xmath299 and assume that @xmath300 ( note that the assumption that @xmath301 $ ] implies that @xmath302 ) then , by the previous lemma , we have @xmath303 since @xmath304 then the first term dominates the other two terms , and we can write @xmath305 hence , @xmath306 \bigr\ } .\end{aligned}\ ] ] clearly , we can find an @xmath201 such that for all @xmath307 @xmath308hence , for all @xmath307 @xmath309 > t\biggr\ } \leq \exp \ { -n ( t^{2}/16c^{2 } ) n\ } .\ ] ] the case of the inequality @xmath310
< -t\biggr\ } \leq \exp \
{ -n ( t^{2}/16c^{2 } ) n\}\ ] ] is similar .
finally , note that @xmath311
let us introduce the following assumption .
[ ad ] @xmath312 ^{2}$ ] exists and bounded by a constant that does not depend on @xmath38 let assumptions [ aa ] , [ ab ] and [ ad ] hold .
suppose that for every @xmath313 there exists such an @xmath314 that @xmath315 for all @xmath1 and all @xmath316 let @xmath317 is an arbitrary function of @xmath1 such that @xmath318 then @xmath319 let @xmath320 be such a unit vector that @xmath321 note that @xmath322 has the haar distribution by assumption of rotational invariance . by using the fact that @xmath323 we can write the inequality @xmath324where @xmath325 are independent haar - distributed vectors . by using assumption [ ad ]
, we can conclude that @xmath326 converges in probability to@xmath327 and that the convergence is uniform in @xmath38 this fact and the supposition of the theorem imply that @xmath328 must converge in probability to zero as @xmath329 and that the convergence must be uniform in @xmath38 if the conclusion of the theorem were invalid , then for some @xmath313 and all @xmath117 we could find an @xmath330 such that @xmath331 and this would contradict the uniform convergence of @xmath328 to zero .
in this paper , we found sufficient conditions that ensure that the convergence rate in the furstenberg kesten theorem is uniform with respect to the dimension of the space in which matrices operate .
let us call this phenomenon dimensional uniformity of convergence .
several interesting questions remain to be answered .
first , is it possible to prove the dimensional uniformity of convergence for random matrices which are not rotationally invariant , for example , for wigner matrices ?
second , assuming rotational invariance , what characterizes the laws of singular values @xmath332 for which the dimensional uniformity of convergence holds ?
in other words , what are necessary and sufficient conditions for dimensional uniformity of convergence ? | suppose that @xmath0 are i.i.d .
rotationally invariant @xmath1-by-@xmath1 matrices .
let @xmath2 .
it is known that @xmath3 converges to a nonrandom limit .
we prove that under certain additional assumptions on matrices @xmath4 the speed of convergence to this limit does not decrease when the size of matrices , @xmath1 , grows . . |
numerical simulations of collisions between gas - rich galaxies ( e.g. , barnes & hernquist 1996 and references therein ) and even minor mergers between a gas - rich galaxy and a satellite companion ( e.g. , mihos & hernquist 1994 ) show that these processes are very efficient in transporting large quantities of molecular gas into the galaxy nuclei . as a result of
the concentration of gas , a strong burst ( or various bursts ) of star - formation may occur , and an agn may be activated , as observed in many infrared luminous and ultraluminous galaxies ( lirgs and ulirgs ) .
sanders et al .
( 1988 ) suggested that the lirg and ulirg galaxies are the initial stage for the appearance of a quasar ( see also the review by sanders & mirabel 1996 ) .
however , recent results from mid - infrared spectroscopy seem to indicate that most lirgs and many ulirgs may be powered by star formation ( lutz et al . 1999 ) . these objects therefore allow probing the process of star formation on an extreme scale and intensity .
lcccc nic1 & f110 m & 0.043 & 384 & pseudo-@xmath9 + nic2 & f160w & 0.076 & 192 & @xmath10 + nic2 & f222 m & 0.076 & 640 & @xmath11 + nic2 & f187n & 0.076 & 640 & continuum + nic2 & f190n & 0.076 & 640 & pa@xmath1 + nic2 & f212n & 0.076 & 1792 & continuum + nic2 & f215n & 0.076 & 1792 & h@xmath7 + nic2 & f237 m & 0.076 & 960 & co absorption + wfpc2 & f606w & 0.041 & 500 & @xmath12 + ngc 1614 ( arp 186 ) is relatively nearby ( distance @xmath13mpc for @xmath14 km s@xmath15 mpc@xmath15 ) and has a high infrared luminosity ( @xmath16l@xmath4 , which places this system in the luminous infrared galaxy category ) .
the galaxy shows a spectacular outer structure with tidal tails or plumes ( see e.g. , neff et al .
1990 and also figure 1 ) suggesting that this morphology is the result of an earlier interaction / merger process with another galaxy .
neff et al . ( 1990 ) collected optical , near - infrared and radio observations for this galaxy and found no evidence for the presence of active galactic nucleus ( agn ) , making it an excellent laboratory for study of a very luminous starburst .
we present _
hst_/nicmos near infrared observations of ngc 1614 , giving unprecedented angular resolution on the galaxy for this spectral region .
these new observations are combined with an archival _ hst_/wfpc2 red image , ground - based near infrared spectroscopy , and results from the literature to probe the starburst .
_ hst_/nicmos observations of ngc 1614 were obtained in february 1998 using cameras nic1 and nic2 .
the pixel sizes are 0.043pixel@xmath15 and 0.076pixel@xmath15 respectively .
table 1 lists details of the observations .
standard data reduction procedures were applied ( see alonso - herrero et al .
2000a for more details ) .
the flux calibration was performed using the conversion factors based on measurements of the standard star p330-e during the servicing mission observatory verification ( smov ) program .
the fully - reduced images were rotated to the usual orientation with north up , east to the left .
they are shown in figure 2 .
in addition , we constructed an infrared color map using the nic2 f160w and nic2 f222 m images , which is equivalent to a ground - based @xmath2 color map . in the nic1 f110 m image ( highest angular resolution )
the nucleus of ngc 1614 appears to be slightly resolved with a fwhm of 0.15 ( or 45pc for the assumed distance ) , whereas for the unresolved sources ( the stellar clusters discussed in section 3.2 ) we measured fwhms of approximately 0.11 ( or 35 pc ) .
the continuum subtracted pa@xmath1 image was produced by a straight subtraction of the flux calibrated nic2 f187n image ( continuum ) from the flux calibrated nic2 f190n image ( line+continuum ) .
nicmos only provides narrow continuum bands to one side of the emission line . for ngc 1614
the continuum image lies to the blue , and therefore if some extinction is present the continuum at the emission line wavelength may be slightly over - subtracted .
table 2 gives the photometry of the hii regions in the ring of star formation and the brightest ones along the spiral arms .
a continuum subtracted h@xmath7 image was produced in a similar manner , using the f212n ( continuum ) and f215n ( line+continuum ) filters . because the h@xmath7 line is relatively weak , the image subtraction near the nucleus is not reliable .
artificial images of point sources were generated for both filters and used in experiments to determine where useful information could be obtained .
based on these experiments , we have excluded the data within 0.25 of the nucleus from further analysis .
figure 3 shows the surface brightness in pa@xmath1 compared with that in the ( 1,0 ) s(1 ) h@xmath7 line at @xmath17 m .
we have averaged the data along an artificial slit 0.61 in width and oriented at position angle 170 selected to avoid the secondary nucleus and the brightest hii regions in the circumnuclear ring ( see below ) .
the h@xmath7 extends farther out from the nucleus than does the pa@xmath1 .
lccccc + 1 & 0.05 & 0.10 & @xmath18 & 40.08 & 52.87 + 2 & 0.03 & @xmath19 & @xmath20 & 40.07 & 52.86 + 3 & @xmath21 & 0.24 & @xmath22 & 40.00 & 52.79 + 4 & 0.00 & 0.40 & @xmath23 & 40.04 & 52.83 + 5 & 0.01 & @xmath24 & @xmath25 & 40.03 & 52.82 + total ring & & & @xmath26 & 41.47 & 54.30 + + 6 & @xmath27 & 1.93 & @xmath28 & 39.07 & 51.86 + 7 & @xmath19 & @xmath29 & @xmath30 & 38.74 & 51.53 + 8 & 0.05 & @xmath31 & @xmath32 & 38.65 & 51.44 + 9 & 0.25 & @xmath33 & @xmath34 & 38.54 & 51.33 + 10 & 0.30 & 1.34 & @xmath35 & 39.12 & 51.91 + 11 & 0.44 & @xmath36 & @xmath37 & 38.60 & 51.39 + 12 & 0.57 & 2.13 & @xmath38 & 38.52 & 51.31 + 13 & 0.66 & 3.32 & @xmath39 & 38.64 & 51.43 + alonso - herrero et al .
( 2000a ) discuss a nicmos co photometric index ( defined as @xmath40 , where the fluxes @xmath41(f222 m ) and @xmath41(f237 m ) are in jy ) .
the spatial profile of this index is shown in figure 4 together with the pa@xmath1 line emission ( upper panel ) and the nic2 f222 m ( @xmath11-band ) brightness profile ( lower panel ) .
the errors in the co index account for the background subtraction uncertainties in both the nic2 f222 m and nic2 f237 m filters .
the nominal errors of the pa@xmath1 and nic2 f222 m profiles are smaller than the size of the symbols .
a wide field planetary camera 2 ( wfpc2 ) image of ngc 1614 was retrieved from the _ hst _ data archive .
this image was taken through the f606w filter in november 1994 .
standard pipeline reduction procedures were applied . because only a single exposure was taken for this galaxy , the cosmic rays are particularly difficult to remove , especially the extended ones .
the image is saturated in the center .
we rotated the cleaned image to the usual orientation ( see figure 1 ) .
near - infrared spectra were obtained at steward observatory s 2.3-m bok telescope on kitt peak using fspec ( williams et al .
1993 ) with an orientation @xmath42 .
the data were reduced as described by engelbracht et al .
the spectra were extracted from a @xmath43 region centered on the galaxy nucleus and were flux - calibrated using ground - based imaging data .
the @xmath9-band spectrum at a resolution of @xmath44 is shown in figure 5 ( upper panel ) , while spectra in the @xmath10 and @xmath11 bands at a resolution of @xmath45 are shown in figure 5 ( middle and lower panels respectively ) .
we have used the procedures discussed in engelbracht et al .
( 1998 ) to subtract a stellar continuum , so that accurate measurements can be made of faint lines that might otherwise be affected by stellar absorption features . in table 3
we present emission line fluxes .
our detections of the ( 1 - 0)s(0 ) and ( 2 - 1)s(1 ) lines of h@xmath7 are supported by weak detections of the same lines by puxley & brand ( 1994 ) .
the spectra also allow us to determine the co index to quantify the depth of the first overtone absorption bands at 2.3@xmath46 m . using the method and calibration discussed by engelbracht et al .
( 1998 ) , the co index is 0.23 , substantially stronger than for normal old stellar populations found in the nuclei of non - starbursting galaxies .
our determination that the co absorption is stronger than expected for a `` normal '' galaxy nucleus is in agreement with previous work ( ridgway , wynn - williams , & becklin 1994 ; puxley & brand 1994 as discussed by goldader et al .
1997 ) .
lcc 1.0941 & pa@xmath47 & 8.2 @xmath48 0.6 + 1.2570 & [ ] & 3.9 @xmath48 0.2 + 1.2822 & pa@xmath49 & 13.4 @xmath48 0.3 + 1.5339 & [ ] & 0.5 @xmath48 0.2 + 1.6114 & br13 & 0.9 @xmath48 0.3 + 1.6412 & br12 & 0.7 @xmath48 0.2 + 1.6440 & [ ] & 4.4 @xmath48 0.2 + 1.6774 & [ ] & 0.5 @xmath48 0.2 + 1.6811 & br11 & 1.7 @xmath48 0.3 + 1.7367 & br10 & 1.6 @xmath48 0.1 + 2.0338 & h@xmath7(1,0)s(2 ) & 0.3 @xmath48 0.1 + 2.0587 & hei & 2.0 @xmath48 0.1 + 2.0735 & h@xmath7(2,1)s(3 ) & 0.3 @xmath48 0.1 + 2.1218 & h@xmath7(1,0)s(1 ) & 1.0 @xmath48 0.1 + 2.1542 & h@xmath7(2,1)s(2 ) & @xmath50 + 2.1661 & br@xmath47 & 4.0 @xmath48 0.1 + 2.2233 & h@xmath7(1,0)s(0 ) & 0.3 @xmath48 0.1 + 2.2477 & h@xmath7(2,1)s(1 ) & 0.2 @xmath48 0.03 +
before we interpret the properties of ngc 1614 , we need to apply an extinction correction .
the extinction to the nucleus of ngc 1614 has been estimated a number of times .
some works ( e.g. , neff et al .
1990 ; shier , rieke , rieke 1996 ) have found relatively small values for the visual extinction ( @xmath51mag with a foreground screen model ) from optical and infrared colors , and infrared emission lines .
puxley & brand ( 1994 ) used optical and near infrared hydrogen recombination lines , and claimed that the extinction to ngc 1614 could not be modeled with a simple foreground screen model .
instead they used a composite model ( mixture of dust and gas , and a foreground screen ) and inferred a total extinction in the optical of @xmath52mag . ho ,
beck , & turner ( 1990 ) derived a relatively large extinction between br@xmath47 and br@xmath1 , but their result is subject to significant calibration uncertainties ( see also puxley & brand 1994 ) .
lccccc 1 & @xmath53 & @xmath54 & 18.7 & 0.56 & @xmath55 + 2 & @xmath56 & @xmath57 & 19.1 & 0.32 & @xmath58 + 3 & @xmath59 & @xmath60 & 18.9 & 0.52 & @xmath61 + 4 & @xmath59 & 0.34 & 17.5 & 0.44 & @xmath62 + 5 & @xmath63 & 1.45 & 16.9 & 0.38 & @xmath64 + 6 & @xmath65 & 2.42 & 18.8 & 0.05 & @xmath66 + 7 & @xmath67 & @xmath68 & 18.6 & 0.36 & @xmath55 + 8 & @xmath69 & 2.41 & 18.2 & 0.69 & @xmath70 + 9 & @xmath71 & 2.04 & 20.0 & 1.94 & @xmath72 + 10 & @xmath73 & 1.37 & 17.9 & 0.28 & @xmath74 + 11 & @xmath75 & 2.81 & 18.9 & 0.39 & @xmath76 + 12 & @xmath27 & @xmath77 & 17.5 & 0.47 & @xmath78 + 13 & 0.07 & 1.70 & 17.0 & 0.84 & @xmath79 + 14 & 0.16 & 1.04 & 18.1 & 0.70 & @xmath80 + 15 & 0.30 & 1.44 & 19.3 & 1.18 & @xmath81 + 16 & 0.52 & @xmath27 & 18.9 & 0.52 & @xmath76 + using the fluxes of the two [ feii ] emission lines at @xmath82 m and @xmath83 m ( whose ratio is largely independent of nebular conditions ) and the hydrogen recombination lines pa@xmath49 and br@xmath47 and a foreground dust screen model , we find values of the extinction to the gas in the @xmath11-band of @xmath84mag , corresponding to @xmath85mag . a comparison of dust models ( from witt , thronson , & capuano 1992 ) with the simple foreground screen model ( c.f . , the analysis of ngc 253 by engelbracht et al .
1998 ) shows that the correction needed for all the infrared lines ( [ feii]@xmath86 m , [ feii]@xmath87 m , pa@xmath49 and br@xmath47 ) and the broad - band colors is similar . thus a foreground dust screen model ( with extinction in the @xmath11-band of @xmath88mag ) is a good approximation for the near - infrared extinction to the ionized gas
. this approximation may not give the correct extinction in the optical , but we do not use any optical data for our study of the star formation properties of this galaxy .
the infrared @xmath2 color map ( figure 2 and also the close - up in figure 6 ) shows that the actual distribution of the extinction is quite patchy .
if we assume an intrinsic color @xmath89 for the stellar population and the rieke & lebofsky ( 1985 ) extinction law , we find values of the @xmath11-band extinction to the stars of @xmath90mag to the east of the center of ngc 1614 , whereas the @xmath11-band extinction to the west is higher , ranging between approximately 0.2 and 1.2mag . the @xmath2 color of the nucleus ( through a 0.76-diameter aperture ) is 0.41 , which would imply a @xmath11-band extinction of @xmath91mag .
compared with other dusty starbursts , the levels of extinction in ngc 1614 are low
. we can determine the near infrared properties of the nucleus of ngc 1614 reasonably accurately by taking an average extinction of @xmath92mag and making a first order correction by assuming a simple foreground screen of dust .
this value for the extinction agrees well with foreground screen model of puxley & brand ( 1994 ) , @xmath93mag .
they show that an underestimate of the ly continuum by about 20% may result compared with more sophisticated models .
however , from the combination of our slightly higher assumed extinction level and other minor differences , our derived ly continuum is similar to the value they feel is most accurate . in general ,
if there are regions of very heavy extinction , or where the extinction is optically thick at the wavelengths of our observations , then we would have underestimated the starburst parameters .
our conclusions about the galaxy would become stronger with this error corrected .
the relatively modest extinction in a face - on starburst is an important advantage for studies of ngc 1614 .
the foundation for much of this paper is the resulting possibility to determine the starburst properties more reliably than for most other well - studied examples .
the nicmos images provide important new insights to the morphology of the starburst in ngc 1614 .
figure 2 presents from the upper left corner and clockwise , the optical wfpc2 f606w image , nic2 f160w ( @xmath10-band ) , nic2 f160w - nic2 f222 m ( @xmath2 color map ) , and continuum subtracted pa@xmath1 image .
the field of view of these images is @xmath94 .
figure 6 presents close - ups of the central regions ( @xmath95 ) as in figure 2 .
the nucleus of ngc 1614 is extremely luminous , with an @xmath10-band absolute magnitude measured through a 0.76-diameter aperture of @xmath96mag ( corrected for extinction ) .
the bright infrared nucleus is surrounded by a number of fainter sources and similar sources lie along the spiral arms .
these objects are probably stellar clusters .
we measured @xmath10-band luminosities ( see table 4 ) ranging up to @xmath97mag for the brightest sixteen clusters ( excluding those located in the vicinity of the bright nucleus ) , similar to the cluster luminosities measured in arp 299 ( alonso - herrero et al .
2000a ) and other luminous and ultraluminous infrared galaxies ( scoville et al .
neff et al . ( 1990 ) noted the presence of a knot some 10sw of the nucleus in a @xmath98-band image ( figure 2 in their paper ) , and suggested that it could be interpreted to be fragments of a second galaxy .
this knot ( tentatively identified with cluster # 7 , table 4 ) is on the edge of our nic2 f160w image .
it is most likely to be another stellar cluster from its observed luminosity .
ground - based imaging shows the nucleus of the galaxy to be elongated at position angle 39@xmath99 ( e.g. , mazzarella & boroson 1993 ) .
the wfpc2 f606w image resolves the central 2 of ngc 1614 into two prominent sources at a similar position angle ( see figure 6 ) .
the brighter of these sources coincides with the infrared nucleus .
the other optical source lies at the position of a secondary peak in the infrared images to the ne of the nucleus , but it is far less prominent in the infrared than in the f606w image .
its @xmath10-band luminosity is about 30 times fainter than the primary nucleus ( @xmath100mag ; the error accounts for the uncertainty in the determination of the underlying background because of its proximity to the bright nucleus ) .
the @xmath10-band absolute magnitude of the secondary peak is a magnitude or more brighter than the brightest of the stellar clusters , suggesting that it is the nucleus of an interacting small galaxy .
the small extinction toward the secondary nucleus , as indicated by its blue color from optical to infrared , places it in front of the main nucleus and the activity around it ( see figure 6 ) .
the co band in the primary nucleus is very strong as shown by both our spectroscopy ( section 2.3 ) and imaging ( figure 4 ) , much stronger than expected for an old stellar population .
the most plausible explanation for this behavior is that the light from this nucleus is dominated by red supergiants that are the product of a starburst of age 10 or more million years . as shown in figure 4 , outside the starburst ring
the co index assumes a value appropriate to an old stellar population , presumably that of the underlying galaxy .
the pa@xmath1 line emission image reveals a nuclear ring of star formation with an approximate diameter of 650pc .
the pa@xmath1 morphology is remarkably similar to the 6 cm radio map ( neff et al .
1990 ) . although the two available mid - infrared images ( keto et al . 1992 ; miles et al .
1996 ) do not agree in detail , they do show that the emission in the central part of the galaxy is confined to a 2 " diameter region .
the image of miles et al .
( 1996 ) , which should have higher angular resolution and signal to noise , suggests that the emission arises from the ring seen in pa@xmath1 .
comparison of small aperture ( @xmath101 5diameter ) ground - based photometry ( rieke & low 1972 ; lebofsky & rieke 1979 ; carico et al . 1988 ) with the iras fluxes ( in a 45 by 250 beam ) shows that the infrared emission of the entire galaxy is strongly concentrated into the central region . in fact , if approximate corrections are made from the ground - based to iras effective wavelengths and for the wavelength weighting introduced in iras to reduce color corrections to the fluxes , more than 80% of the total infrared flux at 12 @xmath46 m originates in the nucleus , and hence , from the miles et al .
( 1996 ) image , primarily in the circumnuclear ring . at 25@xmath46 m , making similar corrections to the measurement by lebofsky & rieke ( 1979 ) , we find about 75% of the total flux to originate in the nucleus .
given the relation between far infrared and radio fluxes , we can also estimate the central concentration in the far infrared from radio data .
we compare the survey results by griffith et al .
( 1995 ) in a 4.2 arcmin beam at 6 cm with the vla map of the nucleus reported by neff et al .
we have corrected the flux scale in the latter reference upward by 11% , as they suggest .
we find that @xmath101 60% of the 6 cm flux lies in the nuclear ring .
these three estimates are consistent in stating that about 70% of the far infrared luminosity arises from the nuclear region . in the following ,
we can ascribe most of the far infrared luminosity to the circumnuclear ring of hii regions .
the integrated far infrared luminosity indicates a star formation rate of 52 m@xmath4 yr@xmath15 from equation(3 ) of kennicutt ( 1998 ) .
if 70% of this star formation occurs in the nucleus , we predict a level there of 36 m@xmath4 yr@xmath15 .
the pa@xmath1 luminosity in the circumnuclear ring ( see table 2 ) indicates , after correction for reddening , a star formation rate of @xmath101 36 m@xmath4 yr@xmath15 from equation ( 2 ) of kennicutt ( 1998 ) , in perfect agreement .
the hii regions in the ring of ngc 1614 are extremely luminous with an equivalent ionizing luminosity an order of magnitude greater than that of 30 doradus .
a similar rich population of bright hii regions has been found in the luminous starburst in arp 299 ( see alonso - herrero et al .
the hii regions in the spiral arms are weaker than those in the ring , but still exceedingly luminous .
those listed in table 2 are all similar in luminosity to 30 dor ( for which log l@xmath102 @xmath103erg s@xmath15 ; kennicutt , edgar , & hodge 1989 ) .
30 dor is usually considered the prototype _ super_-hii region .
such objects are very uncommon in normal galaxies ( e.g. , kennicutt et al .
1989 ; rozas , beckman , & knapen 1996 ) .
the existence of so many comparable to 30 dor and even with greater ionizing luminosity appears to be associated with strong starbursts ( see alonso - herrero et al .
2000a ) .
just outside the ring of hii regions , the @xmath2 extinction map shows a partial ring of high extinction .
this ring is obliterated to the ne due to the overlying secondary nucleus , but otherwise it can be seen for more than 220@xmath99 .
it traces the dense interstellar molecular material as discussed in the following section .
the molecular hydrogen appears to extend from the ring of hii regions into the molecular cloud ( see figure 3 ) .
the detection of the ( 2 - 1 ) s(3 ) line at @xmath104 m and the ( 2 - 1 ) s(1 ) line at @xmath105 m ( the latter line both in our spectrum and in that of puxley & brand 1994 ) suggests that a significant fraction of the emission is due to fluorescence in low density gas , for which the ratio of these lines to the @xmath106 m line is @xmath107 ( e.g. black & van dishoek 1987 ; sternberg & dalgarno 1989 ) .
the relative h@xmath7 line strengths are very similar to those in ngc 253 ( engelbracht et al . 1998 ) , for which we derived that the excitation was probably shared between fluorescence and shocks , with nearly 2/3 of the total near infrared h@xmath7 luminosity from the former mechanism .
the morphology suggests that at least some of the h@xmath7 originates at a shock penetrating into the molecular ring from the hii ring .
spectra of higher signal to noise and angular resolution will be needed to sort out the excitation conditions for the h@xmath7 in more detail .
figure 7 summarizes the distribution of starlight on a large scale .
we have used ellipse fitting to generate surface brightness profiles from the nicmos @xmath11-band image ( figure 4 ) and from a ground - based image ( shier et al .
these profiles join smoothly and trace the surface brightness to 20 radius ( @xmath108kpc ) .
the galaxy is dominated by an exponential disk with a scale length of 6.3 ( @xmath109kpc ) from a radius of 5 ( @xmath110kpc ) outward . within a radius of 3(@xmath111kpc ) , there is a bright central source fitted well by an r@xmath112 law with a scale length of 1.14 or 350pc .
at least in the innermost arcsec this region is dominated by the output of the newly formed stars and not by a traditional bulge of old stars . combining the results reported above , ngc 1614
appears to be a textbook example of a propagating starburst that started in the nucleus of a late type , large spiral galaxy , has grown outward to a radius of @xmath101 300 pc , and is potentially still growing into a circumnuclear nuclear ring of molecular material just outside this radius ( see also discussion in section 3.4 ) .
this picture will be made more quantitative in a following section where we carry out evolutionary synthesis modeling of the starburst .
our data re - emphasize the result of neff et al .
( 1990 ) that the activity in this galaxy is dominated by star formation and not by an agn there is no compact nuclear source either in the near or mid infrared , the emission lines are all narrow , and correlations observed in star forming regions such as the ratio of hydrogen recombination lines to far infrared luminosity appear to hold in ngc 1614 .
an interesting additional result is that the starburst is accompanied by a number of _ super_-hii regions in the spiral arms of the galaxy .
the presence of a secondary nucleus out of the plane of ngc 1614 suggests that all this activity has been triggered by an interaction with a smaller galaxy , which has by now largely been destroyed .
there is further evidence for a merger from the tidal tails in the outer structure of ngc 1614 ( figure 1 ) .
an aperture of diameter 2 ( 620pc at the galaxy distance ) includes the starburst and molecular ring . to understand the behavior in these regions , we need to determine the total mass and the mass of various constituents within this aperture .
we first estimate the dynamical mass and then show that it is virtually all accounted for by the population of old , pre - starburst stars .
the mass budget leaves very little for the starburst and molecular gas , as summarized in table 5 .
this budget is strongly incompatible with the standard conversion of co to gas mass and also challenges starburst models using conventional forms of the initial mass function .
lcc total dynamical mass & & 13 + @xmath6 from stellar velocity dispersion & 12 & + @xmath6 from pa@xmath1 ring rotation & 14 & + mass in old stars & & 12 + modeled starburst mass & & 5.5 + molecular gas mass & & 3 + @xmath6 from standard co conversion & 48 & + @xmath6 from extinction & @xmath113 & + @xmath6 from toomre q condition & @xmath114 & + @xmath6 for optically thin co & 2 & + shier et al .
( 1996 ) modeled the velocity dispersion measured in the 2.3@xmath46 m co band head to derive the dynamical mass for a 1000pc ( 3 ) diameter region in ngc 1614 .
they considered three models for the stellar distribution : two bulge models and a disk model .
the mass was only weakly dependent on which model was used for its derivation , although the mass in the disk model would be about twice that in the bulge ones if the disk were assumed to be at the inclination we find for the circumnuclear ring .
however , since the @xmath11-band surface brightness profile of the inner 2-arcsec radius is well fitted with an @xmath115 law ( see figure 7 ) , here we adopt the highest value for the spherical geometry ( bulge ) models of shier et al .
( 1996 ) , 1.8 @xmath116 10@xmath117 m@xmath4 , and make an approximate correction in proportion to diameter to the 2 diameter region which is thus indicated to have a mass of 12 @xmath48 4 @xmath116 10@xmath118 m@xmath4 .
an independent estimate of the mass in the central region of ngc 1614 can be obtained using the br@xmath47 spectroscopy presented in puxley & brand ( 1999 ) , and our knowledge of the morphology and size of the emitting region ( a ring ) .
it is expected that the ring is associated with a lindblad resonance or other similar dynamical feature and is close to round .
puxley & brand ( 1999 ) detected two peaks of line emission separated by @xmath119 km s@xmath15 .
we have obtained the ring rotational velocity by modeling the line profile expected for a rotating ring , convolved with a gaussian to represent both the instrumental profile ( resolution 39 km s@xmath15 ) and the dispersion within the beam .
we find models with dispersions between 38 and 64 km s@xmath15 fit equally well as the double gaussian used by puxley & brand . in all the satisfactory models ,
the peak rotational velocity is very close to 1.4 times the velocity that would have been deduced just from the separation of the gaussians fitted by puxley and brand .
that is , the rotational velocity is 106 km s@xmath15 , uncorrected for inclination .
the total rotational velocity is determined by de - projecting for an inclination of @xmath120 ( determined from the observed ellipticity and an assumption that the ring is round ) . from the @xmath10-band morphology we know that the mass is symmetrically distributed inside the hii ring , and therefore we can use the virial theorem in its simplest form to convert the rotational velocity to a mass .
the conversion depends weakly on the mass distribution .
given the good fit of the r@xmath112 model to the starlight in the center of the galaxy ( see section 3.2 ) , coupled with the large mass represented by these stars ( see next section ) , we have assumed a spherically symmetric mass distribution .
we find that the total mass in a region with diameter 2@xmath121 620 pc is @xmath122 .
the mass for a spherically symmetric distribution lies midway between the minimum possible value , for a thin disk , and the maximum , for a thin and narrow ring .
the total range between these two extremes is a factor of two ( i.e. , @xmath101 @xmath123 ) . both
because the situations at the extremes tend to be unstable and because of the observational evidence for a massive stellar spheroid within the ring , the systematic errors in our mass estimate should be much smaller than this range . for the following we adopt the average of the two measures of @xmath101 13 @xmath116 10@xmath118 m@xmath4 as the best estimate of the mass in the starburst region , with an upper limit of 20 @xmath116 10@xmath118 m@xmath4 .
not only is this value based on two independent and consistent determinations , but also our extinction data indicate that the galaxy is optically thin at the wavelengths and in the regions critical for the mass determinations .
thus , the mass should be measured reliability .
to estimate the mass in old stars within the star forming region , we used the @xmath11-band fit to the surface brightness profile outside the ring of star formation ( see figure 7 ) .
the exponential disk is assumed to account for the old stars present in the galaxy before the strong episode of star formation seen today .
the disk fit was extrapolated to the center of the galaxy to estimate the contribution of old stars hidden by the output of the starburst .
no extinction correction was applied , since the extinction to the disk is unknown and in any case is likely to be less than the already modest extinction ( at @xmath11 ) to the nucleus .
we estimated 1.2mjy of @xmath11-band light is emitted by the old population within the central 2 " . using the light - to - mass ratio derived for spiral galaxy disks by thronson & greenhouse ( 1988 ) ( corresponding to @xmath124 ) , we find that the old stars account for @xmath125 .
there is substantial evidence that the @xmath126 for spiral disks is very similar from galaxy to galaxy and in accordance with the values derived from thronson & greenhouse ( e.g. , bottema 1993 ; 1999 ) .
in fact , the widely applied tully - fisher relation probably depends on the uniformity of @xmath126 in spiral disks ( aaronson , mould , & huchra 1979 ) .
even if we assume that it is at the extreme low side of the dispersion in the tully fisher relation , the mass in the disk of ngc 1614 is unlikely to be more than 30% lower than our estimate . that is , assuming the lower limit @xmath127 , a lower limit to the disk mass in the nuclear region is 9 @xmath128 .
our estimate ignores the mass of any bulge of old stars hidden within the central starburst region , and hence this lower limit is quite conservative .
comparison of the mass in old stars with the dynamical mass of 13 @xmath116 10@xmath118 m@xmath4 provides a difficult constraint .
this problem can not be relieved significantly by enlarging the region under consideration .
for example , within a 3 diameter region , the flux from the exponential disk model is 1.9mjy , corresponding to a lower limit on the mass in old stars of 14.5 @xmath116 @xmath129 , which includes virtually all the dynamical mass within this diameter also .
use of the standard procedure for estimating the mass in the interstellar medium from co measurements is strongly inconsistent with the mass budget implied by the dynamics and the old stellar population ; see table 5 .
scoville et al .
( 1989 ) made observations of the nuclear region in the @xmath5co j=1 - 0 line .
these results have been re - analyzed by scoville et al .
( 1991 ) and bryant & scoville ( 1999 ) ; our discussion is based primarily on the last reference .
the nuclear co source is described as being unresolved with a diameter @xmath130 3 .
most of the material is probably located in the ring of strong extinction revealed in our @xmath131 images , which is 2 in diameter .
bryant & scoville ( 1999 ) use a `` standard '' conversion of co emission to total mass ( h@xmath7 + he ) of molecular gas of n(h@xmath7)/i@xmath132 = 2.25 @xmath116 10@xmath133 @xmath134 ( k km s@xmath15)@xmath15 , corrected by a factor of 1.36 to allow for the mass in helium .
they deduce a nuclear molecular mass of 48 @xmath116 10@xmath118 m@xmath4 , nearly four times the total dynamical mass !
further details of the conversion ( but with a slightly higher standard factor ) can be found in sanders , scoville , & soifer ( 1991 ) .
the use of the standard conversion factor in intense starbursts has come under question previously .
maloney & black ( 1988 ) suggested that it should substantially overestimate the molecular masses because of the relatively high densities and temperatures of the molecular clouds in these regions .
shier , rieke , & rieke ( 1994 ) determined the dynamical mass with near infrared spectroscopy of the 2.3@xmath46 m co bandhead in the nuclei of a number of luminous starburst galaxies and found that the masses of gas were overestimated by the standard conversion by factors of @xmath135 .
solomon et al . (
1997 ) and downes & solomon ( 1998 ) concluded the standard factor may be high by a factor of five in starbursts , based on a comparison of the dynamical mass and gas mass from far infrared measures .
lisenfeld , isaak , & hills ( 2000 ) used submillimeter measurements of the emission by cold dust to argue that the co - deduced mass is typically @xmath136 times higher than would be deduced from the thermal emission .
dunne et al . ( 2000 ) point out that such an increase is one of the possible explanations for apparently large values of m@xmath137/m@xmath138 in some strong merging / starburst galaxies .
however , agreement has not been reached as illustrated , for example , by the use of the standard factor by bryant & scoville ( 1996 , 1999 ) . in ngc 1614
, an independent test of the large molecular mass deduced from the standard conversion can be based on the mm - wave co line profile .
casoli et al .
( 1991 ) have published a profile in a 44 diameter beam that shows the full width at 20% intensity to be @xmath139 km s@xmath15 .
the line profile of sanders et al .
( 1991 ) in a 55 beam has a full width at 20% intensity of 390 km s@xmath15 .
scoville et al .
( 1989 ) show that about 30% of the emission in these beams would have been from the central compact source .
if the mass of this region were @xmath140m@xmath4 , the line width would be @xmath141 km s@xmath15 or greater for this central 30% component , which would be very difficult to reconcile with the lack of strong , broad wings in the observed profile .
we have made another estimate of the amount of molecular gas , based on our discovery of an extinction shadow due to a circumnuclear molecular ring .
we use the ratio of @xmath142co emission to h@xmath7 from the study of molecular clouds by goldsmith , bergin , & lis ( 1997 ) and the relation between a@xmath143 and @xmath142co from alves , lada , & lada ( 1999 ) to relate the extinction in this ring to its mass . from our data
, we estimate that the shadow of the molecular ring subtends an annulus between radii of 0.5 and 1 with a typical extinction level of a@xmath143 @xmath144mag .
because the extinction is only effective in the foreground of the stellar population , we made the assumption that the source is symmetric along the line of sight and therefore doubled the mass estimates derived from the relations cited above .
our value should probably be taken as a lower limit : we would underestimate the extinction if stars are mixed with the obscuring gas ( or if some stars lie in front of it , or the obscuration is clumpy ) .
also , the reddening at @xmath10 and @xmath11 in the thickest parts of the ring is strong enough that optical depth effects might reduce the apparent extinction .
nonetheless , our value of @xmath145 @xmath116 10@xmath118 m@xmath4 is well aligned with the dynamical mass and , in fact , a significantly higher value would create a serious problem in explaining the nuclear properties given the other constituents and their expected contributions to the mass .
an upper limit to the mass of interstellar material can be obtained from the discovery that this material is predominantly distributed in a circumnuclear ring , as shown both by the pa@xmath1 morphology and the extinction shadow of a ring sector seen in the color maps .
although this ring is currently the site of vigorous star formation as the starburst propagates outward , it must have originally been a semi - stable configuration .
the toomre stability parameter for gas in a plane is @xmath146 where @xmath147 is the epicyclic frequency , @xmath148 is the cloud velocity dispersion and @xmath149 is the gas surface density . so that the gas remains in a disk and does not fragment into small clumps , @xmath150 .
thus , we can derive an upper limit for @xmath149 : @xmath151 where @xmath152 is the circular velocity and @xmath153 .
we have used a maximum value of @xmath154 for a solid body rotation curve .
we assume a velocity dispersion of @xmath155 km s@xmath15 .
this value was determined by extrapolating the observed cloud size / line width relation ( falgarone , puget , & prault 1992 ) to a size of 100pc and then doubling the width for good measure , to allow for possible extreme conditions in a galaxy nuclear region .
taking @xmath156 using @xmath157 pc ( based on the ring width in the extinction shadow ) , we find a maximum mass of @xmath158 .
if there is a high dispersion component of material , the actual gas mass could be somewhat higher .
another lower limit of m@xmath159 @xmath160 2 @xmath116 10@xmath118 m@xmath4 can be computed by assuming the co emission is optically thin ( bryant & scoville 1996 , equation a4 ) . combining these arguments , the molecular mass in the nucleus is most likely @xmath101 3 @xmath116 10@xmath118 m@xmath4 , about 25% of the dynamical mass in this region .
this mass is about 15 times lower than the @xmath161 in molecular mass deduced from co and the standard conversion by bryant and scoville ( 1999 ) .
table 5 illustrates how tight the mass budget is .
it seems unlikely that the molecular gas can be more than 1/10 the amount estimated from the standard conversion even if the dynamical mass is set to the upper limit of 20 @xmath116 10@xmath118 m@xmath4 and the mass in old stars to its lower limit of 9 @xmath116 10@xmath118 m@xmath4 . in that case
, it might be necessary to incorporate a significant change in the starburst initial mass function ( see section 3.5.2 ) .
our results are in agreement with those of shier et al .
( 1994 ) , who estimated that the amount of gas in the nucleus of ngc 1614 was a factor of at least nine below the value from the standard conversion . to accompany their estimate of a huge mass of molecular gas , scoville et al .
( 1989 ) inferred the extinction in the nuclear region of ngc 1614 to be @xmath162mag .
they argued that much of the `` action '' in this region would be hidden from optical and even infrared observers unless there happened to be a particularly favorable line of sight .
this picture of lirgs and ulirgs has gained wide credence , but the observations presented here are a serious caution .
not only do we find far less molecular mass , but more significantly for the `` hidden action '' model , we _ do not _ find anything approaching the high extinction deduced from the co observations .
an independent indication that the extinction is not hiding any significant action is the close correspondence between the star formation rates deduced from pa@xmath1 and far infrared luminosities , as discussed in section 3.2 . for our analysis of the star formation properties of ngc 1614
, we will use the rieke et al .
( 1993 ) evolutionary synthesis starburst models because they include a careful calibration of the co stellar absorption band strengths against observational data
. we have considered gaussian bursts of 5myr duration ( fwhm ) with the peak of star formation at 5myr after the beginning time for the burst .
ngc 1614 is a well known wolf - rayet ( wr ) galaxy ( vacca & conti 1992 ; schaerer , contini , & pindao 1999 ) . the presence of wr features in star forming galaxies can only be explained with short bursts of star formation with durations of the order of a few million years ( leitherer & heckman 1995 ) , as used in our fit to the observed properties of the galaxy .
we used a truncated salpeter imf from 1 to @xmath163 , and @xmath164 from 0.1 to @xmath165 . ] which is virtually identical to imf8 which rieke et al .
( 1993 ) decided gave the best fit to the properties of m82 , and has since been shown to fit other starburst galaxies well ( engelbracht et al .
1996 , 1998 ) . in table 6
we summarize the properties of the central starburst of ngc 1614 that will be fitted with the evolutionary synthesis models .
llll dynamical mass & @xmath3 & ir spectroscopy & this work + infrared luminosity & @xmath166 & iras & goldader et al .
( 1997 ) + ionizing photons ( log ) & 54.5 , @xmath167 & [ neii ] , pa@xmath1 & roche et al .
( 1991 ) , this work + @xmath11 absolute magnitude & @xmath168mag & & engelbracht ( 1997 ) + snr & @xmath169 & [ feii ] & this work + co index & 0.23 & ir spectroscopy & this work + the infrared luminosity will be taken as a lower limit to the bolometric luminosity of the galaxy .
this assumption makes two errors in opposite direction . from the aperture measurements discussed in section 3.2 , up to about one third of the infrared luminosity of the galaxy
may be produced outside the nuclear region .
on the other hand , a significant amount of optical / uv energy must escape from the starburst before it is absorbed and re - emitted in the infrared , particularly given the patchy and relatively low level of extinction .
the number of ionizing photons is from both the [ neii]@xmath170 m line flux of roche et al .
( 1991 ) , and the _
pa@xmath1 flux , the latter corrected for an average extinction of @xmath171mag and assuming case b recombination .
our estimate of the number of ionizing photons is in good agreement with the value from puxley & brand ( 1994 ) derived from measurements that included br@xmath1 , analyzed through detailed modeling of the extinction .
it is also consistent with the upper limit established from the 3@xmath148 non - detection of the radio h@xmath192 hydrogen recombination line ( phookun , anantharamaiah , & goss 1998 ) . to allow for the possibility of absorption of ionizing photons by interstellar dust , in the models we take the derived uv flux as a lower limit .
the [ feii ] line ratios are similar to those detected in ngc 253 ( engelbracht et al .
1998 ) and are typical of those in supernova remnants .
we estimated the supernova rate ( snr ) from the [ feii]@xmath83 m luminosity ( corrected for extinction ) from this work and a new calibration between these two quantities ( alonso - herrero et al .
2000b , in preparation ) .
our best fit to the properties of ngc 1614 with a single burst of star formation is presented in figure 8 ( left panel ) . in this diagram
the output of the model is normalized to the properties of ngc 1614 so the target quantity is unity .
the model uses a mass in newly formed stars of @xmath172 ( @xmath173 ) .
the derived age is between 5myr and 8myr after the peak of star formation , or @xmath174myr measured from the rising half power of the assumed gaussian star formation rate .
this age is consistent with our failure to detect the 1.7@xmath46 m he line , which suggests ( although it is not a strong result given the signal to noise ) that there are relatively few stars hotter than 40,000k ( vanzi et al .
all the observables except the co index converge at the young age .
the observed co index appears to be too deep for such a young population .
the strong co of this galaxy is more typical of a stellar population ten million years old or older .
the discrepancy with the age of 5 @xmath6 8 myr for the other starburst indicators suggests that this galaxy has experienced a more extended period of star formation than in our single burst .
the simplest example would be a double burst . in figure 8
( right panel ) we show the best fit model obtained with two gaussian bursts ( both with fwhm = 5myr ) separated by 5myr .
each burst uses 20% of the dynamical mass . as can be seen from this figure
, the fit is excellent .
the derived age is @xmath175yr after the peak of star formation of the first burst .
a variety of other time dependencies of star formation are also likely to fit the data , but the minimum duration of the episode and the minimum mass required are both well constrained by the double burst model .
this starburst model coincides closely with the morphology of the galaxy .
we find that the strong co is concentrated in the nucleus ( figure 4 , upper panel ) , and the modeling associates this feature with the older part of the starburst .
most of the ionizing flux is produced in the younger part of the starburst , and is physically located in the circumnuclear ring ( figure 6 ) . from the starburst modeling and the diameter of the circumnuclear ring of hii regions , we can roughly estimate that the speed at which the starburst is propagating outward is @xmath176 km s@xmath15 .
rieke et al .
( 1980 ) found that evolutionary starburst models using the then - current form for the local imf ( miller & scalo 1979 ) had difficulty fitting the properties of m82 within the available dynamical mass .
they suggested that the imf in the starburst might be biased toward massive stars compared with the proposed local form .
since the low mass stars contribute most of the total mass in the imf , but have no observable properties in a starburst , reducing their numbers is a plausible way to reconcile the masses .
this possibility has been discussed at great length since .
rieke et al .
( 1993 ) reconsidered m82 , experimenting with a broad variety of forms for the imf and histories of star formation .
they confirmed that the currently proposed forms for the local imf ( e.g. , scalo 1986 ) were inadequate and proposed their `` imf8 '' as a more satisfactory solution , imf8 provides @xmath177 times more stars in the @xmath178 range than does scalo ( 1986 ) but drops toward higher masses to avoid overproduction of oxygen in very massive stars ( rieke et al . 1993 ) . in the last few years
, modeling of the ultraviolet spectra of starburst galaxies has led to the conclusion that their imfs have roughly the salpeter slope at high masses ( e.g. , gonzlez - delgado et al .
satyapal et al . (
1997 ) modeled m82 with less complete evolutionary models than used by rieke et al .
they claimed that they could account for the @xmath11-band luminosity with a salpeter imf , avoiding the necessity to suppress formation of low mass stars compared with the local imf .
a problem with their approach is that they did not demonstrate the simultaneous fitting of all the starburst constraints used by rieke et al .
( 1993 ) ; it is relatively easy to fit a subset of constraints if the rest are ignored .
furthermore , because they assumed a foreground screen extinction model , the @xmath11-band luminosity they fitted is in fact a lower limit to the true luminosity , making their fits less demanding .
nonetheless , their success with a modified salpeter imf confirms independently the derivation of imf8 to fit m82 by rieke et al .
( 1993 ) , as we show below .
lcc imf8 & 14% & 5% + modified salpeter & 13% & 9% + original salpeter & 7% & 5% + miller & scalo 1979 & 6.5% & 3% + scalo 1986 & 3% & 1% + basu & rana 1992 & 3% & 3.5% + the critical aspect of a starburst imf constrained by luminosities and dynamical mass is the proportion of mass it places in the @xmath178 range that dominates the observable aspects of the starburst .
table 7 summarizes this proportion and that of more massive stars for the various forms of the imf that have been used to fit starbursts .
it shows the similarity of imf8 and the modified salpeter imf , which are equivalent in fitting the primary starburst characteristics . however , imf8 is to be preferred because its factor of two lower portion of very massive stars results in a proportionate reduction in the production of oxygen .
keeping the abundance of oxygen within the observational limits is a difficult constraint in starburst models ( rieke et al .
1993 ; wang & silk 1993 ) .
the other forms for the imf produce significantly fewer 10 @xmath6 30 m@xmath4 stars .
to first order , the dynamical masses required with them are just in the ratio of the proportions of these stars .
thus , if m82 can be fitted with 2.5 @xmath116 @xmath129 in a modified salpeter imf , it would require about @xmath179m@xmath4 with a miller - scalo imf , about 70% of the dynamical mass available and unlikely to be an acceptable fit once allowance is made for the mass of other constituents of the nuclear region . for ngc 1614 , the miller - scalo imf would take all the dynamical mass and the scalo ( 1986 ) and basu & rana ( 1992 ) ones would require about twice the dynamical mass , all of which is clearly unacceptable . virtually all studies find the local imf to fall more steeply toward high masses than the salpeter form .
even salpeter ( 1955 ) continued the simple power law to high masses only for simplicity , but at the same time called attention to an apparent high mass steepening : `` it is not yet clear whether the steeper drop for masses larger than 10 m@xmath4 is a real effect , since in this region masses and bolometric corrections are not known very accurately and the number of such stars near the galactic plane is small . '' as with the rest of his work this statement has proven prescient .
however , a salpeter imf has been shown to be an approximate fit in many situations outside the local neighborhood ( e.g. , scalo 1998 ) .
it now appears that a better description of the m82 issue might be that the local imf ( or at least our estimates of it ) may differ from the imf we find in most regions of active star formation , including in starbursts like m82 .
we have emphasized that ngc 1614 is particularly well suited to studying starburst properties , given the relatively modest near infrared extinction and our favorable viewing of it at modest inclination .
it has a well determined dynamical mass of 13 @xmath116 10@xmath118 m@xmath4 in the starburst region , as discussed in section 3.2 .
however , the constituents in this region appear to sum to a significantly greater mass , @xmath160 9 @xmath116 10@xmath118 m@xmath4 for the old stars plus @xmath101 3 @xmath116 10@xmath118 m@xmath4 for the molecular gas plus 5.5 @xmath116 10@xmath118 m@xmath4 for the starburst using either a salpeter imf or imf8 .
the total exceeds the dynamical mass by at least 35% . unless there is substantial contamination of the disk outside the starburst by younger stars
, these numbers suggest that the mass in the starburst may be overestimated .
a further indication in the same direction is that the starburst plus remaining molecular gas in our nominal estimates account for 65% of the dynamical mass .
theoretical modeling suggests that instabilities will set in and a starburst will appear once the mass of gas reaches @xmath101 20% of the total in a galaxy nucleus ( wada & habe 1992 ; bekki 1995 ) .
these estimates are somewhat uncertain and may not apply strictly to the conditions in ngc 1614 .
however , they fall so far below the deduced mass of new stars and residual gas to suggest that the imf in the ngc 1614 starburst is weighted toward massive stars , even compared with imf8 or a modified salpeter function . this behavior should be tested in studies of other extreme starbursts .
the mechanism needed to account for the strong star formation in luminous mergers is a matter of debate .
the two main competing models are : ( 1 ) cloud - cloud collisions which predict a large population of high mass stars to give rise to the high infrared luminosities ( scoville , sanders , & clemens 1986 ) ( 2 ) gravitational instability in nuclear gas disks , as favored by taniguchi & ohyama ( 1998 ) .
the former model would predict a schmidt law , schmidt ( 1959 ) ] with index @xmath180 , whereas the latter model predicts indices @xmath181 as found for isolated galaxies and strong starbursts ( kennicutt 1998 ; taniguchi & ohyama 1998 ) . from the pa@xmath1 luminosity of ngc 1614 ( corrected for extinction ) and using the relation between sfr and h@xmath1 luminosity ( kennicutt 1998 ) we measured a @xmath182 in the ring and a sfr surface density of @xmath183 , similar to , although slightly higher than , the earlier estimate by kennicutt ( 1998 ) from the fir luminosity and somewhat different parameters for the size of the emitting region ( @xmath184 ) .
the value of @xmath185 that we found for ngc 1614 is of the same magnitude as that of the giant hii region 30 doradus ( kennicutt 1998 ) , and the global star formation efficiency for this galaxy approaches 100% over a gas consumption period of @xmath186 .
based on the standard conversion , kennicutt ( 1998 ) derived @xmath187 .
if the mass in molecular gas is only @xmath101 3 @xmath116 10@xmath118 m@xmath4 , with a conversion of co to h@xmath7 mass a factor of @xmath101 15 lower than the standard value , one would derive @xmath188 or more , as discussed by kennicutt ( 1998 ) ( with less gas , higher efficiency is needed ) .
thus , our results appear to favor the cloud collision explanation for the high rate of star formation .
mihos & hernquist ( 1994 ) explored via numerical simulations the merger of a large and a smaller galaxy , similar to our proposal for ngc 1614 . from their simulation they predicted that in response to the tidal perturbation of the infalling satellite , the disk galaxy develops strong spiral arms .
this non - axisymmetric perturbation causes large quantities of material to be funneled into the central regions , and as a result a strong burst of star formation is triggered lasting for up to 60myr .
the predicted central starburst is very compact ( @xmath189pc in radius ) and contains some 85% of the total mass in newly formed stars .
hernquist & mihos ( 1995 ) carried out further simulations .
they showed that the degree of central concentration of gas decreases markedly with the presence of a massive bulge .
the relatively low dynamical mass in the central 620 pc of ngc 1614 indicates that it had little bulge prior to the collision , so it would be ideal for a strong central concentration of gas .
hernquist & mihos ( 1995 ) state that the resolution of their calculations is inadequate to determine how the gas behaves once it has sunk into the central few hundred pc .
however , from the behavior of ngc 1614 , it appears that it must be further concentrated until instability is reached in the bottom of the potential well of the galaxy , leading to the initial stages of massive star formation .
as the starburst ages , supernovae begin to explode and winds develop that expand outward , compressing the remaining molecular gas surrounding the nucleus ( e.g. , tenorio - tagle & muoz - tun 1997 ; taniguchi , trentham , & shioya 1998 ) .
gravitationally unstable shocked layers of interstellar gas may be produced in cloud cloud collisions and by the impact of these winds on incoming clouds .
the resulting star formation appears to favor massive stars ( whitworth et al .
1994 ) , possibly accounting for the suggestion of a high abundance of such stars in extreme starbursts .
the evidence we find that the starburst in ngc 1614 has expanded from the nucleus to the ring of hii regions and into the surrounding molecular ring is in agreement with this picture .
we have used _ hst_/nicmos and wfpc2 imaging , ground - based near infrared spectra , and information from the literature to study the luminous infrared galaxy ngc 1614 .
the modest extinction to its starbursting nucleus , combined with our favorable viewing angle , make this galaxy ideal for probing the properties of its powerful starburst .
we show that star formation has propagated outward from the nucleus to a @xmath139pc radius circumnuclear ring of massive hii regions .
these circumnuclear hii regions are an order of magnitude more luminous than 30 doradus , and hii regions in the spiral arms are similar in luminosity to 30 dor .
these regions are probably younger versions of the luminous stellar clusters also around the nucleus and in the spiral arms .
the presence of so many extremely luminous hii regions and stellar clusters is similar to the behavior of other luminous starbursts .
the ring of hii regions is surrounded by molecular gas .
the mass in molecular gas appears to be @xmath190 the prediction from the standard @xmath5co / h@xmath7 ratio , with an upper limit of 1/10 the standard ratio .
the large levels of extinction predicted previously from the standard ratio are absent in ngc 1614 .
the mass budget for the starburst is very tight , suggesting that the proportion of massive stars is larger than predicted by a salpeter imf modified to turn over near 1 m@xmath4 , or by the virtually identical imf8 found to fit the starburst in m82 ( rieke et al .
1993 ) . until now
, such imfs have been adequate to fit the properties of all known starbursts . the high efficiency needed to convert the available molecular gas density into the observed density of star formation suggests that the starburst is propagating by cloud - cloud interactions and winds .
the behavior of ngc 1614 is consistent with simulations for the interaction of a modest sized galaxy with a massive , late type spiral .
we thank j. hinz for a helpful discussion and the anonymous referee for a thorough critique of the first version of this paper .
this work was supported by the national aeronautics and space administration through grant nag 5 - 3042 and by the national science foundation grant ast-9529190 . | the modest extinction and reasonably face - on viewing geometry make the luminous infrared galaxy ngc 1614 an ideal laboratory for study of a powerful starburst . _
hst_/nicmos observations show : 1 . )
deep co stellar absorption , tracing a starburst nucleus about 45pc in diameter ; 2 . ) surrounded by a @xmath0pc diameter ring of supergiant hii regions revealed in pa@xmath1 line emission ; 3 . )
lying within a molecular ring indicated by its extinction shadow in @xmath2 ; 4 . )
all at the center of a disturbed spiral galaxy .
the luminosities of the giant hii regions in the ring are extremely high , an order of magnitude brighter than 30 doradus ; very luminous hii regions , comparable with 30 dor , are also found in the spiral arms of the galaxy .
luminous stellar clusters surround the nucleus and lie in the spiral arms , similar to clusters observed in other infrared luminous and ultraluminous galaxies .
the star forming activity may have been initiated by a merger between a disk galaxy and a companion satellite , whose nucleus appears in projection about 300pc to the ne of the nucleus of the primary galaxy .
the relation of deep stellar co bands to surrounding ionized gas ring to molecular gas indicates that the luminous starburst started in the nucleus and is propagating outward into the surrounding molecular ring .
this hypothesis is supported by evolutionary starburst modeling that shows that the properties of ngc 1614 can be fitted with two short - lived bursts of star formation separated by 5myr ( and by inference by a variety of models with a similar duration of star formation ) .
the total dynamical mass of the starburst region of @xmath3 is mostly accounted for by the old pre - starburst stellar population .
although our starburst models use a modified salpeter initial mass function ( turning over near 1 m@xmath4 ) , the tight mass budget suggests that the imf may contain relatively more 10 - 30 m@xmath4 stars and fewer low mass stars than the salpeter function .
the dynamical mass is nearly 4 times smaller than the mass of molecular gas estimated from the standard ratio of @xmath5co ( 1 @xmath6 0 ) to h@xmath7 .
a number of arguments place the mass of gas in the starburst region at @xmath8% of the dynamical mass , nominally about 1/15 and with an upper limit of 1/10 of the amount estimated from @xmath5co and the standard ratio . |
water maser emission ( @xmath9 cm in the rest frame ) is known to trace warm , dense gas at radii of 0.1 to 1 pc in the accretion disks surrounding supermassive black holes in galactic nuclei ( e.g. , * ? ? ?
it can also trace material heated by jet activity ( e.g. , * ? ? ? * ) and wide - angle nuclear winds @xcite .
emission from disks is visible when they are viewed close to edge - on and amplification paths are longest . several have been mapped with very long baseline interferometry ( vlbi ) : ngc4258 @xcite , ngc1068 @xcite , and the circinus galaxy @xcite .
`` maser disks '' may also exist in ic2560 @xcite , ngc5793 @xcite , and mrk1419 @xcite , though confirmation awaits further study . triply peaked spectra characterize emission from accretion disks that are well populated by masers .
emission close to the systemic velocity of the host galaxies ( i.e. , low - velocity emission ) occurs where orbital motion is transverse to the line of sight .
high - velocity emission is symmetrically offset by the disk orbital velocities and arises in regions where the disk motion is parallel to the line of sight .
velocities as high as @xmath10 have been observed ( i.e. , in ngc4258 ) .
water maser sources in active galactic nuclei ( agn ) are important astrophysical tracers in part because vlbi can provide maps of resolved disk structure ( e.g. , warping ) and dynamics ( e.g. , rotation curves and proper motions ) , as in @xcite .
unfortunately , only @xmath11 h@xmath0o masers are known in agn , and few of these exhibit triply peaked spectra .
the discovery of new masers is a priority and a challenge .
first , the emission is typically weak , and surveys must invest substantial time observing each target with the largest available apertures ( e.g. , 1 hour with a 100 m diameter antenna ) .
second , mean detection rates in surveys are typically @xmath12% for seyfert ii galaxies closer than @xmath13 @xcite .
third , because the orbital speeds of disks ( and concomitant velocity range of maser emission ) can not be known in advance , surveys must have instantaneous bandwidths of thousands of .
sufficiently broadband observing systems have become available to the general community only recently .
we report the detection of seven new masers obtained in a high sensitivity , broad bandwidth survey of 160 nearby agn ( @xmath14 ) .
the survey was unusual for two reasons .
first , we used a 70-m antenna of the nasa deep space network ( dsn ) to achieve high sensitivity ( see also * ? ? ?
second , we used a custom built , portable , 5350-wide spectrometer .
the survey , which is ongoing at dsn facilities in the northern and southern hemispheres , and the hardware will be discussed in detail elsewhere .
the 70-m dsn antenna located at tidbinbilla , near canberra , is equipped with a cooled 18 to 26 ghz hemt receiver .
the left circular polarization channel is limited by a selectable 600 mhz bandpass filter and downconverted to a band centered at 321.4 mhz . a second downconversion and bandpass filter
deliver a 400 mhz baseband ( @xmath15 at @xmath9 cm ) at the input of a 2-bit , four - level , 4096-lag digital autocorrelator .
survey observations were conducted between 2002 may and september .
the zenith system temperature was typically 40k under good winter observing conditions , and zenith opacities were typically @xmath16 .
the temperature was calibrated with respect to an ambient load .
the peak aperture efficiency was @xmath17 , and we determined its dependence on elevation through antenna temperature measurements of pks1830@xmath18211 ( @xmath19 jy ) and pks1921@xmath18293 ( @xmath20 jy ) .
we calibrated these flux densities against 3c286 ( 2.6 jy at 21.8 ghz ) with the australia telescope compact array ( atca ) on 2002 september 9 and 12 .
we measured and corrected for pointing errors ( at the 70-m antenna ) , which were under most circumstances on the order of @xmath21 or 8% of the half - power beamwidth . as a result
, we estimate the flux density calibration of spectra is uncertain by @xmath22 . to construct total - power spectra for each agn , we nodded the antenna between signal and reference positions . switching times of 30 or 45 s were usually sufficient to produce flat baselines in the spectra .
we removed residual fluctuations by subtracting a running mean computed over intervals of 256 channels ( @xmath23 ) .
heliocentric velocities were computed from sky frequencies determined by the digital tuning of receiver elements , using the radio definition of doppler shift .
this calibration should be accurate to better than 0.1 .
we checked it by observing the h66@xmath24 radio recombination line in w33 , which has a known local standard of rest velocity of @xmath25 @xcite , and for which we measured a velocity of @xmath26 .
we observed 160 agn with the tidbinbilla antenna and obtained typical noise levels of 10 to 20 mjy ( @xmath27 ) with 30 minutes of on - source integration .
we discovered seven new h@xmath0o masers ( table1 , figure1 ) .
each detection has been confirmed through observation on more than one day .
we have also measured positions for six of the new masers with the very large array ( vla ) of the national radio astronomy observatory ( nrao ) or the atca .
these observing tracks were also used to estimate continuum flux density levels from the underlying agn at about @xmath28 cm ( table1 ) . where we detected continuum emission , it was coincident with the maser emission , to within estimated uncertainties ( @xmath3 ) .
six masers lie in seyfertii objects , and one lies in a nucleus whose classification is ambiguous ( ngc2824 ; * ? ? ?
the masers in ngc5643 and ngc6300 lie in galaxies that had been targeted in previous searches .
@xcite reported noise levels of 59 to 87 mjy for ngc5643 and 59 mjy for ngc6300 in @xmath29 wide channels , after hanning smoothing .
@xcite report noise levels of 119 and 109 mjy for the two galaxies respectively , with 0.8 wide channels . at the strength we measured ( @xmath30 mjy in a 1.3 channel without hanning smoothing ) , the maser emission in ngc5643 would have been detected at the 3 to @xmath31 level by these earlier observations , from which we conclude the maser emission varies significantly with time .
the emission in ngc6300 is sufficiently weak that it would not have been detected by the earlier observations .
two of the new maser sources exhibit discrete line complexes above and below the systemic velocity , which probably correspond to the high - velocity emission that is a characteristic of maser action in the accretion disks of supermassive black holes .
eso269-g012 displays red and blueshifted high - velocity emission , symmetrically offset by @xmath32 kms@xmath33 from a narrow line near the systemic velocity of the galaxy ( figure2 ) .
the inferred orbital speed is exceeded only by that of ngc4258 among known maser hosts .
the high - velocity line complexes extend over @xmath34 , which may be due to the radial distribution of emitting gas .
if the rotation curve is `` keplerian '' ( @xmath35 ) , then the outer radius of the molecular material in the disk would be @xmath36 times the inner radius .
ngc6926 exhibits high - velocity emission that is symmetrically offset by up to @xmath37 from the systemic velocity ( figure1 ) .
because the blueshifted emission is relatively weak , we have used the vla to confirm it is real .
if the masers in eso269-g012 and ngc6926 lie in disks at radii of 0.1 pc , which is the radius of the innermost masers in ngc4258 , then the enclosed mass is on the order of @xmath38 to @xmath39 m@xmath40 .
the corresponding centripetal acceleration in the eso269-g012 disk ( @xmath41 ) , which would be manifested by a secular velocity drift in the low - velocity emission , is @xmath42yr@xmath33 , which is large enough to measure readily within one year .
the maser in ngc2824 lies in an early - type galaxy and displays an anomalously broad line profile ( @xmath43 half power full width ) .
broad lines are also seen toward other early - type galaxies that host maser sources ( e.g. , ngc1052 , mrk348 ) , where the emission is seen toward radio jets rather than accretion disks @xcite . however , any correspondence between hubble - type and maser characteristics is tentative .
for example , the maser in irasf01063@xmath188034 is broad @xcite , but the galaxy is believed to be type sa ( though it is difficult to classify because it is edge - on ) . in ngc1386 and irasf22265@xmath181826 , the relative positions of the masers and jets therein are unknown because the jets are undetected at about @xmath4 cm @xcite . for ngc2824 , we have established a 1 mjy ( @xmath31 ) upper limit on continuum emission at @xmath44 cm ( table1 ) .
previously reported h@xmath0o maser emission in the circinus galaxy comprises two broad complexes of lines that bracket the systemic velocity ( @xmath45 ; * ? ? ?
* ) , with spectral features offset by as much as @xmath46 @xcite .
vlbi images ( @xmath47 mjy @xmath27 noise ) resolve a warped accretion disk and show emission with somewhat larger @xmath48 doppler shifts @xcite .
this is the maximum detected orbital speed , and it corresponds to a radius of @xmath49 pc .
we have detected weak ( @xmath50 jy ) high - velocity emission that establishes a @xmath51 to @xmath52 range of emission ( i.e. , doppler shifts up to @xmath53 ) .
the rotation curve of the circinus accretion disk is approximately keplerian @xcite .
if the new emission arises in the accretion disk , then it lies at @xmath7 pc radius ( @xmath54 schwarzschild radii for a @xmath55 m@xmath40 black hole ) , and future vlbi studies will be able to map the rotation curve and warp of the accretion disk over an expanded range of radii .
we have detected new h@xmath0o maser sources in seven agn , as well as new high - velocity emission components in the circinus galaxy .
two of the new masers exhibit high - velocity emission , indicative of emission from accretion disks .
it should be possible to map these two disks with vlbi , trace their rotation curves , and weigh the central engines that bind them .
ultimately , it may be possible to measure geometric distances to the two host galaxies , which would be significant because both lie in the hubble flow .
one of the new masers lies in an early - type galaxy , and we speculate that it might therefore be associated with an as yet undetected jet .
we have also discovered new high - velocity water maser emission in the circinus galaxy , vlbi observation of which could greatly improve our understanding of its accretion disk .
we thank c. timoc and spaceborne , inc . for designing and building the correlator , and for technical support in the field .
we are grateful for the invaluable support and assistance given by the management , operations , and technical staff of the cdscc , and in particular thank j. lauf for his technical assistance with our program early on .
we also acknowledge the contributions of c. garcia - miro ( inta ) , j. f. gmez , e. jimnez bailn , and i. de gregorio - monsalvo ( laeff / inta ) to testing of the spectrometer and observing software at the madrid dsn station prior to the observing season in australia . this work was supported in part by r&d funds of the smithsonian astrophysical observatory .
greenhill , l. j. , et al .
2001 , in proc iau symp . 205 , galaxies and their constituents at the highest angular resolutions , eds .
r. schilizzi , s. vogel , f. paresce , & m. elvis ( asp conf .
san francisco : asp ) , 334 - 337 peck , a. p. , falcke , h. , henkel , c. , menten , k. , hagiwara , y. , gallimore , j. , & ulvestad , j. 2001 , in proc iau symp .
206 , cosmic masers : from protostars to black holes eds .
v. migenes ( asp conf .
; san francisco : asp ) , in press ngc2824 & 2735 & 170 & 09 19 02.22 & @xmath5626 16 12.0 & 2370 & 12 & @xmath57 + ngc2979 & 2695 & 152 & 09 43 08.65 & @xmath1810 23 00.0 & 4050 & 6.6 & + ngc5643 & 1194 & 252 & 14 32 40.70 & @xmath1844 10 27.8 & 3840 & 6.8 & 4.1 + ngc6300 & 1106 & 185 & & & 3465 & 5.7 & + ngc6926 & 5851 & 183 & 20 33 06.11 & @xmath1802 01 38.9 & 5490 & 5.1 & 2.9 + eso269-g012 & 4868 & 152 & 12 56 40.51 & @xmath1846 55 34.4 & 4905 & 6.5 &
@xmath58 + irasf19370 - 0131&6060 & 208 & 19 39 38.91 & @xmath1801 24 33.2 & 4995 & 5.8 & @xmath59 + circinus & 439 & 168 & 14 13 09.95 & @xmath1865 20 21.2 & 13068 & 5.1 & + | we report the discovery of h@xmath0o maser emission at 1.35 cm wavelength in seven active galactic nuclei ( at distances up to @xmath1 mpc ) during a survey conducted at the 70-m diameter antenna of the nasa deep space network near canberra , australia .
the detection rate was @xmath2 .
two of the maser sources are particularly interesting because they display satellite high - velocity emission lines , which are a signature of emission from the accretion disks of supermassive black holes when seen edge on .
three of the masers are coincident , to within uncertainties of @xmath3 , with continuum emission sources we observed at about @xmath4 cm .
we also report the discovery of new spectral features in the circinus galaxy h@xmath0o maser that broaden the known velocity range of emission therein by a factor of @xmath5 .
if the new spectral features originate in the circinus accretion disk , then molecular material must survive at radii @xmath6 times smaller than had been believed previously ( @xmath7 pc or @xmath8 schwarzschild radii ) . |
optimization is a challenging problem in economic analysis and risk management , which dates back to the seminal work of markowitz [ 1 ] .
the main assumption is that the return of any financial asset is described by a random variable , whose expected mean and variance are assumed to be reliably estimated from historical data . the expected mean and variance
are interpreted as the reward , and respectively the risk of the investment .
the portfolio optimization problem can be formulated as following : given a set of financial assets , characterized by their expected mean and their covariances , find the optimal weight of each asset , such that the overall portfolio provides the smallest risk for a given overall return [ 1 - 5 ] .
therefore , the problem reduces to finding the `` efficient frontier '' , which is the set of all achievable portfolios that offer the highest rate of return for a given level of risk . using the quadratic optimization mathematical framework
it can be shown that for each level of risk there is exactly one achievable portfolio offering the highest rate of return . here , we consider the standard risk - return portfolio optimization model , when both long buying and short selling of a relatively large number of assets is allowed .
we derive the analytical expression of the efficient frontier for a portfolio of @xmath0 risky assets , and for the case when a risk - free asset is added to the model .
also , we provide an r implementation for both cases , and we discuss in detail a numerical example of a portfolio of several risky common stocks .
a portfolio is an investment made in @xmath0 assets @xmath1 , with the returns @xmath2 , @xmath3 , using some amount of wealth @xmath4 .
let @xmath5 denote the amount invested in the @xmath6-th asset .
negative values of @xmath5 can be interpreted as short selling .
since the total wealth is @xmath4 we have : @xmath7 it is convenient to describe the investments in terms of relative values such that : @xmath8 and @xmath9 to characterize the portfolio we consider the expected return : @xmath10 where @xmath11 is the expected return of each asset , @xmath3 .
also , we use the covariance matrix of the portfolio : @xmath12,\ ] ] where @xmath13 in order to quantify the deviation from the expected return , and to capture the risk of the investment .
the variance of the portfolio is then given by : @xmath14 where @xmath15^{t}$ ] is the vector of weights .
a portfolio is optimal if for a given expected return @xmath16 , the portfolio has the least variance @xmath17 .
finding such a portfolio requires the solution of the following constrained quadratic optimization problem [ 6 ] : @xmath18 subject to : * the constant invested wealth constraint ( equivalent to eq .
2 ) @xmath19 * the expected return constraint ( equivalent to eq .
4 ) : @xmath20 where @xmath21^{t}$ ] and @xmath22^{t}$ ] .
this problem can be solved using the method of lagrange multipliers .
let us define the lagrangian : @xmath23 where @xmath24 and @xmath25 are the lagrange multipliers .
the critical point of the lagrangian can be obtained by solving the system of equations : @xmath26 @xmath27 @xmath28 from the first equation we have : @xmath29 and from the next two equations we have : @xmath30 @xmath31 taking into account that : @xmath32 we can write : @xmath33\left[\begin{array}{c } \mu_{1}\\ \mu_{2 } \end{array}\right]=2\left[\begin{array}{c } 1\\ \rho \end{array}\right],\ ] ] where : @xmath34=\left[\begin{array}{cc } \mathbf{u}^{t}\mathbf{s}^{-1}\mathbf{u } & \mathbf{r}^{t}\mathbf{s}^{-1}\mathbf{u}\\ \mathbf{r}^{t}\mathbf{s}^{-1}\mathbf{u } & \mathbf{r}^{t}\mathbf{s}^{-1}\mathbf{r } \end{array}\right].\ ] ] this system has a solution if : @xmath35 since @xmath36 is a positive definite matrix , the inverse @xmath37 is also positive definite , which means that @xmath38 for any vector @xmath39
. obviously we have @xmath40 and @xmath41 , and : @xmath42 and therefore we also have @xmath43 .
the lagrange multipliers are then given by : @xmath44=\frac{2}{d}\left[\begin{array}{c } a_{22}-a_{12}\rho\\ -a_{12}+a_{11}\rho \end{array}\right],\ ] ] and the weights of the optimal portfolio are : @xmath45 where : @xmath46 @xmath47 the portfolio which minimizes the variance for a specified expected return is called a `` frontier portfolio '' .
it follows that all frontier portfolios @xmath48 are a linear combination of the two portfolios @xmath49 and @xmath50 .
the variance of the frontier portfolio is : @xmath51 which can be further simplified as : @xmath52 this equation represents the `` efficient frontier '' , and it represents a hyperbola in the @xmath53-plane .
from here we obtain the weights of the minimum variance portfolio : @xmath54 and the corresponding risk - return values : @xmath55 @xmath56 an important investment preference on the `` efficient frontier '' is the portfolio with the maximum sharpe ratio [ 1 - 3 ] : @xmath57 the sharpe ratio represents the expected return per unit of risk
. therefore , the portfolio with maximum sharpe ratio @xmath58 gives the highest expected return per unit of risk , and therefore is the most `` risk - efficient '' portfolio .
geometrically , the portfolio with maximum sharpe ratio is the point where a line through the origin is tangent to the efficient frontier , and therefore it is also called the `` tangency portfolio '' . in order to find the tangency point @xmath59
we observe that the slope of the tangency line : @xmath60 should be equal with the derivative of the `` efficient frontier '' at that point : @xmath61 thus , we easily obtain the risk - return pair for the `` tangency - portfolio '' : @xmath62 @xmath63 also , the allocation of the assets for the `` tangency portfolio '' are therefore given by : @xmath64
the covariance matrix @xmath36 is positive definite , and the correlation matrix @xmath65 is given by : @xmath66 where @xmath67 obviously , the correlation matrix @xmath65 has @xmath0 positive eigenvalues : @xmath68 and @xmath0 associated orthogonal eigenvectors : @xmath69,\ ] ] @xmath70 such that : @xmath71 where : @xmath72 in order to define the eigen - portfolios [ 7 ] we divide each eigenvector of the correlation matrix by the volatility of the corresponding asset : @xmath73 and we normalize by imposing a constant invested wealth , such that : @xmath74 where @xmath21^{t}$ ] , @xmath75 . for each eigen - portfolio ,
the weight of a given asset is inversely proportional to its volatility .
also , the eigen - portfolios are pairwise orthogonal , and therefore completely decorrelated , since : @xmath76 thus , any portfolio can be represented as a linear combination of the eigen - portfolios , since they are orthogonal and form a basis in the asset space .
it is also important to emphasize that the first eigen - portfolio , corresponding to the largest eigenvalue , typically has positive weights , corresponding to long - only positions .
this is a consequence of the classical perron - frobenius theorem , which states that a sufficient condition for the existence of a dominant eigen - portfolio with positive entries is that all the pairwise correlations are positive .
one can always get a dominant eigen - portfolio with positive weights using a shrinkage estimate [ 8 - 9 ] , which is a convex combination of the covariance matrix and a shrinkage target matrix @xmath77 : @xmath78 where the shrinkage matrix @xmath77 is a diagonal matrix : @xmath79 in this case , there exists a @xmath80 such that the shrinkage estimator has a dominant eigen - portfolio ( dep ) with all weights positive .
this portfolio is of interest since it provides a long - only investment solution , which may be desirable for investors who would like to avoid short positions and high risk .
let us now assume that one can also invest in a risk - free asset .
a risk - free asset @xmath81 is an asset with a low return @xmath82 , but with no risk at all , i.e. zero variance @xmath83 . the risk - free asset is also uncorrelated with the risky assets , such that @xmath84 for all risky assets @xmath3 .
the investor can both lend and borrow at the risk - free rate .
lending means a positive amount is invested in the risk - free asset , borrowing implies that a negative amount is invested in the risk - free asset . in this case , we consider the following quadratic optimization problem [ 1 - 3 ] : @xmath18 subject to : @xmath85 the lagrangian of the problem is given by : @xmath86.\ ] ] the critical point of the lagrangian is the solution of the system of equations : @xmath87 @xmath88 from the first equation we have : @xmath89 therefore , the second equation becomes : @xmath90 and from here we obtain : @xmath91 where @xmath92 the weights of the risky assets are therefore given by : @xmath93 and the corresponding amount that is invested in the risk - free asset is : @xmath94 also , the standard deviation of the risky assets is : @xmath95 or equivalently : @xmath96 this is the efficient frontier when the risk - free asset is added , or the capital market line ( cml ) , and it is a straight line in the return - risk ( @xmath97 ) space .
obviously , cml intersects the return axis for @xmath98 , at @xmath99 , which is the return when the whole capital is invested in the risk - free asset .
the tangency point of intersection between the efficient frontier and the cml corresponds to the `` market portfolio '' .
this is the portfolio on the cml where nothing is invested in the risk - free asset .
if the investor goes on the left side of the market portfolio , then he invests a proportion in the risk - free asset .
if he chooses the right side of the market portfolio , he borrows at the risk - free rate .
the market portfolio can be easily calculated from the equality condition : @xmath100 the solution of the above equation provides the coordinates of the market portfolio : @xmath101 @xmath102 and the weights of the market portfolio are then given by : @xmath103
the code for portfolio optimization was written in r , which is a free software environment for statistical computing and graphics [ 10 ] . to exemplify the above analytical results
, we consider a portfolio of common stocks .
the raw data can be downloaded from yahoo finance [ 11 ] , and contains historical prices of each stock .
the list of stocks to be extracted is given in a text file , as a comma delimited list .
the raw data corresponding to each stock is downloaded and saved in a local `` data '' directory , using the `` data.r '' script ( appendix a ) , which has one input argument : the file containing stock symbols .
once the raw data is downloaded the correct daily closing prices for each stock are extracted , and saved in another file , which is the main data input for the optimization program .
the extraction is performed using the `` price.r '' script ( appendix b ) , which has three input arguments : the file containing stock symbols included in the portfolio , the number of trading days used in the model , the output file of the stock prices .
the pseudo - code for the case with _
n _ risky assets is presented in algorithm 1 . also , the r script performing the optimization and visualization for the _
n _ risky assets case is `` optimization1.r '' ( appendix c ) .
the script has three input arguments : the name of the data file , the number of portfolios on the efficient frontier to be calculated , and the maximum return considered on the `` efficient frontier '' ( this should be several ( 5 - 10 ) times higher than the maximum return of the individual assets ) .
the pseudo - code for the case with _
n _ risky assets and a risk - free asset is presented in algorithm 2 , and the r script performing the optimization and visualization for the _
n _ risky assets case is `` optimization2.r '' ( appendix d ) .
the script has four input arguments : the name of the data file , the number of portfolios on the cml to be calculated , the daily return of the risk free asset , and the maximum return considered on the `` efficient frontier '' . in order to exectute the code , on unix / linux platforms one can simply run the following script : .... # file containing the list of stock symbols stocks="stocks.txt " # number of trading days t=250 # file containing the daily stock prices prices="portfolio.txt " # number of portfolios on the frontier n=100 # daily return for the risk free asset r=0.0003 # maximum daily return value considered rmax=0.01 ./data.r $ stocks ./price.r $ stocks $ t $ prices ./optimization1.r $ prices $ n $ rmax ./optimization2.r $ prices $ n $ r $ rmax .... on windows platforms one can use a simple batch file , like the following ones : .... @echo off rem file containing the list of stock symbols set stocks="stocks.txt " rem number of trading days set t="250 " rem file containing the daily stock prices set file="portfolio.txt " rem number of portfolios on the frontier set n="100 " rem daily return for the risk free asset set r="0.0003 " rem maximum daily return value considered set rmax="0.01 " rem the path to r set rpath="c:\program files\r\r-3.0.1\bin\x64\rscript.exe " call % rpath% data.r % stocks% call % rpath% price.r % stocks% % t% % file% call % rpath% optimization1.r % file% % n% % rmax% call % rpath% optimization2.r % file% % n% % r% % rmax% .... in these examples : `` stocks.txt '' is a file containing the symbols of some common stocks to be downloaded ; `` portfolio.txt '' is the file where all the relevant stock prices are extracted for the current analysis .
in order to illustrate the above results , we consider the case of a portfolio consisting of @xmath104 common stocks from it industry .
the content of the `` stocks.txt '' file is : @xmath105 a historical record of daily prices of these stocks for the last @xmath106 trading days was used to estimate the mean return and the covariance matrix .
the maximum return considered in computation is 0.01 .
the daily returns of the assets are calculated as : @xmath107 where @xmath108 is the day index , and @xmath109 is the price of asset @xmath1 at the closing day @xmath110 .
the estimate average returns and covariances are : @xmath111 @xmath112[r(j , t)-r_{j}],\ ] ] @xmath113 the asset prices and their expected returns for the considered time period are given in figure 1 .
the resulted efficient frontier is given in figure 2 .
the figure shows also the risk - return values of each stock considered , the minimum variance portfolio mvp1 , the tangency portfolio tgp , and the weights of the efficient frontier portfolios as a function of risk .
figure 3 provides the weights of the mvp1 , tgp and dep portfolios . in order to illustrate the effect of the risk free asset ( rfa )
, we consider that the investor can invest at a daily rate of return of @xmath114 .
the capital market line , together with the frontier line and the position of the new minimum variance portfolio mvp2 and of the market portfolio mp are given in figure 4 .
in this figure we also the risk dependence of the optimal portfolios from the capital market line , and we emphasize the mvp2 and mp portfolios . the weights of mvp2 and mp are plotted in figure 5 .
we have considered the standard risk - return portfolio optimization model , when both long buying and short selling , i.e. positive and negative weights , of a relatively large number of assets is allowed .
we have derived the analytical expression of the efficient frontier for a portfolio of @xmath0 risky assets , and of the capital market line when a risk - free asset is added to the model . also , we have provided an r implementation for both cases , and we have discussed in detail a numerical example of a portfolio of several risky common stocks . extract < - function(args ) { stocks < - t(read.table(args[1],sep=",")[1 , ] ) j < - length(stocks ) dir.create("data " , showwarnings = false ) path < - " http://ichart.finance.yahoo.com / table.csv?s= " for(j in 1:j ) { dat < - read.csv(paste(path , stocks[j],sep= " " ) ) write.csv(dat , file = paste("./data/",stocks[j],sep= " " ) , row.names=false , quote = false ) } } extract_price < - function(args ) { stocks < - t(read.table(args[1 ] , sep=",")[1 , ] ) j < - length(stocks ) n < - as.integer(args[2 ] ) dat < - read.csv(paste("./data/ " , stocks[1 ] , sep= " " ) ) price < - dat[1:n,5 ] if(j > 1 ) { for(j in 2:j ) { dat < - read.csv(paste("./data/ " , stocks[j ] , sep= " " ) ) price < - cbind(price , dat[1:n,5 ] ) } } write.table(price , file = paste("./data/ " , args[3 ] , sep= " " ) , row.names=false , col.names=stocks , quote = false , sep= " , " ) } returns < - function(data ) { dat < - data[[1 ] ] ; n < - nrow(dat ) - 1 ; j < - ncol(dat ) ; ret < - ( dat[1:n,1 ] - dat[2:(n+1),1])/dat[2:(n+1),1 ] ; if(j > 1 ) { for(j in 2:j ) { ret < - cbind(ret , ( dat[1:n , j ] - dat[2:(n+1),j])/ dat[2:(n+1),j ] ) } } return(list(ret , names(data[[1 ] ] ) , data[[2 ] ] , data[[3 ] ] ) ) ; } optimization < - function(returns ) { p < - colmeans(returns[[1 ] ] ) ; names(p ) < - returns[[2 ] ] ; j < - ncol(returns[[1 ] ] ) ; m < - returns[[3 ] ] ; rmax < - returns[[4 ] ] ; s < - cov(returns[[1 ] ] ) ; q < - solve(s ) ; u < - rep(1,j ) ; a < - matrix(rep(0,4),nrow=2 ) ; a[1,1 ] < - u%*%q%*%u ; a[1,2 ] < - a[2,1 ] < - u%*%q%*%p ; a[2,2 ] < - p%*%q%*%p ; d < - a[1,1]*a[2,2 ] - a[1,2]*a[1,2 ] ; f < - ( q%*% ( a[2,2]*u - a[1,2]*p))/d ; g < - ( q%*%(-a[1,2]*u + a[1,1]*p))/d ; r < - seq(0 , rmax , length = m ) ; w < - matrix(rep(0,j*m ) , nrow = j ) ; for(m in 1:m ) w[,m ] < - f + r[m]*g ; s < - sqrt ( a[1,1]*((r - a[1,2]/a[1,1])^2)/d + 1/a[1,1 ] ) ; ss < - sqrt(diag(s ) ) ; minp < - c(sqrt(1/a[1,1 ] ) , a[1,2]/a[1,1 ] ) ; wminp < - f + ( a[1,2]/a[1,1])*g ; tanp < - c(sqrt(a[2,2])/a[1,2 ] , a[2,2]/a[1,2 ] ) ; wtanp
< - f + ( a[2,2]/a[1,2])*g ; q < - sqrt(diag(1.0/ss ) ) ; x < - eigen(q%*%s%*%q ) ; v < - q%*%x$vec ; for(j in 1:j ) v[,j ] < - v[,j]/(u%*%v[,j ] ) ; sv < - rv < - rep(0 , j ) ; for(j in 1:j ) { rv[j ] < - t(v[,j])%*%p ; if(rv[j ] < 0 ) { rv[j ] < - -rv[j ] ; v[,j ] < - -v[,j ] ; } sv[j ] < - sqrt(t(v[,j])%*%s%*%v[,j ] ) ; } return(list(s , r , ss , p , minp , tanp , wminp , wtanp , w , v , sv , rv ) ) ; } plot_results<- function(data , returns , results ) { dat < - log(data[[1 ] ] ) ; m
< - nrow(dat ) ; j < - ncol(dat ) ; ymax = max(dat ) ; ymin = min(dat ) mycolors < - rainbow(j+1 ) ; s < - results[[1 ] ] ; r < - results[[2 ] ] ; ss < - results[[3 ] ] ; p < - results[[4 ] ] ; minp < - results[[5 ] ] ; tanp < - results[[6 ] ] ; wminp < - results[[7 ] ] ; wtanp < - results[[8 ] ] ; f < - t(results[[9 ] ] ) ; v < - results[[10 ] ] ; sv < - results[[11 ] ] ; rv < - results[[12 ] ] ; postscript(file="./results1/fig1.eps " , onefile = false , horizontal = false , height=10 , width=5 ) ; par(mfrow = c(2,1 ) ) ; i d < - c(1:nrow(dat ) ) ; plot(id , rev(dat[,1 ] ) , ylim = c(ymin , ymax ) , type="l " , col = mycolors[1 ] , xlab="day " , ylab="log(price ) " , main = " asset prices " ) ; if(j >
1 ) { for(j in 2:j ) { lines(id , rev(dat[,j ] ) , type="l " , col = mycolors[j ] ) ; } } legend("topleft " , names(dat ) , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; ret < - returns[[1 ] ] ; ymax = max(ret ) ; ymin = min(ret ) ; i d < - c(1:nrow(ret ) ) ; plot(id , rev(ret[,1 ] ) , ylim = c(ymin , ymax ) , type="l " , col = mycolors[1 ] , xlab="day " , ylab="returns " , main = " asset returns " ) ; if(j >
1 ) { for(j in 2:j ) { lines(id , rev(ret[,j]),type="l",col = mycolors[j ] ) ; } } legend("topleft " , returns[[2 ] ] , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; postscript(file="./results1/fig2.eps " , onefile = false , horizontal = false , height=10 , width=5 ) ; par(mfrow = c(2,1 ) ) ; plot(s , r , xlim = c(0,max(s ) ) , ylim = c(min(r , p ) , max(r , p ) ) , type="l " , col="blue " , xlab="risk " , ylab="return " , main = " efficient frontier , mvp1 , tgp " ) ; points(ss , p , pch=19 , col = mycolors ) ; text(ss , p , pos=4 , cex=0.5 , names(p ) ) ; points(sv[1 ] , rv[1 ] , pch=15 , col="black " ) ; text(sv[1 ] , rv[1 ] , pos=4 , cex=0.5 , " dep " ) ; points(minp[1 ] , minp[2 ] , pch=19 , col="black " ) ; text(minp[1 ] , minp[2 ] , pos=2 , cex=0.5 , " mvp1 " ) ; points(tanp[1 ] , tanp[2 ] , pch=19 , col="black " ) ; text(tanp[1 ] , tanp[2 ] , pos=2 , cex=0.5 , " tgp " ) ; lines(c(0,max(s ) ) , c(0,max(s)*tanp[2]/tanp[1 ] ) , lty=3 ) ; abline(h=0 , lty=2 ) ; abline(v=0 , lty=2 ) ; plot(s , f[,1 ] , xlim = c(0,max(s ) ) , ylim = c(min(f),max(f ) ) , col = mycolors[1 ] , type="l " , xlab="risk " , ylab="portfolio weights " , main = " efficient portfolio weights " ) ; if(j >
1 ) { for(j in 2:j ) { lines(s , f[,j ] , type="l " , col = mycolors[j ] ) ; } } abline(h=0 , lty=2 ) ; abline(v = minp[1 ] , lty=3 ) ; abline(v = tanp[1 ] , lty=3 ) ; text(minp[1 ] , min(f ) , pos=4 , cex=0.5 , " mvp1 " ) ; text(tanp[1 ] , min(f ) , pos=4 , cex=0.5 , " tgp " ) ; legend("topleft " , names(p ) , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; postscript(file="./results1/fig3.eps " , onefile = false , horizontal = false , height=10 , width=5 ) ; par(mfrow = c(2,1 ) ) ; barplot(wminp , main="minimum variance portfolio 1 ( mvp1 ) " , xlab="assets " , ylab="weights " , col = mycolors , beside = true ) ; abline(h=0 , lty=1 ) ; legend("topleft " , names(p ) , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; barplot(wtanp , main="tangency portfolio ( tgp ) " , xlab="assets " , ylab="weights " , col = mycolors , beside = true ) ; abline(h=0 , lty=1 ) ; legend("topleft " , names(p ) , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; barplot(v[,1 ] , main="dominant eigen - portfolio ( dep ) " , xlab="assets " , ylab="weights " , col = mycolors , beside = true ) ; abline(h=0 , lty=1 ) ; legend("topleft " , names(p ) , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; } read_data < - function(args ) { data < - read.table(paste("./data/ " , args[1 ] , sep= " " ) , header = true , sep= " , " ) return(list(data , as.integer(args[2 ] ) , as.double(args[3 ] ) , as.double(args[4 ] ) ) ) ; } returns < - function(data ) { dat < - data[[1 ] ] ; n < - nrow(dat ) - 1 ; j < - ncol(dat ) ; ret < - ( dat[1:n,1 ] - dat[2:(n+1),1])/dat[2:(n+1),1 ] ; if(j > 1 ) { for(j in 2:j ) { ret < - cbind(ret , ( dat[1:n , j ] - dat[2:(n+1),j])/ dat[2:(n+1),j ] ) ; } } return(list(ret , names(data[[1 ] ] ) , data[[2 ] ] , data[[3 ] ] , data[[4 ] ] ) ) ; } foptimization < - function(returns ) { p < - colmeans(returns[[1 ] ] ) ; names(p ) < - returns[[2 ] ] ; j < - ncol(returns[[1 ] ] ) ; m < - returns[[3 ] ] ; r < - returns[[4 ] ] ; rmax < - returns[[5 ] ] ; s < - cov(returns[[1 ] ] ) ; q < - solve(s ) ; u < - rep(1,j ) ; a < - matrix(rep(0,4),nrow=2 ) ; a[1,1 ] < - u%*%q%*%u ; a[1,2 ] < - a[2,1 ] < - u%*%q%*%p ; a[2,2 ] < - p%*%q%*%p ; d < - a[1,1]*a[2,2 ] - a[1,2]*a[1,2 ] ; r
< - seq(r , rmax , length = m ) ; s < - sqrt ( a[1,1]*((r - a[1,2]/a[1,1])^2)/d + 1/a[1,1 ] ) ; ss < - sqrt(diag(s ) ) ; cml < - c(sqrt(a[1,1]*r*r - 2*a[1,2]*r + a[2,2 ] ) , r ) ; z < - ( r - r)/cml[1 ] ; f < - q%*%(p - r*u)/(cml[1]*cml[1 ] ) ; wcml < - matrix(rep(0,j*m ) , nrow = j ) ; wf < - rep(0,m ) ; for(m in 1:m ) { wcml[,m ] < - ( r[m ] - r)*f ; wf[m ] < - 1 - wcml[,m]%*%u ; } wcml < - rbind(wcml , t(wf ) ) ; mp < - c(cml[1]/(a[1,2 ] - a[1,1]*r ) , ( a[2,2 ] - a[1,2]*r)/(a[1,2 ] - a[1,1]*r ) ) ; wmp < - q%*%(p - r*u)/(a[1,2 ] - a[1,1]*r ) ; minp < - c(sqrt(1/a[1,1 ] ) , cml[1]*sqrt(1/a[1,1 ] ) + r ) ; wminp < - ( minp[2 ] - r)*f ; wfminp < - 1- t(wminp)%*%u ; wminp
< - rbind(wminp , wfminp ) ; return(list(s , z , r , ss , p , cml , wcml , mp , wmp , minp , wminp ) ) ; } plot_results<- function(data , returns , results ) { dat < - log(data[[1 ] ] ) ; m < - nrow(dat ) ; j < - ncol(dat ) ; ymax = max(dat ) ; ymin = min(dat ) ; mycolors < - rainbow(j ) ; s < - results[[1 ] ] ; z < - results[[2 ] ] ; r < - results[[3 ] ] ; ss < - results[[4 ] ] ; p < - results[[5 ] ] ; cml < - results[[6 ] ] ; mp < - results[[8 ] ] ; wmp < - results[[9 ] ] ; minp < - results[[10 ] ] ; wminp < - results[[11 ] ] ; postscript(file="./results2/fig1.eps " , onefile = false , horizontal = false , height=10 , width=5 ) ; par(mfrow = c(2,1 ) ) ; i d < - c(1:nrow(dat ) ) ; plot(id , rev(dat[,1 ] ) , ylim = c(ymin , ymax ) , type="l " , col = mycolors[1 ] , xlab="day " , ylab="log(price ) " , main = " asset prices " ) ; if(j >
1 ) { for(j in 2:j ) { lines(id , rev(dat[,j]),type="l",col = mycolors[j ] ) ; } } legend("topleft " , names(dat ) , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; ret < - returns[[1 ] ] ; ymax = max(ret ) ; ymin = min(ret ) ; i d < - c(1:nrow(ret ) ) ; plot(id , rev(ret[,1]),ylim = c(ymin , ymax),type="l " , col = mycolors[1 ] , xlab="day " , ylab="returns " , main = " asset returns " ) ; if(j >
1 ) { for(j in 2:j ) { lines(id , rev(ret[,j]),type="l",col = mycolors[j ] ) } } legend("topleft " , returns[[2 ] ] , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; postscript(file="./results2/fig2.eps " , onefile = false , horizontal = false , height=10 , width=5 ) ; par(mfrow = c(2,1 ) ) ; mycolors < - rainbow(length(p)+1 ) ; plot(s , r , xlim = c(0 , max(s)),ylim = c(min(r , p),max(r , p ) ) , type="l " , col="blue " , xlab="risk " , ylab="return " , main = " capital market line , mvp2 , mp " ) ; points(ss , p , pch=19 , col = mycolors ) ; text(ss , p , pos=4 , cex=0.5 , names(p ) ) ; points(mp[1 ] , mp[2 ] , pch=19 , col="black " ) ; points(mp[1 ] , mp[2 ] , pch=19 , col="black " ) ; text(mp[1 ] , mp[2 ] , pos=2 , cex=0.6 , " mp " ) ; points(minp[1 ] , minp[2 ] , pch=19 , col="black " ) ; text(minp[1 ] , minp[2 ] , pos=2 , cex=0.6 , " mvp2 " ) ; text(0 , cml[2 ] , pos=4 , cex=0.5 , " rfa " ) ; lines(c(0 , max(s ) ) , c(cml[2 ] , max(s)*cml[1 ] + cml[2 ] ) , lty=3 ) ; abline(h=0 , lty=2 ) ; abline(v=0 , lty=2 ) ; f < - t(results[[7 ] ] ) ; mycolors < - rainbow(j+1 ) ; plot(z , f[,1 ] , xlim = c(0,max(z ) ) , ylim = c(min(f),max(f ) ) , type="l " , col = mycolors[1 ] , xlab="risk " , ylab="portfolio weights " , main="cml portfolio weights " ) ; if(j >
1 ) { for(j in 2:j+1 ) { lines(z , f[,j],type="l",col = mycolors[j ] ) ; } } abline(h=0 , lty=2 ) ; abline(v = mp[1 ] , lty=3 ) ; text(mp[1 ] , min(f ) , pos=4 , cex=0.5 , " mp " ) ; abline(v = minp[1 ] , lty=3 ) ; text(minp[1 ] , min(f ) , pos=4 , cex=0.5 , " mvp2 " ) ; legend("topleft " , c(names(p ) , " rfa " ) , cex=0.5 , pch = rep(15 , j+1 ) , col = mycolors ) ; postscript(file="./results2/fig3.eps " , onefile = false , horizontal = false , height=10 , width=5 ) ; par(mfrow = c(2,1 ) ) ; barplot(wminp , main="minimum variance portfolio 2 " , xlab="assets " , ylab="weights " , col = mycolors , beside = true ) ; abline(h=0 , lty=1 ) ; legend("topleft " , c(names(p),"rfa " ) , cex=0.5 , pch = rep(15 , j+1 ) , col = mycolors ) ; barplot(wmp , main="market portfolio " , xlab="assets " , ylab="weights " , col = mycolors , beside = true ) ; abline(h=0 , lty=1 ) ; legend("topleft " , names(p ) , cex=0.5 , pch = rep(15 , j ) , col = mycolors ) ; } s. k. mishra , g. panda , b. majhi , r. majhi , `` improved portfolio optimization combining multiobjective evolutionary computing algorithm and prediction strategy , '' _ proceedings of the world congress on engineering 2012 vol i , wce 2012 , july 4 - 6 , 2012 , london , u.k._. | we consider the problem of finding the efficient frontier associated with the risk - return portfolio optimization model . we derive the analytical expression of the efficient frontier for a portfolio of @xmath0 risky assets , and for the case when a risk - free asset is added to the model .
also , we provide an r implementation , and we discuss in detail a numerical example of a portfolio of several risky common stocks .
portfolio optimization , efficient frontier , r. |
sn 1992ar is one of the supernovae ( sne ) discovered by the caln / tololo collaboration ( hamuy et al .
it was found by r. antezana , on a plate taken with the ctio curtis
schmidt telescope on ut 1992 july 27.23 ( hamuy & maza 1992 ) , close to a tight group of three anonymous galaxies ( see figure [ fi : chart ] ) .
the new object was confirmed as a sn 6 days later , through a spectrum taken with ctio 4.0 m telescope .
spectra of the nearby galaxies indicated that the group was at a redshift of @xmath30.15 , which made this the most distant sn discovered during the caln / tololo survey .
surprisingly , the spectrum of the sn revealed that the object more closely resembled a type ic event rather than a type ia .
preliminary calibration of a local sequence of standards and point spread function fitting photometry of the sn showed that sn 1992ar could be even brighter than a type ia sn , depending on k corrections which were unknown at the time .
these brightness estimates were in strong contradiction with the expectation that type ib and ic sne ( those formerly known as `` peculiar '' type ia sne ) were subluminous with respect to type ia .
the belief that type ib and ic were subluminous with respect to type ia sne was rooted in both the observational record and theoretical interpretation .
all sne of these types , so far , had been 1.01.5 magnitudes dimmer than a typical type ia sn .
the fact that they appear always associated with regions of very recent star formation suggested that they originated in stars of large main sequence mass which had lost the outer envelope and , hence , should explode through core collapse like type ii sne .
it is not yet clear what is the `` typical '' amount of that a core collapse explosion ejects .
although some events produced as much as @xmath30.3 ( schmidt et al .
1994 ) , comparative analysis of their late time light curves suggest that most of them eject between 0.04 and 0.10 ( patat et al .
this , the successful interpretation and modeling of sn 1987a in terms of the ejection of 0.07 of ( woosley 1988 ; 1990 ) , and the modeling of other type ii and stripped
envelope sne in terms of explosions ejecting similar amounts of ( wheeler et al .
1993 , shigeyama et al .
1994 , woosley et al .
1994 , woosley et al .
1995 , clocchiatti et al .
1996a ) strengthened the case for low production in typical core collapse sne .
the extreme brightness of sn 1992ar raises the question , therefore , of whether it might actually be a peculiar type ia event which just happens to bear a close spectroscopic resemblance with a sn ic .
ccd observations at only three epochs were obtained from ctio , the latest one 17 days after discovery .
in addition , very late deep images of the field , when the sn brightness had faded beyond detection , were taken with the blanco 4.0 m telescope to allow subtraction of the galaxy background . using this restricted database , kohnenkamp ( 1995 ) ,
did a preliminary study of this sn .
additional data , however , had been obtained at eso , where researchers working under the key program on sne followed sn 1992ar for more than 120 days after discovery .
the ctio and eso data , when combined , provide a well sampled light curve with good signal - to - noise ratio , spanning nearly 105 days in the rest frame of the sn .
these data permit a detailed comparison of sn 1992ar with typical type ia and ib / c events . from this comparison
, the nature of sn 1992ar can be finally established . in section 2
we present the spectrum of sn 1992ar .
we describe the observations , reduction , and the spectrum itself , and compare it with spectra of nearby , well observed , type ia and ic sne . in section 3
we describe the ccd photometric observations , the process used to eliminate the background light , and the calibration of the photometry in a standard system .
section 4 of this paper describes how we obtained a photographic magnitude from the discovery plate .
section 5 describes the k corrections we used to translate the observed photometry into the rest frame of the sn . in section 6 , finally , we reach the point where a light curve of sn 1992ar can be presented . in section 7
we discuss our results and present our conclusions .
the confirming spectrum of sn 1992ar was obtained at the ctio 4.0 m blanco telescope on ut 1992 august 2.34 by m. hamuy and r. williams .
the cassegrain spectrograph and a 1240x400 reticon ccd detector were used .
the wavelength range covered was 32487500 , with a dispersion of 3.6 pix@xmath4 ( approximate resolution @xmath37 ) .
two exposures of 10 minutes each were obtained at an airmass of 1.07 , approximately , bracketed by exposures of hear comparison lamps .
spectra of the three galaxies around the sn ( see fig . [
fi : chart ] ) and two flux standard stars of the list of hamuy et al .
( 1994 ) were also obtained .
all the objects were observed through a 2 arcsec slit , which was not oriented along the parallactic angle .
the relative flux calibration of the spectrum is , hence , uncertain due to light loss by atmospheric dispersion .
the absolute flux calibration is uncertain , as well , since the night was not photometric .
image reduction and calibration followed the typical procedures .
the spectrum images were divided by normalized flat
field images built from several exposures of the dome illuminated by two light sources with different spectral distribution to provide good signal - to - noise ratio in all the spectral range .
spectra of the sn were extracted and sky subtracted .
special care was taken in fitting the sky underneath the sn profile , since this sn appeared superimposed on a complicated background ( see fig . [
fi : chart ] ) .
the spectra were linearized in wavelength using the hear lamp exposures , corrected by atmospheric extinction using the mean extinction curve for ctio , and calibrated in flux using the sensitivity curve for the night , obtained from the spectrophotometric standards .
the two final sn spectra were normalized to the same mean value ( multiplying the weakest by a constant ) , and , finally , combined .
spectra of the three neighboring galaxies were obtained and reduced in the same way .
the spectrum of the sn is plotted both in figures [ fi : spec1 ] , and [ fi : spec2 ] , and the spectra of the neighboring galaxies are given in figure [ fi : specg ] .
the redshifts for each galaxy , obtained from the average of several emission and/or absorption lines , are , 0.1452 , 0.1450 , and 0.1462 , for galaxies a , b and c , respectively .
they agree to within @xmath5360 km s@xmath4 , confirming that the galaxies are physically associated .
note that narrow , weak [ oii ] 3727 , , and [ oiii ] 5007 emission lines at [email protected] are also present in the sn spectrum , pointing at a close association with the group .
although it is not completely clear which of the galaxies was the actual host , circumstantial evidence suggests sn 1992ar was more closely related with galaxies a and/or b than with galaxy c. on the one hand , the projected positions of galaxies a and b are closer to the position of the sn than that of galaxy c. on the other , galaxies a and b show the strong emission lines characteristic of galaxies with starburst ( kennicutt , 1992 ) , and type ic sne are known to be strongly associated with massive star forming regions in nearby galaxies ( van dyk , hamuy & filippenko , 1996 ) . using the b & v transmission curves of bessel ( 1990 )
, we found that the color of the spectrum of sn 1992ar was b@xmath7v= 0.85 .
it is difficult to attach an error bar to this estimate .
the relative spectrophotometry from ctio spectra is typically outstanding ( see for example figure 2 in phillips et al .
1992 ) , with rms uncertainty on the order of a few hundredths of a magnitude , similar to that of ccd photometry .
we expect , however , that this would not be the case for sn 1992ar .
one the one hand , this sn was dimmer than the examples available for comparison , and the background of the parent galaxies was more complicated than usual .
on the other , our spectrum is not corrected for atmospheric dispersion .
although it is difficult to size the impact of the background on the spectrophotometry , an estimate of the atmospheric dispersion can be made .
according to table 1 of filippenko ( 1982 ) the separation of the intensity peaks at the effective wavelengths of the b and v filters is smaller than 0.2 arcseconds ( i.e. , @xmath810% of the slit width ) if the airmass is @xmath31.07 . assuming a gaussian profile for the seeing and integrating the intensity profiles inside of the slit
, this separation implies that the flux in b could have been reduced , at most , by 2% ( i.e. our b@xmath7v color would be at most 0.02 mag too red ) .
to be on the safe side , however , we assume an uncertainty of @xmath9 mag , about 5 times larger than the uncertainty in the photometry taken the day before , and 10 times larger than the expected systematic error due to atmospheric dispersion alone .
finally , an estimate of the reddening from foreground interstellar matter can be made from the absence of nai d lines . according to burstein & heiles ( 1984 ) the galactic extinction in the direction of sn 1992ar is negligible .
regarding interstellar matter in the parent galaxy , the noise peaks in the region of the nai d lines at the redshift of sn 1992ar imply an equivalent width smaller than @xmath30.5 . in the absence of a better solution
, this upper limit can be used to estimate an upper limit to the color excess .
clocchiatti et al .
( 1995 ) find that , in the case of sn 1993j , @xmath10 mag @xmath4 , which would imply in the case of sn 1992ar a maximum e(b@xmath7v ) of @xmath30.1 , or @xmath11 mag .
we will neglect in what follows any correction for reddening .
however , it should be kept in mind that our results could still be consistent with modest amounts of foreground extinction , up to a limit of @xmath30.3 mag in v. the spectrum of sn 1992ar is compared in fig .
[ fi : spec1 ] with spectra of the `` normal '' type ia sn 1989b and the peculiar type ia sn 1991 t .
the phases of the 1989b and 1991 t spectra were chosen to be consistent with the phase implied for our spectrum of 1992ar ( were it a sn ia ) by a spectroscopic matching technique ( see 6.1 ) .
we can see that the similarities of the sn 1992ar spectrum with those of type ia sne are , mainly , the broad features to the blue of @xmath35300 .
these absorption features and p cygni profiles are , from red to blue , due to feii 5169 , feii 5018 , feii `` 4555 '' , feii `` 4274 '' , and caii h & k. since they are present in the spectra of sne of all types , including type ii , their potential to differentiate a sn from another is very limited .
features of caii , feii , and nai , which are common to type ia and ic sne , and the siii 6355 line , which , although weaker than in type ia , also appears in type ic sne , are labeled with lower case letters . even though the similarities in this spectral region are clear , there are a few differences worth noting .
the features of sn 1992ar are shallower and broader , and the spectra of type ia sne show narrow cores in the absorption lines .
these differences are particularly noticeable between 3900 and 4200 , and between 4700 and 5100 . in the former region ,
the main difference is made by the feature labeled with letter a , which corresponds to siii 4130 . in the latter region , there is an additional feature , labeled with letter b , which corresponds to siii 5051 .
it is , however , in the region to the red of @xmath35200 that the distinct character of the spectrum of sn 1992ar becomes clear .
first , the strong siii 6355 line , centered between 6100 and 6200 in sn 1989b and 1991 t , is absent in sn 1992ar .
second , the much weaker but also distinctly present absorption centered near 5500 in the type ia sne , typically attributed to sii 5612 and sii 5654 ( labeled with letter d in the figure ) , is also absent .
there is another , much weaker , feature about 200 to the blue , labeled with letter c in the figure , which corresponds to sii 5468 .
the signal - to - noise ratio of the spectrum of sn 1992ar is not good enough to decide whether this small feature is present or not .
strong siii lines and clearly detectable sii lines are characteristic features of type ia sne .
the fact that sn 1992ar does not show them is a clear indication that , if it were a type ia sn , it would be even more spectroscopically peculiar than sn 1991 t .
consistent with this , fig .
[ fi : spec2 ] shows that there are no major differences between the spectra of typical type ic sne 1983v or 1987 m and that of sn 1992ar . with the exception of the caii h & k lines ,
the spectrum of sn 1983v is remarkably similar to that of sn 1992ar .
however , since type ic sne do show differences in the strength of the caii lines , the weaker spectral features are not a major concern .
paradoxically , one of the notable differences in this figure is , again , the siii 6355 line . although it is clear that the spectra shown fig .
[ fi : spec2 ] are type ic , one sees that this line was stronger in sn 1987 m than in the other two .
we have collected 10 ccd images of sn 1992ar taken from telescopes at ctio and eso .
a description of the dates , telescopes used , observers , and other relevant information is given in table [ ta : ccdobs ] .
the images , taken through standard johnson b , v ( johnson 1955 ) , and cousins r ( cousins 1976a , 1976b , 1978 ) filters were reduced following standard procedures .
the ccd pedestal was corrected using its overscan region , any remainder 2-dimensional pattern was removed using bias images when necessary .
no correction for dark current was required .
finally , pixel to pixel sensitivity variations were removed using median dome or sky flat field images .
sn 1992ar appeared on a complicated background ( see fig .
[ fi : chart ] ) .
the nearby galaxies 1 and 2 mean that the sky near the sn is bright and has a large gradient .
in addition , the sn is centered on a valley of the brightness distribution .
this configuration makes both aperture and psf fitting photometry bad prospects . in order to do unbiased photometry
, we resorted to eliminating the light of the nearby galaxies by taking very late - epoch images of the region , with good signal - to - noise ratio .
information on these images , which we refer to in what follows as the final epoch images , is given in table [ ta : feimages ] .
we used the image matching algorithms described by phillips & davis ( 1995 ) , proceeding in detail as follows : ( a ) we shifted and rotated the final epoch image so that the centers of the stars were in coincidence with those of the image with the sn ; ( b ) we found the difference kernel that , when convolved with the image of the narrowest psf ( usually this is the final epoch image ) , would degrade it to match the psf of the other image , and convolved the image with better seeing with that kernel ; ( c ) we found the linear transformation in intensities to make the final epoch image match the intensities of the image with the sn , and applied that transformation to the final epoch image ; ( e ) we subtracted the region around the sn of the transformed final epoch image from the image with the sn .
after this procedure we found that the background of the galaxies near the sn had been removed , and that the sky underneath the sn was flat .
we did relative photometry of the sn on the images with the background subtracted . using the tasks of the package daophot in iraf we built a mean psf for each frame from the profiles of the stars in the field and found the instrumental magnitudes of the local sequence of standards and the sn by fitting this psf function . the sequence of local standard stars indicated in fig .
[ fi : chart ] was calibrated in johnson b and v colors using observations taken with ctio 0.9 m telescope on august 22 , 1993 . on that night , images of the region of sn 1992ar were taken together with standard stars of the lists of graham ( 1982 ) , menzies et al .
( 1989 ) , and landolt ( 1992 ) .
the zero points , extinction coefficients , and color terms were fitted to find the photometric solution for the night , and the b and v magnitudes given in table [ ta : std ] were computed from this solution .
we did not have a photometric night with enough standard star observations to compute all the coefficients in the photometric solution for the cousins r filter .
the final epoch image , however , was taken on a photometric night and our set of images included three standard stars of the list of landolt ( 1992 ) .
we used these stars , a mean extinction coefficient for la silla from the database of the photometry group of the geneva observatory ( burki et al .
1995 ) , and a typical color term for the instrument used ( turatto , 1996 ) , to find the zero point of the calibration in r. due to the characteristics of this solution , there could be a systematic error in the calibrated r magnitude for the local sequence of standards .
it must be small , however , of the order of the uncertainties given in table [ ta : ccdobs ] , since the two color diagram of the local standard sequence looks reasonable .
we decided to give the r values out of completeness . in the event that these particular observations of sn 1992ar become important ( to estimate the foreground extinction by comparing the late time v - r color with those of future type ic sne , for example ) the sequence can be recalibrated and a correction to the values given in table [ ta : ccdobs ] could be computed . using the instrumental magnitudes of the stars in the frames with the sn and their b , v , and r magnitudes , we computed a zero point for each frame and applied it to the instrumental magnitude of the sn .
the color terms for all the combinations of ccd and instruments used is small , so color term corrections to the magnitude of the sn relative to the local standard stars are expected to be smaller than the photometric uncertainty .
consistent with this , we found the zero points to be independent of the color of the local sequence stars .
the magnitudes of the sn , calibrated in this way , are given in table [ ta : ccdobs ] .
as it will become clear in the next section , to find the nature of sn 1992ar requires us to use all possible data at hand , including its discovery plate .
a complication in this regard is that the search plates of the caln / tololo survey were taken without a filter .
however , the standard kodak iia0 emulsion , with a red cut off very similar to that of the johnson b band , was employed .
this , together with the fairly red color of sn 1992ar , suggest that it should be possible to get a good estimate of its b magnitude from the photographic magnitude .
there were three schmidt plates in the database of cerro caln observatory with images of the region of sn 1992ar ( wischnjewsky 1997 ) .
one of them is the first epoch image , which was taken on june 26.5 1992 , the second is the discovery image , taken on july 27.3 1992 , the third is an additional plate taken on august 2.5 1992 .
ideally , these plates could have been used to digitize the region of the sn and sequence of local photometric standards in all the plates , transform the photographic density to an approximate intensity scale , use the plate of june 26.5 to estimate the background underneath the sn and subtract it from the other two plates , compute magnitudes for the local standards and the sn , analyze the correlation between the instrumental photographic magnitudes and the calibrated magnitudes for the standard stars in all plates , and use this correlation to estimate the photographic magnitude of the sn in a standard passband . since by august 2.5 the light curve is known from ccd observations , the calibration of the august plate would be useful to check the accuracy of the photographic photometry .
we were able to accomplish this project only partially .
on the one hand , we found that the photographic density of the sky in the region of the sn was only marginally detectable in the june image ( negative observation ) and , hence , dominated by noise .
this made it useless to model the background of the sn .
in addition , the seeing of the plate of august 2.5 was worse than that of the discovery plate .
since the sn was fainter on august 2.5 than on july 27.3 , the plate of august provides only a marginal detection of the sn , not good enough to set a meaningful constraint on the precision of the photographic photometry .
we decided , then , to use our final epoch ccd images to model the background of the sn in the digitized discovery plate , and desist in our desire to check the accuracy of the photographic photometry .
we proceeded as follows .
building upon the suggestions of stetson ( 1979 ) , especially his equations 911 we took the photographic density to be proportional to the intensity , an approximation which holds true in the limit of stellar images of low intensity above the background .
we transformed our final epoch ccd images in the b filter to the pixel scale of the plate , studied the correlation ( pixel by pixel ) of the intensity recorded by the ccd and the photographic density in regions of low density , and found that they were actually consistent with a linear transformation .
we found the linear transformation that best fit the correlation , and applied it to the ccd image .
we then took a small region of the transformed ccd image , centered on the sn position , and subtracted the transformed ccd intensity distribution of this region from the photographic density .
we found that the sky around the sn was acceptably well subtracted , leaving residuals only near the cores of the nearby galaxies ( where the assumption of ccd intensity proportional to photographic density was expected to break down ) .
we used this final image to compute the photographic index of the local standard sequence and sn .
we did not have enough stars to model the photographic psf shape and correct for saturation at high photographic density ( as in stetson , 1979 ) .
we resorted , therefore , to computing our photographic index using aperture photometry .
using 7 stars imaged both in the ccd and the plate , we found a tight linear correlation between our photographic index and the b magnitude of the stars in the local sequence .
the scatter of the correlation , which spans 2.6 magnitudes in the photographic index , is 0.05 mag .
the residuals of the fit , in addition , display a small but statistically significant color term . from the correlation and color term obtained ,
the b magnitude of sn 1992ar on the discovery plate is given by , @xmath12 the color term in this equation is the same , within the uncertainty , as those found by pierce & jacoby ( 1995 ) and arp ( 1961 ) , in transforming magnitudes from iiao plates to b. the uncertainty given in equation [ eq : bphot ] results from combining in quadrature the uncertainty in the instrumental magnitude of the sn and the rms of the transformation of the stars in the local sequence .
there is an additional uncertainty , that we can not evaluate , because the sn is fainter than the faintest star in the local sequence of the plate .
the linear correlation was _ extrapolated _ @xmath13 magnitudes to reach the photographic index of the sn .
it is clear , in addition , that in order to derive a b magnitude we need to estimate the ( b@xmath7v ) color of the sn .
this estimate will introduce another source of uncertainty .
at a redshift of [email protected] , it is important to consider the modification of the flux distribution due to the doppler effect , and apply the corresponding k corrections to the observed magnitudes , if comparison with sne observed in the nearby universe is to be done .
we have computed k corrections for the b and v magnitudes of sn 1992ar following the usual prescriptions ( see hamuy et al .
1993 , kim et al .
1996 , or schmidt 1998 , for modern discussions ) , and using spectra of sn 1983v ( clocchiatti et al . 1997 ) , sn 1987 m ( filippenko , porter & sargent 1990 ) , and sn 1994i ( filippenko et al .
1995 , clocchiatti et al .
1996b , jeffery et al .
1994 ) . a deeper consideration of k corrections for type ic sne will be given elsewhere ( clocchiatti 1999 ) .
it is necessary , however , to give a short discussion here .
type ic sne present a more complex problem than type ia sne , when it comes to applying k corrections . consider first that , as emphasized by hamuy et al .
( 1993 ) , the k - corrections for a sn qualitatively behave as a color .
now , even when type ia sne are not as homogeneous as it was thought just a few years ago , their differences in color evolution do not translate into major differences in the k corrections computed from spectra of different sne .
even when intrinsic differences in the sne increase the scatter in the k corrections the results are still reasonably consistent with a unique k correction function for all sne , at least at redshifts smaller than [email protected] ( hamuy et al .
type ic sne , on the other hand , present a much wider range in the speed of the light curve and color evolution . some of them are extremely fast , like sn 1994i ( richmond et al . 1996 ) , while some others are slow , like sn 1983v ( clocchiatti et al .
these two in particular were considerably faster and slower , respectively , than a typical type ia sn .
accordingly , the k corrections for a type ic sn are not a unique function of time but critically depend on the speed of the light curve . in figure
[ fi : kcorr ] we give the k correction in v for a fast and a slow type ic sn at [email protected] as a function of rest frame time since maximum in v , together with the adopted fitting functions .
the rms of the interpolation is 0.06 mag , for fast ic sne , and 0.05 for slow ic sne .
the rms around the fit for the k correction in b ( not given in fig .
[ fi : kcorr ] ) is 0.06 mag .
we will take these values as a measure of the uncertainty introduced by the k corrections .
in addition , it was necessary to test the possibility that sn 1992ar was a type ia sn with a peculiar spectrum . lacking anything better , k corrections based on typical type ia sn spectra
were used for this .
we took those given by hamuy et al .
( 1993 ) , interpolating their values using piece@xmath7wise low degree polynomial functions .
the adopted corrections are also given in fig .
[ fi : kcorr ] .
the rms around the interpolating function is 0.05 mag in this case .
a different k correction is needed for the b magnitude obtained from the discovery plate , since the sensitivity function of the telescope plus plate system is not matched by a standard filter .
we used the sensitivity function computed by pierce & jacoby ( 1995 ) for the kitt peak schmidt telescope plus a iia0 plate , under the assumption that it must be similar to that of the tololo curtis - schmidt telescope with the same kind of plate . an additional problem in this case
is that this sensitivity function is very blue , and it is difficult to find sne spectra blue enough to cover this passband .
according with the different sn type and phase we were trying to replicate , we used literature spectra of the type ic sn 1994i and the type ia sn 1992a , together with their hst spectra ( obtained by the sins collaboration , jeffery et al .
1994 ) , to compute these k corrections . the k correction functions displayed in fig .
[ fi : kcorr ] illustrate the problem of using the same standard filter in the rest and redshifted frames . not only are the resulting k corrections function of time , but the time scales of the variations _ and their amplitudes _ , are similar to those which differentiate one type of sne light curve from the others , even at the comparatively small redshift of sn 1992ar .
this effect emphasizes the importance of observing high redshift sne in filters that provide a better match to the wavelength range where the light curves used for comparison were observed .
these matches can be achieved via the coincidence of effective wavelengths of standard filters for a given redshift ( kim , goobar & perlmuter 1996 ) , or the construction of new photometric systems which are the redshifted counterparts of the standard one ( schmidt et al .
undilating the time scale using the redshift measured from the emission lines in the nearby galaxies , it is clear that the observed photometry is consistent with that of a sn ( see figure [ fi : lcexample ] ) .
it is clear , also , that the light curve does not correspond to that of a typical type ii plateau event .
finally , it would seem that the confirming spectrum was taken close to maximum light in v. the strong time variation of the k corrections , however , as illustrated in fig .
[ fi : lcexample ] , implies that obtaining the correct final light curve requires a decision as to which of the functions plotted in fig .
[ fi : kcorr ] is the correct k correction to apply .
the analysis presented in
2 supports the original identification of sn 1992ar as a type ic sn ( hamuy & phillips 1992 ) .
the sn is , however , surprisingly bright for being a ic . if h@xmath14 km s@xmath4 and q@xmath15 , then the distance modulus is 39.28 magnitudes , and the first ccd point implies m@xmath16 mag , without applying any k correction ( which could make the sn brighter ) , or residual foreground extinction ( which _ will _ make the sn brighter ) .
km s@xmath4 here and throughout the paper in order to compare the absolute brightness of sn 1992ar with those of the other caln / tololo sne given by hamuy et al .
( 1996 ) . ]
according to hamuy et al .
( 1996 ) , the absolute v magnitudes of type ia sne range between @xmath719.5 and @xmath718.5 , approximately , for the same value of h@xmath17 .
sn 1992ar therefore compares in brightness with the brightest type ia sne in the caln / tololo survey .
other type ic sne observed near maximum are clearly subluminous with respect to typical sne ia .
sn 1983v ( clocchiatti et al .
1997 ) , a slow type ic sn , reached m@xmath18 -18.1 @xmath50.2 , with the uncertainty dominated by the reddening estimate .
sn 1994i ( richmond et al .
1996 ) , reached m@xmath18 -18.1 @xmath19 , with the big uncertainty due to a large foreground extinction .
the fact that sn 1992ar was as much as a magnitude brighter than these two well observed nearby type ic sne raises the question of whether it could have been a type ia sn with a peculiar spectrum .
we analyze this possibility in the following subsection .
the spectrum of sn 1992ar , as we discussed before , is more consistent with the spectrum of a type ic sne than with that of a type ia .
this is so , however , mainly because of the missing si ii 6355 line .
there is at least one example of a type ia sn which showed a very weak si ii 6355 line : the luminous , slow declining , and spectroscopically peculiar sn 1991 t ( filippenko et al . 1992 , phillips et al . 1992 ) .
it is important to note that the peculiar character of the spectrum of sn 1991 t , and the weakness of the si ii 6355 line , were more evident in the pre
maximum stage . shortly after maximum ,
the spectrum of sn 1991 t became more similar to the typical spectrum of a sn ia ( note that the spectrum of sn 1991 t plotted in fig [ fi : spec1 ] already shows the si ii 6355 line ) .
hence , if sn 1992ar were a 1991t like event , the spectrum and the photometry points taken shortly before and after it , should most likely correspond to a maximum or premaximum stage . to test
whether sn 1992ar could have been a spectroscopically peculiar type ia sn we compared the photometry of table [ ta : ccdobs ] with typical light curves of type ia sn of different speed classes ( hamuy et al .
we proceeded as follows .
we assumed a reasonable phase for the first point ( between @xmath76 and 12 days after maximum in v ) , computed the k correction for each observed phase , corrected the ccd photometry using these k corrections and fit the resulting data to one of the template sn ia curves given by hamuy et al .
( 1996b ) .
we then computed the rms of the fit , changed the phase by one day , and iterated the process until finding the phase that provided the minimal rms .
the results of this exercise for the light curves of sn 1992bc , a bright type ia sn with a slow decaying light curve similar to that of sn 1991 t , and sn 1992a , a fairly typical type ia sn , are displayed in figure [ fi : lcia1 ] .
usage of a fast decaying type ia light curve , typical of underluminous sne like sn 1991bg , did not provide a better match to the exponential tail ; since it is clear that sn 1992ar was not a low luminosity event , we do not show those fits .
the fits in figure [ fi : lcia1 ] show that , for reasonable values of the phase , the light curves of type ia sne provide a poor match to the observed points .
in particular , it is not possible to obtain a reasonable fit of either the decay after maximum or the exponential tail with the light curve of a slow decaying type ia sn ( top panel in figure [ fi : lcia1 ] ) .
the fit is particularly bad if a a negative phase for the spectrum , taken about one day after the first ccd point , is assumed .
the absolute magnitude at maximum with this fit would be m@xmath18 -19.08 , ( h@xmath14 km s@xmath4 , q@xmath15 ) , on the dim side for a slow declining type ia sn .
using the light curve of a faster declining type ia sn , like sn 1992a ( bottom panel in figure [ fi : lcia1 ] ) , the decay after maximum is reasonably well matched and the absolute magnitude at maximum results m@xmath18 -19.2 , similar to those of the comparable type ia sne in the caln / tololo survey .
however , the well observed exponential tail is overluminous by @xmath3 0.6 mag , and allows us to reject this fit . the only way to obtain a decent fit using the light curve and k corrections of a type ia sn is by releasing the requirement that the phase of the first ccd point is `` reasonable , '' and by assuming that it corresponds to @xmath317 days after v maximum ( see figure [ fi : lcia2 ] ) .
it is at this point that an estimate of the brightness at the time of discovery , based on the discovery plate , could be useful . using equation [ eq : bphot ] plus an estimate of the color of the sn at the time of discovery or the color evolution from the time of the spectrum and the color measured from the spectrum
, we obtained the v magnitudes given in table [ ta : mphot ] .
the color of the sn obtained in this second case is too blue for a sn ia 12 days after maximum in v. as we see in figure [ fi : lcia2 ] ) , however , both of the v estimates are consistent with the light curve , given the large uncertainties involved .
even when the interpretation of sn 1992ar as a typical type ia sn caught at late times can not be ruled out with the sole argument of its light curve , it is questionable on two grounds .
first , the maximum of sn 1992ar would have occurred around july 12 , 1992 , at v@xmath20 .
this implies m@xmath18 @xmath720.2 ( h@xmath14 km s@xmath4 mpc@xmath4 , q@xmath21 ) , more than 0.6 mag brighter than the brightest type ia sn in the caln / tololo survey .
second , the phase of the spectrum would be @xmath318 days after v maximum , which is inconsistent with the spectral features seen in fig .
[ fi : spec1 ] . using a more quantitative approach , spectroscopic dating of the spectrum as described by riess et al .
( 1997 ) , indicates that the phase of this spectrum ( if it were the spectrum a type ia sn ) was @xmath22 days after v maximum , in gross contradiction with the phase implied by this particular fit of the light curve .
these considerations are a strong indication that the photometric evolution and spectral age of sn 1992ar can not be reasonably matched by any known examples of type ia sn light curves and spectra . using the light curves and k corrections of slow and fast type ic sne
, we repeated the procedure of assuming a phase for the first photometric point , computing k corrections , fitting the light curve for a template , computing of the rms , and iterating on the phase until the rms was minimal .
as a model of the light curve of a fast type ic sn we took sn 1994i ( richmond et al .
1996 ) and as an example of a slow type ic sn we chose the light curve of sn 1993j ( richmond et al .
the latter sn was a transitional type ii , which showed strong he i lines starting @xmath320 days after explosion .
the second maximum of its light curve and early exponential decay are a good match to the light curves of the slow type ic sne 1983v ( clocchiatti et al .
1997 ) and 1990b ( clocchiatti et al . 1999 ) .
these kind of light curves are probably representative of a photometric group that includes sne of spectral types ib , ic and ii transition to ib .
the best fitting light curves and phases are displayed in figure [ fi : lcic ] .
it is clear from the figure , that both of the template light curves provide a reasonable match to the ccd photometry within the observational uncertainties and expected intrinsic scattering of each sn subclass , to the point that it is not possible to discard either .
the fitting of a fast ic light curve implies , like in the case of the ia templates , that maximum occurred several days before the first photometric data were obtained ( c.f .
[ fi : lcia2 ] ) . the absolute magnitude implied for
maximum would be m@xmath23 mag , again , more than 0.6 mag brighter than the brightest type ia sne in the caln / tololo survey . in this case
, however , the spectrum is reasonably similar to that of other type ic sne @xmath38 days after v maximum ( the phase implied by the fit ) and thus does not give an argument to discard it .
the fitting of a slow type ic sn is also good ( see bottom panel in fig . [
fi : lcic ] ) , and implies an absolute magnitude at maximum of m@xmath18 -19.26 mag , similar to the brightest sne in the caln / tololo survey .
it is clear from the figure that the goodness of the fit decreases as the phase increases .
this is , however , the expected behavior for stripped
envelope sne and is consistent with the usual picture of them originating in massive stars which have lost their outer layers .
a light curve faster on the tail implies a more rapid decay of the total optical depth and is related with a smaller mass to energy ratio in the ejecta ( see clocchiatti & wheeler , 1997 , for a simple model ) .
again , we can try to decide between this two different interpretation by resorting to brightness estimates based on the discovery plate . from the phases provided by the fits , the b@xmath7v color measured from the spectrum , and the assumption of a b@xmath7v color evolution similar to those of sn 1994i ( richmond et al . 1996 ) , or sn 1983v ( clocchiatti et al .
1997 ) , we can estimate the color of the sn at the time of the discovery plate for the cases of a fast and slow type ic sn , respectively .
another color estimate can be obtained under the assumption that it was similar to that of either sne _ at _ the given phase . with these color estimates and equation [ eq : bphot ]
, we can derive its b and hence v , magnitudes .
these estimates are given in table [ ta : mphot ] .
we have these v estimates for each sn type in fig .
[ fi : lcic ] .
as one can see , they are consistent with both the light curve of a slow type ic sn and that of a fast one , perhaps slightly better with the latter .
hence , surprisingly , the photographic plate with the discovery observation of sn 1992ar does not allow us to discriminate between the fast and slow sn ic templates .
even when the difference between the two interpretations at the time of discovery is about 0.8 mag , the different speed of color and k correction evolution combine in such a manner so as to make the plate consistent ( within the large uncertainties ) , with either hypothesis . to summarize , therefore
, we can only conclude that , if a phase consistent with that of the spectrum is assumed , the light curve of sn 1992ar is best fit by those of type ic sne .
we can not discern , however , whether it was a fast or a slow one .
the rest frame phases , magnitudes , and k corrections implied by the ic sn fits are given in table [ ta : icfits ] .
the rest frame b@xmath7v color on the date of the spectrum ( @xmath24 rest frame days after v maximum for a slow type ic sn , and @xmath25 rest frame days after v maximum for a fast type ic sn ) would have been @xmath30.62 mag , in both cases , in reasonable agreement with those of other type ic sne at these phases ( clocchiatti et al .
1997 , richmond et al .
1996 ) when corrections by foreground extinction ( to the other sne ) are considered .
sne of the types ib , ic , and ii - transition to ib ( hereafter referred to as _ transition type sne _ ) display a great degree of heterogeneity both in the appearance of the spectra and in the photometric behavior .
typically understood as stars originally massive which have lost their light element envelopes through episodes of stellar wind and/or mass transfer to interacting binary companions , this diversity has been taken as indicative of a large variety of possible configurations of the progenitor star .
even the three parameter space of two initial stellar masses and their separation , without consideration of the all the possibilities of the complex physics involved , provides many different branches of interacting stellar evolution , and has the potential of providing a variety of combinations of core and envelope at the time of core collapse .
different subsets of these combinations have been proposed to explain sne of the spectroscopic types ib , ic , ii@xmath7l , and ii@xmath7n .
a handful of these have been studied , with their theoretical light curves and spectra computed , and shown to match the observed light curves and spectra of nearby bright sne like 1983n , 1993j and 1994i ( woosley et al .
1994 ) ; woosley , langer & weaver 1995 ; nomoto et al .
all of these recent studies assume the same basic explosion mechanism : the gravitational collapse of an iron core that creates a shock , produces a last episode of stellar nucleosynthesis and disrupts the star , ejecting a relatively small amount of ( @xmath30.10 ) .
the extreme luminosity of sn 1992ar introduces yet another dimension of heterogeneity in this type of explosion .
we have collected in table [ ta : absmag ] , the absolute magnitudes of some of the best studied transition type sne .
although all of them were affected by foreground extinction in various amounts , and this implies a large uncertainty in some cases , it is clear that sn 1992ar s luminosity is unprecedented . from the spectrum displayed in fig .
[ fi : spec1 ] , the mass to energy ratio of sn 1992ar , as indicated by the square of the expansion velocity , does not appear to be terribly different from that of sn 1987 m or 1983v .
it is not fair to attach a quantitative estimate to the differences that may exist , however , since the spectra compared correspond to an early epoch .
mass to energy ratios are better represented by the expansion velocities when they reach the asymptotic behavior , typically at the beginning of the radioactive tail .
the slope of the light curve , on the other hand , under traditional interpretation of the deposition in expanding spherical shells is also related to the mass to energy ratio . according to the simple model of clocchiatti & wheeler ( 1997 ) , for the same mass density structure and effective opacity per unit mass , the difference in slopes some 100 days after maximum indicates that the mass to energy ratio of sn 1992ar was approximately 1.2 times smaller than that of sn 1993j .
we can also estimate the amount of powering the light curve of sn 1992ar .
modeling of the light curve of sn 1993j by different groups , at the distance given by recent calibration of cepheid variables in the parent galaxy ( m81 ) , suggests an ejected mass of radioactive from 0.073 ( woosley et al .
1994 ) up to 0.1 ( nomoto et al . 1997 ) .
assuming that both the percentage of energy from radioactive decays and the bolometric correction at maximum were similar in sn 1992ar and sn 1993j , the mass of ejected in the latter can be obtained using the absolute magnitudes in table [ ta : absmag ] .
the flux in v of sn 1992ar was about 4 times larger than that of sn 1993j , and thus the implied for sn 1992ar is between 0.3 and 0.4 . in the case of a fast type ic sn light curve fit , we can compare again with sn 1994i .
iwamoto et al .
( 1994 ) estimated an ejected mass of 0.07@xmath26 . according with table [ ta : absmag ] ,
the flux in v of sn 1992ar was approximately 7 times larger than that of sn 1994i so a scaling of iwamoto s estimate will put the mass of sn 1992ar at @xmath30.5 @xmath27 . both of these estimates indicate that the mass ejected by sn 1992ar should have been comparable to that produced by a fairly bright type ia sn ( arnett 1996 ) . a subset of transition type sne , including the slow type ic events , follows a similar photometric evolution even though they display a variety of spectra at maximum .
sn 1993j , which is a good example of the objects in this photometric group , is well understood in terms of a `` traditional '' core collapse explosion .
could one of these explosions produce as much as required by the slow type ic sn fit to sn 1992ar ? or , alternatively , could a core collapse explosion in a @xmath32 c@xmath28o star like the one used by iwamoto et al .
( 1994 ) to match the light curve of sn 1994i , produce half a solar mass of ?
the other option is to conclude that there are some stripped envelope sne which , contrary to conventional wisdom , do explode via a thermonuclear explosions ( branch , nomoto & filippenko 1991 ) .
whether a thermonuclear explosion could give a ic spectrum together with the fast decline after maximum and slow decline on the tail of a slow type ic sn , _ or _ the fast light curve of a fast type ic sn , is a question that merits further study .
the existence of luminous type ic sne such as sn 1992ar has important implications for the the high redshift sne searches currently under way with the aim of measuring the cosmological parameters of the universe ( perlmutter et al .
1998 , schmidt et al . 1998 ) because it exemplifies the problem of sample contamination .
sn 1992ar shows that there are sne which are as bright as a type ia sn and have a similar light curve near maximum light , but have different intrinsic colors , different color evolution , and do not follow the light curve shape luminosity relation used to calibrate the intrinsic luminosity of a type ia sn .
one of the earliest and clearest indications that sn 1992ar was a peculiar event was given by its spectrum , which did not display the siii 6355 line .
if sn 1992ar had been at a redshift larger than @xmath30.50 , however , this line would have been doppler shifted to @xmath39200 , and the brightness at maximum would have been around r@xmath323 mag .
this , together with the poorer sensitivity and typical fringing of most spectrographs in the far red , would have made it very difficult for the observers to distinguish on spectroscopic grounds such a sn from a typical sn ia ( see riess et al .
1998 , for an example of this problem ) .
the light curve ( which almost surely would have had worse signal - to - noise ratio and shorter time baseline than those of sn 1992ar ) would have probably been acceptably fitted with a typical type ia like curve , using parameters similar to those of the bottom panel in fig .
[ fi : lcia1 ] .
note that this fit implies a phase of @xmath3 7 days after v maximum for the spectrum , in good agreement with the spectroscopic age derived by the method of riess et al .
( 1997 ) for the spectrum of sn 1992ar . using the k corrections for type ia sn ,
the rest frame color at the time of the spectrum implied by this fit ( @xmath37 days after v maximum ) is b@xmath7v= 0.6 mag , somewhat red in comparison with a typical ia sn @xmath29 days after b maximum . this red color , however ,
would have been interpreted as indicative of foreground extinction , and the intrinsic brightness of sn 1992ar would have been increased accordingly . in the end
, the luminosity distance derived from this object ( if it were assumed to be a type ia sn ) would have been too short .
the possibility exists , therefore , depending on the signal - to - noise ratio of the observational sets obtained for a particular object , that the samples of sne with [email protected] can be contaminated by bright ic sne like 1992ar
. with current detectors and telescopes , which make detection of the siii 6355 line difficult for sne with [email protected] , the most reliable means of eliminating bright type ic sne from the high z sne samples sne will be : ( 1 ) to obtain high s / n ratio spectra in the visual and red regions in order to detect the typically much weaker siii lines at 4130 and 5051
( note that the line siii 5972 could appear merged with nai d , so its diagnostic power is limited ) , and the sii lines at 5468 , 5612 and 5654 , which have never been claimed to be present in the spectrum of type ic sne , _ and _ , ( 2 ) to obtain well time sampled two color light curves with good signal - to - noise ratio including at least a pre maximum point and , especially , a couple of good points , with photometric uncertainty smaller than @xmath30.05 mag on the early exponential tail , between 40 and 60 days after maximum light .
a good s / n ratio data set for a high z sn would allow to apply the following criteria . if a sn has a red color , does not show evidence for siii 4130 , siii 5051 , sii 5468 , sii 5612 , and sii 5654 lines , displays a maximum noticeably narrower than that of a type ia sn , _ or _ a light curve which remains brighter than the type ia sne light curve templates on the early exponential tail , it should be suspect of being a type ic sn and excluded from the samples .
how could these luminous type ic sne affect the high z hubble diagrams if not removed from the samples ? according to the current , scarce , database , and opposite to what type ia sne display , the shapes of light curves of transition type sne do not seem to correlate with their intrinsic brightness . as we have shown , the same type of light curves fit sne with very different intrinsic brightness .
in the cases of a slow type ic sn fit for sn 1992ar this means m@xmath31 ( sn 1983n ) , and m@xmath32 ( sn 1992ar ) , while in the case of a fast type ic sn fit for sn 1992ar it means m@xmath33 ( sn 1994i ) , and m@xmath34 ( sn 1992ar ) .
it could well be that the process which determines the amount of ejected in core
collapse explosions is intrinsically chaotic and bears no relation to the mass to energy ratio of the explosions , responsible for the light curve shapes .
it would appear , in principle , that the main effect of not removing objects like sn 1992ar from the high z type ia sne samples will be an increase in the scatter of the final hubble diagrams .
on the other hand , systematic shifts may enter through malmquist bias .
if most of the type ic which go unchecked and are included in the samples are as bright as sn 1992ar ( i.e. slightly brighter than a typical type ia sn ) , then the sample of distant sne will be biased in the sense of appearing brighter than they should .
the result of this will be that the distances estimated will be shorter than they are , and consistent with a positive deceleration parameter q@xmath17 , or alternatively , with higher values of @xmath35 and lower values of @xmath36 .
support for ac , at different stages of this project , was provided by fundacin antorchas argentina under project a-13313 , by p. universidad catlica de chile under dipuc project 97/12e , and by fondecyt , chile , under project 1980803 .
nbs and mmp acknowledge the support from nasa through grants go-2563.01 - 87a and go-6020 from stsci .
while this paper was in the final stages of writing and submission the first results on sn 1998bw became available ( saddler et al .
1998 , kulkarni et al . , 1998 ,
galama et al .
although a detailed comparison of sne 1998bw and 1992ar is beyond the scope of the present paper , a note cross linking them is necessary .
sn 1998bw was discovered well within the error box of grb 980425 and there is compelling evidence that it was associated with the burst .
sn 1998bw displayed peculiarities in its light curve and spectra , with a very fast rise to maximum ( woosley , eastman & schmidt 1999 ) , and photospheric expansion velocities of @xmath328,000 km s@xmath4 ten days after explosion ( iwamoto et al .
sn 1998bw was also recognized as a very bright event . at an absolute magnitude of m@xmath37 ( galama et al .
1998 ) , however , it could have been dimmer than sn 1992ar . if the two events correspond to core collapse explosions , they show very clearly that this mechanism does not result in a correlation between mass ejection and mass to energy ratio in the ejecta .
35.93 & 20.11 @xmath5 0.04 & & & ctio 0.91 m + ccd r. avils / c .
smith + 36.78 & & 21.05 @xmath5 0.20 & & b@xmath7v=0.85 ( spectrum ) and v=20.20 .
+ 40.80 & & 21.47 @xmath5 0.19 & & ctio 0.91 m + ccd e.p .
rubenstein + 40.81 & 20.54 @xmath5 0.06 & & & ctio 0.91 m + ccd e.p .
rubenstein + 47.71 & 21.22 @xmath5 0.12 & & & ctio 0.91 m + ccd r. avils + 67.55 & 21.91 @xmath5 0.05 & & 21.14 @xmath5 0.07 & eso 3.6 m + efosc1 e. cappellaro + 98.78 & & & 22.13 @xmath5 0.07 & eso / mpi 2.2 m + efosc2 m. della valle + 98.79 & 22.31 @xmath5 0.07 & & & eso / mpi 2.2 m + efosc2 m. della valle + 99.65 & 22.30 @xmath5 0.08 & & & eso / ntt + emmi m. della valle + 156.54 & 23.22 @xmath5 0.15 & & & eso 3.6 m + efosc1 m. della valle + 1993 , sep 9 - 6:35 & 409 & b & ctio 4.0 m
cass ccd camera & m. navarrete + 1993 , nov 11 - 3:53 & 472 & v & ctio 4.0 m cass ccd camera & r. avils + 1996 , dec 15 - 2:21 & 1602 & r & eso 2.2 m
efosc 2 & c. zanin i d & v & b & b@xmath7v & r + 92ar-1 & 16.15 @xmath5 0.02 & 16.79 @xmath5 0.02 & 0.65 @xmath5 0.03 & 15.67 @xmath5 0.04 + 92ar-2 & 18.48 @xmath5 0.02 & 19.37 @xmath5 0.03 & 0.89 @xmath5 0.03 & 17.83 @xmath5 0.05 + 92ar-3 & 16.72 @xmath5 0.02 & 17.85 @xmath5 0.03 & 1.14 @xmath5 0.03 & 15.87 @xmath5 0.07 + 92ar-4 & 17.61 @xmath5 0.02 & 19.05 @xmath5 0.03 & 1.44 @xmath5 0.03 & 16.42 @xmath5 0.09 + 92ar-5 & 17.53 @xmath5 0.02 & 18.81 @xmath5 0.03 & 1.28 @xmath5 0.03 & + 92ar-6 & 17.05 @xmath5 0.02 & 17.71 @xmath5 0.03 & 0.67 @xmath5 0.03 & 16.57 @xmath5 0.04 + 92ar-7 & 17.99 @xmath5 0.02 & 18.61 @xmath5 0.03 & 0.62 @xmath5 0.03 & + 92ar-8 & 20.73 @xmath5 0.05 & 21.60 @xmath5 0.12 & 0.87 @xmath5 0.13 & 19.95 @xmath5 0.06 + 92ar-9 & 18.31 @xmath5 0.02 & 19.70 @xmath5 0.03 & 1.39 @xmath5 0.03 & + 92ar-11 & 20.04 @xmath5 0.03 & 21.41 @xmath5 0.09 & 1.38 @xmath5 0.10 & 18.70 @xmath5 0.08 + 92ar-12 & 19.97 @xmath5 0.03 & 20.23 @xmath5 0.05 & 0.26 @xmath5 0.06 & 20.03 @xmath5 0.03 + 92ar-13 & 20.22 @xmath5 0.03 & 21.66 @xmath5 0.11 & 1.44 @xmath5 0.11 & 18.97 @xmath5 0.09 | we present spectroscopic and photometric observations of sn 1992ar , the more distant sn in the caln / tololo survey .
we compare its spectrum with those of nearby type ia and ic sne and conclude that the latter type is a better match to sn 1992ar . using k
corrections based on the spectra of well observed type ic and ia sne we compute different possible rest frame light curves of sn 1992ar and compare them with those of representative sne of each type observed in the nearby universe . from the photometry and the spectra , we are able to conclude that sn 1992ar can not be matched by any known example of a type ia sn .
even though the data set collected is fairly complete ( one spectrum and 10 photometric points ) , it is not possible to decide whether sn 1992ar was a fast type ic sn , like sn 1994i , or a slow one , like sn 1983v .
the absolute v magnitudes at maximum implied by each of these possibilities are @xmath0 and @xmath1 , respectively .
the latter would make sn 1992ar one of the brightest sne on record .
sn 1992ar , hence , illustrates the problem of contamination faced by the high z type ia sne samples whose luminosity distances are used to determine the cosmological parameters of the universe .
we present observational criteria to distinguish the two sn types when the siii@xmath26355 line is redshifted out of the sensitivity range of typical ccd detectors , and discuss the effect that these luminous type ic sne would have on the measured cosmological parameters , if not removed from the high z type ia sn samples .
@xmath2 submitted to _
astrophysical journal _ |
let @xmath2 and consider the @xmath1 hypergeometric differential equation @xmath3 where @xmath4 and @xmath5 .
assuming , as we do , that @xmath6 , @xmath7 , and the @xmath8 s and @xmath9 s are distinct , the @xmath10-functions @xmath11 where @xmath12 denotes omit @xmath13 , are linearly independent solutions to ( [ deq ] ) . here
@xmath1 is the hypergeometric function @xmath14 and @xmath15
. equation ( [ deq ] ) is regular away from @xmath16 and its monodromy group @xmath17 is generated by the local monodromies @xmath18 ( @xmath19 ) gotten by analytic continuation of a basis of solutions along loops about @xmath20 , and @xmath21 respectively , see @xcite for a detailed description .
the local monodromies of equations that come from geometry are quasi - unipotent which is one reason for our restricting @xmath8 and @xmath9 to be rational .
we restrict further to such @xmath17 s which after a suitable conjugation are contained in @xmath22 . according to @xcite ,
this happens if the characteristic polynomials of @xmath23 and @xmath24 , whose roots are @xmath25 and @xmath26 respectively , are products of cyclotomic polynomials .
in particular for each @xmath27 there are only finitely many such choices for the pair @xmath28 in @xmath29 .
@xcite also determine the zariski closure @xmath30 of @xmath17 explicitly in terms of @xmath31 .
furthermore the integrality conditions that we are imposing imply that @xmath17 is self dual so that @xmath32 is either finite , @xmath33 ( @xmath10 even ) or @xmath34 .
the signature of the quadratic form in the orthogonal case is determined by the relative locations of the roots @xmath28 ( see section [ beuksum ] ) .
our interest is whether @xmath17 is of finite or infinite index in @xmath35 $ ] . in the first case
we say that @xmath17 is _ arithmetic _ and in the second case that it is _
thin_. this distinction is important in various associated number theoretic problems ( see @xcite ) and this paper is concerned with understanding which case happens and which is typical . in a given example , if @xmath17 is arithmetic one can usually verify that it is so by producing generators of a finite index subgroup of @xmath36 , on the other hand if @xmath17 is thin then there is no general procedure to show that it is so .
our main result is a robust certificate for showing that certain @xmath17 s are thin . until recently , other than the cases where @xmath17 ( or equivalently @xmath32 ) is finite , there were few cases for which @xmath17 itself was known .
for @xmath37 it is well known that all the @xmath17 s are arithmetic and we show that the same is true for @xmath38 . for @xmath39 brav and
thomas @xcite showed very recently that the dwork family @xcite @xmath40 as well as six other hypergeometrics with @xmath41 which correspond to families of calabi - yau three - folds , are thin .
in fact they show that the generators @xmath23 and @xmath42 of the above @xmath17 s play generalized ping - pong on certain subsets of @xmath43 , from which they deduce that @xmath17 is a free product and hence by standard cohomological arguments that @xmath17 is thin . on the other hand ,
venkataramana shows in @xcite that for @xmath10 even and @xmath44 @xmath17 is arithmetic ( in @xmath45 ) . in particular , there are infinitely many arithmetic @xmath17 s . in @xcite
many more examples with @xmath46 and for which @xmath17 is arithmetic are given .
another example for which @xmath47 can be shown to be thin is @xmath48 , @xmath49 , see @xcite . in this case
@xmath50 is orthogonal and has signature @xmath51 and @xmath36 splits as a product of @xmath52 s .
all of our results are concerned with the case that @xmath32 is orthogonal and is of signature @xmath0 over @xmath53 .
we call these hyperbolic hypergeometric monodromy groups
. there is a unique ( up to a scalar multiple ) integral quadratic form @xmath54 for which @xmath55 , or what is the same thing an integral quadratic lattice @xmath56 with @xmath57 . in section [ quadform ]
we determine a commensurable quadratic sublattice explicitly which facilitates many further calculations . in this hyperbolic setting
@xmath58 acts naturally as isometries of hyperbolic @xmath59-space @xmath60 and we will use this geometry as a critical ingredient to provide a certificate for @xmath17 being thin .
our first result is the determination of the @xmath28 s for which @xmath32 is hyperbolic , see theorem [ theo ] .
firstly , these only occur if @xmath10 is odd and for @xmath61 they are completely described by seven infinite parametric families . for @xmath62
there are sporadic examples which are listed in tables 2 and 3 of section [ numeric ] .
our determination of the seven families is based on a reduction to @xcite s list of families of @xmath32 s which are finite ( i.e. those @xmath32 s for which @xmath63 and have signature ( @xmath64 ) ) . for @xmath38 ,
if @xmath17 is not finite then it is hyperbolic and as we noted all @xmath65 of these hyperbolic groups are arithmetic .
this is verified separately for each case , there being no difficulty in deciding whether a finitely generated subgroup of @xmath66 is thin or not ( the latter is a double cover of @xmath67 , see the appendix .
for @xmath68 the hyperbolic monodromies behave differently .
our certificate of thinness applies in these cases and it is quite robust as exemplified by [ theorem1 ] the two families of hyperbolic monodromies @xmath17 with @xmath68 and odd * @xmath69 * @xmath70 are thin .
in particular infinitely many of the @xmath17 s are thin and as far as we know these give the first examples in the general monodromy group setting of thin monodromy groups for which @xmath71 is high dimensional and simple .
the normalized @xmath1 s corresponding to ( i ) and ( ii ) above ( see @xcite for the normalization ) are @xmath72 respectively .
the second has integral coefficients while the first does not , hence this arithmetic feature of @xmath73 is not reflected in the arithmeticity of the corresponding @xmath47 ( see the end of section [ familyclassification ] for more about the integrality of the coefficients ) .
our certificate for thinness applies to a number of the families and many of the sporadic examples .
the full lists that we can handle thus far are recorded in theorems [ thinfamilythm ] and [ secondthin ] of section [ mindistgraph ] as well as table [ table3 ] of section [ numeric ] .
there remain some families for which the method does not apply ( at least not directly ) . in any case
we are led to [ allthinconj ] all but finitely many hyperbolic hypergeometric @xmath17 s are thin .
we turn to a brief description of our methods . in general
if there is an epimorphism @xmath74 with @xmath75 of infinite index in @xmath76 , then clearly @xmath17 must be of infinite index in @xmath36 . in principle
, @xmath77 could be infinite which would imply that @xmath17 is thin since the latter is essentially generated by involutions , however we know very little about these cohomology groups for the full @xmath78 .
instead we use variations of the quotient @xmath79 where @xmath80 is the vinberg reflective subgroup of @xmath78 .
namely , @xmath80 is the group generated by the elements of @xmath78 which induce reflections of hyperbolic space @xmath60 ( that is the induced action on one of the two sheeted hyperboloids in @xmath10-space , see section [ cartaninvol ] ) .
vinberg @xcite and nikulin @xcite have shown that except for rare cases @xmath76 is infinite , see section [ cartaninvol ] for a review of their results which we use . to apply this
we need to understand the image of @xmath17 in @xmath76 .
the key observation is that the element @xmath42 in @xmath17 , which is a linear reflection of @xmath10-space , induces a cartan involution of @xmath60 , that is it is an isometry of @xmath60 which is an inversion in a point @xmath81 .
the reflection subgroup @xmath82 of @xmath17 is the group generated by the cartan involutions @xmath83 , @xmath84 and @xmath85 is cyclic .
thus essentially up to commensurability the question of whether @xmath17 is thin is a special case of deciding whether a subgroup @xmath86 of @xmath78 generated by cartan involutions is thin or not .
we approach this by examining the image of such a @xmath86 in @xmath87 ( when the latter is infinite ) . at this point
the study is about @xmath78 , @xmath87 , and a general such @xmath86 ( and @xmath10 need nt be odd ) .
we define a graph @xmath88 , the `` distance graph , '' associated with root vectors of @xmath54 which is central to the analysis .
we assume that @xmath54 is even ( that is @xmath89 for @xmath90 ) and for @xmath91 or @xmath92 let @xmath93 be the corresponding root vectors ( in the case that @xmath94 for all @xmath95 which comes up in some cases we also allow @xmath96 to be @xmath97 ) .
the root vectors define linear reflections lying in @xmath78 given by @xmath98 . by our choice of the signature of @xmath54 , for @xmath99 the map @xmath100 induces a cartan involution on @xmath60 while for @xmath101 it induces a hyperbolic reflection on @xmath60 .
assume that the cartan involutions generating @xmath86 come from root vectors in @xmath102 .
@xmath88 has for its vertices the set @xmath102 and we join @xmath103 to @xmath104 if @xmath105 , this corresponds to @xmath103 and @xmath104 having the smallest distance allowed by discreteness , as points in @xmath60 .
the graph @xmath88 is a disjoint union of its connected components @xmath106 and these satisfy ( see section [ graphprop ] for a detailed statement ) * the components @xmath107 consist of finitely many isomorphism types . * [ [ ( ii ) ] each type is either a singleton or an infinite homogeneous graph corresponding to a transitive isometric action of a coxeter group with finite stabilizer . *
if @xmath107 is a connected component of @xmath88 , then the group generated by the corresponding cartan roots , @xmath108 is up to index @xmath109 contained in the reflection group @xmath110 . with this the certificate for showing that @xmath86 is thin is clear : according to vinberg and nikulin
, @xmath111 is infinite except in rare cases ( and nikulin has a classification of thee ) , hence if ( iii ) is satisfied for the generators of @xmath86 then @xmath86 must be thin .
the calculations connected with the minimal distance graph can be carried out effectively in general and even explicitly for some of our families .
the process in the general case invokes an algorithm for computing the fundamental polyhedral cell for a discrete group of motions of hyperbolic space which is generated by a finite number of reflections .
such an algorithm is provided in section [ thinguys ] .
the above leads to the cases discussed in sections [ mindistgraph ] and [ numeric ] for which @xmath17 is shown to be thin . to end
we note that it is possible that a much stronger version of vinberg s theorem in the following form is valid : for all but finituely many rational quadratic forms @xmath54 ( at least if @xmath10 is large enough ) the full `` weyl subgroup '' @xmath112 generated by _ all _ reflections in @xmath113 ( that is , both those inducing hyperbolic reflections and cartan involutions on @xmath60 ) is infinite index in @xmath114 ( see nikulin @xcite ) .
if this is true then conjecture [ allthinconj ] would follow easily from our discussion . in section [ cartaninvol ]
we give an example in dimension @xmath115 with @xmath116 being thin and @xmath117 arithmetic , so there is no general commensurability between these groups .
finally , we note that when our analysis of the thinness of @xmath47 succeeds , it comes with a description of @xmath47 as a subgroup ( up to commensurability ) of a geometrically finite subgroup of @xmath116 and this opens the door to determine the group structure of @xmath47 itself .
we leave this for the future .
we begin by reviewing @xcite which forms the basis of our analysis .
the setup and notation is as in the introduction .
the starting point for studying @xmath17 is the following theorem of levelt .
[ levthm ] for @xmath118 let @xmath119 and @xmath120 be as above .
define the complex numbers @xmath121 to be the coefficients of the polynomials @xmath122 then @xmath17 is the group generated by @xmath123 note that @xmath17 can be conjugated into @xmath22 if and only if the polynomials @xmath124 and @xmath125 above factor as cyclotomic polynomials , and so the roots of @xmath124 and @xmath125 are reciprocal , meaning they are left invariant under the map @xmath126 . since we are interested in integral monodromy groups , we will assume this throughout the article .
also , given a group @xmath127 as above , one obtains another hypergeometric group @xmath128 by taking @xmath129 to be @xmath130 and @xmath131 modulo @xmath21 , respectively .
the group @xmath128 is called a _ scalar shift _ of @xmath17 , and as pointed out in remark 5.6 in @xcite , we have that @xmath132 , up to a finite center . thus when classifying all possible pairs @xmath133 such that @xmath17 is integral and fixes a quadratic form of signature @xmath0 in section [ familyclassification ] we will do so up to scalar shift .
note that for any given @xmath10 there are only finitely many such groups @xmath17 .
we also assume that @xmath47 is _ irreducible _ , meaning that it fixes no proper subspace of @xmath134 , and _ primitive _ , meaning that there is no direct sum decomposition @xmath135 with @xmath136 and @xmath137 for all @xmath138 such that @xmath47 simply permutes the spaces @xmath139 .
finally , we define the reflection subgroup @xmath140 of @xmath47 to be the group generated by the reflections @xmath141 . by theorem 5.3 of @xcite
we have that the primitivity of @xmath47 implies the irreducibility of @xmath140 . with the notation above , beukers - heckman show in @xcite
that @xmath142 falls into one of three categories . given a hypergeometric monodromy group @xmath17 ,
let @xmath143 where @xmath144 and @xmath145 are as in theorem [ levthm ] .
note that in the cases we consider @xmath146 .
then @xmath17 belongs to one of the following categories . * a finite group ( beukers - heckman list such cases completely in @xcite , and we summarize these cases in theorem [ finiteh ] below ) * if @xmath10 is even , @xmath47 is infinite , and @xmath147 then @xmath148 and @xmath149 * if @xmath10 is odd and @xmath47 is infinite ; or if @xmath10 is even , @xmath47 is infinite , and @xmath150 , then @xmath151 for some rational , unique up to scalar multiple quadratic form @xmath152 in @xmath10 variables and @xmath153 as noted in the introduction , in this article we study hypergeometric groups @xmath47 which fall into category ( 2 ) and such that @xmath47 fixes a quadratic form of signature @xmath0 . we should mention that beukers - heckman show that the signature of @xmath154 in category ( 2 ) is given by @xmath155 where @xmath156 and @xmath157 and the @xmath119 and @xmath120 are ordered as described in the introduction : @xmath6 , @xmath7 .
although we concern ourselves with the case where @xmath47 is infinite , irreducible , and primitive , we state below beukers - heckman s classification of _ all _ possible finite groups @xmath17 . in section [ familyclassification ]
we will need this classification to classify all of the primitive , irreducible hypergeometric groups @xmath17 which we consider
i.e. integral orthogonal of signature @xmath0
. the precise statement of the theorem is taken from @xcite .
[ finiteh ] let @xmath158 and @xmath159 and let @xmath17 be as before .
define @xmath160 and @xmath161 and let @xmath162 and @xmath163 .
let @xmath164 corresponding to @xmath17 be as in theorem [ levthm ] .
if @xmath17 is finite then the corresponding polynomials @xmath164 are as in one of the cases below .
\(1 ) the case where @xmath140 is primitive : in this case the corresponding polynomials @xmath124 and @xmath125 belong to one of two infinite families or to one of @xmath165 sporadic examples .
one such infinite family corresponds to @xmath166 where @xmath167 , @xmath168 , @xmath169 .
the other infinite family is obtained by replacing @xmath170 with @xmath171 above .
\(2 ) the case where @xmath140 acts reducibly on @xmath134 : in this case theorem 5.3 of @xcite implies that @xmath140 is primitive , and we have that there is some primitive @xmath172th root of unity @xmath173 with @xmath174 such that @xmath175 .
then @xmath133 gives the pair @xmath176 where @xmath177 and @xmath178 are as in case ( 1 ) or correspond to one of the sporadic examples mentioned there .
\(3 ) the case where @xmath47 is imprimitive and @xmath140 is irreducible : in this case the corresponding polynomials @xmath124 and @xmath125 belong to one of two infinite families .
one such family is @xmath179 where @xmath180 , @xmath181 , and @xmath182 .
the other such family is obtained by replacing @xmath170 with @xmath171 above . in this section
we classify all almost interlacing cyclotomic sequences ( see definition [ almostinterlacing ] ) .
this classification will be used in the next section to classify all hyperbolic hypergeometric monodromies in dimension @xmath61 .
two sequences @xmath183 and @xmath184 are called interlacing cyclotomic sequences if the following three conditions hold : * @xmath185 for every @xmath186 . *
@xmath187 and @xmath188 are products of cyclotomic polynomials . *
@xmath189 or @xmath190 .
define @xmath191 and @xmath192 .
condition 3 in the above definition is equivalent to the following condition : * @xmath193 .
[ almostinterlacing ] two sequences @xmath194 and @xmath195 are called almost interlacing cyclotomic sequences if the following three conditions hold : * @xmath185 for every @xmath186 . *
@xmath187 and @xmath196 are products of cyclotomic polynomials .
* @xmath197 .
note that , by the expression for the signature in ( [ sig ] ) , it is precisely when @xmath8 and @xmath9 are almost interlacing that @xmath17 is hyperbolic ( signature @xmath0 ) .
beukers and hekman classify all interlacing cyclotomic sequences in @xcite .
we use their classification in order to classify all almost interlacing cyclotomic sequences .
we start with the following lemma : [ interlacingtypes ] let @xmath198 and @xmath199 be almost interlacing cyclotomic sequences with @xmath200 . then @xmath10 is odd and @xmath201 and @xmath202 or @xmath203 and @xmath204 .
moreover , if the first possibility happens then one of the following 4 options holds : * the sequences @xmath205 and @xmath206 are interlacing cyclotomic sequences where : @xmath207 * the sequences @xmath208 and @xmath209 are interlacing cyclotomic sequences where : @xmath210 * the sequences @xmath208 and @xmath209 are interlacing cyclotomic sequences where : @xmath211 * the sequences @xmath212 and @xmath213 are interlacing cyclotomic sequences where : @xmath214 the proof is divided into easy steps : * the sequences @xmath198 and @xmath199 are cyclotomic so for every @xmath215 , @xmath216 and @xmath217 * for every @xmath218 denote @xmath219 and @xmath220 .
the equality @xmath221 implies that exactly one of the following possibilities holds : * * @xmath222 * * @xmath223 * * @xmath224 * * @xmath225 * * @xmath226 for some @xmath227 . *
the symmetry of step @xmath228 shows that options ( 3 ) and ( 4 ) are not possible and that : * * if option ( 1 ) happens then @xmath229 . * * if option ( 2 ) happens then @xmath230 .
* * if option ( 3 ) happens then @xmath10 is odd , @xmath231 and @xmath232 .
* denote @xmath233 .
changing the roles of the @xmath119 s and the @xmath234 s we see that @xmath10 is odd and that either @xmath235 and @xmath202 or @xmath236 and @xmath237 .
* from now on we will assume that @xmath235 so @xmath238 and @xmath202 .
* since there are exactly @xmath10 @xmath234
s the assumptions of step ( e ) imply : @xmath239 furthermore , the symmetry of step ( a ) implies that if @xmath240 then also @xmath241 while if @xmath242 then @xmath243 .
* the symmetry of step ( a ) implies that either @xmath244 or @xmath245 .
thus , there are only 4 options : * * @xmath246 @xmath247 * * @xmath248 @xmath249 * * @xmath250 @xmath251 * * @xmath252 @xmath253 in this section we combine lemma [ interlacingtypes ] with the classification of finite hypergeometric monodromy groups in theorem [ finiteh ] to produce seven infinite families of primitive integer hypergeometric groups which fix a quadratic form of signature @xmath0 .
each family is either a two or three parameter family .
as will be clear in the proof of the following theorem , the @xmath10-dimensional two parameter families listed below are derived from the roots of @xmath254 where @xmath255 is taken to be @xmath256 or @xmath10 , while the @xmath10-dimensional three parameter families are derived from the roots of @xmath257 where @xmath258 and @xmath255 is taken to be @xmath256 or @xmath10 .
these families describe all primitive hypergeometric hyperbolic monodromy groups in dimension @xmath61 up to scalar shift .
in addition to groups in these families , there are several sporadic groups in dimensions @xmath259 which are listed in table [ table3 ] of section [ sporadic ] along with all primitive hyperbolic @xmath17 in dimension @xmath259 .
[ theo ] let @xmath167 be odd and let @xmath260 be as defined in ( [ polystate0 ] ) and ( [ polystate ] ) .
let @xmath261 and @xmath262 where @xmath263 for @xmath118 and let @xmath17 be as in theorem [ levthm ] . if @xmath133 belongs to one of the following families or is a scalar shift of a pair in one of these families , then @xmath264 . if @xmath61 , this list of families completely describes ( up to scalar shift ) the groups @xmath265 where @xmath54 is a quadratic form in @xmath10 variables of signature @xmath0 .
* @xmath266 : @xmath267 where @xmath268 is an odd integer .
* @xmath269 : @xmath270 where @xmath271 is an integer and @xmath272 is odd . *
@xmath273 : @xmath274 where @xmath275 is an odd integer . *
@xmath276 : @xmath277 where @xmath278 , @xmath279 , @xmath280 , @xmath281 are the roots of @xmath282 , and @xmath283 are the roots of @xmath284 . * @xmath285 : @xmath286 where @xmath278 , @xmath280 , @xmath287 , @xmath288 are the roots of @xmath289 , and @xmath290 are the roots of @xmath291 . * @xmath292 : @xmath293 where @xmath278 , @xmath280 , @xmath287 , @xmath288 are the roots of @xmath289 , and @xmath290 are the roots of @xmath291 .
* @xmath294 : @xmath295 where @xmath278 , @xmath296 , @xmath297 are the roots of @xmath298 and @xmath299 are the roots of @xmath300 . from the previous section , we have that every family of pairs @xmath133 for which @xmath301 can be derived from a family of pairs @xmath302 for which @xmath128 is finite , but not necessarily primitive or irreducible .
there are several cases to consider .
* case 1 : * @xmath303 . without loss of generality ,
assume @xmath304 for some @xmath96 and @xmath305 . by lemma [ interlacingtypes ] ,
letting @xmath306 @xmath307 @xmath308 @xmath309 we have that @xmath301 .
suppose @xmath128 is finite , imprimitive , and irreducible .
from theorem [ finiteh ] there are two infinite families of @xmath302 , the second of which gives pairs that are scalar shifts by @xmath310 of the pairs coming from the first family .
since we are only interested in determining pairs up to scalar shift , we consider just one of these families , which has corresponding cyclotomic polynomials @xmath311 and @xmath312 from ( [ polystate0 ] ) where @xmath313 is odd .
up to scalar shift , this yields the family @xmath314 : @xmath315 where @xmath313 is odd .
note that , unlike @xmath302 , the pair @xmath133 is no longer imprimitive , and still irreducible .
suppose @xmath128 is finite and reducible or primitive and irreducible .
then for every @xmath316 , theorem [ finiteh ] gives two infinite families of @xmath302 , of which we consider just the first , as the other can be obtained via scalar shift of the first .
the corresponding cyclotomic polynomials in this case are @xmath282 and @xmath284 from ( [ polystate ] ) , where @xmath317 , with @xmath278 and @xmath318 .
let @xmath279 and denote the roots of @xmath282 by @xmath281 .
let @xmath280 and denote the roots of @xmath284 by @xmath283 .
up to scalar shift , this yields the family @xmath319 : @xmath320 note that the pair @xmath133 is primitive and irreducible .
* case 2 : * @xmath321 . as before , we consider the imprimitive irreducible , and the primitive irreducible or reducible cases .
suppose @xmath128 is finite , imprimitive , and irreducible . from theorem [ finiteh
] , we have that up to scalar shift the only family of @xmath302 in this case corresponds to the roots of @xmath322 and @xmath323 in ( [ polystate0 ] ) : @xmath324 where @xmath272 is odd . according to lemma [ interlacingtypes ] , there are two ways to obtain from @xmath302 a pair @xmath133 for which @xmath17 is signature @xmath0 is obtained as above , giving the two families @xmath325 where @xmath326 and @xmath327 where @xmath272 is odd .
however , since these two families are shifts of each other by @xmath310 , we record them under one family @xmath328 .
suppose @xmath128 is finite and reducible or primitive irreducible . from theorem [ finiteh
] , we have that for every @xmath329 that there is only one family up to scalar shift of @xmath302 for which @xmath128 is reducible .
namely , let @xmath317 , with @xmath278 and @xmath330 .
denote the roots of @xmath289 in ( [ polystate ] ) by @xmath288 .
let @xmath280 , @xmath287 , and denote the roots of @xmath291 in ( [ polystate ] ) by @xmath290 .
then the family @xmath331 is the only family up to scalar shift in this case such that @xmath128 is reducible . according to lemma [ interlacingtypes ] , there are two ways to obtain a pair @xmath133 for which @xmath17 is signature @xmath0 .
one way is to remove the term @xmath280 and add two @xmath310 s to @xmath332 , and add the term @xmath333 to @xmath334 , yielding the family @xmath335 : @xmath336 another way is to add the term @xmath310 to @xmath334 , take away the term @xmath287 from @xmath332 , and insert two @xmath337 into @xmath332 , yielding the family @xmath338 : @xmath339 * case 3 : * @xmath340 .
suppose @xmath128 is finite , imprimitive , and irreducible . from theorem [ finiteh
] , we have that up to scalar shift the only family of @xmath302 in this case is obtained from the roots of the cyclotomic polynomials @xmath341 and @xmath342 in ( [ polystate0 ] ) : @xmath343 where @xmath313 is odd .
the corresponding @xmath133 for which @xmath17 is signature @xmath0 is obtained by adding two @xmath310 s to @xmath334 , and adding two @xmath333 s to @xmath332 , yielding the family @xmath344 : @xmath345 where @xmath313 is odd .
suppose @xmath128 is finite and reducible or primitive irreducible . from theorem [ finiteh
] , we have that for every @xmath346 that there is only one family up to scalar shift of @xmath302 for which @xmath128 is reducible .
namely , let @xmath317 , with @xmath278 and @xmath347 .
denote the roots of @xmath298 in ( [ polystate ] ) by @xmath297 .
denote the roots of @xmath300 in ( [ polystate ] ) by @xmath299 .
then the family @xmath348 is the only family up to scalar shift in this case such that @xmath128 is reducible .
the corresponding @xmath133 for which @xmath17 is signature @xmath0 is obtained by adding two @xmath310 s to @xmath334 , and adding two @xmath333 s to @xmath332 , yielding the family @xmath349 : @xmath350 this exhausts all of the possibilities and thus we have the statement in the theorem as desired . since from the previous section we have that any primitive hypergeometric hyperbolic monodromy group is obtained by permuting the coordinates of some interlacing pair @xmath302 , and since there are no sporadic such interlacing pairs in dimensions @xmath351 , we have that the families @xmath352 and @xmath349 completely describe the primitive hyperbolic hypergeometric @xmath17 s in dimension @xmath61
. we end this subsection with a remark about the integrality of the coefficients of @xmath1 for hyperbolic hypergeometrics .
writing @xmath353 , where @xmath177 and @xmath178 are as in theorem [ levthm ] , as @xmath354 with @xmath355 positive integers and @xmath356 .
the coefficients of the corresponding hypergeometric function are @xmath357 if these are to be integers then @xmath358 must be at least @xmath21 .
note that @xmath359 is the multiplicity of the @xmath96 s for which @xmath360 .
then @xmath361 where @xmath362 ( see @xcite , @xcite ) .
it is easy to check using landau s criterion @xcite and our characterization of hyperbolic hypergeometric monodromies , that for the latter @xmath363 iff @xmath364 . from the description of the families of hyperbolic hypergeometric monodromies in theorem [ theo ]
we see that @xmath365 , @xmath366 , @xmath367 , and @xmath368 are the infinite families with integral coefficients while the sporadic ones can be read off from tables 2 and 3 , namely the entries with @xmath369 .
this can be used to give the list of integral factorials ( [ star ] ) which correspond to hyperbolic hypergeometric monodromies .
this is the analogue of @xcite and @xcite who classify the integral @xmath370 s in ( [ star ] ) which correspond to finite monodromy groups ( which in turn correspond to @xmath371 ) . in this section
we calculate the quadratic forms preserved by the primitive hyperbolic hypergeometric monodromy groups described in the previous section
. this will be a necessary ingredient in our minimal distance graph method described in section [ mindistgraph ] . recall that the monodromy group @xmath142 is generated by two matrices @xmath372 where the characteristic polynomials of @xmath23 and @xmath24 , @xmath373 and @xmath374 respectively , are products of cyclotomic polynomials .
denote @xmath375 the eigenvalues of @xmath42 are @xmath21 and @xmath376 and their geometric multiplicities are @xmath59 and @xmath21 respectively .
the first @xmath59 elements @xmath377 of the standard basis are eigenvectors for the eigenvalue @xmath21 while @xmath103 is an eigenvector for the eigenvalue @xmath376 .
[ lemtec1 ] let @xmath378 .
then @xmath379 if and only if @xmath380 for every @xmath381 where @xmath382 .
the only if part follows from comparing the first @xmath59 columns of both sides of @xmath383 . for the if part it is suffices to check that also the last columns of @xmath384 and @xmath385 are equal .
the equalities of the first @xmath59 columns imply that @xmath386 for every @xmath218 .
the characteristic polynomial @xmath387 of @xmath23 is also the characteristic polynomial of @xmath388 since it is a a product of cyclotomic polynomials .
thus , @xmath389 @xmath390 [ lemtec2 ] assume that @xmath378 satisfies @xmath379 .
then @xmath391 if and only if @xmath392 is euclidean - orthogonal to @xmath393 for every @xmath381 .
in particular , if @xmath379 and @xmath391 then @xmath394 is euclidean - orthogonal to @xmath103 for every @xmath218 .
lemma [ lemtec1 ] and its analog with respect to @xmath24 implies that @xmath391 if and only if @xmath395 for every @xmath381 . as the first row of @xmath396
is @xmath397 while the other rows are zeros , @xmath392 must be orthogonal to @xmath103 ( under the usual euclidean scalar product ) .
for every @xmath398 @xmath399 so by induction on @xmath400 we see that @xmath54 satisfies the required condition if and only if @xmath401 which is equivalent to @xmath54 being euclidean - orthogonal to @xmath402 for every @xmath403 . finally , if @xmath379 , @xmath391 then also @xmath404 .
thus , for @xmath218 we have @xmath405 so @xmath406 .
[ prop form ] let @xmath23 , @xmath24 , @xmath42 , @xmath407 , @xmath408 and @xmath103 be as above .
assume that the group @xmath409 is primitive .
then : * the vectors @xmath410 are linearly independent and @xmath411 preserves the lattice @xmath56 spanned by them .
* there exists a unique ( up to a scalar product ) non - zero integral quadratic form @xmath412 such that @xmath23 and @xmath24 belong to its orthogonal group . *
if the quadratic form is normalized to have @xmath413 then @xmath414 equals to minus the @xmath415-coordinate of @xmath416 for every @xmath417 . *
if the coefficients of @xmath418 satisfy @xmath419 then the matrix representing @xmath54 w.r.t @xmath410 is given by @xmath420 we start by proving the existence part of @xmath421 . by lemma [ lemtec2 ] there exists a non - zero @xmath378 such that @xmath422 and @xmath391 .
assume first that @xmath54 is anti - symmetric so @xmath423 .
lemma [ lemtec2 ] implies that @xmath406 for every @xmath218 . since @xmath424 span @xmath425 ,
the set @xmath426 is a non - trivial proper subspace of @xmath425 .
this subspace is preserved by @xmath47 , a contradiction to the irreducibility of @xmath47 .
thus , @xmath54 is not anti - symmetric and the quadratic form @xmath427 has the desired properties .
lemma [ lemtec2 ] implies that @xmath428 for some @xmath429 so @xmath430 holds .
since @xmath47 is primitive , every non - trivial normal subgroup of it is irreducible .
thus in order to prove @xmath431 it is enough to show that the normal subgroup @xmath432 preserves the @xmath433-lattice spanned by @xmath434 .
property @xmath430 implies that @xmath435 for every @xmath417 . since @xmath24 preserves the quadratic form , @xmath436 for every @xmath417 .
thus , @xmath140 preserves the @xmath437-lattice spanned by @xmath438 and the proof of @xmath431 is complete .
the uniqueness part of @xmath421 now follows from @xmath431 together with lemma [ lemtec2 ] .
for @xmath218 the @xmath439-th and the @xmath440-th coordinates of bottom row of @xmath441 are equal to 1 and all the other coordinates of this row are zero .
thus , the last coordinate of @xmath442 is @xmath443 if @xmath444 and @xmath445 if @xmath446 .
from @xmath430 we get @xmath447 for every @xmath448 .
property @xmath449 follows from the fact the @xmath450 depends only on @xmath451 since @xmath24 preserves the quadratic form .
we now record the invariant quadratic form for two cases that we consider in section [ thinguys ] . [ corol form 1 ] let @xmath452 be the normalized quadratic form preserved by @xmath453 . if @xmath186 then @xmath454 [ corol form 3 ] let @xmath452 be the normalized quadratic form preserved by @xmath455 . if @xmath186 then @xmath456
as noted in the introduction we make crucial use of vinberg s @xcite , @xcite as well as nikulin s @xcite results concerning the size of groups generated by hyperbolic reflections . before reviewing this arithmetic theory
we begin with quadratic forms and hyperbolic space over the reals .
a non - degenerate real quadratic form @xmath457 with corresponding symmetric bilinear form @xmath458 is determined by its signature .
our interest is in the case that @xmath54 has signature @xmath459 , which after a real linear change of variable can be brought to the form @xmath460 the null cone @xmath42 of @xmath54 is @xmath461 as depicted in figure [ conefig ] .
@xmath462 consists of @xmath463 components , the outside , @xmath464 , which consists of one component , and the inside , @xmath465 , which consists of @xmath109 components : @xmath466 and @xmath467 .
the quadrics @xmath468 where @xmath469 are two sheeted hyperboloids and either sheet , say the one with @xmath466 and @xmath470 , can be chosen as a model of hyperbolic @xmath59 dimensional space ( we assume that @xmath180 ) , @xmath60 .
this is done by restricting the line element @xmath471 to @xmath60 .
the quadrics @xmath468 with @xmath472 are one sheeted hyperboloids .
if @xmath473 , the real orthogonal group of @xmath54 , then @xmath474 preserves the quadrics , however it may switch the sheets of the two sheeted hyperboloid .
if it preserves each of these , then @xmath474 acts on @xmath60 isometrically , while if @xmath474 switches the sheets then @xmath475 acts isometrically on @xmath60 . in either case
we obtain an induced isometry .
of particular interest to us are linear reflections of @xmath476 given by @xmath103 s with @xmath477 ( i.e. not on the null cone ) .
for such a @xmath103 denote by @xmath100 the linear transformation @xmath478 it is easy to see that @xmath479 , that it is an involution , and that it preserves the components of the complement of the light cone if @xmath480 , and switches them if @xmath481 . in the first case , the induced isometry of @xmath60 is a hyperbolic reflection .
the fixed point set of @xmath100 in @xmath476 is @xmath482 and this intersects @xmath60 in an @xmath483-dimensional hyperbolic hyperplane and the induced isometry is a ( hyperbolic ) reflection about this hyperplane . in the second case , @xmath481
, the induced isometry @xmath484 fixes the line @xmath485 and this line meets @xmath60 in a point @xmath486 .
the induced isometry of @xmath60 in this case inverts geodesics in @xmath486 and is a cartan involution .
these two involutive isometrics of @xmath60 are quite different . in the case of a discrete group of motions generated by hyperbolic reflections
, there is a canonical fundamental domain , namely the connected components of the complement of the hyperplanes corresponding to all the reflections .
this is the starting point to vinberg s theory of reflective groups . for groups generated by cartan involutions
there is no apparent geometric approach .
we turn to the arithmetic theory .
let @xmath487 be a quadratic lattice , that is a rank @xmath10 @xmath433-module equipped with a nondegenerate integral symmetric bilinear form @xmath458 , whose associated quadratic form @xmath488 is rational .
the dual lattice @xmath489 of @xmath56 is given by @xmath490 it contains @xmath56 as a finite sublattice and the invariant factors of @xmath56 are defined to be the invariants of the torsion module @xmath491 .
the product of these invariants is equal to @xmath492 , where @xmath493 is the discriminant of @xmath56 ( that is @xmath494 where @xmath73 is an integral matrix realization of @xmath54 ) .
we assume that @xmath495 is hyperbolic and of signature @xmath459 .
the @xmath433 analogue of the null cone is @xmath496 and that of the quadrics is @xmath497 for @xmath498 , @xmath499 .
we assume throughout that @xmath54 is isotropic over @xmath500 which means that @xmath501 ( and is in fact infinite ) .
the group of integral automorphs of @xmath56 is denoted by @xmath113 or @xmath78 . a primitive vector @xmath103 in @xmath502
is called a @xmath96-root if @xmath503 . in this case
the linear reflection @xmath100 of @xmath29 given by ( [ eq3.2 ] ) is integral and lies in @xmath113 . for @xmath504 or @xmath505
, @xmath506 is always in @xmath489 and we will have occasions where @xmath507 yields root vectors as well .
as discussed above , the root vectors with @xmath472 induce hyperbolic reflections on @xmath508 ( which we choose to be one of the components of @xmath509 ) while for @xmath469 they induce cartan involutions .
the vinberg reflective group @xmath510 is the subgroup of @xmath113 generated by all hyperbolic reflections coming from root vectors with @xmath472 .
it is plainly a normal subgroup of @xmath113 and the main result in @xcite asserts that if @xmath511 then @xmath512 .
vinberg s proof is based on studying the fundamental domain in @xmath60 of @xmath510 and relating it to the cusps of @xmath513 which correspond to null vectors @xmath514 .
in particular it uses the assumption that @xmath54 is isotropic .
nikulin s results @xcite are concerned with the case that @xmath56 is even , that is @xmath515 for all @xmath90 , and the subgroup @xmath116 generated by all the @xmath109-root vectors in @xmath56 ( note he chooses @xmath54 to have signature @xmath516 so that his @xmath92-root vectors are our @xmath109-root vectors ) .
@xmath116 is a subgroup of @xmath510 and it is also normal in @xmath113 . for these he gives a complete classification of all @xmath56 s for which @xmath517 is finite .
there are only finitely many such and for @xmath68 the list is quite short .
@xmath56 is called two elementary if @xmath491 is @xmath518 for some @xmath519 .
if @xmath56 is not two elementary and @xmath68 and odd ( the last condition on @xmath10 is what is of interest to us for our applications ) then @xmath520 unless @xmath56 is isomorphic to @xmath521 and @xmath76 is one of @xmath522 @xmath523 or @xmath56 is isomorphic to @xmath524 here the positive definite ( @xmath109-reflective ) lattices @xmath144 , @xmath525 , @xmath526 are the standard ones ( see @xcite ) while @xmath527 $ ] and @xmath528 $ ] . as a consequence unless @xmath56 is two elementary or its invariant factors are among the list below , @xmath529 .
[ cols="<,<",options="header " , ] h. park @xcite has recently shown that all the @xmath80 s in table 3 are thin except for the second entry , which is arithmetic . moreover , in that case @xmath140 is still contained in a thin reflection subgroup of @xmath80 . in particular , all the @xmath140 s in table 3 are thin . ] while our certificate for being thin succeeds in many cases as demonstrated in sections [ mindistgraph ] and [ numeric ] , there are families such as @xmath294 for which we found almost no paths between the basis vectors in the minimal distance graph .
if indeed the corresponding components of @xmath88 are singletons for members of this family , then clearly a variation of the minimal distance graph is needed if this approach is to work .
a crude , and it appears mostly reliable , test for the size of @xmath142 in @xmath530 is to simply count the number of elements in @xmath47 in a large ball .
that is , let @xmath531 be a large number and let @xmath532 if @xmath47 is arithmetic then it is known ( @xcite ) that @xmath533 thus to probe the size of @xmath47 we generate elements in @xmath47 using our defining generators and try to determine @xmath534 for @xmath531 quite large .
the difficulty is that we do nt know which elements of @xmath113 are in @xmath47 and the above procedure ( by going to large generations of elements in @xmath47 gotten from the generating set ) only gives a lower bound for @xmath534 . in any case we can compute in this way a lower bound for @xmath535 for @xmath531 large .
if this is essentially @xmath483 then this suggests that @xmath47 is arithmetic , while if this number is less than @xmath483 it suggests that @xmath47 is thin ( if @xmath47 is geometrically finite then this quantity should be an approximation to the hausdorff dimension of the limit set of @xmath47 ( see for example @xcite ) .
we have run this crude test for many of our hyperbolic hypergeometric monodromies and it gives the correct answer in the cases where we have a rigorous treatment .
for example , in figures [ arith1 ] and [ arith2 ] we give a plot of @xmath536 versus @xmath537 for the arithmetic examples @xmath538 and @xmath539 .
the similar plots for four examples with @xmath540 , and for which our certificate failed are given in figures [ num1 ] , [ num2 ] , [ num3 ] , and [ num4 ] below .
based on these , it appears that these @xmath47 s are all thin .
further such experimentation is consistent with conjecture [ allthinconj ] in the introduction and even that for @xmath68 , every hyperbolic hypergeometric monodromy group is thin . versus @xmath537 graph for @xmath17 where @xmath541 , @xmath542.,height=207 ] versus @xmath537 graph for @xmath17 where @xmath541 , @xmath543.,height=207 ] versus @xmath537 graph for @xmath17 where @xmath544 , @xmath545.,height=207 ] versus @xmath537 graph for @xmath17 where @xmath546 , @xmath547.,height=207 ] versus @xmath537 graph for @xmath17 where @xmath548 , @xmath549.,height=207 ] versus @xmath537 graph for @xmath17 where @xmath550 , @xmath545.,height=207 ]
in this appendix , we consider the six @xmath463-dimensional hyperbolic hypergeometric monodromy groups and show that they are all arithmetic .
our method is to determine the preimage of the groups in the spin double cover @xmath66 of @xmath551 and work in the setting of @xmath52 .
once we pull back to @xmath52 , we rely on one of the following strategies . if the quadratic form @xmath54 that is fixed by hyperbolic hypergeometric monodromy group @xmath47 is isotropic over @xmath500 then one can find @xmath552 such that @xmath553 is some scalar multiple of @xmath554 below .
moreover , under such a change of variable , the preimage of @xmath555 in the spin double cover of @xmath556 can be made to sit inside @xmath557 , and in this setting showing that the group is finite index is usually straightforward .
out of the six @xmath463-dimensional hyperbolic hypergeometric monodromy groups , four fix an isotropic form @xmath54 , and in those cases we show that the preimage of the group in the spin double cover of @xmath556 contains a principle congruence subgroup of @xmath558 if the form @xmath54 fixed by @xmath47 is anisotropic , we again pull back to the spin double cover @xmath559 of @xmath556 in @xmath66 and , after generating several elements of @xmath559 , construct a region in the upper half plane which contains a fundamental domain for @xmath559 acting on @xmath560 ( this method works for the isotropic case as well ) .
specifically , a dirichlet fundamental domain for @xmath559 is given by @xmath561 where @xmath562 is any base point and @xmath563 is a closed half space .
one can prove that @xmath564 is compact ( and hence @xmath559 is arithmetic ) if the intersection @xmath565 is compact even when taken over a _
finite _ number of elements @xmath566 .
note that this method uses the same idea as described in section [ numeric1 ] : generating elements of @xmath559 to say something about thinness , but in this case it can actually prove whether the group is thin or not .
note that @xmath576 where @xmath577\ ] ] and conjugation by @xmath578 gives an isomorphism of the subgroup @xmath579 with the subgroup @xmath580 where @xmath581 , b'=\left[\begin{array}{ccc } 1&0&0\\ 1&1&0\\ 1&2&1 \end{array}\right].\ ] ] the preimage of this group in @xmath557 under the spin homomorphism in ( [ spin2 ] ) is the group @xmath582 where @xmath583 , y=\left[\begin{array}{cc } 1&1\\ 0&1 \end{array}\right].\ ] ] this group contains the generators @xmath584 , \left[\begin{array}{cc } \minus 15&4\\ \minus 4&1 \end{array}\right ] , \left[\begin{array}{cc } 5&\minus 4\\ 4&\minus 3 \end{array}\right ] , \left[\begin{array}{cc } 9&\minus
16\\ 4&\minus 7 \end{array}\right ] , \left[\begin{array}{cc } 13&\minus 36\\ 4&\minus 11 \end{array}\right]\ ] ] of @xmath585 .
hence @xmath17 is itself arithmetic .
in addition , our second strategy of finding a region which contains a fundamental domain for @xmath47 yields the compact region shown in figure [ fundy1 ] .
@xmath587 , a=\left[\begin{array}{ccc } 0&0&\minus 1\\ 1&0&\minus 2\\ 0&1&\minus 2 \end{array}\right ] , b=\left[\begin{array}{ccc } 0&0&1\\ 1&0&\minus 3\\ 0&1&3 \end{array}\right].\ ] ] note that @xmath588 where @xmath589\ ] ] and conjugation by @xmath578 gives an isomorphism of the subgroup @xmath579 with the subgroup @xmath580 where @xmath590 , b'=\left[\begin{array}{ccc } 1&\minus 2&1\\ 0&1&\minus 1\\ 0&0&1 \end{array}\right].\ ] ] the preimage of this group in @xmath557 under the spin homomorphism in ( [ spin2 ] ) is the group @xmath582 where @xmath591 , y=\left[\begin{array}{cc } 1&0\\ \minus 1&1 \end{array}\right].\ ] ] this group contains the generators @xmath592 , \left[\begin{array}{cc } \minus 8&3\\ \minus 3&1 \end{array}\right ] , \left[\begin{array}{cc } 4&\minus 3\\ 3&\minus 2 \end{array}\right]\ ] ] of @xmath593 . hence @xmath17 is itself arithmetic .
in addition , our second strategy of finding a region which contains a fundamental domain for @xmath47 yields the compact region shown in figure [ fundy2 ] .
@xmath595 , a=\left[\begin{array}{ccc } 0&0&\minus 1\\ 1&0&\minus 1\\ 0&1&\minus 1 \end{array}\right ] , b=\left[\begin{array}{ccc } 0&0&1\\ 1&0&\minus 3\\ 0&1&3 \end{array}\right].\ ] ] note that @xmath596 where @xmath597\ ] ] and conjugation by @xmath578 gives an isomorphism of the subgroup @xmath579 with the subgroup @xmath598 where @xmath599 , b'=\left[\begin{array}{ccc } 1&\minus 2&\minus 2\\
2&\minus 1&\minus 2\\ \minus 2&2&3 \end{array}\right].\ ] ] the preimage of this group in @xmath557 under the spin homomorphism in ( [ spin1 ] ) is the group @xmath582 where @xmath600 , y=\left[\begin{array}{cc } 0&1\\ \minus 1&2 \end{array}\right]\ ] ] and @xmath601 generate the principal congruence subgroup @xmath602 . hence @xmath17 is itself arithmetic .
in addition , our second strategy of finding a region which contains a fundamental domain for @xmath47 yields the compact region shown in figure [ fundy3 ] .
@xmath604 , a=\left[\begin{array}{ccc } 0&0&\minus 1\\ 1&0&0\\ 0&1&0 \end{array}\right ] , b=\left[\begin{array}{ccc } 0&0&1\\ 1&0&\minus 3\\ 0&1&3 \end{array}\right].\ ] ] note that @xmath596 where @xmath605\ ] ] and conjugation by @xmath578 gives an isomorphism of the subgroup @xmath579 with the subgroup @xmath606 where @xmath607 , b'=\left[\begin{array}{ccc } 1/2&\minus 1&\minus 1/2\\ 1&1&1\\ 1/2&1&3/2 \end{array}\right].\ ] ] the preimage of this group in @xmath557 under the spin homomorphism in ( [ spin1 ] ) is the group @xmath582 where @xmath608 , y=\left[\begin{array}{cc } 1&1\\ 0&1 \end{array}\right]\ ] ] and @xmath609 are a well - known generating set for all of @xmath557 . hence @xmath17 is itself arithmetic .
in addition , our second strategy of finding a region which contains a fundamental domain for @xmath47 yields the compact region shown in figure [ fundy4 ] .
@xmath610 , a=\left[\begin{array}{ccc } 0&0&\minus 1\\ 1&0&\minus 2\\ 0&1&\minus 2 \end{array}\right ] , b=\left[\begin{array}{ccc } 0&0&1\\ 1&0&\minus 1\\ 0&1&1 \end{array}\right].\ ] ] in this case @xmath54 is anisotropic over @xmath500 , and hence we turn to our strategy of approximating the fundamental domain to prove that @xmath47 is arithmetic : we produce the compact region shown in figure [ fundy5 ] which contains the fundamental domain of @xmath47 .
@xmath611 , a=\left[\begin{array}{ccc } 0&0&\minus 1\\ 1&0&\minus 2\\ 0&1&\minus 2 \end{array}\right ] , b=\left[\begin{array}{ccc } 0&0&1\\ 1&0&\minus 2\\ 0&1&2 \end{array}\right].\ ] ] in this case @xmath54 is anisotropic over @xmath500 , and hence we turn to our strategy of approximating the fundamental domain to prove that @xmath47 is arithmetic : we produce the compact region shown in figure [ fundy6 ] which contains the fundamental domain of @xmath47 . 1 f. beukers , g. heckman , _
monodromy for hypergeometric function @xmath1_. invent . math . *
95 * , pp .
325 - 354 ( 1989 ) j. bober , _ integer ratios of factorials , hypergeometric functions , and related step functions_. ph.d .
thesis , university of michigan ( 2009 ) c. brav , h. thomas , _ classical and derived monodromy of the quintic threefold_. preprint at http://people.maths.ox.ac.uk/brav/monodromy.pdf j.w.s .
cassels , _ rational quadratic forms_. dover publications , inc . ,
mineola , ny ( 1978 ) j.h .
conway , n.j.a .
sloane , _ sphere packings , lattices and groups_. springer - verlag ( 1998 ) b. dwork , _
@xmath612-adic cycles_. publ .
ihes , * 37 * , pp.27 - 115 ( 1969 ) j. elstrodt , f. grunewald , f and j. mennicke , _ groups acting on hyperbolic space . harmonic analysis and number theory .
_ springer monographs in mathematics .
springer - verlag , berlin ( 1998 ) e. fuchs , _ the ubiquity of thin groups .
_ proceedings of msri conference on thin groups and super - strong approximation .
lax , r. phillips , _ the asymptotic distribution of lattice points in euclidean and non - euclidean spaces_. jnl .
46 * , pp .
280 - 350 ( 1982 ) e. landau , _ sur les conditions de divisibilite dun produit de factorielles par un autre _ , nouv .
, * 3 * , 19 , pp .
344 - 362 ( 1900 ) a.h.m levelt , _ hypergeometric functions . _ ph.d .
thesis , university of amsterdam ( 1961 ) g. margulis , _ discrete subgroups of semi - simple lie groups _ , springer ( 1991 ) j. mckee , c. smyth , _ single polynomials that correspond to pairs of cyclotomic polynomials with interlacing zeros .
eur . j. math . ,
* 11 * , pp .
882 - 899 ( 2013 ) v. nikulin , _ on factor groups of the automorphism groups of hyperbolic forms modulo subgroups generated by 2-reflections_. soviet math
* 20 * , pp .
1156 - 1158 ( 1979 ) v. nikulin , _ self correspondences of @xmath613-surfaces via moduli of sheaves and arithmetic hyperbolic reflection groups_. arxiv : 0810.2945 h. park , ph.d .
thesis , princeton university ( august 2013 ) .
f. rodriguez - villegas , _ hypergeometric families of calabi - yau manifolds_. fields inst .
commun . * 38 * , ams ( 2003 ) p. sarnak , _ notes on thin matrix groups_. arxiv : 1212.3525 p. sarnak , _ reciprocal geodesics_. clay math . * 7 * , pp . 217 - 237 ( 2007 ) s. singh , t.n .
venkataramana , _ arithmeticity of certain symplectic hypergeometric groups_. arxiv : 1208.6460 ( 2012 ) t.n .
venkataramana , _ arithmeticity of the braid group at roots of unity_. arxiv : 1204.4778 ( 2012 ) e. vinberg , _ on groups of unit elements of certain quadratic forms_. math .
ussr sbornik * 16 * , no.1 , pp .
17 - 35 ( 1972 ) e. vinberg , _ the absence of crystallographic reflection groups in lobachevsky spaces of large dimension_. trudy moscow
47 * , pp .
68 - 102 ( 1984 ) | we give a criterion which ensures that a group generated by cartan involutions in the automorph group of a rational quadratic form of signature @xmath0 is `` thin '' , namely it is of infinite index in the latter .
it is based on a graph defined on the integral cartan root vectors , as well as vinberg s theory of hyperbolic reflection groups .
the criterion is shown to be robust for showing that many hyperbolic hypergeometric groups for @xmath1 are thin . |
in this paper , all graphs considered are simple , finite and undirected .
we refer to book @xcite for undefined notation and terminology in graph theory .
let @xmath0 be a connected graph on @xmath4 vertices with minimum degree @xmath5 .
a path in an edge - colored graph @xmath0 is a _ rainbow path _ if its edges have different colors .
an edge - colored graph @xmath0 is _ rainbow connected _ if any two vertices of @xmath0 are connected by a rainbow path of @xmath0 .
the _ rainbow connection number _ , denoted by @xmath20 , is defined as the smallest number of colors required to make @xmath0 rainbow connected .
chartrand et al .
@xcite introduced these concepts .
notice that @xmath21 if and only if @xmath0 is a complete graph and that @xmath22 if and only if @xmath0 is a tree .
moreover , @xmath23 . a lot of results on the rainbow connection have been obtained ; see @xcite .
from @xcite we know that to compute the number @xmath20 of a connected graph @xmath0 is np - hard .
so , to find good upper bounds is an interesting problem .
krivelevich and yuster @xcite obtained that @xmath24 .
caro et al .
@xcite obtained that @xmath25 .
finally , chandran et al .
@xcite got the following benchmark result .
[ thm 1]@xcite for every connected graph @xmath0 of order @xmath4 and minimum degree @xmath5 , @xmath26 .
the concept of rainbow vertex - connection was introduced by krivelevich and yuster in @xcite .
a path in a vertex - colored graph @xmath0 is a _ vertex - rainbow path _ if its internal vertices have different colors .
a vertex - colored graph @xmath0 is _ rainbow vertex - connected _ if any two vertices of @xmath0 are connected by a vertex - rainbow path of @xmath0 .
the _ rainbow vertex - connection number _ , denoted by @xmath27 , is defined as the smallest number of colors required to make @xmath0 rainbow vertex - connected .
observe that @xmath28 and that @xmath29 if and only if @xmath0 is a complete graph .
the problem of determining the number @xmath27 of a connected graph @xmath0 is also np - hard ; see @xcite .
there are a few results about the upper bounds of the rainbow vertex - connection number .
krivelevich and yuster @xcite proved that @xmath30 .
li and shi @xcite improved this bound and showed the following results .
[ thm 2]@xcite for a connected graph @xmath0 of order @xmath4 and minimum degree @xmath5 , @xmath31 for @xmath15 , @xmath32 for @xmath17 and @xmath33 for @xmath19 . for sufficiently large @xmath5 , @xmath34 , where @xmath35 is any constant exceeding 2.5 .
[ thm 3]@xcite a connected graph @xmath0 of order @xmath4 with minimum degree @xmath5 has @xmath36 for @xmath37 and @xmath38 , while @xmath39 for @xmath40 and @xmath41 for @xmath42 , where @xmath13 . recently , liu et al .
@xcite proposed the concept of total rainbow connection .
total - colored graph _ is a graph @xmath0 such that both all edges and all vertices of @xmath0 are colored .
a path in a total - colored graph @xmath0 is a _
total rainbow path _ if its edges and internal vertices have distinct colors .
a total - colored graph @xmath0 is _ total - rainbow connected _ if any two vertices of @xmath0 are connected by a total rainbow path of @xmath0 .
the total rainbow connection number , denoted by @xmath1 , is defined as the smallest number of colors required to make @xmath0 total - rainbow connected .
it is easy to observe that @xmath43 if and only if @xmath0 is a complete graph .
moreover , @xmath2 .
the following proposition gives an upper bound of the total rainbow connection number .
[ prop 1]@xcite let @xmath0 be a connected graph on @xmath4 vertices and @xmath44 vertices having degree at least @xmath45 . then , @xmath46 , with equality if and only if @xmath0 is a tree . from theorem [ thm 1 ] and [ thm 3 ] , one can see that @xmath20 and @xmath27 are bounded by a function of the minimum degree @xmath5 , and that when @xmath5 is in linear with @xmath4 , then both @xmath20 and @xmath27 are some constants . in this paper , we will use the same idea in @xcite to obtain upper bounds for the number @xmath1 , which are also functions of @xmath5 and imply that when @xmath5 is in linear with @xmath4 , then @xmath1 is a constant .
let @xmath0 be a connected graph on @xmath4 vertices with minimum degree @xmath5
. denote by @xmath47 the maximum number of leaves in any spanning tree of @xmath0 . if @xmath15 , then @xmath48 which was proved by linial and sturtevant ( unpublished ) . in @xcite
, it was proved that @xmath49 for @xmath17 .
moreover , griggs and wu @xcite showed that if @xmath19 , then @xmath50 .
for sufficiently large @xmath5 , @xmath51 , where @xmath35 is any constant exceeding 2.5 , which was proved in @xcite .
thus , we can get the following results .
[ thm 4 ] for a connected graph @xmath0 of order @xmath4 with minimum degree @xmath5 , @xmath14 for @xmath15 , @xmath16 for @xmath17 and @xmath18 for @xmath19 . for sufficiently large @xmath5 , @xmath52 , where @xmath35 is any constant exceeding 2.5 .
we can choose a spanning tree @xmath53 with the most leaves .
denote @xmath54 the maximum number of leaves .
then color all non - leaf vertices and all edges of @xmath53 with @xmath55 colors , each receiving a distinct color .
hence , @xmath56 . given a graph @xmath0 , a set @xmath58
is called a _
@xmath45-step dominating set _ of @xmath0 if every vertex of @xmath0 which is not dominated by @xmath59 has a neighbor that is dominated by @xmath59 .
a @xmath45-step dominating set @xmath60 is _
@xmath61-strong _ if every vertex which is not dominated by @xmath60 has at least @xmath61 neighbors that are dominated by @xmath60 .
if @xmath60 induces a connected subgraph of @xmath0 , then @xmath60 is called a _ connected @xmath61-strong @xmath45-step dominating set_. these concepts can be found in @xcite .
[ lem3]@xcite ( @xmath65sz local lemma ) let @xmath66 be the events in an arbitrary probability space .
suppose that each event @xmath67 is mutually independent of a set of all the other events @xmath68 but at most @xmath69 , and that @xmath70\leq p$ ] for all @xmath71 . if @xmath72 , then @xmath73>0 $ ] .
* proof of theorem 5 : * the proof goes similarly for the main result of @xcite .
we are given a connected graph @xmath0 of order @xmath4 with minimum degree @xmath5 .
suppose that @xmath0 has less than @xmath57 edges by lemma [ lem1 ] .
let @xmath60 denote a connected @xmath63-strong @xmath45-step dominating set of @xmath0 .
then , we have @xmath74 by lemma [ lem2 ] .
let @xmath75 denote the set of all vertices at distance exactly @xmath61 from @xmath60 .
we give a partition to @xmath76 as follows .
first , let @xmath77 be a new graph constructed on @xmath76 with edge set @xmath78 or @xmath79 such that @xmath80 is a path of @xmath81 .
let @xmath82 be the set of all isolated vertices of @xmath77 .
moreover , there exists a spanning forest @xmath83 of @xmath84 .
finally , choose a bipartition defined by this forest , denoted by @xmath85 and @xmath86 .
partition @xmath87 into three subsets : @xmath88 , @xmath89 and @xmath90 ; see figure 1(a ) .
next we give a coloring to the edges and vertices of @xmath0 .
let @xmath91 and @xmath53 be a spanning tree of @xmath92 $ ] .
color the edges and vertices of @xmath53 with @xmath61 distinct colors such that @xmath92 $ ] is total rainbow connected .
assign every @xmath93 $ ] edge with color @xmath94 , every @xmath95 $ ] edge with color @xmath96 and every edge in @xmath76 with color @xmath97 .
since the minimum degree @xmath62 , every vertex in @xmath82 has at least two neighbors in @xmath60 .
color one edge with @xmath94 and all others with @xmath96 .
assign every @xmath98 $ ] edge with color @xmath97 , every @xmath99 $ ] edge with color @xmath100 and every vertex of @xmath101 with color @xmath102 .
we assign seven new colors from @xmath103 to the vertices of @xmath85 such that each vertex of @xmath85 chooses its color randomly and independently from all other vertices of @xmath85 .
similarly , we assign another seven colors to the vertices of @xmath86 .
assign seven colors from @xmath104 to the edges between @xmath105 and @xmath85 as follows : for every vertex @xmath106 , let @xmath107 denote the set of all neighbors of @xmath108 in @xmath85 ; for every vertex @xmath109 , if we color @xmath110 with @xmath111 , then color @xmath112 with @xmath113 . in a similar way , we assign seven new colors to the edges between @xmath114 and @xmath86 .
all other edges and vertices of @xmath0 are uncolored .
thus , the number of all colors we used is @xmath115 [ claim1 ] for any @xmath116 , we have a coloring for the vertices in @xmath117 with seven colors such that there exist two neighbors @xmath118 and @xmath119 in @xmath120 that receive different colors .
hence , the edges @xmath121 and @xmath122 are also colored differently .
notice that for every vertex @xmath123 , @xmath124 has two neighbors in @xmath125 .
moreover , @xmath126 .
thus , @xmath124 has less than @xmath127 neighbors in @xmath105 . for every vertex @xmath128
, @xmath108 has at least @xmath63 neighbors in @xmath85 since @xmath60 is a connected @xmath63-strong @xmath45-step dominating set of @xmath0 .
let @xmath129 denote the event that @xmath107 receives at least two distinct colors .
fix a set @xmath130 with @xmath131 .
let @xmath132 denote the event that all vertices of @xmath133 are colored the same .
hence , @xmath134\leq 7^{-\lceil\delta/3\rceil+1}$ ] .
moreover , the event @xmath132 is independent of all other events @xmath135 for @xmath136 but at most @xmath137 of them .
since @xmath138 , for all @xmath7 and @xmath8 , we have @xmath139>0 $ ] by lemma [ lem3 ] . therefore , @xmath140>0 $ ] .
we will show that @xmath0 is total - rainbow connected .
take any two vertices @xmath108 and @xmath141 in @xmath142 .
if they are all in @xmath60 , there is a total rainbow path connecting them in @xmath92 $ ] .
if one of them is in @xmath76 , say @xmath108 , then @xmath108 has a neighbor @xmath110 in @xmath60 .
thus , @xmath143 is a required path , where @xmath144 is a total rainbow path in @xmath92 $ ] connecting @xmath110 and @xmath141 .
if one of them is in @xmath145 , say @xmath108 , and the other is in @xmath146 , say @xmath141 , then @xmath108 has a neighbor @xmath110 in @xmath60 and @xmath141 has a neighbor @xmath147 in @xmath60 .
hence , @xmath148 is a required path , where @xmath144 is a total rainbow path connecting @xmath110 and @xmath147 in @xmath92 $ ] .
if they are all in @xmath85 , then there exists a @xmath149 such that @xmath108 and @xmath110 are connected by a single edge or a total rainbow path of length two . we know that @xmath110 and @xmath141 are total - rainbow connected .
therefore , @xmath108 and @xmath141 are connected by a total rainbow path .
if one of them is in @xmath150 , say @xmath108 , and the other is in @xmath151 , say @xmath141 , then @xmath108 has a neighbor @xmath110 in @xmath85 , and @xmath141 has a neighbor @xmath147 in @xmath86 .
thus , they are total - rainbow connected .
if they are all in @xmath105 , by claim 1 @xmath108 has two neighbors @xmath118 and @xmath119 in @xmath85 such that @xmath152 and @xmath122 are colored differently .
similarly , we also have that @xmath141 has two neighbors @xmath153 and @xmath154 in @xmath85 such that @xmath155 and @xmath156 are colored differently .
hence , @xmath108 and @xmath141 are total - rainbow connected .
we can check that @xmath108 and @xmath141 are total - rainbow connected in all other cases .
we partition @xmath85 into two subsets @xmath158 and @xmath159 . for any @xmath160 , if @xmath108 has at least @xmath161 neighbors in @xmath105 , then @xmath162 ; otherwise , @xmath163 .
similarly , we partition @xmath86 onto two subsets @xmath164 and @xmath165 .
note that @xmath166 since @xmath0 has less than @xmath167 edges .
partition @xmath105 into two subsets @xmath168 and @xmath169 . for any @xmath128 , if @xmath108 has at least one neighbor in @xmath158 , then @xmath170
; otherwise , @xmath171 . in a similar way , we partition @xmath114 into two subsets @xmath172 and @xmath173 ; see figure 1(b ) . for @xmath10 ,
assume that @xmath174 ; for @xmath175 , assume that @xmath13 .
now we give a coloring to the edges and vertices of @xmath0 .
let @xmath91 and @xmath53 be a spanning tree of @xmath92 $ ] .
color the edges and vertices of @xmath53 with @xmath61 distinct colors .
assign every @xmath176 $ ] edge with color @xmath94 , every @xmath177 $ ] edge with color @xmath96 and every edge in @xmath76 with color @xmath97 .
since every vertex in @xmath82 has at least two neighbors in @xmath60 , color one edge with @xmath94 and all others with @xmath96 . assign every @xmath178 $ ] edge with color @xmath97 , every @xmath179 $ ] edge with color @xmath100 and every vertex of @xmath101 with color @xmath102 .
assign distinct colors to each vertex of @xmath180 and @xmath181 new colors from @xmath182 to the vertices of @xmath159 such that each vertex of @xmath159 chooses its color randomly and independently from all other vertices of @xmath159 .
similarly , we assign @xmath181 new colors to the vertices of @xmath165 . for every vertex @xmath183 ,
if @xmath124 has at least two neighbors in @xmath158 , color one edge with @xmath184 and all others with @xmath185 ; if @xmath124 has only one neighbor in @xmath158 , then it has another neighbor in @xmath159 since @xmath60 is a connected @xmath63-strong @xmath45-step dominating set .
thus , color the edge incident with @xmath158 with @xmath184 and all edges incident with @xmath159 with @xmath185 .
we assign @xmath181 colors from @xmath186 to the edges between @xmath169 and @xmath159 . for every vertex @xmath187 ,
let @xmath188 denote all the neighbors of @xmath108 in @xmath159 .
for every vertex @xmath189 , if we color @xmath110 with @xmath190 , then color @xmath112 with @xmath113 . in a similar way
, we assign another @xmath191 colors to the edges between @xmath114 and @xmath86 . all other edges and vertices of @xmath0
are uncolored .
hence , the number of all colors we used is @xmath192 [ claim2 ] for any @xmath193 , we have a coloring for the vertices in @xmath194 with @xmath181 colors such that there exist two neighbors @xmath118 and @xmath119 in @xmath195 that receive different colors .
thus , the edges @xmath121 and @xmath122 are also colored differently .
notice that every vertex @xmath108 of @xmath169 has at least @xmath63 neighbors in @xmath159 since @xmath60 is a connected @xmath63-strong @xmath45-step dominating set of @xmath0 .
let @xmath129 denote the event that @xmath188 receives at least two distinct colors .
fix a set @xmath196 with @xmath197 .
let @xmath132 denote the event that all vertices of @xmath198 are colored the same .
therefore , @xmath134\leq ( c(\delta)+2)^{-\lceil\delta/3\rceil+1}$ ] .
moreover , the event @xmath132 is independent of all other events @xmath135 for @xmath136 but at most @xmath137 of them .
since @xmath199 , we have @xmath200>0 $ ] by lemma 3 .
hence , we have @xmath140>0 $ ] . | a total - colored graph is a graph @xmath0 such that both all edges and all vertices of @xmath0 are colored .
a path in a total - colored graph @xmath0 is a total rainbow path if its edges and internal vertices have distinct colors .
a total - colored graph @xmath0 is total - rainbow connected if any two vertices of @xmath0 are connected by a total rainbow path of @xmath0 .
the total rainbow connection number of @xmath0 , denoted by @xmath1 , is defined as the smallest number of colors that are needed to make @xmath0 total - rainbow connected .
these concepts were introduced by liu et al .
notice that for a connected graph @xmath0 , @xmath2 , where @xmath3 denotes the diameter of @xmath0 and @xmath4 is the order of @xmath0 . in this paper
we show , for a connected graph @xmath0 of order @xmath4 with minimum degree @xmath5 , that @xmath6 for @xmath7 and @xmath8 , while @xmath9 for @xmath10 and @xmath11 for @xmath12 , where @xmath13 .
this implies that when @xmath5 is in linear with @xmath4 , then the total rainbow number @xmath1 is a constant .
we also show that @xmath14 for @xmath15 , @xmath16 for @xmath17 and @xmath18 for @xmath19 .
furthermore , an example shows that our bound can be seen tight up to additive factors when @xmath7 .
* keywords * : total - colored graph ; total rainbow connection ; minimum degree ; 2-step dominating set . *
ams subject classification 2010 * : 05c15 , 05c40 , 05c69 , 05d40 . |
the recent advocacy of the use of bayesian statistics for the analysis of data from lattice simulations , in the guise of the methods of constrained curve fitting @xcite , or maximum entropy @xcite , has eased considerably the ambiguity and irritation associated with estimating the systematic errors due to curve fitting , especially when extracting masses , spectral weights and matrix elements from monte carlo estimates of correlation functions .
previously , monte carlo estimates , @xmath6 , of two - point hadronic correlators had been fit to a theoretical model , such as @xmath7 where @xmath8 is the spectral weight of the @xmath9 state , by the maximum - likelihood procedure of minimizing the @xmath0 @xmath10 with covariance matrix @xmath11 traditionally , these had been fit only at large euclidean times @xmath12 , where contributions from excited states are exponentially damped .
the art had been to choose a value of @xmath13 which compromises between unnecessarily high statistical errors for large @xmath13 and high systematic errors ( from contamination from excited states ) for small @xmath14 .
lattice alchemy provided various recipes for making the compromise and estimating the systematic errors , but the procedures were often suspect and always frustrating .
the truncation of the data set to only a few large @xmath15 was deemed necessary because the alternative ( of including more time slices but also more terms in the fit model ) resulted in unacceptably unstable fits to the sum of decaying exponentials ( traditionally a bane of numerical analysts ) .
success was achieved in some cases by enlarging the data set by including more channels , e.g. diagonalization of multi - source multi - exponential fits .
indeed when correlators from very many sources could be calculated cheaply , such as for glueballs or static quarks , the improvement was dramatic .
but most often , when only a couple of channels at best could be fit simultaneously , the competition between increased statistical errors for large @xmath13 and large systematic errors for small @xmath13 remained ; although the final statistical and systematic errors were reduced , the effort and uncertainty in obtaining a reliable systematic error remained .
constrained curve fitting @xcite offers the alternative of minimizing an augmented @xmath0 , @xmath16 where @xmath17 denotes the collective parameters of the fit ( e.g. @xmath18 for a sum of exponentials ) , as a way of achieving stability by `` guiding the fit '' with the use of bayesian priors , that is , values of the parameters obtained from _ a priori _ estimates @xmath19 . with
improved stability , the data sets can be enlarged to include small @xmath15 and the theory can be enlarged by including many more terms in the fit model until convergence is obtained . the systematic error associated with the choice of @xmath13 is thereby largely absorbed into the statistical error .
the advantage that the constrained curve fitting of lattice data has over a typical data set that a numerical generalist would consider , is that often we have reliable estimates of , or at least constraints on , the fit parameters from outside the data ( for example , the masses must be positive , or the level spacing is expected to be such - and - such from reliable models ) which can then be used as bayesian priors .
examples , such as upsilon spectroscopy @xcite where the level spacing can be reliably estimated from quark models and experiments , are impressive . remarkably , constrained curve fitting with bayesian priors on such data has been able to give satisfactory fits for local - local correlation functions , i.e. when multi - source fits are unavailable ( presumably due to prohibitive cost ) .
but with our recent data @xcite , we enter previously unexplored territory .
we work with overlap fermions with exact chiral symmetry at unprecedented small quark mass and large spatial volume . the literature , from which to obtain estimates to be used as priors , is limited .
furthermore , the details of the level spacings ( e.g. the roper resonance and the @xmath20 ) are hotly debated between advocates of quark models versus those of chiral models .
the use of priors in standard constrained curve fitting tends to `` lock in '' the fit ( within a sigma or so ) ; if one gets them badly wrong , then the fitted results may be misleading .
furthermore , the stability of the fit results against choice of prior must be tested this reintroduces an element of subjectivity . as a modification of the basic bayesian - prior constrained - curve fitting ( augmented @xmath0 ) procedure
, we propose to make it more automatic , and to further absorb systematic errors associated with choice of prior into statistical errors . in section
ii , we give an overview of the `` sequential empirical bayes method '' detailing our extension of constrained curve - fitting . in section iii , we add some further improvements to better assess and reduce systematic errors , and study fits to artificial data where the true values of the parameters are known . in section iv we give , as an illustration of the efficacy of the algorithm , some results from our low quark mass overlap fermion data for the excited states of the pion , present preliminary fits for the @xmath1 where ghost states ( a quenched artifact ) must be dealt with , and comment on the details of fits of the roper resonance and @xmath2 previously presented elsewhere @xcite . our summary and conclusions follow in section v.
bayesian statistics is an entire field in itself , with an old and broad history ; for an introduction , see @xcite . empirical bayes methods are intermediate between classical ( `` frequentist '' ) and bayesian methods .
the core ideas of a sequential analysis of the data originate with robbins @xcite .
we propose the sequential empirical bayes ( seb ) method as a refinement especially well suited for the special properties of lattice monte carlo correlation functions .
we begin in subsection [ sect : synopsis ] by giving a brief description of the standard constrained - curve fitting approach emphasizing its basis in bayesian probability theory .
our sketch relies heavily on a few recent and relevant summaries @xcite ; indeed , our notation is an amalgam of theirs .
we follow with a descriptive overview of our method in subsection [ sect : overview ] and a more detailed rendering of the basic algorithm in subsection [ sect : basicalgorithm ] . to account for the more general case of multi - source fits , as well as to allow the time dependence to be non - exponential ( periodic boundary conditions result in @xmath21 or @xmath22 , and quenched artifacts can give rise to more complicated temporal dependence ) ,
we write @xmath23 where @xmath24 are the collective parameters of the fit ( e.g. @xmath18 for a sum of exponentials ) , the indices @xmath25 , @xmath26 distinguish different values of the independent variable ( time for correlation functions ) and different interpolating fields , @xmath27 are the monte carlo data , @xmath28 is the fit model , and @xmath29 is the covariance matrix . minimizing the @xmath0 in eq . [ generalchisquare ] is the solution of the problem of determining the set of fit parameters @xmath24 which maximizes @xmath30 , the conditional probability of measuring the data @xmath27 given a set of parameters , also known as the `` likelihood '' of the data .
bayesian inference turns this question around and demands that the solution of the curve - fitting problem consist of determining the set of parameters @xmath24 which maximizes @xmath31 , the conditional probability that @xmath24 is correct given the measured data @xmath27 .
that is , bayesian inference asks which fit - model parameters are most likely given the data .
the computation of the latter conditional probability is possible because of the celebrated bayes theorem @xmath32 which follows directly from the elementary properties of probability theory @xmath33 the unconditional probability @xmath34 is the plausibility one assigns to the parameters @xmath24 before the additional information of the fit is provided , and is known as the `` bayesian prior distribution '' . the conditional probability @xmath31 is known as the `` posterior probability distribution '' ; it is the reassessment of the likelihood of the parameters after the fit incorporates the data .
the unconditional probability @xmath35 , known as the `` prior predictive probability '' of the data , is independent of @xmath24 and thus serves only as a normalization constant ; it is determined from eq .
[ bayestheorem ] and the normalization condition @xmath36 .
heuristically , bayes theorem can be thought of as `` posterior probability '' @xmath37 `` likelihood '' @xmath38 `` prior probability '' .
traditional inference ( `` frequentist theory '' ) uses the likelihood @xmath30 to quantify all statistical measures of inference ( mean , standard deviation , @xmath0 , confidence limits , @xmath39 ) , while bayesian inference uses the posterior distribution @xmath31 to determine the statistics . from eq .
[ bayestheorem ] , bayesian theory requires augmenting the information that the frequentist theory uses with the additional estimate of the prior distribution @xmath34 .
of course , the two methods agree when the prior distribution is chosen to be constant .
the likelihood @xmath30 , needed by both frequentists and bayesians , can be estimated if the monte carlo data set is sufficiently large .
then , regardless of the statistics of the underlying population distribution ( the ensemble of all possible configurations ) , the sample distribution ( monte - carlo - generated averages over finite sets of configurations ) will have gaussian statistics , as assured by the central limit theorem .
thus @xmath40 with the @xmath0 given by eq .
[ generalchisquare ] .
the bayesian prior distribution @xmath34 is the probability that a particular set of parameters @xmath24 is correct , _ a priori _ to the data analysis .
it is implicit in that the bayesian probabilities are conditional on some background information .
operationally , @xmath34 serves as a constraint on the parameters @xmath24 . for the physicist
, these might be constraints on the ranges of parametric values that are physically feasible ( e.g `` positive energy splittings '' ) , or perhaps something stronger as dictated by experience from previous similar fits or guidance from models of qcd or experiments .
a simple choice for the prior distribution is gaussian @xmath41 with @xmath42 defined in eq .
[ chiprior ]
. then @xmath43 of eq .
[ bayestheorem ] is maximized by minimizing the augmented @xmath0 ( @xmath44 ) of eq .
[ augmentedchisquare ] .
this is the approach outlined in @xcite where they choose values for the priors @xmath45 and @xmath46 based on physicists intuition .
the opposite extreme is the use of the `` entropic prior '' , based on the view that for some fits where the number of parameters is larger than the number of measured data ( such as for the spectral density function @xcite , or in a wider context image restoration ) only minimal information about the priors is available .
we seek an approach between the two extremes .
we use the language of augmented @xmath0 but seek to use a subset of the available data to estimate the priors in an orderly fashion , which will then be used on new data .
so how do we obtain our priors ?
for concreteness , consider again the fit model of the sum of decaying exponentials in eq .
[ exp ] . apportion the data into a nested set ( picture the shells in an onion or a kewpie doll ) and
extract estimates for priors in a progressive manner , at each stage obtaining estimates of one or two new priors upon each expansion of the data set .
an especially natural and well - suited nesting is to partition the hadronic correlator data via euclidean time slices .
that is , the first and smallest data set includes all configurations but with times slices @xmath15 restricted to @xmath47 ( and perhaps @xmath48 depending on boundary conditions ) .
subsequently , the @xmath49 data set includes all time slices such that @xmath50 , where @xmath51 and @xmath52 is most simply chosen to be the constant @xmath53 .
( a more general choice is made in practice , as described later . )
the time @xmath54 is chosen by the same criteria used in traditional fits so that an unconstrained fit is well approximated by a single exponential .
the output values for two parameters , the ground state mass and spectral weight , are then used as priors for the next fit on the augmented data set . for the second fit , two ( or more , in general ) additional time slices are included , as are two more parameters in the fit model , the weight and mass of the first excited state .
the new fit is constrained with regard to the ground state but unconstrained with regard to the first excited state .
the four fitted parameters are then used as priors in a third fit on a larger data set , and so on until all desired time slices and/or terms in the fit are included .
in this way , we estimate the priors from a subset of the data .
furthermore , we choose that subset of the data which is best suited for making the estimation .
the choice is determined with the same compromise as is used for traditional unconstrained fits , namely between minimizing statistical errors ( which grow as @xmath15 increases ) and minimizing the contamination of higher excited states ( which grows as @xmath15 decreases ) .
thus we propose an adaptive self - contained constrained curve - fitting procedure , dubbed the `` sequential empirical bayes '' method . in a nutshell
, we obtain the priors gradually ( allowing them to change as needed ) , from the ground state up , as the data set is monotonically enlarged by including earlier and earlier time slices .
its advantages include that it is usable whenever external reliable estimates of the priors are not available , and it is as automatic as one could hope for , thereby reducing the potential to introduce bias and of course decreasing the frustrating busy - work of fitting . especially for the low - lying states ,
if the initial priors are estimated incorrectly , there are several subsequent steps by which they may change ( by about a sigma each time ) . from experience
we have discovered that for some data , either `` real '' ( actual lattice gauge theory data ) or artificially constructed , a very basic algorithm which incorporates the seb philosophy is adequate for producing reliable priors .
however , some of our data have exposed deficiencies in the basic algorithm . these have been corrected without violating the spirit of the seb , but make the resulting final algorithm rather mysterious and complicated to describe in one pass . accordingly , in this paper , we begin here by outlining the simplest algorithm as a template . in sect .
[ sect : systematic ] we will introduce modifications as necessary . [
cols="^,^,^,^,^ " , ] table [ table : parameters ] describes for simplicity the very basic `` fixed @xmath52 '' algorithm , in which after each pair of steps , @xmath55 new ( earlier ) time slices are added to the data to be fitted .
the algorithm is as follows : * choose @xmath56 and @xmath13 , the maximum window over which the fits will be done .
the window is to be chosen as large as possible . for correlation functions expected , on theoretical grounds , to be positive
then if a value of a correlation function at a time slice is not within one sigma of being positive , then that time slice and all greater time slices are eliminated from consideration .
this prevents noisy correlation functions at large @xmath15 from giving grossly inaccurate estimates of the fitted values to be subsequently used as priors .
* determine the number of the terms we want to use in the fit model , and determine @xmath54 as the starting point for the fit .
ensure that @xmath57 * choose central trial values @xmath58 and @xmath59 ( or @xmath60 ) equal to those obtained from the effective mass relations .
loop on various trial values around these central values . for each , use an unconstrained fit on the one - mass - term model to fit the correlator data including time slices @xmath61 to @xmath56 and obtain @xmath62 and @xmath63 .
choose as input for the next step those values which yield the lowest ( but reasonable ) @xmath64 .
* using these values of @xmath65 and @xmath66 as both priors and initial values , do a constrained curve fit ( using the one - mass - term model on the data set enlarged to include @xmath67 ) to obtain @xmath68 and @xmath69 .
* loop on a wide range of trial values for @xmath70 and @xmath71 . with a two - mass - term model ,
constrain the first mass and weight ( using the previous output as both priors and initial values ) but leave the second mass and weight unconstrained . loop on various trial values for the latter .
do this half - constrained fit on the data set enlarged to include @xmath72 and obtain @xmath73 and @xmath74 .
choose as input for the next step those values which yield the lowest ( but reasonable ) @xmath75 .
* using these values of @xmath76 , @xmath66 , @xmath77 , @xmath78 as both priors and initial values , do a fully - constrained fit ( using the two - mass - term model on the data set enlarged to include @xmath79 ) to obtain @xmath80 , @xmath81 , @xmath82 , and @xmath83 .
* repeat the last two steps until all desired mass terms and time slices are included .
one thus obtains a complete set of priors . *
add the final time slice @xmath13 and do a fully - constrained fit using previously obtained values for priors and initial guesses . all fits are correlated using the full covariance matrix . furthermore , the entire process is bootstrapped ( or jackknifed ) .
final quoted errors are bootstrap ( or jackknife ) errors . within a bootstrap sample , at intermediate steps in the algorithm
, the sigmas of the priors , @xmath84 , are obtained from the fitting errors of the previous step . as a variation , rather than deciding _ a priori _ on the number of terms in the fit and adding time slices one at time
, one can let the data decide how many time slices to include with each enlargement of the data by choosing the minimum @xmath75 over a range of reasonable possibilities .
thus , for example , if the data is dominated by the ground state for many time slices , then many time slices will be automatically added before an attempt is made to fit the first - excited state .
we will refer to this improvement as the `` variable @xmath52 '' approach , and to the basic template described in this section as the `` fixed @xmath52 '' approach .
the @xmath85 minimization can be cast more generally as minimizing a functional @xmath86 of the vector @xmath24 ( see , for example , @xcite ) . in the extreme case , with more unknown parameters than data points
, the minimization procedure becomes degenerate and there is enough freedom to drive @xmath85 down to unreasonably small values . with more data points than unknown parameters , although the agreement with the data typically is good ( often too good to believe ) ,
the solution is unstable and often wildly oscillating .
this signals a proximity to the degeneracy of the underlying minimization problem . constrained
curve fitting can be cast more generally as the problem of minimizing @xmath87 @xmath88 if @xmath89 is degenerate , but @xmath90 is not , the degeneracy is lifted .
the stability is improved in general .
the solution @xmath91 varies along a `` trade - off curve '' ( see , for example , @xcite ) .
@xmath92 ( e.g. @xmath0 ) measures the agreement of the model to the data , while @xmath93 ( e.g. @xmath94 ) plays the role of a stabilizing functional .
the standard constrained - curve fitting procedure selects @xmath95 .
to estimate the systematic errors associated with the choice of priors , it is common to repeat the procedure with all @xmath84 replaced by @xmath96 where @xmath97 is a typical choice .
this moves one along the trade - off curve from @xmath95 to @xmath98 , giving less weight to the priors .
more generally , one could have an independent weighting factor , @xmath99 , multiplying the @xmath100 of each prior . for concreteness , we return to the simple case of a sum of decaying exponentials of eqs .
[ exp ] , [ chisquare ] @xmath101 by choosing @xmath102 , one relaxes the constraint of the priors and thus tests the sensitivity of the fitted values to the choice of prior .
we present three ways , in order of increasing sophistication , to implement the choice of each @xmath103 . 1 .
_ global static weight : _ the canonical approach is to keep it unchanged as a global constant for the fit , @xmath104 .
the variation of the dependence of the output fitted results on this global input parameter may be used to estimate the systematic error .
local variable weight : _ allow each @xmath103 to be a local variable . at each step ,
loop on various values of @xmath103 and choose the value which minimizes the @xmath75 .
monitor against runaway solutions .
+ plot of ground - state pion mass , @xmath105 ( left ) and spectral weight @xmath106 ( right ) , versus global static weight , @xmath107 , for bare quark mass @xmath108.,title="fig : " ] plot of ground - state pion mass , @xmath105 ( left ) and spectral weight @xmath106 ( right ) , versus global static weight , @xmath107 , for bare quark mass @xmath108.,title="fig : " ] + same as in fig .
[ fig : pionlambdatest ] but for the nucleon for bare quark mass @xmath109 ( @xmath110).,title="fig : " ] same as in fig .
[ fig : pionlambdatest ] but for the nucleon for bare quark mass @xmath109 ( @xmath110).,title="fig : " ] 3 .
_ local dynamical weight : _ for some quantities , such as for the ground - state mass and spectral weight of the pion displayed in fig .
[ fig : pionlambdatest ] , the dependence of the fitted results on the value of @xmath107 is quite mild . for others , such as for the nucleon of fig .
[ fig : nucleonlambdatest ] , the dependence is stronger .
accordingly , we propose an adaptive procedure of absorbing any systematic error associated with the choice of @xmath111 into the statistical error , by upgrading each @xmath99 to a dynamical variable to be determined by the fit
. it is incorporated as a dynamical fit variable by further augmenting the @xmath85 : @xmath112 the choice of @xmath113 and @xmath114 is also somewhat arbitrary , and results in some uncertainty in assessing the systematic error ; however , the dependence on these parameters is gentler than on the corresponding static global values and any remaining systematic error is masked by statistical error .
global dynamical weight : _ in our fits presented here , we made the simplest choice of a special case of ( c ) : we took all @xmath103 for different parameters to be the same global @xmath111 , and took @xmath115 and @xmath116 .
+ often , the dependence of the results of the fit on the global static weight is sufficiently smooth that the use of global dynamical weight is unnecessary .
a way to mitigate the potentially aggressive nature of the fully - constrained fit in the sequential empirical bayes method is the following : when one is interested in estimating a particular parameter ( i.e. the mass or weight of the ground state or @xmath49 excited state ) one takes the priors from the previous method and does a partially constrained fit : all other parameters are constrained except the one in question which is unconstrained ( or perhaps only very lightly constrained ) .
we say that the parameter to be estimated is `` released '' ( from the constraint ) .
the statistical error estimate obtained this way is larger and a more conservative choice
. we will , by default , release the constraint for all of the final results presented . referring back to our algorithm ,
if a prior is available , we use it as the value of the initial guess .
but as a new parameter is introduced there is no prior , and an initial guess must be obtained by some other criterion . in practice
, it may happen that different initial guesses lead to different values for the fit parameters , especially for the unconstrained fit .
this may signal the presence of multiple local minima in the complicated @xmath85 topography . to address this
we introduce and advocate the use of `` scanning '' over a grid of reasonable initial values and selecting that choice which leads to a descent to the lowest @xmath85 . by extending the scan to a rather large region of parameter space ,
we ensure that the initial guesses for the parameters will not confine the @xmath85 minimization search to a single basin of local attraction , but rather will extend over many , including that of the global minimum .
recovery of parameters from artificially constructed data.,title="fig : " ] recovery of parameters from artificially constructed data.,title="fig : " ] as there are a lot of ingredients to our modification of the standard constrained fit method ( `` onion - shell data '' , `` scanning '' , `` global dynamical weight '' , `` releasing the constraint '' ) it is comforting to learn that the method can successfully reconstruct the parameters of artificially - constructed data where the true results are known independently of the fit .
we created a sample of artificial data as a sum of five decaying exponentials , with means for masses and weights fixed at values close to those extracted from real data .
then , for simplicity , given this function , @xmath117 , we added an independent gaussian noise at each value of @xmath15 . when run through our fitting procedure
, we were able to reconstruct the masses and weights for the ground state , first - excited state , and second - excited state ( see fig .
[ fig : fakedata ] ) ; for these , the actual values were within one measured standard deviation of the measured masses .
test of partitioning the data .
ground and first excited state pion mass , @xmath105 , as a function of the bare quark mass @xmath118 . ]
we have constructed an automated and natural way of obtaining the priors from a subset of the data . however , one worries that there is a significant amount of `` data snooping '' which may make it difficult to estimate the systematic errors .
( strictly speaking , using any of the data to obtain the priors violates the bayesian approach . ) to alleviate these worries , we have implemented the following test : we partition the data into two non - intersecting sets of configurations , @xmath119 and @xmath120 , with an equal number , @xmath121 of configurations in each set ( which must still be large enough to permit stable covariant fits ) . using the set @xmath119 of configurations
, we perform our procedure outlined above of obtaining the priors gradually , from the ground state up , as the data set is monotonically enlarged by including earlier and earlier time slices . now , regarding this entire procedure as a black box solely for the purposes of obtaining the priors , we next use this fixed set of priors in the canonical way @xcite to perform a constrained fit separately on data set @xmath120 ( for which there is no data snooping ) , on data set @xmath119 ( maximal data snooping ) , and on the full set @xmath122 ( partial data snooping but with greater statistics ) .
any disagreement beyond statistical errors can help assess systematic errors due to data snooping .
in our case we found no appreciable differences beyond expected statistical fluctuations .
figure [ fig : partition ] shows a plot of the ground and first excited state pion mass , @xmath105 , as a function of the bare quark mass @xmath118 .
however priors are obtained , they can then be used in a standard constrained curve fitting procedure , wherein the fit window @xmath13@xmath56 is held fixed ( and as big as feasible ) , while the number of terms in the fit model is increased one - by - one until the fit results converge for the lowest few parameters of interest .
to test whether the sequential empirical bayes method has produced reliable priors , we subsequently use its priors in the standard way @xcite as described above .
figure [ fig : stability ] illustrates the passing of this test .
in a test of stability , priors had previously been selected by the sequential empirical bayes method .
now they are used in the standard constrained fit .
the figure shows fit values for the lowest two masses from constrained fits ( for all @xmath15 s ) with different numbers of terms in the fit model .
the data is from the pion two - point correlation function @xmath123 at bare quark mass @xmath109 ( @xmath110 ) . ]
given the posterior probability distribution , @xmath31 , all statistics may be evaluated by computing integrals . in practice , such evaluation is often difficult to perform directly ; monte carlo methods such as simulated annealing @xcite can be used .
even so , the computational cost is often daunting and so approximations are sought .
most commonly , one may assume that the data set is sufficiently large that the central limit theorem implies that @xmath124 is approximately quadratic about its minimum .
we in fact use this approximation , and furthermore use the resulting fitting errors from a fit as the priors @xmath125 for the next fit .
to test this , we plot in fig .
[ fig : fittingerrors ] the @xmath124 as a function of @xmath24 in the neighborhood of the minimum .
superimposed is the parabola which is the quadratic approximation about the minimum .
the two agree very well . also calculated and plotted as ranges are the `` fit error '' , obtained from the quadratic approximation , and the second moment of the actual distribution .
these too agree very well , and indicate that using the naive fit error as provided by the minimization routine is quite adequate .
although fit errors are used to produce priors for the inner loops of the algorithm , all final errors reported are bootstrap errors .
plot of the augmented @xmath124 in the vicinity of its minimum .
it agrees well with a quadratic approximation , ensuring that the fit error provided by the minimization routine is adequate .
the data is from the pion two - point correlation function at bare quark mass @xmath108 ( @xmath126 ) . ]
it is crucial to be extremely conservative when making fits , especially when relying on bayes inference methods .
one must be sure to avoid announcing a `` false positive '' , that is making claim which can not be guaranteed to be correct .
we can understand in more detail how a false positive may emerge and how to design the procedure to avoid them , by testing variants of the seb algorithm against a number of `` toy models '' where the actual values of the function s parameters are known .
suppose the artificial data is created with the three - state model function @xmath127 and we want to perform a fit of the data sample generated by this function with fixed values of input parameters and gaussian statistical error @xmath128 at each time slice @xmath15
. then seb should work if the following holds : 1 .
there exists a @xmath129 such that in the time range @xmath130 , @xmath131 so that in this range , the data can be fitted by @xmath132 .
that is , there is a `` plateau '' in the effective mass plot for large time . 2 .
there exists a @xmath133 such that in the time range @xmath134 , @xmath135 so that in this range , the data can be well fitted by @xmath136 .
3 . in the time range @xmath137
, the third state becomes important , and the full three - state model must be used in the fit .
now the seb is very adaptive ; in the course of the iterative procedure , the ground state is fitted several times and allowed to float within a sigma or so of its current prior with each new fit .
thus the original criterion of the existence of a clear plateau need not be strictly enforced .
but if the condition for the plateau is badly violated , or if the similar conditions for the excited states are badly violated , there is a possible danger of the algorithm gravitating toward a local minimum in the @xmath0 which is not the true value . to illustrate , consider artificial data , constructed from the following three - state toy model @xmath138 in the time range @xmath139 , with relative errors ( uncorrelated and gaussian ) increasing with time @xmath15 to mimic actual lgt data .
in keeping with the spirit of the seb , we fit first to a single exponential over the time interval @xmath140 $ ] , then use the fitted ground state values as priors for a two - state partially - constrained fit in the time interval @xmath141 $ ] , and then use the fitted ground and first - excited state fit values as priors for a final three - state partially - constrained fit in the time interval @xmath142 $ ] , where @xmath143 . but to stress the caveat about the dangers of `` false positives '' , we explore all values for the pair @xmath144 .
[ [ lowest - precision - case ] ] `` lowest precision '' case : + + + + + + + + + + + + + + + + + + + + + + + + + + with statistical errors ranging from @xmath145 at @xmath146 , to @xmath147 at @xmath148 , the lowest @xmath149 fit is for @xmath150 with @xmath151 , @xmath152 , @xmath153 , @xmath154 , @xmath155 , @xmath156 , that is , the fit fails to produce the input parameters of the toy model .
[ [ low - precision - case ] ] `` low precision '' case : + + + + + + + + + + + + + + + + + + + + + + + if the gaussian noise is reduced by about a factor of two to @xmath157 at @xmath146 , @xmath158 at @xmath159 , then there are two solutions with the lowest @xmath160 : a false positive with @xmath161 and @xmath162 , @xmath163 , @xmath164 , @xmath165 , @xmath166 , @xmath167 , and a fit which agrees with the input model : @xmath168 with @xmath169 , @xmath170 , @xmath171 , @xmath172 , @xmath173 , @xmath174 .
[ [ high - precision - case ] ] `` high precision '' case : + + + + + + + + + + + + + + + + + + + + + + + + if the gaussian noise is further reduced to @xmath175 at @xmath146 , @xmath176 at @xmath177 then the lowest @xmath178 does reproduce the input parameters : @xmath179 , @xmath180 , @xmath181 , @xmath182 , @xmath183 , @xmath184 , @xmath185 .
let @xmath186 be the absolute value of the difference of the function @xmath187 of input parameters and the function @xmath188 of fitted parameters @xmath189 a fit @xmath188 with reasonable small @xmath64 implies that the relation @xmath190 roughly holds at most times @xmath15 . in other words , with this statistical error @xmath191 , we can not distinguish the fitted function @xmath188 from the original @xmath187 . the relative errors of three data sets of increasing precision ( `` lowest '' , `` low '' , and `` high '' ) are plotted with points .
the curved solid ( black ) curve is the plot of the function @xmath192 , which is the relative difference between original function @xmath193 and the false - positive fitted function @xmath194 from the data set ( `` lowest precision '' ) . ] from fig .
[ fig : relativeerror ] we see that the statistical errors of the data set ( `` lowest precision '' ) are all larger than the difference of the two functions in the full time range .
this is the reason why the false positive can be chosen by the fit routine .
the dashed ( blue ) straight line ( @xmath195 ) and the dotted ( red ) straight line ( @xmath196 ) intersect with the curve of relative errors at @xmath129 and @xmath133 , respectively .
this shows that the ground state dominates in the range @xmath197 , the first - excited state plays a role in the range @xmath198 , and the three states should all be included in the time range @xmath199 . when the precision is higher , the statistical errors are smaller than the difference of two functions , in some time range , so that the fit can reject the false positive @xmath200 . returning to the false positive of the `` lowest precision '' case ,
notice that when forced to fit to a three - state model , the fit prefers a two - state solution with the one weight essentially zero , and the remaining weight consistent with the sum of the true second and third weights and the remaining mass interpolating between the true masses .
in short , the fit is content with averaging the two higher states into one .
also notice that it is the second weight which is zero .
this is because , at the intermediate stage , a two - state model was forcing a fit to data completely dominated by a single state .
subsequently , as earlier times were added , the data could be described by two states , but since the would - be second state weight was by now constrained near zero , its role was supplanted by the third state in the model . that is , by introducing the new ( second ) state into the model before the data was precise enough to discern it , spurious results were obtained .
this suggests the cure : postpone the introduction of new states in the model until the data demands it ( as evidenced by a sudden increase in the @xmath149 ) . in more detail ,
if the time range is large enough there exists a largest time @xmath129 beyond which the contamination of higher states will be obscured by the statistical errors and can be neglected so that in the time range @xmath197 the ground state dominates and the data can be well described by the single exponential @xmath201 . thus the time range @xmath197 is a good choice for the first step of seb , and we can get a reasonably small @xmath64 over this ground - state effective - mass plateau . if we include more time slices with @xmath134 , the contribution of the first excited state becomes significant , but the second excited state contributes little to the correlation function in this range .
the presence of the first excited state results in a noticeable increase of the @xmath64 if we continue to force a single exponential as a fit model ; the inclusion of one more term in the fit model is necessary .
continue this procedure until the time - slices are exhausted . _ the penultimate algorithm : `` not - too - soon '' _ 1 .
the available data are in the time range @xmath202 .
2 . from effective mass plots , choose an initial time range @xmath203 $ ] to do an unconstrained one - mass fit .
use `` scanning '' see sect .
[ sect : scanning ] .
3 . include one more time slice and repeat an ( independent ) unconstrained one - mass fit , and monitor the fitted parameters @xmath204 and the @xmath64 .
4 . if the fitted parameters and the @xmath64 do not change much , include one more time slice and repeat the previous step .
this iteration stops if there is a noticeable change of @xmath64 and the values of the fitted parameters indicating a breakdown of the one - state model .
then set @xmath205 equal to the time at which the @xmath64 jumps , indicating the necessity of a two - mass fit for @xmath206 .
set the priors for the ground state mass and weight equal to the fitted values from the last low@xmath64 fit over @xmath207 $ ] .
5 . include one more mass term in the fit model and do a partially - constrained two - mass fit ( with scanning ) for @xmath208 $ ] .
the ground state priors are fixed at the values determined by the previous step .
the first - excited state is unconstrained but scanned .
repeat , adding time slices until the two - state model breaks down as indicated by a jump in the @xmath64 . then set @xmath209 equal to the time at which the @xmath64 jumps , indicating the necessity of a three - mass fit for @xmath199 . set the priors for the ground state and first - excited state mass and weight equal to the fitted values from the last low@xmath64 fit over @xmath210 $ ] .
note that the ground - state priors are refreshed .
repeat , adding more time slices and more fit - model terms ( one at a time and only when necessary ) and more until all time slices are used . 8 .
the highest state in the fit model will be absorbing all the contributions from higher states in the true function , and thus its fitted parameters will differ from the true values .
thus the highest state in the fit model must be rejected .
now we return to the artificial data from the toy model `` lowest precision '' . recall that when we assume a three - mass fit and independently try all pairs @xmath144
, we obtained a false positive .
but the statistical errors are so large that they obscure the contributions from excited states at all but the earliest time slices . adding terms to the fit model
prematurely led to spurious results .
now we apply the new method outlined immediately above to the same data .
a one - state fit works well for @xmath211 $ ] until the @xmath0 suddenly jumps at the ( proposed value of ) @xmath212 .
thus we set @xmath213 .
the values of the one - mass term fit over time range @xmath214 $ ] are used to set priors for @xmath215 and @xmath216 , and a second term is added to the fit model .
the two - term model works well for a fit in the range @xmath217 $ ] with @xmath133 varying from 3 to 0 .
the two - mass model fit the data very well ( @xmath218 ) in the whole time range with fitted parameters @xmath219 , @xmath220 , @xmath221 , @xmath222 .
since the two - mass fit exhausts all the time slices and does not show a noticeable increase of @xmath64 , we do nt go ahead to include the third mass term in the fit model .
but this fit is a `` false positive '' ! since the statistical errors are larger than the difference between the false positive and the true solution ( eq . [ deltag ] ) , the false positive
can not be rejected on the basis of its @xmath64 , by this or any other method .
so what have we gained ?
the difference this time is that this solution is exposed as a two - mass - term fit rather than a spurious three - term fit . and
as always , the highest state in the fit must routinely be dropped since in general it can be contaminated by contributions from higher excited states .
we conclude that with the `` lowest precision '' data set , we can not get an unambiguous estimate of the second and the third state .
the fit is comfortable predicting the ground state parameters only ( and they are correct ) .
thus the algorithm has not made a `` false positive '' claim .
now , we illustrate this technique with another set of artificial data for which it will be possible to extract excited states . it will be instructive to monitor at the behavior of the @xmath64 , but also to see how the fit parameters of each term in the fit model stabilize as time slices are added .
[ fig : toychisq1 ] ( left ) plots the @xmath64 for data artificially constructed as a sum of four decaying exponentials .
a one - term fit is adequate for time slice 13 ( @xmath223 ) .
but it is exposed as being entirely inadequate as time slice 12 is added to the data set to be fit .
( the @xmath64 jumps tremendously because we have made the statistical errors on the data to be fit quite small , to illustrate the point . )
so for time slices 6 - 12 a two - term fit model is used , and the fit is quite adequate . at time slice 5 , however , the @xmath64 jumps tremendously , exposing the two - term fit as inadequate .
thus a third term is added at time slice 5 .
the three - term fit works down through time slice 2 . at time slice 1
, a fourth term must be added to keep the @xmath64 from being unreasonably large .
behavior of the @xmath64 of the fit as earlier time slices are added to the fit range .
a sudden jump indicates that the fit model is inadequate and that a new mass term should be added to the model .
accordingly , then the @xmath64 drops only to increase again as further time slices are added .
`` not - too - soon '' addition of terms to the fit model prevents a single state being erroneously identified as two .
also shown are the behavior of the ground and first few excited states , which stabilize as more terms are added .
, title="fig : " ] behavior of the @xmath64 of the fit as earlier time slices are added to the fit range .
a sudden jump indicates that the fit model is inadequate and that a new mass term should be added to the model .
accordingly , then the @xmath64 drops only to increase again as further time slices are added .
`` not - too - soon '' addition of terms to the fit model prevents a single state being erroneously identified as two .
also shown are the behavior of the ground and first few excited states , which stabilize as more terms are added .
, title="fig : " ] meanwhile , fig .
[ fig : toychisq1 ] ( right ) monitors the masses of the lowest few states as more time slices are added to the data set and more terms are added ( as necessary ) to the fit model . at time slice 13 ,
the one - term model is fitted with a ground - state mass ( open circle ) .
( its value is consistent with the value obtained much later when time slices 112 have been added to the fit model and three more terms are added to the fit model .
that is , time slices above 13 are sufficient to determine the ground - state mass , i.e the effective mass plot has `` plateaued '' .
but more importantly for the efficacy of the algorithm , this fitted value remains stable later as the data and fit model are expanded . ) at time slice 12
, the second term is added to the fit model ( solid diamonds ) . as each earlier time
is added , the first - excited state mass changes by a sigma or so , as the augmented data is refit .
this is important .
it is saying that if a relatively small amount of data from time slices 12 and above suggested a somewhat inaccurate guess for the prior for this mass , then subsequent fits including data at time slices 8 - 11 can influence the value and bring it toward the correct value ( at the horizontal line ) . by time slice 8 ,
the first excited state mass has stabilized ( at the correct value ) and will not subsequently deviate through time slice 6 . at time slice 5 ,
the first - excited state mass deviates from the correct answer .
but this is where the jump in the @xmath64 warns that the two - state fit model is inadequate ; the first - excited state fitted mass is contaminated from higher states in the data . adding a third term returns the fitted first - excited state mass to its correct value .
it remains stably at this value with the addition of earlier time slices 24 to the three - state fit model ( open triangles ) . before the second - excited state has stabilized , we have run out of earlier time slices to add , and we can make no claim as to whether its fitted value is correct .
( it is nt . ) but the algorithm has successfully fitted the first - excited state ( and of course the ground state ) .
same as in fig .
[ fig : toychisq1 ] but the artificial data is less precise ( 1% ) .
accordingly , the @xmath64 does not jump to such extreme values when the number of the terms in the fit model becomes inadequate.,title="fig : " ] same as in fig .
[ fig : toychisq1 ] but the artificial data is less precise ( 1% ) .
accordingly , the @xmath64 does not jump to such extreme values when the number of the terms in the fit model becomes inadequate.,title="fig : " ] figs .
[ fig : toychisq2 ] and [ fig : toychisq3 ] tell a similar story for artificial data which is less precise ( 1% and 5% respectively ) . the @xmath64 jump ( indicating that another term should be added to the fit model ) but , as one would expect , not as extremely as for the more precise data of figs .
[ fig : toychisq1 ] .
same as in fig .
[ fig : toychisq1 ] but the artificial data is even less precise ( 5%).,title="fig : " ] same as in fig .
[ fig : toychisq1 ] but the artificial data is even less precise ( 5%).,title="fig : " ] as we saw in the last section , the `` not - too - soon '' algorithm avoids `` false positives '' by delaying the addition of a new term to the fit model before the data were able to discern it .
however , there is the logical possibility of the opposite danger : if the new term is added too late , then there could potentially be an erroneous @xmath224-state fit of data which is in fact better described by @xmath225 states .
if so , then the highest two states will be `` averaged '' into one .
( so , for example , a fitted first excited state mass might take a value intermediate between the true first and second excited state masses . )
the addition of further states to the fit model might then cover up this misidentification , and allow for this different kind of false positive .
one would think that an increased @xmath64 would warn of this danger ; however , since the @xmath64 is an average over many time slices , it s warning may come a time slice or two too late . if the data is such that a new term is discernible every couple of slices , then this may indeed lead to erroneous results .
ratio of the contribution to the fit model from a proposed new term , @xmath226 , compared to the statistical error @xmath227 ) at the newest time slice @xmath228 added to the data set .
the new term is deemed not needed at time slices @xmath229 and @xmath230 . at time slice @xmath231 , the ratio exceeds @xmath232 and the new term is conditionally accepted . the ratio remains above @xmath232 for time slice @xmath233 and below , confirming that the new term is needed .
the data is from a fit of the roper resonance at the same quark mass as in fig .
[ fig : roper_stability_previous ] . ] a final tweak of the algorithm addresses this issue : the `` just - in - time '' algorithm is the same as the `` not - too - soon '' algorithm outlined in the last section , except that the previous criterion for adding a new term to the fit model , namely that without this new term the @xmath64 would suddenly increase , is replaced by a more sensitive criterion : as a new time slice @xmath228 is added to the data set , a tentative fit is made with a proposed extra term , \{@xmath234,@xmath235}. then the contribution of this fitted term is compared to the statistical error at the new time slice @xmath236 ( see fig .
[ fig : relativecontribution ] . ) if eq .
[ acceptcriterion ] does not hold , then the new term is deemed not needed .
if it does hold , then the new term is accepted ( but only conditionally since the equation may hold only by statistical fluctuation ) .
as new time slices are added , the new term must continue to pass this test .
( this is stable : typically , if it passes the test once , it is likely to continue to pass the test at earlier time slices , since for these the relative error gets smaller and the fractional contribution gets larger . )
if the test is failed before a higher - order term is conditionally added , then one returns to time slice @xmath228 and it is deemed that the @xmath224th term is not included in the fit . then time @xmath237 is added to the data , a new proposal is made to add the @xmath224th term , and the process continues .
we have presented a list of algorithms with steadily increasing complexity but with steadily increasing robustness .
the earliest and simplest `` fixed @xmath238 '' algorithm , most easily described and presented first here as a template , was originally used ( prematurely in retrospect ) for some preliminary conference presentations .
we now deem it too unsophisticated and do not use it anymore .
the plain vanilla `` variable @xmath52 '' algorithm is adequate provided that the level spacings are well separated and each term is saturated by several time slices .
it is perfectly adequate to describe , for instance , the pion ground and excited state , as explained in sect .
[ sect : pion ] .
the plain `` variable @xmath52 '' was used in the first version of our roper paper and caused some erroneous results at high quark mass .
this has been rectified by the use of the `` just - in - time '' algorithm for the updated version of our roper paper and is briefly described in sect . [
sect : roper ] . on a @xmath239 lattice ,
we use the overlap fermion @xcite and the iwasaki gauge action @xcite with @xmath240 .
the lattice spacing , determined from the measured pion decay constant @xmath241 , is determined to be @xmath242 , and thus the lattice has spatial size of @xmath243 .
we adopt the following form for the massive dirac operator @xcite @xmath244 where @xmath245 so that @xmath246 where @xmath247 is the matrix sign function and @xmath248 is taken to be the hermitian wilson - dirac operator , i.e. @xmath249 . here
@xmath250 is the usual wilson fermion operator , except with a negative mass parameter @xmath251 in which @xmath252 .
we take @xmath253 in our calculation which corresponds to @xmath254 .
the massive overlap action is defined so that the tree - level renormalization of mass and wavefunction is unity .
we adopt the zolatarev implementation @xcite of the optimal rational approximation @xcite to approximate the matrix sign function .
the inversion of the quark matrix involves nested do loops in this approximation .
further details of the procedure are given elsewhere @xcite .
two - point correlation function @xmath255 for the pion for three bare quark masses . ]
figure [ fig : picor ] shows the correlation function for the correlator @xmath123 .
one can see that it is dominated by the ground state of the pseudoscalar channel ( pion ) over all but the few earliest time slices .
this presents a problem for the default `` fixed @xmath52 '' approach . referring back to the algorithm of section [ sect : basicalgorithm ] ,
we see that at each step of the algorithm , @xmath52 new time slices are added to the data while a new excited state is added to the fit model .
thus for the pion correlator , one would be trying to fit to a model with many states when the data is saturated by just the ground state . forcing
a fit may give misleading estimates of excited state parameters which are subsequently used as priors .
with the `` variable @xmath52 '' refinement of the algorithm , rather than deciding _ a priori _ on the number of terms in the fit and adding time slices a fixed number at a time , one lets the data decide how many time slices to include with each enlargement of the data by choosing the minimum @xmath85 over a range of reasonable possibilities .
thus since the pion correlator is dominated by the ground state for many time slices , then many time slices will be automatically added before an attempt is made to fit the first - excited state .
in fact , we find that the `` variable @xmath52 '' method works very well for the pion correlators . figure [ fig : pionmass ] shows the results of the fit , the ground- and first - excited - state weight and mass as a function of the bare quark mass .
ground and first - excited state pion mass @xmath105 , as a function of the bare quark mass @xmath118 ( left ) . ground and
first - excited stated pion weight , as a function of the bare quark mass @xmath118 ( right ) .
notice that the ground state weight does not diverge as the quark mass approaches zero for this @xmath123 correlator as it would for the @xmath256 correlator , where @xmath257 is the pseudoscalar density.,title="fig : " ] ground and first - excited state pion mass @xmath105 , as a function of the bare quark mass @xmath118 ( left ) .
ground and first - excited stated pion weight , as a function of the bare quark mass @xmath118 ( right ) .
notice that the ground state weight does not diverge as the quark mass approaches zero for this @xmath123 correlator as it would for the @xmath256 correlator , where @xmath257 is the pseudoscalar density.,title="fig : " ] studies using standard curve fitting have heretofore failed to satisfactorily identify the `` roper resonance '' @xmath258 of the nucleon on the lattice @xcite .
our analysis is the first lattice calculation to obtain the masses of the roper and @xmath259 at low quark masses well in the chiral regime ( with a pion mass as low as @xmath260 ) .
however , the effects of quenched artifacts , specifically the presence of ghost @xmath261 states , complicates the physics as well as the functional form of fit model , and so the details of the calculation are presented elsewhere @xcite . to be sure , our calculation benefits from going to unprecedented low quark mass with the full chiral symmetry provided by overlap fermions , but the sequential empirical bayes method plays a crucial role .
it also presented a challenge and led to the refinement of the seb from the `` variable @xmath52 '' version to the `` just - in - time '' version .
the first version of our roper results @xcite used the `` variable @xmath238 '' method which , as we ve seen , had until then been perfectly adequate to expose excited states ( such as for the pion in sec . [
sect : pion ] ) where the density of excited states is not too high . compared to our final analysis with the `` just - in - time '' version @xcite , the results for the nucleon and @xmath262 did not change .
the only change is for the roper state at medium and heavy quark masses .
indeed our first results at medium - heavy masses passed the `` stability plot '' test ( fig .
[ fig : roper_stability_previous ] ) and our focus returned to the more interesting light quark mass regime . here
the `` variable @xmath52 '' criterion for when to add new states remains perfectly adequate for pion masses below @xmath263 .
stability test for the roper resonance at bare quark mass @xmath264 ( @xmath265 ) .
priors had previously been selected by the plain `` variable @xmath52 '' version of the sequential empirical bayes method .
now they are used in the standard constrained fit .
the figure shows fit values for the lowest two masses ( nucleon and roper ) from constrained fits with different numbers of terms in the fit model .
( at this mass , ghost states make no contribution , and so the fit model contains only exponentials . ) ] subsequently , we came to realize that the `` variable @xmath52 '' criterion that we adopted to introduce the excited state was not adequate for medium and heavy quarks , although it is adequate for the light quarks . in fitting someone else s proprietary charmonium data
, we realized that the crucial difference between the heavy quark spectrum and that of the light is that the ratio of the excited state mass to that of the ground state is much smaller in the heavy system , that is , the excitation spectrum in the heavy quark system is more dense .
the earlier criterion erroneously resulted in obtaining the average of the higher excited states as the first excited state for the heavy system . in view of this
, we devised the new and final `` just - in - time '' criterion ( sec .
[ sect : finalalgorithm ] ) , wherein a new term is added if its contribution is statistically significant at the current time slice .
in fact , the penultimate `` not - too - soon '' algorithm ( sec . [ sect : finalalgorithm ] ) , where the criterion for adding a new term to the fit model depends on a sudden jump in the @xmath75 , works just as well as `` just - in - time '' for the roper resonance at both medium - high and low quark masses .
but for some charmonium data , or other test data constructed with a dense excited - state spectrum , `` just - in - time '' is somewhat better than `` not - to - soon '' , and because of its potential theoretical advantages is our method of choice .
[ fig : roper_stability ] shows that our final algorithm passes the `` stability test '' .
however , so did the earlier `` variable @xmath52 '' algorithm .
this and the fact that the extracted excited state mass changed alarmingly means that the `` stability test '' @xcite is only a necessary but not a sufficient test for the reliability of constrained - curve fitting .
we emphasize , however , that the source of the discrepancy has been identified and cured ; the final algorithm can robustly protect against the averaging of excited states .
same as for fig .
[ fig : roper_stability_previous ] but for our best and final `` just - in - time '' algorithm . ] in the quenched approximation , there are artifacts associated with the absence of quark loops .
one of the more interesting consequences is that the would - be @xmath266 propagator involves only double @xmath111 poles in `` hairpin diagrams '' .
these lead to the chiral - log terms contributing to hadron masses ; we see these clearly in our recent lattice calculation of pion and nucleon masses @xcite .
another quenched artifact is the contribution of ghost states in the hadron propagators as first seen in the @xmath1 meson channel @xcite where the ghost @xmath267-wave @xmath268 state lies lower in mass than the @xmath1 ( for sufficiently small quark mass ) .
we have seen a similar effect in analysis of the excited nucleon spectrum where a @xmath257-wave @xmath261 lies close in mass to the roper resonance @xcite , and its presence must be carefully disentangled by the fitting code .
more dramatically , in the negative - parity channel @xmath262 ( @xmath269 ) , the lowest @xmath267-wave @xmath261 state has a mass lower than that of the @xmath262 for small quark mass .
since the @xmath111@xmath111 coupling in the hairpin diagram is negative @xcite , the @xmath262 correlator changes sign with increasing time separation @xcite .
this effect is only seen at small enough quark masses , and thus is not seen in most lattice simulations at much higher masses . as a result ,
the form of the fitting function changes ; there are extra non - exponential terms .
nevertheless , the seb can successfully handle this situation .
it is crucial to enforce the physical constraint that the weight of the ghost state be negative ; otherwise there would be no way to distinguish it from the physical state ( e.g. roper ) lying nearby .
the details of the fitting model and the results of the fitting ( using the `` just - in - time '' seb algorithm ) for the roper and @xmath262 are in @xcite .
@xmath1 correlators for our lowest quark mass for which @xmath270 ( left ) and @xmath271 ( right ) .
the negative dip of the correlators is an indication of the domination of the ghost @xmath267-wave @xmath268 state over the physical @xmath272 .
the curves are contributions to the fit model and are explained in the text.,title="fig : " ] @xmath1 correlators for our lowest quark mass for which @xmath270 ( left ) and @xmath271 ( right ) .
the negative dip of the correlators is an indication of the domination of the ghost @xmath267-wave @xmath268 state over the physical @xmath272 .
the curves are contributions to the fit model and are explained in the text.,title="fig : " ] another example is displayed in fig .
[ fig : a0_prop_1 ] where we show the @xmath1 propagator at two low quark masses , for which @xmath270 ( left ) and @xmath271 ( right ) .
the most dramatic feature of the data in each figure is the negative dip of the correlator at time slice 3 and above .
the solid ( red ) line passing through the data is the result of the seb fit .
the two dominant contributions to the fit model are negative and are indicated by the dashed ( green ) line labeled `` term 1 '' and by the solid ( cyan ) curve labeled `` term 2 '' .
they are modeled by the expression @xmath273 where @xmath216 is constrained to be negative and the @xmath274 factor reflects the double - pole nature of the hairpin diagram @xcite .
we fit @xmath275 , the mass of the interacting would - be @xmath266 and @xmath276 state .
since we work in a finite box , the @xmath277 states are discrete and they are constrained to be near the energy of the two non - interacting particles , each with @xmath278 for discrete lattice momentum @xmath279 where @xmath224 is an integer . for `` term 1 '' , @xmath280 and for `` term 2 '' , @xmath281 .
the contribution of the physical state @xmath1 is positive and is indicated by dotted magenta curve in each figure . in summary ,
our seb algorithm is quite capable of handling non - standard forms of the fit model including negative - normed ghost - state contributions and can fit successfully the data .
we have advocated refinements of bayesian - inspired constrained - curve fitting which we believe better stabilizes fits at low quark mass .
this permits analysis where the values of fit parameters are not reliably known _ a priori _ , such that their use as priors might be dangerously biased . in the `` sequential empirical bayes method '' ,
we have constructed an automated and natural way of reliably obtaining the priors from naturally - nested subsets of the data ( `` onion shells '' ) . for the prototype described here , a time - dependent two - point correlation function @xmath117
, we first obtain estimates of the ground state mass and weight from unconstrained fits to a subset of data restricted to large times .
these are then used as the priors in a subsequent constrained fit .
a sequence of ( fully or partially ) constrained fits follow . at each step , more data from adjacent earlier times are added and another term is added to the fit model .
when a new parameter is added , its first estimate is obtained from a fit in which it is unconstrained but previously introduced parameters are constrained .
after each fit , the fitted values of each parameter are used as the priors for the next fit .
thus the value of each prior is free to float from one fit to the next , and thus is not prevented from changing significantly over the course of the several steps of the iteration as more and more data are added .
this mitigates any bias which may have been introduced wherein the first estimates of a prior are inaccurate due to statistical fluctuations in the smaller initial data set . in the final step , as the last data are added , no new terms are added to the fit model . a final fit for each parameter leaves it unconstrained ( or very weakly constrained ) with all other parameters constrained by priors set equal to the latest fitted values .
our final bootstrapped errors obtained from `` releasing the constraint '' in this way are larger than for a completely - constrained fit and are selected as a conservative estimate .
we have found some further refinements to be fruitful : ( a ) we use a `` scanning '' technique to automate the selection of an initial value for a new parameter for its first unconstrained fit .
we thus avoid the pitfall of the @xmath85 minimization routine becoming trapped in the attraction basin of a local , but not global minimum .
( b ) the weight factor ( essentially a lagrange multiplier ) , @xmath107 , balances the influence of the data versus the priors in determining the output of a fit .
decreasing @xmath107 from its canonical bayesian value of 1 , can be used to assess the systematic error associated with the choice of prior(s ) .
we advocate absorbing this systematic error into a statistical error by promoting @xmath107 to be `` global dynamical weight '' , that is , a fit parameter with its own prior mean and standard deviation .
we have tested that our algorithm can successful recover the correct fit parameters of an artificial data set , constructed as a sum of decaying exponentials with realistic values of the parameters . for one channel ( corresponding to using only the local - local two - point correlation function ) , the masses and weights of the ground and first - excited states are fit to within a measured standard deviation of the true values .
thus , for our real data , we restrict ourselves to measuring only the lowest two states , even though we use four or five states in the fit model .
this is sufficient for our present purposes .
our method is not in strict accord with the bayesian philosophy , as we use subsets of the data to guide the selection of the priors .
nevertheless , the following mitigates bias from such data snooping : ( a ) the data is naturally nested in that a subset of data restricted to large times can provide a fair estimate of the lowest state parameters .
adiabatically increasing the data set with the introduction of new terms in the fit model gives fair estimates of excited - state parameters to be used as priors .
( b ) the priors are given ample opportunity to change in accordance with the data in the several steps of the algorithm .
is this enough ? to test this we `` partition the data '' by configuration into equally - sized but disjoint sets and apply our algorithm to one half to obtain priors and then use these priors in a standard constrained fit on the other half ( in strict accord with the bayesian philosophy ) and on the full data set .
we see no difference beyond expected statistical errors ; that is , no bias is seen . in further studies using
artificially constructed data with known means and errors ( `` toy models '' ) we have presented some caveats against the careless application of the algorithm which might result in a state being spuriously identified as a combination of two , leading to a `` false positive '' prediction .
these false positives were avoided in the toy model studies by insisting that a new term be added to the fit model `` just - in - time '' , that is only after a fit with fewer terms is deemed inadequate . for this pilot study , we have analyzed the ground and first - excited masses and weights for the pion from overlap fermions on a quenched @xmath3 lattice with spatial size @xmath4 and pion mass as low as @xmath282 , commented on the history of the use and suitability of variations of the seb algorithm for extracting the roper resonance @xcite , and given another example ( @xmath1 ) where the method can handle ghost states . | we introduce the `` sequential empirical bayes method '' , an adaptive constrained - curve fitting procedure for extracting reliable priors .
these are then used in standard augmented-@xmath0 fits on separate data .
this better stabilizes fits to lattice qcd overlap - fermion data at very low quark mass where _ a priori _ values are not otherwise known .
lessons learned ( including caveats limiting the scope of the method ) from studying artificial data are presented . as an illustration , from local - local two - point correlation functions ,
we obtain masses and spectral weights for ground and first - excited states of the pion , give preliminary fits for the @xmath1 where ghost states ( a quenched artifact ) must be dealt with , and elaborate on the details of fits of the roper resonance and @xmath2 previously presented elsewhere .
the data are from overlap fermions on a quenched @xmath3 lattice with spatial size @xmath4 and pion mass as low as @xmath5 . |
low - luminosity active galactic nuclei ( llagns ) , operationally defined as a h@xmath0 luminosity of @xmath1 , reside in about 40 per cent of nearby bright galaxies , and occupy majority of the population of agns @xcite .
its low luminosities are thought to be caused by very low accretion rate on to super - massive black holes .
such accretion has often been explained by the model of optically - thin disc such as an advection - dominated accretion flow ( adaf ; @xcite ) , rather than optically - thick disc ( ` standard disc ' ; ) .
it is predicted that the adaf has a ` submillimetre bump ' in its spectral - energy distribution ( sed ) , while the standard disc is responsible for a big - blue bump in the sed of luminous agns such as quasars . since a high brightness temperature of @xmath2 k in radio bands
is expected from the adaf model , the first imaging investigations for accretion discs must be promising with future very - long - baseline - interferometry ( vlbi ) instruments that will provide micro - arcsec resolutions @xcite .
therefore , the observational evidence of the adaf is crucial in the llagn radio sources .
although the adaf model successfully explains the broadband spectra of llagns , there is a significant gap between observations and the adaf spectrum especially at low - frequency radio bands ( e.g. , @xcite ) .
this indicates that additional unknown emission does exist , and putative emission from the accretion disc may be buried under it .
the submillimetre bump means a highly inverted spectrum at centimetre - to - submillimetre wavelengths of spectral index @xmath3 , where @xmath4 , @xmath5 is flux density at frequency @xmath6 @xcite
. observations for llagns have been carried out exclusively at centimetre bands where high sensitivities are available , because most of llagns are very faint radio sources .
about half of low - luminosity seyfert galaxies and low - ionization nuclear emission - line regions ( liners ; @xcite ) hold a compact radio core @xcite , and at least 25 per cent of transition objects , which are hypothesized to be liner/-nucleus composite systems @xcite , also hold a compact core @xcite , at 15 or 8.4 ghz in 0.152.5 arcsec resolutions . most of the cores show nearly flat spectra ( @xmath7 ) in 0.5-arcsec resolution : that is evidence for jet domination @xcite .
although slightly inverted spectra have been found in unresolved compact cores of several llagns in milli - arcsec resolutions of vlbis , their radio - to - x - ray luminosity ratios suggest that another significant radio component seems to contribute in addition to original fluxes from the adaf model @xcite .
to explain this radio excess , an jet adaf model has been proposed @xcite .
thus , at centimetre bands , the contamination that is believed to be from jets has prevented the adaf from being revealed . in the present paper ,
we report millimetre survey and submillimetre data analyses for many llagns .
although technically difficult , a high - frequency observation is very promising to detect the inverted spectrum of the adaf because of following two advantages : ( 1 ) spectral measurements at high frequencies are less affected by non - thermal jets , which generally show power - law spectra ( @xmath8@xmath9 ) , and ( 2 ) the adaf predicts larger flux densities at higher frequencies .
in fact , it has been obsevationally confirmed that flux densities at millimetre wavelengths are 510 times larger than those at centimetre wavelengths in sgr a * ( e.g. @xcite ) , which is the nearest low - accretion rate system from us .
however , contamination from diffuse dust may be harmful at @xmath10 ghz when we use a poor spatial resolution .
therefore , the use of a beam as small as possible and the estimations of every conceivable contamination are essential .
the present paper is structured as follows .
sample selection is described in section [ section : sample ] .
observations and data reduction are described in section [ section : obs&reduction ] . in section
[ section : results ] , we report these results and the estimations of diffuse contamination in our millimetre measurements .
our results of submillimetre photometry are utilized only for estimation for dust contribution .
the origins of the spectra are discussed in section [ section : discussion ] from the relation between the high - frequency radio spectrum and low - frequency radio core power .
finally , we summarize in section [ section : summary ] .
our sample , ` vlbi - detected llagn sample ' , covers 20 out of 25 all - known llagn radio sources that have been detected with vlbis ( column [ 1 ] of table [ table1 ] ) .
the other five llagns are ngc 4235 , ngc 4450 @xcite , ngc 524 , ngc 5354 , and ngc 5846 @xcite , which were recently newly vlbi - detected , but whose reports had not yet been published when we plan the survey .
it is difficult to estimate resultant selection bias including into our sample , because of multiplicity in selection criteria across the past vlbi surveys .
however , at least we can say that 16 out of 20 sources of our sample are all known llagns that are at @xmath11 mpc with @xmath12 mjy at 15 ghz , and the other four sources ( ngc 266 , ngc 3147 , ngc 3226 , and ngc 4772 ) are more distant .
it may be , therefore , as a moderately good distance - limited and radio - flux - limited sample for llagns with a compact radio core .
we made millimetre continuum observations for 16 out of 20 targets using the nobeyama millimetre array ( nma ) , at the nobeyama radio observatory ( nro ) , with d configuration that is the most compact array configuration .
we had precluded the other four sources from our observation list because they had already been observed at @xmath13 mm in the past ( ngc 266 , @xcite ; ngc 3031 , @xcite ; ngc 4258 , doi et al . in prep . ; ngc 4486 , @xcite ) .
our campaign spent more than 20 observational days between 2002 november 28 and 2003 may 25 .
most of weak targets were observed over several days for integration .
visibility data were obtained with double - sided - band receiving system at centre frequencies of 89.725 and 101.725 ghz , which were doppler - tracked .
we used the ultra wide - band correlator ( uwbc ; * ? ? ?
* ) , which can process a bandwidth of 1 ghz per each sideband , i.e. , 2 ghz in total . the wide bandwidth gives the nma a very high sensitivity for continuum observations .
even if the systemic velocities of our targets and galaxies rotations of several hundred km s@xmath14 are accounted for , the observing band can avoid possible contaminations from several significant line emissions : @xmath15co(@xmath16 ) , c@xmath17o(@xmath16 ) , hcn(@xmath16 ) , hco@xmath18(@xmath16 ) , and sio(@xmath19 ) .
a system noise temperature , @xmath20 , was typically about 150 k in the double - sided bands . for observations of each target , we made scans of a reference calibrator close to the target every 20 or 25 minutes for gain calibration .
bandpass calibrators of very bright quasars were scanned once a day .
the data were reduced using the uvproc - ii package @xcite , developed at the nro , by standard manners , including flagging bad data , baseline correction , opacity correction , bandpass calibration and gain calibration . to achieve higher sensitivity , visibilities of both
the sidebands were combined in the same weight , which resulted in a centre frequency of 95.725 ghz .
each daily image was individually made in natural weighting and deconvolved using the aips software , developed at the national radio astronomy observatory ( nrao ) .
statistically significant variability was detected from the daily images in several llagns , which will be reported in a separate paper . in the present paper ,
the visibilities of all days were combined by weighting with @xmath21 , and then imaged .
the half - power beam widths ( hpbws ) of synthesized beams were typically @xmath22 arcsec , corresponding to @xmath23660 pc at a mean distance of 18.9 mpc in our sample .
the flux scales of the gain calibrators were derived with the uncertainty to 10 per cent by relative comparisons to the flux - known calibrators , such as uranus , neptune , or mars .
these were scanned quasi - simultaneously when they were at almost the same elevations .
the rms of noise on the images were estimated from statistics on off - source blank sky with the imean task of the aips .
source identifications and measurements of flux densities were done on image domain by elliptical gaussian profile fitting with the jmfit task of the aips .
we derived total errors in the flux measurements from root sum square of the errors in the gaussian fitting ( including thermal noise ) and the flux scaling .
we attempted to measure flux densities at submillimetre band from the nuclear regions using archival data .
we searched the data that had been obtained at 347.38 ghz ( @xmath24 ) using submillimetre common user bolometer array ( scuba ; @xcite ) on the james clerk maxwell telescope ( jcmt ) .
the data of 14 out of 20 objects of the vlbi - detected sample were available to the public on 2004 .
the jcmt measurements for five objects have already been reported ( ngc 266 , @xcite ; ngc 4258 , doi et al . in prep .
; ngc 4374 , @xcite ; ngc 4472 , @xcite ; ngc 4486 , @xcite ) .
four objects had been observed in jiggle - map mode .
we reduced these data using the scuba user reduction facility ( surf ) package .
standard reduction procedures were used , including flat fielding , flagging of transient spikes , correction for extinction , pointing correction , sky removal and flux - density scaling .
flux calibrators ( uranus , mars , or crl 618 ) gave us flux scaling factors ( we assumed its uncertainty to @xmath2315 per cent ) and an effective beam size of 15.1 arcsec .
the derivations of the rms of noise on images , source identifications , measurements of flux densities , and total error estimations were done in the same manners as the nma reductions described in section [ subsection:3 mm ] .
the other five objects had been observed in photometry mode .
we followed standard reduction procedures , including flat fielding , flagging of transient spikes , correction for extinction , sky removal , averaging , and flux - density scaling , using the surf package .
the photometry can measure a flux density from a beam - size region at the nucleus , i.e. , an intensity .
we derived total errors in flux measurements from root sum square of the errors of the image noise and the flux scale .
[ cols="<,>,^,^,>,>,^,^,^ " , ] [ table4 ] note .
column are : * ( 1 ) * galaxy name ; * ( 2 ) * core power detected in vlbi observations at 5 ghz , except for ngc 4258 at 22 ghz ; * ( 3 ) * presence of jet feature in the vlbi image . ` s ' represents sub - pc - scale jets .
` p ' represents pc - scale jets ; * ( 4 ) * reference for the vlbi observation , as listed below ; * ( 5 ) * presence of jet feature in vla image . `
k ' represents kpc - scale jets ; * ( 6 ) * reference for the vla observation , as listed below ; * ( 7 ) * spectral category , the same as table [ table3 ] .
the sources showing steep spectra have exclusively high vlbi - core powers , @xmath25 w hz@xmath14 . note .
reference : * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ; * * : @xcite ) .
the dot - dashed line represents the radio power limit that the adaf on a black hole with @xmath26 m@xmath27 can generate . ]
we have suggested in the previous section that jets and the adaf seem to coexist in llagns .
however , only an unresolved structure has been found in milli - arcsec images for most of the vlbi - detected llagns @xcite . on the other hand ,
pc / sub - pc - scale jet structures have been found in several llagns .
we find from the core power ranking in table [ table4 ] that the presence of vlbi - detectable pc - scale jets depends on the core radio powers .
the pc - scale jets have been found exclusively in high - core - power sources . in low - core - power sources ,
no pc - scale jet has been detected in vlbi images even with moderately high dynamic ranges of @xmath28 @xcite .
can not the low - core - power sources generate high - brightness jets ?
the cases of ngc 3031 and ngc 4258 are very suggestive .
sub - pc - scale jets have been revealed in their nuclei ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , nevertheless , they are low - core - power llagns .
note that the two llagns are the nearest two sources in the vlbi - detected sample ( table [ table3 ] ) .
we speculate that all low - core - power sources also have sub - pc - scale or smaller jets , and that such structures would be revealed if they were at less than several mpc like ngc 3031 and ngc 4258 .
the detectability of small - scale jets might merely depend on distance .
if that is true , the 5-ghz vlbi - core power should contain jet contributions significantly .
additionally , ngc 3031 and ngc 4258 are very suggestive cases also about coexistence with the adaf disc .
although they show clear evidence of jets , their high - frequency spectra are highly inverted , @xmath29 and @xmath30 , respectively .
these are the highest two spectral indices in the vlbi - detected llagns with negligible diffuse contaminations .
if their millimetre emissions originate in the adaf discs , the two llagns would be examples for the coexistence of the jets and adaf .
we mention apparent exceptions .
ngc 266 and ngc 3147 show steep spectra , but have no detectable jet in spite of their high core powers ( table [ table4 ] ) . since the two llagns are the most two distant llagns in our sample , it could be somewhat hard to reveal their putative small - scale jet structures .
their high radio powers might be also responsible for the dominance of such small - scale jets .
ngc 5866 show a definitely steep spectrum , in spite of a low core power .
its nuclear activity type , ` transition ' , might be related to the jet domination ; another transition object , ngc 4552 , also shows a definitely steep spectrum .
however , our limited sample of transition objects can not allow us to discuss it in detail
. there appears a clear difference in the spectral properties between type 1s ( s1.5 , s1.9 , and l1.9 ) and type 2s ( s2 , l2 , and t2 ) , as seen in table [ table3 ] .
most of the type 1s show inverted spectra , while most of type 2s show steep spectra .
we suggest this difference conflicts with the unified scheme @xcite . in this scheme , many observed properties are determined basically only by a viewing angle with respect to the obscuring dusty torus surrounding the nucleus .
type-1 seyferts , with broad permitted optical emission lines , are those in which the torus is viewed closer to face - on orientation .
type-2 seyferts lack broad permitted lines , except in polarized flux , and thought to be seen with the line of sight passing through the torus in an orientation closer to edge - on .
this classification has been expediently used in liners and transitions ( e.g. , @xcite ) .
since the radio emission is transparent through the dusty torus , luminosities and spectral indices of type 1s and 2s should look just the same to the observer unless affected by relativistic beaming .
however , radio powers of type 2s tend to be significantly larger than those of type 1s in our sample .
it seems that the difference is due to larger contribution of jets in type 2s , nevertheless , which would be less affected by relativistic beaming than type 1s .
there are also the other reports about differences between type 1s and type 2s in llagns .
@xcite found that low - luminosity type-1 seyferts rather show somewhat stronger centimetre radio emission than low - luminosity type-2 seyferts .
@xcite have reported that type-2 llagns are underluminous in an x - ray band compared to type-1 ones with the same h@xmath0 luminosity .
there are ` intrinsically true type-2 seyferts ( without broad lines even in polarized flux ) ' especially in the low - luminosity range ( @xcite , and reference therein ) .
thus , the discrepancy in the unified scheme is commonly seen in llagns .
we suggest a possible idea for the discrepancy with the unified scheme in terms of radio properties of llagns .
if outflows carry accreting mass away and decrease a density at the innermost disc , then the spectral - peak frequency could decrease to several tens of ghz @xcite .
although such a central engine could generate strong jets , neither high - frequency inverted spectra nor a lot of up - scattered photons that will ionize broad - line clouds can be produced .
that is , llagns with strong jets would tend to have weak broad - line emission , which could not have been detected in a limited sensitivity of the optical survey , then they could be classified into type 2s , even if the obscuring torus is in a face - on orientation .
we have found evidence for inverted high - frequency radio spectra in a number of llagns .
such a high - frequency property is potentially the long - sought signatures of the adaf .
it is an apparent observational confirmation for the submillimetre bump theoretically - predicted in the adaf model . in past observations ,
the radio spectra of llagns had been investigated exclusively at centimetre bands where high sensitivities are available .
@xcite reported the radio spectra of 16 llagns that had been observed with the vla at 515 ghz ; @xcite reported the radio spectra of three llagns that had been observed with the vlba at 1.78.4 ghz ; @xcite reported the radio spectra measured with the vlba for three llagns at 2.315 ghz and three llagns at 1.743 ghz .
the frequency range of submillimetre - bump hunting has greatly expanded by our high - frequency radio survey .
our findings will encourage further llagn surveys at higher - frequency ranges with , e.g. , the coming atacama large millimetre and submillimetre array ( alma ) , to prove the adaf .
the existence of the adaf will give radio astronomers a chance to make the direct imaging for accretion discs .
their area of study had been limited to the nonthermal phenomena in relativistic jets of agns , although optical and x - ray astronomers have observed the direct fluxes from the accretion discs as unresolved sources .
radio astronomers have vlbi techniques , which provide the greatest spatial resolutions all over the electromagnetic wavelengths .
a high - frequency vlbi instrument will report the first image of morphology of accretion discs in micro - arcsec resolutions .
the vsop-2 , a next generation space vlbi mission currently being planned , will provide a 38 micro - arcsec resolution at 43 ghz and a brightness - temperature sensitivity of 10@xmath31 k @xcite . because an electron temperature of 10@xmath32 k at the region within several tens schwarzschild radius is predicted by the adaf model
, the vsop-2 can reveal disc structures of more than a dozen nearby llagns .
if the accretion disc were an optically - thick standard disc , it will have never been detected with the vlbi because of low brightness temperature . from our results , both the jets and disc seem to coexist in the core of llagns , and their flux ratios are different from object to object .
it suggests that relative radiative efficiencies between the jet and disc would determined by some kind of physical parameter(s ) in the central engine . from the theoretical point of view @xcite
, it has been suggested that radiatively inefficient flows of low accretion rate tend to stimulate strong convective instabilities and powerful outflows .
the central engines of llagns probably can produce jets .
we will discuss the physical parameters that control jet productions in our llagn sample in a separate paper ( doi et al .
in prep ) .
our millimetre survey and analyses of submillimetre archival data for the 20 vlbi - detected llagns reveal high - frequency radio spectra of llagns with compact cores as follows : ( i ) : : at least half of the sources show inverted spectra between 15 and 96 ghz .
we used the published data at 15 ghz with the vla in a 0.15-arcsec resolution and our measurements at 96 ghz with the nma in a 7-arcsec resolution .
( ii ) : : submillimetre fluxes measured with the scuba on the jcmt may be from the dust associated with host galaxies .
these data have been useful to estimate dust contamination in the nma measurements .
( iii ) : : we confirm little diffuse contamination in the nma measurements from dust , free
free , and extended nonthermal emissions for llagns showing inverted spectra .
therefore , the inverted spectra are not caused by resolution effect , but intrinsic properties of the cores .
( iv ) : : in our sample of 20 llagns , radio cores with inverted high frequency radio spectra reside in at least ten llagns .
( v ) : : we find an inverse correlation between the high - frequency radio spectrum and the low - frequency radio core power .
such a property can be explained by an idea that flux ratios between the jet showing steep spectrum and the adaf disc showing inverted spectrum are different from llagn to llagn .
the disc components could be seen only if the jets are faint .
consequently , we have found evidence for inverted high - frequency radio spectra in a number of llagns .
such a high - frequency property is potentially the long - sought signatures of the adaf .
we acknowledge s. okumura for helpful advices in observations with the nma at the nobeyama radio observatory . the nobeyama radio observatory ( nro ) is a branch of the national astronomical observatory of japan ( naoj ) , which belongs to the national institutes of natural sciences ( nins ) .
we have made extensive use of the nvss and first data from the national radio astronomy observatory that is a facility of the national science foundation operated under cooperative agreement by associated universities , inc .
we have also made extensive use of jcmt archival data from canadian astronomy data centre , which is operated by the dominion astrophysical observatory for the national research council of canada s herzberg institute of astrophysics . | we investigate the high - frequency radio spectra of twenty low - luminosity active galactic nuclei ( llagns ) with compact radio cores .
our millimetre survey with the nobeyama millimetre array ( nma ) and analyses of submillimetre archival data that had been obtained with the submillimetre common user bolometer array ( scuba ) on the james clerk maxwell telescope ( jcmt ) reveal the following properties .
at least half of the llagns show inverted spectra between 15 and 96 ghz ; we use published data at 15 ghz with the very large array ( vla ) in a 0.15-arcsec resolution and our measurements at 96 ghz with the nma in a 7-arcsec resolution .
the inverted spectra are not artificially made due to their unmatched beam sizes , because of little diffuse contamination from dust , regions , or extended jets in these llagns .
such high - frequency inverted spectra are apparently consistent with a ` submillimetre bump ' , which is predicted by an advection - dominated accretion flow ( adaf ) model .
we find a strong correlation between the high - frequency spectral index and low - frequency core power measured with very - long - baseline - interferometry ( vlbi ) instruments .
the inverted spectra were found exclusively in low - core - power sources , while steep spectra were in high - core - power ones with prominent pc - scale jets .
this suggests that the adaf and nonthermal jets may coexists .
the flux ratios between disc and jet seem to be different from llagn to llagn ; disc components can be seen in nuclear radio spectra only if the jets are faint .
[ firstpage ] galaxies : active galaxies : seyfert radio continuum : galaxies submillimetre . |
in continuum non abelian field theories , most popular choices of fixing the gauge ( e.g. landau , coulomb ) suffer from the gribov ambiguity @xcite .
it is now well established that this problem also affects the lattice formulation of these theories @xcite-@xcite .
this problem has been neglected for a long time because , in principle , the computation of gauge invariant operators in compact lattice theories does not require gauge fixing . fixing the gauge is , however , necessary in several cases .
monopole studies in su(2 ) pure gauge theory have been done in the unitary gauge and the effect of the gribov ambiguity on the number of su(2 ) monopoles has been investigated @xcite .
the authors conclude that , in their case , the gribov noise does not exceed the statistical uncertainty . in su(3 )
gauge theory , gauge fixing is essential in the the computation of gauge dependent quantities , such as gluon and quark propagators .
there are now several studies of lattice propagators .
the gluon propagator has been calculated in @xcite-@xcite with the aim of studying the mechanism through which the gluon may become massive at long distances .
more recent attempts have investigated its behaviour as a function of momentum @xcite .
analogous studies have also been performed on the quark propagator ( see , for example @xcite ) . in practice , there are also cases in which it is convenient to implement a gauge dependent procedure for the computation of gauge invariant quantities @xcite-@xcite . for example , smeared fermionic interpolating operators are widely being used in lattice qcd spectroscopy and phenomenology , in order to optimise the overlap of the lower - lying physical state with the operator .
the point - splitted smeared operators are gauge dependent , and therefore the gauge must be fixed before they are calculated . in particular , the calculation of the decay constant of the @xmath5 meson in the static approximation , in which the @xmath6-quark has infinite mass , requires the computation of the two point correlation function of the axial current .
the isolation of the lightest state at large times is not possible if local ( gauge invariant ) operators are used . a nice way out consists in smearing the bilocal operator over a small cube and extracting @xmath7 by forming suitable ratios of smeared and local correlation functions @xcite .
this is an explicitly gauge dependent procedure which is most naturally carried out in the coulomb gauge . in ref.@xcite the smeared - smeared correlation functions on a few individual configurations were computed .
two gribov copies were produced per configuration .
the gribov noise on individual configurations was found to vary from @xmath8 to @xmath9 depending on the time - slice , which implies that it may still be a considerable effect after averaging over configurations .
however , it was not possible to estimate its effect beyond individual configurations .
the reason is that in such a study other sources of error dominate , such as the systematic error arising from fitting the exponential decay of the correlation function with time .
thus the isolation of the gribov noise is difficult . in the static limit @xcite
uses a different method for constructing ratios of smeared and local correlators which avoids fitting .
this method , however , requires a large temporal extention of the lattice . ] in this paper we study a different physical quantity , namely the renormalisation constant @xmath0 of the lattice axial current .
a knowledge of these renormalisation constants is necessary for matching the matrix elements computed using lattice simulations to those required in a definite continuum renormalisation scheme .
provided that the lattice spacing is sufficiently small it is possible to calculate these renormalisation constants in perturbation theory . for a more reliable determination of these constants
it has been suggested to impose the chiral ward identities of @xmath10 non - perturbatively @xcite .
here we focus our attention on the determination of the r@xmath11le of the gribov ambiguity in the calculation of @xmath0 , obtained from quark state correlation functions . a recently proposed method to determine @xmath0 and other renormalisation constants , based on truncated quark green functions in momentum space @xcite
can also in principle be afflicted by gribov fluctuations .
since reasonably small errors are expected , in this kind of calculations , it is crucial to investigate the r@xmath12le of the gribov noise .
moreover , the renormalisation constant @xmath0 of the axial current is particularly well suited to the study of the gribov fluctuations , mainly for two reasons .
firstly , @xmath0 can be obtained from chiral ward identities in two distinct ways : a gauge independent one , which consists in taking the matrix elements between hadronic states , and a gauge dependent one , which consists in taking the matrix elements between quark states .
hence , there is an explicitly gauge invariant estimate of @xmath0 which is free of gribov noise and which can be directly compared to the gauge dependent , gribov affected , estimate .
the second advantage is that @xmath0 is obtained by solving a first degree algebraic equation for each lattice time slice , thus avoiding the usual systematic errors arising from fitting exponentially decaying signals in time .
the theoretical framework for the non - perturbative evaluation of @xmath0 for wilson fermions , has been developed in @xcite .
the renormalisation constant is obtained through ward identities generated by axial transformations .
a first application of these techniques in numerical simulations using the wilson action was attempted in @xcite .
the extension of these methods to the @xmath13 improved clover action @xcite ( @xmath14 is the lattice spacing ) was presented in @xcite , which we follow most closely .
here we only give a brief outline of the results which are essential to our work . in this study terms
that , close to the continuum limit , are effectively of @xmath13 are eliminated by using the clover action @xcite and rotating all quark fields of the matrix elements according to the `` improved improvement '' prescription of @xcite : @xmath15 ( @xmath16 is the wilson term parameter ; in this work @xmath17 ) .
the two fermion local operators considered in the following are the axial and vector currents and the pseudoscalar density : @xmath18 ( @xmath19 is a flavour label and the notation is generic for any quark fields @xmath20 and @xmath21 ) . in order to ensure that the lattice axial current @xmath22 has the correct chiral properties , it is normalised by a renormalisation constant @xmath0 @xcite ; this implies that @xmath23 ( m is the bare quark mass and @xmath24 the pseudoscalar state ) . as can be seen from the above equation , @xmath25 is gauge invariant . the gauge dependent calculation of @xmath0 relies on the following ward identity for quark green functions @xcite @xmath26 \nonumber \\ = \left ( \frac{1}{z_{a}}-{\rho ra}\right ) tr \left [ \int d^3\vec y < ( \gamma_5 d(y)\bar d(0)\ , + \ , u(y)\bar u ( 0)\gamma _ 5 ) > _ { \alpha,\beta } \right ] \label{eq : wiqp}\end{aligned}\ ] ] in eq.([eq : wiqp ] ) we work explicitly with up and down quark fields with spinor labels @xmath27 and @xmath28 .
the trace is over colour indices .
the expectation values on both sides of ( [ eq : wiqp ] ) are evaluated as functions of @xmath29 .
taking the value of @xmath30 obtained using eq.([eq : twomzal ] ) , @xmath31 can then be determined . in order to enhance the signal , we add in both sides of ( [ eq : wiqp ] ) the four contributions @xmath32=(1,3 ) , ( 3,1 ) , ( 2,4 ) and ( 4,2 ) , which were found to give the clearest signal @xcite .
a plateau in @xmath29 is typically obtained and @xmath0 is estimated from it .
the crucial point is that both sides of eq.([eq : wiqp ] ) are gauge dependent , and thus this determination of @xmath0 is in principle sensitive to the gribov noise .
a gauge invariant determination of @xmath0 is obtained through the ward identity @xmath33 in the above equation , the vector current renormalisation constant @xmath34 is also needed .
for the clover action , @xmath34 is calculated with the aid of the so - called conserved and improved vector current @xmath35 \nonumber\\ & & \ \ \ \ \ \ \ \ + ( x\rightarrow x-\hat\mu ) + \frac{r}{2 } \sum_{\rho}\partial_\rho\left(\bar\psi(x)\sigma _ { \rho\mu}\psi(x)\right ) \label{eq : vci}\end{aligned}\ ] ] since the current is conserved , its renormalisation constant is precisely 1 , and , since it is improved " , its matrix elements have no corrections of @xmath13 .
the normalisation constant @xmath34 of the local vector current @xmath36 is determined through the ratio of the vacuum to vector state matrix elements of the two vector currents @xcite : @xmath37 ( here @xmath38 denotes the vector state ) . by calculating all correlation functions of eq.([eq : wivaint ] ) and
@xmath34 from eq.([eq : zvrat ] ) and by requiring that eq.([eq : wivaint ] ) holds at each @xmath29 , we can determine @xmath0 in an explicitly gauge invariant fashion for which the gribov ambiguity is irrelevant .
one more comment is in place here : the terms proportional to @xmath39 on the right - hand - side of eqs.([eq : wiqp ] ) and ( [ eq : wivaint ] ) arise from the rotations of the fermion fields defined in eq.([rot ] ) , which are inherent to clover action improvement @xcite .
the @xmath40 rotations , combined with the equations of motion , generate contact terms which , to @xmath13 , give rise to the terms proportional to @xmath39 @xcite .
gauge fixing and the generation of gribov copies on the lattice is by now a standard procedure .
given a thermalised configuration generated by a monte carlo simulation , the landau gauge is fixed by minimising the functional @xcite @xmath41 \equiv - \frac{1}{v } re \ tr \sum_{\mu=1}^{4 } \sum_{n } u_{\mu}^{g}(n ) \label{eq : bigf}\ ] ] with respect to @xmath42 . here
@xmath43 is the lattice volume and @xmath44 is the compact @xmath1 gauge field , gauge transformed by a local gauge transformation @xmath45 .
the extrema of @xmath46 correspond to configurations that satisfy the gauge condition @xmath47 in discretised form .
the minimisation of @xmath46 is obtained through iteration : each lattice site is visited and @xmath46 is minimized locally .
after several lattice sweeps @xmath46 becomes constant , and the gauge is fixed .
this is nothing else but a discretised analogue of the continuum formulation @xcite , according to which the local minima of the functional @xmath48 = - \sum_{\mu=1}^4 tr \int d^4x ( a^g_\mu)^2 $ ] with respect to @xmath49 correspond to configurations in the landau gauge . even if there is a straightforward correspondence between the lattice and continuum gauge fixing procedures , it is interesting to note that on the lattice , because of discretisation , there can be more minima than in the continuum @xcite .
the production of gribov copies consists in generating gauge equivalent configurations , by applying random gauge transformations to our thermalised link configuration @xcite , @xcite .
the gauge is then fixed and @xmath46 is measured . since it is a gauge dependent quantity , a different value of @xmath46 for two gauge equivalent , gauge fixed configurations means that they are gribov copies .
the over - relaxation technique of @xcite ( which consists in accelerating the gauge fixing algorithm by raising the gauge transformation @xmath49 to a real tunable power @xmath50 at every iteration ) has been implemented at fixed @xmath50 .
the over - relaxation itself can be used for the generation of gribov copies , by varying the value of @xmath50 , as proposed in @xcite . in this work
we have opted for the random gauge transformation method .
we work in the framework of the quenched approximation with the clover action of su(3 ) gauge theory .
the lattice volume is @xmath51 and @xmath3 .
after @xmath52 thermalising sweeps , 36 configurations were generated , separated by @xmath53 sweeps .
an @xmath54 hit metropolis algorithm was used . for each thermalised configuration
, we generated @xmath55 gribov copies .
this was done by fixing the gauge both on the original configuration and on @xmath56 gauge equivalent replicas , obtained by applying random gauge transformations .
it is remarkable that , in our case , each random gauge transformation produced a gribov copy .
this high probability to find gribov copies is a characteristic of large volume lattices in the confined region , as discussed in @xcite .
we fix the landau gauge , using the over - relaxation algorithm suggested in @xcite at fourth order in the over - relaxation parameter @xmath57 .
the stopping condition we have imposed is that the relative variation @xmath58 , of the minimised functional @xmath46 , between two consecutive gauge fixing sweeps be less than @xmath59 .
this value guarantees a good quality of the gauge fixing , allows us to distinguish gribov copies , and it is typically reached after a number of sweeps which varies between 500 and 1500 .
the gauge fixing was done on an ibm risc 6000/550 equipped with 128 mbyte of ram memory and with a cpu working at 42 mhz ; with this machine a single gauge fixing sweep takes about 40 s. the quark propagators , rotated as indicated by eq.([rot ] ) were obtained at a wilson hopping parameter value of @xmath60 , which , for the clover action , corresponds to a pion of roughly 900 mev . before passing to a detailed discussion of the gribov noise ,
it is appropriate to present a comparison of the results for @xmath0 , obtained in a gauge invariant way ( see eq.([eq : wivaint ] ) ) , to those based on the gauge dependent identity of eq.([eq : wiqp ] ) .
this calculation has been already presented in @xcite , on the first 18 configurations of our ensemble ; here we have doubled the statistics .
moreover , we have re - gauged the first 18 configurations in order to reach the better quality of gauge fixing required for this study . in fig.(1 ) we show the behaviour of the two estimates of @xmath0 , as a function of @xmath29 , calculated on the same ensemble of 36 configurations .
the gauge invariant values of @xmath0 show a flat behaviour already at @xmath61 and with small error bars .
the new value of @xmath0 , obtained with the gauge independent technique , over 36 configurations , is @xmath62 to be compared with the old one obtained over 18 configurations : @xmath63 ( see ref.@xcite ) . in the gauge dependent case
, we have taken the average over the gribov copies , in the way that will be discussed below . in this case , the @xmath0 behaviour becomes flat only at @xmath64 showing a large sensitivity to the contact terms of the ward identity at small @xmath29 values .
the new value of @xmath0 , obtained from the gauge dependent ward identity , as can be seen from fig.(1 ) is @xmath65 ; to be compared to the value obtained from 18 configurations , @xmath66 ( see ref.@xcite ) .
as already stressed in @xcite , this method gives results that are compatible within the errors with the gauge independent ones .
moreover , the error of the gauge invariant calculation of @xmath0 is always smaller than the error of its gauge dependent counterpart .
this is due to the fact that the quark state correlation functions fluctuate more than the gauge invariant correlation functions , but it may also indicate the presence of another effect , which is probably the gribov ambiguity . in the hypothetical case of two determinations of @xmath0 , affected by the same statistical error , the gauge dependent estimate should fluctuate more due to the gribov noise .
then the amount of gribov noise could be estimated as the difference ( in quadrature ) of the two errors . normally , in a standard simulation of gauge dependent quantities
, one does not generate gribov copies .
consequently , one measures a given quantity by taking the average and error over the ensemble of the gauge fixed configurations that have been generated .
this implies a particular and arbitrary choice of gribov copies .
the error estimated , for example , by a jacknife method , is not purely statistical as it implicitly contains the uncertainty due to the particular choice of a gribov copy . in our case , having generated @xmath67 copies for each of the @xmath68 thermalised configurations , there are @xmath69 possible combinations that we may consider when forming a particular ensemble .
to the best of our knowledge , the distribution of the gribov copies of a given configuration is unknown ; thus the weight to be associated to it is arbitrary .
moreover , any technique used to generate different gribov copies selects a particular copy in a completely uncontrolled way .
hence , we assume that the particular choice of different combinations of gribov copies when forming a statistical ensemble is arbitrary . in order to exhibit the effect of such arbitrariness , we show in fig.(2 ) the behaviour of @xmath70 for @xmath71 arbitrary choices of copies .
the @xmath71 different behaviours are compatible , and the same is true for the jacknife errors .
it is clear , however , that the presence of gribov copies is a visible effect ; each of the six estimates of @xmath0 shown has a slightly different profile as a function of @xmath29 .
we now expose the procedure we implemented for taking into account the gribov ambiguity .
the gauge dependent traces of the two and three point correlation functions appearing in eq.([eq : wiqp ] ) , calculated on a single configuration @xmath72 and for a particular gribov copy @xmath42 are denoted by : @xmath73 \nonumber \\ t_3(t_y ; c , g ) = tr \left [ \int d^4x \int d^3\vec y \ , u(y)\,\left(\bar u(x)\gamma_5 d(x)\right ) \,\bar
d(0)\ \right ] \label{eq
: t2t3a}\end{aligned}\ ] ] ( in the above equations the dirac indices have been implicitly averaged over , as explained in sect.(2 ) ) .
then , for a given configuration @xmath72 , we consider our `` best estimate '' of these matrix elements to be their average over the @xmath67 gribov copies : @xmath74 the average of the above @xmath75 s over the @xmath76 configurations will be taken as our `` best estimate '' @xmath77 for the gauge dependent traces . on the other hand ,
@xmath25 , being gauge invariant , does not depend on a particular choice of gribov copies , but only on the configuration ensemble .
then @xmath70 is obtained by applying eq.([eq : twomzal ] ) as follows @xmath78 the error is obtained by a standard jacknife method performed on the quantities @xmath79,@xmath80 and @xmath81 , by decimating one configuration at a time .
this is the gauge dependent @xmath70 result shown in fig.(1 ) .
we want to stress that the values of @xmath82 and @xmath83 fluctuate more than their ratio @xmath84 .
the latter quantity is gauge invariant , according to eq.([eq : zam1 ] ) . for example , at @xmath85 , @xmath86 , @xmath87 and @xmath88 .
the strong reduction of the relative error indicates a great sensitivity of @xmath89 to the gribov fluctuation , as opposed to a relative stability of the gauge invariant quantities . in order to estimate the uncertainty arising from a particular choice of copies , out of the @xmath90 possible ones , we have applied a procedure which takes this arbitrariness into account
we have chosen randomly @xmath91 combinations of copies of our ensemble and have calculated @xmath70 for each one of these combinations at fixed @xmath29 .
the histogram of the values obtained for @xmath70 is well fitted by a gaussian , the r.m.s .
width of which is taken as an estimate of the fluctuation . in fig.(3
a ) we compare the jacknife error of our `` best estimate '' to the width of the gaussian .
we see that for all @xmath29 the width is smaller than the jacknife error .
this implies that the fluctuations induced by the particular choice of gribov copy when forming the ensemble are small and do not overcome the statistical uncertainty .
as it is also important to understand how the above behaviour scales with increasing the number of configurations , we have performed the same analysis for the first @xmath92 configurations ( half of our ensemble ) .
the result is shown in fig.(3 b ) .
comparing fig.(3 a ) to fig.(3 b ) , we note that both errors scale at least as @xmath93 . thus , within our moderate statistics , we find that the error , even if affected by the gribov ambiguity , decreases with increasing configuration number . in conclusion , our investigation , even within its limitations , shows that , on the lattice , the residual gauge dependence associated with the gribov copies does not generate an overwhelming fluactuation of the @xmath0 measurements performed by the gauge dependent method .
nevertheless the arbitrariness associated to a particular choice of gribov copies is a visible effect which manifests itself , especially in the behaviour of @xmath0 as a function of @xmath29 .
the jacknife error is greater in the gauge dependent determination than in the gauge independent one .
however , in the former case , even though the error is afflicted by the gribov uncertainty , it is still decreasing with increasing statistics .
this implies that the uncertainty arising from the gribov ambiguity may be a secondary effect .
analogous studies on different physical quantities and renormalisation constants could further support this belief .
we warmly thank g.martinelli , c.t.sachrajda and m. testa for many discussions and partecipation throughout this work .
useful discussions with v.n .
gribov and a.j .
van der sijs are also greatfully acknowledged .
we also thank the ibm - semea for providing us with 32 mbytes of memory .
figure 1 : comparison of the gauge independent calculation of @xmath70 ( diamonds ) to the gauge dependent one ( crosses ) .
the errors are jacknife .
the crosses have been slightly displaced to help the eye .
+ figure 2 : the gauge dependent calculation of @xmath70 for 4 arbitrarily chosen gribov copies .
the errors are jacknife . + figure 3 : comparison of the jacknife error ( crosses ) to the r.m.s .
gaussian width ( diamonds ) due to the gribov ambiguity ( see text ) .
the crosses have been slightly displaced to help the eye .
case ( a ) is with 36 configurations ; case ( b ) with the first 18 configurations | we study the influence of gribov copies , in the landau gauge , on the lattice renormalisation constant @xmath0 of the axial current , obtained from a ward identity on quark state correlation functions , with the clover action , in quenched @xmath1 gauge theory .
a comparison between the gauge invariant determination of @xmath0 and the gauge dependent one is discussed . on a @xmath2 lattice at @xmath3 and with @xmath4 , the values , on a sample of 36 configurations , are : @xmath0 = 1.08(5 ) ( gauge dependent calculation ) and @xmath0 = 1.06(2 ) ( gauge independent calculation ) .
we find that the residual gauge freedom associated to gribov copies induces observable effects , which , at the level of numerical accuracy of our simulation , are included in the statistical uncertainty inherent in a monte carlo simulation .
doubling the statistics suggests that the fluctuation due to the lattice gribov ambiguity scales down at least as fast as a pure statistical error . |
tidal interactions between short - period planets and their host stars are thought to play an important role in the evolution of the planetary orbit as well as the stellar and planetary spins ( e.g. zahn @xcite ; mathis et al .
@xcite ; ogilvie @xcite ) .
the clearest evidence of tidal evolution in extrasolar planetary systems is the eccentricity distribution of the approximately jupiter - mass planets ( here taken to mean masses @xmath0 ) , which is shown in fig .
planets with @xmath1 d have a wide range of eccentricities , whereas the population with @xmath2 d has much lower eccentricities and displays a strong preference for circular orbits .
tidal dissipation inside the planet ( and perhaps partly also the star ) is thought to be responsible for this dichotomy , since it tends to damp the orbital eccentricities of the shortest - period planets ( it may also have played a role in their formation if they underwent a period of high eccentricity migration " to attain such periods e.g. wu & lithwick @xcite ; naoz et al .
@xcite ; anderson et al .
the timescale for this evolution depends on the mechanisms of tidal dissipation , which are only partly understood theoretically .
here , i will focus on nonlinear tidal mechanisms that require finite amplitude tidal deformations .
this is partly because nonlinear effects are the least well understood , but more importantly it is because their consideration is likely to be essential for understanding the tidal evolution of short - period planets .
and @xmath3 d ) .
planets with @xmath1 d have a wide range of eccentricities , whereas the population with @xmath2 d has much lower eccentricities and displays a strong preference for circular orbits .
this provides strong evidence of the important role of tidal dissipation in shaping planetary orbits.,scaledwidth=60.5% ] the ( dimensionless ) tidal deformations of short - period planets can be estimated by ( the height of the tide is approximately @xmath4 ) @xmath5 where @xmath6 and @xmath7 are the stellar and planetary masses , @xmath8 is the planetary radius , @xmath9 is the orbital semi - major axis , and @xmath10 is the orbital period ( taking @xmath11 , @xmath12 and @xmath13 on the right hand side ) .
the most extreme current example is wasp-19 b ( hebb et al .
@xcite ) , with its @xmath14 d orbital period , giving @xmath15 .
this is no longer a small parameter , indicating that nonlinear effects could be important even for large - scale tidal flows in such a body .
this can be compared with the tides in jupiter and saturn due to their most massive satellites ( @xmath16 ) , where nonlinear effects may be much less important for the largest - scale tidal flows ( though they could still be important in damping tidally - excited short - wavelength waves ) . in this paper , i will discuss two related nonlinear tidal mechanisms : the elliptical instability and the precessional instability .
the former occurs in fluids with elliptical streamlines ( see also the related paper by favier @xcite ) , such as in tidally deformed planets , and the latter occurs in fluid bodies that undergo axial precession , such as planets with misaligned spins and orbits ( nonzero obliquities ) . both are parametric instabilities driven by the periodic time - dependence of fluid properties around a streamline , leading to the excitation of inertial waves ( restored by the coriolis force ) in the planetary interior .
and both mechanisms are likely to be important for tidal evolution of short - period planets .
is pointing towards us ) , also indicating the local model considered by barker & lithwick ( @xcite ) .
right : results of local cartesian ( mhd ) and global ellipsoidal ( hydrodynamic ) simulations of the elliptical instability , showing that the turbulent ( volume and time - averaged ) dissipation is consistent with a simple cubic scaling with @xmath17.,scaledwidth=46.0% ] the elliptical instability is a fluid instability of elliptical streamlines , such as the large - scale non - wave - like tidal flow in a planet that is non - synchronously rotating or has an eccentric orbit ( see the left panel of fig . [ ei ] for illustration ) .
the simplest case for illustration is a planet on a circular orbit but that is non - synchronously rotating ( with an aligned or anti - aligned spin axis ) . in the frame rotating at the rate @xmath18 about the spin axis , the elliptical deformation has frequency @xmath19 , where @xmath20 .
inertial waves exist with frequencies @xmath21 . if a pair of these waves ( with subscripts 1 & 2 ) has frequencies such that @xmath22 , then the superposition of one of these waves with the deformation can excite the other wave , and vice versa , leading to instability .
consideration of the spatial structure of the waves leads to the additional requirement that the azimuthal wavenumbers and harmonic degrees satisfy @xmath23 ( since the deformation has @xmath24 ) and @xmath25 .
the maximum growth rate ( which typically occurs when @xmath26 ) is ( e.g. kerswell @xcite ) @xmath27 where @xmath28 is a dimensionless function of @xmath29 and @xmath30 . in the limit @xmath31
, instability occurs if @xmath32 , but is also possible outside of this range ( particularly for anti - aligned spins with @xmath33 , if @xmath34 is sufficiently large , due to the finite resonant widths e.g. barker et al .
@xcite ) .
the instability leads to turbulence in the planetary interior that produces enhanced tidal dissipation . in hydrodynamical simulations ( barker @xcite ) , the instability generates differential rotation in the planetary interior in the form of zonal flows ( or columnar vortices in a local cartesian model , at least if @xmath35 ; barker & lithwick @xcite ) , which control the saturation of the instability , leading to cyclic , predator - prey - like dynamics ( where zonal flows act as the predator and waves the prey ) with intermittent dissipation .
zonal flows have also been observed in experiments of the elliptical instability ( e.g. le bars et al .
@xcite ; favier et al .
@xcite ) . in the presence of weak magnetic fields ( barker & lithwick @xcite ) , or an alternative frictional mechanism that damps zonal flows ( e.g. le reun et al .
@xcite ) , the instability leads to sustained turbulence with significantly enhanced dissipation over that of the laminar tidal flow .
the ( volume and time - averaged ) turbulent dissipation rate @xmath36 resulting from sustained instability can be simply estimated .
nonlinearities saturate growth of the unstable waves if @xmath37 , where @xmath38 is a typical velocity and @xmath39 is the wavelength of the dominant waves .
kinetic energy is dissipated at the rate @xmath40 , where we have taken @xmath41 for simplicity . for numerically accessible @xmath42 ( including the relevant regime for the shortest - period planets such as wasp-19 b ) , both local and global simulations are consistent with this simple argument , as we show in fig .
we find @xmath43 , where @xmath44 provides an approximate upper bound on @xmath36 ( though @xmath45 does vary to some extent with @xmath29 , as does @xmath46 ) .
it remains to be seen whether this scaling would persist for much smaller @xmath34 , when the instability is more likely to excite very short - wavelength waves ( @xmath47 ) .
also , these arguments neglect the inhibiting effects of zonal flows that lead to intermittent rather than sustained turbulence .
hence , this scaling is best thought of as an approximate upper bound on the dissipation ( due to bulk turbulence ) if we want to extrapolate to smaller @xmath34 . in the frequency range where the instability operates ,
the resulting turbulent dissipation would drive tidal synchronisation ( and evolution of the planetary spin - orbit angle @xmath48 ) on a timescale @xmath49 where @xmath50 is the planetary rotation period , and circularisation on a timescale @xmath51 where we have taken the planet to have jupiter s mass and radius ( and radius of gyration ) and the star to be sun - like for both estimates ( or constant lag - time ) model . ] ( barker @xcite ) .
hence , this instability could be important for tidal evolution of the shortest - period hot jupiters , and we would predict orbital circularisation for @xmath52 d , and spin synchronisation and alignment for @xmath53 d , due to this mechanism acting in isolation . however , additional mechanisms appear to be required to explain tidal circularisation for @xmath54 ( fig .
[ evsp ] ) , such as tidal dissipation of linearly - excited inertial waves ( and their linear or nonlinear damping ) in planets with a core ( e.g. ogilvie & lin @xcite ) , or dissipation in the solid core itself ( remus et al .
@xcite ; storch & lai @xcite ) .
the spin axis of a rotationally deformed planet is forced to precess about its orbital angular momentum vector , due to the tidal gravity of its host star , if these directions are misaligned ( if the spin - orbit angle @xmath55 ) .
a typical period of axial precession is @xmath56 where @xmath57 is the precession rate ( @xmath58 is a dimensionless version ) , for a jupiter - mass and radius planet orbiting a sun - like star .
this precession is potentially observable for some transiting planets through transit depth variations ( e.g. carter & winn @xcite ) , which would provide important constraints on the oblateness and interior structure of the planet .
however , axial precession would only occur in single planet systems ( and be potentially observable ) if @xmath48 can remain significantly misaligned .
tidal evolution is expected to ultimately lead to alignment ( @xmath59 ) .
one mechanism that could play a role in driving this tidal evolution is the precessional instability , which will now be introduced .
axial precession induces internal flows inside the planet that are time - periodic , with frequency @xmath60 in the frame rotating with the planet .
pairs of inertial waves can be driven unstable if @xmath61 , and the fastest growing modes typically have @xmath62 , with a maximum growth rate ( kerswell @xcite ) @xmath63 the nonlinear evolution of this instability in a local cartesian model behaves just like the elliptical instability ( barker & lithwick @xcite , barker @xcite ) . for small @xmath58 , the purely hydrodynamical instability generates columnar vortices which inhibit sustained instability and lead to cyclic , predator - prey - like dynamics , with intermittent turbulent dissipation .
the addition of a weak magnetic field eliminates these vortices and leads to sustained turbulence .
global simulations are currently in progress , where we expect the instability to generate zonal flows , and for these to subsequently control the dissipation ( this was demonstrated already in the specific case of an initially anti - aligned spin , once it had evolved away from pure anti - alignment , in barker @xcite , where the elliptical instability was also present ) . the turbulent dissipation resulting from this instability can be estimated in the same way as for the elliptical instability outlined above .
the local mhd ( and also , approximately , the bursty " hydrodynamic ) simulations are consistent with @xmath64 ( with turbulent velocities @xmath65 ) , with @xmath66 , over the range @xmath67 , as we show in fig .
this range is somewhat larger than the realistic values of @xmath58 , requiring us to extrapolate the scaling to probe the astrophysical regime .
it remains to be seen whether this scaling would persist for much smaller @xmath58 , so it is best thought of as an approximate upper bound on the bulk turbulent dissipation when @xmath68 .
the consequences of this turbulent dissipation are tidal evolution of the spin - orbit angle @xmath48 for short - period planets , which evolves due to this mechanism acting in isolation according to ( lai @xcite , ogilvie @xcite ) @xmath69 where @xmath70 is the ratio of planetary spin to orbital angular momentum and @xmath71 is the timescale for @xmath48 to evolve towards alignment ( @xmath72 ) if @xmath73 is small ( barker @xcite ) , for a jupiter - mass and radius planet around the current sun .
this is also the timescale for @xmath48 to evolve towards anti - alignment ( @xmath74 ) , if @xmath75 is small .
the corresponding timescale if @xmath76 is larger by a factor @xmath77 , indicating that most planets with nearly perpendicular spins will not undergo appreciable evolution by this mechanism acting in isolation .
however , other tidal mechanisms will also contribute , and in general we expect the ultimate tidal evolution to be towards alignment .
nevertheless , this crude estimate suggests that the precessional instability could be important for tidal evolution of the spin - orbit angle for the shortest - period hot ( and warm ) jupiters with @xmath78 d.
i have discussed two related nonlinear mechanisms of tidal dissipation that require finite - amplitude tidal deformations for their operation , which could both be important for the spin - orbit evolution of short - period extrasolar planets .
the elliptical instability occurs in tidally deformed fluid bodies .
local and global simulations suggest that it could drive tidal spin synchronisation ( and spin - orbit alignment ) for hot ( and warm ) jupiters with with @xmath34 that takes into account that @xmath79 when @xmath31 ( barker @xcite ) , which is also consistent with simulations for accessible values of @xmath80 . ]
@xmath81 d , and tidal circularisation for @xmath82 d. other mechanisms are probably required to explain the small eccentricities of planets with @xmath83 , shown in fig .
[ evsp ] .
the precessional instability occurs in axially precessing fluid bodies , such as in rotating planets with spin - orbit misalignments .
simulations using a local model suggest that it can drive tidal evolution of the spin - orbit angle ( towards alignment or anti - alignment , depending on the initial value ) for planets with @xmath84 d ( barker @xcite ) .
global simulations of this instability are currently in progress . the models that have been adopted so far to study these instabilities are extremely idealised , mainly restricted to homogeneous fluid bodies .
real planets and stars are certainly not homogeneous , but exploring such models constituted a necessary first step .
future work is required to simulate the instability in more realistic planetary models ( i.e. with a realistic density and entropy stratification ) , including the presence of an inner core ( e.g. by extending favier et al .
@xcite to adopt more realistic boundary conditions and consider larger tidal amplitudes ) , as well as the influence of magnetic fields in global models .
the effect of zonal flows on the saturation of these instabilities in the astrophysically relevant regime of small deformations ( or precession rates ) , as well as vanishingly small viscosities should be considered , as should the coexistence of both types of instability , and these instabilities with turbulent convection .
clearly , much further work is required to understand tidal dissipation in planets and stars , and for us to be able to explain the astrophysical observations .
ajb was funded by a leverhulme trust early career fellowship .
anderson , k. r. , storch , n. i. , lai , d. 2016 , mnras , 456 , 3671 barker , a. j. & lithwick , y. 2013 , mnras , 435 , 3614 barker , a. j. & lithwick , y. 2014 , mnras , 437 , 305 barker , a. j. , braviner , h. j. & ogilvie , g. i. 2016 , mnras , 459 , 924 barker , a. j. 2016 , mnras , 459 , 939 ( 2016a ) barker , a. j. 2016 , mnras , 460 , 2339 ( 2016b ) carter , j. a. & winn , j.n . 2010 , apj , 716 , 850 hebb , l. , et al .
2010 , apj , 708 , 224 favier , b. 2017 , this volume favier , b. , barker , a. j. , baruteau , c. & ogilvie , g. i. 2014 , mnras , 439 , 845 favier , b. , grannan , a. m. , le bars , m. , aurnou , j. , m. 2015 , physics of fluids , 27 , 6 le bars , m. , cbron , d. , le gal , p. 2015 , ann .
, 47 , 163 le reun , t. , favier , b. , barker , a. j. , le bars , m. 2017 , submitted kerswell , r. r. 1993 , gafd , 72 , 107 kerswell , r. r. 2002 , ann .
mech . , 34 , 83 lai , d. 2012 , mnras , 423 , 486 mathis , s. et al .
2013 , in _ tides in astronomy and astrophysics _ ,
lecture notes in physics , volume 861 , springer - verlag berlin heidelberg , 2013 , p. 255 naoz , s. , farr , w. m. , lithwick , y. , rasio , f. 2011 , nature , 473 , 7346 remus , f. , mathis , s. , zahn , j - p . , lainey , v. 2012 , a&a , 541 , 165 storch , n. , lai .
, d. 2014 , mnras , 438 , 1526 ogilvie , g. i. 2004 , apj , 610 , 477 ogilvie , g. i. 2014 , ara & a , 52 , 171 wu , y. , lithwick , y. 2011 , apj , 735 , 109 zahn , j - p .
2013 , in _ tides in astronomy and astrophysics _ ,
lecture notes in physics , volume 861 , springer - verlag berlin heidelberg , 2013 , p. 301 | i discuss two related nonlinear mechanisms of tidal dissipation that require finite tidal deformations for their operation : the elliptical instability and the precessional instability .
both are likely to be important for the tidal evolution of short - period extrasolar planets .
the elliptical instability is a fluid instability of elliptical streamlines , such as in tidally deformed non - synchronously rotating or non - circularly orbiting planets .
i summarise the results of local and global simulations that indicate this mechanism to be important for tidal spin synchronisation , planetary spin - orbit alignment and orbital circularisation for the shortest period hot jupiters .
the precessional instability is a fluid instability that occurs in planets undergoing axial precession , such as those with spin - orbit misalignments ( non - zero obliquities ) .
i summarise the outcome of local mhd simulations designed to study the turbulent damping of axial precession , which suggest this mechanism to be important in driving tidal evolution of the spin - orbit angle for hot jupiters .
avenues for future work are also discussed . |
the one - electron models of solids are based on the study of schrdinger operator with periodic potential .
there are a lot of studies on the periodic potential , in particular , for periodic point interactions , we can show the spectral set explicitly ( albeverio et .
@xcite is the best guide to this field for readers ) .
most fundamental case is the one - dimensional schrdinger operator with periodic point interactions , called the _ kronig
penney model _
( see kronig
penney @xcite ) , given by @xmath5 where @xmath6 is a non - zero real constant , and @xmath7 is the dirac delta measure at @xmath8 .
the positive sign of @xmath6 corresponds the repulsive interaction , while the negative one corresponds the attractive one .
more precisely , @xmath0 is the negative laplacian with boundary conditions on integer points : @xmath9 where @xmath10 here @xmath11 and @xmath12 is the usual sobolev space of order @xmath13 on the open set @xmath14 . from sobolev s embedding theorem @xmath15 ,
every elements of @xmath16 are continuous ( classical sense ) and uniformly bounded functions .
it is well - known that @xmath0 is self - adjoint and is a model describing electrons on the quantum wire .
the spectrum of this model is explicitly given by @xmath17 where @xmath18 @xmath19 is so - called _ discriminant _ and @xmath20 can be regarded as an entire function with respect to @xmath21 .
the spectrum @xmath22 of @xmath0 consists of infinitely many closed intervals ( spectral bands ) and is purely absolutely continuous . on the other hand , for the schrdinger operator @xmath23 on @xmath24 with decaying potential @xmath6 ,
the _ dispersive estimate _ for the schrdinger time evolution operator @xmath2 is stated as follows : @xmath25 where @xmath26 denotes the spectral projection to the absolutely continuous subspace for @xmath0 .
the dispersive estimate is a quantitative representation of the diffusion phenomena in quantum mechanics , and is extensively studied recently , because of its usefulness in the theory of the non - linear schrdinger operator ( see e.g. journ
sogge @xcite , weder @xcite , yajima @xcite , mizutani @xcite , and references therein ) .
the estimate ( [ dispersive_estimate ] ) is also obtained in the case of the one - dimensional point interaction .
sacchetti @xcite obtain ( [ dispersive_estimate ] ) when @xmath6 is one point @xmath27 potential , and so do kovak
sacchetti @xcite when @xmath6 is the sum of @xmath27 potentials at two points .
the motivation of the present paper is to obtain a similar estimate for our periodic model ( [ kronig - penney0 ] ) .
though this problem is quite fundamental , we could not find such kind of results in the literature , probably because the deduction of the result requires a detailed analysis of the band functions , as we shall see below .
since the spectrum of the schrdinger operator with periodic potential is absolutely continuous , one may expect some dispersion type estimate holds also in this case
. however , there seems to be few results about the dispersive type estimate for the time evolution operator of the differential equation with periodic coefficients .
an example is the paper by cuccagna @xcite , in which the klein
gordon equation @xmath28 is considered , where @xmath29 is a smooth real - valued periodic function with period @xmath30 .
cuccagna proves the solution @xmath31 satisfies @xmath32 for @xmath33 , where @xmath34 is some bounded discrete set .
the peculiar power @xmath35 comes from the following reason .
the integral kernel for the time evolution operator is written as the sum of oscillatory integrals @xmath36 where @xmath37+x'$ ] , @xmath38+y'$ ] , @xmath39,[y]\in\mathbf{z}$ ] , @xmath40 , and @xmath41-[y])/t$ ] .
the function @xmath42 is called the _ band function _ for the @xmath43-th band , which is a real - analytic function of @xmath44 with period @xmath45 . in the large time limit @xmath4 , it is well - known that the main contribution of the oscillatory integral ( [ intro0 ] ) comes from the part nearby the _ stationary phase point _
@xmath46 ( the solution of @xmath47 ) , and the _ stationary phase method _ tells us the principal term in the asymptotic bound is a constant multiple of @xmath48 ( see e.g. stein ( * ? ? ?
* chapter viii ) or lemma [ lemma_stein ] below ) .
however , since @xmath42 is a periodic function , there exists a point @xmath49 so that @xmath50 .
if the stationary phase point @xmath46 coincides with @xmath49 , then the previous bound no longer makes sense .
instead , the stationary phase method concludes the principal term is a constant multiple of @xmath51 .
since the integral kernel of our operator has the same form as ( [ intro0 ] ) ( see ( [ st01 ] ) below ) , we expect a result similar to ( [ cuccagna ] ) also holds in our case .
let us formulate our main result .
let @xmath0 be the hamiltonian for the kronig - penney model given in ( [ kronig - penney1 ] ) and ( [ kronig - penney2 ] ) . as stated above
, the spectrum of @xmath0 has the band structure , that is , @xmath52 where the @xmath43-th band @xmath53 is a closed interval of finite length ( for the precise definition , see ( [ bloch13 ] ) below ) .
our main result is as follows .
[ theorem_main ] let @xmath54 be the spectral projection onto the @xmath43-th energy band @xmath53 .
then , for sufficiently large @xmath43 , there exist positive constants @xmath55 and @xmath56 such that @xmath57 for any @xmath58 and any @xmath59 , where @xmath60 .
the coefficients obey the bound @xmath61 as @xmath62 .
the power @xmath63 in the first term of the coefficient in ( [ main1 ] ) is the same as in ( [ dispersive_estimate ] ) with @xmath64 and @xmath65 , since it comes from the states corresponding to the energy near the band center , which behaves like a free particle .
this fact can be understood from the graph of @xmath66 ( figure [ figure_band_function ] ) .
the part of the graph corresponding to the band center is similar to the parabola @xmath67 or its translation , which is the band function for the free hamiltonian @xmath68 . on the other hand ,
the power @xmath35 comes from part of the integral ( [ intro0 ] ) given by @xmath69 where @xmath70 is some open set including two solutions @xmath49 to the equation @xmath71 .
notice that @xmath72 is an inflection point in the graph of @xmath66 ; see figure [ figure_band_function ] .
the estimates for the coefficients are obtained from the lower bounds for the derivatives of @xmath42 .
actually , we can choose @xmath70 so that @xmath73\setminus j_n}|\lambda_n''(\theta)|\geq c,\quad \inf_{\theta\in j_n}|\lambda_n^{(3)}(\theta)|\geq c n^{1/3},\ ] ] where @xmath74 is a positive constant independent of @xmath43 . by ( [ intro2 ] ) and the estimates for the amplitude function ( proposition [ proposition_amplitude1 ] ) , we can prove theorem [ theorem_main ] by using a lemma for estimating oscillatory integrals , given in stein s book ( see stein @xcite or lemma [ lemma_stein ] below ) .
we can also prove the power @xmath35 is optimal , by considering the case @xmath75 ( so , @xmath49 is a stationary phase point ) , and applying the asymptotic expansion formula in the stationary phase method ( see e.g. stein ( * ? ? ?
* page 334 ) ) .
( @xmath76 ) for @xmath77 .
the range of @xmath42 is the @xmath43-th band @xmath53 .
we find two inflection points of @xmath66 near @xmath78 , for every @xmath43 .
, width=302 ] the physical implication of the result is as follows . by definition , the parameter @xmath41-[y])/t$ ] represents the _ propagation velocity _ of a quantum particle .
the wave packet with energy near @xmath79 has the maximal group speed in the @xmath43-th band , and the speed of the quantum diffusion is slowest in that band .
thus such state has a bit longer life - span ( in the sense of @xmath80-norm ) than the ordinary state has ; the state is in some sense a _ resonant state _ , caused by the meeting of two stationary phase points @xmath46 as @xmath81 tends to @xmath82 .
it is well - known that the existence of resonant states makes the decay of the solution with respect to @xmath83 slower ( see jensen kato @xcite or mizutani @xcite ) . since
the estimate ( [ main1 ] ) is given bandwise , it is natural to ask we can obtain the dispersive type estimate for the whole schrdinger time evolution operator @xmath2 , like cuccagna s result ( [ cuccagna ] ) .
however , it turns out to be difficult in the present case , from the following reason . a reasonable strategy to prove such estimate is as follows .
first , we divide the integral @xmath84 into two parts @xmath85 and the rest , where @xmath85 is given in ( [ intro1 ] ) with @xmath70 some open set including @xmath86 .
next , we show the sum of @xmath85 converges and gives @xmath3 , and the sum of the rests also converges and gives @xmath87 .
however , for fixed @xmath88 and @xmath89 , we find that our upper bound for @xmath90 is not better than @xmath91 , and the sum of the upper bounds does not converge ( see the last part of section 4 ) .
one reason for this divergence is very strong singularity of our potential , the sum of @xmath27-functions .
because of this singularity , the width of the band gap , say @xmath92 ( @xmath43 is the band number ) , does not decay at all in the high energy limit @xmath93 .
implies @xmath94 as @xmath93 . ]
then we can not take the open set @xmath70 so small , in the proof of theorem [ theorem_main ] . ] and the sum of the lengths @xmath95 diverges ; if this sum converges , we can use a simple bound @xmath96 to control the sum .
thus we do not succeed to obtain a bound for the sum of @xmath90 at present . on the other hand , for the schrdinger operator @xmath97 on @xmath98 with real - valued periodic potential @xmath6
, it is known that the decay rate of the width of the band gap @xmath92 reflects the smoothness of the potential @xmath6 .
hochstadt @xcite says @xmath99 if @xmath6 is in @xmath100 , and trubowitz @xcite says @xmath101 ( @xmath102 is some positive constant ) if @xmath6 is real analytic .
so , if @xmath6 is sufficiently smooth , it is expected that we can control the sum of @xmath85 , and obtain the dispersive type estimate for the whole operator @xmath2 ( i.e. ( [ main1 ] ) without the projection @xmath54 ) .
we hope to argue this matter elsewhere in the near future .
the paper is organized as follows . in section 2 , we review the floquet bloch theory for our operator @xmath0 and give the explicit form of the integral kernel of @xmath2 . in section 3 , we give more concrete analysis for the band functions , especially give some estimates for the derivatives . in section 4
, we prove theorem [ theorem_main ] , and give some comment for the summability with respect to @xmath43 of the estimates ( [ main1 ] ) .
in this section , we shall calculate the integral kernel of the operator @xmath2 by using the floquet - bloch theory .
most results in this section are already written in another literature ( e.g. reed - simon ( * ? ? ?
* xiii.16 ) and albeverio et .
* iii.2.3 ) ) , but we shall give it here again for the completeness .
first we shall calculate the generalized eigenfunctions for our model , i.e. , the solutions to the equations @xmath103 the condition ( [ bloch02 ] ) comes from the requirement @xmath104 , and we use the abbreviation @xmath105 in ( [ bloch03 ] ) .
[ proposition_generalized_eigenfunction ] let @xmath106 , @xmath107 , and take @xmath108 so that @xmath109
. then , the equations ( [ bloch01])-([bloch03 ] ) have a solution @xmath110 of the following form .
@xmath111 where @xmath112 and @xmath113 are constants .
when @xmath114 , we interpret @xmath115 .
the coefficients @xmath112 and @xmath113 satisfy the following recurrence relation .
@xmath116 @xmath117 the matrix @xmath118 satisfies @xmath119 and the discriminant @xmath120 is @xmath121 the proof is a simple calculation , so we shall omit it . notice that @xmath122 is an entire function with respect to @xmath123 , since @xmath19 is an even function .
next we shall calculate the _
bloch waves _ , the solution to ( [ bloch01])-([bloch03 ] ) with the quasi - periodic condition @xmath124 for some @xmath125 .
[ proposition_blochwave ] 1 . for @xmath126
, there exists a non - trivial solution @xmath127 to ( [ bloch01])-([bloch03 ] ) satisfying the bloch wave condition ( [ bloch04 ] ) if and only if @xmath128 2 .
when ( [ bloch09 ] ) holds , a solution @xmath110 to ( [ bloch01])-([bloch03 ] ) and ( [ bloch04 ] ) is given by ( [ bloch05 ] ) with the coefficients @xmath129 \(i ) it is easy to see a non - trivial solution to ( [ bloch01])-([bloch03 ] ) and ( [ bloch04 ] ) exists if and only if @xmath118 has an eigenvalue @xmath130 , and the latter condition is equivalent to ( [ bloch09 ] ) , since @xmath119 and @xmath131 .
( ii ) when ( [ bloch09 ] ) holds , the vector @xmath132 given in ( [ bloch10 ] ) is an eigenvector of @xmath118 with the eigenvalue @xmath130 .
thus the second equation in ( [ bloch10 ] ) follows from ( [ bloch06 ] ) .
proposition [ proposition_blochwave ] and the bloch theorem imply @xmath133 if and only if ( [ bloch09 ] ) holds for some @xmath125 , that is , @xmath134 as already stated in ( [ kronig - penney3 ] ) . for @xmath135
, we have @xmath136 if @xmath77 , the right hand side of ( [ bloch12 ] ) is larger than @xmath137 and ( [ bloch11 ] ) does not hold for @xmath135 .
thus , there is no negative part in @xmath22 .
if @xmath138 , then some negative value @xmath123 belongs to @xmath22 , and the corresponding @xmath139 is pure imaginary .
however , we concentrate on the high energy limit in the present paper , and the existence of the negative spectrum does not affect our argument .
so we sometimes assume @xmath77 in the sequel , in order to simplify the notation . in this case
, the results for @xmath138 will be stated in the remark . by an elementary inspection of the graph of @xmath140
, we find the following properties .
[ proposition_discriminant ] assume @xmath77 .
then , 1
. @xmath141 and @xmath142 for @xmath143 .
2 . the equation @xmath144 has a unique solution @xmath145 in the open interval @xmath146 for @xmath147 .
3 . the equation @xmath148 has a unique solution @xmath149 in the open interval @xmath146 for @xmath143 , and @xmath150 .
4 . for convenience
, we put @xmath151 .
then , @xmath19 is monotone decreasing on @xmath152 $ ] for even @xmath43 , and monotone increasing on @xmath152 $ ] for odd @xmath43 . * remark . *
when @xmath138 , we denote the solution to @xmath153 in @xmath154 by @xmath155 , and the solution to @xmath148 in @xmath154 by @xmath156 , for @xmath157 . [ cols="^,^ " , ] though the formulas ( [ der1])-([der6 ] ) are explicit , it is still not easy to obtain the precise lower bound for the derivatives of @xmath42 , especially when @xmath42 is near the band edge .
for this reason , we employ the _ puiseux expansion _ of the inverse function @xmath158 , which makes our analysis clear .
this kind of expansion is studied in the classical work by kohn @xcite .
first let us analyze the asymptotics of @xmath155 and @xmath156 given in proposition [ proposition_discriminant ] .
a related result is written in albeverio et .
* theorem 2.3.3 ) .
[ proposition_knln ] let @xmath155 and @xmath156 as in proposition [ proposition_discriminant ] and its remark .
then , @xmath159 @xmath160 we assume @xmath77 for simplicity ( in the case @xmath138 , we only need to change the sign of @xmath161 given below ) .
first we prove ( [ asymptotic_kn ] ) .
the number @xmath155 is the solution to @xmath162 put @xmath163 .
then ( [ ba02 ] ) is equivalent to @xmath164 then , for sufficiently large @xmath43 , it is easy to see that the solution to ( [ ba03 ] ) is the limit of the sequence @xmath165 given by$ ] , @xmath166 , @xmath167 and assume @xmath168 and @xmath169
. then @xmath170 has the unique fixed point in @xmath171 which is the limit of the sequence ( [ ba04 ] ) ' .
if we take @xmath170 as in ( [ ba03 ] ) and @xmath172 $ ] , we can apply the contraction mapping theorem for sufficiently large @xmath43 . ]
@xmath173 by a simple calculation , we have @xmath174 then twice substitution of ( [ ba05 ] ) into ( [ ba04 ] ) gives the formula ( [ asymptotic_kn ] ) ( three times substitution gives the same formula ) .
next we prove ( [ asymptotic_ln ] ) .
put @xmath175 .
then the defining equation of @xmath156 @xmath176 is equivalent to @xmath177 then we can obtain the formula ( [ asymptotic_ln ] ) by using the following expansion recursively .
@xmath178 in order to calculate the puiseux expansion of @xmath179 , first we calculate the taylor expansion of @xmath140 .
[ proopsition_taylor_dk ] the taylor expansion of @xmath140 near @xmath149 is given as follows .
@xmath180 @xmath181,\\ \label{d2 } d_2 & = & ( -1)^n\left[-1 - \left(v+\frac{v^2}{8}\right)(n\pi)^{-2}+o(n^{-4})\right],\\ \label{d3 } d_3 & = & ( -1)^n\left [ \left(v+\frac{v^2}{6}\right)(n\pi)^{-3}+o(n^{-5})\right],\\ \label{d4 } d_4 & = & ( -1)^n\left[\frac{1}{12}+ \left(\frac{v}{6}+\frac{v^2}{96}\right)(n\pi)^{-2}+o(n^{-4})\right],\\ \label{d5 } d_5 & = & ( -1)^n\left [ -\left(\frac{v}{6}+\frac{v^2}{60}\right)(n\pi)^{-3}+o(n^{-5})\right ] , \\
\label{d6 } d_6 & = & ( -1)^n\left[-\frac{1}{360}- \left(\frac{v}{120}+\frac{v^2}{2880}\right)(n\pi)^{-2}+o(n^{-4})\right].\end{aligned}\ ] ] notice that @xmath182 by definition .
the results ( [ yn])-([d5 ] ) can be obtained by substituting the following formulas into ( [ d0k])-([d6k ] ) . from ( [ asymptotic_ln ] )
, we have @xmath183 next we shall calculate the puiseux expansion of @xmath158 near @xmath184 . since @xmath145 is the zero of order 2 of @xmath19 , @xmath158 is a double - valued function with respect to the complex variable @xmath89 .
[ proposition_puiseux_k ] the puiseux expansion of @xmath179 near @xmath184 is written as follows .
@xmath185 put @xmath186 and @xmath187 , where @xmath188 are the coefficients in ( [ taylor_dk ] ) . then the taylor expansion ( [ taylor_dk ] )
is rewritten as @xmath189 substituting the puiseux expansion ( [ puiseux_k ] ) into ( [ pk01 ] ) , and comparing the coefficient of each power @xmath190 , we have @xmath191 since the expansions of @xmath187 are obtained from ( [ d2])-([d6 ] ) , we can calculate the coefficients ( [ e1])-([e5 ] ) by solving the above equations .
* remark . * 1 .
the branch points of @xmath179 are @xmath184 , and @xmath192 is nearby @xmath193 by ( [ yn ] ) .
thus the radius of convergence of the puiseux expansion ( [ puiseux_k ] ) is about @xmath194 , for sufficiently large @xmath43 .
this fact justifies the calculus in the sequel .
\2 . in the case
@xmath195 , the expansion ( [ puiseux_k ] ) coincides with the puiseux expansion @xmath196 where @xmath43 is any integer satisfying @xmath197 . substituting @xmath198 into ( [ puiseux_k ] ) , we obtain the function @xmath199 .
since the puiseux expansion ( [ puiseux_k ] ) has two branches , we obtain two functions @xmath200 notice that @xmath161 is positive for real @xmath44 , so @xmath190 in ( [ ba07 ] ) is uniquely defined as a positive function . for notational simplicity
, we put @xmath201 then we have the following formulas , useful for the calculation below .
@xmath202 then we obtain expansions of the band functions and their derivatives near the band edge , as follows .
[ proposition_puiseux_lambda ] let @xmath203 , @xmath161 and @xmath204 as in ( [ ba07])-([zndef ] ) , and put @xmath205 .
then the following expansions hold near @xmath206 ( or @xmath207 ) .
1 . @xmath208 + @xmath209 2 .
@xmath210 3 .
@xmath211 + @xmath212 4 .
@xmath213 the proof can be done simply by substituting the expansion ( [ ba07 ] ) , ( [ asymptotic_ln ] ) and ( [ e1])-([e5 ] ) into @xmath214 and taking the derivative with respect to @xmath44 ( or @xmath204 ) repeatedly . the formulas ( [ formula1])-([formula5 ] ) help the calculation . by construction , we have @xmath215 for small @xmath204 . using ( [ ba09 ] ) and the expansion ( [ pu19 ] ) , we can find the asymptotics of the solution to @xmath71 .
[ proposition_inflection ] the equation @xmath216 has a unique solution @xmath217 for sufficiently large @xmath43 .
the asymptotics of @xmath49 is given as @xmath218 before the proof , we give a numerical result by using the explicit formulas ( [ der1])-([der6 ] ) , in figure [ figure_inflection ] .
the result shows the formula ( [ asymptotics_theta0 ] ) gives a good approximation .
near @xmath219 , @xmath220 . here
we take @xmath221 , @xmath222 , @xmath223 , and @xmath224 . , width=226 ] put @xmath220 .
we divide the interval @xmath225 $ ] into three intervals @xmath226,\quad i_2=[(n-1)\pi+\delta_n^{1/4 } , n\pi -\delta_n^{1/4}],\\ & & i_3=[n\pi -\delta_n^{1/4 } , n\pi].\end{aligned}\ ] ] for @xmath227 , we apply the first equality of ( [ ba09 ] ) with @xmath43 replaced by @xmath228 , so @xmath229 and @xmath230
. then the expansion ( [ pu19 ] ) for @xmath231 implies @xmath232 for @xmath227 , for sufficiently large @xmath43 .
moreover , the formula ( [ inner_lambda2 ] ) given later in proposition [ proposition_inner_taylor ] shows @xmath232 for @xmath233 , for sufficiently large @xmath43 . for @xmath234
, we apply the second equality of ( [ ba09 ] ) , so @xmath235 and @xmath236 .
then @xmath237 , and ( [ pu25 ] ) for @xmath238 implies @xmath239 for @xmath240 , for sufficiently large @xmath43 .
thus @xmath241 is monotone decreasing on this interval , and ( [ pu19 ] ) for @xmath242 implies @xmath243 and @xmath244 for sufficiently large @xmath43 .
, the latter fact can also be proved by the explicit value @xmath245 ; see ( ( [ pm11 ] ) below ) .
] thus there exists a unique @xmath246 $ ] with @xmath50 for sufficiently large @xmath43 , and @xmath247 .
let us find more detailed asymptotics of @xmath49 .
by ( [ pu19 ] ) for @xmath242 , @xmath248 if @xmath234 , then @xmath249 and @xmath250 by ( [ formula2 ] ) .
so , if @xmath161 is the solution to ( [ ba11 ] ) , the expansions of the coefficients ( [ pu21])-([pu24 ] ) imply @xmath251 , and again by ( [ ba11 ] ) @xmath252 thus ( [ ba12 ] ) and ( [ formula2 ] ) imply @xmath253 since @xmath254 , we have the conclusion .
let us prove @xmath42 is similar to the parabola @xmath67 near the band center , that is , for @xmath44 in the interval @xmath255.\ ] ] notice that the asymptotics ( [ asymptotics_theta0 ] ) implies @xmath49 is not in the interval ( [ inner06 ] ) for large @xmath43 .
[ lemma_taylor_inner ] let @xmath256 given in ( [ ba01 ] ) , and @xmath44 in the interval ( [ inner06 ] ) . then , we have the following expansion .
@xmath257 * remark .
* the explicit forms of the derivatives are given in ( [ d0k])-([d6k ] ) .
the taylor expansion of @xmath179 around @xmath258 is @xmath259 substituting the equality @xmath260 into ( [ ba15 ] ) and ( [ ba16 ] ) , we obtain the conclusion since @xmath261 where we used ( [ dkformula2 ] ) .
[ proposition_inner_taylor ] let @xmath43 be a sufficiently large integer .
then , for @xmath44 in the interval ( [ inner06 ] ) , we have the expansion of @xmath262 given in ( [ ba01 ] ) as follows .
@xmath263 moreover , for @xmath264 we have @xmath265 let us rewrite the formula ( [ taylor_k_inner ] ) as @xmath266 when @xmath44 is in the interval ( [ inner06 ] ) , we have @xmath267 thus @xmath268 thus the first three terms in ( [ inner15 ] ) coincide the formula ( [ inner_k0 ] ) .
let us show that the remainder term @xmath269 is negligible . by the differentiation of the inverse function , we can prove that @xmath270 the polynomial part is bounded uniformly with respect to @xmath43 . by the expansion ( [ puiseux_k ] ) and the monotonicity of @xmath271
, we see that @xmath271 satisfies @xmath272 and @xmath273 put @xmath274 @xmath275 , then @xmath276 lies between @xmath262 and @xmath277 , and by ( [ inner20])-([inner22 ] ) @xmath278 thus , we once have a rough estimate @xmath279 so the equation ( [ inner_k0 ] ) holds with the worse remainder term @xmath280
. however , this conclusion implies @xmath281 which also implies @xmath282 thus we have @xmath283 and ( [ inner_k0 ] ) holds .
other estimates can be obtained by differentiating the formula ( [ taylor_k_inner ] ) .
then we find the estimate for the remainder term becomes worse by the power @xmath284 per one differentiation .
for example , the leading term in the remainder in ( [ inner17 ] ) is up to constant multiple @xmath285 differentiating this term , we get @xmath286 and the result is worse than @xmath287 by @xmath284 . as for @xmath269 , @xmath288 for the first term of ( [ inner25 ] ) , one @xmath289 in the numerator changed into @xmath290 by differentiation , and the estimate becomes worse by @xmath284 , because of ( [ inner16 ] ) .
for the second term , that ` @xmath291 turned into @xmath292 ' makes two @xmath293 in the denominator ( see ( [ inner20 ] ) ) , one of which cancels with @xmath294 appeared next . thus the estimate also becomes worse by @xmath284 , in total .
we can treat the other remainder terms similarly
. we shall give the bound for the amplitude function in ( [ bloch16 ] ) , that is , @xmath295 where @xmath296 and @xmath297 given in ( [ ba01 ] ) .
[ proposition_amplitude1 ] let @xmath298 given in ( [ amp01 ] ) . 1 . the function @xmath299 is bounded uniformly with respect to @xmath126 , @xmath296 and @xmath143 .
2 . put @xmath300 , @xmath301 , and @xmath302,\quad i_2 = [ ( n-1)\pi + \delta_1 , ( n-1)\pi + \delta_2],\\ & & i_3 = [ ( n-1)\pi + \delta_2 , n\pi - \delta_2],\quad i_4 = [ n\pi - \delta_2 , n\pi - \delta_1],\quad i_5 = [
n\pi - \delta_1 , n\pi].\end{aligned}\ ] ] for sufficiently large @xmath43 , the derivative @xmath303 obeys the following bound @xmath304 where @xmath74 is a positive constant independent of @xmath44 , @xmath88 , @xmath89 , and @xmath43 .
especially , @xmath305 ) } = \int_{(n-1)\pi}^{n\pi}|a_n'(\theta , x , y)|d\theta\ ] ] is bounded uniformly with respect to @xmath43 , @xmath88 , @xmath89 .
\(i ) first we prove @xmath306 is bounded uniformly with respect to @xmath43 and @xmath307 $ ] . for @xmath308 , we have @xmath309 , and @xmath310 by ( [ formula2 ] ) .
then we have by the expansions ( [ taylor_dk ] ) and ( [ ba07 ] ) @xmath311 let @xmath312 as in ( [ zndef ] ) . since @xmath313 , @xmath314 thus @xmath315 and the right hand side
is uniformly bounded .
we can prove @xmath316 is uniformly bounded for @xmath317 in a similar way . for @xmath234 , the bounds ( [ inner16 ] ) and ( [ inner24 ] ) hold , so ( [ inner16 ] ) holds even if @xmath44 is replaced by @xmath139 .
then @xmath318 thus @xmath316 is uniformly bounded on all the intervals @xmath319 .
similarly we can prove @xmath320 and @xmath321 are uniformly bounded .
the remaining factors are clearly bounded , so we have the conclusion .
\(ii ) it is sufficient to show the three functions @xmath322 obey the bound ( [ bound_amplitude ] ) , since the derivatives of the remaining factors are bounded . first ,
by ( [ d1k ] ) , ( [ d2k ] ) and ( [ bloch17.5 ] ) @xmath323 for @xmath324 , the expansion ( [ amp02 ] ) implies @xmath325 for @xmath326 , by ( [ formula1 ] ) @xmath327 thus ( [ amp04 ] ) , ( [ amp05 ] ) and ( [ amp06 ] ) imply @xmath328 for @xmath329 . for @xmath330
, we have @xmath331 and by ( [ amp021 ] ) @xmath332 for large @xmath43 .
thus ( [ amp04 ] ) , ( [ amp05 ] ) and ( [ amp07 ] ) imply @xmath333 for @xmath334 , for some positive constant @xmath74 independent of @xmath43 . for @xmath234 , we have by ( [ amp03 ] ) @xmath335 so ( [ amp04 ] ) implies @xmath336 for @xmath234
. similarly we can prove @xmath337 for @xmath233 and @xmath338 for @xmath227 , thus @xmath339 obeys the bound ( [ bound_amplitude ] ) .
next , we shall estimate @xmath340 . by ( [ dkformula1 ] ) , ( [ d0k])-([d2k ] ) and ( [ bloch17.5 ] )
, we have @xmath341 this equality and ( [ amp02 ] ) , ( [ amp06 ] ) , ( [ amp07 ] ) and ( [ inner16 ] ) ( with @xmath44 replaced by @xmath139 ) gives the same conclusion for @xmath340 ( actually , we obtain a bit faster decay for @xmath342 ) . finally , by ( [ dkformula1 ] ) , ( [ d0k])-([d2k ] ) and ( [ bloch17.5 ] ) , @xmath343 this estimate is the same as ( [ amp04 ] ) .
so the same conclusion holds for @xmath344 .
the @xmath1 norm of the operator @xmath345 is just the supremum with respect to @xmath346 of the absolute value of the integral kernel @xmath347 given in ( [ bloch16 ] ) . put @xmath37+x'$ ] ,
@xmath38+y'$ ] , @xmath39,[y]\in \mathbf{z}$ ] , @xmath348 , and @xmath41-[y])/t$ ] , then ( [ bloch16 ] ) is rewritten as @xmath349 where @xmath299 is the amplitude function given in ( [ amp01 ] ) .
the following lemma , taken from stein s book @xcite , gives the decay order of the oscillatory integral with respect to @xmath83 .
[ lemma_stein ] let @xmath350 be a finite open interval and @xmath351 .
let @xmath352 , and assume @xmath353 if @xmath354 , we additionally assume @xmath355 is a monotone function on @xmath171 .
let @xmath356 and assume @xmath357 .
then we have @xmath358 for any @xmath359 , where @xmath360 .
* here , we say @xmath361 is monotone if @xmath362 ( @xmath363 ) or @xmath364 ( @xmath363 ) .
the assumption ` @xmath171 is finite ' can be removed if the integral in the left hand side converges . the proof is done by integration by parts and a mathematical induction ( see stein @xcite ) .
let @xmath49 as in proposition [ proposition_inflection ] .
let @xmath367 be the maximum of the function @xmath368 .
put @xmath369 and @xmath370 . put @xmath220
. then we have by ( [ asymptotics_theta0 ] ) and ( [ formula2 ] ) , @xmath371 since @xmath372 near the upper edge , we have by ( [ pm01])-([pm03 ] ) and ( [ pu14 ] ) ) is consistent with figure [ figure_lambda1 ] . ] @xmath373 if @xmath374 , then we have @xmath375 by the periodicity of the integrand , we can divide the integral ( [ st01 ] ) as the sum of two integrals so that @xmath376 is a monotone function on each interval
. then we can apply lemma [ lemma_stein ] with @xmath354 , and we have by ( [ pm05 ] ) and proposition [ proposition_amplitude1 ] @xmath377 where @xmath378 $ ] and @xmath74 is a positive constant independent of @xmath43 .
if @xmath379 , then we slide the interval of integration by periodicity , and divide it into four intervals @xmath380,\quad i_2 = \left [ n\pi - \delta_n^{1/3}/2 , n\pi + \delta_n^{1/3}/2 \right],\quad\\ i_3 = \left[n\pi + \delta_n^{1/3}/2 , n\pi + 2 \delta_n^{1/3 } \right],\quad i_4 = \left[n\pi + 2 \delta_n^{1/3 } , ( n+2)\pi -2 \delta_n^{1/3}\right].\end{aligned}\ ] ] put @xmath381 if @xmath227 or @xmath382 , then @xmath383 and by ( [ formula2 ] ) @xmath384 by ( [ pu19 ] ) , @xmath385 ( [ pm06 ] ) and ( [ pm07 ] ) imply ) and ( [ pm09 ] ) in figure [ figure_inflection ] . ]
@xmath386 next , for @xmath387 , we have from ( [ formula2 ] ) @xmath388 and by ( [ pu25 ] ) @xmath389 for sufficiently large @xmath43 , since @xmath250 and @xmath390 .
thus @xmath391 is monotone on the left half of @xmath392 .
this fact and ( [ pm09 ] ) imply @xmath393 for sufficiently large @xmath43 , for some positive constant @xmath74 independent of @xmath43 .
moreover , ( [ pm08 ] ) , ( [ pm10 ] ) and ( [ inner_lambda2 ] ) imply @xmath394 is also uniformly bounded from below on @xmath395 .
then lemma [ lemma_stein ] implies @xmath396 and @xmath397 is @xmath87 , uniformly with respect to @xmath43 .
moreover , for @xmath398 , we have @xmath399 , and ( [ formula2 ] ) and ( [ pm10 ] ) imply @xmath400 for sufficiently large @xmath43 , where @xmath74 is a constant independent of @xmath43 .
thus lemma [ lemma_stein ] implies @xmath401 and the conclusion holds .
we conclude the paper by arguing the summability of the bandwise estimates . if we fix @xmath88 , @xmath89 and take the limit @xmath93 , then @xmath402 for sufficiently large @xmath43 , by ( [ pm04 ] ) ( see also figure [ figure_lambda1 ] ) .
so there always exists the stationary phase point @xmath46 ( the solution to @xmath47 ) nearby @xmath403 , for sufficiently large @xmath43 . for simplicity ,
assume @xmath77 in the sequel .
when @xmath206 , we have by the formulas ( [ der1])-([der6 ] ) , ( [ dkformula2 ] ) , ( [ d1k ] ) and ( [ d2k ] ) @xmath404 thus , even if we cut out a small interval @xmath70 around @xmath206 , the bound for the integral over @xmath70 is not better than @xmath91 , which is not summable with respect to @xmath43 .
however , we already know the sum @xmath405 converges in the strong operator topology in @xmath406 .
thus it seems that the sum of the integral kernels converges only conditionally .
we have to analyze the cancellation between the integral kernels for two adjacent bands more carefully , in order to obtain a better estimate .
we hope to argue this subject in the future . | in this paper we consider the 1d schrdinger operator @xmath0 with periodic point interactions .
we show an @xmath1 bound for the time evolution operator @xmath2 restricted to each energy band with decay order @xmath3 as @xmath4 , which comes from some kind of resonant state .
the order @xmath3 is optimal for our model .
we also give an asymptotic bound for the coefficient in the high energy limit .
for the proof , we give an asymptotic analysis for the band functions and the bloch waves in the high energy limit .
especially we give the asymptotics for the inflection points in the graphs of band functions , which is crucial for the asymptotics of the coefficient in our estimate . |
one of the primary goals of the current generation of colliders is the discovery of the mechanism by which the mass scale of the weak interactions is generated and stabilized .
whether that mechanism is supersymmetry ( susy ) , technicolor , extra dimensions or something not yet considered , it will generically require a number of new particle states .
one or more of these particles are often stable , at least on experimental time scales , thanks to discrete symmetries in the model or suppressed couplings .
the archetypes for such particles are the `` lsp '' ( lightest susy particle ) in the most susy models , and the nlsp ( next - to - lightest susy particle ) in gauge - mediated models .
if a new particle produced at a collider lives long enough to escape the detector , there is no simple way to measure its lifetime . of utmost importance
will be determining whether the lifetime is indeed finite or if the particle is stable on cosmological timescales .
it has become standard lore that any stable remnant of new weak - scale physics must be , at most , weakly interacting .
the possibility that the stable particle is electrically charged or strongly interacting has long been dismissed based on a combination of astrophysical and cosmological arguments . if such a particle , often called a champ @xcite , exists in abundance in our galaxy , then a steady flux of champs would have been captured by the earth over the course of our planet s lifetime
. such stable , charged particles would form hydrogen - like atoms which could in turn form superheavy water molecules . these superheavy atoms and molecules
would collect in our oceans in significant abundances .
searches for superheavy isotopes of water in ocean water have all yielded negative results @xcite .
these searches have effectively ruled out charged dark matter at the weak scale .
this result is obviously dependent on the flux of champs onto the earth .
there are two theoretical avenues for determining this flux .
the first assumes that these relics represent the majority component of the galactic dark matter halo ; then their local abundance can be estimated directly from astrophysical data . from this
we easily deduce the rate at which the champs would be captured and accumulate in our oceans .
the second approach is to use standard cosmological assumptions and techniques to calculate the relic abundance of the champs from first principles . then using calculations of dark matter clumping
, a local density can be estimated .
these calculations also predict far greater abundances of superheavy water than are observed experimentally , even when the abundance is too small to account for the dark matter .
there has been recent interest in whether charged stable relics are ruled out in more cosmologies . motivated in part by theories with large extra dimensions @xcite , studies have been done to calculate relic abundances for the case of extremely low reheat temperatures ( as low as an mev ) .
the surprising result of these studies is that particles with tev masses can still be created during the reheat process even if the final temperature is around the mev scale @xcite .
when applied to electrically charged relics , a bound of roughly 1 tev on the relic s mass can be deduced using experimental search bounds @xcite .
but can we bound , exclude or search for stable , charged relics without recourse to cosmology ? in some models , physics near the tev scale is so non - canonical that it is unclear whether existing cosmological bounds apply at all ( see for example the model of ref .
@xcite ) . in this paper
, we will approach the same problem but from an original point of view .
we will consider charged relics created by cosmic rays in the earth s upper atmosphere . in order to be specific
, we will consider susy as our template model ; however extension of these bounds to any other model of weak - scale physics with a stable charged relic would be trivial .
we will not place a bound on the mass of the susy relic itself .
instead we will place a bound on the susy scale , or more specifically , the masses of the squarks and gluinos . direct production of the charged relic in cosmic ray collisions
may be very small ( for example , the relic could be a stau with only small electromagnetic cross - sections ) .
however , production of any susy state would , after a cascade of decays , result in one or more of the charged lsps .
thus the production cross - section will generally be dominated by production of squarks and gluinos , not the lsp itself .
none of these results depend in any way on new models which attempt to explain cosmic rays at the very highest energies .
these results are generated using the usual spectrum of incident protons and heavy nuclei .
our only assumption is that this spectrum has remained essentially unchanged over the last 3 billion years of the earth s history .
while the energy frontier in particle accelerators is slowing moving higher and higher , collisions with center - of - mass energies in the tev range have been common throughout the history of the earth in the form of cosmic rays scattering off atmospheric nuclei .
while the vast majority of these cosmic rays lose their energy through one or more hard qcd interactions , a small fraction can undergo new - physics interactions . with energies ranging up to @xmath1 ( in the earth s frame ) , the incident cosmic rays are capable of generating center - of - mass energies above @xmath2 when scattering off a proton . and with the exception of the very highest energies , the incident cosmic ray energy spectrum is well - measured . while these weak interactions in the upper atmosphere are useless for studying short - lived susy states , such states will be produced nonetheless . and
if r - parity is conserved , those susy states will eventually decay down to the lsp .
thus lsps are produced in our atmosphere in standard particle physics processes , independent of cosmological or astrophysical assumptions the all - particle spectrum for cosmic rays is well - measured up to energies of about @xmath1 . at energies up to about @xmath3
the spectrum follows a simple @xmath4 power law .
around @xmath3 , the so - called `` knee '' is hit at which point the spectrum begins falling more dramatically , as @xmath5 .
finally , at energies around @xmath6 ( the `` ankle '' ) the spectrum flattens out slightly to @xmath7 @xcite .
the composition of the incident cosmic rays is not as well known .
the primary spectrum is composed of a number of elements including protons , helium , iron , etc . for energies below the `` knee '' ( @xmath8 per nucleus ) , protons are the most abundant constituents , representing roughly 50% to 80% of the spectrum . at the knee ,
iron nuclei begin to dominate the spectrum ; however , only qualitative details of the composition can be inferred using , for example , the depth of shower maximum .
the data seems to indicate that there is a relative rise in heavier elements and then a gradual decrease again as the `` ankle '' is approached @xcite . above @xmath9 per nucleus
, lighter elements appear to be more abundant again , consistent with fragmentation of the heavier elements by the cmbr .
extractions from the data of the average particle number as a function of primary energy vary from 5 to 15 for different experiments @xcite . given this level of ignorance , we assume a fixed fraction of protons ( 50% ) at energies below @xmath10 , exponentially decreasing to a constant 1% at energies above @xmath11 .
we will also assume that the remainder of the primary composition ( the part that is not protons ) has on average @xmath12 nucleons / incident nucleus ; we will choose @xmath13 as a typical experimental value @xcite .
we will model these nuclei as a collection of @xmath12 loosely bound nucleons , each carrying @xmath14 of the incident energy @xmath15 of the entire nucleus ; in the center - of - mass frame , the total energy of the collision , @xmath16 , is therefore reduced by a factor @xmath17 . to be conservative , we assume that any one nuclei can only participate in a single hard interaction and we do not consider the interactions of the daughter nuclei , if any are formed , at all .
we find that nuclei heavier than hydrogen are responsible for only about 10% of the susy interactions and therefore our result is almost independent of how we model their interactions .
we believe the above choices to be very conservative .
looser constraints with more protons in the spectrum could allow our calculated rates to more than double .
protons incident on our atmosphere will usually lose their energy either by qcd or qed processes .
typical qed energy loss rates through brehmstrahlung and ionization are about 2 mev/(g/@xmath18 ) @xcite so that charged particles lose only a few @xmath19 traversing 15 km of atmosphere .
this is negligible compared to losses in hard qcd interactions .
energetic primary protons have a nuclear interaction length of @xmath20g/@xmath18 , corresponding to 12 qcd interactions over the depth of the atmosphere @xcite . for simplicity ,
we only consider the first such interaction and neglect secondaries produced in the resulting air showers .
that is , we assume that once an incident proton has had a single hard qcd interaction , its energy is degraded beyond the point at which it is kinematically possible to create susy particles . in effect
, we are discarding the possibility that a secondary could participate in a susy interaction , a choice which again results in a conservative bound on the number of susy states produced .
a complete analysis , unneccesary here , involves solving a set of coupled cascade equations with appropriate boundary conditions , taking into account energy loss processes and energy - dependent cross sections .
our method reduces to analyzing a fixed - target @xmath21 collision with one of the protons at very high energies .
now we present the details of our calculation . after passing through a distance @xmath22 of the atmosphere with local density @xmath23 , the flux of protons with energies between @xmath15 and @xmath24 which have undergone hard qcd interactions and are therefore `` lost '' for susy interactions is simply given by @xmath25 where @xmath26 is the flux of protons measured in gev@xmath27@xmath28s@xmath27sr@xmath29 , @xmath30 is the number density of protons in the atmosphere at a depth @xmath31 , and @xmath15 is the energy of the incident proton in the earth s frame . at the energies under consideration here , the qcd @xmath21 cross - section , @xmath32 , is roughly 100 mb , independent of energy . over that same slice of atmosphere ,
the total number of susy interactions ( integrated over all energies ) , corresponding to the flux of lsps , is @xmath33 here @xmath34 represents the inclusive cross - section for @xmath35 , independent of whether @xmath36 is squarks , sleptons , etc .
( @xmath34 is a function of @xmath37 . )
this system of equations is simple to solve . for a spectrum of only protons ,
the total number of susy interactions in @xmath28s@xmath27sr@xmath29 is then : @xmath38 where @xmath26 and @xmath34 are implicitly functions of @xmath15 . accounting for the observation that only a fraction @xmath39 of the incident cosmic rays are protons , and using our stated assumption that the remaining cosmic rays contain on average @xmath12 nucleons , then the above discussion generalizes to : @xmath40\ ] ] where @xmath41 is the all - particle incident flux of cosmic rays .
there is another , simpler way to understand the above calculation . since we only allow each incident cosmic ray one hard interaction ( and only 1 lsp per susy interaction ) , we can simply calculate the probability that that one interaction will be either qcd or susy .
since @xmath42 , the probability of a susy interaction is then roughly @xmath43 .
that probability is then integrated over the entire flux of incident cosmic rays to find the flux of lsps produced .
the only unknown quantity that remains is @xmath34 .
since this represents a total cross - section , summed over all possible susy final states , there is considerable room for model - dependence .
however , we can make several simplifying observations / assumptions .
first , because the cosmic rays and the atmospheric nuclei are baryonic , strong interactions should dominate the susy production processes .
second , in most `` realistic '' susy models , there is an approximate degeneracy among the strongly - interacting sparticles , that is , among the squarks and the gluinos .
these two statements allow us to greatly simplify the calculation by only considering superqcd interactions and by assuming that all squarks and gluinos have a common mass @xmath44 .
independent of the details , the parton - level cross - section for susy production obviously goes like @xmath45 . to be slightly more realistic , we will use actual calculated cross - sections for the range of superqcd processes . the processes that we consider
are : @xmath46 the relevant cross - sections have been calculated and tabulated in ref .
@xcite ; we confine our calculation to tree level .
to go to the proton - proton cross - section we use the cteq5 m parton distribution functions @xcite .
the total @xmath47susy cross - section is shown in figure [ fig1 ] as a function of @xmath16 for several choices of @xmath44 . over the energy range of interest ,
the susy cross - section is dominated by @xmath48 and @xmath49 .
there is a subtlety associated with exact versus approximate degeneracy of the squarks and gluinos and so we consider both the case in which @xmath50 and @xmath51 .
we find a difference of about 50% in our calculated number densities as we vary over this range , and so we show a range of limits .
= 3truein once a squark or gluino has been produced , it will decay in a cascade down to the the lsp .
presumably , one lsp of positive charge will be produced for each of negative charge .
searches for the latter are more difficult and we will concentrate only on the former .
further , we will assume conservatively that only one positively charged lsp is produced per susy interaction , though the number can be significantly higher as the cascades of decays progress .
one expects that these positively charged lsps form superheavy hydrogen by attracting a nearby electron and that this superheavy hydrogen eventually ( over the lifetime of the earth ) bonds into a superheavy water molecule in the earth s oceans .
so as a final step we calculate the concentration of superheavy water in the oceans . considering the age of the oceans , @xmath52 , to be roughly 3 billions years , and assuming that the flux of cosmic rays has remained essentially unchanged over that time period , the number of superheavy water molecules per usual water molecule in the oceans ( the `` anomalous concentration '' , @xmath53 ) is @xmath54 where @xmath55 is the average depth of the oceans .
the constraints on @xmath53 come from searches for superheavy molecules in large samples of water .
experiments then place bounds on @xmath53 as a function of the mass of the stable , charged particle . in our approach , this mass in unknown , though it is bounded from above by @xmath44 .
the strongest bound comes the experiment of smith @xcite , who find limits ranging from @xmath56 for @xmath57 to @xmath58 for @xmath59 . because their bound monotonically weakens as @xmath60 increases
, we can place a very conservative bound on @xmath44 by setting @xmath61 . for that case , we find @xmath62 if the lsp is stable and charged .
( our bound has an uncertainty of roughly @xmath63 due to a @xmath64 uncertainty that comes from the subtleties for defining @xmath44 discussed above . )
these results are summarized in figure [ fig2 ] where we have shown the predicted anomalous concentration as a function of @xmath65 and the experimental bound of ref .
@xcite .
we have shown that there exist bounds from cosmic ray production in the upper atmosphere on charged stable relics ( a charged susy lsp in particular ) which are independent of cosmological constraints . under the assumption that the incident cosmic ray flux has remained constant over the last 3 billion years
, we have calculated a conservative lower bound on the scale of new physics ( @xmath44 ) , using nothing more than standard particle physics .
if susy has a stable and charged lsp , then we can place a lower bound on the mass scale of the squarks and gluinos at 230 gev .
this procedure can easily be extended to other models of weak - scale physics .
in such a case , limits similar to those found here could be placed on the masses of new strongly - interacting particles .
importantly , these bounds will not change if the basic paradigms of cosmology at and below the weak scale are questioned , such as happens in models with large extra dimensions .
for example , it is unclear how standard cosmological bounds can be used to constrain models such as that of ref .
@xcite which predicts a light , stable top squark but becomes non - perturbative and higher - dimensional at the tev scale .
the bound presented here should hold even in these highly non - standard cases .
finally , if a charged particle is discovered which is stable on collider timescales , ruling out or verifying that it is stable on cosmological timescales will require that searches for superheavy water be examined again , though with larger initial samples .
unfortunately , the steeply falling cosmic ray spectrum requires us to go to exponentially larger samples in order to significantly increase our sensitivity .
for example , the method used in ref .
@xcite would require an initial heavy water sample equal to that contained in the sno experiment in order to probe squark masses up to 700 gev .
while such a large - scale search seems unnecessary at present , any future discovery of a new stable , charged particle might require just such an effort .
otherwise there may be no other way to study the stability of that state on long timescales .
ck would like to thank n. arkani - hamed , g. domokos , l. hall , h. murayama and j. poirier for enlightening conversations .
this research was supported by the national science foundation under grant phy00 - 98791 .
99 a. de rujula , s. l. glashow and u. sarid , nucl .
b * 333 * , 173 ( 1990 ) ; + s. dimopoulos , d. eichler , r. esmailzadeh and g. d. starkman , phys .
d * 41 * , 2388 ( 1990 ) .
p. f. smith and j. r. bennett , nucl .
b * 149 * , 525 ( 1979 ) ; + t. k. hemmick _ et al .
_ , phys .
d * 41 * , 2074 ( 1990 ) ; + p. verkerk , phys .
lett . * 68 * , 1116 ( 1992 ) ; + t. yamagata , y. takamori and h. utsunomiya , phys .
d * 47 * , 1231 ( 1993 ) .
f. smith , nucl .
b * 206 * , 333 ( 1982 ) .
n. arkani - hamed , s. dimopoulos , n. kaloper and j. march - russell , nucl .
b * 567 * , 189 ( 2000 ) .
d. j. chung , e. w. kolb and a. riotto , phys .
d * 60 * , 063504 ( 1999 ) ; + g. f. giudice , e. w. kolb and a. riotto , phys
. rev .
d * 64 * , 023508 ( 2001 ) .
a. kudo and m. yamaguchi , phys .
b * 516 * , 151 ( 2001 ) .
r. barbieri , l. j. hall and y. nomura , phys .
d * 63 * , 105007 ( 2001 ) .
d. j. bird _
[ hires collaboration ] , phys .
lett . * 71 * , 3401 ( 1993 ) .
data from a variety of original experiments has been compiled in p. greider , _ cosmic rays at earth _ , elsevier , amsterdam ( 2001 ) . | supersymmetric models often predict a lightest superpartner ( lsp ) which is electrically charged and stable on the timescales of collider experiments .
if such a particle were to be observed experimentally , is it possible to determine whether or not it is stable on cosmological timescales ? charged , stable particles are usually considered to be excluded by cosmological arguments coupled with terrestrial searches for anomalously heavy water molecules .
but when the cosmology is significantly altered , as can happen in models with large extra dimensions , these arguments are in turn significantly weakened . in this paper we suggest an alternate way to use searches for superheavy water to constrain the lifetimes of long - lived , charged particles , independent of most cosmological assumptions . by considering susy production by cosmic rays in the upper atmosphere , we are able to use current bounds on superheavy water to constrain the mass scale of squarks and gluinos to be greater than about @xmath0 , assuming a stable
, charged lsp .
this bound can be increased , but only by significantly increasing the size of the initial water sample tested .
@=11 caption#1[#2]#3 @=12 .7ex .7ex gev mev ev kev tev m_z m_pl _ _ fc ^-1 ^-1 * d * * d^ * * u * * u^ * * y_d * * y_d^ * * y_u * * y_u^ * * v * * v^ * * v^0 * * v^0 * * x * b^0-|b^0 # 1#1 | # 1| # 1 * m * m_susy m_unif # 1#2#3nucl . phys .
* b#1 * ( 19#2 ) # 3 # 1#2#3phys . lett . * b#1 * ( 19#2 ) # 3 # 1#2#3phys . lett .
* # 1b * ( 19#2 ) # 3 # 1#2#3phys . rev .
* d#1 * ( 19#2 ) # 3 # 1#2#3phys .
. lett . *
# 1 * ( 19#2 ) # 3 # 1#2#3phys . rep . *
# 1 * ( 19#2 ) # 3 # 1#2#3ann .
rev .
. astrophys . *
# 1 * ( 19#2 ) # 3 # 1#2#3ann .
rev .
nucl . part .
sci . * # 1 * ( 19#2 ) # 3 # 1#2#3mod .
. lett . *
# 1 * ( 19#2 ) # 3 # 1#2#3zeit .
fr physik * c#1 * ( 19#2 ) # 3 # 1#2#3ap .
j. * # 1 * ( 19#2 ) # 3 # 1#2#3ann .
phys . * # 1 * ( 19#2 ) # 3 # 1#2#3rev .
mod .
phys . *
# 1 * ( 19#2 ) # 3 # 1#2#3comm .
math .
phys . *
# 1 * ( 19#2 ) # 3 _ i.e. _ _ et al . _
_ e.g. _ _ etc .
_ _ c.f .
_ /#1/ |#1 # 1#1 february 2002 + * bounds on charged , stable superpartners from cosmic ray production * + 0.5 cm _ department of physics , university of notre dame + notre dame , in 46556 , usa +
_ .5 cm |
in field theories we often encounter infrared ( ir ) divergences of different kind .
some of them have physical meaning while others are just artifacts of the perturbation theory .
a common feature , however , is that the ir divergences appear at each order of perturbation theory , usually being more and more severe at higher loop orders . in order to consistently define the theory ,
these ir divergences must be summed up .
2pi resummations provide a consistent resummation framework known for a long time @xcite .
the basic idea is to replace the free propagator in the perturbation theory with the exact one which is approximated self - consistently with fixed - loop skeleton diagrams .
the so - defined perturbation theory is renormalizable @xcite-@xcite , and can be applied to study different physical questions from non - equilibrium @xcite , @xcite , thermodynamics @xcite , @xcite,@xcite,@xcite and different systems like o(n ) model @xcite , @xcite or gauge theories @xcite .
although the 2pi approximation is constructed by physical arguments , and we expect better results ( ie .
closer to the exact one ) after 2pi resummation , a priory it is not sure that one really achieves this goal .
probably the finite lifetime effects are well represented by 2pi resummation both in equilibrium @xcite as well in non - equilibrium , where the 2pi is close to the boltzmann - equation approximation @xcite .
but if the deep ir regime is important where multi - particle excitations also play crucial role , the picture is far to be so clean . to make the case even worse , in most gauge theory models
there is hard to make exact statements about the ir behavior of the model . in this paper
we aim to study the bloch - nordsieck model , which is an exactly solvable 3 + 1d gauge theory @xcite .
it is the eikonal approximation model of the qed , and one can argue @xcite that in the deep ir regime it describes correctly qed .
therefore it is used to estimate ir properties of qed from this model , for example the soft photon creation processes @xcite or finite temperature electron damping @xcite .
this model is therefore a motivated case study where the accuracy of the 2pi resummation can be assessed .
we therefore perform a series of approximations for the bloch - nordsieck model : a one - loop perturbation theory , a 2pi resummation and finally the solution of the schwinger - dyson equations with an ansatz for the vertex function . in this model
all levels of the approximations can be treated analytically .
we show that the last method is exact in the model under consideration although that is not expected in general .
this observation , however , leads us to a proposal how the 2pi resummation can be improved in a generic model in order to catch the ir physics correctly .
the structure of the paper is as follows .
we first overview the standard solution of the bloch - nordsieck propagator at zero temperature in section [ sec:1 ]
. then we compute the one loop level fermion propagator in section [ sec : oneloop ] .
next , in section [ sec:2pi ] we write up the 2pi equations for this model , perform renormalization on that , and give the analytical solution as far it can be done
. then we study the problem also numerically , determine the fermion propagator and compare the result with the exact one .
finally , in section [ sec : sd ] we study the truncated schwinger - dyson equations , renormalize it , and show that for the bloch - nordsieck model the so - defined approximation scheme is exact . for conclusion ( section [ sec : conc ] )
we summarize our approach again and speculate about the possible generalizations .
the bloch - nordsieck model is a simplification of the qed lagrangian where the dirac matrices @xmath0 are replaced by a four - vector @xmath1 @xmath2 the singled - out four vector @xmath3 represents the velocity of the rest frame of the fermion .
the fermion wave function here has only one component and @xmath4 .
we are interested in the fermion propagator which reads in the path integral representation @xmath5}\psi(x)\bar\psi(0).\ ] ] at the tree level it reads : @xmath6 since it has a single pole , there is no antiparticles in the model , and also the feynman propagator is the same as the retarded propagator .
the lack of antiparticles also means that all closed fermion loops are zero . in a closed loop @xmath7 ,
therefore the product of theta functions is zero almost everywhere . ] . as a consequence
the photon self - energy is zero , the free photon propagator is the exact one . in feynman gauge
therefore the _ exact _ photon propagator is @xmath8 now we shortly review the exact solution for the fermion propagator , cf .
@xcite , @xcite .
we first define the gauge field dependent propagator : @xmath9}\psi(x)\bar\psi(0).\ ] ] this satisfies the schwinger - dyson equation @xmath10 we will need the solution in an exponential form for @xmath11 , and this is achieved by introducing the auxiliary quantity @xmath12 which satisfies @xmath13 by integration of the above equation by @xmath14 and assuming @xmath15 ( for which we need a convergence factor @xmath16 ) we see that @xmath17 we perform fourier transformation and separate the free time dependence @xmath18 , then we obtain @xmath19 the linear @xmath20 behavior is the consequence of the linearity in the kernel .
if the kernel is non - linear or it is not scalar ( has a matrix structure ) then this form is not true any more . from here an inverse fourier transformation yields @xmath21 once we have an exponential representation for the background - dependent propagator , we can perform the gaussian @xmath11-integration . as a result we obtain in the exponent the factor @xmath22
this integral is uv divergent ; in dimensional regularization one finds the result @xmath23 where @xmath24 and @xmath25
. then the fermion propagator reads @xmath26 where @xmath27 .
this is uv divergent which means that we need a wave function renormalization .
the renormalized propagator reads @xmath28 where @xmath29 is a finite quantity .
we can determine the discontinuity of this formula , for simplicity choosing @xmath30 : @xmath31 with this spectral function the sum rule @xmath32 , which is the consequence of the equal time anticommutation relations , can not be fulfilled , since the integral is divergent .
this divergence should be compensated with the choice @xmath33 , but then we are faced with a @xmath34 expression . therefore one should always use a regularized version of the spectral function ( or propagator ) , maintaining the sum rule , and only at the end of the calculation is one allowed to release the regularization .
the lagrangian is lorentz - invariant in the sense that we must also transform @xmath35 .
so we can choose a lorentz - transformation where @xmath36 .
if @xmath1 is a 4-velocity then @xmath37 ; if it is of the form @xmath38 , then it is @xmath39 .
after rescaling the field @xmath40 and the mass as @xmath41 , the lagrangian reads @xmath42 this lagrangian will be used mostly in this work later . if necessary , the complete @xmath35 dependence is easily recoverable .
the goal of our investigations is to see , how the different levels of resummations improve the result .
thus first we start with the one loop perturbation theory . here
we need the renormalized lagrangian ; in feynman gauge it reads ( using the fact that the photon self - energy is zero ) : @xmath43 for the fermion self - energy the one loop diagram is the bubble with the contribution : @xmath44 moreover we have wave function and mass renormalization counterterms @xmath45 the complete one loop self - energy is @xmath46 . in the calculation we have to take care of the non - standard form of the free fermion propagator .
the details of the computation can be found in the appendix , as a result we obtain @xmath47,\ ] ] where @xmath24 and @xmath48 for renormalization we have to subtract the divergences with help of the counterterms , the finite parts are fixed by the renormalization scheme . in the @xmath49 scheme
we choose the counterterms like @xmath50 this results in @xmath51 the discontinuity of the renormalized self - energy reads @xmath52 for the one - loop propagator we obtain @xmath53 this is consistent with the exact result in the leading order of @xmath54 .
the spectral function @xmath55 reads @xmath56 this spectral function is normalizable , since @xmath57 on the other hand the one - loop result is not reliable when @xmath58 , ie . in the vicinity of the mass shell as well as in the large @xmath59 regime . in order to have a better description of these kinematical regimes , we need resummation of certain class of diagrams .
as it is discussed in the introduction , the next level of our approximations is the 2pi resummation .
the idea is to use the exact propagators in the perturbation theory , this propagator is determined self - consistently using skeleton diagrams as resummation patterns .
the one - loop bubble diagram in the present case generates the resummation of all the `` rainbow '' diagrams . to obtain an expression for the 2pi resummation we use the technique of @xcite
: we use the 1loop formula , interpret the appearing propagators as full propagators , and finally perform renormalization with the same _ form _ of divergent parts of the counterterms as in the 1-loop case ( the actual values will be different ) .
the tree level photon propagator is exact , therefore we can write @xmath60 using a spectral representation for the fermion propagator ( using that now the feynman propagator is the retarded one and that the fermion spectral function is @xmath61 ) we find @xmath62 from this form it is clear that we obtain the weighted one - loop result , ie .
@xmath63 in particular , if @xmath64 , then we get back the one - loop result . at this point
it is worth to examine the uv divergence structure of the 2pi approximation .
uv divergences may occur in for large values of @xmath65 : using we find that the large @xmath65 behavior of the one - loop self - energy reads : @xmath66 since @xmath67 is integrable for large @xmath65 values , therefore the @xmath68 is already finite .
therefore the divergence structure of the self - energy is @xmath69 , just like for the free case , and so the same type of counterterms are needed ( although the values are different ) .
this is a manifestation of the general case of counterterm renormalizability of 2pi resummations @xcite .
first we try to analyze with analytic methods .
we differentiate it with respect to @xmath59 to find @xmath70 since @xmath71 , we find for @xmath72 : @xmath73 to solve the equation we first should realize that the @xmath74 and @xmath75 cases are very different . if @xmath74 then @xmath76 and the propagator behaves as @xmath77 with some wave function renormalization constant @xmath78 and mass @xmath79 .
this agrees with the free case .
we also see that the integration constants correspond to the renormalization scheme ( here the wave function and mass renormalization ) .
if @xmath80 then we can redefine the variables with an arbitrary @xmath81 scale as @xmath82 then we find @xmath83 this equation does not depend on the coupling any more .
the coupling constant dependence shows up in the integration constants which are the manifestation of the renormalization scheme .
we shall also note that the equation does not give information about the sign of @xmath84 and @xmath85 , because for @xmath86 or @xmath87 the equation remains the same .
the chosen signs in turn out later to be the physical choice .
we introduce @xmath88 this means that we can write for @xmath89 : @xmath90 with an integration constant @xmath91 .
therefore @xmath92 there could appear an integration constant also here on the right hand side : @xmath93 .
but recalling that @xmath94 , we see that @xmath95 corresponds to a mass shift : if the mass remains the tree level @xmath96 then @xmath97 .
this is the ( implicit ) solution of the 2pi equations .
we see that for real @xmath85 the left hand side is real and positive , moreover for @xmath98 .
the @xmath99 part corresponds to imaginary values of @xmath85 .
since the equation itself is real , if @xmath85 is a solution , it is @xmath100 , too .
this means that the imaginary part is in fact the ( half ) discontinuity of the solution .
we see that irrespective of the value of @xmath91 , at @xmath101 , ie . on the mass shell @xmath102 and so @xmath103 finite .
this yields difficulties when we try to apply renormalization conditions on the self - energy .
namely , if we keep the mass shell unchanged ( this would correspond to the choice of @xmath95 above ) , then the renormalization of the self - energy would mean @xmath104 and @xmath105 finite .
then , however , near the mass shell the propagator should always behave as @xmath106 , ie .
_ infinite _ at the mass shell .
this means that the physical renormalization process requires @xmath107 . in this case
the propagator behaves near the mass shell as : @xmath108 because if @xmath85 is close to @xmath109 then the log term can be neglected in , and we find @xmath110 . for large values of @xmath85 ,
on the other hand , @xmath91 can be neglected .
then the integral can be evaluated as @xmath111 for large @xmath85 values it behaves as @xmath112 now let us turn to the numerical study of the system , based on @xcite and @xcite : we determine the discontinuity of the self - energy self - consistently .
the discontinuity of now reads @xmath113 knowing the discontinuity of the self - energy , we can use the kramers - kronig relation to restore the complete self - energy : @xmath114 while is a completely finite expression , in the kramers - kronig relation we will find divergences .
this corresponds to the divergences of the self - energies which must be made finite by applying the appropriate counterterms .
technically one can regularize the integral in and then make it finite with counterterms , or use the ( twice ) subtracted form of the kramers kronig relation . to see how it works , we determine the one - loop result from the tree level spectral function and the dimensional regularization of the kramers - kronig equations ( interpreting @xmath115 ) : @xmath116.\ ] ] the divergence structure is the same , and also the @xmath49 scheme result is the same as in ( the different finite parts are due to the different regularization method ) .
now we can set up an algorithm to solve .
we choose an arbitrary spectral function as a starting one ( practically the free spectral function ) , then follow the following steps : step 1 : : : compute the discontinuity of the self - energy using step 2 : : : compute the complete self - energy using the kramers - kronig relation step 3 : : : renormalize the self - energy with local counterterms . to fix the counterterms we used on - mass - shell ( om ) renormalization scheme , ie .
the real part of the self - energy at the mass shell is zero and its derivative is also zero @xmath117 we note here that releasing the first condition yields a mass shift , releasing the second condition yields a finite wave function renormalization .
but in all renormalization schemes it will remain true that near the ( renormalized ) mass shell the propagator behaves as @xmath118 .
step 4 : : : construct the new spectral function from the discontinuity of the propagator knowing the real and imaginary part of the self - energy as @xmath119 step 5 : : : continue with step 1 until the process converges .
integrations in the above algorithm are performed numerically .
this strategy was applied successfully for the @xmath120 model in @xcite .
the direct application of this strategy , however , this times fails .
numerically what we can observe is that the spectral function becomes more and more shallow , and pointwise it goes to zero @xmath121 . in order to see a convergence
, we had to use a supplementary step in the iteration after step 4 : step 4 : : : use a rescaling of the generated spectral function : @xmath122 with appropriate @xmath11 and @xmath123 which can ensure convergence .
the appropriate values can be found by inspection , but the actual values are not too important ( we used @xmath124 and @xmath125 in our numerics ) . in this way
finally we succeeded to see convergence in the spectral function .
the numerical reason of this behavior is that the exact spectral function has a discontinuity at the mass shell , and
apart from this single point
it has always negative derivative .
numerically , however , we can not have a jump , since in all regularizations equation yields @xmath126 where @xmath127 .
since the exact curve starts to bend downwards , the recursion tries to lower the spectral function in order to have smaller derivative near the mass shell .
since the spectral function has to be positive , these requirements can be satisfied only with @xmath128 . with the continuous rescaling we can achieve that
the numerically badly conditioned part , the vicinity of the mass shell , becomes smaller and smaller .
the numerical results can be seen on figure [ fig:1 ] .
the expected asymptotics can be nicely identified on the calculation ( cf .
[ fig:2 ] ) .
this also proves implicitly that the strategy to resolve the aforementioned numerical problem with the 2pi equation was correct .
if we compare the 1-loop , the 2pi and the exact results we see that there is not too much improvement .
the ir problem near the mass shell which made the 1-loop calculation unreliable , _ seems _ to be cured , but in fact the result is not closer to the exact one as the one - loop result .
the physics of the deep infrared photons can not be described by the 2pi approximation .
the next level of the approximations is based on the schwinger - dyson equations . for the bloch - nordsieck model in feynman gauge
it can be written as @xmath129 where @xmath130 is the vertex function .
for the vertex function there is another exact equation , coming from the current conservation .
this results in the ward - identity analogous to the qed case @xcite : @xmath131 in this model , however , the vertex function is proportional to @xmath1 . in principle
the lorentz - index in this model can come from @xmath1 or from any of the momenta .
but , since the fermion propagator depends on the 4-momentum in the form @xmath132 , the fermion - photon vertex does not depend on the momentum components which are orthogonal to @xmath3 .
therefore the lorentz - index which comes from @xmath133 in fact comes from the longitudinal part of @xmath133 , ie .
proportional to @xmath1 .
so we can write @xmath134 .
this gives us the possibility that from the ward identities we _ exactly _ determine the vertex function .
the ward - identity for the current conservation yields then in case when @xmath135 : @xmath136 therefore we find @xmath137 this is an exact equation in the bloch - nordsieck model .
now we will solve this equation in the renormalized theory , and demonstrate that the solution is indeed identical with the bolch - nordsieck solution presented in section [ sec:1 ] .
in the second term @xmath138 drops out , resulting in an integral @xmath139 because of @xmath140 symmetry .
what remains is @xmath141 this form is true in the original model , we shall now find the renormalized form .
first we adapt the wave function renormalization for the fermionic fields which changes the bare propagator to @xmath142 where @xmath143 .
we will assume that the mass shell remains the same , then @xmath144 , and the free propagator will be @xmath145 .
we will use also the notation @xmath146 .
the full propagator then reads @xmath147 using we find the equation @xmath148 where @xmath149 @xmath150 can be interpreted as a running wave function renormalization constant . with a spectral representation @xmath151 in the appendix
we evaluate @xmath152 , and we find @xmath153.\ ] ] we rewrite it into , then , assuming normalizable spectral function , after some algebraic manipulation we find @xmath154 we may assume that the explicit integral is not uv divergent ( it can be checked a posteriori , or , as in the present case , knowing the exact solution )
. then the above equation can be made finite by requiring @xmath155 where @xmath156 and @xmath157 are finite .
this form can be interpreted physically as the appearance of the renormalized coupling @xmath156 and the finite wave function renormalization @xmath157 .
we note that the coupling constant renormalization equation agrees with the nonperturbative coupling constant renormalization in the o(n ) models @xcite .
now we find @xmath158 this function depends on the arbitrary scale @xmath159 , but the physics , of course , must be @xmath159 independent .
this can be achieved by appropriately changing the @xmath157 and @xmath156 constant when we change @xmath159 .
the @xmath159-independence of @xmath150 requires ( using the sum rule for @xmath67 ) : @xmath160 this can be satisfied if @xmath161 the second equation means @xmath162 where @xmath163 is a constant ; the first equation then reads @xmath164 using the normalizability of @xmath67 we finally find @xmath165 the @xmath163 and the scale @xmath166 are renormalization group independent quantities ( ie .
independent of the scale @xmath159 ) , these characterize the renormalization scheme .
the appearance of a scale @xmath166 is the manifestation of dimensional transmutation .
now , instead of that scale @xmath166 it is worth to use @xmath167 for which @xmath168 .
clearly @xmath169 if @xmath170 .
then with differentiating @xmath29 with respect to @xmath59 we find @xmath171 this gives finally @xmath172 by differentiation with respect to @xmath59 we find @xmath173 where @xmath174 is an arbitrary constant .
this is indeed the solution of bloch and nordsieck , now in terms of the renormalized quantities .
but we also see that the condition @xmath175 can be satisfied only with @xmath176 .
this is in close relation with the fact that at the mass shell @xmath177 , the propagator ( and its discontinuity ) is not integrable .
the lesson of this analysis is that the deep ir physics is well describable by the schwinger - dyson equation , truncated in a way which respects the ward - identities .
as we have seen , this strategy is renormalizable and exact in case of the bloch - nordsieck model .
a big advantage of this approach is that , besides being exact in the ir , it can be easily generalized to other theories .
so we expect that in qed the schwinger - dyson equations truncated in the way we have done it in the bloch - nordsieck theory will represent the exact result well in the problematic deep ir regime .
in this paper we examined the exactly solvable bloch - nordsieck model from the point of view of different perturbative methods .
we first reviewed the known method to obtain the exact solution @xcite .
then the different levels of approximations , like the one - loop level perturbation theory , the 2pi resummation and the truncated schwinger - dyson equations were studied .
the 1-loop result exhibits an ir sensitivity when we approach the mass shell which renders the theory ill - defined . the self - energy ( 2pi ) resummation reorganizes the perturbative series in a way that this ir problem disappears .
this does not mean , however , that the result itself would be closer to the exact one , only the explicit ir sensitivity can not be seen .
on the other hand , the schwinger - dyson equations , truncated in a way that the ward - identities are satisfied yield the exact result in the bloch - nordsieck model .
this is a new way of obtaining the exact solution in the bloch - nordsieck model . and , while the original solution method is very hard to generalize to other theories , the generalization of the specially truncated schwinger - dyson equations is straightforward .
the authors thank useful discussions with t.s .
br , f. csikor and a. patks .
this work is supported by the hungarian research fund ( otka ) under contract no .
the one - loop contribution to the self - energy reads , with a generic @xmath35 vector in feynman gauge : @xmath178 this is lorentz - invariant , if we a lorentz transformation both on @xmath35 and @xmath179 .
so we may choose a special frame where @xmath180 . if @xmath35 is a proper 4-velocity , then @xmath37 ; if it is @xmath38 , then @xmath39 , but still constant , since @xmath181 is a parameter of the theory .
we find then @xmath182 thus it is enough to consider @xmath183 only .
there we transform to positive frequency integrals @xmath184 where in the last step we performed wick rotation ( the choice of the imaginary part of @xmath185 is crucial for the direction of the rotation on the complex plane ) .
now we can write up the integral in @xmath186 and @xmath187 space , in the latter using @xmath188 dimensions : @xmath189 we use the relation @xmath190 to proceed as @xmath191.\end{aligned}\ ] ] we write it as @xmath192,\ ] ] where @xmath193 therefore @xmath194.\ ] ] j. m. luttinger and j. c. ward , phys .
118 ( 1960 ) 1417 .
g. baym , phys .
127 ( 1962 ) 1391 .
j. m. cornwall , r. jackiw and e. tomboulis , phys .
d 10 ( 1974 ) 2428 .
h. van hees an j. knoll , phys .
d65 ( 2002 ) 025010 [ hep - ph/0107200 ] ; phys .
d65 ( 2002 ) 105005 [ hep - ph/0111193 ] ; phys .
d66 ( 2002 ) 025028 [ hep - ph/0203008 ] j .- p .
blaizot , e. iancu and u. reinosa , phys .
b568 ( 2003 ) 160 [ hep - ph/0301201 ] ; nucl .
a736 ( 2004 ) 149 [ hep - ph/0312085 ] j. berges , sz .
borsanyi , u. reinosa and j. serreau , annals phys . 320
( 2005 ) 344 [ hep - ph/0503240 ] u. reinosa , j. serreau , jhep 0607 ( 2006 ) 028 [ hep - th/0605023 ] ; a. jakovac , phys .
rev . * d76 * , 125004 ( 2007 ) .
[ hep - ph/0612268 ] .
a. patkos , z. .szep , nucl .
phys . * a811 * , 329 - 352 ( 2008 ) .
[ arxiv:0806.2554 [ hep - ph ] ] .
j. berges and j. cox , phys . lett .
b 517 ( 2001 ) 369 .
j. berges , s. borsanyi and c. wetterich , phys .
93 ( 2004 ) 142002 .
j. berges , s. roth , nucl.phys.b847:197-219,2011 , [ arxiv:1012.1212 ] m.e .
carrington , e. kovalchuk , phys.rev.d76:045019,2007 , [ arxiv:0705.0162 ] n.n .
bogoliubov and d.v .
shirkov , _ introduction to the theories to the quantized fields _
( john wiley & sons , inc . , 1980 ) h.m
. fried , _
green s functions and ordered exponentials _ ( cambridge university press , 2002 ) j. -p .
blaizot and e. iancu , phys .
d * 56 * , 7877 ( 1997 ) [ hep - ph/9706397 ] , j. -p .
blaizot and e. iancu , phys .
d * 55 * , 973 ( 1997 ) [ hep - ph/9607303 ] .
weldon , phys .
d44 ( 1991 ) 3955 ; phys
d49 ( 1994 ) 1579 . m. le bellac , _ thermal field theory _ , ( cambridge univ . press , 1996 . ) p. mati diplomawork j.o .
andersen , d. boer , h.j .
warringa , phys .
d70 ( 2004 ) , 116007 [ hep - ph/0408033 ] , a. jakovac , phys . rev
. d * 78 * , 085013 ( 2008 ) [ arxiv:0808.1800 [ hep - th ] ] , g. fejos , a. patkos and z. .szep , phys .
d * 80 * , 025015 ( 2009 ) [ arxiv:0902.0473 [ hep - ph ] ] .
peskin , d.v .
schroeder , _ an introduction to quantum field theory _ , ( perseus books publishing , 1995 . ) | we studied different levels of resummations of the exactly solvable bloch - nordsieck model in order to be able to compare the approximations with an exact result .
we studied one - loop perturbation theory , 2pi resummation and schwinger - dyson equations truncated in a way to maintain ward - identities .
at all levels we carefully performed renormalization .
we found that although the 2pi resummation does not exhibit infrared ( ir ) sensitivity at the mass shell , as the one - loop perturbation theory does , but it is still far from the exact solution .
the method of truncated schwinger - dyson equations , however , is exact in this model , so it provides a new way of solving the bloch - nordsieck model .
this method can also be generalized to other , more complicated theories . |
the performance of atomic clocks has improved more than seven orders of magnitude since the first cesium clock was demonstrated by essen and parry over 50 years ago @xcite . in the past two decades
, improvement has been driven by several major scientific developments .
most recently , optical frequency combs , narrow line width lasers , and methods of confining atoms and ions without perturbing the clock frequency have led to optical clocks , with reference frequencies and corresponding @xmath7s orders of magnitude higher than previous systems .
prior to optical clocks , the ground breaking techniques of laser cooling and trapping of atomic gases brought about dramatic improvements in atomic clock performance by enabling the creation of ultra - cold samples of atoms , providing longer interaction times and reduced systematic uncertainties .
the first clocks based on cold atoms were atomic fountains , which are in some sense the natural progression of atomic beam clocks .
atomic fountains are the basis of contemporary primary frequency standards , in which the microwave frequency of the hyperfine transition in cesium-133 is used as the reference for a clock @xcite . as the most precise and accurate frequency standards , cold - atom clocks ( optical or microwave ) find application in frequency metrology @xcite , precise tests of fundamental symmetries @xcite , and as test beds for quantum scattering @xcite , entanglement @xcite and information processing @xcite . however , the application of atomic frequency standards as clocks for timing applications is still dominated by clocks based on older but more mature technology , with far greater levels of engineering and evaluation .
the workhorses of the timing community , commercial cesium beams and hydrogen masers , which do not incorporate any laser technology , run continuously and require little maintenance and upkeep , often running for years without requiring service .
international atomic time ( tai ) utilizes data from hundreds of these clocks , and many timing labs rely exclusively on commercial cesiums and hydrogen masers for their local ensembles .
it is to be expected that the best optical clocks will require significant engineering before being compatible with user - free operation in a timing ensemble .
( in fact , many of the best accuracy and stability evaluations carried out with optical clocks use measurements made at optical frequencies , without dividing the signal down to rf , which would be a requirement for most timing applications @xcite . ) on the other hand , atomic fountains are being used much more regularly for timing applications .
as primary frequency standards , cesium fountains have been contributing to tai for close to a decade , with the number and precision of reports increasing over time , and recently lne - syrte began reporting a rubidium fountain to the bipm as a secondary standard @xcite .
and many timing labs are relying more on fountain clocks for their local timescales , with atomic fountains in use or under development in at least 10 nations worldwide .
even with this increased use , the role of fountain clocks for timing has typically been different from the commercial clocks that dominate ( in number ) timing ensembles .
serving as primary frequency standards , the fountains provide calibrations of the frequencies of other clocks . in this role , they do not necessarily need to run continuously for months or years at a time without user intervention , though some have achieved close to continuous operation @xcite . and for a local timescale , it can suffice to use only a single primary standard , which ensures a constant long - term frequency , at least on the time frame of the accuracy evaluations . for example , the ptb has produced an excellent timescale based on a single cesium fountain @xcite .
still , there are advantages to having multiple clocks .
more clocks translate to more robustness and reliability . and
it is sometimes the case that a fountain clock is characterized by comparing to a second one . for these reasons ,
most institutions with fountains have or are developing at least two ; lne - syrte has three and npl is working on a third ( in each case one of the three fountains uses rubidium ) .
additional advantages of multiple clocks , particularly for timekeeping , are related to reducing noise .
the white - frequency - noise level of an ensemble with @xmath8 clocks decreases as @xmath9 . and since continuously running atomic clocks exhibit non - stationary behavior , timing ensembles based on such clocks require multiple devices to optimize performance . the u.s .
naval observatory ( usno ) , the largest contributor of clock data to the bipm , one of the official sources of time for the united states and the official time and frequency reference for critical infrastructure such as gps , has designed and built an ensemble of atomic fountain clocks that has been in operation for more than 2.5 years .
the ensemble consists of four rubidium fountains housed in washington , dc @xcite .
these clocks run continuously , with little operator intervention required , and are used to enhance the local timing ensemble , contributing to the usno master clock in a manner similar to commercial cesium beams and hydrogen masers .
the specific role of the fountains is to provide a superior long - term frequency reference , in the past solely the job of the commercial cesiums , while the short - term reference continues to be provided by an ensemble of hydrogen masers .
all four rubidium fountains have been regularly reported to the bipm for over two years as continuously running clocks , the first cold - atom clocks so reported . here
we present the performance of these clocks over the first 2 years of operation , the first characterization of continuously running atomic fountains over such time frames .
the four fountains in operation at usno are designated nrf2 , nrf3 , nrf4 and nrf5 .
the systems were built in two generations , according to slightly different designs @xcite .
all four fountains began continuous operation in a dedicated clock facility in march 2011 .
since then , some of the improvements in design used for the second generation , nrf4 and nrf5 , have been incorporated as retrofits to the first generation , nrf2 and nrf3 .
additionally , both routine and unexpected maintenance were required on occasion . since december of 2011 ,
fountain data for all four devices have been reported without interruption to the bipm , as clock type 93 . in may of 2012 , after the traditional evaluation period , the fountains were weighted in tai for the first time , and all four clocks have received maximum weight in every evaluation since .
each fountain has a local oscillator ( lo ) comprised of a quartz crystal phase - locked to a dedicated maser , which serves as the reference for the 6.8 ghz drive for rubidium spectroscopy as well as the reference for a precise frequency synthesizer , an auxiliary output generator ( aog ) .
every 20 seconds , the average relative frequency of the fountain and maser is used to steer the aog output , providing a continuous nominal 5 mhz signal that reflects the fountain frequency and is measured against the usno master clock and other clocks in the ensemble . during each fountain cycle
, auxiliary data such as the number of atoms participating in the frequency measurement and the fraction of atoms making the transition between clock states are used to determine whether the fountain is operating normally .
if these data fall outside of a predetermined healthy range , the fountain automatically switches to holdover mode , in which the maser serves as a flywheel ; in this case , the control computer applies a steer that is the median frequency difference between the fountain and maser over the previous hour .
the same holdover steer is applied until the auxiliary data reflect healthy values and normal operation resumes .
the robustness of the fountains and the ability to rely on a maser as a flywheel have resulted in a high uptime over the 2 year period . the percentage of time that each fountain has generated a good , steered output is 99.2% , 99.7% , 98.5% and 100% for nrf2 through nrf5 .
a `` good output '' can include brief intervals where the system is in holdover ; so long as the interval corresponds to an averaging time for which the relative frequency variations of the maser and fountain are still white ( _ i.e. _ the frequency fluctuations are still limited by fountain performance ) .
periods in holdover long enough that the steered output reflects maser behavior is treated as `` down time '' and is not included in analysis of fountain behavior ; it is handled in the data analysis by interpolating the frequency record and integrating to generate the phase record .
the quickest degradation of performance of the ( nominal ) 5 mhz output signal is the occurrence of a problem with the processor controlling fountain operation ; if the steering algorithm is interrupted , the aog will no longer be updated , and its output frequency will reflect a steer determined from 20 s of averaging ( rather than the holdover value ) .
processor refresh is usually a simple process .
problems that are more difficult to solve , but do not show up as quickly in the steered output , include shutter failure and diode laser aging .
aside from these three problems that have arisen , routine maintenance is required on occasion .
this consists almost exclusively of occasional tweaks to optical alignment , sometimes not needed for months at a time .
very rarely an adjustment to the vco on the reference maser or aog is required to maintain phase lock .
there is no maintenance required for the laser frequency locks , which perform without incident .
perhaps in part due to the refinement in design , the highest performing fountains have been nrf4 and nrf5 .
each device exhibits a white - frequency noise level of @xmath10 and excellent long - term stability .
( there is an obvious enhancement to the microwave chain which should improve the short - term stability to a level of about @xmath11 , which was demonstrated in an engineering prototype , nrf1 . ) in fig .
[ f.45good](a ) , the relative phase of nrf4 and nrf5 is plotted over an uninterrupted period of more than 1 year , after removing a relative frequency of @xmath12 and a meaningless relative phase .
residual peak - to - peak phase deviations stay within about 1 ns for the entire interval .
a limit on the relative frequency drift can be obtained by fitting a line to the frequency , with the result that the relative drift is zero at the level of @xmath13/day ( @xmath14/year ) .
a sigma - tau plot , shown in ( b ) , includes overlapping allan deviation , total deviation and theo statistics , along with a white - frequency noise reference line .
there is no indication of deviation from white - frequency noise behavior out to the total averaging time of more than 8 months .
the average ( in)stability for each fountain is consistent with an overlapping allan deviation of @xmath15 at @xmath16 s , a total deviation of @xmath17 at @xmath18 s and a theo of @xmath19 at @xmath20 s ( values correspond to the white - frequency reference level divided by @xmath21 , not to the data points which lie below this line ) . figure [ f.45good ] ( c ) shows the time deviation between the two fountains for the @xmath22 year interval ; the average tdev for each fountain is about 20 ps at a day .
it is likely that this represents record performance for microwave clocks . for each fountain .
( c ) time deviation for data in ( a).,scaledwidth=60.0% ] the data analyzed in fig .
[ f.45good ] correspond to an interval when both fountains exhibited excellent stability , which also coincided with a period of time during which there was little or no user intervention .
there were times during the two years when some final refinements were made to nrf4 , upgrades were made to nrf2 and nrf3 , and maintenance consisting of laser and shutter replacement were carried out on nrf2 and nrf3 . when the entire two year data set is analyzed without any restrictions , and with no corrections or re - evaluations of the frequency across the service or upgrade , we see long - term behavior that deviates from white frequency noise .
figure [ f.pairwise ] shows the overlapping allan deviations for all pair - wise comparisons using the entire two - year data set .
the long - term behavior in all cases shows similar non - stationary behavior .
when the nrf4 vs nrf5 comparison in ( a ) is extended to longer averaging times using the theo deviation ( darker symbols ) , the most obvious fit is to random - walk frequency noise , with a @xmath23 dependence .
the reference line in the figure corresponds to a random - walk level of @xmath24 per fountain .
but , as we now show , the behavior leading to these stability plots is entirely consistent with discrete changes in frequency , at a rate that makes it impossible at this point to distinguish between inherent random - walk frequency noise and frequency changes arising from user intervention . in fig .
[ f.45white](a ) , we show the relative phase of nrf4 and nrf5 over the entire two year period .
the record of the relative phase shows two different linear regions with different slopes , indicating that the relative frequency changed ( the data in fig .
[ f.45good ] correspond to the second frequency shown ) .
this bears out when we adjust the data starting from 0.9 years ( mjd 55975 ) to the end by removing a single frequency , of size @xmath25 , in order to make the last part of the frequency record match the first .
the subsequent data look white , as illustrated by the sigma - tau plot in fig .
[ f.45white](b ) , indicating that the non - stationary behavior is indeed consistent with a single frequency change .
similar behavior was observed for nrf3 . in fig .
[ f.35white](a ) , we show the relative phase of nrf3 and nrf5 over the entire two year period . here
, the relative frequency appears to take on three different values . adjusting the data at the two times when the frequency changed ( 0.6 yrs ( mjd 55840 ) and 1.1 yrs ( mjd 56030 ) ) , thereby adding two frequency adjustments , the resulting behavior is consistent with integration as white - frequency noise , as shown by the sigma - tau plot in fig .
[ f.35white](b ) .
so for nrf3 , nrf4 and nrf5 , the long - term behavior under continuous operation is consistent with rare , discrete frequency changes , at an average rate of 1 every 2 years and an average size of @xmath26 .
we can unambiguously assign the frequency changes in fig .
[ f.35white ] to nrf3 , as they appear as well in comparisons with nrf4 .
the frequency change in fig .
[ f.45white ] is more difficult to assign to one fountain or the other , because the other two fountains do not support a particular position .
we suspect that the frequency change can be assigned to nrf4 , however , because the time at which the change occurred coincides within the error bar to a time at which some finishing touches were completed on nrf4 . similarly , one of the changes in frequency exhibited in fig .
[ f.35white](a ) occurred close in time to one of the upgrades made to bring the design of nrf3 closer to that of the second generation .
it is impossible to conclude that these frequency changes were caused by user intervention , but it is a possibility , and it is something we intend to continue analyzing . almost all of the most likely sources of significant frequency shifts in a fountain have been ruled out as drivers of the observed behavior the magnetic field is regularly measured and shows no correlated changes ; light shifts do not seem to be a problem ; the ambient temperature is monitored and is constant .
it is hard to conclusively rule out a microwave effect , particularly within the microwave chain
. it would be interesting to compare to primary standards , in terms of whether biases measured in accuracy evaluations change occasionally over time . because these frequency changes are rare
, it is straightforward to recharacterize a fountain using the others as a reference .
this introduces the possibility of producing a paper fountain timescale that conceivably integrates as white - frequency noise to on order of @xmath27 .
we plan to consider this further .
the fourth fountain , nrf2 , has exhibited the least stable behavior thus far .
it has shown discrete frequency changes , of larger size and at a higher rate than those discussed above , as well as intervals that more likely corresponded to drift .
some of the problematic behavior in nrf2 does not correlate with episodes of significant intervention .
we hope to be able to improve the performance of nrf2 and in the process identify the cause of at least some of this non - stationary behavior .
finally , we point out that these changes in the output frequencies of the fountains are not observable using other clocks or timescales at usno .
neither the cesium ensemble , nor the maser ensemble , nor individual masers independently show that a fountain has changed frequency ; these events are only observed with inter - fountain comparisons .
for years the usno master clock , the physical representation of utc(usno ) , has been generated by incorporating the long - term frequency stability of commercial cesium beams with the superior short - term stability of hydrogen masers .
masers are well known to display time - variable drift and other behavior different from white - frequency noise after integrating to a level between low @xmath28s and high @xmath1s .
each maser is re - characterized using the cesiums ; the sooner that non - stationary behavior is detected , the better .
one role of the rubidium fountains is to improve on the process of maser re - characterization . in fig .
[ f.usno_timescale ] we show the allan deviation of an exceptionally stable maser versus a crude fountain timescale , created by averaging the frequencies of the four rubidium fountains .
the short term relative stability of @xmath29 at 1 s is consistent with a white frequency noise level of @xmath30 for the maser and @xmath5 for the fountain timescale , as expected for four fountains with an average white - frequency noise of @xmath31 .
the white frequency noise level achieved with the usno ensemble of about 70 commercial cesium beams is about @xmath32 .
so the fountain ensemble shows a factor of 10 improvement in short - term stability , which , in terms of averaging time , translates to a factor of 100 reduction . even a single fountain is more than 50 times better than the entire ensemble of cesiums . , corresponding to @xmath5 for the fountain timescale and @xmath30 for the maser.,scaledwidth=60.0% ]
as stated previously , it should be possible to improve the short - term stability of each fountain to about @xmath33 , making the fountain timescale white frequency noise level @xmath34 .
we expect that this would be the limit for our four - fountain ensemble .
we have investigated using the same lo for a pair of fountains to enable reduction of the lo noise contribution .
removing the entire lo noise contribution by differencing the fountain frequencies is possible for a paper clock and is useful for measuring atomic physics effects more quickly , but it can not be incorporated into a physical output . to generate a physical output with reduced lo noise
, we implemented what we call `` cooperative steering '' , where two fountains steer the same lo . by synchronizing the load - measurement cycles so that the lo is always being measured and compensated for , the additional noise from not continually sampling the lo , the dick effect , can be reduced @xcite .
because of the high quality of the los used for the rubidium fountains , the reduction in the dick effect does not improve the stability greatly , while using a single lo for different fountains increases risk significantly .
this increase in operational risk led us to abandon this technique .
as confidence in the reliability of the fountains has been gained , the clocks are now being used operationally as the long - term reference in the generation of utc(usno ) . the specific algorithm for including the fountains as well as a quantitative assessment of their impact on utc(usno )
is an area of ongoing research @xcite .
eight cesium fountains worldwide contributed on some basis to tai during the 2 years we are considering .
perhaps the best metric of long - term frequency stability of the usno rubidium fountains is by comparing to these primary standards @xcite .
we do this by using the usno master clock , the physical realization of utc(usno ) , as a transfer oscillator to allow us to determine the frequency difference between a fountain and tai , @xmath35 . along with the values of the average frequency of the primaries compared to eal , @xmath36 , and the steers of eal to tai , @xmath37 , which are published monthly by the bipm in their circular t report , the frequency of a fountain or fountain timescale compared to the primaries can be obtained .
figure [ f.primaries](a ) shows the monthly average of the frequency of nrf5 versus the average of the primaries .
the line in the plot is a linear fit , @xmath38/day , showing that we can place a limit on the drift of nrf5 with respect to the primaries at the level of @xmath39/per day . in fig .
[ f.primaries](b ) , a sigma - tau plot shows the stability of each fountain versus the primaries .
/per day , demonstrating that nrf5 exhibits zero drift with respect to the world s primaries at the level of @xmath39/per day .
( b ) total deviation for each individual fountain vs the cesium fountains .
theo deviation and sample error bars are included for nrf5.,scaledwidth=60.0% ] because of the continuous operation of our rubidium fountains , in addition to comparing to the monthly average of the primaries , we can obtain a measurement against each individual primary standard reported or against standards at a particular institution .
this allows us to analyze trends with respect to specific standards labs .
this is illustrated in fig .
[ f.ind_primaries ] , where the relative frequency of nrf5 and each primary standard reported to tai over the 2 years is plotted .
the ( three ) fountains at lne - syrte and ( the two ) at ptb are differentiated from the other labs reporting ( nist , npl and one report from nict ) .
to summarize , the usno has put into operation an ensemble of continuously running rubidium fountain clocks to incorporate into timing applications and timescale generation in a manner similar to hydrogen masers and commercial cesium beam clocks .
we have analyzed performance of the four fountains in the ensemble over the first two years of operation , demonstrating intervals reflecting record performance for microwave clocks and observing rare , discrete frequency changes .
comparison to the world s primary standards suggest that the rubidium fountains provide a stable , long - term frequency reference on par with any local reference .
we benefited from assistance from many members of the usno time service department , particularly paul koppang , jim skinner and demetrios matsakis .
scott crane played a significant role in the design and construction of the fountains .
atomic fountain development at usno has been funded by onr and spawar .
10 g. j. dick , j. d. prestage , c. a. greenhall and l. maleki , in _ proceedings of the 19th precise time and time interval ( ptti ) applications and planning meeting _ , redondo , california ( national aeronautics and space administration , washington , dc 1987 ) , p. 133 | an ensemble of rubidium atomic fountain clocks has been put into operation at the u.s . naval observatory ( usno ) .
these fountains are used as continuous clocks in the manner of commercial cesium beams and hydrogen masers for the purpose of improved timing applications .
four fountains have been in operation for more than two years and are included in the ensemble used to generate the usno master clock .
individual fountain performance is characterized by a white - frequency noise level below @xmath0 and fractional - frequency stability routinely reaching the low @xmath1s .
the highest performing pair of fountains exhibits stability consistent with each fountain integrating as white frequency noise , with allan deviation surpassing @xmath2 at @xmath3 s , and with no relative drift between the fountains at the level of @xmath4/day . as an ensemble ,
the fountains generate a timescale with white - frequency noise level of @xmath5 and long - term frequency stability consistent with zero drift relative to the world s primary standards at @xmath6/day .
the rubidium fountains are reported to the bipm as continuously running clocks , as opposed to secondary standards , the only cold - atom clocks so reported . here
we further characterize the performance of the individual fountains and the ensemble during the first two years in an operational environment , presenting the first look at long - term continuous behavior of fountain clocks . |
the study of changes in the natural world , dynamics , is divided among several distinct disciplines .
thermodynamics , for example , considers changes between special states , the so - called states of equilibrium , and addresses the question of which final states can be reached from any given initial state .
mechanics studies the changes we call motion , chemistry deals with chemical reactions , quantum mechanics with transitions between quantum states , and the list goes on . in all of these examples we want to predict or explain the observed changes on the basis of information that is codified in a variety of ways into what we call the states . in some cases
the final state can be predicted with certainty , in others the information available is incomplete and we can , at best , only assign probabilities .
the theory of thermodynamics holds a very special place among all these forms of dynamics . with the development of statistical mechanics by maxwell , boltzmann , gibbs and others , and eventually culminating in the work of jaynes @xcite
, thermodynamics became the first clear example of a fundamental physical theory that could be derived from general principles of probable inference .
the entire theory follows from a clear idea of the subject matter , that is , an appropriate choice of which states one is talking about , plus well - known principles of inference @xcite , namely , consistency , objectivity , universality and honesty .
these principles are sufficiently constraining that they lead to a unique set of rules for processing information : these are the rules of probability theory @xcite and the method of maximum entropy @xcite@xcite
. there are strong indications that a second example of a dynamics that can be deduced from principles of inference is afforded by quantum mechanics @xcite .
many features of the theory , traditionally considered as postulates , follow from the correct identification of the subject matter plus general principles of inference . briefly
, the goal of quantum mechanics is not to predict the behavior of microscopic particles , but rather to predict the outcomes of experiments performed with certain idealized setups .
thus , the subject of quantum theory is not just the particles , but rather the experimental setups .
the variables that encode the information relevant for prediction are the amplitudes or wave functions assigned to the setups .
these ingredients plus a requirement of consistency ( namely , that if there are two ways to compute an amplitude , the two results should agree ) supplemented by entropic arguments are sufficient to derive most of the standard formalism including hilbert spaces , a time evolution that is linear and unitary , and the born probability rule .
if quantum mechanics , deemed by many to be _ the _ fundamental theory , can be derived in this way , then it is possible , perhaps even likely , that other forms of dynamics might ultimately reflect laws of inference rather than laws of nature .
should this turn out to be the case , then the fundamental equations of change , or motion , or evolution as the case might be , would follow from probabilistic and entropic arguments and the discovery of new dynamical laws would be reduced to the discovery of what is the necessary information for carrying out correct inferences . unfortunately , this search for the right variables has always been and remains to this day the major stumbling block in the understanding of new phenomena . the purpose of this paper is to explore this possible connection between the fundamental laws of physics and the theory of probable inference : can dynamics be derived from inference ? rather than starting with a known dynamical theory and attempting to derive it , i proceed in the opposite direction and ask : what sort of dynamics can one derive from well - established rules of inference ? in section 2
i establish the notation , define the space of states , and briefly review how the introduction of a natural quantitative measure of the change involved in going from one state to another turns the space of states into a metric space @xcite .
( such metric structures have been found useful in statistical inference , where the subject is known as information geometry @xcite , and in physics , to study both equilibrium @xcite and nonequilibrium thermodynamics @xcite . ) typically , once the kinematics appropriate to a certain motion has been selected , one proceeds to define the dynamics by additional postulates .
this is precisely the option i want to avoid : in the dynamics developed here there are no such postulates .
the equations of motion follow from an assumption about what information is relevant and sufficient to predict the motion . in a previous paper @xcite
i tackled a similar problem .
there i answered the question : q1 : : : given the initial state and that the system evolves to other states , what trajectory is the system expected to follow ?
this question implicitly assumes that there is a trajectory and that information about the initial state is sufficient to determine it .
the dynamical law follows from the application of a principle of inference , the method of maximum entropy ( me ) , to the only information available , the initial state and the recognition that motion occurred .
nothing else .
the resulting ` entropic ' dynamics is very simple : the system moves continuously and _ irreversibly _ along the entropy gradient .
thus , the honest , correct answer to the inference problem posed by question q1 has been given , but the equally important question ` will the system in fact follow the expected trajectory ? ' remained unanswered . whether the actual trajectory is
the expected one depends on whether the information encoded in the initial state happened to be sufficient for prediction . indeed , for many systems , including those for which the dynamics is _ reversible _ , more information is needed . in section 3
we answer the question : q2 : : : given the initial and the final states , what trajectory is the system expected to follow ?
again , the question implicitly assumes that there is a trajectory , that in moving from one state to another the system will pass through a continuous set of intermediate states . and
again , the equation of motion is obtained from a principle of inference , the principle of maximum entropy , and not from a principle of physics .
( for a brief account of the me method in a form that is convenient for our current purpose see @xcite . )
the resulting ` entropic ' dynamics also turns out to be simple : the system moves along a geodesic in the space of states .
this is simple but not trivial : the geometry of the space of states is curved and possibly quite complicated .
important features of this entropic dynamics are explored in section 4 .
we show that there are some remarkable formal similarities with the theory of general relativity ( gr ) .
for example , just as in gr there is no reference to an external physical time .
the only clock available is provided by the system itself .
it turns out that there is a natural choice for an ` intrinsic ' or ` proper ' time .
it is a derived , statistical time defined and measured by the change itself .
intrinsic time is quantified change .
this entropic dynamics can be derived from a jacobi - type principle of least action , which we explore both in lagrangian and hamiltonian form . just as in gr there is invariance under arbitrary reparametrizations a form of general covariance and the entropic dynamics is an example of what is called a constrained dynamics .
in this section we briefly review how to quantify the notion of change ( for more details see @xcite ) .
the idea is simple : since the larger the change involved in going from one state to another , the easier it is to distinguish between them , we claim that change can be measured by distinguishability .
next , using the me method one assigns a probability distribution to each state .
this transforms the problem of distinguishing between two states into the problem of distinguishing between the corresponding probability distributions .
the solution is well - known : the extent to which one distribution can be distinguished from another is given by the distance between them as measured by the fisher - rao information metric @xcite@xcite@xcite .
thus , change is measured by distinguishability which is measured by distance .
let the microstates of a physical system be labelled by @xmath0 , and let @xmath1 be the number of microstates in the range @xmath2 .
we assume that a state of the system ( _ i.e. _ , a macrostate ) is defined by the expected values @xmath3 of some @xmath4 appropriately chosen variables @xmath5 ( @xmath6 ) , @xmath7 a crucial assumption is that the selected variables codify all the information relevant to answering the particular questions in which we happen to be interested .
this is a point that we have made before but must be emphasized again : there is no systematic procedure to choose the right variables . at present the selection of relevant variables
is made on the basis of intuition guided by experiment ; it is essentially a matter of trial and error .
the variables should include those that can be controlled or observed experimentally , but there are cases where others must also be included .
the success of equilibrium thermodynamics , for example , derives from the fact that a few variables are sufficient to describe a static situation , and being few , these variables are easy to identify . in fluid dynamics , on the other hand ,
the selection is more dificult .
one must include many more variables , such as the local densities of particles , momentum , and energy , that are neither controlled nor usually observed .
the states form an @xmath4-dimensional manifold with coordinates given by the numerical values @xmath8 . to each state
we can associate a probability distribution @xmath9 .
the distribution that best reflects the prior information contained in @xmath10 updated by the information @xmath3 is obtained by maximizing the entropy @xmath11=-\int \,dx\,p(x)\log \frac{p(x)}{m(x)}. \label{s[p]}\ ] ] subject to the constraints ( [ aalpha ] ) .
the result is @xmath12 where the partition function @xmath13 and the lagrange multipliers @xmath14 are given by @xmath15 next , we argue that the change involved in going from state @xmath16 to the state @xmath17 can be measured by the extent to which the two distributions can be distinguished . as discussed in @xcite , except for an overall multiplicative constant , the measure of distinguishability we seek is given by the ` distance ' @xmath18 between @xmath9 and @xmath19 , @xmath20 where @xmath21 is the fisher - rao metric @xcite@xcite .
it turns out that this metric is unique : it is the only riemannian metric that adequately reflects the fact that the states @xmath16 are not ` structureless points ' , but happen to be probability distributions . to summarize : the very act of assigning a probability distribution @xmath9 to each state @xmath16 , automatically provides the space of states with a metric structure .
given the initial and the final states , what trajectory is the system expected to follow ?
the key to answering this question lies in the implicit assumption that there exists a trajectory or , in other words , that large changes are the result of a continuous succession of very many small changes .
thus , the difficult problem of studying large changes is reduced to the much simpler problem of studying small changes .
let us therefore focus on small changes and assume that the change in going from the initial state @xmath22 to the final state @xmath23 is small enough that the distance @xmath24 between them is given by @xmath25 to find which states are expected to lie on the trajectory between @xmath22 and @xmath26 we reason as follows . in going from the initial to the final state the system must pass through a halfway point , that is , a state @xmath16 that is equidistant from @xmath22 and @xmath26 ( see fig.1a ) .
the question is which halfway state should we choose ?
an answer to this question would clearly determine the trajectory : first find the halfway point , and use it to determine ` quarter of the way ' points , and so on .
next we notice that there is nothing special about halfway states .
we could equally well have argued that in going from the initial to the final state the system must first traverse a third of the way , that is , it must pass through a state that is twice as distant from @xmath26 as it is from @xmath22 .
in general , we can assert that the system must pass through intermediate states @xmath27 such that , having already moved a distance @xmath18 away from the initial @xmath22 , there remains a distance @xmath28 to be covered to reach the final @xmath26 .
halfway states have @xmath29 , ` third of the way ' states have @xmath30 , and so on ( see fig.1b ) .
it appears that each different value of @xmath31 provides a different criterion to select the trajectory . if there is a trajectory and there are several ways to determine it , consistency demands that all these ways should agree : in the end we must verify that the selected trajectory is independent of @xmath31 or else we have a problem .
our basic dynamical question q2 can be rephrased as follows : q2@xmath32 : : the system is initially in state @xmath33 and we are given the new information that the system has moved to one of the neighboring states in the family @xmath34 .
which @xmath35 do we select ? phrased in this way it is clear that this is precisely the kind of problem to be tackled using the me method . recall @xcite : the me method is a method for processing information and change our minds .
it allows us to go from an old set of beliefs , described by the prior probability distribution , to a new set of beliefs , described by the posterior distribution , when the information available is just a specification of the family of distributions from which the posterior must be selected @xcite . in the more traditional applications of the method this family of posteriors
is constrained or defined by the known expected values of some relevant variables , but this is not necessary , the constraints need not be linear functionals . here
the constraints are defined geometrically . an important question that should arise whenever one contemplates using the me method is which entropy should one maximize .
since we want to select a distribution @xmath9 the entropies to be considered must be of the form @xmath36=-\int \,dx\,p(x|a)\log \frac{p(x|a)}{q(x)}.\ ] ] this is the entropy of @xmath9 relative to the _ prior _ @xmath37 .
the interpretation of @xmath37 as the prior follows from the logic behind the me method itself .
recall @xcite : in the absence of new information there is no reason to change one s mind . when there are no constraints the selected posterior distribution should coincide with the prior distribution .
since the distribution @xmath38 that maximizes @xmath39 $ ] subject to no constraints is @xmath40 , we must set @xmath37 equal to the prior . coming back to our dynamical problem ,
suppose we know that the system is initially in state @xmath33 and we are not given the information that the system moved , then we have no reason to suspect that any change has occurred
. therefore the prior @xmath37 should be chosen so that the maximization of @xmath39 $ ] subject to no constraints yields the posterior @xmath41 .
the correct choice is @xmath42 .
now we are ready to tackle the question q2@xmath32 : the answer is obtained by maximizing the entropy @xmath43=-\int \,dx\,p(x|a)\log \frac{p(x|a)}{p(x|a_i)},\ ] ] subject to the constraint @xmath44 .
this presents no problems .
it is convenient to write @xmath45 and @xmath23 so that @xmath46 $ ] simplifies to @xmath47=-\frac 12g_{\alpha \beta } \,da^{{}\alpha } da^{{}\beta } , \ ] ] and the distances @xmath48 and @xmath49 from @xmath27 to @xmath22 and @xmath26 are given by @xmath50 then , to maximize @xmath51 $ ] under variations of @xmath52 subject to the constraint @xmath53 introduce a lagrange multiplier @xmath54 , @xmath55 we get @xmath56 the multiplier @xmath54 , or equivalently the quantity @xmath57 , is determined substituting back into the constraint ( [ w constr ] ) . from eq.([distances ] ) we get @xmath58 and @xmath59 , and therefore @xcite @xmath60 thus , the intermediate state @xmath27 selected by the maximum entropy method is such that @xmath61 the geometrical interpretation is obvious : the triangle defined by the points @xmath22 , @xmath27 , and @xmath26 ( fig.1 ) degenerates into a straight line .
this is sufficient to determine a short segment of the trajectory : all intermediate states lie on the straight line between @xmath22 and @xmath26 .
the generalization beyond short trajectories is immediate : if any three nearby points along a curve lie on a straight line the curve is a geodesic .
note that this result is independent of the value @xmath31
so the potential consistency problem we had identified earlier does not arise . to summarize , the answer to our question q2 is simple and elegant : ed : : the expected trajectory is the geodesic that passes through the given initial and final states .
this is the main result of this paper .
as promised , in entropic dynamics the motion is predicted on the basis of a ` principle of inference ' , the principle of maximum entropy , and not from a ` principle of physics ' .
the dynamics ed was derived in an unusual way and one should expect some unusual features .
indeed , they become evident as soon as one asks any question involving time . for example
, ed determines the vector tangent to the trajectory @xmath62 , but not the actual ` velocity ' @xmath63 .
the reason is not hard to find : nowhere in question q2 nor in any implicit background information is there any reference to an external time @xmath64 .
additional information is required if one is to find a relation between the distance @xmath65 along the trajectory and the external time @xmath64 . in conventional forms of dynamics
this information is implicitly supplied by a ` principle of physics ' , by a hamiltonian which fixes the evolution of a system relative to external clocks . but
q2 makes no mention of any external universe ; the only clock available is the system itself , and our problem becomes one of deciding how this clock should be read .
we could , for example , choose one of the variables @xmath8 , say @xmath66 , as our clock variable and arbitrarily call it intrinsic time .
ultimately it is best _ to define intrinsic time so that motion looks simple_. a very natural definition consists in stipulating that the system moves with unit velocity , then the intrinsic time @xmath67 is given by the distance @xmath68 itself , @xmath69 . _
intrinsic time is quantified change_. a peculiar consequence of this definition is that intervals between events along the trajectory are not a priori known , they are determined only after the equations of motion are solved and the actual trajectory is determined .
this reminds us of the theory of general relativity ( gr ) . an important feature of gr
is the absence of references to an external time .
given initial and final states , in this case the initial and final three - dimensional geometries of space , the proper time interval along any curve between them is only determined after solving the einstein equations of motion @xcite .
the absence of an external time has been a serious impediment in understanding the classical theory @xcite because it is not clear which variables represent the true gravitational degrees of freedom and also in formulating a quantum theory @xcite because of difficulties in defining equal - time commutators . in the following section we rewrite the entropic dynamics in lagrangian and hamiltonian forms and we point out some further formal similarities between ed and gr .
the entropic dynamics can be derived from an ` action ' principle .
since the trajectory is a geodesic , the ` action ' is the length itself , @xmath70=\int_{\eta _ i}^{\eta _ f}d\eta \,l(a,\dot{a } ) , \label{j[a]}\ ] ] where @xmath71 is an arbitrary parameter along the trajectory , the lagrangian is @xmath72 the action @xmath73 $ ] is invariant under reparametrizations @xmath74 provided the end points are left unchanged , @xmath75 and @xmath76 . indeed , when the transformation is infinitesimal , @xmath77 , the corresponding change in the action , @xmath78 vanishes provided @xmath79 .
as pointed out in @xcite there is an important distinction between the symmetries of a generally covariant theory such as gr and the internal symmetries of a proper gauge theory .
the action of a generally covariant theory is invariant under those reparametrizations that are restricted to map the boundary onto itself ; for proper internal gauge transformations there are no such restrictions .
thus ed shares with gr the fact that both are generally covariant theories .
it is instructive to consider the analogous principle of least action for a nonrelativistic particle .
the standard hamilton s principle requires extremizing the action @xmath80 where @xmath64 is ` physical ' time , and the interval between initial and final states @xmath81 is given .
in contrast , jacobi s principle of least action for a particle with energy @xmath82 moving in a potential @xmath83 determines the trajectory by extremizing the action @xmath84=\int_{\eta _ i}^{\eta _ f}d\eta \left ( 2m\delta _ { ab}\frac{dx^{{}a}}{% d\eta } \frac{dx^{{}b}}{d\eta } \right ) ^{1/2}\left ( e - v(x)\right ) ^{1/2}. \label{jacobi}\ ] ] there is no reference to any time @xmath64 , the time interval between initial and final states is not given , and the parameter @xmath71 is unphysical and arbitrary . to determine the temporal evolution along the trajectory requires an additional supplementary condition , @xmath85 thus the ed action , eq.([j[a ] ] ) ,
is an action of the jacobi type .
the natural choice for a supplementary condition that defines @xmath67 and determines the evolution along the trajectory is @xmath86 it is interesting that gr is also described by a jacobi - type action @xcite . to explore this similarity
further it is convenient to construct the canonical ( _ i.e. _ , hamiltonian as opposed to lagrangian ) version of jacobi s action .
the canonical momenta are given by @xmath87 and have unit magnitude , @xmath88 the canonical hamiltonian vanishes identically , @xmath89 because the lagrangian is homogeneous of first degree in the @xmath90 s .
physically this is not surprising : the generator of time evolution can be expected to vanish whenever there is no external time with respect to which the system could possibly evolve .
we are led to consider the canonical action @xmath91 but eq.([mom mag ] ) implies that unconstrained variations of the momenta @xmath92 are not allowed .
the correct variational principle requires to extremize the action @xmath93=\int_{\eta _ i}^{\eta _ f}d\eta \,\left [ \dot{a}^{{}\alpha } \pi _ { { } \alpha } -n\,h(a,\pi ) \right]\ ] ] where @xmath94 and @xmath95 are lagrange multipliers that enforce the constraint @xmath96 for each value of @xmath71 .
the overall factor of @xmath97 in eq.([h ] ) is introduced for later convenience ; it amounts to rescaling @xmath98 .
variation of @xmath99 $ ] with respect to @xmath16 , @xmath100 , and @xmath98 yields the equations of motion , @xmath101 and eq.([h constraint ] ) .
naturally there is no equation of motion for @xmath98 and it must be determined from the constraint .
we obtain , @xmath102 which , using the supplementary condition eq.([suppl cond ] ) , implies @xmath103 the analogue of @xmath98 in gr is called the lapse function , it gives the increase of ` intrinsic ' time per unit increase of the unphysical parameter @xmath104 . in terms of @xmath67 the equations of motion become @xmath105 in reparametrization invariant or generally covariant theories there is no canonical hamiltonian ( it vanishes identically ) but there are constraints .
it is the constraints that play the role of generators of evolution , of change .
accordingly , the analogue of eq.([h constraint ] ) in gr is called the hamiltonian constraint .
i have provided an answer to the question ` given the initial and final states , what is the trajectory followed by the system ? '
the answer follows from established principles of inference without invoking additional ` physical ' postulates .
the entropic dynamics thus derived turns out to be formally similar to other generally covariant theories : the dynamics is reversible ; the trajectories are geodesics ; the system supplies its own notion of an intrinsic time ; the motion can be derived from a variational principle that turns out to be of the form of jacobi s action principle rather than the more familiar principle of hamilton ; and the canonical hamiltonian formulation is an example of a dynamics driven by constraints .
to conclude one should point out that a reasonable physical theory must satisfy two key requirements : the first is that it must provide us with a set of mathematical models , the second is that the theory must identify real physical systems to which the models might possibly apply .
the entropic dynamics we propose in this paper satisfies the first requirement , but so far it fails with respect to the second ; it may be a reasonable theory but it is not yet ` physical ' .
there are formal similarities with the general theory of relativity and one wonders : is this a coincidence ? whether gr will in the end turn out to be an example of ed is at this point no more than a speculation .
a more definite answer hinges on the still unsettled problem of identifying those variables that describe the true degrees of freedom of the gravitational field @xcite@xcite .
j. e. shore and r. w. johnson , ieee trans .
theory * it-26 * , 26 ( 1980 ) ; y. tikochinsky , n. z. tishby and r. d. levine , phys . rev . lett .
* 52 * , 1357 ( 1984 ) and phys
. rev . * a30 * , 2638 ( 1984 ) ; j. skilling , ` the axioms of maximum entropy ' , in _ maximum - entropy and bayesian methods in science and engineering _ , ed . by g. j. erickson and c. r. smith ( kluwer , dordrecht , 1988 ) .
a. caticha , phys . lett . *
a244 * , 13 ( 1998 ) ; phys . rev . *
a57 * , 1572 ( 1998 ) ; ` probability and entropy in quantum theory ' , in _ maximum entropy and bayesian methods _ , ed . by w. von der linden et al .
( kluwer , dordrecht , 1999 ) ; found .
phys . * 30 * , 227 ( 2000 ) .
a. caticha , ` change , time and information geometry ' , in _ bayesian methods and maximum entropy in science and engineering _ , ed . by a. mohammad - djafari , aip conf
. proc . * 568 * , 72 ( 2001 ) ( online at arxiv.org/abs/math-ph/0008018 ) .
s. amari , _ differential - geometrical methods in statistics _
( springer - verlag , 1985 ) ; c. c. rodrguez , ` the metrics generated by the kullback number ' , in _ maximum entropy and bayesian methods _ , ed . by j. skilling ( kluwer , dordrecht , 1989 ) ; ` objective bayesianism and geometry ' , in _ maximum entropy and bayesian methods _ , ed . by p. f. fougre ( kluwer , dordrecht , 1990 ) ; and ` bayesian
robustness : a new look from geometry ' , in _ maximum entropy and bayesian methods _ , ed .
by g. r. heidbreder ( kluwer , dordrecht , 1996 ) .
f. weinhold , j. chem .
* 63 * , 2479 ( 1975 ) ; g. ruppeiner , phys . rev .
a * 20 * , 1608 ( 1979 ) and * 27 * , 1116 ( 1983 ) ; l. disi and b. lukcs .
a * 31 * , 3415 ( 1985 ) and phys
112a * , 13 ( 1985 ) ; r. s. ingarden , tensor , n. s. * 30 * , 201 ( 1976 ) .
a. caticha , ` maximum entropy , fluctuations and priors ' , in _ bayesian methods and maximum entropy in science and engineering _ , ed . by a. mohammad - djafari , aip conf .
* 568 * , 94 ( 2001 ) ( online at arxiv.org/abs/math-ph/0008017 ) .
the terms ` prior ' and ` posterior ' are normally used to refer to the distributions appearing in bayes theorem ; we retain the same terminology when using the me method because both bayes and me are concerned with the similar goal of processing new information to upgrade from a prior to a posterior .
the difference lies in the nature of the information involved : for bayes theorem the information is in the form of data , for the me method it is a constraint on the family of allowed posteriors .
in many important applications of the me method the prior happens to be uniform ( _ e.g. _ , the ` postulate ' of equal a priori probabilities in statistical mechanics ) and it is easy to overlook its presence and the important role it plays . j. w. york , ` kinematics and dynamics of general relativity ' , in _ sources of gravitational radiation _
, ed . by l. smarr ( cambridge univ
. press , 1979 ) ; a. ashtekar , ` quantum mechanics of geometry ' online as gr - qc/9901023 ; c. rovelli , ` strings , loops and the others : a critical survey on the present approaches to quantum gravity ' , in _ gravitation and relativity : at the turn of the millenium _ , ed . by n. dadhich and j. narlikar ( poona univ .
press , 1998 ) ; j. barbour , class .
* 11 * , 2853 and 2875 ( 1994 ) .
k. kuchar , ` time and interpretations of quantum gravity ' , in _ proceedings of the 4thcanadian conference on general relativity and astrophysics _ , ed . by g. kunsatter , d. vincent , and j. williams ( world scientific , singapore , 1992 ) ; c. isham ` canonical quantum gravity and the problem of time ' , in _ integrable systems , quantum groups and quantum field theories _ , ed . by l. a. ibort and m. a. rodriguez
( world scientific , singapore , 1993 ) . | i explore the possibility that the laws of physics might be laws of inference rather than laws of nature
. what sort of dynamics can one derive from well - established rules of inference ?
specifically , i ask : given relevant information codified in the initial and the final states , what trajectory is the system expected to follow ?
the answer follows from a principle of inference , the principle of maximum entropy , and not from a principle of physics .
the entropic dynamics derived this way exhibits some remarkable formal similarities with other generally covariant theories such as general relativity . |
simulations of the local group s formation ( klypin et al . 1999 )
predict that an order of magnitude more satellites should be associated with the milky way and m31 than are actually observed .
this discrepancy is a significant challenge to hierarchical clustering scenarios
. an intriguing suggestion as to the whereabouts of the `` missing '' satellites is provided by the local group infall model of blitz et al .
( 1999 ) , who speculate that a large fraction of the classical ensemble of hvcs are these local group building blocks .
the continuing infall of the hvcs onto the disk of the galaxy would then provide the bulk of the fuel necessary to maintain ongoing star formation .
hvcs are ubiquitous ( @xmath020% sky covering fraction ) clouds seen in hi emission , whose velocities are incompatible with simple models of galactic rotation . because the majority of their distances are effectively unconstrained , rampant speculation exists as to their exact nature and origin , ranging from solar metallicity galactic fountain gas ( @xmath1kpc and z@xmath0z@xmath2 ) , to magellanic cloud tidal debris ( @xmath3kpc and [email protected]@xmath2 ) , to the blitz / spergel local group formation remnants ( @xmath5kpc and 0.0@[email protected]@xmath2 ) . the fact that each scenario makes specific predictions regarding the distance and metallicity for the `` typical '' hvc means that , in principle , the above models could be distinguished from one another wth appropriate observations . in practice ,
the definitive observational discriminant has been difficult to obtain .
the cleanest discriminant between the competing hvc models is that of their distance . if it could be shown that the majority of hvcs reside in the galactic halo , as opposed to being distributed throughout the local group
, one could sound the death knell for the blitz / spergel model .
unfortunately , direct distance determinations for hvcs are few and far between ; to set a useful upper limit requires a suitably bright background halo star of known distance to lie directly behind a high hi column density hvc .
the dearth of catalogued blue horizontal branch stars and early subdwarfs in the outer halo ( rr lyrae stars can sometimes be employed , in a pinch ) is one immediate problem ; those bright enough to obtain high s / n , high - resolution spectra ( to actually search for hvc absorption features ) are rarer still .
non - detections ( both for foreground and background probes ) are more difficult to interpret , as fine - scale hi structure may conspire to make the probe `` miss '' any intervening hi . to date , there are only five hvcs for which either an upper limit or distance bracket exists . as table 1 shows , of these five hvcs none is consistent with an intra - local group residence , as might be expected under the blitz / spergel picture . an ongoing attempt to detect complex wd in absorption towards a distant halo rr lyrae star
may soon add a sixth entry to table 1 ( comeron 2000 ) .
a few other hvcs have solid lower distance limits , but they do not provide any discriminant between halo and local group residence ( being only @xmath41@xmath75kpc ) . these are therefore not reported here .
lcl hvc & distance ( kpc ) & reference + 100@xmath7@xmath87@xmath9110 & @xmath101 & bates et al .
( 1991 ) + complex m & @xmath104 & ryans et al .
( 1997 ) + 328@xmath716@xmath9100 & @xmath1111 & sembach et al .
( 1991 ) + complex a & 4@xmath710 & van woerden et al .
( 1999 ) + 279@xmath733@xmath9120 & @xmath1150 & richter et al .
( 1999 ) + the background stellar probe technique described above is virtually impossible to apply to any potential local group hvc at @xmath12400kpc .
perhaps the most promising method for attempting to prove an hvc truly lies at @xmath0mpc distances lies in the detection of the tip of the red giant branch in any putative stellar population associated with the hvc ( grebel et al .
2000 ) .
recently , combes & charmandaris ( 2000 ) have shown that both the wakker & schwarz ( 1991 ) and braun & burton ( 1999 ) hvcs ( at 1@xmath13 and 30@xmath13 resolution , respectively ) follow closely the size - linewidth relation defined by galactic molecular clouds , _ provided that their mean distances are @xmath020kpc .
this is _ indirect evidence against the blitz / spergel picture , but concerns regarding the use of the size - linewidth technique as a distance determinator must be heeded ( wakker & van woerden 1997 ; 4.1 ) . _ _
both blitz et al . ( 1999 ) and braun & burton ( 1999 ) have used the fact that the dispersion @xmath14 in the hvc distribution relative to @xmath15 is greater than the dispersion @xmath16 relative to @xmath17 or @xmath18 as support for preferring the galactic and local group standards of rest , over the local standard of rest . they use this as indirect support for an extragalactic origin for many hvcs .
it should be stressed that , while @xmath19 _ is a necessary condition for the blitz / spergel picture , it does _ not constitute sufficient proof .
any model that predicts a sinusoidal @xmath15 vs. galactic longitude distribution , necessarily satisfies the same @xmath20 condition , a wholly underappreciated fact .
specifically , @xmath19 for _ all local group infall _ and galactic fountain _ and magellanic stream disruption models .
in addition , there is a significant selection effect at play in these @xmath16 vs @xmath14 comparisons in that @xmath21@xmath6100kms@xmath22 hi is not included in the @xmath16 @xmath23 @xmath14 conversion .
any effect this `` missing '' gas might have upon the resulting distribution was neglected by blitz et al .
( 1999 ) and braun & burton ( 1999 ) . _
_ _ _ _ the superposition of wakker s ( 1990 ; ch .
5 ) galactic fountain and magellanic stream models results in an hvc flux distribution indistinguishable from that observed . specifically , sum figures 9(b ) and 9(d ) of ch . 5 in wakker ( 1990 ) and contrast with figure 5 of ch .
2 . not only is the flux distribution reproduced , but so are many secondary signatures , including the asymmetry in the velocity extrema ( blitz et al .
1999 ; 5.2.3 ) . to be fair
, it should be noted that the spatial deployment of the hvcs ( in the @xmath24 aitoff projection ) in wakker s galactic fountain @xmath9 magellanic stream model ( top panels of fig . 3 and 8 of ch .
5 of wakker 1990 ) appears inferior to that of the blitz et al .
( 1999 ; fig . 13 ) model .
new galactic fountain models from de avillez ( 2000 ) may reduce this discrepancy .
high spectral resolution and high s / n ultraviolet spectra taken with spectrometers aboard hst and fuse have enabled major advances in quantifying the distribution of hvc metallicities . preferred atomic lines for such abundance analyses include those of sii and oi ( @xmath25-elements , minimal dependency upon ionization corrections , relatively insensitive to dust depletion ) ; other popular lines include those of feii , siii , and mgii ( albeit subject to large depletion corrections ) and ni ( dominated by secondary production at low metallicities , however ) . regardless of atomic species considered , _
many abundance determinations are susceptible to uncertain dust depletion corrections , _ most are susceptible to uncertain ionization corrections , and _ all are subject to ( potentially considerable ) spatial resolution uncertainties .
the latter can not be emphasized strongly enough .
the absorption line is probing sub - arcsecond scales in the intervening hvc gas , while the 21 cm hi is probing a scale 2@xmath73 orders of magnitude greater ! substantial systematic uncertainties are typically incorporated into the final quoted result to reflect this `` resolution '' uncertainty , but one can never be certain that a pathological case has not been encountered . _ _ _ recall from 1 that in the blitz / spergel picture , metallicities of the order @xmath60.1z@xmath2 are to be expected ( but not necessarily primordial ) , while under a galactic fountain scenario values of @xmath0z@xmath2 should be more prevalent .
magellanic cloud debris might be expected to lie in the @[email protected]@xmath2 regime .
table 2 lists eight recent hvc metallicity determinations ; it is by no means complete , but it does provide a representative sample .
lcl hvc & abundance & reference + complex c & [email protected] / h@xmath2 & wakker et al .
( 1999 ) ; + & & gibson et al .
( 2001a ) + complex wb & @xmath40.1ca / h@xmath2 & robertson et al .
( 1991 ) + 287@xmath922@xmath9240 & 0.3s / h@xmath2 & lu et al .
( 1998 ) + magellanic stream & 0.3s / h@xmath2 & gibson et al .
( 2000 ) + 225@xmath783@xmath7200 & @xmath40.3fe / h@xmath2 & gibson et al .
( 2001b ) + 279@xmath733@xmath9120 & @xmath40.5fe / h@xmath2 & richter et al .
( 1999 ) + 258@xmath739@xmath9232 & @xmath40.6si / h@xmath2 & sahu & blades ( 197 ) + 100@xmath7@xmath87@xmath9110 & 0.7o / h@xmath2 & bates et al .
( 1990,1991 ) + an immediate conclusion to be drawn from table 2 is that , _ to date , no hvc shows unequivocal evidence for [email protected]@xmath2 ( blitz / spergel ) .
conversely , no hvc shows unequivocal evidence for [email protected]@xmath2 ( galactic fountain ) , an interesting conundrum !
future hst / stis and fuse analyses are sorely needed , particularly in light of the fact that the anti - local group barycentre clouds , where the blitz / spergel picture predicts an excess of intra - local group hvcs , have yet to be sampled in uv absorption . _ according to simple models of galactic chemical evolution with gas infall ( tosi 1988 ) , z@xmath27 must be @xmath60.1z@xmath2 ; for z@xmath28 , the resulting stellar population distribution violates the present - day galactic disk constraints .
table 2 would then imply that the majority of hvcs are not representative of the class of infalling gas clouds invoked by theorists to explain the g - dwarf problem ( the status of complex c is still being debated at this time ) .
while not listed in table 2 , hvc 125@xmath941@xmath7207 deserves special mention .
this hvc has received attention from braun & burton ( 1999 ) owing to its exceptionally narrow core hi linewidth ( @xmath62kms@xmath22 ) .
an indirect argument based upon thermal and pressure characteristics suggests that the distance to this hvc is of the order @xmath0700kpc ( the quoted distance is based upon an unpublished modification of the wolfire et al .
1995 model , and so a formal uncertainty in this number is not yet available ) .
very low abundances have been claimed for this hvc based upon mgii@xmath292796,2803 derived from the hst / ghrs spectrum of bowen & blades ( 1993 ) and bowen et al .
( 1995 ) along the line of sight to the background agn mrk 205 .
the former group found w@xmath30(mgii2796)=76@xmath3112m and w@xmath30(mgii2803)=46@xmath3117m , while the latter found w@xmath30(mgii2796)=169@xmath3128m and w@xmath30(mgii2803)=73@xmath3130m , based upon the same spectrum ; the source of the discrepancy is not stated .
our conservative analysis of their dataset yields w@xmath30(mgii2796)=130@xmath3140m and w@xmath30(mgii2803)=90@xmath3140m .
a @xmath32-value of 6@xmath311kms@xmath22 was found by all groups .
constructing curves of growth , our analysis implies a mgii abundance ( 68% c.l . ) of ( [email protected])mg / h@xmath2 .
dust depletion similar to that seen in the galactic halo clouds would increase this value by a factor of 3@xmath74 .
unfortunately , the low s / n fuse spectrum for mrk 205 has ( thus far ) been unable to shed new light on the metallicity of hvc 125@xmath941@xmath7207 , but a scheduled hst / stis programme ( pid#8625 ) should clarify the situation .
special mention should be made of the most spectacular of hvcs , the magellanic stream ( ms ) , an @xmath01000deg@xmath33 hi feature trailing the magellanic clouds .
the ms is the one hvc for which we know the origin : the clouds themselves .
what remains controversial is the exact mechanism by which this gas was extracted from said clouds : tides or ram pressure stripping ?
the key discriminants between the two scenarios are : ( i ) the prediction of a leading counterpart to the trailing stream ( tidal models ) ; and ( ii ) the tip of the stream ( msiv@xmath34msvi ) being 2@xmath73@xmath35 further away in the case of tides , than in the case of drag .
a consequence of ( ii ) is the prediction that , in the mean , the intensity of h@xmath25 emission i(h@xmath25 ) from the stream should decrease ( increase ) the further one moves down - stream , under the tidal ( drag ) scenario [ under the inherent assumption that i(h@xmath25 ) is driven primarily by photo - ionization , as opposed to collisional ionization ( c.f .
weiner et al .
2001 ) ] .
the extant observational data on the stream favours the tidal scenario , although there are admittedly one or two weak links : * existence of leading arm feature ( laf ) @xmath23 tides ( putman et al .
1998 ) * metallicity of leading arm hvc 287@xmath922@xmath9240 @xmath23 tides ( lu et al .
1998 ) * _ if i(h@xmath25)@xmath36@xmath10i(h@xmath25)@xmath37 @xmath23 tides _ * i(h@xmath25)@xmath38@xmath39i(h@xmath25)@xmath40 @xmath23 not tides ( bland - hawthorn et al . 2001 ) * stars not observed in ms @xmath23 tides _ or drag ( yoshizawa 1998 ; mihos 2001 ) _ * star streams in vicinity of lmc @xmath23 tides ( majewski et al .
1999 ) * _ if n@xmath41(65 kpc)@xmath610@xmath42@xmath43 _ or if n@xmath44(50kpc)@xmath45(5@xmath311)@xmath3510@xmath46@xmath47@xmath23 not moore & davis ( 1994 ) drag _ _
perhaps the most important conclusion to take from the study of hvcs is that they are _ not all the same beasts . to categorize all hvcs as due to a fountain ( galactic waste ) or infalling remnants of the local group s formation ( galactic fuel )
is doomed to failure from the start .
no doubt there are a plethora of origins scattered about the hvc family tree .
what can be concluded is that ( to date ) there is no unequivocal evidence for the existence of intra - local group hvcs , based upon limited distance , metallicity , and optical emissivity arguments . concerning the latter , neither weiner et al .
( 2001 ) nor tufte et al .
( 2001 ) have thus far observed an hvc consistent with the i(h@xmath25 ) predictions of the blitz / spergel model ( c.f .
bland - hawthorn et al . 2001 ) . on the other hand , neither is there unequivocal evidence for the existence of z@xmath4z@xmath2 hvcs , as might be expected for a galactic fountain origin ( intermediate - velocity clouds on the other hand do occasionally show z@xmath0z@xmath2 , but that is a story left for another day ) .
_ we acknowledge the financial support at the university of colorado of the nasa ltsa program ( nag5 - 7262 ) , the fuse science team ( nas5 - 32985 ) , and the fuse gi program ( nag5 - 9018 ) .
a special thanks to all those who helped bkg and jts consume excess quantities of excellent pisco sours , cigars , and sub - standard scotch
bland - hawthorn , phil maloney , sylvain veilleux , kevin mclin , d. j. pisano , ben weiner , and todd tripp .
bates , b , catney , m.g . , keenan , f.p .
1990 , , 242 , 267 bates , b , catney , m.g . , gilheany , s. , et al .
1991 , , 249 , 282 bland - hawthorn , j. , et al .
2001 , these proceedings blitz , l. , spergel , d.n . ,
teuben , p. , hartmann , d. & burton , w.b . 1999 , , 514 , 818 bowen , d.v .
& blades , j.c .
1993 , , 403 , l55 bowen , d.v .
, blades , j.c . &
pettini , m. 1995 , , 448 , 662 braun , r. & burton , w.b . 1999 , , 341 , 437 combes , f. & charmandaris , v. 2000 , , 357 , 75 comeron , f. 2000 , private communication de avillez , m.a .
2000 , preprint ( astro - ph/0001297 ) gibson , b.k . ,
giroux , m.l . ,
penton , s.v .
, et al . 2000 , , in press ( astro - ph/0007078 ) gibson , b.k . ,
giroux , m.l . ,
penton , s.v .
, et al . 2001a , , submitted gibson , b.k .
, et al . 2001b , in preparation grebel , e.k . ,
braun , r. & burton , w.b . 2000 , , 196 , # 28.09 klypin , a. , kravtsov , a.v . , valenzuela , o. & prada , f. 1999 , , 522 , 82 lu , l. , sargent , w.l.w . , savage , b.d .
, et al . 1998 , , 115 , 162 majewski , s.r . ,
ostheimer , j.c . , kunkel , w.e .
, et al . 1999 , in new views of the magellanic clouds , ed .
y .- h . chu et al .
( dordrecht : kluwer ) , 508 mihos , c. 2001 , these proceedings moore , b. & davis , m. 1994 , , 270 , 209 putman , m.e . , gibson , b.k . , staveley - smith , l. , et al .
1998 , nature , 394 , 752 richter , p. , de boer , k.s . , widmann , h. , et al .
1999 , nature , 402 , 386 robertson , j.g . ,
schwarz , u.j .
, van woerden , h. , et al .
1991 , , 248 , 508 ryans , r.s.i . ,
keenan , f.p .
, sembach , k.r .
, et al . 1997 , , 289 , 83 sahu , m.s . & blades , j.c . 1997 , , 484 , l125 sembach , k.r . , savage , b.d . & massa , d. 1991 , , 372 , 81 tosi , m. 1988 , , 197 , 47 tufte , s.l .
, et al . 2001 , in preparation van woerden , h. , schwarz , u.j . , peletier , r.f .
, et al . 1999 , nature , 400 , 138 wakker , b.p . 1990 , phd dissertation , groningen wakker , b.p . &
schwarz , u. 1991 , , 250 , 484 wakker , b.p . &
van woerden , h. 1997 , , 35 , 217 wakker , b.p . ,
howk , j.c . ,
savage , b.d .
, et al . 1999 , nature , 402 , 388 weiner , b.j .
, vogel , s.n . & williams , t.b .
2001 , these proceedings wolfire , m.g . , mckee , c.f . , hollenbach , d. , et al .
1995 , , 453 , 673 yoshizawa , a. 1998 , phd dissertation , tohoku university | tantalizing evidence has been presented supporting the suggestion that a large population of extragalactic gas clouds permeates the local group , a population which has been associated with the galactic high - velocity clouds ( hvcs ) .
we comment on both the strengths and weaknesses of this suggestion , informally referred to as the blitz / spergel picture .
theoretical predictions for the spatial and kinematic distributions , metallicities , distances , and emission properties of blitz / spergel hvcs will be confronted with extant observational data . |
in this paper , we study site percolation on the triangular planar lattice . recall that this can be viewed as a random coloring of the hexagonal cells of a honeycomb lattice , where the color ( black or white ) of each cell is chosen independently of the others : each of these cells has a probability @xmath6 to be black and @xmath7 to be white , for some parameter @xmath6 between @xmath8 and @xmath9 . in percolation theory ,
one is interested in the connectivity properties of the set of black hexagons ( or the set of white ones ) .
they can be regrouped into connected components ( or clusters ) .
the phase transition for percolation on this lattice occurs at @xmath10 .
often , it is described mathematically as follows , in terms of almost sure properties of percolation in the infinite lattice : when @xmath11 , there exists with probability @xmath9 no infinite cluster of black sites ( subcritical regime ) and an infinite cluster of white sites , and conversely when @xmath12 , there is an infinite cluster of black sites ( supercritical regime ) but no infinite cluster of white sites . in the critical case where @xmath10 , there exists neither an infinite white cluster , nor an infinite black cluster
but if one takes a finite large piece @xmath13 of the lattice , one will see white and black clusters of size comparable to that of @xmath13 .
see e.g. @xcite for an introduction to percolation .
a lot of progress has been made recently in the understanding of the large - scale behavior of critical percolation : in particular , smirnov @xcite proved conformal invariance of the connection probabilities , which allowed to make the link @xcite with the schramm - loewner evolution ( sle ) with parameter 6 introduced in @xcite , and to use the sle technology and computations @xcite to derive further properties of critical percolation , such as the value of some critical exponents , describing the asymptotic behavior of the probabilities of certain exceptional events ( arm exponents ) @xcite .
we refer to @xcite for a survey .
one precise relation to sle goes as follows : we consider the large equilateral triangle @xmath14 with even side length @xmath1 on the triangular grid such that the middle of the bottom part is the origin and the top point is the point at distance @xmath15 above the origin .
we decide to color all cells on the boundary of the triangle , in white if their @xmath16-coordinate is positive and in black if their @xmath16-coordinate is negative , and we perform critical percolation in the inside of @xmath14 .
then , we consider the interface @xmath17 ( viewed as a path on the hexagonal lattice dual to the triangular lattice ) between the set of black sites attached to the left part of the triangle and the set of white sites connected to the right part of the triangle ( see figure [ triangle ] ) . when @xmath18 , the law of the rescaled interface @xmath19 converges ( in an appropriate topology ) to that of the sle(6 ) process from @xmath20 to @xmath21 in the equilateral triangle with unit side length .
see @xcite for details , and e.g. @xcite for an introduction to sle .
thanks to this convergence result , one is able to deduce properties of critical percolation from the properties of sle . for instance ( and we shall come back to this later ) , one can prove that the typical number of steps of the path @xmath17 is of the order @xmath22 ( more precisely , for each @xmath23 , the probability that the number of steps is between @xmath24 and @xmath25 goes to @xmath9 as @xmath18 ) . and the interface @xmath17 ( sketch).,width=340 ] understanding the behavior of critical percolation allows also to derive some properties of percolation when the parameter @xmath6 is very close to @xmath26 .
these are the scaling ( and hyperscaling ) relations that were first developed in the physics literature ( the names of fischer , kadanoff and wilson are often associated to these ideas ) , and later rigorously derived in the case of percolation by kesten @xcite ( see also @xcite ) . before explaining these relations in a little more detail ,
let us first make a trivial observation to help the newcomer to get a feeling of what goes on .
suppose that we consider the percolation interface in the large equilateral triangle as before , but choose @xmath27 in such a way that @xmath28 .
then , one can couple it with a critical percolation interface ( i.e. with the case when @xmath10 exactly ) in the same triangle in such a way that the two paths are identical except on an event of probability that goes to @xmath8 as @xmath29 ( this is just because the probability that one of the neighbors of the interface changed its color is bounded by @xmath30 times the number of neighbors of the interface and we have just seen that this number is @xmath31 ) . hence ,
if one lets @xmath1 go to infinity , the scaling limit of the interface is still sle(6 ) . in particular , the probability
that the interface @xmath32 hits the right side of the triangle before the left side of the triangle let us call @xmath33 this probability goes to @xmath26 as @xmath18 . on the other hand , for any fixed @xmath34
, one can define @xmath35 for this choice of @xmath36 , if one looks at the possible limiting behavior of @xmath32 , it is clear that the law can not be exactly sle(6 ) anymore because it will hit the right side of the triangle before the left one with probability at least @xmath37 .
it is therefore natural to ask what can happen to the scaling limit of this curve when @xmath18 in this regime , and to see how it is related ( or not ) to sle(6 ) .
one can equivalently define the so - called correlation length @xmath38 in such a way that @xmath39 .
in other words , for @xmath40 , @xmath41 ( note that for @xmath42 fixed , @xmath43 as @xmath18 ) .
kesten @xcite has shown that it is possible to deduce from the arm exponents of critical percolation the behavior of @xmath44 as @xmath45 .
this derivation relies on the `` four - arm exponent '' : the crossing probability increases exactly when `` pivotal '' sites are flipped . combining kesten s results with the exponents computed using sle
, one gets @xcite that @xmath46 when @xmath47 , for any fixed choice of @xmath34 .
this is ( see @xcite ) a crucial step in the rigorous proof of the fact that the `` density '' @xmath48 of the infinite cluster for percolation with parameter @xmath6 decays like @xmath49 as @xmath6 decays to @xmath50 .
we shall see ( and this is quite easy ) that in order to get a non - trivial limit for @xmath32 ( i.e. neither sle(6 ) nor a path that just sticks to the boundary of @xmath51 ) , one has to take @xmath52 in such a way that @xmath1 is of the order of the correlation length @xmath44 i.e. that @xmath53 $ ] for some @xmath54 . the terminology to describe this regime depends on the circumstances and the papers : near - critical , off - critical , finite - size scaling .
it is also very closely related to the scaling and hyperscaling relations .
anyway , it has been the subject of numerous and interesting works , see e.g. @xcite and the references therein .
one general rule behind the derivation of these results is that as long as @xmath55 , the behavior remains roughly the same as at criticality .
we will remind some related results in the next sections , but for example , the probabilities of existence of crossings of annuli for near - critical percolation are bounded by constant times those at criticality .
this implies that the exponents describing critical percolation also describe off - critical percolation and lead to kesten s scaling relations . in @xcite , camia ,
fontes and newman have suggested that the scaling limit of the near - critical picture could be obtained from that of the critical picture by just opening a poissonian family of `` pivotal '' points .
we understand that @xcite have succeeded in proving this fact .
a consequence would be that the subsequential limits that we are discussing in the present paper are in fact limits , and that they are all related to each other by scaling .
let us also mention that near - critical percolation is closely related to the questions concerning the noise - sensitivity of percolation , such as studied in @xcite . before describing the results of the present paper ,
let us just emphasize the relevance of such near - critical models for physical applications .
suppose for instance that one is considering a percolation model in a large box , and that one `` heats '' progressively the system i.e. that cells become progressively black one after the other ( and independently ) .
then , we know that the percolation parameter at which a left - to - right crossing of the big box will appear is ( with a large probability ) close to the critical one .
but , if one looks at it more precisely , it is easy to see that the value at which the crossing occurs will be @xmath56 where @xmath57 is a uniform random variable in @xmath58 $ ] .
hence , at this moment , the picture and its fine properties are that of near - critical percolation .
another example where this near - critical picture appears spontaneously is that of gradient percolation , an inhomogeneous percolation model introduced in @xcite and studied mathematically in @xcite
. our focus here will be on the scaling limits of near - critical interfaces .
here is a list of results that we shall derive in the present paper . choose a sequence @xmath59 and study the behavior of the law of @xmath32 as @xmath18 : * for any @xmath34 , if we choose @xmath0 in such a way that @xmath0 is close to @xmath60 , then there exist subsequential limits for the law of @xmath32 . * the limiting laws are all singular with respect to the law of sle(6 ) . * the only other possible scaling limits are sle(6 ) itself ( when @xmath0 is too close to @xmath2 ) or a degenerate case where @xmath61 follows ( the right part of ) the boundary of the triangle ( this is when @xmath0 is too large ) . * the critical exponents associated to these non - degenerate scaling limits are the same as those of sle(6 ) .
in particular , the hausdorff dimension of the curves is almost surely @xmath5 . and
at @xmath62 of figure [ nicepic ] depicted separately.,width=384 ] the first result is a rather direct application of the tightness arguments of aizenman and burchard @xcite and the definition of the correlation length .
the last two statements follow from the ideas of kesten s paper @xcite .
our main result is probably the second one , the fact that the laws of non - trivial scaling limits are singular with respect to sle(6 ) .
one way to explain it is the following . in the off - critical regime , one sees on a macroscopic scale a difference between the law of the interface and that of the critical percolation interface ( i.e. the non - critical interface is more to the `` right '' for instance ) .
if one zooms in by a factor @xmath63 , one still sees a difference , but this difference tends to disappear , because one is not looking at a picture of size ca . @xmath44 any more , but at a picture of size @xmath64 .
the question is whether this difference disappears sufficiently fast when @xmath65 or not .
note that one can show that ( just as for the critical interface ) the number of boxes of size @xmath66 visited by the path is of order @xmath67 when @xmath63 is large ( and @xmath1 very large ) .
either this difference vanishes fast with @xmath63 and one is not able to almost surely detect a difference between the two macroscopic interfaces , or the difference between these two behaviors can be detected by averaging them out over the @xmath67 parts of the path . in the end
, one has to compare certain critical exponents to decide which scenario is correct and it turns out that it is the second one .
a `` flavor '' of supercritical percolation is therefore still present in the scaling limit .
we would like to emphasize that we do not really use any complex analytic sle technology to derive our results .
our considerations are based on percolation techniques , kesten s scaling ideas and the knowledge of the exact value of the critical exponents ( the derivation of which however used sle ) .
this paper will build on earlier results on near - critical percolation , and in particular on kesten s paper @xcite . in this section
, we recall some results that we shall use .
all these results are stated and derived in @xcite .
see also the last section of @xcite for a rough survey . throughout the paper , we will use the notation @xmath68 to say that the ratio between two quantities remains bounded away from @xmath8 and @xmath69 . recall our definition of the correlation length @xmath70 note that @xmath71 is non - increasing with respect to @xmath6 and non - decreasing with respect to @xmath72 .
we define the event @xmath73 that there exist four paths of alternating colors ( when ordered clockwise around the origin ) , each of them joining a neighbor of the origin to the circle of radius @xmath74 .
similarly , we define the event @xmath75 ( resp .
@xmath76 ) that there exist two paths of different colors ( resp .
one black path ) from a neighbor of the origin to the circle of radius @xmath74 .
we will use the following properties , that hold for any fixed @xmath34 : 1 .
the russo - seymour - welsh estimates remain valid below @xmath77 : for all @xmath78 , there exists @xmath79 such that for all @xmath6 , for all @xmath80 , the probability ( for @xmath81 ) that there is a black ( resp .
white ) horizontal crossing of a @xmath82 parallelogram is at least @xmath83 .
2 . let @xmath84 denote the event that there exist two arms of different colors joining the circles of radii @xmath85 and @xmath86 .
then ( this follows from the `` arm - separation lemmas '' ) , @xmath87 uniformly for @xmath88 and @xmath89 ( this is known as the _ quasi - multiplicativity _ property ) .
the same is true for four arms and one arm , _
i.e. _ with @xmath90 or @xmath91 instead of @xmath92 .
[ multip ] 3 .
it can be used to prove that @xmath93 uniformly for @xmath94 and @xmath95 .
[ summation ] 4 .
we have @xmath96 uniformly for @xmath94 and @xmath97 . [ arms_near ] 5 .
finally , @xmath98 as @xmath47
. [ equiv ] the last property , which holds for each @xmath99 , implies in particular that for any fixed @xmath100 , @xmath101 as @xmath18 , and that @xmath102 as @xmath47 . note that the combination of items [ multip ] . and [ arms_near ] .
shows that @xmath103 uniformly for @xmath94 , @xmath104 ( and similarly for @xmath90 ) .
recall also @xcite ( see also @xcite ) that for any @xmath105 , @xmath106 as @xmath107 , where @xmath108 and @xmath109 as @xmath110 , which in turn implies that @xmath111 as @xmath107 .
analogous statements hold for any number of arms ( see @xcite ) .
note that the correlation length introduced here differs slightly from the usual one let us denote it here by @xmath112 that is defined in terms of crossing probabilities of rectangles or rhombi ( the one introduced in @xcite , and used for instance in @xcite ) .
using russo - seymour - welsh considerations ( and also the fact that for any @xmath100 , @xmath113 see @xcite ) , these two definitions can easily be shown to be equivalent : for any @xmath99 , @xmath114 .
we are thus allowed to use for @xmath71 results established for @xmath115 .
alternatively , we could have chosen to work with rhombi instead of triangles and then , we would have used directly the usual definition of correlation length .
these two definitions are also known to be equivalent to other `` natural '' correlation lengths , describing for example the mean radius of a finite cluster or the rate of decay of connectivity properties ( see e.g. the discussion in section 2.2 of @xcite ) .
for simplicity of presentation , we will stick to the setup that we described in the introduction , even if the shape of the considered domain ( in our case , it is the triangle ) could be chosen arbitrarily .
let us describe this setup more precisely .
we follow here the definitions of @xcite , and refer the reader to this paper for more details .
we consider @xmath51 the ( filled ) equilateral triangle of unit side length , with corners @xmath116 , @xmath117 and @xmath21 .
the rescaled interfaces will be elements of @xmath118 , the space of curves in @xmath51 : these are equivalence classes of continuous functions from @xmath119 $ ] to @xmath51 , where two functions @xmath120 and @xmath121 represent the same curve if and only if there exists a continuous increasing bijection @xmath122 \to [ 0 , 1]$ ] such that @xmath123 .
we endow this space with the quotient metric : @xmath124 } |f_1 - f_2 | \bigg)\ ] ] where the infimum is over the set of increasing bijections @xmath125 from @xmath126 $ ] onto itself .
we call @xmath127 the law of the rescaled interface @xmath128 of percolation with parameter @xmath6 in our triangle ( with mesh size @xmath129 ) : this is a probability measure on @xmath118 .
endowed with the previous metric , @xmath118 is a complete separable space , so that tightness and relative compactness are equivalent ( by prohorov s theorem ) .
the family @xmath130 , n \ge 1)$ ] is relatively compact in @xmath131 .
* let us consider a sequence @xmath132 in this family .
our goal is to find a converging subsequence : * by symmetry , we can assume that @xmath133 for all @xmath134 .
* if @xmath135 remains bounded along a subsequence , proving convergence of a subsequence of @xmath136 is trivial .
* we can therefore restrict ourselves to the case where @xmath137 .
suppose that for all @xmath34 , @xmath138 along a subsequence ; then it is easy to check ( using russo - seymour - welsh arguments ) that along that subsequence , @xmath139 converges in law to the concatenation of the two parametrized segments @xmath140 \cup [ 1 / 2 , \sqrt{3 } i / 2]$ ] . *
we now suppose that for some @xmath141 , @xmath142 remains bounded . in particular , we get that for some fixed @xmath143 , @xmath144 for all large enough @xmath145 ( indeed , by russo - seymour - welsh , one can increase @xmath71 by at least any constant factor by choosing a larger @xmath146 ) .
we can then use the machinery developed by aizenman and burchard ( theorem 1.2 in @xcite ) : russo - seymour - welsh estimates hold uniformly for all @xmath6 and @xmath147 , so that for any annulus @xmath148 , @xmath149 for two universal constants @xmath150 ( traversed means here that it visits the inner and the outer boundary of the annulus ) .
the bk inequality ( see e.g. @xcite ) then leads to the fact that @xmath151 which is exactly the hypothesis ( h1 ) of @xcite ( uniform power bounds on the probability of multiple crossings in annuli ) . + for the sake of completeness , let us just sketch why tightness can be derived from this .
first , the hypothesis ( h1 ) implies regularity properties ( theorem 1.1 in @xcite ) for the random curves @xmath152 : for any @xmath153 , there is some hlder continuity bound such that for each @xmath145 , there is a probability at least @xmath154 that @xmath152 can be parametrized and satisfy this bound .
in particular , we can exhibit some equicontinuous set such that for each @xmath145 , @xmath152 belongs to this set with probability at least @xmath154 .
tightness follows ( theorem 1.2 in @xcite ) using ( a slight adaptation of ) arzela - ascoli s characterization of compactness for continuous functions on a compact set .
hence , in all cases , one can find a converging subsequence of @xmath155 .
in this section , we will study the non - degenerate case where @xmath156 is not too large : we assume that @xmath157 $ ] for some fixed @xmath99 ( note that this includes the critical case ) .
we exhibit properties that near - critical interfaces share with critical ones .
we will first focus on an estimate for discrete interfaces and then study the scaling limits .
the length ( number of edges ) @xmath158 of the discrete interface @xmath17 in the finite - size scaling regime is roughly the same as that of the critical interface : [ proplength ] assume that @xmath157 $ ] .
then for any fixed @xmath159 , @xmath160 ) = 0.\ ] ] * proof . *
note that the hypothesis on @xmath0 implies in particular that @xmath161 , so that we can use the results on near - critical percolation that we have recalled in the preliminaries .
let us first derive the upper bound for @xmath162 : we estimate its expectation .
we note that for any edge @xmath163 on the dual hexagonal lattice , @xmath164 if and only if there exist two arms as depicted on figure [ 2arms ] , one black joining a neighbor of @xmath16 to the negative real half - axis , and one white joining a neighbor of @xmath16 to the positive real half - axis . , being on @xmath17 corresponds to the existence of two arms . for the lower bound
, we consider semi - annuli @xmath165 and we restrict to the edges at distance at least @xmath166 from the boundary of @xmath167 ( dashed).,width=302 ] in particular , @xmath168 it allows us to get an upper bound on @xmath169 : for each @xmath170 , and for some constants @xmath171 ( using the estimates on near - critical percolation and the fact that there are never more than @xmath172 points at distance @xmath173 from @xmath174 ) , @xmath175 the upper bound for @xmath162 follows immediately by markov s inequality . for the lower bound
, we use the standard second moment method . for that purpose , take @xmath176 such that @xmath177 and decompose the triangle @xmath14 into concentric `` semi - annuli '' @xmath178 , @xmath179 . note that the curve @xmath17 has to cross each of these semi - annuli . for any edge @xmath16 in @xmath167 at distance at least @xmath166 from the boundary of @xmath167 , the results that we have recalled together with the `` separation lemmas ''
imply that the probability @xmath180 that there exist two paths as before , from @xmath181 to the positive and negative real half - axes , that stay in @xmath167 , satisfies @xmath182 uniformly for such @xmath16 and @xmath183 .
such an edge @xmath16 is necessarily on @xmath17 by the previous remark .
let @xmath184 denote the number of these edges : we have @xmath185 for large @xmath173 ( and any fixed @xmath153 ) .
now , using the quasi - multiplicativity property and then item [ summation ] .
we get @xmath186 it follows that for some ( universal ) constant @xmath187 , @xmath188 consider some @xmath189 : since the events @xmath190 for @xmath191 are independent , we get that @xmath192 the lower bound follows readily because @xmath193 and our lower bound for @xmath194 .
we now show that scaling limits of near - critical interfaces have the same hausdorff dimension as in the critical regime : assume that the law of the curve @xmath61 is the weak limit of @xmath132 , with @xmath195 and @xmath196 $ ] .
then the hausdorff dimension of @xmath61 is almost surely @xmath5 .
similar statements for @xmath74-point correlation functions , multiple crossings , multifractral spectrum etc . could be derived in the same way .
we leave this to the interested reader ( see the informal discussion at the end of the paper ) and we focus here only on this fractal dimension
. we might remark that the a priori estimates for the existence of two or three arms near a half - plane , or for the existence of five arms in the interior of the domain still hold ( uniformly ) for @xmath197 , as these are consequences of the russo - seymour - welsh estimates only .
hence , exactly as in the convergence to sle(6 ) case ( e.g. @xcite ) , the discrete hitting probabilities converge to the continuous hitting probabilities . * proof . *
the argument goes along similar lines as for the discrete length . for the upper bound , it suffices to prove that for all @xmath153 , the expected number of balls of radius @xmath198 needed to cover @xmath61 is bounded by @xmath199 .
but , we know that this expected number is bounded uniformly by this quantity for the discrete paths @xmath200 , and so this also holds in the scaling limit . for the lower bound , we decompose the triangle into concentric semi - annuli , and we exhibit a family @xmath201 of independent events of probability at least @xmath202 , such that on each @xmath203 , the curve @xmath61 has dimension @xmath5 .
again , we follow exactly the same idea ( with second moment estimates ) as in the discrete case , using the fact ( noticed in section [ prel ] ) that the discrete estimates for @xmath204 hold uniformly .
we now prove that for the scaling limits of percolation interfaces , there are only three main options . [
dichotomy ] suppose that @xmath205 is the weak limit of a sequence @xmath136 when @xmath206 , with @xmath195 and @xmath207 .
then : * either @xmath205 is the dirac mass on the concatenation of the two boundary segments @xmath208 $ ] and @xmath209 $ ] , * or @xmath205 is the law of sle(6 ) , * or for some @xmath210 , @xmath211 for all large @xmath145 .
* suppose now that @xmath205 is the limiting law of a sequence @xmath212 defined under @xmath213 , where @xmath195 and @xmath214 .
* let us first suppose that for each @xmath34 , @xmath215 for all sufficiently large @xmath145 .
this implies readily ( by standard russo - seymour - welsh arguments ) that for each fixed @xmath146 , @xmath216 .
it follows readily that the probability that there exists a closed cluster of radius greater than any fixed positive @xmath57 for @xmath213 in the rescaled triangle goes to @xmath8 as @xmath206 and the @xmath205 must be supported on the path that goes along the right boundary of the triangle .
* let us now suppose that for each @xmath217 , @xmath218 for all sufficiently large @xmath145 .
we assume that the law @xmath205 is not the law of an sle(6 ) and try to reach a contradiction . in the discrete
setting , it is always possible to couple @xmath219 ( corresponding to @xmath220 ) with a realization of the interface @xmath221 for @xmath222 in such a way that the sites that are black for the realization of @xmath222 are also black for the realization of @xmath223 . in this way , the path @xmath219 is more `` to the right '' than @xmath221 . letting @xmath206
, we see that we can couple a realization @xmath224 of @xmath205 with an sle(6 ) @xmath225 in the triangle @xmath51 with the same property .
+ it follows from our results on hausdorff dimensions ( in fact the uniform russo - seymour - welsh bounds would suffice ) that @xmath224 is almost surely a continuous curve with zero lebesgue measure .
clearly , one can divide the set of connected components of @xmath226 into two parts . intuitively speaking , those `` to the right of @xmath224 '' and those `` to the left of @xmath224 '' .
let us call @xmath227 and @xmath228 the respective unions of these components .
similarly , we can define @xmath229 and @xmath230 .
then , we have that @xmath231 .
+ the fact that the laws of @xmath224 and @xmath225 are not identical implies that with positive probability the two open sets @xmath230 and @xmath228 are not equal .
it follows ( using also the fact that @xmath224 is a curve that does not backtrack on its own past ) that for some @xmath232 with rational coordinates and for some @xmath233 with rational coordinates , if we denote by @xmath234 the disc centered at @xmath235 with radius @xmath236 , @xmath237 we also choose @xmath236 small enough so that the distance between @xmath235 and the boundary of the triangle is at least @xmath238 . from now on , @xmath235 and @xmath236
will be fixed .
+ the convergence of the discrete interfaces to the continuous ones imply that for some positive constant @xmath239 , @xmath240 for any large @xmath145 , with obvious notation .
note that these discrete events do not depend on the state of the sites inside @xmath241 as they imply that neither @xmath219 nor @xmath221 did intersect the disc .
so , it follows from the uniform rsw estimates that if we define the event @xmath242 that there exist a white circuit in @xmath243 for @xmath223 and a black circuit for @xmath222 in the same annulus ( where both circuits disconnect @xmath235 from @xmath244 ) , we get that for some absolute positive constant @xmath245 and all sufficiently large @xmath145 , @xmath246 + it follows readily that if one increases continuously @xmath6 from @xmath2 to @xmath247 , there will exist with probability bounded from below
a value @xmath6 at which the interface jumps from the left to the right of @xmath241 without hitting it . at this @xmath6 , there exist necessarily four arms of alternating colors originating at the site @xmath16 that has been flipped .
the different possibilities are depicted in figure [ flip ] .
+ .,height=412 ] + for each given site @xmath16 on the grid @xmath248 , let us evaluate the probability that this occurs .
when @xmath16 is not close to the boundary of the triangle , i.e. if it is at distance at least @xmath249 , then we see that four arms of alternating colors must originate from @xmath16 all the way to distance @xmath249 from @xmath16 .
the probability is then bounded by @xmath250 ( recall that @xmath236 is fixed and that @xmath251 ) .
+ when @xmath16 is at distance @xmath252 from the boundary , then one must have four arms of alternating colors up to distance @xmath253 , and then two arms that cross a semi - annulus ( in a half - plane ) from radius @xmath253 to @xmath249 . combining this with quasi - multiplicativity and
the known bounds ( 2 arms in a half - plane , 5 arms in the plane that give an upper - bound for the four - arm probability ) leads to a bound of the type @xmath254 + for any @xmath173 , there are at most a constant times @xmath135 points on our grid that are at distance exactly @xmath253 from the boundary of the triangle .
hence , when we sum over all possible sites @xmath16 in the triangle , we get that the probability that between @xmath6 and @xmath255 a site @xmath16 is flipped in such a way that the curve jumps from the left to the right of the small disc is bounded by @xmath256 times @xmath257 ( for some constants @xmath258 that do not depend on @xmath145 ) . +
it finally follows that for some positive constant @xmath239 , @xmath259 but if we apply russo s formula in a similar way to evaluate @xmath260 together with the fact that the four - arm probabilities are uniformly controlled for @xmath55 , we get that for some positive @xmath261 , @xmath262 hence , there exists a positive constant @xmath263 such that @xmath264 for all sufficiently large @xmath145 . in other words ,
@xmath265 for all sufficiently large @xmath145 , and this contradicts our assumption .
* finally to conclude the proof of the proposition , let us suppose that we are not in the last scenario described in its statement i.e. that it is not true that for some @xmath266 , @xmath267 for all large @xmath145 .
then it means that a subsequence of @xmath268 falls into one of the two cases that we have just studied .
hence , the limiting probability measure is either the law of sle(6 ) or the dirac mass on the union of the two boundary segments .
in this section , we are going to assume that we are in the intermediate regime where the discrete percolation parameter @xmath6 used at scale n satisfies @xmath269\ ] ] for some fixed @xmath270 .
the previous alternative tells us that these are the `` non - trivial '' cases in the large - scale limit .
we are going to exhibit asymmetry properties that enable to detect the difference with a critical interface .
we will first derive a result in the discrete setting , and then a result in the continuous setting for the scaling limit . on the discrete level
, it is possible to explore the interface dynamically : each time one discovers a new hexagon , one tosses a coin to decide if it is black or white , so that with probability @xmath6 ( resp .
@xmath271 ) , one makes a 60 degree turn to the right ( resp . to the left ) .
of course , one often bumps into an already discovered hexagon , and in this case , the turn is already determined by a previous tossing .
hence , the percolation interface is the image under some map of a sequence of coin tosses .
let us call @xmath272 and @xmath273 the number of black ( resp .
white ) cells neighboring the path @xmath17 .
[ das ] suppose that @xmath274 $ ] .
then for any fixed @xmath153 , @xmath275 on the other hand , for critical percolation , @xmath276 * proof .
* define @xmath277 .
clearly the difference between the number of black neighbors discovered by @xmath17 and the number of white discovered neighbors evolves like a biased simple random walk with drift @xmath278 ( at each step the probability to add one is @xmath279 ) stopped after a certain random number of steps @xmath280 that is roughly of the order @xmath22 ( see proposition [ proplength ] ) .
hence , we see that this difference will be of the order @xmath281 ( with high probability ) i.e. more precisely that for all @xmath159 , @xmath282 in the off - critical case where @xmath283 , we see that the drift term dominates because of proposition [ proplength ] so that the first statement follows readily . similarly , when @xmath284 , the result follows directly from proposition [ proplength ] .
hence , the discrete interface in the off - critical regime where @xmath285 will have ( on a microscopical scale ) anomalously more `` black neighbors '' than `` white neighbors '' . moreover , the proof shows that one can detect a good approximation of the actual value of @xmath0 by looking at this difference because @xmath286 will be close to @xmath287 .
but this result is not sufficient to deduce asymmetry of scaling limits . indeed ,
if we had for instance taken @xmath288 for some @xmath289 , then an analogous discrete asymmetry result would hold i.e. for some small @xmath159 @xmath290 but we have proved in the previous section that in this case , the interface still converges to sle(6 ) as @xmath18 ( and we know that sle(6 ) is symmetric ) . the goal of this section is to prove the following result : suppose that @xmath61 is the limit ( in law ) of a sequence of @xmath200 s where @xmath291 $ ] ( and @xmath195 ) .
then the law of @xmath61 is singular with respect to that of sle(6 ) .
* let us fix a very small positive @xmath292 .
define the rectangle @xmath293\times [ 0 , 1/4]$ ] for a small fixed @xmath294 .
define the segments @xmath295 \times \{1/8\}$ ] and @xmath296\times \{1/4\}$ ] , and the smaller triangle @xmath297 ( see figure [ subsets ] ) . for each small equilateral triangle @xmath298 of size @xmath198 ( i.e. that one can obtain by translating @xmath299 )
, we define analogously the sets @xmath236 , @xmath300 , @xmath301 and @xmath302 .
we say that the triangle @xmath298 is _ good _ for @xmath61 , if at the first hitting time @xmath303 of @xmath304 , @xmath305 , and if @xmath61 does exit @xmath236 through @xmath301 after its first hitting time of @xmath300 .
, we consider the subsets @xmath306 , @xmath307 , @xmath308 and @xmath309 . in any small equilateral triangle @xmath298 of size @xmath198 , we define similarly the corresponding sets @xmath236 , @xmath300 , @xmath301 and @xmath302.,width=453 ] let us define the connected component @xmath310 of @xmath311 $ ] that has the top boundaries of @xmath302 on its boundary .
we consider the following three marked points on the boundary of @xmath310 : suppose that we go along the boundary of @xmath310 clockwise starting from the bottom right corner of @xmath302 ( that we call @xmath312 ) .
then @xmath313 is the right - most point on @xmath314 and @xmath315 is the left - most point on @xmath314 ( see figure [ pivotal ] ) .
we call @xmath316 ( resp .
@xmath317 ) the part of the boundary of @xmath310 between @xmath312 and @xmath313 ( resp . between @xmath315 and @xmath312 ) . is very good.,width=340 ] when @xmath298 is a good triangle , we define @xmath318 to be the probability given by cardy s formula of the existence in @xmath310 of a crossing from @xmath316 to the part @xmath319 of the boundary between @xmath320 and @xmath315 .
if after @xmath321 , @xmath61 does hit @xmath316 before @xmath317 , we then say that @xmath298 is _ very good _ for @xmath61 . if @xmath61 was an sle(6 ) , then @xmath318 would just be the conditional probability given @xmath322)$ ] that @xmath298 is very good for @xmath61 . but here , the law of @xmath61 is the limit of the law of near - critical percolation interfaces @xmath139 defined under @xmath213 , where @xmath323 and @xmath195 . by skorokhod
s representation theorem , we can couple on the same probability space samples of all @xmath219 s and @xmath61 in such a way that @xmath324 almost surely just as in the proof of convergence of critical percolation interfaces to sle(6 ) ( as in @xcite for instance ) . for @xmath219 , we can also check if a triangle @xmath298 is good and very good .
in fact , for each fixed triangle , the probability that its status for @xmath61 is not identical to its status for @xmath219 goes to zero as @xmath206 .
indeed , if we suppose for instance that @xmath298 is good for @xmath61 but not for @xmath219 , then it means that a three - arm event occurs for @xmath61 and a priori bounds do exclude this .
if we define the stopping time @xmath325 associated to the triangle @xmath298 and the path @xmath219 , this means that when @xmath298 is good , then @xmath326 $ ] converges almost surely to @xmath327 $ ] .
so , the domains @xmath328 ( with obvious notation ) converge in carathodory topology towards @xmath310 almost surely .
we now couple each realization of @xmath219 with a realization @xmath329 of critical percolation on the same lattice , in such a way that the realization used to define @xmath219 dominates @xmath329 ( _ i.e. _ its set of black sites is larger ) .
then ( see e.g. @xcite ) , it follows that the probability @xmath330 that there exists a crossing from @xmath331 to @xmath332 in @xmath328 for @xmath329 ( with natural notation ) converges almost surely to @xmath318 .
suppose now that the triangle @xmath298 is good for @xmath219 and consider the configurations as depicted on figure [ pivotal ] : we restrict ourselves to the case where there exist sites in a smaller rhombus of size comparable to @xmath333 from which four arms of alternating colors originate .
for each such site @xmath16 , and conditionally on @xmath326 $ ] , the probability that this four - arm configuration exists for percolation of parameter @xmath334 $ ] in the domain not yet explored by @xmath219 is bounded from below by a constant @xmath239 times @xmath335 , uniformly over @xmath6 , @xmath145 and the choice of @xmath6 ( this follows from the uniform estimates for four arms , the separation lemmas and russo - seymour - welsh ) .
the probability that this site is flipped from white to black when one increases @xmath6 from @xmath2 to @xmath247 , and that at this value @xmath6 the configuration is as described above , is therefore larger than @xmath336 finally , note that this event can happen for only one site ( when @xmath6 increases , since the crossing event is increasing , there can exist only one value of @xmath6 and therefore one site @xmath16 such that the state of @xmath16 is flipped exactly at @xmath6 and @xmath16 at that moment is pivotal ) .
hence , we can sum over all sites @xmath16 in the small rhombus , and deduce that the conditional probability given @xmath337 $ ] that @xmath298 is very good is bounded from below by @xmath338 ( we have used the quasi - multiplicativity to estimate the ratio on the second line ) for all small @xmath198 and uniformly with respect to @xmath145 and to the good configurations @xmath326 $ ] . letting @xmath206 , we conclude that for any good triangle @xmath298 , the conditional probability given @xmath339 $ ] that @xmath298 is very good for @xmath61 is at least @xmath340 we now want to estimate the number of good triangles @xmath298 .
we divide the triangle @xmath51 into @xmath341 small deterministic triangles . using the same method as when studying the dimension of @xmath61 ( and the arm - separation ideas that allow to estimate the probability that a triangle is good i.e. to say that the probability for a triangle @xmath298 to be good is comparable to the probability to be hit ) , we see that with probability going to @xmath9 as @xmath342 , the number of good triangles is at least @xmath343 .
we now count how many triangles are very good among the first @xmath344 good ones , and we subtract the sum of the @xmath345 for these good triangles . if @xmath61 was an sle(6 ) , the obtained quantity @xmath346 would satisfy @xmath347 and @xmath348 .
but , in our case , we see that @xmath349 and @xmath350 .
if the sequence @xmath351 has been chosen to converge sufficiently fast to zero , we see that almost surely ( using borel - cantelli ) , for all sufficiently large @xmath176 , @xmath352 whereas if @xmath61 had been an sle(6 ) , we would almost surely have had @xmath353 for infinitely many @xmath176 s .
we thus found an event that holds with probability one for @xmath61 , but with probability zero for sle(6 ) .
seemingly , the fact that the laws of near - critical interfaces are singular with respect to the law of sle(6 ) surprises some theoretical physicists who work on this topic .
recall that one important aspect of the sle approach to critical systems was precisely to show that critical conformally invariant models in the same `` universality class '' give rise to exactly the same curves in the scaling limit . for near - critical models that are not strictly conformally invariant such as near - critical percolation here
, this strong `` universality '' can fail to be true .
suppose that a random curve @xmath61 is the scaling limit of a near - critical percolation interface .
the tightness arguments , together with the estimates for existence of multiple arms indicate that the curve @xmath61 is almost surely a loewner chain , and that it can be defined via its random driving function @xmath354 .
it is of course a natural question to ask what this driving function could be ( recall that in the case of sle(6 ) , @xmath355 is a standard brownian motion ) .
note that after each ( fixed or stopping ) time , we expect the curve @xmath61 to turn `` more right '' as what a sle(6 ) would do in the same situation .
this leads naturally to the conjecture that ( if one extends properly the probability space ) one can couple @xmath354 with a standard brownian motion @xmath356 in such a way that @xmath357 is an increasing continuous process adapted to the filtration generated by the processes @xmath358 and @xmath359 .
one may think that this contradicts the fact that the law of @xmath360 is singular with respect to that of @xmath292 ( since the curves are generated by the driving functions , this is equivalent to say that the law of @xmath61 is singular with respect to that of sle(6 ) ) because the law of brownian motion with drift is known to be absolutely continuous with respect to that of brownian motion .
but this last fact is only valid under some regularity properties for the drift , and in the present case , one can expect this drift to be quite complicated ( its derivative measure might be supported on a set of exceptional times i.e. a fractal - type set ) .
we have seen that important properties ( such as the dimension of the curve ) are shared by the critical interfaces and the near - critical interfaces in the scaling limit , so that the technology based on conformal invariance of the critical model still provides the correct description of the near - critical interfaces in terms of exponents ( this was used in @xcite to describe the `` percolation front '' ) .
in fact , the near - critical percolation technology allows to prove the following result ( this is a direct consequence of the quasi - multiplicativity property and the fact that arm probabilities are comparable uniformly to the arm probabilities at @xmath4 ) : for any @xmath145 , there exists a constant @xmath361 , such that for any @xmath362 in @xmath51 , and for any sufficiently small @xmath72 , the ratio between the probability that @xmath61 visits the @xmath72-neighborhood of these @xmath145 points and the probability of the same event for sle(6 ) stays in @xmath363 for all small @xmath72 and uniformly over the choices of @xmath364 . loosely speaking ,
the `` finite - dimensional marginals '' of the law of @xmath61 are uniformly comparable to those of sle(6 ) ( even when the @xmath365 s are very close to each other ) .
note that this is the type of functions that is computable via the conformal field theory technology .
but this result does not contradict the fact that the law of @xmath61 is singular with respect to that of sle(6 ) .
note for instance that the constant @xmath361 does depend on the number of considered points .
our proof used the fact that we did not zoom on the behavior of the curve at just one point , but that we detected the asymmetry by looking at the mean of the behavior of the curve near to more and more points .
laboratoire de mathmatiques , bt .
425 , universit paris - sud 11 , 91405 orsay cedex , france dpartement de mathmatiques et applications , cole normale suprieure , 45 rue dulm , 75230 paris cedex 05 , france | we study the possible scaling limits of percolation interfaces in two dimensions on the triangular lattice . when one lets the percolation parameter @xmath0 vary with the size @xmath1 of the box that one is considering , three possibilities arise in the large - scale limit .
it is known that when @xmath0 does not converge to @xmath2 fast enough , then the scaling limits are degenerate , whereas if @xmath3 goes to zero quickly , the scaling limits are sle(6 ) as when @xmath4 .
we study some properties of the ( non - void ) intermediate regime where the large scale behavior is neither sle(6 ) nor degenerate .
we prove that in this case , the law of any scaling limit is singular with respect to that of sle(6 ) , even if it is still supported on the set of curves with hausdorff dimension equal to @xmath5 . |
in this paper we scratch an intelligent variable neighbourhood search ( int - vns ) aimed to achieve further improvements of a successful vns for the minimum labelling spanning tree ( mlst ) and the @xmath1-labelled spanning forest ( @xmath1lsf ) problems .
this approach integrates the basic vns with other complementary intelligence tools and has been shown a promising strategy in @xcite for the mlst problem and in @xcite for the @xmath1lsf problem .
the approach could be easily adapted to other optimization problems where the space solution consists of the subsets of a reference set ; like the feature subset selection or some location problems .
first we introduced a local search mechanism that is inserted at top of the basic vns to get the complementary variable neighbourhood search ( co - vns ) .
then we insert a probability - based constructive method and a reactive setting of the size of shaking process .
a labelled graph @xmath2 consists of an undirected graph where @xmath3 is its set of nodes and @xmath4 is the set of edges that are labelled on the set @xmath0 of labels . in this paper
we consider two problems defined on a labelled graph : the mlst and the @xmath1lsf problems .
the mlst problem @xcite consists on , given a labelled input graph @xmath2 , to get the spanning tree with the minimum number of labels ; i.e. , to find the labelled spanning tree @xmath5 of the input graph that minimizes the size of label set @xmath6 .
the @xmath1lsf problem @xcite is defined as follows .
given a labelled input graph @xmath2 and an integer positive value @xmath7 , to find a labelled spanning forest @xmath8 of the input graph having the minimum number of connected components with the upper bound @xmath7 for the number of labels to use , i.e. @xmath9 , the labelled subgraph @xmath10 may contain cycles , but they can arbitrarily break each of them by eliminating edges in polynomial time until a forest or a tree is obtained . therefore in both problems , the matter is to find the optimal set of labels @xmath11 . since a mlst solution would be a solution also to the @xmath1lsf problem if the obtained solution tree would not violate the limit @xmath7 on the used number of labels , it is easily deductable that the two problems are deeply correlated .
the np - hardness of the mlst and @xmath1lsf problems was stated in @xcite and in @xcite respectively .
therefore any practical solution approach to both problems requires heuristics @xcite .
the first extension of the vns metaheuristic that we introduced for these problems is a local search mechanism that is inserted at top of the basic vns @xcite .
the resulting local search method is referred to as _ complementary variable neighbourhood search _
( co - vns ) @xcite .
given a labelled graph @xmath12 with @xmath13 vertices , @xmath14 edges , and @xmath15 labels , co - vns replaces iteratively each incumbent solution @xmath11 with another solution selected from the _ complementary space _ of @xmath11 defined as the sets of labels that are not contained in @xmath11 ; @xmath16 .
the iterative process of extraction of a complementary solution helps to escape the algorithm from possible traps in local minima , since the complementary solution lies in a very different zone of the search space with respect to the incumbent solution .
this process yields an immediate peak of diversification of the whole local search procedure . to get a complementary solution , co
- vns uses a greedy heuristic as constructive method in the complementary space of the current solution . for the mlst and @xmath1lsf problems
the greedy heuristic is the maximum vertex covering algorithm ( mvca ) @xcite applied to the subgraph of @xmath17 with labels in @xmath18 .
note that co - vns stops if either the set of unused labels contained in the complementary space is empty ( @xmath19 ) or a final feasible solution is produced .
successively , the basic vns is applied in order to improve the resulting solution . at the starting point of vns
, it is required to define a suitable series of neighbourhood structures of size @xmath20 . in order to impose a neighbourhood structure on the solution space @xmath21
we use the hamming distance between two solutions @xmath22 given by @xmath23 where @xmath24 consists of labels that are in one of the solutions but not in the other .
vns starts from an initial solution @xmath11 with @xmath25 increasing iteratively from 1 up to the maximum neighborhood size , @xmath20 .
the basic idea of vns to change the neighbourhood structure when the search is trapped at a local minimum , is implemented by the shaking phase .
it consists of the random selection of another point in the neighbourhood @xmath26 of the current solution @xmath11 .
given @xmath11 , we consider its @xmath27 neighbourhood @xmath26 comprised by sets having a hamming distance from @xmath11 equal to @xmath25 labels , where @xmath28 . in order to construct the neighbourhood of a solution @xmath11
, the algorithm proceeds with the deletion of @xmath25 labels from @xmath11 .
the proposed intelligent metaheuristic ( int - vns ) is built from co - vns , with the insertion of a probability - based local search as constructive method to get the complementary space solutions .
in particular , this local search is a modification of greedy heuristic , obtained by introducing a probabilistic choice on the next label to be added into incomplete solutions . by allowing worse components to be added to incomplete solutions ,
this probabilistic constructive heuristic produces a further increase on the diversification of the optimization process .
the construction criterion is as follows .
the procedure starts from an initial solution and iteratively selects at random a candidate move .
if this move leads to a solution having a better objective function value than the current solution , then this move is accepted unconditionally ; otherwise the move is accepted with a probability that depends on the deterioration , @xmath29 , of the objective function value .
this construction criterion takes inspiration from simulated annealing ( sa ) @xcite .
however , the probabilistic local search works with partial solutions which are iteratively extended with additional components until complete solutions emerge . in the probabilistic local search , the acceptance probability of a worse component into a partial solution
is evaluated according to the usual sa criterion by the boltzmann function @xmath30 , where the temperature parameter @xmath31 controls the dynamics of the search .
initially the value of @xmath31 is large , so allowing many worse moves to be accepted , and is gradually reduced by the following geometric cooling law : @xmath32 , where @xmath33 and @xmath34 $ ] , with @xmath35 being the current best solution .
this cooling law is very fast and produces a good balance between intensification and diversification capabilities .
in addition , this cooling schedule does not requires any intervention from the user regarding the setting of its parameters , as it is guided automatically by the best solution @xmath35 .
therefore the whole optimization process is able to react in response to the search algorithm s behavior and to adapt its setting on - line according to the instance of the problem under evaluation @xcite .
the probabilistic local search has the purpose of allowing also the inclusion of less promising labels to incomplete solutions .
probability values assigned to each label are decreasing in the quality of the solution they give . in this way , at each step , labels with a better quality will have a higher probability of being selected ; the progressive reduction of the temperature in the adaptive cooling law produces , step by step , an increasing of this diversity in probabilities . at the beginning of int - vns
, the algorithm generates an initial feasible solution at random , that is the first current best solution @xmath35 , and set parameter @xmath20 to the number of labels of the initial solution ( @xmath36 ) .
then the _ complementary _ procedure is applied to @xmath35 to obtain a solution @xmath11 from the complementary space of @xmath35 by means of the probabilistic local search .
the complementary procedure stops if either a feasible solution @xmath11 is obtained , or the set of unused labels contained in the complementary space is empty producing a final infeasible solution .
subsequently , the shaking phase used for the basic vns is applied to the resulting solution @xmath11 .
it consists of the random selection of a point @xmath37 in the neighbourhood @xmath26 of the current solution @xmath11 , as in co - vns .
the successive local search corresponds also to that of co - vns .
however , since either co - vns , or the deletion of labels in the shaking phase , can produce an incomplete solution , the first step of the local search consists of including additional labels in the current solution in order to restore feasibility , if needed .
the addition of labels at this step is according to the probabilistic procedure .
then , the local search tries to drop labels in @xmath37 , and then to add further labels following the greedy rule , until @xmath7 labels emerge . at this step ,
if no improvements are obtained the neighbourhood structure is increased ( @xmath38 ) producing progressively a larger diversification .
otherwise , the algorithm moves @xmath11 to solution @xmath37 restarting the search with the smallest neighbourhood ( @xmath39 ) .
this iterative process is repeated until the maximum size of the shaking phase , @xmath20 , is reached .
the resulting local minimum @xmath11 is compared to the current best solution @xmath35 , which is updated in case of improvement ( @xmath40 ) . at this point a reactive setting for the parameter @xmath20 is used @xcite . in case of an improved solution , @xmath20 is decreased ( @xmath41 ) in order to raise the intensification factor of the search process .
conversely , in case of none improvement , the maximum size of the shaking is increased ( latexmath:[$q_{max } \leftarrow min(q_{max}+1 ; 2 \cdot algorithm . in each case , the adaptive setting of @xmath20 is bounded to lie in the interval between @xmath43 and @xmath44 to avoid a lack of balance between intensification and diversification factors .
the algorithm proceeds with the same procedure until the user termination conditions are satisfied , producing at the end the best solution to date , @xmath35 , as output .
the achieved optimization strategy seems to be highly promising for both labelling graph problems . ongoing investigation
consists in statistical comparisons of the resulting strategy against the best algorithms in the literature for these problems , in order to quantify and qualify the improvements obtained .
further investigation will deal with the application of this strategy to other problems . | in this paper we describe an extension of the variable neighbourhood search ( vns ) which integrates the basic vns with other complementary approaches from machine learning , statistics and experimental algorithmic , in order to produce high - quality performance and to completely automate the resulting optimization strategy .
the resulting intelligent vns has been successfully applied to a couple of optimization problems where the solution space consists of the subsets of a finite reference set .
these problems are the labelled spanning tree and forest problems that are formulated on an undirected labelled graph ; a graph where each edge has a label in a finite set of labels @xmath0 .
the problems consist on selecting the subset of labels such that the subgraph generated by these labels has an optimal spanning tree or forest , respectively .
these problems have several applications in the real - world , where one aims to ensure connectivity by means of homogeneous connections . |
@xmath1 is the low energy effective theory of the strong interactions .
it is given as a power expansion of the external four - momenta of the pseudo - goldstone bosons @xmath13 , @xmath14 and @xmath15 on the scale @xmath161 gev .
as a result , the expansion is typically valid up to @xmath17500 mev .
however , the constraints coming from the spontaneous / explicit chiral symmetry are not restricted to the low energy region @xcite . in this work
, we present a way of resummation of the @xmath1 series that in fact can be applied to any other system whose dynamics can be described by low energy chiral lagrangians .
we describe the successfull application of such approach to meson - meson interactions which are well reproduced up to @xmath171.2 gev .
let us a consider a partial wave amplitude @xmath18 with definite isospin ( @xmath19 ) .
we use a matrix formalism in order to deal with coupled channels . in this way
@xmath18 will be a matrix whose element @xmath20 represents the scattering of @xmath21 with angular momentum @xmath22 and isospin @xmath19 . if we consider only two body intermediate states unitarity with coupled channels reads in our normalization : @xmath23 where @xmath24 is a diagonal matrix with elements @xmath25 with @xmath26 the center mass three - momentum , @xmath27 and @xmath28 are the masses of the particles in the state @xmath29 and @xmath30 is the usual heaviside function .
( [ uni ] ) is a well known result and is the basis of the @xmath14 matrix formalism since all the dynamics is embodied in re@xmath5 which is @xmath31 .
the former equation shows clearly that , when considering @xmath5 , unitarity is exactly satisfied with two body intermediate states . from the @xmath1 expansion of @xmath32 , where @xmath33 and @xmath34 are the @xmath3 and @xmath4 contributions respectively , we work out the expansion of @xmath5 . in this way
we will obtain our approach for the @xmath14 matrix ( or re@xmath5 ) .
@xmath35^{-1}= t_2^{- 1}\cdot [ 1 + t_4 \cdot t_2^{- 1}+ ... ]^{- 1}\nonumber \\ & = & t_2^{- 1}\cdot [ 1 - t_4 \cdot t_2^{- 1}+ ...
]=t_2^{-1}\cdot [ t_2-t_4]\cdot t_2^{-1}\end{aligned}\ ] ] inverting the former result , one obtains : @xmath36^{-1}\cdot t_2 \nonumber \\
k&=&t_2\cdot \left[t_2-\hbox{re}t_4 \right]^{-1}\cdot t_2\end{aligned}\ ] ]
in @xcite we study the @xmath39 and @xmath40 partial waves . to make use of eq .
( [ t ] ) one needs the lowest and next to leading order @xmath41 amplitudes . in our case the @xmath42 and @xmath43
are taken from @xcite and the @xmath44 is also given in @xcite .
our amplitudes depend on six parameters @xmath45 , @xmath46 , @xmath47 , @xmath48 , @xmath49 and @xmath50 which are fitted to the elastic @xmath37 @xmath51 and @xmath52 phase shifts . in the following table
we show the resulting values for the @xmath53 coefficients comparing them with the @xmath1 values .
.@xmath53 coefficients . [ cols="^,^,^",options="header " , ]
we have presented a method of resummation of the @xmath1 series based in the expansion of @xmath5 . in this way
unitarity is fulfilled to all orders and resonances are well reproduced .
the method is rather general and could be applied to any system whose dynamics is described by chiral lagrangians .
we have applied it successfully to describe the s and p - wave meson - meson amplitudes giving rise to the resonances : @xmath6 , @xmath7 , @xmath8 , @xmath9 , the octet contribution to the @xmath10 , @xmath11 and @xmath12 .
200 j. a. oller , e. oset and j. r. pelez , phys .
( 1998 ) 3452 .
j. gasser and h. leutwyler , nucl .
b 250 ( 1985 ) 465 , 517 , 539 . f. guerrero and j. a. oller , nucl .
b 537 ( 1999 ) 459 .
j. a. oller , e. oset and j. r. pelez , phys .
d 59 ( 1999 ) 074001 . j. v. steele , h. yamagishi and i. zahed , nucl .
phys . a 615 ( 1997 ) 305 ; m. r. pennington and j. portoles , phys .
b 344 ( 1995 ) 399 .
v. bernard , n. kaiser and u. g. meissner , nucl .
b 357 ( 1991 ) 129 .
g. ecker , j. gasser , a. pich and e. de rafael , nucl .
b 321 ( 1989 ) 311 . | a non - perturbative method @xcite which combines constraints from chiral symmetry breaking and coupled channel unitarity is used to describe meson - meson interactions up to @xmath0 gev , extending in this way the range of applicability of the information contained in chiral perturbation theory ( @xmath1 ) @xcite , since this perturbative series is typically restricted to @xmath2 mev .
the approach uses the @xmath3 and @xmath4 @xmath1 lagrangians .
the seven free parameters resulting from the @xmath4 lagrangian are fitted to the experimental data .
the approach makes use of the expansion of @xmath5 instead of the amplitude itself as done in @xmath1 .
the former expansion is suggested by analogy with the effective range approximation in quantum mechanics and it appears to be very useful .
the results , in fact , are in good agreement with a vast amount of experimental analyses @xcite . the amplitudes develop poles corresponding to the @xmath6 , @xmath7 , @xmath8 , @xmath9 , the octet contribution to the @xmath10 , @xmath11 and @xmath12 @xcite . the total and partial decay widths of the resonances
are also well reproduced . |
the signature of a @xmath11-dimensional fractional brownian motion is a sequence of iterated stratonovich integrals along the paths of the fractional brownian motion ; it is an object taking values in the tensor algebra over @xmath12 .
+ signatures were firstly studied by k.t .-
chen in 1950 s in a series of papers @xcite , @xcite and @xcite . in the last twenty years
the attention devoted to signatures has increased rapidly .
this has been caused by the pivotal role they have in rough path theory , a field developed in the late nineties by terry lyons culminating in the paper @xcite , which is also at the base of newly developed theory of regularity structures @xcite .
the signature of a path summarises the essential properties of that path allowing the possibility to study spdes driven by that path .
+ + in 2015 hao ni and weijun xu @xcite computed the sharp rate of convergence for expected brownian signatures .
they obtained a rate of convergence of @xmath2 , where @xmath2 is the size of the mesh of the dyadic approximation .
+ however , for fractional brownian motion no progress has been made in this direction . in particular ,
the rate of convergence for the expected signature of the fractional brownian motion is not known for any value of the hurst parameter @xmath13 $ ] .
this article address this problem , obtaining the rate of convergence for @xmath0 $ ] . in order to obtain it we used the results of baudoin and coutin @xcite .
indeed , in 2007 baudoin and coutin computed the expected signature for fractional brownian motion for @xmath10 and also for small times for @xmath14 .
further works that analyse the properties of the signature of the fbm are @xcite and @xcite among others .
+ in this work we focus on the weak rate of convergence and we refer to the work of friz and riedel @xcite for the strong rate of convergence .
they obtained a rate of @xmath15 , while here we obtain a weak convergence rate of @xmath1 . + + moving away from the dyadic approximation and focusing just on the expected signature , we recall the work of chevyrev and lyons @xcite . in @xcite they showed that the expected signature has infinite radius of convergence , but not a factorial decay . in this work
we show that the expected signature has a factorial decay , indeed the bound for the @xmath4-th term of the signature is simply given by @xmath8 for all @xmath16 .
the sharp decay rate is expected to be @xmath17 , but it remains an open problem . in the @xmath18 case , our result gives an alternative proof , with sharper estimates , that the expected signature of fractional brownian motion has infinite radius of convergence , which by @xcite implies that the expected signature determines the signature in distribution .
our estimate is also sharper than the one obtained by friz and riedel @xcite , who proved a factorial decay of @xmath9 . + + in 2003 lyons and victoir @xcite developed a numerical methods for solving parabolic pdes and general sdes driven by brownian motions called cubature method on weiner space . in the cubature method the first step is to obtain the cubature formula in which the truncated signature of a path ( in the case of @xcite is the brownian motion ) is matched to a finite sum of dirac delta measures applied to the iterated integrals of deterministic paths . in this work
we give an extension of this result by obtaining the cubature formula for the fractional brownian motion for a particular case .
+ + this paper is structured in the following way . in section 2
we introduce some notations and state the main results . in section 3
we will discuss about preliminaries involving definitions and the results of other authors . in section 4 , 5 and 6
we prove the first three main results of the article . in section 7
we discuss the fourth result whose proof is in the appendix due to its long computations .
in this section we introduce the main results of the article . but first , we introduce some notations .
the notations used in this work are in line with the ones used in the papers of baudoin and coutin @xmath19 , lyons @xmath20 , and lyons and victoir @xmath21 and in the book by lyons , caruana , and lvy @xmath22 . + the fractional brownian motion is defined as follows .
let h be a constant belonging to ( 0,1 ) . a fractional brownian motion ( fbm ) @xmath23 of hurst index h is a continuous and centered gaussian process with covariance function @xmath24=\frac{1}{2}(t^{2h}+s^{2h}-|t - s|^{2h})\ ] ] from now on we will denote @xmath25 .
+ for @xmath3 then the fbm is a bm .
further , the multi - dimensional fractional brownian motion has coordinate components that are independent and identically distributed copies of one dimensional fbm .
+ now , we define the simplex @xmath26 $ ] , @xmath27:=\{(t_{1}, ... ,t_{k})\in[0,1]^{k}:t_{1}< ... <t_{k}\}.\ ] ] further , we define the following iterated integrals .
let @xmath28 be a word with length @xmath7 then @xmath29}db^{i}:=\int_{0\leq t_{1}< ... <t_{k}\leq 1}db^{i_{1}}_{t_{1}}\cdot\cdot\cdot db^{i_{k}}_{t_{k}}\ ] ] and @xmath29}db^{m , i}:=\int_{0\leq t_{1}< ...
<t_{k}\leq 1}db^{m , i_{1}}_{t_{1}}\cdot\cdot\cdot db^{m , i_{k}}_{t_{k}}\ ] ] where @xmath30 is the fractional brownian motion with hurst parameter @xmath16 and @xmath31 is its dyadic approximation .
in addition , @xmath32 is the @xmath33-th coordinate component of the fbm @xmath30 and the iterated integrals can be defined in the sense of young @xcite .
+ moreover , the dyadic approximation @xmath31 is defined as follows .
let @xmath34 for @xmath35 .
if @xmath36 $ ] then @xmath37 from now on we will denote @xmath38 .
+ we can now present our main results .
the first result is about the rate of convergence of the expected signature of the dyadic approximation of the fbm to its exact value .
[ pr1 ] let @xmath39 .
letting @xmath40 be a word where @xmath41 for @xmath42 , then for all m @xmath43}db^{i}\right)-\mathbb{e}\left(\int_{\delta^{2k}[0,1]}db^{m , i}\right)\bigg|\leq c2^{-2hm}\ ] ] where c is a finite constant and independent of m. it is important to stress that for the brownian motion case , i.e. @xmath44 , ni and xu in @xcite proved that the sharp rate of convergence is given by @xmath2 , which is in line with the result presented here .
moreover , the proof used in @xcite can not in principle be used to prove our result since ni and xu used the independence of the increments property ( i.e. the semigroup property ) of the brownian motion , which does not hold for the fbm .
conversely , the proof used to prove * theorem 2.2 * is based on the integral form of the covariance function of the fbm which is valid only for @xmath39 , hence our proof can not in principle be used to prove the result in @xcite .
+ in the next theorem we focus on and refine the value of the coefficient @xmath45 of the previous theorem and we provide a bound for it .
[ pr2 ] let @xmath39 .
letting @xmath40 be a word where @xmath41 for @xmath42 , then @xmath46}db^{i}\right)-\mathbb{e}\left(\int_{\delta^{2k}[0,1]}db^{m , i}\right)\bigg|\leq \dfrac{ah(2h-1)}{(k-1)!2^{k}}\ ] ] where a is a finite constant and independent of m and k. the following theorem provides a bound for the value for the expected signature of the fbm . as mentioned in the introduction the following result is a sharper estimates than the one obtained by chevyrev and lyons @xcite and the one by friz and riedel @xcite .
our result in particular implies , in the @xmath18 case , chevyrev - lyonsresult that the expected signature of fbm has infinite radius of convergence .
this in turns implies that the expected signature determines the signature in distribution .
[ pr3 ] let @xmath39 .
letting @xmath40 be a word where @xmath41 for @xmath42 , then @xmath47}db^{i}\right)\leq \dfrac{1}{k!2^{k}}\ ] ] we move now to the cubature method .
we provide the cubature formula for the one dimensional fbm up to certain degrees .
moreover , we will discuss in more details the concept of _ degree _ of a cubature formula in chapter [ cub ] .
[ pr4 ] let @xmath48 .
define @xmath49 and @xmath50 to be @xmath51 for @xmath52 .
the weights @xmath53 and the paths @xmath54 will satisfy the cubature formula @xmath55=\sum_{j=1}^{3}\lambda_{j}\int_{0<t_{1}<, ... ,<t_{k } < t}d\hat{\omega}_{t_{1},j}^{i_{1}}\cdot\cdot\cdot d\hat{\omega}_{t_{1},j}^{i_{k}},\ ] ] for the following degree @xmath56 if @xmath57 and @xmath58 , and @xmath59,\\ ( \alpha-\beta)+(2\beta-\alpha)t , \qquad t\in[\frac{1}{3},\frac{2}{3}],\\ ( \beta-\alpha)+(2\alpha-\beta)t , \qquad t\in[\frac{2}{3},1],\\ \end{cases}\ ] ] where @xmath60 and @xmath61 and @xmath62 for @xmath63 $ ] .
it is important to remark that for @xmath64 , i.e. when the fractional brownian motion is just a brownian motion , the results obtained in this theorem perfectly coincide with the results of lyons and victoir obtained in @xcite .
one of the main concepts of this paper is the signature of the fractional brownian motion . in order to give a definition of signature we need to introduce first the following concept : the _ p - variation _ of a path .
[ p - var ] let @xmath65 be a real number .
let @xmath66 be continuous path , where j is a compact interval and e is a banach space .
the p - variation of x on the interval j is defined by @xmath67^{\frac{1}{p}}\ ] ] where the supremum is taken over all the partitions of j. a path @xmath68 is said to be of finite @xmath69-variation over the interval @xmath70 if @xmath71 . + for signature of a stochastic process we mean the following let e be a banach space .
let @xmath72\rightarrow e$ ] be a continuous path with finite p - variation for some @xmath73 .
the signature of x is the element @xmath74 of @xmath75 defined as follows @xmath76})=(1,x_{[0,t]}^{1},x_{[0,t]}^{2}, ... )\ ] ] where , for each @xmath77 , @xmath78}^{n}=\int_{0<u_{1}< ... <u_{n}<t}dx_{u_{1}}\otimes ... \otimes dx_{u_{n}}.\ ] ] the integration in the previous definition is defined in the sense of young @xcite .
+ the fractional brownian motion is a @xmath69-variation path for all @xmath79 .
hence , for @xmath0 $ ] we have @xmath73 .
notice that for @xmath80 , a word composed by the letters @xmath81 where each @xmath82 , we have that @xmath83 .
therefore , we have that the @xmath4-th element of the signature of the fbm and its dyadic approximation are respectively given by @xmath84}db^{i}\right)e_{i_{1}}\otimes\cdot\cdot\cdot \otimes e_{i_{2k}}\ ] ] and by @xmath85}db^{m , i}\right)e_{i_{1}}\otimes\cdot\cdot\cdot \otimes e_{i_{2k}}\ ] ] where @xmath86 is the basis of @xmath12 . + + we recall two results which can be found for example in baudoin and coutin paper @xmath19 .
first , we present a reformulation of the isserlis ( or wick s ) theorem : let @xmath87 be a centered gaussian vector .
we have @xmath88 where @xmath89 is the group of the permutations of the set @xmath90 . therefore from the previous lemma we have @xmath47}db^{m , i}\right)=\dfrac{1}{k!2^{k}}\sum_{\sigma\in\mathcal{g}_{2k}}\int_{\delta^{2k}[0,1]}\prod_{l=1}^{k}\mathbb{e}\left(\dfrac{db^{m , i_{\sigma(2l-1)}}}{dt_{\sigma(2l-1)}}\dfrac{db^{m , i_{\sigma(2l)}}}{dt_{\sigma(2l)}}\right)dt_{1}\cdot\cdot\cdot dt_{2k}\ ] ] notice that if @xmath91 $ ] and @xmath92 $ ] then @xmath93 further , recall theorem 31 of @xmath19 .
[ 31 ] assume @xmath39 . letting @xmath40
be a word , then @xmath47}db^{i}\right)=\dfrac{h^{k}(2h-1)^{k}}{k!2^{k}}\sum_{\sigma\in\mathcal{g}_{2k}}\int_{\delta^{2k}[0,1]}\prod_{l=1}^{k}\delta_{i_{\sigma(2l)},i_{\sigma(2l-1)}}|t_{\sigma(2l)}-t_{\sigma(2l-1)}|^{2h-2}dt_{1}\cdot\cdot\cdot dt_{2k}\ ] ] we now move toward the cubature method and we introduce two definitions of the cubature formula .
both are introduced in the work of lyons and victoir @xcite . + we denote by @xmath94 $ ] the space of polynomials in @xmath11 variables and of degree less than @xmath95 .
let @xmath96 be a positive measure on @xmath12 , @xmath97 and m be a natural number .
we say that the points @xmath98 in the support of @xmath96 , and the positive weights @xmath99 define a cubature formula of degree m with respect to @xmath96 if and only if , for all polynomials @xmath100 $ ] , @xmath101 when @xmath102 , one talks about quadrature formulae rather than cubature formulae .
+ now we move from a finite dimensional space to the wiener space , which is infinite dimensional .
[ cubature - definition ] let m be a natural number .
let define @xmath103 .
we say that the paths @xmath104 , \mathbb{r}^{d})\ ] ] and the positive weights @xmath105 define a cubature formula on wiener space of degree m at time t , if and only if , for all @xmath106 , @xmath55=\sum_{j=1}^{n}\lambda_{j}\int_{0<t_{1}<, ...
,<t_{k } < t}d\hat{\omega}_{j , t_{1}}^{i_{1}}\cdot\cdot\cdot d\hat{\omega}_{j , t_{1}}^{i_{k}}.\ ] ] notice that we wrote @xmath107 instead of @xmath108 for @xmath109 in order to stress the fact that it is not a brownian motion , since it is a brownian motion for @xmath110 but for @xmath111 we have @xmath112 .
+ in general , we say that the paths @xmath113 and the weights @xmath114 define a cubature formula on wiener space of degree @xmath95 if the expectations of the ( stratonovich)-iterated integrals of degree less than @xmath95 are , under the wiener measure and under the probability measure @xmath115 the same . in this article , we do not deal with the concatenation step of the cubature method since we focus mainly in obtaining the cubature formula for the fractional brownian motion .
indeed , while the concatenation for the brownian motion case takes advantage of the independence of the increments ( and so of the markov semigroup property ) , for the fractional brownian motion case these properties do not hold .
hence , the problem of the concatenation is an open problem and appears to be not at all trivial .
[ rem ] for the fractional brownian motion we have the same definition of cubature formula as for the brownian motion except for the definition of @xmath116 .
indeed , for the fbm we have @xmath117 as it is possible to see the two definitions coincide for @xmath3 .
the first result we show regards rate of convergence of the expected signature of the dyadic approximation of the fractional brownian motion ( fbm ) to its exact value .
the sharpness of the rate of convergence is a delicate matter and it will be discussed at the end of this chapter .
+ in this chapter we prove theorem @xmath118 . as mentioned before , this theorem covers the weak convergence rate of the signature of the fbm .
a previous result on the strong convergence rate was obtained by friz and riedel in @xmath119 .
+ + recall from theorem @xmath120 that we have @xmath43}db^{i}\right)-\mathbb{e}\left(\int_{\delta^{2k}[0,1]}db^{m , i}\right)\bigg|\ ] ] @xmath121}\prod_{l=1}^{k}\delta_{i_{\sigma(2l)},i_{\sigma(2l-1)}}|t_{\sigma(2l)}-t_{\sigma(2l-1)}|^{2h-2}\ ] ] @xmath122\times[t_{j},t_{j+1}]}(t_{\sigma(2l)},t_{\sigma(2l-1)})2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdydt_{1}\cdot\cdot\cdot dt_{2k}\bigg|\ ] ] in order to understand this as a whole we need to first understand the single parts of it . in other words
let us focus on @xmath123\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\ ] ] now we are going to split this double integral into different parts according to our dyadic approximation . in particular , let @xmath124 be the lowest dyadic time above @xmath125 and @xmath126 the highest dyadic time below @xmath127 , as shown in the figure 1 .
then we have that our double integral can be rewritten in the following way : @xmath128 ) . ]
to better understand this decomposition look at the upper triangle in figure 1 ( i.e. the triangle from the red diagonal to the up and left blue edges ) .
the first integral is the small triangle in the lower left . from here
you first move up and then move right .
the last integral is the central yellow triangle in figure 1 . for this integral the dyadic approximation and the exact value
are equal , hence their difference is zero .
this is because the dyadic approximation is equal to the exact value if the points in time considered coincide with the dyadic points .
+ + what we want to show now is that for the other integrals the difference between the dyadic approximation and the exact value is going to zero as the dyadic time step goes to zero ( i.e. @xmath129 ) .
+ + in order to facilitate the reading and understanding of the content of this chapter , instead of having a 7 page long proof we decided to split the proof of theorem [ pr1 ] in different small propositions .
+ the following propositions regards the rate of convergence of each double integral of the right hand side of equation @xmath130 . + +
so let us focus on the integrals : @xmath131 and @xmath132 .
we have the following proposition .
the differences @xmath133\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\bigg|\ ] ] and @xmath134\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\bigg|\ ] ] are bounded by @xmath135 @xmath136\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\ ] ] @xmath137 @xmath138 @xmath139 notice that @xmath140 @xmath141 the same reasoning done for this integral applies _ mutatis mutandis _ to the integral @xmath132 .
let us focus on the integrals : @xmath142 and @xmath143 .
we have the following proposition the differences @xmath144\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\bigg|\ ] ] and @xmath145\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\bigg|\ ] ] are bounded by @xmath146 for @xmath147 , and by @xmath148 for @xmath149 .
we have @xmath150\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\ ] ] @xmath151}(t)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{q-1}}^{t_{q}}|x - y|^{2h-2}dxdy\big)dsdt\ ] ] @xmath152 here we need to split the sum in two parts . the first part is composed by the first element of the sum , while the second is composed by the whole sum without the first element
let us first consider the first part .
@xmath153 @xmath154 @xmath155 notice that every element in the numerators is smaller or equal than @xmath1 except @xmath156 and @xmath157 for which the following relation holds @xmath158 .
therefore , taking absolute value of each element we have @xmath159 now let s focus on the second part , that is @xmath160 @xmath161 @xmath162 by applying mean value theorem we have @xmath163}(2 - 2h)\xi^{2h-3}|t - s -(x - y)|dydx\big)dsdt\ ] ] @xmath164 notice
that for @xmath149 then the equation above is just zero . thus , for the following we are in the case @xmath165 . notice also that @xmath166 and @xmath167 , hence @xmath168
thus , we have @xmath169 @xmath170 notice that @xmath171 @xmath172 where @xmath173 is a finite constant independent of @xmath95 .
notice that @xmath174 for @xmath175 . however , here we are only in the case @xmath175 as discussed above . + a similar reasoning applies to the integral @xmath143 .
the last integral to be analysed is @xmath176 .
the difference @xmath177\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\ ] ] is bounded by @xmath178 here we need to distinguish two cases .
one in which @xmath179 and the other in which @xmath180 .
+ let s focus first on the case @xmath180 .
@xmath181\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\ ] ] @xmath182 @xmath183 by mean value theorem we have @xmath184}(2 - 2h)\xi^{2h-3}|t - s -(x - y)|dxdy\big)dsdt\ ] ] @xmath185 since @xmath180 then @xmath186 .
@xmath187 @xmath188 finally consider the second case , i.e. @xmath179 .
@xmath189 @xmath190 @xmath191 now since @xmath179 then @xmath192 so @xmath193 . therefore ,
we have @xmath194 moreover , since @xmath195 then we can use the right hand side of the above inequality as the coefficient for the rate of convergence for the integral @xmath176 independently of the value of @xmath125 and @xmath127 .
therefore , we have the following proposition [ propositionintegrals ] the difference for the double integral @xmath196 @xmath197\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\ ] ] is bounded by @xmath198 where @xmath6 is given by @xmath199 for @xmath175 , and @xmath200 for @xmath149 .
it is immediate from the previous propositions , from the fact that for the double integral @xmath201 the difference is zero as discussed before and by considering equation @xmath130 .
indeed , the difference for the double integral @xmath201 is given by @xmath202\times[t_{j},t_{j+1}]}(t , s)2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\big)dsdt\ ] ] notice that this difference can be expressed as the sum of two classes of integrals .
the first class contains the off - diagonal terms , which can be expressed as @xmath203 where @xmath204 .
it is possible to see that this difference it is immediately zero .
+ the second case is @xmath205 where @xmath206 .
this difference is equal to @xmath207 @xmath208 @xmath209 this result could also be seen from the symmetric properties with respect to the diagonal of the function @xmath210 .
therefore , we can now do the proof of theorem @xmath118 .
@xmath43}db^{i}\right)-\mathbb{e}\left(\int_{\delta^{2k}[0,1]}db^{m , i}\right)\bigg|=\ ] ] @xmath121}\prod_{l=1}^{k}\delta_{i_{\sigma(2l)},i_{\sigma(2l-1)}}|t_{\sigma(2l)}-t_{\sigma(2l-1)}|^{2h-2}\ ] ] @xmath122\times[t_{j},t_{j+1}]}(t_{\sigma(2l)},t_{\sigma(2l-1)})2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdydt_{1}\cdot\cdot\cdot dt_{2k}\bigg|\ ] ] now by using the relation @xmath211 where @xmath212 for @xmath213 , and by using proposition @xmath214 , this absolute value is bounded by @xmath215 @xmath216}\prod_{l=2}^{k}\delta_{i_{\sigma(2l)},i_{\sigma(2l-1)}}\sum_{i , j=1}^{2^{m}}\textbf{1}_{[t_{i},t_{i+1}]\times[t_{j},t_{j+1}]}(t_{\sigma(2l)},t_{\sigma(2l-1)})2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\ ] ] @xmath217\times[t_{j},t_{j+1}]}(t_{\sigma(2l)},t_{\sigma(2l-1)})2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\ ] ] @xmath218 consider now the single addends in the formula above .
it is possible to see that each element is bounded by @xmath219 the reason why we have @xmath220 to @xmath7 in the product does not matter . indeed ,
once we take the integral to be over @xmath221^{k}$ ] instead of @xmath222 it is not important which permutation we consider .
we can move the integrands around using tonelli s theorem since they are positive value .
further , the above bound is the same as considering @xmath223 .
this integral reduces to @xmath224 therefore , we can bound @xmath225 by @xmath226 where we have used the fact that @xmath227 , which is the sum over all permutations of the set @xmath228 , corresponds to @xmath229 .
therefore we can rewrite the formula above as @xmath230 where @xmath231 this concludes the proof of the rate of convergence .
the sharpness of the bound is a delicate matter .
from one side we always used sharp estimates for each value we considered in the proof of the theorem .
further , it has been proved by @xmath232 that for the brownian motion the sharp rate of convergence is @xmath2 .
hence , we are confident that the rate of convergence obtained here is sharp .
however , in order to be completely sure about it we need to prove that @xmath46}db^{i}\right)-\mathbb{e}\left(\int_{\delta^{2k}[0,1]}db^{m , i}\right)\bigg|>0.\ ] ] this will prove sharpness of the rate of convergence .
however , the dyadic approximation makes this idea extremely difficult to put in practice .
this is because , when we focus on the double integral @xmath196 , the dyadic approximation and the exact value for certain times are equal , in particular when the times @xmath125 and @xmath127 correspond to the times of the mesh of the dyadic approximation @xmath124 and @xmath126 respectively .
hence as @xmath129 the difference between the exact value and the dyadic approximation goes to zero discontinuously ( or better continuously but with jumps / discontinuous point , i.e. stepwise ) .
this makes the problem of the sharpness of the rate of convergence to appear impossible .
a possible solution is to consider the iterated integral as a whole without focusing on the general double integral .
but this seems a difficult way of dealing with the problem .
indeed , we proceeded on with this way at the beginning of this work but the result was a bound with a rate of convergence of @xmath1 but with a non integrable coefficient . for these reasons
we are confident that the rate of convergence of @xmath1 is sharp . + a final remark regards the fact that @xmath233 , which means that the difference goes to zero faster for the fbm than the bm .
this is an expected result since the fbm with @xmath10 is a smoother path than the bm .
this higher smoothness is related to the positive correlation of the increments of the fbm with @xmath10 .
indeed , we can see that the rate of convergence becomes smaller ( hence the convergence faster ) as the positive correlation ( represented by @xmath16 ) increases .
as it is possible to see from equation @xmath234 the coefficient of the rate of convergence goes to infinity as the degree of the truncated signature @xmath7 goes to infinity . in other words , @xmath235 as @xmath236 . in the following
we show that the coefficient in fact goes to zero as @xmath7 goes to infinity .
+ in this chapter we prove the theorem [ pr2 ] .
first , multiply our main object of study @xmath43}db^{i}\right)-\mathbb{e}\left(\int_{\delta^{2k}[0,1]}db^{m , i}\right)\bigg|\ ] ] by @xmath237 .
we can use our estimates obtained in ( @xmath238 ) to get : @xmath239 @xmath216}\prod_{l=2}^{k}\delta_{i_{\sigma(2l)},i_{\sigma(2l-1)}}\sum_{i , j=1}^{2^{m}}\textbf{1}_{[t_{i},t_{i+1}]\times[t_{j},t_{j+1}]}(t_{\sigma(2l)},t_{\sigma(2l-1)})2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\ ] ] @xmath217\times[t_{j},t_{j+1}]}(t_{\sigma(2l)},t_{\sigma(2l-1)})2^{2m}\int_{t_{i}}^{t_{i+1}}\int_{t_{j}}^{t_{j+1}}|x - y|^{2h-2}dxdy\ ] ] @xmath240 we have not written it in absolute value because it is a positive value .
now , taking the limit as @xmath129 . by dominated convergence theorem
we obtain : @xmath241}\prod_{l=2}^{k}\delta_{i_{\sigma(2l)},i_{\sigma(2l-1)}}|t_{\sigma(2l)}-t_{\sigma(2l-1)}|^{2h-2}\ ] ] @xmath242 @xmath243 notice that @xmath244 @xmath245 @xmath246 moreover notice that the function @xmath247 defined as @xmath248 is symmetric with respect to the diagonal .
hence we have , @xmath249}f(t_{1}, ...
,t_{2k})dt_{1}\cdot\cdot\cdot dt_{2k}=\dfrac{1}{2k!}\int_{[0,1]^{2k}}f(t_{1}, ... ,t_{2k})dt_{1}\cdot\cdot\cdot dt_{2k}\ ] ] therefore we have that the formula @xmath250 is equal to @xmath251^{2k}}\prod_{l=2}^{k}\delta_{i_{\sigma(2l)},i_{\sigma(2l-1)}}|t_{\sigma(2l)}-t_{\sigma(2l-1)}|^{2h-2}dt_{1}\cdot\cdot\cdot dt_{2k}\ ] ] @xmath252 notice that @xmath253 thus , the coefficient of the rate of convergence goes to zero as the number of iterated integrals increases ( i.e. as the order of the signature increases ) and it goes there very fast .
in this chapter we shift the focus we had in the last two chapters .
our object is now just the expected signature of the fractional brownian motion ( fbm ) .
we prove a simple but precise estimate for it .
+ indeed , we prove theorem [ pr3 ] .
the proof of this theorem is very short since it relies on the computations done in the proof of theorem [ pr2 ] .
consider the ( truncated ) signature of the fractional brownian motion , we have @xmath47}db^{i}\right)=\dfrac{h^{k}(2h-1)^{k}}{k!2^{k}}\sum_{\sigma\in\mathcal{g}_{2k}}\int_{\delta^{2k}[0,1]}\prod_{l=1}^{k}\delta_{i_{\sigma(2l)},i_{\sigma(2l-1)}}|t_{\sigma(2l)}-t_{\sigma(2l-1)}|^{2h-2}dt_{1}\cdot\cdot\cdot dt_{2k}.\ ] ] @xmath254 by the proof of theorem [ pr2 ] . + further , we have @xmath255 in the next chapter we shift again our attention and we focus on the cubature method which uses the signature and its approximation in order to obtain numerical solutions of sdes driven by the fbm .
+ indeed , one of the main reasons why we focused on these problems and tried to obtain these results is to provide good rigorous estimates ( about the rate of convergence , the coefficient , the signature , etc .. ) to improve the theoretical capability of the numerical methods ( mainly the cubature method ) of obtaining good results in an efficient amount of time .
this is fundamental since the cubature method can be a cumbersome method in some cases .
hence , having good theoretical estimates tells us the level of performance we can achieve , at least theoretically .
in this chapter we obtain the cubature formula for the one dimensional fractional brownian motion up to degree 5 . + as discussed before ( see section [ preliminaries ] and in particular definition [ cubature - definition ] ) , in order to obtain the cubature formula what we need to do is to find @xmath256 paths and @xmath256 weights such that @xmath55=\sum_{j=1}^{n}\lambda_{j}\int_{0<t_{1}<, ... ,<t_{k } < t}d\hat{\omega}_{j , t_{1}}^{i_{1}}\cdot\cdot\cdot d\hat{\omega}_{j , t_{1}}^{i_{k}}.\ ] ] in our case , since we focus on the one dimensional case , @xmath257 for @xmath109 . for @xmath258 we mean that @xmath259 and @xmath260 are both the path @xmath261 ( i.e. @xmath262 ) . on the other hand , for @xmath263
we mean that @xmath264 is the fractional brownian motion at time @xmath261 and @xmath265 is the deterministic path ( that we need to find ) at time @xmath261 that we denote @xmath266 . + further , since we focus on degree up to 5 then @xmath267 . in the appendix
@xmath268 we provide a complete description of the iterated integrals we have to consider for this cubature formula , they are 17 in total .
further , again in the appendix @xmath268 we provide their values and the corresponding equations for the deterministic paths @xmath266 .
+ in this chapter we should prove theorem [ pr4 ] . however , due to the fact that the proof is mainly based on linear algebra computations , we decided to put the proof of this theorem in the appendix ( see appendix @xmath269 ) .
+ + now , we need to discuss some important remarks about this result .
given the definition of @xmath116 ( see remark [ rem ] ) we have that the cubature formula obtained in theorem [ pr4 ] is valid for the following degree @xmath270 the reason why we have different degrees for different values of @xmath16 is because we wanted to have the same cubature formula for different values of @xmath16 . in particular , we wanted a cubature formula which was the same obtained by lyons and victoir in @xcite for the brownian motion of degree 5 . in order to have it we needed to consider the same iterated integrals they considered .
+ in other words , we decided to fix the number of iterated integrals and so to fix the system of 17 equations we needed to solve in order to have a cubature formula that was exactly as the one of @xcite for @xmath3 .
this had as a counterpart the fact that for high value of @xmath16 the degree have to be decreased according to the definition of @xmath116 ( see remark [ rem ] ) . + further , in case we wanted to have a cubature formula for degree 5 ( independently of @xmath16 ) then we would had to compute more iterated integrals than the ones we computed .
in this way we would add more constraints ending up with a cubature formula different from the one of @xcite .
+ + given that the concept of _ degree _ for a general stochastic process is not usually known , we introduce now a more general concept which is related to the degree and can be adapted to any stochastic process : the concept of _
order_. let @xmath271 be a stochastic process .
we say that a set of words @xmath70 is of order @xmath272 with respect to the stochastic process w if for any @xmath273 , where @xmath274 is any smooth vector field , we have @xmath275 @xmath276 , for @xmath277 small , and @xmath278 smooth and bounded . where c is a constant .
further , we introduce the concept of _ cubature formula of order l_. a cubature formula of order l is obtained by matching @xmath279 @xmath280 , for @xmath70 a set of words of order l with respect to the process w. where @xmath281 are positive weights and @xmath282 are continuous and deterministic paths . using these definitions
we can state that the cubature formula that we have obtained in theorem [ pr4 ] is of order @xmath283 .
this is because the lowest order is obtained for @xmath3 , which is the brownian motion case for which we have an order of @xmath283 . indeed , for the fractional brownian motion we can make the following remark . for the fractional
brownian motion , and so also for the brownian motion , the relation between the order and the degree of a cubature formula is given by @xmath284 where h is the hurst parameter .
hence , precisely the order of the cubature formula of theorem [ pr4 ] is @xmath285 we conclude with a final remark on the solution(s ) obtained for the cubature formula . from the proof of this theorem ( see appendix 2 ) we obtain two solutions of our system of equations .
in particular , we have only used one solution of the equation @xmath286 the other solution is given by @xmath287 the reason why we focused on the other solution and not this one is because lyons and victoir focused on that solution in their paper .
they used mathematica to produce a solution , without explicitly justify their decision .
however , both solutions are feasible for our system of equations @xmath288 . hence , we have two valid solutions for the cubature formula of the fractional brownian motion for this kind of structure ( i.e. piecewise linear paths with change of slopes at @xmath289 and @xmath290 ) .
+ the reason why they did not justify their decision is probably due to the fact that the structure adopted is already arbitrary and hence it is important to have a solution and not to have a particular or unique solution .
the author would like to thank dr .
h. boedihardjo ( university of reading ) and prof .
d. crisan ( imperial college london ) for numerous and useful comments , suggestions and discussions during the whole production of this work .
further , the author would like to thank dr .
t. kuna ( university of reading ) for useful comments regarding problem 2 and problem 3 of this work and dr .
j. broecker ( university of reading ) for different and useful discussions .
finally , the author would like to thank the centre for doctoral training in mathematics of planet earth of the university of reading and of the imperial college london for providing funding for this research .
in this appendix we study the iterated integrals with respect to the path @xmath291 .
notice that we will use the notation @xmath292 .
we will not consider the case of iterated integrals of time solely since they bring no information for the construction of the cubature .
further , we will use many times the following formula for the fractional brownian motion : @xmath293 in particular , we have : + degree@xmath294 : @xmath295 + degree@xmath296 : @xmath297 + degree@xmath298 : @xmath299 and @xmath300 + degree@xmath301 : @xmath302 + degree@xmath303 : @xmath304 and @xmath305 @xmath306 and @xmath307 @xmath308 @xmath309 @xmath310 + degree@xmath311 : @xmath312 + degree@xmath313 : @xmath314 and @xmath315 and @xmath316 degree@xmath317 : @xmath318 and @xmath319 @xmath320 and @xmath321 @xmath322 @xmath323 and @xmath324 @xmath325 @xmath326 @xmath327 degree@xmath328 : @xmath329 + + now we need to match them with the corresponding deterministic iterated integrals . in other words , we have + + for degree@xmath294 : @xmath330 + for degree@xmath296 : @xmath331 + for degree@xmath298 : @xmath332 and @xmath333 for degree@xmath301 : @xmath334 + for degree@xmath303 : @xmath335 and @xmath336 and @xmath337 for degree@xmath311 : @xmath338 + for degree@xmath313 : @xmath339 and @xmath340 and @xmath341 for degree@xmath317 : @xmath342 and @xmath343 @xmath344 and @xmath345 and @xmath346 for degree@xmath328 : @xmath347
in this appendix we present the proof of * theorem 2.5*. let us now start to investigate the form of the functions @xmath348 for @xmath349 .
+ from the appendix it is possible to see that we have 17 equations , hence we need to have 17 unknowns . + following the work done by lyons and victoir we are going to choose two symmetric paths and one path which has constant value zero .
this reduces the number of equations to 5 .
hence assume that there are two continuous functions , @xmath350 and @xmath351 , with the property that @xmath352 for @xmath353 $ ] .
further assume that there is a third path @xmath354 for @xmath353 $ ] . with this formulation
only 5 equations need to be taken into consideration since the other 12 are already satisfied .
the 5 equations are the following .
first , @xmath355 second , @xmath356 third , @xmath357 fourth , @xmath358 fifth , @xmath359 the other 12 equations are automatically zero by the symmetric properties of the @xmath350 and @xmath360 , and by the fact that @xmath354 , since the 12 equations involve odd integrals of the @xmath361s .
+ now we need to have maximum 5 unknowns in order to solve the system of equations .
we have actually have 6 equations since the sum of the weights @xmath362 , because we are considering a probability measure .
assume that @xmath363,\\ b_{1}s+b_{0}\qquad for \qquad s\in[\frac{1}{3},\frac{2}{3}],\\ c_{1}s+c_{0}\qquad for \qquad s\in[\frac{2}{3},1].\\ \end{cases}\ ] ] with this formulation we have 5 unknowns which are @xmath364 , @xmath365 , @xmath366 , @xmath367 and @xmath368 .
the reason why @xmath369 and @xmath370 are not unknowns is because they have to take certain values in order to make the path @xmath350 continuous .
notice that @xmath371 is not an unknowns since @xmath372 .
we have 5 unknowns for 6 equations there is a risk that the system can not be solved .
however , we hope that two of the 5 equations are the same .
an alternative approach is to let the points where the slope changes , which we fixed to be at @xmath373 and @xmath374 , be two unknowns .
+ let us now solve the system . consider equation @xmath375 , we have @xmath376 consider equation @xmath377 , we have @xmath378 @xmath379 @xmath380 @xmath381 consider now equation @xmath382 , we have @xmath383 by fubini s theorem we have @xmath384 @xmath385 @xmath386 @xmath387\ ] ] @xmath388 @xmath389\ ] ] @xmath390 @xmath391 @xmath392 @xmath393 consider now equation @xmath394 , we have @xmath395 by fubini s theorem we have @xmath396 @xmath397\ ] ] @xmath398 @xmath399\ ] ] @xmath400 @xmath401 @xmath402 @xmath403 consider now equation @xmath404 , we have @xmath405 therefore we have the following system of equations @xmath406,\\ 2\lambda_{1}(c_{1}+c_{0})^{2}=1,\\ \dfrac{a^{2}}{81}+\dfrac{1}{3}\left(\dfrac{7}{27}b_{1}^{2}+b_{1}b_{0}+b_{0}^{2}\right)+\dfrac{1}{3}\left(\dfrac{19}{27}c_{1}^{2}+\dfrac{5}{3}c_{1}c_{0}+c_{0}^{2}\right)=\dfrac{1}{2\lambda_{1}(2h+1)},\\ \dfrac{a^{2}}{81}-2(c_{1}+c_{0})\left(a\dfrac{1}{18}+b_{1}\dfrac{1}{6}+b_{0}\dfrac{1}{3}\right)+\dfrac{1}{3}\left(\dfrac{7}{27}b_{1}^{2}+b_{1}b_{0}+b_{0}^{2}\right)+\dfrac{55c_{1}^{2}}{81}+\dfrac{2c_{0}^{2}}{3}+\dfrac{4c_{1}c_{0}}{3}=\dfrac{1}{2\lambda_{1}(2h+1)},\\ ( c_{1}+c_{0})\left(a\dfrac{1}{18}+b_{1}\dfrac{1}{6}+b_{0}\dfrac{1}{3}\right)-\dfrac{a^{2}}{81}-\dfrac{1}{3}\left(\dfrac{7}{27}b_{1}^{2}+b_{1}b_{0}+b_{0}^{2}\right)+\dfrac{7c_{1}^{2}}{162}+\dfrac{c_{1}c_{0}}{18}=\dfrac{2h-1}{4\lambda_{1}(2h+1)},\\ \lambda_{1}(c_{1}+c_{0})^{4}=\dfrac{3}{2 } , \end{cases}\ ] ] with the following unknowns @xmath407 .
+ + consider the two equations in the system above @xmath408 we have @xmath409 and @xmath410 further , using @xmath411 we have @xmath412 now consider the equations @xmath413 and @xmath414 by summing them , we have @xmath415 @xmath416 further , by taking the difference of the two equations @xmath417 and @xmath413 we obtain @xmath418 @xmath419 now using equations @xmath420 and @xmath421 we have @xmath422 @xmath423 we already knew this information from equation @xmath424 , but we did not use it here .
this means that two of our constraints are in fact only one .
hence we have one constraint less than expected , meaning that we are able to find a solution of our system since now the number of equations is the same as the number of unknowns .
+ recall that there are two underlying equations , which comes from the continuity condition of @xmath361 , which are @xmath425 let us continue to focus on the equation @xmath420 .
substitute @xmath426 , @xmath427 and @xmath428 .
in other words , take @xmath429 and @xmath430 hence we have that equation @xmath420 becomes @xmath431 @xmath432 @xmath433 @xmath434 we can now consider the third equation of our system ( i.e. equation @xmath435 ) and rewrite it in terms of @xmath366 .
first , let us rewrite all the unknowns that we need in terms of @xmath366 : @xmath436 and @xmath437 notice that @xmath438 and @xmath439 , which is in accordance with the results of lyons and victoir .
+ we can now proceed with the substitution @xmath440 @xmath441 @xmath442 @xmath443 @xmath444 @xmath445 if @xmath3 then @xmath446 which is in accordance with the results of lyons and victoir . + therefore
, we know all the unknowns . in particular , the path @xmath447 is given by : @xmath448,\\ \dfrac{2h\sqrt{3}+\sqrt{3}-\sqrt{-96h^{2}+66h+57}}{2h+1}+\dfrac{2\sqrt{-96h^{2}+66h+57}-2h\sqrt{3}-\sqrt{3}}{2h+1}t , \qquad t\in[\frac{1}{3},\frac{2}{3}],\\ \dfrac{\sqrt{-96h^{2}+66h+57}-2h\sqrt{3}-\sqrt{3}}{2h+1}+\dfrac{4h\sqrt{3}+2\sqrt{3}-\sqrt{-96h^{2}+66h+57}}{2h+1}t ,
\qquad t\in[\frac{2}{3},1],\\ \end{cases}\ ] ] if you let @xmath449 and @xmath450 to be : @xmath451 then we can rewrite our path @xmath447 can be written in the following form : @xmath452,\\ ( \alpha-\beta)+(2\beta-\alpha)t , \qquad t\in[\frac{1}{3},\frac{2}{3}],\\ ( \beta-\alpha)+(2\alpha-\beta)t , \qquad t\in[\frac{2}{3},1],\\ \end{cases}\ ] ] now we proceed with a check of our result .
indeed we substitute the values obtained for our unknowns in our system of equations to check the correctness of these values .
+ in particular , let us focus first on the equation @xmath413 let us rewrite it in terms of @xmath366 : @xmath453 @xmath454 @xmath455 @xmath456 as before .
now let us focus on the equation @xmath457 following the same procedure we have @xmath458 @xmath459 @xmath460 @xmath461 as before .
the other equations are straightforward to check .
+ therefore , our solution is consistent and , as we have seen before , when @xmath3 our solution is the same solution obtained by lyons and victoir in their paper .
t. lyons and n. victoir .
cubature on wiener space .
ser . a math .
_ 460 , no . 2041 , ( 2004 ) , 169198 .
stochastic analysis with applications to mathematical finance . | in this work we present four different results concerning the signature and the cubature of fractional brownian motion ( fbm ) .
the first result regards the rate of convergence of the dyadic approximation of the expected signature of the fbm to its exact value , for a value of the hurst parameter @xmath0 $ ] .
we show that the rate of convergence is given by @xmath1 , where @xmath2 is the size of a single time step of the dyadic approximation .
we believe that this rate is sharp as it is consistent with the result of ni and xu @xcite , who showed that the sharp rate of convergence for the brownian motion ( i.e. fbm with @xmath3 ) is given by @xmath2 .
the second result regards the bound of the _ coefficient _ of the rate of convergence obtained in the first result . for this result
we also analyse the behaviour of this bound as the number of iterated integrals of the signature of fbm goes to infinity , showing that the bound of the @xmath4-th term in the signature is uniformly bounded by @xmath5 , where @xmath6 is a constant and @xmath7 is the number of iterated integrals .
the third result regards the bound of the expected signature of the fbm .
we show that the bound for the @xmath4-th term of the signature is simply given by @xmath8 .
this is a sharper estimate than the one obtained by chevyrev and lyons in @xcite , who showed that the expected signature has infinite radius of convergence , but not a factorial decay , and than the one by friz and riedel @xcite , who proved a factorial decay of @xmath9 .
the last result concerns the cubature formula for the one dimensional fbm for @xmath10 up to degree 5 .
this result extends the work of lyons and victoir @xcite who focused on the multidimensional brownian motion cubature formula .
+ + * key words : * fractional brownian motion , signature , rate of convergence , cubature method . |
scattering , absorption , and reradiation of photons by dust grains affect the propagation of starlight and play a key role in regulating the energy balance in the interstellar medium .
studies of these processes have provided a great deal of insight into the physical and chemical properties of dust grains ( for reviews of dust properties , see mathis 1990a ; witt 2003 ) .
it is also important to disentangle the effects of intervening dust from many kinds of astronomical observations , ranging from photometry and spectroscopy of galactic stars and external galaxies to mapping small - scale anisotropy in the cosmic microwave background radiation ( cmbr ) .
the degree of absorption and scattering by dust depends , in general , on the wavelength of the incident radiation ( whitford 1958 ) .
this dependence can be quantified in terms of the normalized dust extinction law @xmath20 ( or reddening law ) , where extinction is the sum of absorption plus scattering .
the ratio of total to selective extinction , @xmath21 , is a commonly - used measure of the slope of the extinction / reddening law . using data along many different galactic sight - lines , cardelli , clayton , & mathis ( 1989 , hereafter ccm ) found a tight correlation between the overall shape of the reddening law and the slope @xmath0 .
ccm devised an @xmath0-based , one - parameter family of empirical fitting functions to characterize the observed range of extinction law shapes ; this parameterization was later refined by fitzpatrick ( 1999 ) .
galactic dust displays a wide range of behavior . while diffuse interstellar dust is observed to have @xmath22 _ on average _
( savage & mathis 1979 ; ccm ) , this value is by no means universal
. significant deviations from this canonical @xmath22 value ( and corresponding differences in the overall shape of the extinction / reddening law ) are known to exist for a variety of interstellar dust clouds , and these variations appear to be generally correlated with environment ( fitzpatrick 1999 ) .
for example , @xmath0 has been found to range from 4 to 5 in dense molecular clouds ( mathis 1990b ; larson , whittet , & hough 1996 ; whittet et al .
2001 ) , whereas there are indications that more diffuse , high - latitude cirrus clouds may have @xmath0 values as small as @xmath23 ( fitzpatrick & massa 1990 ; larson et al . 1996
; szomoru & guhathakurta 1999 ) .
studies based on stellar photometry from the optical gravitational lens experiment ( ogle ) and massive compact halo objects ( macho ) projects point to @xmath0 being substantially smaller than 3 in the direction of the galactic bulge ( popowski 2000 ; popowski , cook , & becker 2003 ; udalski 2003 ; sumi 2004 ; popowski 2004 ) .
barbaro et al .
( 2001 ) find departures from the ccm parameterization in the ultraviolet portion of the galactic extinction law .
it is worth noting that the above studies sample dust at a wide range of distances from the sun : some probe the diffuse interstellar medium in its immediate vicinity whereas others , especially those at low galactic latitudes , can probe dust at much larger distances well outside the solar neighborhood ( e.g. , the rcra cloud studied by szomoru & guhathakurta 1999 ) .
some studies have suggested that @xmath0 variations are tied to variations in the size distribution of dust grains from one line of sight to another . in diffuse
cirrus clouds , a relative abundance of small grains might explain the steep rise of the extinction curve into the ultraviolet ( ccm ; larson et al.1996 ) .
by contrast , a high abundance of large grains , which grow readily by coagulation in dense molecular clouds , may explain the larger than average values of @xmath0 observed in these regions ( whittet et al .
whittet et al
. suggest that , in dense molecular clouds , @xmath0 remains close to the standard value of 3.1 except for lines of sight with unusually high extinction ( @xmath24 ) .
this may indicate a more complicated relationship between @xmath0 and @xmath25 .
moreover , it has been argued that chemical composition can also play a role in determining the shape of the extinction law ( rhoads , malhotra , & kochanski 2004 ) . to improve our understanding of the dependence of @xmath0 on environment ,
measurements are necessary along many sight - lines through a broad range of cloud types .
clusters with differential extinction / reddening have long been used to measure the @xmath0 of intervening dust ( mihalas & routly 1968 ) .
the traditional method of spectral typing individual cluster stars ( e.g. , on the morgan - keenan system ) , in conjunction with photometric measurements , allows for a direct and precise measurement of extinction and reddening .
however , this method of measuring @xmath0 requires high - quality , flux - calibrated spectra and a good understanding of the effects of metallicity on the energy output of stars .
our study focuses instead on the use of broad - band photometry in light of the fact that an extensive suite of photometric data sets is currently available . in this paper , we propose three methods for measuring the @xmath0 of dust in the foreground of differentially - reddened globular clusters .
the first two methods are closely related to each other and rely on accurate three - band optical photometry of cluster stars .
the third relies on two - band optical photometry and a map of the dust thermal emission ( schlegel , finkbeiner , & davis 1998 , hereafter sfd ) and is particularly well suited to wide - field data .
all three methods are based on the premise that the width of the blue horizontal branch ( hb ) is minimized when the appropriate @xmath0 value is used to correct the photometry for extinction and reddening .
the methods are applied to ngc 4833 , a low - latitude galactic globular cluster with variable extinction across its face .
realistic monte carlo simulations of the cluster data set are used to estimate the accuracy with which @xmath0 can be recovered .
nearly two decades ago , low et al .
( 1984 ) noticed diffuse background emission in the infrared astronomical satellite ( iras ) 60 and 100@xmath26 m maps and termed it ` infrared cirrus ' due to its complex texture .
the cirrus has been associated with thermal emission from dust grains in the diffuse interstellar medium ( beichman 1987 ) . combining the angular resolution of the iras observations with the photometric accuracy of the cosmic background explorer ( cobe)/diffuse infrared background experiment ( dirbe ) data , sfd derived all - sky maps of the dust column density and mean temperature .
one of the primary uses of these maps has been to correct extragalactic sources for extinction and reddening .
the sfd maps show structure in the cirrus down to the smallest angular scales resolved by iras ( @xmath27 ) .
gautier et al .
( 1992 ) described the structure in the iras data in terms of its angular power spectrum , @xmath28 where @xmath29 is the fourier power , @xmath30 is the reciprocal of the angular scale , and @xmath31 is the spectral index . similarly , guhathakurta & cutri ( 1994 ) examined both iras 100@xmath26 m maps and optical ccd surface photometry of reprocessed starlight from dust grains : they found that power on small scales was dominated by stars and galaxies , and since it was impossible to completely separate out these components , they merely cleared out the obvious compact `` objects '' ( stars and galaxies ) and smoothed the residual image over arcminute scales ; the resulting power spectrum had an index @xmath32 .
more recently , kiss et al . ( 2003 ) observed 13 fields with infrared space observatory ( iso)/isophot and found that the power spectrum index varies from field to field over the range @xmath33 for angular scales larger than @xmath34 .
this work extends the finding of @xmath35 by herbstmeier et al .
( 1998 ) in their earlier isophot - based study .
these last two studies indicate that @xmath36 is not universal . in this study
, the scatter in the hb of the differentially - reddened globular cluster ngc 4833 is used to measure the angular power spectrum of dust on scales smaller than the angular resolution of the thermal emission maps ( e.g. , iras , dirbe , isophot ) . for uncrowded data such as ours ,
the smallest angular scale down to which this method can be applied is set by the surface density of hb tracers which in turn determines the typical nearest - neighbor separation ; in crowding - limited situations , the smallest angular scale is a few times the size of the stellar point spread function : few arcseconds for ground - based images and sub - arcsecond for _ hubble space telescope _
images .
because interstellar dust causes extinction and reddening of starlight and reprocesses the energy , dust patchiness has a profound impact on the accuracy of many astronomical measurements even at high galactic latitudes . thermal emission from cold dust
can confuse cmbr anisotropy measurements .
measurements of large - scale structure in galaxy surveys may be influenced by spatial non - uniformities in the foreground dust .
studies of stellar populations are at risk because measurements of @xmath37 can be impacted by reddening .
extinction can affect the photometry of distance indicators and thereby produce systematic errors in the distance - scale ladder .
it is therefore important to quantify extinction / reddening variations on the smallest scales possible . while the sfd study does an excellent job of characterizing moderate- to large - scale dust variations , their maps miss power on scales smaller than a few arcminutes .
+ + the photometry of ngc 4833 is presented in 2 .
the construction of simulated cluster data sets in discussed in 3 .
the @xmath0 measurement methods are outlined in 4 , along with their application to ngc 4833 and simulated data sets to estimate the accuracy of each method . in 5 , we describe the angular power spectrum of cirrus in the direction of ngc 4833 .
possible future extensions of this work are discussed in 6 .
the main conclusions of this paper are summarized in 7 .
the globular cluster ngc 4833 is located near the equatorial plane of our galaxy ( @xmath38 ) in a dusty region of the constellation musca . the cluster has a mean reddening of @xmath39 mag and reddening variations of order @xmath40 mag across the field ( melbourne et al .
figure 1 of melbourne et al
. shows an image of the dusty features in the region around the cluster including the coalsack nebula .
the johnson - cousins @xmath1 stellar photometry of ngc 4833 used in this study are drawn from melbourne et al .
a detailed description of the data set is given in that paper so only a brief discussion appears here .
the ccd images of the cluster were obtained in 1996 using the 0.9-m telescope at the cerro tololo inter - american observatory .
the data reduction was carried out with the daophot ii ( stetson 1994 ) stellar photometry package .
figure [ fig : fullcmd ] shows the ( @xmath4 , @xmath5 ) and ( @xmath6 , @xmath5 ) cmds ( left and right panels , respectively ) for the full @xmath41 field .
boxes are drawn in the former cmd around the 260 objects that are considered to be candidate cluster hb stars for the purposes of this study ( possible non - hb interlopers in the sample are discussed in 4.1.1 ) .
extinction / reddening variations across the field affect the width of the principal sequences of the cmds .
figure [ fig : radcmd ] shows that the hb is relatively tight and narrow for the central @xmath42 region of the cluster , but increases in width for stars distributed over a larger area in the outer region of the cluster .
this suggests that extinction / reddening variations across the field , rather than photometric errors , are the cause of the large scatter in the hb ; if anything , photometric errors tend to be slightly worse in the inner region of the cluster because of crowding .
a color - color diagram of cluster hb stars is plotted in the top panel of figure [ fig : cc ] , while the bottom panel shows the same after correcting for reddening on a star - by - star basis .
the @xmath43 value for each star is taken from the sfd dust map , and an assumed @xmath0 value of 3.0 is plugged into the ccm parameterization of the dust extinction law ( eqn .
[ eqn : cardel ] ) to derive @xmath44 and the corresponding @xmath45 value .
the reddening - corrected data ( and the uncorrected data , for that matter ) show a linear relationship between @xmath4 and @xmath6 . a least - squares fit ( solid line )
is made to the reddening - corrected data : @xmath46 this equation , in conjunction with the two - polynomial fiducial fit to the hb in the ( @xmath4 , @xmath5 ) cmd , is used to simulate @xmath1 photometry of hb stars ( 3.2 ) .
figure [ fig : photerr ] shows the typical @xmath47 photometric errors in the @xmath1 bands for ngc 4833 stars as a function of their apparent visual magnitude ( melbourne et al .
the photometric errors , estimated using the daophot ii software package ( stetson 1994 ) , range from @xmath48 mag at the bright end of the hb ( @xmath49 ) to @xmath50 mag at the faint end ( @xmath51 ) , and contribute noticeably to the width of the hb .
the characteristic photometric error is defined to be that at @xmath52 , @xmath53 mag , and this is what is referred to throughout the rest of this paper .
a key aspect of our study is the use of simulated cluster data sets to fine tune and test the @xmath0 measurement methods ( 4 ) .
these simulated data sets are designed to mimic the general properties of the ngc 4833 data set as closely as possible .
since we are also interested in applying these @xmath0 measurement methods to a broad range of data , the main set of simulations are constructed over a two - dimensional grid of parameter space : three _ input _ @xmath0 values and three levels of photometric precision . at each of these 9 ( @xmath54 ) grid points ,
100 monte - carlo realizations are carried out
, a total of 900 simulated clusters are constructed each containing 260 hb stars .
the rest of this section describes the step - by - step procedure used to construct each simulated data set .
three input @xmath0 values , 2.0 , 3.0 and 4.0 , are chosen for the simulated cluster data sets .
these mimic a range of extinction law properties for the foreground dust and span the range of values observed in lines of sight with moderate to light reddening ( 1.1 ) .
the ccm parameterization of the galactic extinction laws : @xmath55 is adopted here . for a given choice of @xmath0 , the visual extinction @xmath25
is translated to its corresponding @xmath43 and @xmath45 color excesses .
each simulated star is assigned @xmath56- and @xmath5-band photometry via a random drawing from a fiducial hb curve in the cmd .
the fiducial is based on stitching together two second - order polynomials that are fit to the extinction / reddening - corrected ( @xmath4 , @xmath5 ) hb of ngc 4833 , where the @xmath43 correction for each star is obtained from the sfd dust map and @xmath0 is assumed to be 3.0 .
examples of such two - polynomial fiducial hbs are shown in figures [ fig : av_method ] and [ fig : schlegel_method ] .
the `` sfd - corrected '' cluster hb used for the fiducial fit is similar to the one shown in the left - middle panel of figure [ fig : cmd6 ] , except for the fact that the one illustrated has had the mean extinction / reddening reapplied to it .
when drawing 260 stars from the fiducial , the run of stellar density along the simulated hb is made to mimic that observed along the cluster hb .
this is done by dividing the fiducial into sections and demanding that the number of stars in each section match that in the corresponding section of the cluster hb .
simulated @xmath57-band photometry is obtained from the linear fit to the ( @xmath4 , @xmath6 ) color - color diagram of cluster hb stars ( fig .
[ fig : cc ] , eqn .
[ eqn : cc ] ) . the width of the hb can be used to quantify the effect of extinction / reddening variations .
figure [ fig : hbscat ] shows the rms scatter of hb stars in ngc 4833 as a function of their @xmath5 magnitude ( solid line ) .
the rms scatter is calculated based on the _ shortest distance _ of each hb star from the fiducial curve ( i.e. , distance to the nearest tangent point on the curve ) .
the fiducial consists of two second - order polynomials fit to hb stars within @xmath58 of the cluster center .
the dotted line in figure [ fig : hbscat ] is a model prediction for the hb scatter that would result from photometric errors alone , the dashed line shows the effect of large - scale extinction / reddening variations from sfd plus photometric error , and the dot - dashed line the effect of small- and large - scale extinction / reddening variations plus photometric error .
it is clear from the scatter at the bright end of the hb ( @xmath59 ) that only the dot - dashed line matches the ngc 4833 data .
the following subsections describe in turn the simulations behind the three model lines . a special set of 100 photometric - error - only ( `` @xmath60-only '' ) simulations containing _ no _ reddening variations is constructed for illustration purposes only ; this set is not used in any of the @xmath0 determination exercises described in 4 .
each monte carlo realization in this special set involves three steps : ( * 1 * ) an artificial hb is constructed as described in 3.2 above ; ( * 2 * ) the mean extinction and reddening of the cluster , @xmath61 and @xmath39 ( based on @xmath62 ) , are applied to all simulated stars to bring them into the same range of apparent @xmath1 magnitudes as the ngc 4833 hb stars ; and ( * 3 * ) gaussian photometric errors are added assuming @xmath63 .
the details of step ( 3 ) are given in 3.4 below .
the dotted line in figure [ fig : hbscat ] is the scatter measured in @xmath64 simulated hb stars .
the mismatch with the ngc 4833 data ( solid line ) demonstrates that photometric errors alone can not explain the width of the cluster s hb and there must be additional sources of scatter in the hb .
( the dotted line is much smoother than the solid line because it is based on 100@xmath65 as many stars . )
the obvious additional source of hb scatter is differential extinction / reddening across the cluster .
large - scale variations based on the sfd map are explored in this section .
their maps provide @xmath43 color excesses as a function of sky position , but have an angular resolution of @xmath66 and therefore contain power only on scales larger than this .
the sfd map includes ngc 4833 and the region around it , and indicates that the mean reddening along this line of sight is @xmath67 mag with variations on the order of @xmath40 mag .
following the procedure described above for the `` @xmath60-only '' simulations ( 3.3.1 ) , another special set of 100 simulations is constructed containing only large - scale extinction / reddening variations and photometric errors .
again , these special `` sfd+@xmath60 '' simulations are created for illustrative purposes alone .
the only difference is in step ( 2 ) of the procedure : instead of applying the same ( mean ) extinction and reddening values to all simulated hb stars , each star is randomly assigned the sky position of one of the 260 hb stars in ngc 4833 and the corresponding sfd @xmath43 value , and associated @xmath25 and @xmath45 values based on @xmath62 , are applied to the @xmath1 magnitudes of the simulated star . by assigning actual sky positions to the simulated hb stars ,
we ensure that they mimic the spatial distribution of ngc 4833 s hb stars and range of large - scale extinction / reddening variations across the cluster data set .
the dashed line in figure [ fig : hbscat ] , based on @xmath64 simulated hb stars , shows the scatter resulting from a realistic amount of large - scale extinction / reddening variations and photometric error .
even this line falls a little short of explaining the hb scatter observed in the cluster ( solid line ) .
the most probable cause of this shortfall is the coarse resolution of the sfd map which averages over structure in the dust distribution on angular scales smaller than @xmath66 .
small - scale variations are incorporated into the simulated data sets by adding random gaussian fluctuations with an amplitude proportional to the sfd reddening value at that location : @xmath68\ ] ] where @xmath69 is a random number ( for the @xmath9-th star ) drawn from a gaussian distribution with zero mean and width @xmath70 .
this brings us back to the main set of simulations used for testing the @xmath0 determination methods in 4 .
they are constructed using a procedure similar to that used for the special set of `` sfd+@xmath60 '' simulations ( 3.3.2 ) but with two changes .
first , the sfd reddening values in step ( 2 ) of the procedure are replaced by @xmath71 from eqn .
[ eqn : smallscale ] above .
second , a range of @xmath0 values ( 3.1 ) and photometric error levels ( 3.4 ) are used . with the above choice of gaussian width for the small - scale dust variations , @xmath70 , and for @xmath63 and @xmath62 , the scatter in the simulated hb approaches that of ngc 4833 s hb ( dot - dashed vs. solid lines in fig .
[ fig : hbscat ] ) .
this is a reasonable estimate of the amount of smoothing of dust fine - scale structure in the sfd maps ( 5.1 ; fig .
[ fig : powspec ] ) , and is well within the 10% uncertainty in reddening values estimated by the authors .
the final step in the construction of the main set of simulated hb data sets is the inclusion of photometric error .
since the photometric uncertainty is a function of stellar brightness , this step must be carried out after the inclusion of ( large- and small - scale ) dust effects so that each star is at the appropriate apparent brightness .
gaussian random noise is added to the @xmath1 magnitudes of each star .
three levels of photometric error are simulated : ( 1 ) @xmath63 , with the latter being the daophot - based errors shown in figure [ fig : photerr ] ; ( 2 ) @xmath72 ; and ( 3 ) @xmath73 .
we employ three methods for measuring @xmath0 in the direction of differentially - reddened globular clusters .
each method relies on the premise that the width of the blue hb will be minimized when the appropriate @xmath0 value is used to correct for extinction and reddening .
a description of each method and its application to the ngc 4833 and simulated cluster data sets follows .
the `` @xmath25 rms '' method relies on three - color photometry of differentially - reddened globular clusters .
figure [ fig : av_method ] shows the ( @xmath4 , @xmath5 ) and ( @xmath6 , @xmath5 ) cmds of a simulated hb ( left and right panels , respectively ) .
a fiducial , consisting of two second - order polynomials stitched together ( solid curve ; 3.2 ) , is fit to the hb stars within @xmath58 of the cluster center where differential extinction / reddening is relatively small ; this is done independently for the ( @xmath4 , @xmath5 ) and ( @xmath6 , @xmath5 ) cmds .
the relative visual extinction of each star is calculated to be the vertical distance between the star and the point where the reddening vector drawn from the star intersects the fiducial ( fig .
[ fig : av_method ] inset ) ; it is designated @xmath74 and @xmath75 for the ( @xmath4 , @xmath5 ) and ( @xmath6 , @xmath5 ) cmds , respectively . ''
is used to denote the relative visual extinction of a star based on a single fiducial / cmd , whereas `` @xmath76 '' is used to denote the _ difference _ between two @xmath3 values , say between two cmds for a given star or between a pair of stars . ] the @xmath74 measurement is based on an assumed / test value of @xmath0 , the slope of the reddening vector in the ( @xmath4 , @xmath5 ) cmd . for the @xmath75 calculation ,
the ccm parameterization given in eqn .
[ eqn : cardel ] is used to translate @xmath0 to @xmath44 , the slope in the ( @xmath6 , @xmath5 ) cmd .
the difference between the two @xmath3 values is defined to be @xmath77 . for a given test value of @xmath0 ,
the rms scatter of @xmath78 is computed over all hb stars .
this is done in two iterations , with 3-@xmath79 rejection of outliers after the first iteration to eliminate any non - hb stars in the sample ( e.g. , cluster blue stragglers or galactic field main sequence turnoff stars in the foreground encircled symbols in upper panels of fig . [
fig : radcmd ] ) .
non - hb interlopers are expected to have large deviations from the hb fiducial with effective @xmath3 values that are different between the two cmds .
the calculation of the rms of @xmath78 is repeated for a series of test @xmath0 values between 0.5 and 5.5 in steps of 0.25 .
the @xmath3 parameters from the two cmds are expected to approach each other when the test value approaches the true @xmath0 of the dust in the foreground of the cluster . in other words ,
the rms of @xmath78 should go through a minimum at the correct @xmath0 value .
a polynomial fit is made to the rms of @xmath78 vs.test @xmath0 data points to interpolate to the minimum .
figure [ fig : av_results ] shows the rms scatter in @xmath78 vs. test @xmath0 value , the result of applying the @xmath25 rms method to ngc 4833 ( filled diamonds ) and three sample simulated cluster data sets ( open triangles ) . the simulated data sets are all based on an input @xmath0 value of 3.0 , but have different levels of photometric error : 10% , 50% , and 100% that of the ngc 4833 data set ( 3.4 ) , with the rms sequences arranged in order of @xmath80 increasing upwards .
as expected , each sequence of rms of @xmath78 vs. test @xmath0 goes through a clear minimum ( large open square ) near @xmath62 .
the depth / sharpness of the rms minimum decreases with increasing photometric error and the power of the method diminishes .
the results for all @xmath81 simulated cluster data sets are summarized in table [ tbl : sim ] . for @xmath73 ,
the @xmath25 rms method reproduces the correct @xmath0 to within @xmath82 , as measured by the standard deviation of measured ( output ) @xmath0 values for 100 simulated data sets .
the scatter in output @xmath0 increases to @xmath83 for @xmath63 .
the ( absolute ) precision of the method appears to decrease with increasing input @xmath0 , again judging from the scatter of output @xmath0 values , although the fractional error @xmath84 remains roughly constant .
it is important to note that the output @xmath0 from this method tends to be biased low for moderate to large photometric errors and input @xmath85 .
for example , simulated data sets with input @xmath62 and @xmath63 ( a good match to the real cluster data ) have a mean output @xmath86 .
this bias of @xmath87 in @xmath88 is significant compared to the error in the mean , @xmath89 , and must be taken into account in the determination of @xmath0 for an actual cluster data set .
the cause of the bias towards low @xmath0 may be explained as follows .
photometric error produces scatter in the hb which , unlike the scatter caused by variable extinction / reddening , is _ uncorrelated _ between the two cmds i.e . , the resulting values of @xmath74 and @xmath75 are different .
these errors are particularly important at faint magnitudes where the hb is vertical and the effect is a spread in color . all else
being equal , choosing a smaller @xmath0 ( and @xmath44 ) translates these color errors into smaller @xmath3 values and hence to smaller ( @xmath76)@xmath90 differences as well .
this biases the rms minimum towards lower @xmath0 values .
the rms of @xmath78 for the ngc 4833 data set displays a significant minimum at @xmath91 ( open square associated with filled diamonds in fig .
[ fig : av_results ] ) , where the error bar is taken directly from the scatter in output @xmath0 for comparable simulated data sets . correcting for the bias in the measured @xmath88 for such simulations ,
our best guess for the @xmath0 towards ngc 4833 is @xmath92 .
the `` @xmath25 slope '' method is closely related to the @xmath25 rms method in that it relies on the same @xmath74 and @xmath75 values determined for each star in 4.1.1 above . instead of computing the difference between the two @xmath3 values as in the @xmath25 rms method ,
one is plotted against the other and a linear fit made .
this is illustrated in figure [ fig : av_slope_method ] for a simulated cluster data set created with ( input ) @xmath62 ; four test @xmath0 values are shown .
the linear fit in each panel is effectively a bivariate one : one of the dashed lines is a fit treating the @xmath93- and @xmath94-axes as the independent and dependent variables , respectively , and vice versa for the other dashed line ; their slopes are averaged to get the solid line .
as expected , the slope of the linear fit approaches unity when the test @xmath0 approaches the true ( input ) @xmath0 value . as in 4.1.1
, a polynomial fit is made to the slope vs. test @xmath0 data to interpolate to unit slope .
figure [ fig : slope_results ] shows results from the application of the @xmath25 slope method : the slope of the relation between @xmath74 and @xmath75 is plotted versus test @xmath0 .
the format of this figure is the same as for figure [ fig : av_results ] , except that a slope of unity ( rather than an rms minimum ) corresponds to the best - fit @xmath0 value .
the @xmath25 slope method results for ngc 4833 and simulated data sets are summarized in table [ tbl : sim ] . applying this method to @xmath81 simulated data sets
, we learn that , like the @xmath25 rms method , the power of the @xmath25 slope method decreases with increasing photometric errors .
the scatter in output @xmath0 is @xmath82 for @xmath73 , but increases to @xmath95 for @xmath63 .
some of the 9 simulation categories display a small but significant bias ( positive or negative ) in the measured ( output ) @xmath88 , but not the category directly relevant to ngc 4833 .
the cause of this bias is not as readily apparent as for the `` @xmath25 rms '' method ( 4.1.2 ) . for the ngc
4833 data set , the @xmath25 slope method gives @xmath96 .
the `` optical / ir '' method is different from the previous two methods in that it relies on two - color photometry ( i.e. , a single cmd which , in our case , is @xmath5 vs. @xmath4 ) and the sfd reddening map . using a test @xmath0 value ,
@xmath25 values are derived for all hb stars by scaling the corresponding sfd @xmath43 values .
these color and magnitude corrections are applied on a star - by - star basis .
figure [ fig : schlegel_method ] shows the ( @xmath4 , @xmath5 ) cmd of a simulated hb before ( left ) and after ( right ) making the extinction / reddening corrections .
the latter hb is slightly tighter , but not dramatically so because the sfd reddening values smooth over small - scale dust variations .
the rms scatter of the corrected hb stars is computed with respect to a best - fit fiducial curve ( two second - order polynomials solid line in fig .
[ fig : schlegel_method ] ) based on `` shortest distance '' ( see 3.3 ) .
this process is repeated for a range of test @xmath0 values as for the other two methods .
the rms scatter is expected to go through a minimum when the test @xmath0 approaches the true @xmath0 value . as in 4.1.1
, a polynomial fit is made to the hb scatter vs. test @xmath0 data to interpolate to the minimum .
figure [ fig : schlegel_results ] shows results from the application of the optical / ir method : rms scatter of hb stars versus test @xmath0 .
the format of this figure is the same as for figure [ fig : av_results ] . the optical / ir results for ngc 4833 and simulated data sets are summarized in table [ tbl : sim ] .
tests of the method on @xmath81 simulated cluster data sets indicate that the method has a low level of predictive power ( @xmath97 in @xmath0 ) over the full range of photometric errors simulated .
this is because the main source of error here is uncertainty in the sfd reddening values , including small - scale dust variations that get smoothed out by the relatively coarse resolution of their map .
the optical / ir method gives @xmath98 for ngc 4833 .
our best estimate of the extinction law slope in the direction of ngc 4833 is @xmath7 , based on a weighted average of the values from the three determination methods ( see table [ tbl : sim ] ) . before averaging ,
the output @xmath0 value from the @xmath25 rms method is corrected for measurement bias : @xmath99 ( 4.1.2 ) .
the weights are proportional to @xmath100 ^ 2 $ ] where @xmath101 is the uncertainty in output @xmath0 for a given determination method as derived from the simulations .
figure [ fig : cmd6 ] illustrates the effectiveness of the star - by - star extinction / reddening corrections based on the @xmath0 determination methods described earlier in this section .
the top two panels show the uncorrected ngc 4833 hb in the ( @xmath4 , @xmath5 ) and ( @xmath6 , @xmath5 ) cmds ( left and right panels , respectively ) .
the photometry in the middle two panels has been corrected for extinction / reddening based on the sfd @xmath43 maps , using a best - fit @xmath62 and the corresponding @xmath44 from eqn .
[ eqn : cardel ] to translate the @xmath4 color excess to @xmath25 and @xmath45 .
note , the mean extinction and reddening of the cluster have been added back to the data for ease of comparison with the top panels .
the hb is somewhat tighter after the sfd - based correction but the effect is fairly subtle , a result of missing small - scale dust structure in the sfd map and possibly other sources of uncertainty as well .
the correction in the bottom two panels of figure [ fig : cmd6 ] is based on information derived from @xmath1 photometry of the cluster hb .
the relative visual extinction measures derived from the two cmds , @xmath74 and @xmath75 ( 4.1.1 ) , are averaged to obtain a @xmath102 value for each hb star .
this is translated to color excesses in the two cmds using the best - fit @xmath62 and its corresponding @xmath44 ( eqn .
[ eqn : cardel ] ) .
the dramatic tightening of the hb indicates that the scatter is strongly correlated between ( @xmath4 , @xmath5 ) and ( @xmath6 , @xmath5 ) cmds
e.g . , if a star is displaced to the right of / below the hb fiducial in one cmd , it tends to be displaced in the same direction and by the same amount ( or by a proportional amount in color ) in the other cmd .
the most natural explanation of this correlated hb scatter is of course differential extinction / reddening in the ngc 4833 data set .
there is a handful of stars at each extremity of the hb in the bottom two panels of figure [ fig : cmd6 ] that do not follow the tight sequence delineated by the rest of the stars in fact , their scatter is unchanged relative to the uncorrected cmds in the top panels .
this is because the polynomial hb fiducial used to determine @xmath3 does not extend all the way to the ends of the sequence ( to avoid extrapolation of the fit ) ; as a result no @xmath3 measurement or correction is available for these stars .
a couple of stars near the middle of the hb continue to show large deviations in the @xmath3-corrected cmds but these are probably non - hb stars ( see 4.1.1 and fig .
[ fig : radcmd ] ) . before closing out this discussion ,
two additional sources of hb scatter are considered .
the first of these is photometric error .
errors in @xmath56 and @xmath57 magnitudes produce uncorrelated shifts between the two cmds .
errors in @xmath5 produce shifts with slope @xmath103 in the ( @xmath4 , @xmath5 ) cmd , more or less orthogonal to the reddening vector , and shifts of slope @xmath104 in the ( @xmath6 , @xmath5 ) cmd , roughly along the reddening vector though its slope @xmath44 is significantly steeper than unity . only at the very top of the hb where it becomes horizontal can @xmath5 magnitude errors mimic differential extinction / reddening effects . to summarize
, photometric errors can not explain the correlated scatter seen in ngc 4833 s hb . moreover , as will be demonstrated in 5 , the difference in @xmath3 between pairs of hb stars scales with their angular separation in a manner consistent with the dust power spectrum measured at larger angular separations ; photometric errors have no natural way of explaining this trend either . another factor to consider
is _ intrinsic _ scatter in the hb caused by astrophysical effects such as variations in chemical abundance , age , and mass loss between cluster stars .
it is possible that these can produce correlated scatter between the two cmds and thereby mimic differential extinction / reddening .
however , this scenario can not explain the scaling of the @xmath3 differences with angular separation ( 5 ) .
if there was a radial gradient in intrinsic stellar properties within the cluster , one might expect to see a shift of the hb sequence in the cmd between inner and outer regions .
instead , figure [ fig : radcmd ] shows no such shift ( the increased scatter in the hb in the outer cmd is best explained in terms of dust variations ; see 2.1 ) .
in earlier sections , we demonstrated the effects of spatial extinction / reddening variations on the photometry of ngc 4833 .
we now turn to a statistical description of the strength of these extinction / reddening variations as a function of angular scale .
figure [ fig : dustmap ] is a map of the _ relative _ visual extinction across the face of ngc 4833 .
each point on the map represents the sky position of a hb star in the cluster .
the size of each point is scaled by the measured strength of the relative visual extinction @xmath74 at that location ; it varies from @xmath105 to @xmath106 mag ( smallest@xmath107largest ) .
as described in 4.1.1 , @xmath74 is the vertical distance between the star s cmd location and the point where the reddening vector drawn from the star intersects the fiducial hb ; the fiducial is a fit to the central region of the cluster , and therefore corresponds to a mean visual extinction of @xmath108 . in keeping with the findings of melbourne et al .
( 2000 ) , there is a clear overall n - s gradient in @xmath74 across ngc 4833 , but the extinction can vary significantly from point to point even for points relatively close to each other .
these variations can be characterized in terms of the angular power spectrum of the foreground dust complex .
a direct measurement of the power spectrum is difficult however because of the sparse and irregular spatial sampling of the data points .
the following approach is adopted instead .
first , the `` variance spectrum '' as a function of angular scale is calculated from the cluster @xmath74 data ( 5.1 ) .
next , a variance spectrum and conventional power spectrum are measured from the uniformly - sampled sfd maps , and an empirical transformation is calculated from the former to the latter ( 5.2 ) .
finally , this empirical transformation is applied to the cluster variance spectrum to convert it to a power spectrum , and this allows us to study the composite power spectrum over a wide range of angular scales ( 5.3 ) .
the best - fit value of @xmath62 for ngc 4833 is used to determine the relative visual extinction corresponding to the @xmath9-th hb star , @xmath109 . the difference in relative @xmath25 between stars @xmath9 and @xmath10 , @xmath110 , and their angular separation @xmath11
are then calculated for all possible pairs of hb stars .
figure [ fig : powspec ] ( open diamonds ) shows the variance ( mean square deviation ) of @xmath111 as a function of angular separation @xmath112 in a log - log plot ; the radial bins are chosen so as to include roughly equal numbers of pairs in each bin .
the variance decreases with decreasing @xmath112 before levelling off at log@xmath113\bigr]=-2.3 $ ] for @xmath114 .
the behavior at low @xmath112 may be attributed to photometric errors and possibly the natural width of the hb . in order to quantify and compensate for the effect of photometric error on the variance spectrum , we resort to the special set of `` @xmath115-only '' simulations containing _ no _ reddening variations ( 3.3.1 ) .
the variance of this `` @xmath60-only '' simulated data set , analyzed in the same way as the real cluster data set , is : log@xmath116\bigr]=-2.4 $ ] , and is of course independent of stellar angular separation ( dashed horizontal line in fig .
[ fig : powspec ] ) .
the variance due to photometric errors is subtracted from the variance spectrum of ngc 4833 , equivalent to statistical subtraction of errors in quadrature .
the resulting variance spectrum is shown as filled diamonds in figure [ fig : powspec ] .
a power - law fit to the corrected variance spectrum : @xmath117 yields an index of @xmath14 , over the radial range @xmath118@xmath119 .
the upper end of this range corresponds to the nyquist frequency of the @xmath41 area covered by the ngc 4833 data set .
the first radial bin of the corrected variance spectrum ( @xmath120 ) is excluded from the power - law fit because it is strongly affected by uncertainties in the photometric error correction .
the cluster variance spectrum is used to make a rough estimate of how much small - scale power is missing from the sfd dust map . the sfd map , like the iras 100@xmath26 m data on which they are based , have an angular resolution of about @xmath121 , or @xmath122 .
this point in the cluster variance spectrum corresponds to var@xmath123=4.4\times10^{-3}$ ] .
the rms scatter at this point is @xmath124}=6.6\times10^{-2}$ ] . since the mean visual extinction @xmath2 in the direction of ngc 4833
, the rms is about 7% of the mean .
this provides a good sanity check : it corroborates the result from the analysis of the cluster hb width ( 3.3.2 ) that the sfd map is missing small - scale variations on the order of 6% of the reported reddening value .
it also gives an indication of how accurately one can correct for extinction / reddening using the sfd map , with general implications for precision photometry in dusty regions of the sky . in this section ,
the sfd map is used to investigate the relationship between the variance spectrum and conventional angular power spectrum .
this is done by extracting a @xmath125 sfd reddening image centered on ngc 4833 . the @xmath126 reddening values are converted to @xmath25 using the best - fit @xmath0 of 3.0 .
the difference in visual extinction between pixels @xmath9 and @xmath10 , @xmath127 , is calculated for all possible pairs of pixels .
the variance of @xmath128 is computed as a function of angular separation @xmath112 .
the ` @xmath129 ' symbols in figure [ fig : pwvr ] show @xmath130 for the sfd map , where @xmath19 .
the two - dimensional power spectrum of dust fluctuations is computed via a fast fourier transform ( fft ) of the sfd reddening image .
this is then azimuthally averaged to produce a one - dimensional angular power spectrum @xmath131 , shown by asterisks in figure [ fig : pwvr ] . over the radial range @xmath132 to @xmath133 [ @xmath134 to @xmath135 , @xmath136 and @xmath131
have comparable logarithmic slopes ( power - law index : @xmath137 ) such that : @xmath138 with a best - fit scale factor of @xmath139 . the above transformation ( eqn .
[ eqn : var_to_ps ] ) is applied to the photometric - error - corrected variance spectrum derived from the ngc 4833 @xmath140 data ( 5.1 ) to convert it to an angular power spectrum .
the result is shown as filled diamonds in figure [ fig : pwvr ] .
this small - scale power spectrum derived from optical data matches smoothly onto the power spectrum derived from the sfd map at larger angular scales .
a power law with an index @xmath141 provides an adequate fit to the data over the full range of projected separations shown in figure [ fig : pwvr ] , @xmath17 to @xmath142 .
our best - fit index @xmath143 is significantly shallower than the index of @xmath144 found by gautier et al .
( 1992 ) and guhathakurta & cutri ( 1994 ) and at the shallow end of the range of indices found by kiss et al .
( 2003 ) in their isophot - based study of 13 fields : @xmath145 to @xmath146 ( @xmath82 ) .
the shallower the power - spectrum index , the larger the _ fraction _ of power missed by the sfd map on small scales relative to that on larger scales .
another possible implication relates to kiss et al.s finding that shallower @xmath143 values are associated either with low hi column density , @xmath147 @xmath148 , and/or warm dust .
for example , in the region of the draco nebula , they find that the map of 90@xmath26 m emission ( dominated by warmer dust ) yields @xmath149 while the 170@xmath26 m map ( dominated by cooler dust ) yields @xmath150 .
while ngc 4833 has an extended blue hb , the methods presented in this paper should be usable for clump hbs or the rgb or any other tight cmd feature for that matter .
for instance , law et al .
( 2003 ) used the width of the rgb to measure differential reddening in the direction of three galactic globular clusters , von braun et al .
( 2002 ) used the main sequence for their study , and udalski ( 2003 ) used clump stars to probe dust along the line of sight to the galactic bulge .
our methods should work for any three bands , not just @xmath1 .
additional bands ( beyond three ) can provide : ( 1 ) a longer wavelength baseline and therefore more leverage in @xmath0 determination ; ( 2 ) a larger number of independent measures of the star - by - star extinction / reddening which can be averaged to beat down the errors ; and ( 3 ) most importantly , an empirical check of the ccm parameterization of the shapes of galactic extinction laws .
there is at least one generalization of the methodology presented here that might be worth exploring in the future . in determining @xmath3
, we effectively use @xmath5 magnitude to mark where along the hb sequence the extinction / reddening - corrected star lies . a more general definition of a `` locator '' parameter , to mark the position of the unreddened star along any given cmd sequence , may be particularly useful .
for example , multi - wavelength data sets may not have any single observable in common across all of the cmds in question ( unlike the @xmath5 mag for our two cmds ) , and a color parameter may be more sensitive than @xmath5 mag when using a not - so - vertical cmd feature ( such as the upper portion of ngc 4833 s hb ) .
several high - quality , multi - band ( three or more ) , ground- and space - based photometric data sets have long been available for many star clusters , so the methods presented here can be readily implemented to probe the nature of dust along many lines of sight .
these include milky way clusters , as well as others near enough to allow clean photometry in relatively uncrowded fields , such as _ hubble space telescope _
images of clusters in , and even somewhat beyond , the magellanic clouds .
many of these clusters have detectable extinction / reddening variations across them e.g . , @xmath151 centauri , ngc 6388 , ngc 6441 ( law et al .
2003 ) , m10 , and m12 ( von braun et al .
the sloan digital sky survey ( sdss , stoughton et al .
2002 ) five - band ( @xmath152 ) data set , with its high photometric accuracy , minimal systematic errors , excellent uniformity / homogeneity , and large areal coverage , is opening up lines of sight towards differentially - reddened star clusters across the entire high - latitude north galactic cap for this type of analysis .
near - infrared @xmath153 photometry from the two micron all - sky survey ( 2mass ) data should also prove very fruitful , both for probing highly - reddened lines of sight and for extending the sdss wavelength coverage .
making @xmath0 and power spectrum measurements along many sight - lines should lead to a better understanding of grain formation and chemistry and the effect of interstellar dust on astronomical observations .
while our @xmath54 grid of simulations covers some amount of parameter space in photometric error and @xmath0 , it is fairly limited in scope : it can not simply be translated to any arbitrary cluster data set .
another data set may be different from the ngc 4833 data set in terms of photometric bandpasses , level of photometric error , areal coverage , detailed morphology of cmd features , degree of extinction / reddening variations , and/or extinction law slope .
new simulations must be carried out , matched to the exact parameters of each new cluster data set in question , in order to test the efficacy of the @xmath0 measurement methods presented here .
* we have demonstrated the use of three methods , `` @xmath25 rms '' , `` @xmath25 slope '' , and `` optical / ir '' methods , for determining the dust extinction law slope @xmath0 in the direction of differentially - reddened galactic globular clusters , and have tested the methods on an extensive suite of simulated cluster data sets .
* for cluster data sets with low photometric error , @xmath154 mag , the `` @xmath25 rms '' and `` @xmath25 slope '' methods can be used to determine @xmath0 to an accuracy of @xmath1550.3 . for cluster data
sets with @xmath156 mag , the two methods yield @xmath0 to within @xmath83 , with some systematic biases . * the `` optical / ir '' method generally provides relatively imprecise estimates of @xmath0 , @xmath157 , over the full range of photometric errors explored . *
combining the results from all three methods gives a mean extinction law slope of @xmath7 for the line of sight towards the low - latitute galactic globular cluster ngc 4833 .
* the scatter in the cluster hb is used to estimate the amount of small - scale structure in the dust complex in the foreground of ngc 4833 .
the schlegel et al .
( 1998 ) iras+dirbe - based map of the dust thermal emission averages over small - scale reddening variations that are @xmath158% of the mean reddening value .
* star - to - star variations in relative visual extinction across the face of ngc 4833 provide a measure of the foreground dust angular power spectrum for projected separations in the range @xmath17@xmath18
. this small - scale power spectrum derived from cluster optical data matches smoothly onto the larger - scale power spectrum derived from the sfd reddening map of the region .
the overall power spectrum is well fit by a power law : @xmath15 , where @xmath30 is the reciprocal of the angular scale @xmath112 , with spectral index @xmath137 .
we would like to thank neil balmforth , david burstein , sandra faber , and peter stetson for comments that significantly improved this work .
we would also like to thank ata sarajedini for providing the observations of ngc 4833 .
cc|c|ccc + & & 2.0 & @xmath159 & @xmath160 & @xmath161 + 0.003 & 10 & 3.0 & @xmath162 & @xmath163 & @xmath164 + & & 4.0 & @xmath165 & @xmath166 & @xmath167 + & & 2.0 & @xmath168 & @xmath169 & @xmath170 + 0.015 & 50 & 3.0 & @xmath171 & @xmath172 & @xmath173 + & & 4.0 & @xmath174 & @xmath175 & @xmath176 + & & 2.0 & @xmath177 & @xmath178 & @xmath179 + 0.030 & 100 & 3.0 & @xmath180 & @xmath181 & @xmath182 + & & 4.0 & @xmath183 & @xmath184 & @xmath185 + + 0.030 & & & @xmath186 & @xmath187 & @xmath188 | we present three methods for measuring the slope of the galactic dust extinction law , @xmath0 , and a method for measuring the fine - scale structure of dust clouds in the direction of differentially - reddened globular clusters .
we apply these techniques to @xmath1 photometry of stars in the low - latitude galactic globular cluster ngc 4833 which displays spatially - variable extinction / reddening about a mean @xmath2 .
an extensive suite of monte carlo simulations is used to characterize the efficacy of the methods .
the essence of the first two methods is to determine , for an assumed value of @xmath0 , the _ relative _ visual extinction @xmath3 of each cluster horizontal branch ( hb ) star with respect to an empirical hb locus ; the locus is derived from the color - magnitude diagram ( cmd ) of a subset of stars in a small region near the cluster center for which differential extinction / reddening are relatively small .
a star - by - star comparison of @xmath3 from the ( @xmath4 , @xmath5 ) cmd with that from the ( @xmath6 , @xmath5 ) cmd is used to find the optimal @xmath0 . in the third method , @xmath0 is determined by minimizing the scatter in the hb in the ( @xmath4 , @xmath5 ) cmd after correcting the photometry for extinction and reddening using the schlegel , finkbeiner , & davis ( 1998 ) dust maps .
the weighted average of the results from the three methods gives @xmath7 for the dust along the line of sight to ngc 4833 .
the fine - scale structure of the dust is quantified via the difference , @xmath8 , between pairs of cluster hb stars ( @xmath9,@xmath10 ) as a function of their angular separation @xmath11 .
the variance ( mean square scatter ) of @xmath12 is found to have a power - law dependence on angular scale : @xmath13 , with @xmath14 .
this translates into an angular power spectrum @xmath15 , with the index @xmath16 for @xmath17@xmath18 , where @xmath19 .
the dust angular power spectrum on small scales ( from optical data ) matches smoothly onto the larger - scale power spectrum derived from schlegel et al.s far - infrared map of the dust thermal emission . |
the origin of binary stars is still a puzzle in our understanding of star formation .
numerous theoretical simulations support the hypothesis that the fragmentation of collapsing molecular cloud cores , promoted by either rotation or turbulence , is the main mechanism for the formation of binary / multiple stellar systems ( see reviews by bodenheimer et al .
2000 , tohline 2002 , and goodwin et al . 2007 ) . nevertheless
, many key questions concerning this fragmentation process , e.g. , when exactly does the fragmentation start , are still under debate ( see tohline 2002 ) .
although it is generally assumed that cloud cores do not fragment during the free - fall collapse phase , several groups ( e.g. , bate & burkert 1997 ; tsuribe & inutsuka 1999 ; boss et al . 2000 ) found that fragmentation can occur at the end of the isothermal phase ( generally referred to as prompt fragmentation ) , while others ( e.g. , truelove et al . 1997 , 1998 ; machida et al .
2005 , 2008 ) argue that the isothermal gas is stable against fragmentation and found that fragmentation only occurs during / after the adiabatic phase ( see figure 9 in andr et al .
2009 for a discussion of the different evolutionary phases of core collapse ) . on the observational side , a handful of young protostellar ( i.e. , class0 )
binary systems have been found ( e.g. , looney et al .
2000 ; launhardt 2004 ) , and there are increasing kinematics studies of binarity in the protostellar phase ( e.g. , chen et al .
2007 , 2008 , 2009 ; volgenau et al . 2006 ; maury et al . 2010 ) .
however , the number of observed and well - studied protostellar binary ( protobinary ) systems is still small , and these systems have already gone through the fragmentation process .
the observational link between the initial conditions in a fragmenting cloud core and the final stellar systems formed therein is still missing .
it is therefore critical to observe , at high angular resolution , more dense cores in nearby molecular clouds to search for the earliest phase of fragmentation and study in detail their properties , in order to put direct constraints on fragmentation models . in this letter
, we present submillimeter array ( sma ; ho et al .
2004 ) dust continuum observations toward the r corona australis ( rcra ) region . at a distance of @xmath5170pc ( knude & hg 1998 ) ,
the cra dark cloud is one of the nearest star - forming regions . as the most active star formation site in this cloud
, the rcra region has been extensively observed in the past two decades ( see review by neuh@xmath6user & forbrich 2008 , and references therein ) . using scuba on the james clerk maxwell telescope ( jcmt ) , nutter et al .
( 2005 ) found a prestellar core candidate termed smm1a in the rcra region , which is elongated in the east - west direction and has a gas mass of @xmath510@xmath7 , and an effective radius of @xmath53000au . with a maximum velocity dispersion of about 0.8kms@xmath8 ( harju et al .
1993 ) , the smm1a core is gravitationally bound .
infall motions in this region of the cloud further confirm that this is a prestellar core ( groppi et al .
2004 ) . in this letter ,
we report the discovery of a multiple system within the smm1a core , based on high angular resolution millimeter observations .
this may represent the earliest phase of core fragmentation observed thus far .
the rcra region was observed with the sma on 2006 august 20 in the compact configuration .
six antennas were used in the array , providing baselines from 5k@xmath9 to 52k@xmath9 at 220ghz .
the sma primary beam is about 55@xmath10 at this frequency .
two positions next to each other at ( r.a . , decl.)@xmath11=(19:01:53.3 , @xmath1236:57:21.0 ) and ( r.a . ,
decl.)@xmath11=(19:01:56.4 , @xmath1236:57:27.0 ) were observed .
the digital correlator was set up to cover the frequency ranging [email protected] and [email protected] in the lower and upper sidebands , respectively . the 1.3 mm dust continuum emission
was recorded with a total bandwidth of @xmath53.3ghz ( @xmath51.8ghz usb and @xmath51.5ghz lsb ) .
system temperatures for rcra ranged from 100 to 280k ( depending on elevation ) , with a typical value of @xmath5200k .
the visibility data were calibrated with the mir package ( qi 2005 ) , with quasars 3c454.3 and 1924 - 292 as the bandpass and gain calibrators , respectively .
uranus was used for absolute flux calibration , from which we estimate a flux accuracy of @xmath520% , by comparing the measured quasar fluxes with those listed in the sma calibration database .
the calibrated visibility data were imaged using the miriad toolbox ( sault et al .
the sma synthesized beam size at 1.3 mm dust continuum , with robust _ uv _ weighting 0 , is 5.7@[email protected]@xmath10 .
figure 1a shows the sma 1.3 mm dust continuum image of rcra , overlaid with the scuba 850@xmath14 m dust continuum contours ( from groppi et al .
the northern part of this image shows clear detection of the dust continuum emission associated with the herbig - ae star rcra and protostellar cores smm2 , smm1b and smm1c these last two are also referred to as sma1 and sma2 in groppi et al .
( 2007 ; see also choi et al .
2008 ) . to the south ,
the scuba 850@xmath14 m image shows the smm1a core , which is elongated in the east - west direction ( see fig.1a ) . at higher angular resolution
, the scuba 450@xmath14 m image in figure 1b shows that the smm1a core is clearly resolved into two sub - cores ( see also van den ancker 1999 ) . at even higher resolution
, the sma 1.3 mm continuum observations reveal that the smm1a core is divided into three condensations , aligning roughly from east to west , which we refer to here as smm1a - a , smm1a - b , and smm1a - c ( fig.1b ) .
all are detected with signal - to - noise ratio of 4 or more .
source smm1a - a coincides with the eastern sub - core observed in the scuba 450@xmath14 m image , while smm1a - b and smm1a - c are coincident with the western sub - core ( fig.1b ) .
the single - dish and interferometer observations are in general agreement with each other , which indicates that the three continuum sources detected with the sma are neither artifacts nor are due to noise in the interferometer image46@xmath10 at 270ghz ) ; ( 2 ) more flux was resolved out in their extended configuration observations ; and ( 3 ) the 1.1 mm observations lack the sensitivity to detect these weak sources . ] . also note that the sio(54 ) line and several other molecular lines , e.g. , dcn(32 ) , were included in this sma correlator setup , but no line emission was detected from the three condensations .
figure 2 shows the mid - infrared images of the rcra region , taken by the _
spitzer space telescope_. the mips1 band image at 24@xmath14 m ( fig.2a ) clearly shows infrared emission at the position of cores smm1b ( associated with the infrared source irs7b , see wilking et al .
1997 ) , smm2 , and the herbig stars rcra and tcra .
there is also 24@xmath14 m emission from the smm1c core , which may be associated with the radio source b9 centered on the core ( see choi et al .
2008 ) , and the infrared source irs7a located @xmath55@xmath10 to the south .
the mips2 band image at 70@xmath14 m ( fig.2b ) has a lower angular resolution and is dominated by the emission from the smm1c / irs7a region .
the mips3 ( 160@xmath14 m ) image of rcra is extremely saturated , and is thus not shown here .
none of the @xmath15 images at bands from 3.6 to 70@xmath14 m shows compact infrared emission in the smm1a core , which suggests that the core is extremely cold . from the sma 1.3 mm images , we derived the positions , fluxes , and ( deconvolved ) fwhm sizes of the three sources in rcra smm1a using a multi - component gaussian fitting routine ( see table 1 ) .
angular separations between sources smm1a - a and
-b and smm1a - b and -c are 122@xmath1603 and 58@xmath1603 , corresponding to projected separations of @xmath52100au and @xmath51000au ( at a distance of 170pc ) , respectively . assuming that the 1.3 mm dust continuum emission is optically thin , the total gas mass ( @xmath17 ) of the three sources is calculated with the same method as described in launhardt & henning ( 1997 ) . in the calculations ,
we adopt a dust opacity of @xmath18 ( ossenkopf & henning 1994 ) , a typical value for dense and cold molecular cloud cores , and a mass - averaged dust temperature of @xmath518k ( see below ) .
the gas masses of the three sources , derived from the sma 1.3 mm dust continuum observations , range from @xmath50.10 to @xmath50.23@xmath7 ( see table 1 ) . with these mass and the fwhm size of the condensations
, we estimate their average densities to be in the order of @xmath19 to @xmath20@xmath21 .
note that these are average densities and the local peak densities in these sources may be orders of magnitude higher .
these high densities , as well as the infall motions detected in this region ( see groppi et al .
2004 ) , suggest that the three condensations in the smm1a are on their way to form low - mass stars .
considering that the smm1a core has a mass of approximately 10@xmath22 , and assuming a core - to - star efficiency of @xmath23% ( evans et al .
2009 ) , it is probable that a multiple stellar system with a total mass of 3 @xmath22 will eventually form in smm1a .
table 1 lists the 450@xmath14 m and 850@xmath14 m fluxes estimated from the scuba images , and the 800@xmath14 m fluxes adopted from van den ancker ( 1999 ) . since sources smm1a - b and smm1a - c are not resolved in the jcmt observations , we treat them as one sub - core here . the scuba 450@xmath14 m fluxes
are derived by fitting two - dimensional gaussians toward the two sub - cores ( smm1a - a and smm1a - b+c ) , while the 850@xmath14 m fluxes are estimated from the fluxes enclosed within roughly one beam around each sub - core .
based on the sma 1.3 mm and jcmt submm data points , as well as the 3@xmath24 upper limits in the @xmath15 mips images , we construct the spectral energy distributions ( seds ) of the two sub - cores ( see figure 3 ) . in order to derive luminosities and temperatures , we first interpolate and then integrate the seds , always assuming spherical symmetry .
interpolation between the flux densities is done by a @xmath25@xmath26 single - temperature grey - body fitting to all points ( including the upper limits ) .
the resulting bolometric luminosities are @xmath270.08@xmath28 for smm1a - a and @xmath270.09@xmath28 for smm1a - b+c .
the dust and bolometric temperatures are between 17 and 19k for the sources .
the low luminosities and temperatures , as well as the fact that no compact infrared emission is detected from these sources in the @xmath15 images , resemble the typical properties of prestellar cores ( see review by andr et al .
2009 ) . in the vla survey of the rcra region ,
no radio source is detected at the positions of the three condensations presented here ( choi et al . 2008 ) .
in contrast , the protostars smm1b ( class0/i ) and smm1c ( class0 ) to the north of smm1a are bright and centrally peaked at millimeter wavelengths ( fig.1a ) , and are associated with hard x - ray ( forbrich & preibisch 2007 ) , infrared ( fig.2a ) , and centimeter radio counterparts ( choi et al . 2008 ) .
therefore , although the chemical and kinematic properties of smm1a are still poorly known , the observations thus far suggest that the three sources in smm1a are in an earlier evolutionary stage than that of class 0 protostars . as predicted by theoretical studies , there are different phases in the evolution from an initial isothermal ( 10k ) prestellar core to a class 0 protostar ( andr et al .
the low luminosities ( @xmath270.1@xmath2 ) and temperatures ( close to 20k ) of the three sources in smm1a are similar to the predicted properties of so - called first hydrostatic ( adiabatic ) cores , which are formed after the isothermal collapse of prestellar cores ( e.g. , masunaga et al .
1998 ; chen et al . 2010 ) .
however , it must be noted that large uncertainties remain in our estimates due to the limited observations available .
further high angular resolution observations at different wavelengths are needed to constrain the seds of the three sources in order to address more precisely their evolutionary statuses .
regardless of this , it is fair to say that _ the rcra smm1a multiple system represents the earliest phase of core fragmentation observed thus far_. prompt fragmentation of rotationally flattened cloud cores with initially flat density profiles , immediately after a phase of free - fall collapse , has long been suggested as one of the most efficient mechanisms for binary star formation ( tohline 2002 ) .
since this fragmentation is expected to occur at the end of the isothermal collapse phase , it is also referred to as isothermal fragmentation .
the separation of the fragments formed via prompt fragmentation has a scale of 10@xmath26@xmath1210@xmath29au , which corresponds to the jeans length in this phase .
in contrast , other groups found that isothermal gas is stable against fragmentation and that core fragmentation can only occur during ( and after ) the adiabatic phase ( e.g. , truelove et al .
1998 ; machida et al . 2008 ) .
because this fragmentation occurs after the formation of the first adiabatic core , it is also called the first core fragmentation ( machida et al .
2005 ) . here
, the separation of the fragments formed generally has a scale of 3@xmath12300au ( machida et al .
clearly , observing the earliest phase of core fragmentation is the most effective way to discriminate between prompt ( isothermal ) and adiabatic fragmentation scenarios . if prompt fragmentation does occur in the isothermal phase
, we may expect to find a binary / multiple system consisting of adiabatic cores with the separation of @xmath51000au . as shown in the scuba images ( see figure 1b ) , the rcra smm1a core is highly elongated with an aspect ration of roughly 3 , showing a morphology similar to the bars or filaments seen in the simulations of prompt ( isothermal ) fragmentation ( see , e.g. , bate & burkert 1997 ; tsuribe & inutsuka 1999 ; boss et al .
2000 ) . from the mass and radius of the smm1a core ( see nutter et al . 2005 ) , the average volume density is calculated to be 1@xmath122@xmath1310@xmath30@xmath21 .
the corresponding jeans length is @xmath51000au for a temperature of @xmath510k .
the separations among the three sources found in the smm1a core are also consistent with this estimated jeans length .
moreover , our results from the sed fitting ( see above ) suggest that the three sources are at an evolutionary stage earlier than that of class0 protostars , perhaps in the first ( adiabatic ) core stage .
all these results appear to support the view that the three sources are formed through the prompt fragmentation of an elongated collapsing core during the isothermal phase . in this scenario
, we would expect the rotational axis of smm1a to be perpendicular to the direction of the core s major axis , which would be roughly perpendicular to the large scale rotation observed in the rcra region by groppi et al .
this is not necessarily inconsistent with our picture as other cores show drastic differences in the rotation axes measured at different scales ( e.g. , ohashi et al .
we also note that so far most observed protostellar binary / multiple systems ( e.g. , l723 , launhardt 2004 ; or ngc2264 d - mm1 , teixeira et al . 2007 ) have separations in the scale of 1000au , which also supports the scenario of prompt fragmentation , although these protostellar systems have had at least a few 10@xmath29 yrs to evolve from their prestellar stage and thus their separations may not represent real initial conditions . on the other hand ,
as introduced above , prompt ( isothermal ) fragmentation has trouble accounting for the close binary systems with separations @xmath4100au .
the fragmentation of cores at the adiabatic phase is still an attractive mechanism for close binary star formation . considering that the binary separation distribution in ( pre-)main sequence stars peaks at @xmath530au
, we actually would expect to find frequently close protobinary systems with separations of @xmath4100au to support the adiabatic fragmentation scenario . in practice , however , the number of this kind of systems is very small .
this observational result can be understood by the fact that current large millimeter interferometers ( e.g. , sma and iram - pdbi ) normally reach 12@xmath10 angular resolution under general conditions , and thus mostly resolve wide protobinary systems with separations of 200au or more in nearby star - forming clouds .
we believe that routine observations at angular resolution better than 0.5@xmath10 will reveal more close protobinary systems , which will then allow us to study in detail a statistically significant number of protostellar binary / multipe systems with a wide range of separations ( from @xmath510au to 10@xmath29au ) .
we present sma 1.3 mm dust continuum observations toward the rcra region .
the 1.3 mm dust continuum emission is detected from dense cores smm1b , smm1c , and smm2 , and the herbig - ae star rcra in this region .
we discover within the prestellar core candidate smm1a a multiple system made up of three condensations with masses in the range of 0.1 to 0.2@xmath0 .
the angular separations between the three new sources are @xmath512.2@xmath10 and @xmath55.8@xmath10 , corresponding to the projected separations of @xmath52100au and @xmath51000au , respectively . based on sma observations and complementary data from the jcmt and @xmath15 telescopes ,
we construct their spectral energy distributions ( seds ) and find that all three sources have extremely low bolometric luminosities ( @xmath270.1@xmath2 ) and temperatures ( @xmath2720k ) .
we suggest that the three condensations in smm1a resulted from the fragmentation of an elongated core in the isothermal phase .
the rcra smm1a system appear to be the earliest phase of low - mass core fragmentation observed thus far .
smm1a - a & 19:01:55.86 & @xmath1236:57:48.8 & 46@xmath169 & 56@xmath1352 & [email protected] & [email protected] & 1.5 & [email protected] + smm1a - b & 19:01:54.87 & @xmath1236:57:45.2 & 64@xmath1613 & 68@xmath1363 & [email protected] & [email protected]@xmath31 & 0.8@xmath31 & [email protected]@xmath31 + smm1a - c & 19:01:54.56 & @xmath1236:57:40.6 & 110@xmath1620 & 78@xmath1331 & [email protected] & [email protected]@xmath31 & 0.8@xmath31 & [email protected]@xmath31 + | we report the discovery of multiple condensations in the prestellar core candidate smm1a in the rcra cloud , which may represent the earliest phase of core fragmentation observed thus far .
the separation between the condensations is between 1000 and 2100au , and their masses range from about 0.1 to 0.2@xmath0 .
we find that the three condensations have extremely low bolometric luminosities ( @xmath1@xmath2 ) and temperatures ( @xmath3k ) , indicating that these are young sources that have yet to form protostars .
we suggest that these sources were formed through the fragmentation of an elongated prestellar core .
our results , in concert with other observed protostellar binary systems with separations in the scale of 1000au , support the scenario that prompt fragmentation in the isothermal collapse phase is an efficient mechanism for wide binary star formation , while the fragmentation in the subsequent adiabatic phase may be an additional mechanism for close ( @xmath4100au ) binary star formation . |
titan s thick atmosphere is simultaneously analogous to and extraordinarily different from that of earth .
both atmospheres are composed mostly of nitrogen .
both atmospheres have a major component that can exist as a solid , liquid , or gas ( water on earth , methane on titan ) . both have similar vertical temperature - pressure structures , although titan s atmosphere is @xmath6200 k cooler and is greatly extended relative to earth due to much lower surface gravity . on titan
a complicated network of photochemical reactions leads from methane ( ch@xmath7 ) to the formation of numerous heavier hydrocarbons , including propane ( @xmath8 ) .
measuring the abundances of titan s hydrocarbons provides an important test of models of titan s atmospheric chemistry and , more generally , of our fundamental understanding of atmospheric chemical and physical processes .
photochemical models ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) predict that propane is formed via c@xmath9h@xmath10 + ch@xmath0 + m @xmath11 c@xmath0h@xmath1 + m where ch@xmath0 is created by methane photolysis and c@xmath9h@xmath10 by c@xmath9h@xmath7 + h + m @xmath11 c@xmath9h@xmath10 + m. the primary removal mechanism for propane is condensation at the cold tropopause , although photolytic destruction and reaction with c@xmath9h consume @xmath65 - 15@xmath12 of the propane formed . in the model of @xcite propane condensed onto titan s surface accounts for @xmath61@xmath12 of the carbon removed from titan s atmosphere .
detections of propane s @xmath3 rotation - vibration band near 748 @xmath4 were based on voyager i and infrared space observatory ( iso ) spectra .
voyager spectra had 4.3 @xmath4 resolution ( @xmath13 at @xmath14 @xmath4 ) and initial claims of a propane abundance of 2@xmath15 @xcite were later reduced to 7@xmath16 @xcite .
the resolution of these spectra left every band of propane blended with other emission features . at 748 @xmath4 the @xmath3 band of propane
was blended with features of acetylene ( @xmath17 ) and the derived abundance ( or even detection ) of propane is strongly dependent on the assumed @xmath17 vertical abundance profile .
spectra taken by iso were of higher resolution ( @xmath18 at @xmath14 @xmath4 ) but were still unable to resolve propane from acetylene .
from these data @xcite derived a nominal propane abundance of 2@xmath19 , although they acknowledged the substantial difficulties of disentangling propane emission from that of acetylene .
observations were made at the 3.0 meter nasa infrared telescope facility ( irtf ) using the texes mid - infrared high - resolution grating spectrograph @xcite on 13 , 14 , and 20 december 2002 ut in several overlapping spectral settings around propane s @xmath3 fundamental vibration at 748 @xmath4 . at this wavenumber a single spectral setting covers @xmath65 @xmath4 at a resolution of @xmath20 with slight gaps between spectral orders .
the strongest @xmath3 propane lines are blocked by telluric absorption at 748.3 @xmath4 .
the diffraction limit of irtf at these wavelengths and titan s diameter are both @xmath21 . with a slit width of @xmath22
our data represent a disk average spectrum of titan . the reduced spectrum from each night , corrected for the appropriate doppler shift , is shown in fig . [ fig1 ] along with the final combined spectrum .
all data were reduced using the standard pipeline reduction detailed in @xcite .
telluric transmission was corrected with callisto on the first two nights and the becklin - neugebauer object @xcite on the third night .
we flux calibrated using observations of @xmath23 gem @xcite on the final night , and estimate that this calibration is accurate to 10 - 20% .
we use a new line - by - line code @xcite , dividing titan s atmosphere below 1000 km into 50 spherical shells evenly spaced in @xmath24 and using spectral bins of @xmath25 @xmath4 in order to resolve even the narrowest emission lines .
line parameters for acetylene ( c@xmath9h@xmath9 ) and hcn are from hitran , with the hcn line positions adjusted to agree with the observations of @xcite .
the temperature - pressure profile is the ` recommended ' profile of @xcite . scattering effects are ignored and titan s haze is modeled as a single layer with a lower cutoff of 50 km and with haze opacity scale height equal to the gas density scale height .
the abundances of hcn and @xmath17 are each parameterized with their fractional abundances at 1 mbar ( @xmath26 , @xmath27 ) and with the slope of @xmath28(abundance)-@xmath28(pressure ) ( @xmath29 , @xmath30 ) , i.e. @xmath31 .
we investigated both a constant propane vertical profile as well as scaled versions of an abundance profile predicted by photochemical modeling @xcite .
species are held to saturation vapor pressures below their condensation altitudes . in order to calculate titan s total flux
, the model is run at 32 points from the center of titan s disk to off the edge of titan s solid limb , with the calculation points more closely spaced near the limb .
the modeling of propane s emission spectrum requires additional discussion .
several linelists exist for propane s @xmath3 band at 748 @xmath4 .
these include : the geisa databank ; an unpublished list based on fits to the laboratory spectra of @xcite and @xcite ( blass et al .
1988 ) ; and a list generated using the code of @xcite and the parameters of @xcite .
figure [ fig2 ] shows calculated spectra using each of these linelists compared to low - resolution laboratory spectra of @xcite and to the high - resolution spectra used in @xcite .
given the poor fit of the calculated spectra to the observed spectra in fig .
[ fig2 ] at the wavenumbers we observed , we opted to use the low - pressure ( 1 - 3 torr ) low - temperature ( @xmath6175 k ) high - resolution ( r@xmath32 ) transmission spectra of @xcite to construct an empirical absorption coefficient ( @xmath33 ) spectrum for propane .
this involved fitting for the smoothly varying baseline and then converting to @xmath33 units ( @xmath4 amagat@xmath34 ) using the sample s known pathlength ( 30 cm ) , pressure , and temperature . in using this empirical @xmath33 spectrum we are implicitly assuming that @xmath33 is independent of temperature .
our propane observations primarily probe regions of titan s atmosphere that are at 135 to 175 k and are therefore reasonably well matched by the conditions of these laboratory data used to construct @xmath33 .
most of our spectral fitting was performed using this empirical @xmath33 for propane . as a test of the importance of the mismatch between titan atmospheric and laboratory temperatures we re
- fit our data using the blass et al.linelist scaled to fit our empirical @xmath33 at 175 k , along with the partition sum parameterization of @xcite , to represent propane .
we fit model to data with an ` amoeba'-type minimization while varying six parameters : total haze opacity ( @xmath35 ) , propane ( @xmath36 ) , acetylene ( @xmath27 , @xmath30 ) , and hcn ( @xmath26 , @xmath29 ) .
figure [ fig3 ] compares the observed spectrum with the best - fit model with and without propane , in one case using the blass et al .
linelist and in the other using our empirical @xmath33 to represent propane .
the detection of propane is unambiguous and the best - fit abundance assuming a constant vertical profile and using our empirical @xmath33 is @xmath37 . for acetylene and hcn
our best fits are characterized by @xmath38 and @xmath39 .
the haze is best fit by @xmath40 , although this value is extremely dependent on the assumed vertical haze profile and should not be over interpreted .
we found that reasonable fits to the hcn and @xmath17 features required a non - constant vertical abundance .
even with our two variable log - log parameterization we see non - zero residuals on these features , suggesting a more complicated vertical structure which we will explore further in future work with similar spectra covering many more transitions .
other significant non - zero residuals are explained as : an unidentified species at 747.35 @xmath4 , an imperfectly corrected telluric feature at 748.68 @xmath4 , not fitting for the @xmath41cch@xmath9/c@xmath9h@xmath9 ratio at 748.88 @xmath4 , and an h@xmath41cn feature at 749.02 @xmath4 that is missing from the model .
our focus here is propane and thus we do not present uncertainties on the vertical abundance parameters for hcn and @xmath17 . as is apparent in fig .
[ fig1 ] , random errors are negligible in our spectra .
the other sources of uncertainty are : flux calibration , errors associated with our empirical absorption coefficient method , the extent to which propane abundance is not constant with altitude , the extent to which titan s temperature - pressure profile deviates from that derived by @xcite , and the extent to which propane s abundance varies with latitude .
the uncertainty in flux calibration is at worst 20% , which maps almost linearly into a 20% uncertainty in propane abundance .
the primary potential error from our empirical absorption coefficient method is the mismatch of laboratory sample temperature ( 175 k ) to the temperature of the propane in titan s stratosphere . in our model spectra @xmath42% of the propane flux
arises from the pressure range 13 - 0.24 mbar ( 90 - 250 km altitude ) , where temperatures range over 135 - 175 k. to investigate this uncertainty we re - fit using the blass et al .
linelist ( see fig .
[ fig3 ] . ) .
the overall fit is lower quality than with our empirical @xmath33 and the best - fit propane abundance is only [email protected] times more .
recent photochemical models ( e.g. * ? ? ?
* ; * ? ? ?
* ) predict that the propane abundance is relatively constant over 400 - 850 km ( 11 - 1.5 @xmath44bar ) . above these altitudes propane
is depleted by photolysis and reaction with c@xmath9h , while below this range the effect of eddy diffusion and the cold - trap of the tropopause is to reduce the propane abundance .
our observations are primarily sensitive to the pressure range 13 - 0.24 mbar ( 90 - 250 km altitude ) , as shown in fig .
we re - fit our spectra using the model predicted propane vertical profile and found a best fit when the model predicted profile is multipled by a factor of 2.9 .
figure [ fig4]b shows this best - fit scaled profile of @xcite along with the best - fit constant vertical abundance profile .
the quality of the fits to our data do not distinguish between these two cases .
the uncertainty in the @xcite temperature - pressure profile is not well quantified , and therefore our reported uncertainties do not account for this source of error .
roughly , a change in stratospheric temperature of @xmath435 k requires a @xmath4520@xmath12 change in propane abundance .
there is no strong evidence that propane abundance is constant with latitude on titan . a first attempt at modeling the seasonal variation of photochemistry on titan @xcite predicts that propane abundance will vary by a factor of a few from north to south depending on season . with texes on an 8 - 10 meter telescope we would be able to measure the abundance as a function of latitude , which will place new strong constraints on models of seasonal photochemistry .
we present the first spectrally resolved detection of propane in titan s atmosphere .
we measure propane s fractional abundance to be @xmath5 , assuming the recommended temperature - pressure profile of @xcite and a constant abundance with altitude and latitude .
our observations are primarily sensitive to 13 - 0.24 mbar ( 90 - 250 km altitude ) .
alternatively , our data require the predicted propane profile of @xcite to be increased by a factor of 2.9 .
the current theoretical understanding of propane s vibration - rotation spectrum is inadequate and available linelists do not fit our titan spectra nor laboratory propane spectra well .
therefore we fit for titan s propane abundance using an empirical absorption coefficient spectrum derived from low temperature laboratory spectra of propane .
the most important observational advance that can be made towards detecting new species in planetary atmospheres is to increase the spectral resolving power to @xmath46few@xmath47 such that individual lines are resolved and separated .
high spectral resolution allows us to separate the contribution of a minor species ( propane ) from a strongly emitting species ( acetylene ) .
we thank several people who shared their time and advice , including especially l. lara , b. blass , g. bjoraker , r. gamache , v. typke , and the referee .
observations with texes are supported by nsf grant ast-0205518 .
tkg and mjr were supported by the sofia project through usra grant 8500 - 98 - 005 . | we present the first observations of propane ( c@xmath0h@xmath1 ) on titan that unambiguously resolve propane features from other numerous stratospheric emissions .
this is accomplished using a @xmath2 spectrometer ( texes ) to observe propane s @xmath3 rotation - vibration band near 748 @xmath4 .
we find a best - fit fractional abundance of propane in titan s stratosphere of @xmath5 in the altitude range to which we are sensitive ( 90 - 250 km or 13 - 0.24 mbar ) . |
the international linear collider ( ilc ) will be a tev - scale lepton collider that will require non - invasive beam size monitors with micron and sub - micron resolution for beam phase space optimisation @xcite .
laser - wire monitors operate by focussing a laser to a small spot size that can be scanned across the electron beam , producing compton - scattered photons ( and degraded electrons ) .
these photons can then be detected further downstream using the total energy observed as a function of the laser spot position to infer the transverse profile of the electron bunch .
the laser - wire system installed in the petra ring is part of an ongoing effort in the r&d of producing a feasible non - invasive beam size diagnostic tool .
the petra accelerator was chosen for the installation of the laser - wire experiment because it is capable of producing bunch patterns similar to the ilc .
laser - wire tests are run using a @xmath0 positron beam with a single bunch with a charge of typically @xmath1 . from the optics lattice
the average beam size is @xmath2 for the horizontal and @xmath3 for the vertical dimension .
preliminary simulations showed that the compton - scattered photons loose the majority of their energy in the material of the dipole magnet s beampipe due to hitting the wall with a shallow angle , resulting in an effective length of @xmath4 of aluminium .
an exit window was therefore designed and installed ( by desy ) to allow these photons to reach the detector with little deterioration ( see fig . [ fig - petra_window ] ) .
the laser pulses are created in a q - switched nd : yag laser operating at @xmath5 .
the pulses are then transported via a matched gaussian relay made up of two lenses over a distance of @xmath6 from the laser hut via an access pipe into the tunnel housing the accelerator .
the laser beam is then reflected off the scanning mirror before it reaches a focusing lens with @xmath7 back - focal length .
the scanner is a piezo - driven platform with an attached high - reflectivity mirror which has a maximum scan range of @xmath8 .
the peak power at the laser exit was measured to be @xmath9 . at the ip
the peak power is reduced to @xmath10 as higher order modes carry some fraction of the beam power but these are focussed out of beam transport , which is only matched for the fundamental mode .
the longitudinal profile was measured using a streak camera with @xmath11 time resolution .
the data revealed a pulse length of @xmath12 fwhm with a sub - structure of roughly @xmath13 peak - to - peak and @xmath13 peak width at full contrast due to mode - beating .
this causes the compton signal amplitude to vary between zero and full signal for different laser shots . in order to reduce the data taking time
the current laser will be replaced with an injection seeded system enabling faster data taking .
the laser - wire set up makes use of a calorimeter composed of 9 lead tungstate ( @xmath14 ) crystals arranged in a @xmath15 matrix fixed with optical grease to a square faced photomultiplier .
the individual crystals have dimensions of @xmath16 .
the complete detector set up was tested with a testbeam from the desy ii accelerator using electrons from @xmath17 to @xmath18 .
energy resolution was shown to be better than @xmath19 for individual crystals and @xmath20 for the overall set up .
simulations show that for the @xmath15 matrix , @xmath21 of the total energy deposit is collected for an incoming compton - scattered photon with @xmath22 energy @xcite .
the laser - wire daq system has two main components : the hardware trigger which synchronises the laser and daq components to the electron ( positron ) bunch , and the software which controls the acquisition and collation of data from each sub - component of the system .
the hardware trigger operates with two inputs from the petra integrated timing system ( pit ) and produces the necessary signals to fire the laser .
the trigger card also produces a signal to trigger the ccd cameras and a signal to start the software data acquisition . when the signal from the trigger card is received a counter which runs for approximately @xmath23
is started .
after this time a signal is sent to the integrator card , lasting around @xmath24 , to integrate the output from the calorimeter .
the integrated signal is read by an adc channel .
the daq software also produces a programmable signal , up to a peak of @xmath25 , which is amplified by a factor of 10 and this is used to drive the piezo - electric scanner .
a scaled version of the scanner amplifier output is read by an adc channel .
the other sub - components of the daq system : the bpm monitor , the petra data monitor and the ccd cameras are also read out .
communication with each component is performed by a messaging system using tcp / ip .
in order to determine the transverse size of the electron beam , it is necessary to know the properties of the laser that is being used to scan .
particular attention is paid to the spot size at the laser waist , @xmath26 , and the rayleigh range , @xmath27 , ( the distance from the waist at which the beam size @xmath28 ) .
these properties are related by eq .
[ equ - laserprofile ] : @xmath29 where @xmath30 .
+ the laser is focused using the same final focus lens as described previously .
a cmos camera is placed on a track rail so that it can be moved through the focal plane parallel to the beam direction . due to the high power of the laser ,
the beam was first passed through a @xmath31 reflective mirror , and then through a variable amount of neutral density filter in order to prevent saturation and damage to the camera pixels .
the camera was moved along the track rail to a number of positions , and 100 images were taken in each location . the images taken by the camera are stored as 8-bit greyscale bitmap files .
the pixel data is projected onto the y - axis , and fitted to a gaussian on a linear background in the region around the signal peak .
the width at each location is then plotted , and fitted to eq .
[ equ - laserprofile ] . from this
we obtain @xmath32 , which is within the expected range , and @xmath33 , as shown in fig .
[ fig - laserwaist ] .
the laser is scanned across the electron beam by tilting a mirror on a piezo - electric stack to produce a deflection of @xmath34 .
focusing through the lens produces a travel range for the focal spot at the ip of @xmath35 .
the scanner voltage is applied in a stepped sinusoidal pattern ; @xmath36 triggers are taken at each of @xmath37 voltages over a whole @xmath38 .
the trigger signal is taken from the laser trigger card running at @xmath39 , so a full scan takes approximately @xmath40 . the signal from the adc
is expected to display two peaks ; one as the laser crosses the electron beam on a rising voltage to the scanner , and one on a falling voltage .
the trigger number exactly half way between the peaks should correspond to a turning point in the scanner position .
the mean of the background subracted adc counts at each voltage is then fitted to a gaussian whose width , @xmath41 is given by @xmath42 .
[ fig - scanplots ] shows the typical results observed for a single scan and the results of multi - scan shifts are presented in table [ tbl - eresults ] . note that the large signal variation in fig .
[ fig - scanplots]a is partly due to the sub - structure of each laser pulse and will be removed by a better laser .
.,width=245 ] .data run results for extracted electron beam size , @xmath43 .
the errors are the rms from several scans [ cols="^,^,>",options="header " , ]
the entire petra laser - wire set up has been simulated using bdsim @xcite , which is a fast tracking code utilising the geant4 @xcite physics processes and framework .
the simulation is a full model of the accelerator components including beampipe , magnets , and cooling water channel . for each simulated event
a compton scattered photon is generated with an energy based upon the compton spectrum predicted for the petra laser - wire parameters .
this photon is tracked to the detector whilst fully simulating any interactions with materials such as the beampipe wall .
this process is repeated to create an effective single compton energy distribution and its corresponding distribution at the detector after passing through any matter along the photon path .
the single photon distribution in the detector is extrapolated to the @xmath44 spectrum using poisson statistics whilst also accounting for the energy resolution of the calorimeter and the longitudinal sub - structure of a typical laser pulse .
the simulated spectrum is compared directly to the experimental data ( see fig . [ fig - sim ] ) , where the laser and electron beam were well aligned .
the expected number of compton - scattered photons , @xmath44 , per shot with the laser - wire setup parameters is approximately @xmath45 photons , which agrees with the theoretical value .
the experimental data show an energy resolution of 34% which is dominated by the longitudinal fluctuations in the laser power .
the simulation models these fluctuations using relatively old streak camera data as described above and so does not account for degradation in the quality of the laser since then .
the calorimeter has also not been calibrated for the range of energy deposits now incident upon it and has been in the petra radiation environment for three years .
this could explain why the simulation fails to completely model the experimental data in the lower energy region .
the future strategy for the laser - wire project can be characterised in the short term to concentrate on non - laser issues like data acquisition , signal detection , vertical scanning , and implementation into a linac beamline .
this aims at the development of a standard diagnostic tool to be placed at many locations along the accelerator beamline . in the long run r&d work
is planned to develop a laser system producing pulses matching the ilc micro pulse structure . here
the target is to have a beamsize monitor with full flexibilty . to meet the short - term targets it is planned to purchase an injection seeded q - switch laser with second harmonic generation having excellent longitudinal and transverse mode quality .
a complimentary project concentrating on the achievement of micron - scale laser spot - sizes is underway at the accelerator test facility ( atf ) at kek . | the laser - wire will be an essential diagnostic tool at the international linear collider .
it uses a finely focussed laser beam to measure the transverse profile of electron bunches by detecting the compton - scattered photons ( or degraded electrons ) downstream of where the laser beam intersects the electron beam .
such a system has been installed at the petra storage ring at desy , which uses a piezo - driven mirror to scan the laser - light across the electron beam .
latest results of experimental data taking are presented and compared to detailed simulations using the geant4 based program bdsim . |
interference effects have been found to play an important role in the dynamics of qubits @xcite . for their understanding , the interference phenomena can be considered from different points of view : as interference between successive landau - zener ( lz ) tunneling events @xcite , analogous to mach - zehnder interference @xcite , or in terms of spin dynamics @xcite . in charge - phase qubits ,
strongly coupled to a microwave resonator , analogy with vibrational transitions in diatomic molecules has been pointed out @xcite .
the fact that the latter phenomena are related to interference effects has its origin in the phase - space dynamics @xcite .
the state of a superconducting charge - phase qubit or a cooper - pair box ( cpb ) can be monitored continuously by measuring its reactive response , either as an effective capacitance or inductance , which both are due to the curvature of the qubit s energy bands with respect to charge or phase degrees of freedom .
when such a system is made as a part of an electric @xmath4-oscillator circuit , any change of reactance , caused by the evolution of occupancies of the qubit s energy levels , will shift the resonance frequency of the resonator .
this scheme has been used in several recent experiments , typically having the qubit , with splitting @xmath5 , connected to a coplanar stripline cavity near resonance @xcite .
we have , however , worked in the fully detuned limit where @xmath6 . at weak drive , such a read - out scheme perturbs the investigated system only weakly , allowing studies of inherent interference phenomena of the periodically driven two - level system .
we have investigated interference effects in the cooper - pair box and in a charge - phase qubit circuit , configured as dual to the cpb ( i.e. , connected to the resonator via the phase , rather than the charge port ) , see fig .
[ fig : setup ] .
both circuits provide good model systems of interference effects in periodically driven qubits . in the latter case
, a clearly stronger coupling between the qubit and the resonator could be achieved , which made a difference in the observed effects .
the interference effects in these two circuits are diverse mostly due to the difference in the coupling between the resonator and the qubit . on the whole , interference phenomena can be understood in terms of multiphoton transitions . in strongly driven qubits ,
clear multiphoton transitions have been observed when the energy quantum of the rf - drive , multiplied by a small integer , matches the qubit level splitting @xcite .
in fact , multiphoton transitions allow a clear - cut distinction between our main findings .
we may classify our observations as transitions where : 1 ) several quanta from the drive at @xmath2 and 2 ) one quantum at @xmath2 and several at @xmath3 are exchanged between the qubit and its surrounding entities , i.e. the classical rf - drive port and the @xmath4 resonator .
the first class can be treated as a sequence of phase - sensitive landau - zener transitions which produce a characteristic interference pattern , dependent on the `` finesse '' of the interferometer , namely the decoherence of the system .
the latter class of transitions can be described in terms of vibronic transitions in diatomic molecules where potential changes are fast compared to the vibrational frequencies and non - adiabatic coupling between levels differing by a large number of quanta becomes possible .
these phenomena include creation of several resonator quanta from the ground state , which is the non - adiabatic behavior referred to as the dynamical casimir effect @xcite . of the tank circuit .
this change causes a shift in phase of the reflected , constant - frequency microwave signal as illustrated in c ) . in the illustration
, @xmath7 grows which leads to a reduction of @xmath8 and , consequently , to a decrease of the reflection phase arg(@xmath9 ) .
[ fig : setup],width=377 ] in this paper , we summarize some of our recent results @xcite on interference effects in a driven superconducting charge qubit coupled to a resonator . the original references , in particular , used theoretical considerations and numerical simulations of the bloch equations and the linear - response approximation for a detailed analysis of the shape of interference fringes when dissipation is important . instead
, here we try to concentrate on robust features which are amenable to general conclusions .
for example , part of our analysis will be dealing with the phenomenon of destruction of tunneling .
the paper is organized as follows : first , we start with basics of the cooper - pair box in sect . 2 and describe the dynamics of a two - level system in sect .
[ s.l ] , paying special attention to the destruction of tunneling from the ground state to the excited state .
[ s.m ] covers our experimental work and results on cpbs , and it includes some considerations aiming towards understanding of the stokes phase and its relation to geometric phases in a spin-1/2 system .
we present data at phase bias @xmath10 ( minimum band gap ) , and we analyze , especially , data in the regime of destruction of tunneling at the charge degeneracy
. results at zero phase bias can be found in ref .
section 5 describes the analogy of spin flips in strongly coupled qubits with vibronic transitions in diatomic molecules .
we do not describe much our experimental techniques , but to make the paper self - contained , we have explained some of the main technical points when presenting the experimental results .
a split cooper - pair box ( cpb ) is formed by a single - cooper - pair transistor ( scpt ) embedded into a small superconducting loop @xcite , see fig .
[ fig : setup ] . the charging energy of the cpb , @xmath11 kelvin ,
is given by the total capacitance @xmath12 which includes the junction capacitances , the gate capacitance @xmath13 , and the self - capacitance of the island .
the effective josephson energy is given by the sum of the energies of the individual junctions @xmath14 , which is tunable by magnetic flux @xmath15 , _
i.e. _ by the superconducting phase across the two junctions , @xmath16 . here , @xmath17 is the superconducting magnetic flux quantum .
we may write the hamiltonian in the charge basis as @xmath18.\end{aligned}\ ] ] here @xmath19 denotes the number of extra electron charges on the island , and @xmath20 is the charge in electron units induced by the gate voltage @xmath21 on the gate capacitor with capacitance @xmath13 .
@xmath19 is conjugate to @xmath22 , where @xmath23 is the superconducting phase on the island .
the asymmetry of the two josephson junctions of the cpb is described by @xmath24 . by assuming that @xmath25 and that @xmath26 is close to 1 , one can reduce the circuit to a two - state system , with the hamiltonian @xmath27 \\
-e_{\rm j}[\cos{(\phi/2)}-id\sin{(\phi/2 ) } ] & 4e_{\rm c}(1-n_{\rm g } ) \end{array}\right)\ ] ] in the basis of the relevant charge states @xmath28 and @xmath29 . when this energy operator is expressed as a linear combination of the pauli matrices @xmath30
+ @xmath31 we get the magnetic field components as @xmath32 , @xmath33 and @xmath34
. the energies of the two states as a function of @xmath26 are illustrated in fig .
[ f.lzfig ] .
( in units of @xmath35 ) . in sweeping @xmath26 ( horizontal arrows )
, landau - zener tunneling between the bands takes place close to the point where the energy difference is at minimum ( curved arrows ) .
successive sweeps can lead to interference of the tunneling events .
the interference depends on the phase ( [ dynamic ] ) accumulated in the shaded area .
[ f.lzfig],width=188 ] in the experiments @xmath36 k and @xmath37 , which means that the magnitude for the off - diagonal components @xmath38 ranged over 1.46.2 ghz @xcite .
the non - linear parametric capacitance , which has been the cornerstone of our interference studies , has been experimentally studied in refs .
we study the effect of sweeping the gate charge @xmath26 . we denote the energy eigenstates by @xmath39 and @xmath40 , and suppose the system is initially in the lower energy state @xmath39 , see fig .
[ f.lzfig ] .
two cases can be distinguished .
when the sweep rate is small compared to the energy level difference , the system stays in the lower state .
the alternative case is that the sweep rate is comparable to the level spacing .
this may take place close to the degeneracy points , where the energy bands would cross in the absence of the josephson coupling .
there the system can tunnel from the lower state to the upper in a process known as landau - zener tunneling .
the tunneling probability in a single sweep is given by @xcite @xmath41 here @xmath42 is the speed at which the sweep passes the crossing point @xmath43 and @xmath44 is the energy difference in the absence of the josephson coupling . in eq .
( [ eq : plz ] ) the adiabaticity parameter @xmath45 defines whether the development is adiabatic ( @xmath46 ) or sudden ( @xmath47 ) .
when the system is coherent , and the degeneracy point is crossed several times , the transition amplitudes for each subsequent pass have to be added for evaluating the transition probability .
these amplitudes may be tracked by the ` scattering ' matrix @xmath48 , defined by ( cf .
@xcite ) : @xmath49 where phase factors are chosen to simplify the matrix . here
the scattering phase is @xmath50 , where the stokes phase @xmath51 depends on the adiabaticity parameter @xmath45 : @xmath52 in the adiabatic limit , @xmath53 , while in the sudden limit , @xmath54 .
away from the degeneracy point , the eigenstates @xmath39 and @xmath40 accumulate the relative dynamical phase @xmath55 dt\ , .\ ] ] thus , the condition for constructive interference is that @xmath56 and @xmath57 are multiples of @xmath58 , where @xmath59 and @xmath60 refer to the dynamical phases accumulated on the left and right sides of the degeneracy point , respectively .
for example , in the adiabatic limit , @xmath61 have to be odd multiples of @xmath10 .
the lz interference can also be employed to suppress the tunneling to the upper level @xcite .
in fact , the contrast for destructive interference in the experimental data looks often better than for constructive interference , and these conditions can be employed more simply to determine the behavior of the system as the ground - state response in the measurement is well known . in this case the interference conditions are @xmath62 with an integer @xmath63 . in the sudden limit , @xmath47 , the periodically driven two - level system can be solved in a rather straightforward manner @xcite . at degeneracy ,
the probability of being in the upper state varies periodically with time according to @xmath64\,,\ ] ] where @xmath65 denotes the zeroth bessel function , and @xmath66 ( @xmath67 below ) is the drive amplitude at frequency @xmath68 .
one notices that the destruction of tunneling takes place , when the ratio @xmath69 coincides with a zero of @xmath65 .
the description of interference patterns in a cooper - pair box using the scattering matrix formalism has been discussed in refs .
@xcite . in general , relaxation phenomena
should be included as they have strong influence on the sharpness of the interference fringes .
for this reason , our main analysis method has been based on spin-1/2 nmr simulations @xcite , which also includes linear response calculations in order to obtain the measured , effective capacitance .
alternatively , the dressed - state approach may be employed as has been done by c. wilson _
we have performed low - dissipation microwave reflection measurements @xcite on a series @xmath4 resonator in which the box effective capacitance , @xmath70 is a part of the total capacitance @xmath71 , where the superscript @xmath72 refers to ground and exited states of the qubit . ,
@xcite ] the resonator is formed by a surface mount inductor of @xmath73 nh . with a stray capacitance of @xmath74 ff due to the fairly big lumped resonator , the resonant frequency is @xmath75 mhz , and the quality factor @xmath76 is limited by the external @xmath77 . when @xmath78 varies , the phase and amplitude of the reflected signal @xmath79 change , which is measured by the reflection coefficient @xmath80 . here , @xmath81 is the resonator impedance as seen from the end of the @xmath82 coaxial cable used for the reflection measurement .
the variation in @xmath83 due to modulation in @xmath78 is up to 40@xmath84 in our measurements , corresponding to a shift of resonance frequency @xmath85 mhz . in all the measurements ,
the weak probing signal @xmath86 at frequency @xmath87 was continuously applied , in addition to the dc - bias and the rf - drive .
thus , the total gate charge variation , in units of @xmath35 , can be written as @xmath88 , where the @xmath89 , @xmath90 , and @xmath91 terms correspond to the dc , rf , and measurement drives , respectively .
ghz using phase bias @xmath92 , corresponding to the level repulsion of @xmath93 ghz .
@xmath94 specifies the amplitude of the rf drive in electrons and @xmath95 is the dc charge bias .
the vertical bars give the color scales in degrees and db for the left and right frames , respectively.[fig : piflux],width=453 ] we have made extensive scans of the cpb reflection by varying the lz drive frequency @xmath9620 ghz and its amplitude @xmath973 electrons , as well as the qubit dc - bias : @xmath95 and @xmath98 . in fig .
[ fig : piflux ] we present the reflection phase @xmath99 and magnitude @xmath100 measured at @xmath101 ghz on the plane spanned by @xmath94 and @xmath95 .
the josephson capacitance @xmath102 deduced from the data of fig .
[ fig : piflux ] has been given in ref .
we observe in fig .
[ fig : piflux ] a clear interference pattern whose main features confirm the coherent lz tunneling picture : 1 ) there is an onset of the interference speckles , where the rf drive just reaches the avoided crossing ( charge degeneracy ) , @xmath103 odd integer , with a linear dependence between @xmath95 and the ac drive amplitude .
additionally we have shown in ref .
@xcite that 2 ) the density of the dots is proportional to @xmath104 in the direction of @xmath105 as well as @xmath94 , and 3 ) the pattern loses its contrast below a certain drive frequency , in this circuit around @xmath106 ghz . in fig .
[ fig : piflux]a at charge degeneracy , we observe a clear sequence of blue dots that signifies a similar response as in the ground state .
these dots correspond to the destruction of tunneling due to periodic rf - drive of our two level system .
note also that these locations correspond to a large reflection magnitude in fig .
[ fig : piflux]b .
in fact , the reflection is enhanced compared with the undriven level , which indicates transfer of energy from the microwave drive , via the qubit , to the resonator .
apparently , the destruction of tunneling prevents the microwave energy from being deposited to the two - level system , and the ` extra ' energy is dumped out at both drive frequencies . the order number of the destructive interference dots in fig .
[ fig : piflux ] is plotted in fig .
[ fig : stokes ] . at large rf - drives ,
the interference dots display a linear dependence on the drive amplitude .
the linear dependence is a sign of asymptotic behavior that can be obtained either from eq .
( [ phasecondition ] ) or from eq .
( [ bessel ] ) . starting from eq .
( [ phasecondition ] ) , we may evaluate @xmath107 under a drive of @xmath108 .
this yields @xmath109 where @xmath110 indexes the fringe number , and @xmath111 , where @xmath112 are the integers from eq .
( [ phasecondition ] ) .
thus , we find @xmath113\ ] ] ( note that the order of resonance grows with @xmath114 , and eq .
( [ fit ] ) holds , of course , only for sufficiently large @xmath114 , when it gives @xmath115 ) . for the destructive interference at charge degeneracy in the sudden limit this yields @xmath116 . by fitting eq .
( [ fit ] ) to three / four highest drive points of the data in fig .
[ fig : stokes ] we get @xmath117 , which is close to the theoretical expectation of 0.25 . however , there is always an additional contribution from the dynamical phase picked up near the degeneracy point where the bands are curving from linear .
this contribution was estimated to be about 1/3 from the expected stokes phase .
this correction will drop the vertical offset , and reduce the stokes phase down to @xmath118 .
hence , taking all the uncertainties into account , our final estimate for the stokes phase becomes @xmath119 in the sudden limit .
a fit using the plain dynamic phase of eq .
( [ dynamic ] ) , and the two level approximation for the adiabatic energy levels , is given by the dashed curve in fig .
[ fig : stokes ] .
the @xmath120-scale of the calculated curve has been adjusted to match the experimental results at large drive amplitudes .
the solid curve , on the other hand , is the dynamical phase corrected using the calculated stokes phase in the sudden limit .
the curve agrees with data except at the lowest points where a possible inaccuracy in the asymmetry parameter has its strongest influence .
the open circles display the result of eq .
( [ bessel ] ) , _ i.e. _ the zeroes of @xmath65 , which are also seen to coincide well with the data . ) as a function of the microwave drive amplitude @xmath94 for phase bias @xmath121 . for comparison , the dynamical phase calculated from eq .
( [ dynamic ] ) is given by the solid and dashed curves with and without stokes phase , respectively .
the zeroes of bessel function @xmath65 are given by the open circles .
for details , see text.[fig : stokes ] , width=302 ] the stokes phase is related to the non - adiabatic geometric phase , the aharonov - anandan phase @xcite .
it is rather straightforward to show that @xmath122 from this relation one gets at the charge degeneracy ( @xmath123 ) : @xmath124 , which is nearly the same relation as what we derived for the asymptotic fit in fig .
[ fig : stokes ] , cf .
( [ fit ] ) .
consequently , the geometric aharonov - anandan phase could be approximately determined by taking the measured drive amplitude at the point of destructive interference and multiplying the result by @xmath125 . for the data in fig .
[ fig : stokes ] at 4 ghz , we get a nearly constant value ( 0.200.21 ) @xmath126 at drive amplitudes around @xmath1271 .
in the inductive set , lset , of fig . [ fig : setup]b , a stronger coupling between the qubit and the resonator can be achieved than in the cooper - pair - box configuration .
this leads to more involved multiphonon phenomena which are reminiscent of the transitions in diatomic molecules .
these transitions can also be viewed as interference effects , and it is convenient and instructive to describe them as interference in the phase space @xcite . in order to understand the molecular analogy , we consider the single - cooper - pair transistor as a two - state system .
the two states are the analog of two electronic states in a diatomic molecule .
the transistor is coupled in parallel with an @xmath4 oscillator .
the @xmath4 oscillator is the analog of nuclear vibrations in the molecule . because of the coupling between the two - state system and the oscillator ,
a transition between the electronic levels is often accompanied by a change of the vibrational state .
such transitions , where both vibrational and electronic quantum numbers change simultaneously , are known as _
vibronic_. for quantitative analysis , we write the hamiltonian in the basis of two relevant charge states of the qubit : @xmath128 here the flux @xmath129 and the charge @xmath130 in the @xmath4 oscillator are conjugate variables , and @xmath131 are the pauli matrices
. the capacitive energy @xmath132 can be controlled by the gate voltage .
another control parameter is the flux bias @xmath133 through the loop containing scpt and the inductor . in the following we concentrate on the limit , where the oscillator frequency @xmath134 is much lower than the qubit level difference
this means that all changes in the qubit system are much faster than in the oscillator system .
therefore , one can diagonalize the qubit part separately from the oscillator .
the two energies of the qubit are obtained by diagonalizing the @xmath135 matrix part of eq .
( [ e.hscowei ] ) : @xmath136 this form is valid when the josephson coupling is a small perturbation compared to the capacitive energy @xmath137 .
it is also possible to consider the general case of an arbitrary ratio of the josephson coupling to @xmath137 , in which case the energies of the two lowest states are given by mathieu characteristics .
this means that the dependence of the eigenenergies on @xmath129 would be more complicated than in eq .
( [ e.equbit ] ) but otherwise the following analysis remains intact .
considering now the oscillator , we can think about the capacitive and inductive terms in the hamiltonian as the kinetic and potential energies .
in addition to the inductive potential , there is a potential arising from the qubit , since the qubit energy ( [ e.equbit ] ) depends on the flux @xmath129 .
thus there are two potential curves for the oscillations @xmath138 such a potential is illustrated in fig .
[ f.oscpot]a .
the different slopes of the qubit energies give rise to a relative shift of the minima of the oscillator potentials .
the difference in the curvatures of the qubit energies shifts the vibrational frequencies .
figure [ f.oscpot]a , which is drawn to scale with realistic parameters of our circuit , shows that vibronic transitions can be induced by microwave radiation .
( solid curves ) for vibrations correspond to two qubit states @xmath72 .
the qubit energy ( [ e.equbit ] ) is given by dashed lines .
the energies of six lowest vibrational states corresponding to both qubit states are drawn .
the bars represent energy shifts induced by high and low frequency drives .
( b ) the overlap of @xmath139 states with @xmath140 vibrational quanta with state @xmath141 in the phase space .
the dots show the centers of the ellipses corresponding to qubit states @xmath72 , which are displaced from each other by @xmath142 . ]
the intensities of vibronic transitions are traditionally analyzed in terms of the franck - condon principle , and this analysis turns out to be useful for understanding of our data . in its classical form , the franck - condon principle says that transitions are possible between vibrational states , the trajectories of which intersect in phase space , and they are most intense between states with coincident turning points .
this is based on the idea , already stated , that the transitions between electronic states are fast and the vibrations are slow , so that the vibrational coordinates , here the flux @xmath129 and the charge @xmath130 , have no time to change during the transition .
in addition , @xmath143 is slowest near the turning points , and therefore transitions there are most likely . in the following
we discuss how this picture can be extended by semiclassical analysis in phase space , as shown in refs .
@xcite , and how this interpretation is related to the data obtained .
the phase space formed by the coordinate @xmath129 and the canonical momentum @xmath130 is illustrated in fig .
[ f.oscpot]b .
the vibrational states correspond to elliptic rings . by scaling the coordinates properly , a state with @xmath144 vibrational quanta in a harmonic oscillator
can be associated with a circular ring between radii @xmath145 and @xmath146 ( planck - bohr - sommerfeld bands @xcite ) .
this is the case for the upper qubit state in fig .
[ f.oscpot]b . due to a different resonance frequency
, the vibrational states corresponding to the lower qubit state then appear as slightly squeezed in the @xmath130 direction .
the shift of the minima of the potentials @xmath147 ( fig .
[ f.oscpot]a ) appears in the phase space so that the ellipses corresponding to the states @xmath139 , where @xmath148 means the upper qubit state , have different centers than the ellipses corresponding to the states @xmath149 .
vibronic transitions can occur only between states whose ellipses overlap in the phase space . in the example of fig.[f.oscpot]b
, the state @xmath141 has overlap only with states @xmath139 with @xmath150 . in general , the transition rates are determined by the area of the intersection of the respective bands .
this area is largest when the ellipses touch each other tangentially , i.e. , when the turning points coincide , and this leads to the classical version of the franck - condon principle . in the present case
though , the overlap consists of two crossings of the ellipses , and one has to add the transition amplitudes taking into account the relative phase accumulated between the crossings .
the matrix element of the transitions , @xmath151 , is given by the formula @xcite @xmath152 here @xmath153 is the area of one of the crossing of the elliptic rings , which sets the maximum probability of the transitions .
the phase in the phase factor is determined by the area @xmath154 between the two alternative paths between the crossing points .
( [ e.wheelerf ] ) reproduces well the results of full quantum calculation except near the classical turning points ( where the two crossing areas merge ) . in the case of fig .
[ f.oscpot ] the oscillations are nearly harmonic and the relative difference of frequencies is small .
when , in addition , the difference in the vibrational quanta is small , @xmath155 , the probability @xmath156 scales as @xmath157 , where @xmath158 is the dimensionless separation of the potential minima .
owing to the parabolic bands in the region near the minimum gap , @xmath159 , when @xmath160 .
the measurements on the artificial molecule are made by analyzing the reflection of microwaves at a frequency @xmath161 close to the resonance frequency @xmath3 of the circuit . in order to induce vibronic transitions , a microwave excitation at frequency @xmath162
the resonance condition for vibronic transition between states @xmath163 and @xmath164 is @xmath165 where @xmath5 is the qubit energy splitting . a measurement of reflection coefficient @xmath9 in the bias plane @xmath166 is shown in fig .
[ fig : fringes ] . .
they correspond to @xmath167 , -1 , and -2 and are denoted by @xmath168 , @xmath169 , and @xmath170 , respectively .
the fringe behavior is fitted to bessel functions @xmath171 , @xmath172 , and @xmath173 . for details , see text . [
fig : bessel ] ] .
they correspond to @xmath167 , -1 , and -2 and are denoted by @xmath168 , @xmath169 , and @xmath170 , respectively .
the fringe behavior is fitted to bessel functions @xmath171 , @xmath172 , and @xmath173 . for details , see text . [
fig : bessel ] ] the vibronic transitions are seen as fringes , making half - circles in the bias plane . the fringes are located within a crescent - shaped area .
the language of interference in phase space allows one to easily understand , for instance , the shape of this area , the positions of the fringes and the modulation along the fringes . at @xmath174
only the pure electronic transition is seen .
this agrees with the analysis above since there the displacement of the oscillation minima vanishes . with increasing @xmath133
more fringes appear , which is in agreement with the growing separation of the minima , and the range of the fringes at a given @xmath133 can be determined as in fig .
[ f.oscpot]b .
in addition , the above picture of enhanced probability of the extremal trajectories is reflected in the shape of the absorption pattern in fig .
[ fig : fringes ] , where the edges of the fan - like structure are rather clearly expressed .
the strength of the fringes in the reflection measurement depends on two separate factors .
one is the transition probability to the upper qubit state . in large part of the bias plane ( below the the v - shaped light region in fig .
[ fig : fringes ] ) the averaged frequency of being part time in the upper and part time in the lower qubit state matches better to the measuring frequency than being in the lower state only .
this leads to absorption which is approximately proportional to the transition rate to the upper qubit state .
the other factor is the heating or cooling effect caused by the high - frequency @xmath2 radiation on the vibrations . for a vibronic transition with @xmath175 ( i.e. @xmath176 ) , the resonance deposits energy into the resonator causing heating of the vibrations .
this leads to decreased absorption of the measurement wave . in the opposite case @xmath177
, the absorption of a high frequency photon cools the oscillator by reducing its quanta which leads to increased absorption of the measurement wave .
this asymmetry of the fringes is clearly visible in fig . [
fig : fringes ] , where the cooling fringes at larger radii are stronger than the heating fringes at smaller radii .
numerical simulations of the artificial molecule has been done in ref .
these are based on bloch equations describing the qubit and classical equations describing the circuit .
they reproduce well the observed vibronic spectrum , both the matrix elements and the heating / cooling asymmetry . here
we present a simplified analysis by comparing three of the fringes with the transition probability @xmath178 , neglecting the heating / cooling effect .
[ fig : bessel ] displays the measured reflection magnitude along fringes with @xmath167 , -1 , and -2 as a function of the flux bias of the qubit .
squared bessel functions @xmath171 , @xmath172 , and @xmath173 have been fitted to the data .
for the @xmath120-scale argument we have taken @xmath179 although this approximation will be rather crude when @xmath180 .
nevertheless , the comparison can be employed to look for the presence of basic interference phenomena .
the agreement between the data and the fits is quite good , though the modulation of the pure electronic fringe is a bit weaker in the measurement than given by @xmath171 .
the reason for smearing of the bessel modulation is the variation of the number of quanta in the resonator . assuming the resonator is in a coherent state , it is a superposition of the number states and this should be taken into account in the analysis of the interference fringes
this diminishes the sharpest features of the interference , but does not remove it .
consequently , we may conclude that the dip at @xmath181 in the fringe magnitude corresponding to the pure electronic transition , is due to destructive interference in phase space in formula eq .
( [ e.wheelerf ] ) . in summary ,
we have discussed interband transitions in superconducting cooper - pair boxes and charge - phase qubits and argued , how interference effects in various forms can be found to underlie the observed phenomena .
fruitful discussions with m. feigelman , t. heikkil , f. hekking , and m. paalanen are gratefully acknowledged . | we investigate phase - sensitive interference effects in a periodically @xmath0-driven , artificial two - state system connected to a microwave resonator at @xmath1 mhz .
we observe two kinds of multiphoton transitions in the two - state system , accompanied by : 1 ) several quanta from the drive at @xmath2 and 2 ) one quantum at @xmath2 and several at @xmath3 .
the former are described using phase - sensitive landau - zener transitions , while the latter are discussed in terms of vibronic transitions in diatomic molecules .
interference effects in the vibronic transitions governed by franck - condon coefficients are also considered . |
urban vegetation is receiving a significant amount of attention from researchers in recent years .
this interest stems from its impacts on the environment , affecting the pedestrian comfort and mitigating the negative health effects of the air pollution @xcite .
microscale modelling using the computational fluid dynamics ( cfd ) proved to be an indispensable tool for assessing the impacts of the vegetation in the urban settings .
some numerical studies focused only on the effects of the vegetation on the flow @xcite or on the thermal comfort of the pedestrian @xcite .
others investigated pollutant dispersion in the presence of the vegetation , but without taking the deposition of the pollutant into account @xcite . when including the deposition , authors opted both for simplified model with constant deposition velocity @xcite and complex models expressing the dependence of the deposition velocity on various parameters such as the particle size or the wind speed @xcite .
dispersion of the particles with a fixed size can be described by one scalar partial differential equation .
when the behaviour of the particle size distribution is of interest , straightforward approach - so called _ sectional _ approach - is to divide the size range into a number of discrete bins and then model the appropriate number of scalar pdes , i.e. one for each bin .
other option is to use the transport equation for the moments of the particle size distribution .
such approach can reduce the number of pdes to be solved , and therefore reduce the computational demands .
this class of methods , here referred to as the _ moment method _ , has been used for the simulation of the aerosol behaviour for a long time @xcite .
usage of the moment method in the air quality models is also widespread ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
adapting the deposition velocity models to the moment method framework is not straightforward , since the mathematical formulation of the moment method requires all terms in the equation to be in the form of the power law of the particle size .
@xcite simplified the problem by using the resistance model with brownian particle diffusivity and settling velocity averaged over the particle size range .
@xcite developed a deposition velocity model based on the model proposed by @xcite .
this model , however , only includes the processes of brownian diffusion , impaction and gravitational settling , and does not take into account the processes of interception and turbulent impaction , which play an important role in the dry deposition process @xcite .
this study aims to fix this shortcoming by adapting the model by @xcite for the use in the moment method .
the developed model is then used in the microscale cfd solver to solve the problems of a pollutant dispersion in the presence of a vegetation .
to the authors best knowledge , the moment method has not yet been applied to the microscale urban vegetation problems .
comparison of the obtained results with the results from the sectional model shows the applicability of such approach .
the governing equation for the transport and the deposition of the aerosol particles of a diameter @xmath0 in the flow field given by the velocity @xmath1 can be formulated as @xmath2 where @xmath3 is the number concentration of the particles @xcite .
diffusion coefficient @xmath4 is expressed as a fraction of the turbulent viscosity and the turbulent schmidt number .
effects of the gravitational acceleration @xmath5 are captured in the terminal settling velocity of a particle , given by the stokes equation , @xmath6 where @xmath7 is the density of the particle , @xmath8 is the dynamic viscosity of air , and @xmath9 is the cunningham correction factor @xcite .
the formula used for the correction factor is discussed in section [ sec : brown - dif ] .
the removal of the particles via dry deposition is modelled by a product of three parameters : _ leaf area density _ ( lad ) , defined as a leaf area per unit volume ( ) , deposition velocity @xmath10 ( ) measuring the filtration efficiency of the vegetation under given conditions , and the particle concentration @xcite .
its form is discussed in section [ sec : depvel ] .
the moment method is based on the idea that in order to model the size distribution of the particles , we can investigate the behaviour of the moments of the distribution .
moment of the distribution is defined as @xmath11 where @xmath12 is the order of the moment .
some moments have straightforward physical interpretation : @xmath13 is the total number concentration , @xmath14 is proportionate to the surface area concentration and @xmath15 is proportionate to the volume concentration . assuming @xmath3 is sufficiently smooth in space and time , moment equations are obtained by multiplying eq .
( [ eq : n - eq ] ) by @xmath16 , integrating over the whole size range and interchanging the derivatives and the integrals : @xmath17 now we are left with the evaluation of the integrals in ( [ eq : moment - eq ] ) .
this can be done easily if the multiplicative terms are in a form of a polynomial function of @xmath0 .
such is the case with the gravitational term , if we take into account that gravity plays significant role only for larger particles , where the cunningham correction factor @xmath9 in ( [ eq : us ] ) can be left out . using ( [ eq : us ] ) in the third term on the rhs of ( [ eq : moment - eq ] ) , the term can be rewritten as @xmath18 here we introduced a dependence on the moment of a higher order .
that necessitates that we either solve a separate moment equation also for this higher order moment , or that this moment can be calculated from the moments that we solve for .
the task of integrating the deposition term is more difficult and will be examined in the following section .
variety of models describing the rate of particle transport from the air to the vegetation surface has been proposed @xcite .
dry deposition schemes used in the large - scale air quality models are usually formulated in terms of a friction velocity or a wind speed at some height above the canopy , and do not explicitely model the behaviour inside the canopy .
as such , they are not directly applicable in the microscale cfd models , but the same principles governing the deposition still apply . in this study we adopted the deposition velocity expression given by @xcite for needle - like obstacles .
@xcite formulated the expressions for deposition velocities ( or collection velocities in the original terminology ) associated with each of the five main processes driving the dry deposition inside the canopy .
the model then assumes that these processes , which are the brownian diffusion , interception , impaction , turbulent impaction , and sedimentation , are acting in parallel and independently .
the deposition velocity thus can be written as a sum of the deposition velocities of all processes , @xmath19 contributions of each process to the deposition velocity for an exemplary set of parameters are shown on fig .
[ fig : depvel - petroff - contribs ] . .
see sections [ sec : brown - dif ] - [ sec : sedim ] for the description of the parameters).,scaledwidth=50.0% ] the assumption of the parallel and independent acting is advantageous for adapting the model to the moment method , since it allows us to split the rightmost integral in eq .
( [ eq : moment - eq ] ) into integrals pertaining to the every physical process separately . in this section
, we will describe each process in more detail and show how it can be adapted to the moment method .
few simplifications were made to the original model to make the subsequent analysis simpler : we considered only plagiophile canopies and dirac distribution of the needle sizes .
also note that here we focus only on needle - like obstacles . similar model for broadleaf canopies , given in @xcite , could be adapted to the moment method as well .
brownian diffusion is the dominant process driving the deposition of the particles smaller then 0.1 @xcite .
original model formulates the contribution to the deposition velocity due to the brownian diffusion as @xmath20 where @xmath21 is the magnitude of the wind velocity , @xmath22 is the schmidt number ( with @xmath23 being the kinematic viscosity of air and @xmath24 the brownian diffusion coefficient , @xmath25 ) , @xmath26 is the reynolds number , and @xmath27 is the needle diameter .
assuming laminar boundary layer around the obstacles , the model constants hold the values @xmath28 and @xmath29 .
cunningham correction factor @xmath30 is used in the original model @xcite , where @xmath31 is the mean free path of the particle in the air . in this whole study
we use simpler approximation @xmath32 @xcite .
comparison of these expressions is on fig .
[ fig : correction - factor ] , where it can be seen that their difference peaks to 12% for particle diameter around 0.2 .
.,scaledwidth=80.0% ] furthermore , for the brownian diffusion we take into account only the size - dependent part of the correction factor , dominant in the particle size range where the diffusion is significant , @xmath33 .
putting the expressions above into eq .
( [ eq : vd - b ] ) , we obtain @xmath34 where @xmath35 and @xmath36 . using this formula ,
the contribution to the moment equation can be written as @xmath37 interception denotes the process where the particle follows the streamline , but too close to the obstacle so that it is captured on the surface .
the original model parameterizes its contribution to the deposition velocity as @xmath38 where @xmath39 is the ratio of the leaf surface projected on the plane perpendicular to the flow to the total leaf surface .
since there is a linear dependence on the particle diameter , the expression can be integrated as is , resulting in @xmath40 with @xmath41 .
inertial impaction occurs when particles do not follow the streamlines due to their inertia , resulting in the collision with the obstacle .
the deposition velocity due to the inertial impaction is written as @xmath42 where @xmath43 is the impaction efficiency with the constant @xmath44 . to use this deposition velocity in the moment equations , we approximate the impaction efficiency as @xmath45 , where the coefficients @xmath46 and @xmath47 were obtained by minimizing the quadratic error of the original function and the approximation for @xmath48 $ ] . the original function and the approximation
are shown on fig .
[ fig : impaction - efficiency ] .
it is clear that this approximation introduces overprediction of the impaction efficiency for stokes numbers between 1 and 10 .
this was however deemed to be an acceptable compromise between the accuracy and the computational efficiency . without the approximation
, one could use the numerical integration to evaluate the integral to obtain more precise approximation of the original formula , but at a significantly higher computational cost .
approximated deposition velocity can then be formally written as @xmath49 after some algebraic manipulations the corresponding term in the moment equation can be written using the incomplete moments @xmath50 and @xmath51 as @xmath52 with the threshold @xmath53 and @xmath54 .
effect of the particle impaction due to the canopy turbulence is described by the deposition velocity @xmath55 here @xmath56 is the dimensionless particle relaxation time , @xmath57 is the particle relaxation time , @xmath58 is the local friction velocity , @xmath59 , and @xmath60 .
the contribution to the moment equation can be again expressed using the incomplete moments , @xmath61 with the threshold @xmath62 and @xmath63 .
sedimentation plays the major role for the particles with the diameter above .
the sedimentation contribution to the deposition velocity is expressed as @xmath64 where @xmath65 is the ratio of the leaf surface projected to the horizontal plane to the total leaf surface . substituting this expression to the integral in the moment equation , after some algebraic manipulations we obtain @xmath66 with @xmath67 and @xmath36 as before .
the developed model can be formally written as @xmath68 using the equations ( [ eq : vd - b - apprx ] ) , ( [ eq : vd - in ] ) , ( [ eq : vd - im - apprx ] ) , ( [ eq : vd - ti ] ) and ( [ eq : vd - se ] ) . in the moment method solver
it is however implemented via the equations ( [ eq : mom - b ] ) , ( [ eq : mom - in ] ) , ( [ eq : mom - im ] ) , ( [ eq : mom - ti ] ) and ( [ eq : mom - se ] ) .
two major approximation were made : first , cunningham approximation factor is replaced by its size - dependent part in the brownian diffusion term .
secondly , an inexact power law formula is used for the impaction efficiency .
comparison of this model with the original for an exemplary set of parameters is shown on fig .
[ fig : depvel - comparison ] .
the higher values of the deposition velocity for particles around are the consequence of the inexact approximation to the inertial impaction term described in section [ sec : vd - im ] .
the maximal difference of the deposition velocity given by the two models was determined by evaluating the deposition velocity for every combination of the parameters in the ranges expected in real - world situations ( @xmath69\ \si{\kg\per\m\cubed}$ ] , @xmath70\ \si{\mm}$ ] , @xmath71\ \si{\m\per\s}$ ] , @xmath72\ \si{\um}$ ] ) .
each interval was discretized using 50 points .
local friction velocity @xmath58 was set to 0 @xmath73 , as the turbulent impaction is implemented exactly and its contribution can only reduce the relative difference of the deposition velocities .
the largest relative difference @xmath74 was found to be 4.56 for the parameter values at the end of the expected ranges ( @xmath75 , @xmath76 , @xmath77 ) and particle diameter @xmath78 , giving the deposition velocities @xmath79 and @xmath80 . while this difference is certainly significant , measured values shows even higher variability @xcite . considering this ,
the costs of the alternative to this approximation - i.e. the numerical integration of the impaction term integral - does not outweigh the better fit to the original model . )
.,scaledwidth=50.0% ] before we move on to the description of the implementation , it is necessary to provide some assumptions on the particle size distribution .
size distributions of the atmospheric aerosols are often well fitted by a multimodal lognormal distribution @xcite .
this is the distribution we will use from now on .
we restrict ourselves only to the case of unimodal distribution , noting that the multimodal distribution can be modelled by a superposition of several unimodal distributions .
unimodal lognormal distribution can be described by three parameters : total number concentration @xmath81 , geometric mean size @xmath82 and geometric standard deviation @xmath83 .
its probability density function is @xmath84 knowing the three parameters , @xmath12-th moment can be calculated using the formula @xmath85 from the three moments of order 0 , @xmath86 and @xmath87 the three parameters can be obtained using the relations @xmath88 where @xmath89 and @xmath90 @xcite
. for the incomplete higher order moments following holds : @xmath91 where @xmath92 is the normal cumulative distribution function .
now we turn our attention to the choice of the moments . for
which orders we decide to solve the moment equation ( [ eq : moment - eq ] ) is to a degree an arbitrary decision .
when this problem is discussed in literature , cited reasons for a certain choice include the mathematical simplicity and ease of the formulation of the modelled processes or the physical interpretation of some moments @xcite .
choices of the moments used in the field of atmospheric aerosol modelling in the selected literature are summarized in tab .
[ tab : moments - choice ] .
.choices of the moments in the selected literature [ cols="<,^,<",options="header " , ] the recurrent usage of zeroth order moment brings substantial advantage , as it is equal to the total number concentration , and it is the order we will use as well . on the choice of the other moments authors differ . to assess the influence of the choice of the moments , following numerical experiment was performed .
we investigated the particle deposition in a 1d tube , spanning between 0 and 300 m. homogenous vegetation block of lad = 3 was placed between 100 a 150 m. velocity of the air in the whole tube was set to constant 1 , unaffected by the vegetation .
source of the pollutant was placed at 50 m from the inlet with the intensity of number of particles 1 and the distribution parameters @xmath93 .
the tube was discretized using 400 cells . beside the choices mentioned in tab .
[ tab : moments - choice ] , we tested also a variant with a negative order moment : 0 , -1 , 1 .
non integer choices of the orders would also be possible to use , but we saw no advantage that such choice could bring .
transport and the deposition of the pollutant was calculated by the sectional model based on the eq .
( [ eq : n - eq ] ) and by the moment method based on the eq .
( [ eq : moment - eq ] ) ( see section [ sec : implementation ] for details on the implementation ) . to discard possible errors due to the inexact approximation of the deposition velocity , only the sedimentation contribution , adapted exactly , was taken into account .
the numerical experiment is not meant to model any real - world situation , rather just demonstrate the behaviour of the moment method in a simple setting .
number and volume concentration distribution behind the barrier ( at 150 m ) are shown on fig .
[ fig : choice - dist ] . as a reference , calculated distributions are complemented by the distribution for a case without the vegetation present .
evolution of the zeroth moment ( equal to the number concentration ) and the third moment ( proportional to the volume concentration ) through the vegetation block are shown on fig .
[ fig : choice - horiz ] .
effect of the vegetation , while small in number concentration , is significant in volume concentration .
only the variant using the moments of orders 0 , -1 , and 1 reproduces well the number concentration distribution , but overpredicts the peak of the volume concentration .
variants using the orders 0 , 1 , 2 and 0 , 2 , 3 produce result closer to the sectional model in volume concentration , but with larger differences in number concentration .
variant using the orders 0 , 3 , 6 shows no advantages over the other variants .
choosing between the orders 0 , 1 , 2 and 0 , 2 , 3 , we opted for the latter variant , as the third moment is proportionate to the main quantity of interest - volume ( and mass ) concentration of the pollutant .
both the sectional model and the moment model were implemented using the openfoam platform @xcite .
second order upwind scheme was used for convective terms in equations ( [ eq : n - eq ] ) and ( [ eq : moment - eq ] ) and second order scheme based on the gauss theorem was used for the diffusive terms .
residual levels of @xmath94 were used to test for convergence of the steady state solver . when using the moment method , we have to solve the discretized eq .
( [ eq : moment - eq ] ) for the three selected moments .
these equations are coupled through the gravitational settling term and the deposition term , which depends on the moments of a different order than the one solved by the equation .
the coupling is dealt with the following way . in every iteration ,
first the parameters of the lognormal distribution @xmath81 , @xmath82 and @xmath83 are calculated using the equations ( [ eq : n]-[eq : ln2sigmag ] ) from the values in the preceeding iteration .
three moment equations are then solved one after another with the coupling terms resulting from the deposition being treated explicitly .
fully explicit treatment of the gravitational settling term ( [ eq : mom - grav ] ) can result in numerical instability , unless low values of the relaxation factors are used .
that would however lead to slower convergence , therefore we employed a semi - implicit treatment .
moment @xmath95 in ( [ eq : mom - grav ] ) is rewritten as @xmath96 with @xmath97 and the term @xmath98 is then treated explicitly and @xmath99 implicitly .
relaxation factors 0.95 were used both for the sectional equations and for the moment equations . for the first five iterations of the moment method the relaxation factors for the moment equations
were however set to lower value 0.8 , as the computations proved to be less stable at the beginning .
calculation of the distribution parameters @xmath82 and @xmath100 via eq .
( [ eq : dgn ] ) and ( [ eq : ln2sigmag ] ) includes the division of the moments , potentially very small far away from the source of pollutant . to avoid this problem , small background concentration in the whole domain
is set as an initial condition and used as a boundary condition where zero would be used otherwise .
here we describe two example problems of microscale flows through and around the vegetation and assess the applicability of the developed moment method to the simulation of pollutant dispersion .
two vegetation elements that could be encountered in the urban settings are investigated in this test : small patch of full grown trees and a dense hedgerow .
the flow field in both cases was precomputed by an in - house finite volume cfd solver .
the solver is based on the navier - stokes equations in the boussinesq approximation and utilizes @xmath101 turbulence model .
inlet profiles of velocity and the turbulence quantities , as well as the wall functions , are prescribed by the analytical expressions given by @xcite .
vegetation model for the momentum and @xmath101 equations described by @xcite is employed . for further details
we refer to @xcite , where the solver is described in more detail .
turbulent schmidt number was set to @xmath102 in both cases , based on the analysis by @xcite . in both cases presented below
, we simulated the dispersion of a coarse mode particles from a point or a line source .
the coarse mode is chosen as the mode that contains , together with the accumulation mode , majority of the volume of the particles in the urban environment @xcite , but is affected more strongly by the dry deposition than the accumulation mode .
the number distribution at the source is assumed to be lognormal with the parameters @xmath103 and @xmath104 , typical for the urban environment @xcite .
evaluation of the developed moment method was based on the comparison with the results obtained by the sectional model . in the sectional model , eq .
( [ eq : n - eq ] ) is solved for 41 particle sizes distributed uniformly between and .
the interval is chosen so that the behaviour of the number distribution as well as the volume distribution can be captured by the sectional model .
first case investigates the filtering properties of a small patch of full grown conifer trees .
a simplified 2d model is constructed as follows .
the 30 meters wide and 15 meters high tree patch is represented as a horizontally homogeneous vegetation block .
pollutant source is placed 15 meters upstream from the vegetation , 5 meters above the ground .
lad profile of the vegetation is prescribed by a formula given by @xcite , @xmath105 where @xmath106 m is the height of the trees , @xmath107 is the maximum lad , chosen so that leaf area index , @xmath108 , is equal to 5 , and @xmath109 is the corresponding height of maximal lad .
the sketch of the domain and the lad profile of the vegetation is shown on fig .
[ fig : veg2d - domain ] . trees are modelled as generic conifers with @xmath110 .
the drag coefficient is chosen as @xmath111 @xcite .
intensity of the point source is set to a normalized value in terms of number of particles . since all terms in eq .
( [ eq : n - eq ] ) and eq .
( [ eq : moment - eq ] ) are linear with respect to the number concentration , results can be simply scaled to other value of the source intensity if needed .
inlet wind profile is set as logarithmic with @xmath112 at height 20 m and @xmath113 m. for the number concentration in the sectional model and for all moments in the moment method the neumann boundary conditions are used on the ground , at the top and at the outlet .
no resuspension of the particles is allowed , i.e. any particle that falls on the ground stays on the ground indefinitely .
small value of the concentration and of the moments calculated from the lognormal distribution with the parameters @xmath114 is prescribed at the inlet .
domain is discretized using a cartesian grid with 220 cells in horizontal direction and 100 cells in vertical direction , graded so that the grid is finer near the ground and around the tree patch .
the near ground cells are 0.25 m high , and the vegetation block itself consists of 42 x 40 cells . flow field obtained by the cfd solver
is shown on fig .
[ fig : veg2d - streamlines ] . as visible
, the vegetation block slows the wind down , but allows the air to pass through .
results from the sectional and the moment model are compared in terms of the third moment of the particle size distribution , proportionate to the volume concentration of the particles . as we assume that the density is the same for particles of all sizes , third moment is also proportionate to the mass concentration of the particles . calculated field of the third moment by the moment method
is shown on fig .
[ fig : veg2d - m3 ] ( left ) . the relative difference @xmath115 of the results obtained by the moment method , @xmath116 , and by the sectional model , @xmath117 ,
is shown on the right panel of fig .
[ fig : veg2d - m3 ] . of the third moment calculated by the moment method and the sectional approach.,title="fig:",scaledwidth=70.0% ] of the third moment calculated by the moment method and the sectional approach.,title="fig:",scaledwidth=70.0% ]
the source of the largest discrepancies between the two methods is the vegetation block .
the relative difference raises up to 4% at the downstream edge of the vegetation block , and then decreases with the increasing distance from the vegetation .
the moment method overpredicts the deposition inside the vegetation due to the inexact approximation of the impaction efficiency described in section [ sec : vd - im ] .
( right ) volume concentration at [ 200 ; 2 ] .
discrete points calculated by the sectional method and the distribution calculated by the moment method are shown . for reference ,
the distribution calculated without the size dependent deposition and gravitational settling terms is shown as well.,title="fig:",scaledwidth=40.0% ] ( right ) volume concentration at [ 200 ; 2 ] .
discrete points calculated by the sectional method and the distribution calculated by the moment method are shown . for reference ,
the distribution calculated without the size dependent deposition and gravitational settling terms is shown as well.,title="fig:",scaledwidth=40.0% ] further insights can be obtained from fig .
[ fig : veg2d - dist ] .
it shows the volume concentration distribution at the downstream edge of the tree patch , and at 120 m downstream from the tree patch , both at height 2 m above ground .
the vegetation has negligible effect on the particles smaller than , but significantly reduces the mass of the particles above .
this is captured well both by the sectional and the moment method . at the downstream edge of the tree
patch the moment method predicts lower peak of the volume concentration distribution than the sectional approach .
the difference is reduced by the mixing of the filtered air with the unfiltered air flowing above the vegetation further away from the tree patch . next we tested the method on a 3d model of a dense hedgerow placed near a line source of the pollutant .
this case is a three dimensional extension of the 2d situation investigated in @xcite .
the yew hedge is 10 m wide , 3.2 m deep and 2.4 m high .
it is placed in the 40 m wide , 40 m long , and 20 m high computational domain .
two meters upstream from the hedge is a line source at height 0.5 m above ground .
intensity of the line source is set to a value in terms of number of particles , noting as in section [ sec : veg2d ] that the results can be scaled if other value is desired .
sketch of the domain is shown on the left panel of fig .
[ fig : veg3d - domain ] .
right panel shows the lad profile of the hedge , taken from the original article .
vegetation is further described by the needle diameter is , @xmath118 , and the vegetation drag coefficient which is set to @xmath119 as in @xcite .
the computational mesh was created using the openfoam _ snappyhexmesh _ generator .
the domain consist of 376 000 cells , refined near the ground and around the hedge .
the near - ground cells are 0.07 m high and the hedge itself is discretized using 54 x 20 x 22 cells . wind profile at the inlet is set as logarithmic with @xmath120 at height 2.4 m and @xmath113 m. boundary conditions for the sectional solver and moment method solvers are set similarly as in section [ sec : veg2d ] : neumann boundary conditions are used at the ground , top , sides , and at the outlet .
no resuspension of the particles fallen to the ground is allowed .
again , small amount of the particles given by the lognormal distribution with the parameters @xmath121 is prescribed at the inlet .
streamlines of the flow field calculated by the separate cfd solver are shown on fig .
[ fig : veg3d - streamlines ] . as in the 2d simulation in @xcite ,
recirculation zone is developed behind the dense hedge .
unlike the 2d case , here we can observe part the of the flow to be deflected to the sides .
third moment of the particle size distribution obtained by the moment method is shown on the left panels of fig .
[ fig : veg3d - m3-horiz ] and fig .
[ fig : veg3d - m3-vert ] . while a portion of the pollutant penetrates the barrier , part is deflected to the sides of the hedgerow , creating a zone with a reduced pollutant concentration behind it .
m. quantities shown are as on fig .
[ fig : veg2d - m3].,title="fig:",scaledwidth=70.0% ] m. quantities shown are as on fig .
[ fig : veg2d - m3].,title="fig:",scaledwidth=70.0% ] m. quantities shown are as on fig .
[ fig : veg2d - m3].,title="fig:",scaledwidth=70.0% ] m. quantities shown are as on fig .
[ fig : veg2d - m3].,title="fig:",scaledwidth=70.0% ] relative difference between the solution obtained by the moment method and sectional approach is shown on the right panels of fig .
[ fig : veg3d - m3-horiz ] and fig .
[ fig : veg3d - m3-vert ] . as in the tree patch case ,
the moment method overpredicts the deposition and consequently underestimates the volume concentration behind the barrier .
the difference is below 10% , and decreases away from the barrier .
effects of the coarser mesh in the upper part of the computational domain are visible on fig .
[ fig : veg3d - m3-vert ] .
however , it does not negatively affect the difference between the two methods .
volume concentration distribution at two points - inside the vegetation and downstream from the vegetation - is shown on fig .
[ fig : veg3d - dist ] . due to the smaller size of the vegetation than in the 2d tree patch case , the effect of the vegetation is less pronounced .
the moment method is able to reproduce the shape of the distribution well , but again produces a lower peak than the sectional method .
similarly as before , better fit can be observed further from the barrier due to the mixing with unfiltered air . .
( right ) volume concentration at [ 30 ; 0 ; 2 ] . for reference ,
the distribution calculated without the size dependent deposition and gravitational settling terms is shown.,title="fig:",scaledwidth=40.0% ] .
( right ) volume concentration at [ 30 ; 0 ; 2 ] . for reference ,
the distribution calculated without the size dependent deposition and gravitational settling terms is shown.,title="fig:",scaledwidth=40.0% ] to compare the computational performance of the developed model , we measured the runtime of the sectional approach and the moment method approach for the 3d case described in section [ sec : case3d ] .
both solvers were run on a single core of an intel xeon x5365 processor . the sectional model , comprised of 41 scalar pdes , finished in 9120 seconds .
the average runtime per each equation was thus 222 seconds .
moment model , comprised of 3 coupled pdes , finished in 1128 seconds .
that gives us eightfold acceleration compared to the sectional model .
even though the high number of bins used in this study might not be necessary to obtain sufficiently accurate results , to get an equivalent workload as the moment method in this case , only 5 bins could be used in the sectional model .
such number is insufficient to model the behaviour of the number distribution as well as the volume distribution well .
two points can be made in favor of the sectional method though .
first , the solution process of every equation is independent on the other equations , therefore the approach offers effortless parallelization for the number of cores up to the numbers of bins used .
this is not especially advantageous in our implementation , as the openfoam solvers are already parallelizable , but it could be an important factor for other implementations .
secondly , the relaxation factor 0.95 used for all simulations in the sectional approach was needed only for the bins representing the larger particles . using different values of this parameter for different bins can provide some reduction of the runtime .
in this study , we introduced a formulation of a dry deposition model suitable for implementation in a moment method .
as the original model by @xcite , our approximation includes five main processes of the dry deposition : brownian diffusion , interception , impaction , turbulent impaction , and sedimentation .
the developed deposition velocity model was implemented in a microscale finite volume solver based on the openfoam platform .
the solver employs the moment method to calculate the particle size distribution in the domain .
the deposition model was tested on two example problems of microscale pollutant dispersion .
comparison with the sectional method using the original dry deposition model revealed that the moment method is able to reproduce the shape of the particle size distribution well .
the relative differences between the sectional and the moment method in terms of the third moment of the distribution were below 10 % .
this difference was caused by the higher deposition velocity in our model , which in turn was caused by the inexact representation of the impaction process .
the impaction term could be represented better by employing numerical integration , that would however result in a higher computational costs , and reduce the advantages of using the moment method .
the moment method , described by three coupled pdes , proved to be more computationally efficient than the sectional model using 41 bins .
the speedup was approximately eightfold , and a workload equivalent to the moment method would be achieved by running a sectional model that uses only 5 bins .
this performance improvement together with the reliable results shows that the moment methods , often used in large scale atmospheric models , can be useful also for the microscale problems of pollutant dispersion in the urban environment . the developed method
as formulated here is applicable only when the particle size distribution can be approximated as a lognormal distribution .
here we used only unimodal distribution , but the usage of multimodal distribution would be also possible by superposition of several unimodal ones . furthermore , the method could be reformulated for other distributions , provided that algebraic relations between the moments and distribution parameters are known .
this work was supported by the grant sgs16/206/ohk2/3t/12 of the czech technical university in prague .
authors would like to thank to jan halama and ji frst for helpful discussions .
gromke , c. , blocken , b. , 2015 .
influence of avenue - trees on air quality at the urban neighborhood scale .
part i : quality assurance studies and turbulent schmidt number analysis for rans cfd simulations .
196 , 214223 .
mochida , a. , lun , i. y. , 2008 .
prediction of wind environment and thermal comfort at pedestrian level in urban area .
j. wind eng .
96 ( 1011 ) , 14981527 , 4th international symposium on computational wind engineering ( cwe2006 ) .
pirjola , l. , kulmala , m. , wilck , m. , bischoff , a. , stratmann , f. , otto , e. , 1999 .
formation of sulphuric acid aerosols and cloud condensation nuclei : an expression for significant nucleation and model comparison . j. aerosol sci .
30 ( 8) , 10791094 .
raupach , m. , briggs , p. , ford , p. , leys , j. , muschal , m. , cooper , b. , edge , v. , 2001 .
endosulfan transport ii . modeling airborne dispersal and deposition by spray and vapor .
. qual . 30 ( 3 ) , 729740 .
vranckx , s. , vos , p. , maiheu , b. , janssen , s. , 2015 .
impact of trees on pollutant dispersion in street canyons : a numerical study of the annual average effects in antwerp , belgium .
total environ .
532 , 474483 .
p , v. , bene , l. , 2016 .
cfd optimization of a vegetation barrier . in : karaszen , b. , manguoglu , m. , tezer - sezgin , m. , gktepe , s. , mr ugur ( eds . ) , numerical mathematics and advanced applications - enumath 2015 .
springer international publishing , cham , to appear . | a dry deposition model suitable for use in the moment method has been developed .
contributions from five main processes driving the deposition - brownian diffusion , interception , impaction , turbulent impaction , and sedimentation - are included in the model .
the deposition model was employed in the moment method solver implemented in the openfoam framework .
applicability of the developed expression and the moment method solver was tested on two example problems of particle dispersion in the presence of a vegetation on small scales : a flow through a tree patch in 2d and a flow through a hedgerow in 3d .
comparison with the sectional method showed that the moment method using the developed deposition model is able to reproduce the shape of the particle size distribution well .
the relative difference in terms of the third moment of the distribution was below 10% in both tested cases , and decreased away from the vegetation .
main source of this difference is a known overprediction of the impaction efficiency .
when tested on the 3d test case , the moment method achieved approximately eightfold acceleration compared to the sectional method using 41 bins .
dry deposition , vegetation , microscale modeling , moment method , particle dispersion |
a circulant graph of @xmath7 vertices is fully described by an @xmath7-by-@xmath7 symmetric circulant adjacency matrix @xmath67 defined as follows .
@xmath68\end{aligned}\ ] ] where @xmath69 .
obviously , every circulant matrix can be generated given any row of the matrix conventionally we use the first row of the matrix , denoted as @xmath70 .
it is clear that @xmath67 has at most @xmath7 distinct eigenvalues which are given by @xmath71 , where @xmath72 and @xmath73 @xcite . if @xmath67 is singular , some of the eigenvalues of @xmath67 are zeros .
the complete graph and complete bipartite graph are straightforward examples of circulant graphs with few distinct eigenvalues .
there are also some other interesting examples of circulant graph such as self - complementary circulant graphs and paley graphs with prime order @xcite
. both of these two families of graphs are also strongly regular graphs which have only three distinct eigenvalues .
for example , the paley graph on 13 vertices has three distinct eigenvalues : 6 ( with multiplicity 1 ) and @xmath74 ( both with multiplicity 6 ) , and thus the diagonal unitary @xmath75 can be implemented efficiently .
we note here it is required to implement qft ( and its inverse ) for the dimension of 13 , which does not have the form of @xmath27 .
the qft on general dimensions can be implemented by means of amplitude amplification with extra qubit registers to perform the computation @xcite .
alternatively , approximate versions of the qft on general dimensions have also been developed @xcite .
we say that the eigenvalues of a circulant graph can be characterised efficiently , if they can be calculated efficiently classically . in other words , the eigenvalue matrix @xmath76 of the given circulant hamiltonian can be efficiently computed , and thus the diagonal unitary operator @xmath75 can be efficiently implemented @xcite . specifically , there exists a quantum circuit shown in figure [ diagonalunitarycircuit ] , which transforms a computational basis state @xmath77 , together with a @xmath16-qubit ancilla @xmath78 for @xmath79 , as @xmath80 where @xmath81 .
note that here we assume @xmath82 can be expressed exactly as a rational number with @xmath16 bits of precision .
if this is not the case , truncating @xmath82 to @xmath16 bits of precision will introduce an error which can be made arbitrarily small by taking large enough @xmath79 . the function
@xmath83 returns @xmath84 for any given @xmath85 .
@xmath84 is always a real number since the adjacency matrix is symmetric .
for example , for the case of the cycle graph of @xmath27 vertices , there are essentially @xmath86 distinct eigenvalues simply given by @xmath87 , where @xmath81
. and then @xmath83 will be the cosine function that can be computed with a number of operations polynomial in @xmath4 , using a reversible equivalent of classical algorithms to compute trigonometric functions , e.g. the taylor approximation .
in general , given a sparse circulant graph which has only @xmath37 @xmath88s in the first row @xmath70 of its adjacency matrix , an efficient function @xmath83 can be given as @xmath89 where @xmath90 is the set of positions for which the first row in nonzero .
@xmath83 is a sum of @xmath91 numbers , taking @xmath33 time to compute . for a non - sparse circulant graph ,
its eigenvalues are still possible to be calculated efficiently classically .
some straightforward examples are complete graph , complete bipartite graph @xmath92 and cocktail party graph .
therefore , together with the quantum circuits of qft and the inverse of qft , we construct an efficient quantum circuit for implementing ctqw on the circulant graph whose eigenvalues can be computed efficiently classically .
unlike the sampling problem we discussed in the main text , the scenario of `` swap test '' , where we compare two unitary processes @xmath36 and @xmath93 , could sometimes be easier for a classical computer .
imagine we start each process in the state @xmath94 .
then the overlap @xmath95 between the resulting output states approximated by the swap test satisfies @xmath96 where @xmath97 is the value at position @xmath85 on the diagonal of @xmath30 .
@xmath95 can be approximated by a classical algorithm up to @xmath48 additive error .
the algorithm simply takes the average of @xmath37 values of the product @xmath98 for uniformly random @xmath85 . for each @xmath85
, this value can be computed exactly in polynomial time .
this highlights that the complexity of comparing @xmath99 and @xmath100 depends on the choice of input states @xmath101 and @xmath102 . in full generality
, one could allow these to be arbitrary states produced by a polynomial - time quantum computation ; the state comparison problem would then be bqp - complete , but for rather trivial reasons .
we expect that the problem would remain classically hard for choices of initial states relevant , for example , to quantum - chemistry applications . on the other hand ,
the swap test can still be used as in scenario ( b ) in the main text to compare the evolution of two hamiltonians , one of which is not circulant but is efficiently implementable . in this case , the comparison problem is also bqp - complete , and hence expected to be hard for a classical computer .
we present the ideal and experimentally sampled probability distributions of ctqw with initial states @xmath58'$ ] , @xmath59'$ ] ( mentioned in the main text ) , @xmath103'$ ] , @xmath104'$ ] , @xmath105'$ ] , @xmath106'$ ] , in table i. the achieved average fidelities between ideal and experimental probability distributions are [email protected]% , [email protected]% , [email protected]% , [email protected]% , [email protected]% and [email protected]% respectively . in the main text , we reconstructed the density matrices for the two quantum states @xmath107 and @xmath108 , through performing quantum state tomography .
here we also present the reconstructed density matrices for another two evolution states @xmath109 and @xmath110 , with the achieved fidelities of [email protected]% and [email protected]% respectively .
see in figure [ densitym12 ] .
the four reconstructed density matrices @xmath111 , @xmath112 , @xmath113 and @xmath114 for quantum states @xmath107 , @xmath108 , @xmath115 and @xmath116 are shown as follows . | the random walk formalism is used across a wide range of applications , from modelling share prices to predicting population genetics .
likewise quantum walks have shown much potential as a framework for developing new quantum algorithms . in this paper , we present explicit efficient quantum circuits for implementing continuous - time quantum walks on the circulant class of graphs .
these circuits allow us to sample from the output probability distributions of quantum walks on circulant graphs efficiently .
we also show that solving the same sampling problem for arbitrary circulant quantum circuits is intractable for a classical computer , assuming conjectures from computational complexity theory .
this is a new link between continuous - time quantum walks and computational complexity theory and it indicates a family of tasks which could ultimately demonstrate quantum supremacy over classical computers . as a proof of principle we have experimentally implemented the proposed quantum circuit on an example circulant graph using a two - qubit photonics quantum processor .
quantum walks are the quantum mechanical analogue to the well - known classical random walk and they have established roles in quantum information processing @xcite . in particular , they are central to quantum algorithms created to tackle database search @xcite , graph isomorphism @xcite , network analysis and navigation @xcite , and quantum simulation @xcite , as well as modelling biological processes @xcite . meanwhile , physical properties of quantum walks have been demonstrated in a variety of systems , such as nuclear magnetic resonance @xcite , bulk @xcite and fiber @xcite optics , trapped ions @xcite , trapped neutral atoms @xcite , and photonics @xcite .
almost all physical implementations of quantum walk so far followed an analog approach as for quantum simulation @xcite , whereby the apparatus is dedicated to implement specific instances of hamiltonians without translation onto quantum logic .
however , there is no existing method to implement analog quantum simulations with error correction or fault tolerance , and they do not scale efficiently in resources when simulating broad classes of large graphs . in this paper , we present efficient quantum circuits for implementing continuous time quantum walks ( ctqws ) on circulant graphs with an eigenvalue spectrum that can be classically computed efficiently .
these quantum circuits provide the time - evolution states of ctqws on circulant graphs exponentially faster than best previously known methods @xcite .
we report a proof - of - principle experiment , where we implement ctqws on an example circulant graph ( namely the complete graph of four vertices ) using a two - qubit photonics quantum processor to sample the probability distributions and perform state tomography on the output state of a ctqw .
we also provide evidence from computational complexity theory that the probability distributions that are output from the circuits of this circulant form are hard to sample from using a classical computer , implying our scheme also provides an exponential speedup for sampling .
efficient quantum circuit implementations of ctqws have been presented for sparse and efficiently row - computable graphs @xcite , and specific non - sparse graphs @xcite . however , the design of quantum circuits for implementing ctqws is in general difficult , since the time - evolution operator is time - dependent and non - local @xcite .
a subset of circulant graphs have the property that their eigenvalues and eigenvectors can be classically computed efficiently @xcite .
this enables us to construct a scheme that efficiently outputs the quantum state @xmath0 , which corresponds to the time evolution state of a ctqw on corresponding graphs .
one can then either perform direct measurements on @xmath1 or implement further quantum circuit operations to extract physically meaningful information .
for example the `` swap test '' @xcite can be used to estimate the similarity of dynamical behaviors of two circulant hamiltonians operating on two different initial states , as shown in figure [ cswaptest_sampling](a ) .
this procedure can also be adapted to study the stability of quantum dynamics of circulant molecules ( for example , the dna mbius strips @xcite ) in a perturbational environment @xcite . on the other hand , when measuring @xmath1 in the computational basis we can sample the probability distribution @xmath2 that describes the probability of observing the quantum walker at position @xmath3an @xmath4-bit string , corresponding to the vertices of the given graph , as shown in figure [ cswaptest_sampling](b ) .
sampling of this form is sufficient to solve various search and characterization problems @xcite , and can be used to deduce critical parameters of the quantum walk , such as mixing time @xcite .
it is unlikely for a classical computer to be able to efficiently sample from @xmath5 .
we adapt the similar methodology of refs .
@xcite to show that if there did exist a classical sampler for a somewhat more general class of circuits , this would have the following unlikely complexity - theoretic implication : the infinite tower of complexity classes known as the polynomial hierarchy would collapse .
this evidence of hardness exists despite the classical efficiency with which properties of the ctqw , such as the eigenvalues of circulant graphs , can be computed on a classical machine . for an undirected graph @xmath6 of @xmath7 vertices ,
a quantum particle ( or `` quantum walker '' ) placed on @xmath6 evolves into a superposition @xmath8 of states in the orthonormal basis @xmath9 that correspond to vertices of @xmath6 .
the exact evolution of the ctqw is governed by connections between the vertices of @xmath6 : @xmath10 where the hamiltonian is given by @xmath11 for hopping rate per edge per unit time @xmath12 and where @xmath13 is the @xmath7-by-@xmath7 symmetric adjacency matrix , whose entries are @xmath14 , if vertices @xmath15 and @xmath16 are connected by an edge in @xmath6 , and @xmath17 otherwise @xcite .
circulant graphs are defined by symmetric circulant adjacency matrices for which each row @xmath15 when right - rotated by one element , equals the next row @xmath18for example complete graphs , cycle graphs and mobius ladder graphs are all subclasses of circulant graphs .
it follows that hamiltonians for ctqws on any circulant graph have a symmetric circulant matrix representation , which can be diagonalized by the unitary fourier transform @xcite , i.e. @xmath19 , where @xmath20 and @xmath21 is a diagonal matrix containing eigenvalues of @xmath22 , which are all real and whose order is determined by the order of the eigenvectors in @xmath23 .
consequently , we have @xmath24 , where the time dependence of @xmath25 is confined to the diagonal unitary operator @xmath26
. the fourier transformation @xmath23 can be implemented efficiently by the well - known qft quantum circuit @xcite . for a circulant graph that has @xmath27 vertices ,
the required qft of @xmath7 dimension can be implemented with @xmath28 quantum gates acting on @xmath29 qubits . to implement the inverse qft ,
the same circuit is used in reverse order with phase gates of opposite sign .
@xmath30 can be implemented using at most @xmath7 controlled - phase gates with phase values being a linear function of @xmath31 , because an arbitrary phase can be applied to an arbitrary basis state , conditional on at most @xmath32 qubits .
given a circulant graph that has @xmath33 non - zero eigenvalues , only @xmath33 controlled - phase gates are needed to implement @xmath30 . if the given circulant graph has @xmath34 distinct eigenvalues , which can be characterised efficiently ( such as the cycle graphs and mobius ladder graphs ) , we are still able to implement the diagonal unitary operator @xmath30 using polynomial quantum resources .
a general construction of efficient quantum circuits for @xmath30 was given by childs @xcite , and is shown in the appendix for completeness .
thus , the quantum circuit implementations of ctqws on circulant graphs can be constructed , which have an overall complexity of @xmath33 , and act on at most @xmath29 qubits .
compared with the best known classical algorithm based on fast fourier transform , that has the computational complexity of @xmath35 @xcite , the proposed quantum circuit implementation generates the evolution state @xmath1 with an exponential advantage in speed . consider a circuit of the form @xmath36 , where @xmath30 is a diagonal matrix made up of @xmath37 controlled - phase gates .
define @xmath38 @xmath39 represents a family of circuits having the following structure : each qubit line begins and ends with a hadamard ( @xmath22 ) gate , and , in between , every gate is diagonal in the computational basis .
this class of circuits is known as instantaneous quantum polynomial time ( iqp ) @xcite .
it is known that computing @xmath40 for arbitrary diagonal unitaries @xmath30 made up of circuits of @xmath37 gates , even if each acts on @xmath41 qubits , is # p - hard @xcite .
this hardness result even holds for approximating @xmath40 up to any multiplicative error strictly less than @xmath42 @xcite , where @xmath43 is said to approximate @xmath40 up to multiplicative error @xmath44 if @xmath45 towards a contradiction , assume that there exists a polynomial - time randomized classical algorithm which samples from @xmath46 .
then a classic result of stockmeyer @xcite states that there is an algorithm in the complexity class fbpp@xmath47 which can approximate any desired probability @xmath5 to within multiplicative error @xmath48 .
this complexity class fbpp@xmath47described as polynomial - time randomized classical computation equipped with an oracle to solve arbitrary np problems sits within the infinite tower of complexity classes known as the polynomial(-time ) hierarchy @xcite . combining with the above hardness result of approximating @xmath40
, we find that the assumption implies that an fbpp@xmath49 algorithm solves a # p - hard problem , so p@xmath50 would be contained within fbpp@xmath49 , and therefore the polynomial hierarchy would collapse to its third level .
this consequence is considered very unlikely in computational complexity theory @xcite .
we therefore conclude that a polynomial - time randomized classical sampler from the distribution @xmath46 is unlikely to exist .
further , this even holds for classical algorithms which sample from any distribution @xmath51 which approximates @xmath46 up to multiplicative error strictly less than @xmath42 in each probability @xmath5 .
it is worth noting that if the output distribution results from measurements on only @xmath52 qubits @xcite , or obeys the sparsity promise that only a @xmath37-sized , and a priori unknown , subset of the measurement probabilities are nonzero @xcite , it could be classically efficiently sampled .
the proof of hardness here does not hold for the approximate sampling from @xmath46 up to small _ additive _ error .
it is an interesting open question whether similar techniques to @xcite can be used to prove hardness of the approximate case , perhaps conditioned on other conjectures in complexity theory . as an experimental demonstration
, we used a photonic quantum logic to simulate ctqws on the @xmath53 graph a complete graph with self loops on four vertices ( figure [ expsetup](a ) ) .
complete graphs are a special kind of circulant graph , with an adjacency matrix @xmath13 where @xmath54 for all @xmath55 .
the hamiltonian of a complete graph on @xmath7 vertices has only 2 distinct eigenvalues , 0 and @xmath56 .
therefore , the diagonal matrix of eigenvalues of @xmath53 is @xmath57 . following the aforementioned discussion
, we can readily construct the quantum circuit for implementing ctqws on @xmath53 graph based on diagonalization using the qft matrix .
however , the choice of using the qft matrix as the eigenbasis of hamiltonian is not strictly necessary any equivalent eigenbasis can be selected . through the diagonalization
using hadamard eigenbasis , an alternative efficient quantum circuit for implementing ctqws on @xmath53 graph is shown in figure [ expsetup](b ) , which can be easily extended to the complete graph on @xmath7 vertices .
we built a configurable two - qubit photonics quantum processor ( figure [ expsetup](c ) ) , adapting the entanglement - based technique presented in @xcite , and implemented ctqws on @xmath53 graph with various evolving times and initial states .
specifically , we prepared two different initial states @xmath58'$ ] and @xmath59'$ ] , which represent the quantum walker starting from vertex 1 , and the superposition of vertices 1 and 2 respectively .
we chose the evolution time following the list @xmath60 , which covers the whole periodical characteristics of ctqws on @xmath53 graph .
for each evolution , we sampled the corresponding probability distribution with fixed integration time , shown in figure [ expdata](a ) and ( b ) . to measure how close the experimental and ideal probability distributions are , we calculated the average fidelities defined as @xmath61 .
the achieved average fidelities for the samplings with two distinct initial states are [email protected]% and [email protected]% respectively .
through the proposed circuit implementation , we are also able to examine the evolution states using quantum state tomography , which is generally difficult for the analog simulations . for two specific evolution states @xmath63 and @xmath64 , we performed quantum state tomography and reconstructed the density matrices using the maximum likelihood estimation technique .
the two reconstructed density matrices achieve fidelities of [email protected]% and [email protected]% respectively , shown in figure [ expdata](c ) and ( d ) . in this paper , we have described how ctqws on circulant graphs can be efficiently implemented on a quantum computer , if the eigenvalues of the graphs can be characterised efficiently classically .
in fact , we can construct an efficient quantum circuit to implement ctqws on any graph whose adjacency matrix is efficiently diagonalisable , in other words , as long as the matrix of column eigenvectors @xmath23 and the diagonal matrix of the eigenvalue exponentials @xmath30 can be implemented efficiently .
we have shown that the problem of sampling from the output probability distributions of quantum circuits of the form @xmath65 is hard for classical computers , based on a highly plausible conjecture that the polynomial hierarchy does not collapse .
this observation is particularly interesting from both perspectives of ctqw and computational complexity theory , as it provides new insights into the ctqw framework and also helps to classify and identify new problems in computational complexity theory . for the ctqws on the circulant graphs of @xmath66 non - zero eigenvalues ,
the proposed quantum circuit implementations do not need a fully universal quantum computer , and thus can be viewed as an intermediate model of quantum computation . although the hardness of the approximate case of the sampling problem is unknown , the evidence we provided for the exact case indicates a promising candidate for experimentally establishing quantum supremacy over classical computers , and further evidence against the extended church - turing thesis .
compared with other intermediate models such as the one clean qubit model @xcite , iqp and boson sampling @xcite , the quantum circuit implementation of ctqws is also more appealing due to available methods in fault tolerance and error correction , which are difficult to implement for other models @xcite .
this may also lead onto other practical applications through the use of ctqws for quantum algorithm design .
10 farhi , e. & gutmann , s. quantum computation and decision trees .
* 58 , * 915 ( 1998 ) .
kempe , j. quantum random walks : an introductory overview .
* 44 , * 307 - 327 ( 2003 ) .
childs , a. m. , gosset , d. & webb , z. universal computation by multiparticle quantum walk . * 339 , * 791 - 794 ( 2013 ) .
childs , a. m. & goldstone , j. spatial search by quantum walk . * 70 , * 022314 ( 2004 ) .
douglas , b. l. & wang , j. b. a classical approach to the graph isomorphism problem using quantum walks .
* 41 , * 075303 ( 2008 ) .
gamble , j. k. , friesen , m. , zhou , d. , joynt , r. & coppersmith , s. n. two - particle quantum walks applied to the graph isomorphism problem .
* 81 , * 052313 ( 2010 ) .
berry , s. d. & wang , j. b. two - particle quantum walks : entanglement and graph isomorphism testing .
* 83 , * 042317 ( 2011 ) .
berry , s. d. & wang , j. b. quantum - walk - based search and centrality .
* 82 , * 042333 ( 2010 ) .
snchez - burillo , e. , duch , j. , gmez - gardees , j. & zueco , d. quantum navigation and ranking in complex networks .
* 2 , * 605 ( 2012 ) .
lloyd , s. universal quantum simulators .
* 273 , * 1073 - 1078 ( 1996 ) .
berry , d. w. & childs , a. m. black - box hamiltonian simulation and unitary implementation .
* 12 , * 29 - 62 ( 2012 ) .
schreiber , a. _ et al . _ a 2d quantum walk simulation of two - particle dynamics .
* 336 , * 55 - 58 ( 2012 ) .
engel , g. s. _ et al .
_ evidence for wavelike energy transfer through quantum coherence in photosynthetic systems .
* 446 , * 782 - 786 ( 2007 ) .
rebentrost , p. _ et al . _ environment - assisted quantum transport .
* 11 , * 033003 ( 2009 ) .
du , j. _ et al .
_ experimental implementation of the quantum random - walk algorithm .
* 67 , * 042316 ( 2003 ) .
ryan , c. a. _ et al .
_ experimental implementation of a discrete - time quantum random walk on an nmr quantum - information processor .
* 72 , * 062317 ( 2005 )
. do , b. _ et al . _
experimental realization of a quantum quincunx by use of linear optical elements .
* 22 , * 499 - 504 ( 2005 ) .
schreiber , a. _ et al .
_ photons walking the line : a quantum walk with adjustable coin operations . * 104 , * 050502 ( 2010 ) .
xue , p. , sanders , b. c. & leibfried , d. quantum walk on a line for a trapped ion .
* 103 , * 183602 ( 2009 ) .
schmitz , h. _ et al .
_ quantum walk of a trapped ion in phase space .
* 103 , * 090504 ( 2009 ) .
zhringer , f. _ et al .
_ realization of a quantum walk with one and two trapped ions .
* 104 , * 100503 ( 2010 ) .
karski , m. _ et al .
_ quantum walk in position space with single optically trapped atoms .
* 325 , * 174 - 177 ( 2009 ) .
perets , h. b. _ et al .
_ realization of quantum walks with negligible decoherence in waveguide lattices .
* 100 , * 170506 ( 2008 ) .
carolan , j. _ et al .
_ on the experimental verification of quantum complexity in linear optics .
* 8 , * 621 ( 2014 ) .
manouchehri , k. & wang , j. b. .
springer - verlag , berlin , ( 2014 ) .
exceptions , such as @xcite , adopted the qubit model , but there was no discussion on potentially efficient implementation of quantum walks .
ng , m. k. .
oxford university press new york , ( 2004 ) .
aharonov , d. & ta - shma , a. adiabatic quantum state generation and statistical zero knowledge . in _ proceedings of the thirty - fifth annual acm symposium on theory of computing _ 20 - 29 .
acm ( 2003 ) .
berry , d. w. , ahokas , g. , cleve , r. & sanders , b. c. efficient quantum algorithms for simulating sparse hamiltonians .
* 270 , * 359 - 371 ( 2007 ) .
childs , a. m. & kothari , r. limitations on the simulation of non - sparse hamiltonians .
* 10 , * 669 - 684 ( 2009 ) .
childs , a. m. on the relationship between continuous - and discrete - time quantum walk .
* 294 , * 581 - 603 ( 2010 ) .
gray , r. m. . now publishers inc .
, ( 2006 ) .
buhrman , h. , cleve , r. , watrous , j. & de wolf , r. quantum fingerprinting .
* 87 , * 167902 ( 2001 ) .
han , d. _ et al . _ folding and cutting dna into reconfigurable topological nanostructures .
* 5 , * 712 - 717 ( 2010 ) .
peres , a. stability of quantum motion in chaotic and regular systems .
* 30 , * 1610 ( 1984 ) .
prosen , t. & znidaric , m. stability of quantum motion and correlation decay .
* 35 , * 1455 ( 2002 ) .
aaronson , s. & arkhipov , a. the computational complexity of linear optics . in _ proceedings of the forty - third annual acm symposium on theory of computing _ 333342 .
acm ( 2011 ) .
bremner , m. j. , jozsa , r. & shepherd , d. j. classical simulation of commuting quantum computations implies collapse of the polynomial hierarchy . , rspa.2010.0301 ( 2010 ) .
bremner , m. j. , montanaro , a. & shepherd , d. j. average - case complexity versus approximate simulation of commuting quantum computations , preprint at http://arxiv.org/abs/1504.07999 ( 2015 ) .
nielsen , n. a. & chuang , i. l. .
cambridge university press , ( 2010 ) .
childs , a. m. .
phd thesis , massachusetts institute of technology ( 2004 ) .
shepherd , d. & bremner , m. j. temporally unstructured quantum computation . * 465 , * 1413 - 1439 ( 2009 )
. fujii , k. & morimae , t. quantum commuting circuits and complexity of ising partition functions , preprint at http://arxiv.org/abs/1311.2128 ( 2013 ) .
goldberg , l. a. & guo , h. the complexity of approximating complex - valued ising and tutte partition functions , preprint at http://arxiv.org/abs/1409.5627 ( 2014 ) .
stockmeyer , l. j. on approximation algorithms for # p. * 14 , * 849 - 861 ( 1985 ) .
papadimitriou , c. .
addison - wesley , ( 1994 ) .
nest , m. simulating quantum computers with probabilistic methods .
* 11 , * 784 - 812 ( 2011 ) .
schwarz , m. & nest , m. simulating quantum circuits with sparse output distributions , preprint at http://arxiv.org/abs/1310.6749 ( 2013 ) .
zhou , x. q. _ et al .
_ adding control to arbitrary unknown quantum operations . * 2 , * 413 ( 2011 ) .
knill , e. , & laflamme , r. power of one bit of quantum information .
* 81 , * 5672 ( 1998 ) .
rohde , p. p. & ralph , t. c. error tolerance of the boson - sampling model for linear optics quantum computing .
* 85 , * 022332 ( 2012 ) .
rangarajan , r. , goggin , m. & kwiat , p. optimizing type - i polarization - entangled photons .
* 17 , * 18920 - 18933 ( 2009 ) . |
cyg x-1 is an unavoidable target for any high energy instrument .
being one of the most luminous sources ( up to the mev range ) , it represents an ideal lab to study the mechanism at work in the direct environment of a black hole .
there , the accretion flow is thought to form an optically thick disk and/or an optically thin corona , while the observed radio jets could originate from the same area ( see e.g.the review by done , gierliski & kubota 2007 ) .
the high energy radiation provides insights on the physical processes at work in this region .
thanks to its persistent high flux ( @xmath0 1 crab ) and usually hard spectrum , cygnus x-1 is easy to observe and there are numerous measurements of its the spectrum and variability in the x - rays . at higher energies however , the results are very scarce .
very few instruments were able to explore the properties of the emission above 200 kev .
the high energy emission of cyg x-1 is relatively well known from soft x - rays up to a few hundreds kev .
the shape of the spectrum is variable and can change dramatically on time scales as short as a day .
there are however two main , relatively stable , spectral states : the hard state ( hs , corona dominated ) and the soft state ( ss , disk dominated ) .
their description can be found in various papers ( see for example liang & nolan 1984 ; gierlinski et al . , 1997 , 1999 ) .
so far , the mev region of the spectrum was best explored by the compton gamma - ray observatory ( cgro ) .
mcconnell et al ( 2002 ) have shown that , even though the hs hard x - ray emission is dominated by a thermal comptonization component , both hs and ss spectra present a non thermal powerlaw component extending above 1 mev with a slope of 2.6 in the ss and @xmath1 3 in the hs
. moreover , broad features around 1 mev have been observed in several occasions in the past ( see for example , bassani et al .
1989 and references therein ) .
since cgro observatory however , only the _ integral _ mission contains instrument exploring the same energy domain .
recently , laurent et al . (
2011 ) , stacked all the _ integral / ibis _ data available for cygnus x-1 and detected the presence of a non - thermal power law component between 400 kev and 2 mev .
interestingly they found that contrary to the thermal comptonization component present below 400 kev , the non - thermal power - law emission appears to be strongly polarized .
this non - thermal component appears to have a flux that is stronger than that measured by cgro by a factor 5 - 10 and a much harder spectral slope @xmath2 . here
, we use another _ integral _ instrument operating in this energy range , the spectrometer spi , to investigate the high energy spectral shape of cyg x-1 and test the presence of non - thermal high energy excess .
+ our first goal is to take advantage of the sensitivity achieved with the spi detector and the large amount of data and perform a detailed analysis of the energy extent of the hard x - ray emission together with its spectral variability . + moreover , the spectroscopic capability of the germanium crystals allow us to seek for the presence of spectral features linked to the annihilation process . hereafter , we present briefly the instrument , data set and the method followed for the analysis .
then , we report on our results and start by examining the total mean spectrum to determine the emission above a few hundreds of kev , where scarcity of photons imposes exposure as long as possible . in a second step ,
we analyse the source behavior during individual revolutions , and build several averaged spectra , following some hardness criteria which can be considered as characteristic of the spectral state of the source .
we conclude with a comparison with previous results .
spi is a spectrometer aboard the integral observatory operating between 20 kev and 8 mev . the description of the instrument and its performance can be found in vedrenne et al .
( 2003 ) and roques et al .
the main features of interest for our study are a large fov ( @xmath3 ) with an angular resolution of @xmath4 ( fwhm ) based on a coded aperture mask .
the germanium camera , beyond an excellent spectroscopic capability , offers a good sensitivity over more than 2 decades in energy with a unique instrument .
it is surrounded by a 5-cm thick bgo shield ( acs , anti - coincidence shield ) which measures the particle flux .
this latter can be used as a good tracers of the background level .
+ during a 3-day orbit , the usual dithering strategy ( jensen et al . , 2003 )
consists of a hundred of 30 - 40 min exposures ( also called scw for science window ) , with a given region scanned by @xmath5 steps following pre - determined patterns .
the recommended pattern for spi observations is a grid of 5x5 around the chosen target .
unfortunately , in order to content more proposers , except a few exceptions , most of the cyg x-1 data has been obtained through amalgamated observations , ie with the pattern center somewhere between cyg x-1 and cyg a region . as a consequence
, cyg x-1 appears only in one side of the fov , reducing the mean efficient area ( the source is partially coded ) , with some interruptions in the observation sequences , when the source goes out from the field of view . for our analysis ,
we have selected in the whole integral observation plan , those revolutions in which cyg x-1 is included in the @xmath6 fov during more than 20 scw ( 50 ks ) .
this gives a total of 42 revolutions .
these observations encompass 4 ms of effective time , from 2003 , june to 2009 december .
they were grouped together according to temporal proximity into 13 periods .
exposures with high background level ( entry / exit of radiation belts , solar activity ) or large source off - angle ( source / pointing axis angle beyond 13@xmath7 ) have been removed from our dataset .
table 1 gives some details about the observation pattern , beginning , end and useful duration , for each of the defined periods .
we follow the analysis method described in jourdain & roques ( 2009 ) , based on a sky model fitting , through a @xmath8 minimization .
this methods makes use of the pulse shape discrimination ( psd ) system , a second electronic chain operating in parallel , to eliminate spurious events occurring in the mev region . in the 650 kev-2.2 mev energy range , we use
only the psd tagged events , while below 650 kev , all the single events are analysed .
this procedure has been validated with the crab nebula observations .
+ in this work , we consider a common sky model for all the observations , which consists in 4 sources , namely : cyg x-1 , cyg x-3 , exo2023 + 375 and grs1915 + 105 . to keep the number of degrees of freedom to a minimum , we consider that the three latter have a constant flux over the revolution timescale . for cyg x-1
, we consider a variability timescale of 1 scw ( @xmath0 2 ks ) for the lightcurve and hardness ratio studies , while the spectra have been built assuming a constant source flux during each revolution .
+ the background determination relies on empty field observations . because of detector failures , it is important to use proper empty fields for each configuration .
once the uniformity map of the detector plan is fixed , we allow its normalisation to vary on a timescale of @xmath0 10 - 20 ks .
+ the incident photon spectra have been deconvolved and fitted through the spectral fitting procedure from xspec 11 ( arnaud 1996 ) .
due to some uncertainties on the energy response in the lowest channels and threshold effects , we exclude the first channels ( @xmath9 23.5 kev ) from the fit process .
the spectra presented contain thus 41 ( more or less logarithmic ) channels between 23.5 kev and 2.2 mev .
note that there are two strong narrow background lines ( 10 to 40 times higher than the continuum emission ) at 139 and 198 kev .
we use very narrow channels around these energies to isolate the line emission and prevent contamination of the neighbouring channels .
they may appear as offset data points with large error bars in the spectra .
the total mean spectrum gathers the whole set of available clean data ( 4 ms of observation distributed over more than 6 years ) .
it has been built by extracting one spectrum per revolution then summing them . due to the impressive signal to noise ratio at low energy , we are unable to ensure that the response matrices are known with a sufficient precision
. we have thus added 0.5% of systematic errors to the data during the fit procedure . +
the data have been first fitted with a simple analytical cutoff power law and a comptonization + reflection model ( reflect*comptt in xspec ) .
residuals clearly show deviation from these models above 200 - 300 kev where an excess of emission appears above the model prediction .
to go further , we keep a comptonization law ( + its reflected component ) in the low energy part and focus on the high energy emission .
following laurent et al .
( 2011 ) , we first try to model this additional component by a single power law .
the fit converges toward a photon index of @xmath10 ( close to the 1.6 @xmath11 0.2 value reported by these authors ) but the @xmath8 value remains unacceptable , with a huge contribution of the last channels .
indeed , the high energy part doesnot follow such a power - law : instead , the emission presents a rather sharp decrease after 700 kev and the source is not detected above 1 mev .
a better result is obtained when modeling this high energy component by a second comptonization region with a temperature around 100 kev . as several couples ( kt , @xmath12 )
can reproduce the data , we fixed @xmath12 to 0.5 and obtained a best fit value of kt=123 kev ( see table 2 ) .
[ fig : tot.comptt1et2 ] presents the observed spectrum with one and two comptonization models .
even though the set of parameters proposed in table 2 is only a possible solution , the two comptonization model provides a good description of the data and more specifically of the curvature observed up to 1 mev .
the mev emission may depend on the source spectral state or intensity .
to follow potential changes in the source emission , we have analysed in details each revolution and used the information to group observations corresponding to similar states
[ fig : clrev ] displays the temporal evolution of cyg x-1 in 30 - 70 kev energy range , with the fluxes averaged by revolution ( 100 - 200 ks timescale ) and fig . [ fig : hrrev ] the evolution of the hardness ( defined in the 30 - 120 kev range by @xmath13 ) . in fig .
[ fig : clscw ] and fig .
[ fig : hrscw ] , we present the same flux and hardness but detailed on time scales of a few days ( with a resolution of one scw ie @xmath0 2000 s ) .
these graphs illustrate clearly the variability of the source
. the hypothesis of constant flux has been tested by @xmath8 tests and is strongly rejected for all periods ( reduced @xmath8 always greater than 5.9 for a number of degrees of freedom ranging from 100 to 300 ) .
the variability appears to be chaotic but in both cases over a limited amplitude : the source is always detected and varies by a factor not greater than 3 .
we can recognize the usual temporal behavior of cyg x-1 ( see e.g. ling et al .
1987 and zdziarski et al .
note that the first period , which corresponds the lowest and softest state in our sample , has been analysed by malzac et al .
they identified the ( rather unusal ) source behavior as a mini or failed transition between soft and hard state , in a so called `` intermediate state '' .
however , no robust ( that is , lasting more than a few hours ) incursion in the soft state can be reported during our observations and we conclude that cyg x-1 was always in a hard ( lh ) or intermediate states . fig .
[ fig : hrflux ] displays the hardness as a function of the 30 - 70 kev source flux ( revolution averaged values ) .
the hardness intensity plane can be divided in three regions corresponding to the main clusters of data points .
those three regions are outlined in fig .
[ fig : hrflux ] .
the first two regions gather the points with hardnesses respectively below 0.24 and between 0.24 and 0.28 , and , incidentally , correspond to the first part of the integral mission ( revolutions 79 to 261 ie june
2003 - november 2004 ) .
+ in the third group , the flux levels span a broader range , extending toward higher values , but the mean hardnesses never decrease below 0.29 . during this period , which covers more than 3 years ( mid 2006 up to end of 2009 ) , the source evolution consists of an intensity variation within a factor of @xmath0 2 without any visible change in the emission hardness ( or spectral shape ) , so without notable change in the underlying processes .
this behavior has already been observed in this source and is well known in transient sources .
conversely , it is worth noting that different spectral shapes are observed for an unchanged intensity level in the 30 - 70 kev band . to illustrate both of these effects ,
we have superimposed in fig .
[ fig:5spec.var ] , five spectra representative of the global course followed by the source in fig .
[ fig : hrflux ] ( the corresponding revolutions are identified in this figure by squares , numbered from 1 to 5 ) .
we recognize the two modes of evolution already identified for this source on the ks timescale ( malzac et al 2006 ) : a pivoting of the spectral shape ( spectra 2 - 4 ) and a change of luminosity at constant hardness ( spectra 1 & 2 and 4 & 5 ) .
note however , that the pivot energy is here at @xmath0 45 kev ( in the middle of the studied energy band , by construction ) .
it could be attributed to an increase of both temperature and optical depth of the comptonising medium . when combined
, the two variability modes give rise to a global complex behavior in the hard x - ray domain , although the source remains in the hs ( and intermediate states ) .
to study in more details the spectral shape and its evolution in the high energy part , we have to accumulate data corresponding to the same hardness .
based on the three hardness levels repered in fig .
[ fig : hrflux ] , we have built the corresponding averaged spectra , hereafter refered to as low hardness , mid hardness and high hardness samples , respectively .
they are displayed on fig .
[ fig:3hr ] , while best fit parameters are given in table 3 .
similarly to the total spectrum , the data are modeled by a comptonization law plus its reflection component ( required when looking at residuals ) and a second hotter comptonization .
the evolution of the slope in the low energy part ( in direct relation with the hardness ) is clear , with an increase of the peak energy from @xmath0 50 kev to @xmath0 150 kev .
this behavior is not related to the reflection component but appears as an evolution of the macroscopic parameters ( kt and @xmath12 ) of the comptonizing medium , which both increase from the soft to the hard sample ( see table 3 ) .
+ looking now to the high energy part , no significant emission is detected above 700 kev .
a degeneracy between parameters being unavoidable , we choose to fix the second comptonizing medium temperature and optical depth to 130 kev and 0.6 respectively for all spectra , but free normalisation factors .
+ even though it represents only a possible description of our data , this model , involving a second ( hotter and thiner ) comptonization medium with constant parameters @xmath14 and @xmath12 , provides an acceptable interpretation of the cyg x-1 behavior in the hard x - ray domain . even if other models could reproduce the data , such two comptonizing regions ( or more generally , variable @xmath14 and/or @xmath12 ) scenario resembles those already applied in previous works to cyg x-1 and several bhbs
( see for example the suzaku observations analysed by makishima et al , 2008 , and references therein ) . beyond the specific sets of best fit parameter formally obtained , the inadequacy of the single comptonization region suggests spatial and/or rapid temporal variation of the temperature or optical depth of the comptonization region .
cygnus x-1 is one of the brightest galactic sources of hard x - rays up to several hundreds of kev and is therefore a good candidate for positron production .
the excellent energy resolution of the spi germanium detector makes it the best instrument to seek for potential annihilation signatures in the observed emission .
+ we have thus looked for any emission feature , narrow ( 10 kev fwhm ) or broad ( 80 kev fwhm ) , transient ( scw ie @xmath0 2 ks timescale ) or more persistent ( revolution ie 1 - 2 day timescale ) . on the scw timescale ,
no significant emission is reported with 2 @xmath15 upper limits of 2 - 3 @xmath16 and 0.5 - 1 @xmath17 for a narrow and broad line , respectively .
on the revolution timescale , no significant excess above the continuum contribution are found , with upper limits ranging between 3 - 6 @xmath18 and 0.7 - 1.1 @xmath16 according to the revolution duration .
finally , considering the whole set of data ( 4 ms ) , persistent emission features , if any , should be below 6 @xmath19 and 1.3 @xmath18 .
we now compare our observations with the available data in the mev region . from fig .
[ fig : spigro ] , we can see that our `` high hardness sample '' observations has a spectral shape that is very close to that of the hs mean spectrum reported by mcconnell et al .
( 2002 ) from gro/ comptel and osse instruments . to compare the whole set of data ( integral / spi + cgro / comptel+osse ) , we use a model consisting in a cutoff power law plus a broken power - law with the first index fixed to -1 ( to avoid any contribution in the low energy part ) .
+ imposing the same photon index for the cutoff power - laws , the fit procedure converges toward slightly different cutoff energies ( @xmath0 140 and 160 kev , for spi and osse data respectively , no contribution in comptel range ) .
we also impose a common second photon index in the broken power law and obtained 3.4 @xmath11 0.5 , while the energy break reflects the true difference between the spectra : close to 900 ( @xmath11 100 ) kev in the cgro data , it means that this component is based essentially on the comptel points . around 420 ( @xmath11 25 ) kev in the spi data , it allows us to recover the additional emission above the low energy component , the very soft index limiting its extension toward high energies .
+ in conclusion , even if the spi data do not show significant emission above 1 mev , the upper limits are marginally consistent with the non - thermal tail observed at several mev by cgro / comptel .
we note however that this emission is not taken into account by the two - zone thermal comptonization model proposed above which would require an additional component to fit the comptel data .
the overall good agreement between the spi and osse spectra indicate that the average spectral properties in the hard state have remained constant between the two epochs .
recently , laurent et al . (
2011 ) presented a stacked spectrum of cygnus x-1 obtained using the ibis data during nearly the same observation period as ours .
ibis is composed of two position sensitive detector layers isgri ( cdte , 15 - 1000 kev ; lebrun et al .
2003 ) , and picsit ( csi , 200 kev-10 mev ; labanti et al .
these two detectors are usually used independently to produce spectra and light curves .
laurent et al . ( 2011 ) used the ibis / isgri data up to @xmath0 400 kev and ibis / picsit at higher energies .
[ fig : laurent ] compares our averaged spi spectra with the results of laurent et al .
we note that most of the published results from ibis actually use only the isgri detector .
with the recent versions of the public software , the results from ibis / isgri are generally compatible with those of spi .
this good agreement is confirmed by our comparison of the stacked spi and ibis / isgri spectra . at higher energies however , the results are clearly different : the spi fluxes are lower than the ibis / picsit points by a factor of about 5 at least .
the origin of this disagreement is unclear .
we were not able to go further in the analysis of the discrepancy as there are very few published results from ibis / picsit .
we used 4 ms of _ integral _ spi data to study the emission of cyg x-1 from 20 kev to the mev region . while the source has been essentially detected in the hs , its presents complex variability on all timescales .
we have studied the luminosity and hardness evolution of the source above 20 kev , on the scw ( 2000 s ) and revolution ( 1 - 2 days ) timescales .
the revolution averaged data give a nice picture of the long term source behavior .
a change by a factor of 2 - 3 in luminosity can be accompanied either by a significant softening in the 20 - 150 kev domain ( see fig . [ fig:3hr ] ) or by the same spectral shape just moved up and down . then , in order to improve the photon statistics at high energies we combined observations to produce long exposure time average spectra .
the analysis of the stacked spectra reveals that the emission of cyg x-1 extends up to @xmath0 700 kev but presents a sharp cutoff around that energy .
whatever the criterion we used to built averaged spectra ( low / mid / high hardness , low / high intensity , all data ) , no emission can be detected above 1 mev .
+ nevertheless , a single comptonization model does not provide a good description of the overall spectral shape .
we have shown that a two temperature model provides a good fit of the spi data , with a minimum of free parameters ( @xmath12 and kt of the second compton component can be considered as constant along the time ) . in a final step to investigate the high energy emission , we compare all the data available above 300 kev for cyg x-1 in the last 2 decades and conclude that while our spi upper limits are marginally compatible with the soft powerlaw reported by comptel in the 90 ( mcconnell et al , 2002 ) , they clearly disagree ( lower by a factor @xmath0 5 ) with the recently reported ibis / picsit emission ( laurent et al , 2011 ) .
+ in other words , the presence of a non - thermal mechanism participating to the power supply can not be excluded , but our results contradict the presence of a hard ( polarised or not ) power law emission .
a last point concerns the potential emission linked to the positron production : no positive detection has been found , in a narrow ( 10 kev ) and broad ( 80 kev ) channel , on 2ks and day timescales , as well as in the total dataset .
the question of the origin and the nature of the hard state high energy emission remains thus of prime interest .
cyg x-1 is clearly the most adequate target to investigate it .
since no mission operating above 300 kev is expected for a while , the only hope is that the integral instruments will be able to accumulate a few more ms of data on this important target to resolve the matter once and for all .
the _ integral _ spi project has been completed under the responsibility and leadership of cnes .
we are grateful to asi , cea , cnes , dlr , esa , inta , nasa and ostc for support .
we thank a. a. zdziarski for providing us with gro / comptel and osse data .
79 - 80 ( 5x5 ) & 2003 - 06 - 07 00:59 & 2003 - 06 - 12 03:35 & 293 + 210 - 214 ( a ) & 2004 - 07 - 03 00:01 & 2004 - 07 - 17 00:25 & 709 + 251 - 252 ( a ) & 2004 - 11 - 03 14:23 & 2004 - 11 - 07 16:26 & 176 + 259 & 261 ( h ) & 2004 - 11 - 26 12:28 & 2004 - 12 - 03 15:43 & 143 + 470 ( exo , h ) & 2006 - 08 - 19 09:19 & 2006 - 08 - 21 16:02 & 159 + 486 ( exo , h ) & 2006 - 10 - 06 00:11 & 2006 - 10 - 08 07:55 & 160 + 498 - 505 ( gp ) & 2006 - 11 - 11 19:31 & 2006 - 12 - 04 06:20 & 535 + 628 - 631 ( a ) & 2007 - 12 - 04 19:05 & 2007 - 12 - 15 21:08 & 388 + 673 ( a ) & 2008 - 04 - 18 17:41 & 2008 - 04 - 19 22:09 & 54 + 682 - 684 ( a ) & 2008 - 05 - 14 08:13 & 2008 - 05 - 22 19:54 & 304 + 739 - 746 ( a ) & 2008 - 11 - 01 02:14 & 2008 - 11 - 24 05:25 & 551 + 803 - 806 ( a ) & 2009 - 05 - 11 08:27 & 2009 - 05 - 22 11:32 & 371 + 875(h * ) & 877(h ) & 2009 - 12 - 12 16:18 & 2009 - 12 - 19 20:57 & 160 + lcccccccc
[ tab : fittot ] refl*comptt & 0.90 @xmath11 0.3 & 75.0 @xmath11 3 & 0.91 @xmath11 0.03 & & & 6.0 ( 37 ) + refl*comptt + power law & 0.88 @xmath11 0.3 & 56 @xmath11 3 & 1.2 @xmath11 0.1 & 1.8 @xmath11 0.2 & & 5.1 ( 35 ) + refl*comptt+comptt & 0.8 ( fixed ) & 38 @xmath11 3 & 1.6 @xmath11 0.15 & 123@xmath11 10 & 0.5 ( fixed ) & 1.7 ( 36 ) + lcccccccccccccc [ tab : fit3hr ] low hardness sample & 0.55 @xmath11 0.25 & 33 @xmath11 3 & 1.2 @xmath11 0.1 & 130 ( fix ) & 0.6 ( fix ) & 0.99 ( 36 ) & + mid hardness sample & 0.73 @xmath11 0.15 & 36 @xmath11 3 & 1.4 @xmath11 0.2 & 130 ( fix ) & 0.6 ( fix ) & 2.13 ( 36 ) + high hardness sample & 0.68 @xmath11 0.15 & 40 @xmath11 3 & 1.6 @xmath11 0.1 & 130 ( fix ) & 0.6 ( fix ) & 1.4 ( 36 ) & + | we report on cyg x-1 observations performed by the spi telescope onboard the integral mission and distributed over more than 6 years .
we investigate the variability of the intensity and spectral shape of this peculiar source in the hard x - rays domain , and more particularly up to the mev region .
we first study the total averaged spectrum which presents the best signal to noise ratio ( 4 ms of data ) .
then , we refine our results by building mean spectra by periods and gathering those of similar hardness . several spectral shapes are observed with important changes in the curvature between 20 and 200 kev , even at the same luminosity level . in all cases ,
the emission decreases sharply above 700 kev , with flux values above 1 mev ( or upper limits ) well below the recently reported polarised flux ( laurent et al .
2011 ) , while compatible with the mev emission detected some years ago by cgro / comptel ( mcconnell et al . , 2002 ) .
finally , we take advantage of the spectroscopic capability of the instrument to seek for spectral features in the 500 kev region with negative results for any significant annihilation emission on 2 ks and days timescales , as well as in the total dataset . + |
electromagnetic transition form factor of light mesons is one of the key ingredients to study hadron properties and it recently gains intensive interests . there are fruitful updated measurements from different experimental collaborations , such as na60 @xcite , snd @xcite and babar @xcite . on the theoretical side ,
the transition form factor provides us an important tool to study the intrinsic properties of light hadrons , including light pseudoscalar mesons and vector resonances . of more importance
, it may also help to reduce the hadronic uncertainties in the light by light scattering , which is an important source of theoretical uncertainties of the muon anomalous magnetic moment @xcite .
the transition form factor of @xmath12 is one of the most important form factors in the light by light scattering .
physics involved in this kind of form factor is quite complicated , since one needs to handle different dynamics within a broad range of energy region . in the very high and low energy regions , we have reliable and model - independent theoretical tools , namely pqcd and chiral perturbation theory ( @xmath13pt ) . while this is not the case in the intermediate energy region , where various resonances enter . in the present work ,
we follow the chiral effective field theory explicitly including resonance states developed in ref .
@xcite to study the radiative transition form factors involved with vector resonances .
@xmath13pt is a model - independent method to describe the qcd dynamics in the very low energy region ( @xmath14 ) , which is based on chiral symmetry and expansions in terms of external momentum and light quark masses @xcite .
however the dynamical degrees of freedom in @xmath13pt are restricted to the light pseudo - goldstone bosons @xmath15 @xcite . in the intermediate energy region ( @xmath16 ) ,
clearly the resonance fields need to be explicitly included
@xcite proposed an approach to incorporate resonances in a chiral covariant way . in this theory , not only the chiral symmetry but also the qcd inspired high - energy behaviors at large @xmath17 are implemented , which makes the resonance chiral theory ( r@xmath13 t ) bear more properties of qcd .
moreover , to implement the high energy constraints in the chiral effective theory also makes it possible to apply the results from this theory directly to some form factors with virtual particles , such as those in light by light scattering , where the qcd high energy behavior can be important @xcite . at the practical level ,
to impose the high energy constraints is an efficient way to reduce the number of free resonance coupling , which makes r@xmath13 t more predictive in the phenomenological discussions @xcite . in our previous work
@xcite , we have performed an extensive study on the electromagnetic transition form factors and decays of light pseudoscalar mesons @xmath18 in the framework of r@xmath13 t . in the present work ,
we focus on the similar form factors and decays but involving light vector resonances .
the relevant resonance operators in these kinds of processes are of the odd - intrinsic - parity type . for the odd - intrinsic - parity sector ,
@xcite introduced a general effective chiral lagrangian containing symmetry allowed interactions between two vector objects ( currents or lowest multiplet of resonances ) and one pseudoscalar meson , by employing the antisymmetric tensor formalism as used in ref .
@xcite to describe the vector resonances .
while in ref .
@xcite , similar study was carried out but the vector resonances was described in terms of the proca vector field representation .
later on , the vector - vector - pseudoscalar ( @xmath19 ) type lagrangian with vector resonances in the antisymmetric tensor representation has been put forward in different aspects : ref .
@xcite introduced a second nonet of vector resonances , ref .
@xcite worked out the complete base of resonance operators that are relevant to the @xmath20 @xmath13pt lagrangian in the anomaly sector and in ref .
@xcite we have made a comprehensive discussion on the inclusion of the singlet @xmath21 . though we see impressive progresses in this research field , one still needs to bear in mind that in the strict large @xmath17 qcd there is an infinite tower of zero - width resonances . in practice
, typically one has to truncate the tower to the lowest multiplet of resonances , which is named as the minimal hadronical ansatz in ref . @xcite . under this approximation ,
r@xmath13 t has been successfully applied to the phenomenological study on many processes where the intermediate resonances play an important role @xcite .
however , we notice that in the previous study of r@xmath13 t , most efforts have been made on the lowest lying resonances . for example , in the vector sector , the lowest nonet of @xmath22 , @xmath23 , @xmath24 and @xmath25 has been extensively studied , while investigation on the higher mass resonances is relatively rare .
so one of the important improvements of our present work is to study the excited vector resonances in chiral effective theory , comparing with our previous work @xcite . on the experimental side ,
recently the babar collaboration has measured the @xmath26 cross sections from the threshold to the energy region around 2 @xmath27 3 gev @xcite , which enables us to study properties of more massive vector resonances , i.e. , @xmath7 , @xmath9 and @xmath8 . in this measurement , the moduli and relative phases of isoscalar and isovector components of the @xmath26 cross sections are provided .
the updated data make a strong constraint on the free parameters in our theory and hence allow us to accurately extract the resonance properties , such the masses and widths of the @xmath7 and @xmath8 ( and also @xmath9 ) , their mixings , as well as their couplings to light mesons .
the @xmath28 mixing angle is estimated and we find that it is far from the ideal mixing case .
another important issue we will address in this article is the @xmath10 symmetry breaking effect in radiative decays . an ideal process
to study this effect is @xmath29 and @xmath30 .
it is well known that the ratio @xmath31 in the @xmath10 limit @xcite , while the world average value from experimental measurements is around @xmath32 @xcite .
the large @xmath10 symmetry breaking effect has been discussed in many previous works @xcite . in the framework of r@xmath13 t
, it is interesting to point out that there exists a special resonance operator @xmath33{\rangle}$ ] @xcite , which contributes exclusively to the charged processes @xmath30 .
hence it provides an important source of @xmath10 symmetry breaking .
in fact , ref @xcite has determined several values for the @xmath34 parameter in the discussion of hadronic @xmath35 decays .
our present work provides another way to determine its value and allows us to check which kinds of values from ref .
@xcite are reasonable
. a careful analysis of the strengths of different @xmath10 breaking terms @xmath36 and @xmath37 will be delivered in our work .
contrary to the @xmath38 process , the @xmath39 form factor is free of of large @xmath10 corrections .
nevertheless , another difficulty arises in this form factor , since it is found that the well - established vector - meson - dominant ( vmd ) model fails to describe @xmath40 @xcite . in the present work
, we discuss this transition form factor in r@xmath13 t by including excited vector resonances in addition to the @xmath22 .
we will also confront our theoretical results with the new measurements from snd collaboration @xcite .
this paper is organized as follows . in sec .
[ theor ] , we introduce the theoretical framework and elaborate the calculations for @xmath41 and @xmath26 , the @xmath42 transition form factor and the spectral function for @xmath43 . in sec .
[ disc ] , we present the fit results and discuss the @xmath10 symmetry breaking mechanism and the @xmath6 form factors in detail .
summary and conclusions are given in sec .
[ conclu ] .
the low - energy ( @xmath44 mev ) dynamics of qcd is ruled by the interaction of the octet of pseudoscalar mesons , which are characterized by the spontaneous breaking of chiral symmetry .
the remarkably successful @xmath13pt @xcite describes the strong interactions among pseudoscalar mesons in a model - independent way .
the effective lagrangian to lowest order , @xmath45 , is given by @xmath46 where @xmath47 , \nonumber\end{aligned}\ ] ] @xmath48 the unitary matrix in flavour space @xmath49 incorporates the pseudo - goldstone octet : @xmath50 the external sources @xmath51 and @xmath52 promote the global symmetry in the flavor space to a local one .
the parameter @xmath53 corresponds to the pion decay constant @xmath54 mev in the chiral limit and @xmath55 is related to the quark condensate via @xmath56 $ ] , being @xmath57 the light quark mass .
the explicit chiral symmetry breaking is implemented in @xmath13pt by taking the vacuum expectation values of the scalar sources as @xmath58diag@xmath59 .
we work in the isospin limit throughout , i.e. taking @xmath60 . in a chiral covariant way
, the lowest multiplet of vector meson resonances was explicitly included in ref .
@xcite in terms of antisymmetric tensor fields . in this formalism
, the kinetic term of the vector resonance lagrangian takes the form @xmath61 with the ground vector nonet @xmath62 and the covariant derivative and the chiral connection defined as @xmath63,\hspace{1.5 cm } \gamma_{\mu}=\frac{1}{2}\{u^{+}(\partial _ \mu - ir_\mu)u+u(\partial _ \mu - il_\mu)u^{+}\}.\end{aligned}\ ] ] at leading order of @xmath64 , the mass splitting of the ground vector multiplet is governed by a single resonance operator @xcite @xmath65 in ref .
@xcite , it has been demonstrated that this single operator can well explain the mass splittings of @xmath22 , @xmath66 and @xmath25 .
in addition to the mass splittings of these vector resonances , this operator is also shown to be able to perfectly reproduce the quark mass dependences of the @xmath22 mass from the lattice simulation @xcite .
we will then employ the single operator in eq . to account for the mass splittings of the lowest vector multiplet in this work .
additional @xmath64 suppressed operators , such as @xmath67 , will not be considered for the well - established ground vectors , due to the fact that @xmath23 and @xmath25 in the ground multiplet are well described by the ideal mixing between the octet and singlet vector states .
while because of the unclear situation for the excited vectors , the @xmath64 suppressed operators will be however introduced for them in later discussion .
combining eqs . and , it can be easily verified that the physical states of @xmath23 and @xmath25 result from the ideal mixing of @xmath68 and @xmath69 @xmath70 then the mass splitting pattern of the ground vector multiplet takes the form @xmath71 where we have used the leading order relations @xmath72 and @xmath73 . for the general interaction lagrangian linear in @xmath74 up to @xmath45 , it reads @xcite @xmath75 @xmath76 with @xmath77 the field strength tensors of the left and right external sources @xmath78 and @xmath79 , respectively .
@xmath80 are real resonance coupling constants and only @xmath81 is relevant in our current study .
the interaction operators containing two - vector and one - pseudoscalar objects have been worked out in ref .
@xcite @xmath82{\rangle}+\frac{c_5}{m_v}\epsilon_{\mu\nu\rho\sigma}{\langle}\{\nabla_\alpha v^{\mu\nu},f_+^{\rho\alpha}\}u^\sigma{\rangle}+\frac{c_6}{m_v}\epsilon_{\mu\nu\rho\sigma}{\langle}\{\nabla_\alpha v^{\mu\alpha},f_+^{\rho\sigma}\}u^\nu{\rangle}\nonumber\\ & & + \frac{c_7}{m_v}\epsilon_{\mu\nu\rho\sigma}{\langle}\{\nabla^\sigma v^{\mu\nu},f_+^{\rho\alpha}\}u_\alpha{\rangle},\end{aligned}\ ] ] @xmath83 with @xmath84 the levi - civita antisymmetric tensor .
similar operators with an excited vector multiplet @xmath85 can be straightforwardly constructed @xcite @xmath86 @xmath87 the kinetic lagrangian for the excited vector takes the same form as that in eq . .
the equivalent operators to eq .
involving the excited vector multiplet @xmath85 happen to be irrelevant to the discussions in the present article .
for latter convenience , we follow the convention introduced in ref .
@xcite to define certain combinations of @xmath88 in eq .
@xmath89 in the present article , since we focus on the processes involving pion and kaon , the additional operators that we introduced in ref .
@xcite to discuss @xmath90 and @xmath91 will be irrelevant .
so we do not elaborate more on this issue .
the flavor structure for the excited vector multiplet @xmath85 is the same as the ground multiplet in eq . . while unlike the well - established entries in the ground vector multiplet , the contents of the excited vector multiplet are still not clear
. therefore we include a general set of mass splitting operators to describe the excited vectors , instead of only using a single operator in the ground multiplet case . up to the linear quark
mass corrections , the pertinent operators read @xcite @xmath92 where the second operator is @xmath64 suppressed compared to the first one and @xmath93 is the mass of excited vector multiplet in chiral limit .
another type operator , e.g. @xmath94 , was also introduced in ref .
@xcite , but we consider this operator is less relevant compared to the two terms in eq . .
the reason is that compared to the @xmath95 term though the @xmath96 term is @xmath64 suppressed , its chiral order is enhanced . while for the @xmath97 term , its chiral and @xmath64 orders are both suppressed .
owing to the @xmath64 suppressed operator in eq . , the excited states @xmath9 and @xmath8 can not be simply described as the ideal mixing of @xmath98 and @xmath99 as in the ground state case .
the physical states of @xmath9 and @xmath8 are related to the octet - singlet basis through @xmath100 the mixing angle @xmath101 and the masses of @xmath9 and @xmath8 can be expressed in terms of the couplings in lagrangian @xmath102 with @xmath103 while for the masses of @xmath7 and @xmath104 , they are only contributed by the @xmath95 term in eq . and
take the same form as the ground states in eq .
@xmath105 to include the quark mass corrections to the resonance masses in the lagrangian approach allows us to incorporate the @xmath10 symmetry breaking effects in a systematic way .
this is important to understand the @xmath10 symmetry breaking mechanism in the @xmath106 transitions .
not only the quark mass corrections to the resonance masses but also to interacting vertexes are important to fully study the @xmath10 symmetry breaking effects .
the @xmath19-type lagrangians introduced in eqs . , which are constructed by using the chiral building bricks of the pseudo - goldstone mesons , provides us an appropriate framework to systematically take into account the relevant interacting vertexes .
though it consists of a large number of free couplings , we will see later that in a given process not all of them are independent .
in fact in our present discussion , at most only six of the combinations of the @xmath19-type couplings will appear after taking into account the high energy constraints dictated by qcd . to impose the high energy constraints will not only make the r@xmath13 t bear more properties from qcd , but will also be helpful to reduce the number of free parameters . to proceed , we match the leading operator product expansion ( ope ) of the @xmath19 green function with the result evaluated within r@xmath13 t and require the vector form factor vanish in the high energy limit .
this leads to the following constraints on resonance couplings @xmath107 which are already given by one of us in ref .
we point out that the previous constraints are obtained in the chiral and large @xmath17 limits , as done in most of the r@xmath13 t study @xcite in the literature . at leading order of @xmath64 in r@xmath13 t ,
the @xmath110 transitions receive two types of contributions as displayed in fig .
[ figvpgamma ] , namely the direct type depicted by the diagram ( a ) and the indirect one by the diagram ( b ) .
the structure of the amplitude for the @xmath111 process can be written as @xmath112 where @xmath113 and @xmath114 denote the four momenta of @xmath115 and @xmath116 respectively , @xmath117 and @xmath118 correspond to the polarization vectors in that order ; @xmath119 stands for the electric charge of a positron . in the previous equation
, we collect the contributions from the direct and indirect diagrams of fig .
[ figvpgamma ] in the quantities @xmath120 and @xmath121 respectively . by using the previously introduced resonance lagrangian in sec .
[ sect.lag ] , it is straightforward to calculate @xmath120 and @xmath121 : @xmath122\ , , \nonumber\\ g_{k^{\ast0}k^{0}\gamma}^{indirect}&=&-\frac{2 f_v}{3f_k m_{k^ * } } ( \frac{1}{m_\omega^2 } -\frac{3}{m_\rho^2}-\frac{2}{m_\phi^2 } ) [ ( d_1 + 8d_2-d_3)m_k^2 + d_3 m_{k^*}^2 ] \ , , \nonumber\\ & & -\frac{f_{v_1}}{3f_k m_{k^ * } } \bigg [ \frac{(2\sqrt{2}\cos\theta_v-\sin\theta_v)\sin\theta_v}{m_{\omega'}^2 } -\frac{3}{m_{\rho'}^2 } -\frac{(\cos\theta_v+2\sqrt{2}\sin\theta_v)\cos\theta_v}{m_{\phi'}^2 } \bigg ] \nonumber\\ & & ( d_m m_k^2+d_m m_{k^*}^2)\ , , \label{eqkstar0k}\end{aligned}\ ] ] @xmath123 \ , , \nonumber\\ g_{k^{\ast+}k^{+}\gamma}^{indirect}&=&-\frac{2 f_v}{3f_k m_{k^ * } } ( \frac{1}{m_\omega^2 } + \frac{3}{m_\rho^2}-\frac{2}{m_\phi^2 } ) [ ( d_1 + 8d_2-d_3)m_k^2 + d_3 m_{k^*}^2]\nonumber\\ & & -\frac{f_{v_1}}{3f_k m_{k^ * } } \bigg [ \frac{(2\sqrt{2}\cos\theta_v-\sin\theta_v)\sin\theta_v}{m_{\omega'}^2 } + \frac{3}{m_{\rho'}^2 } -\frac{(\cos\theta_v+2\sqrt{2}\sin\theta_v)\cos\theta_v}{m_{\phi'}^2 } \bigg]\nonumber\\ & & ( d_m m_k^2+d_m m_{k^*}^2)\ , . \label{eqkstarpk}\end{aligned}\ ] ] physical meson masses are used in the kinematics .
the kaon decay constant @xmath124 has been employed in the amplitude , instead of the pseudo - goldstone decay constant @xmath53 in chiral limit that appears in the resonance chiral lagrangian .
we will take @xmath125 gev from pdg @xcite . within our formalism
, the @xmath126 decay width is given by @xmath127 with @xmath128 the fine structure constant . in the leading order of large @xmath17 ,
the @xmath26 transitions also receives two types of contributions , i.e. direct and indirect types , as displayed in fig .
[ figeevp ] the main difference between @xmath26 and @xmath129 is that now we have an off - shell photon .
the final results for the amplitudes of @xmath130 can be obtained in the same way as the @xmath129 case . in order to confront our theoretical results with the experimental data
, we need to work in the isospin bases for the amplitudes @xmath131 - 4f_{v}\bigg [ \frac{1}{3(m_{\omega}^2-s -im_{\omega}\gamma_{\omega } ) } -\frac{2}{3(m_{\phi}^2-s - im_{\phi}\gamma_{\phi } ) } \bigg ] \nonumber \\ & & [ ( d_1 + 8d_2)m_k^2+d_3(m_{k^{\ast}}^2+s - m_k^2 ) ] -2f_{v_1 } \bigg [ \frac{(2\sqrt{2}\cos\theta_v-\sin\theta_v)\sin\theta_v } { 3(m_{\omega'}^2-s - im_{\omega'}\gamma_{\omega'}(s ) ) } \nonumber \\ & & -\frac{(\cos\theta_v+2\sqrt{2}\sin\theta_v)\cos\theta_v } { 3(m_{\phi'}^2-s - im_{\phi'}\gamma_{\phi'}(s ) ) } \bigg ] ( d_m m_k^2+d_m m_{k^{\ast}}^2+d_s s ) \bigg\}\nonumber \\ \triangleq&&i\epsilon_{\alpha\beta\rho\sigma}\epsilon_{k^\ast}^{\alpha}\epsilon_{\gamma}^{\beta}q^{\rho}k^{\sigma}ea_0,\end{aligned}\ ] ] @xmath132 + 4f_{v } \frac{1}{m_{\rho}^2-s - im_{\rho}\gamma_{\rho}(s ) } [ ( d_1 + 8d_2)m_k^2\nonumber \\ & & + d_3(m_{k^{\ast}}^2+s - m_k^2 ) ] + 2f_{v_1 } \frac{1}{m_{\rho'}^2-s - im_{\rho'}\gamma_{\rho'}(s ) } ( d_m m_k^2+d_m m_{k^{\ast}}^2+d_s s ) \bigg\}\nonumber \\ \triangleq&&i\epsilon_{\alpha\beta\rho\sigma}\epsilon_{k^\ast}^{\alpha}\epsilon_{\gamma}^{\beta}q^{\rho}k^{\sigma}ea_1\,,\end{aligned}\ ] ] where @xmath113 and @xmath114 stand for the momenta of @xmath115 and @xmath133 respectively and @xmath134 is the energy square of the @xmath135 system in center of mass frame .
the energy - dependent decay widths for intermediate resonances , such as @xmath136 and @xmath137 , can be important , since the widths of the relevant resonances are typically not so narrow and the final results can be sensitive to the off - shell width effects . to rigorously include the off - shell width effects one needs to carry out the next - to - leading - order of @xmath64 computation in r@xmath13 t , which is beyond the scope of this work , since we focus on the calculations of the various amplitudes at leading order of @xmath64 .
nevertheless , in order to quantitatively estimate to what extent the off - shell width effects will affect the final outputs , we will study two different forms of the energy - dependent widths , which will be discussed in detail in next section .
we see that only the @xmath138like and @xmath139like intermediate mesons contribute to the @xmath140 and only the @xmath141like intermediate mesons contribute to the @xmath142 . while for the local terms or the so - called background terms in many other works @xcite
, they are obtained from a lagrangian based theory in this work , not simply introduced by hand .
the cross sections are related to the isospin amplitudes through @xmath143 where the masses of electron and positron have been neglected .
the amplitude for the radiative decay @xmath146 can be written as : @xmath147 where @xmath134 , @xmath148 and @xmath149 denote the polarization vectors of the @xmath150 resonance and the off - shell photon respectively .
the explicit expression of the electromagnetic transition form factor @xmath151 is @xmath152 \nonumber \\ & & + \frac{4 f_v}{f_\pi m_\omega } \ , \frac{1}{m_{\rho}^2-s - im_{\rho}\gamma_{\rho}(s ) } \ , \big [ ( d_1 + 8d_2-d_3)m_\pi^2 + d_3 ( m_\omega^2 + s ) \big ] \nonumber \\ & & + \frac{2 f_{v_1}}{f_\pi m_\omega } \ , \frac{1}{m_{\rho'}^2-s - im_{\rho'}\gamma_{\rho'}(s ) } \ , \big [ d_m m_\omega^2 + d_s s + d_m m_\pi^2 \big]\,.\end{aligned}\ ] ] in the phenomenological discussion , it is common to normalize the form factor @xmath151 by its value at @xmath153 .
so the interesting quantity that we will study later is @xmath154 the charged @xmath6 form factor or the @xmath6 spectral function can be measured in the semileptonic decays of the @xmath35 lepton . in the eq.(35 ) of ref .
@xcite , one of us has calculated the @xmath43 spectral function .
we give the explicit expression here for completeness @xmath155^{3/2}\nonumber\\ & & \times \bigg|(2d_3 f_v+d_s f_{v_1})\frac{m_\omega^2}{m_v^2}+f_{v_1}(d_m m_\pi^2+d_m m_\omega^2 + d_s s)\frac{1}{m_{\rho'}^2-s - im_{\rho'}\gamma_{\rho'}(s)}\nonumber\\&&+2f_v[(d_1 + 8d_2)m_\pi^2 + d_3 ( s+m_\omega^2-m_\pi^2)]\frac{1}{m_{\rho}^2-s - im_{\rho}\gamma_{\rho}(s)}\bigg|^2,\end{aligned}\ ] ] where the electroweak correction factor @xmath156 has been analyzed in @xcite , and its value will be taken as @xmath157 .
the experimental data that we consider here consist of eight different types : @xmath158 @xcite , @xmath159 @xcite , the @xmath160 cross sections @xcite , the moduli and relative phases of isoscalar and isovector components of the @xmath26 cross sections @xcite , the @xmath161 transition form factor @xcite and the @xmath43 spectral function @xcite . for the @xmath162 data in ref .
@xcite , since in the energy region above 1.8 gev the contributions coming from other excited vector resonances with higher mass and spin , certainly affect the @xmath163 channel , we only fit the data below 1.8 gev . for the @xmath9 [ @xmath164
, we fix its mass and width at their world average values , @xmath165=1.42 gev and @xmath166gev @xcite , because its mass is quite close to the @xmath167 threshold and its contribution only marginally shows up in the left tale of the @xmath162 cross sections . due to the fact that the masses of the ground vector resonances , @xmath168 , can be perfectly described by eq . ,
see refs .
@xcite , we simply employ the physical masses of @xmath168 in the numerical discussions , instead of fitting the @xmath169 coupling .
we point out that if the value of @xmath169 is fitted , it turns out to be very close to the values given in refs .
@xcite . for the mass of the ground vector in chiral limit
, we take the value @xmath170 mev from ref . @xcite . about the value of @xmath81
, it has been accurately determined by taking into account a large amount of experimental data in our previous work @xcite .
hence we will take the value from that reference in this work , which is @xmath171 gev . for @xmath172 in eq .
, since what appears in our discussion is always the combination of @xmath173 or @xmath174 , which is already noticed in ref .
@xcite , we can fix the sign of @xmath172 and leave @xmath175 and @xmath176 for free . in ref .
@xcite @xmath177 gev has been determined and we will take this value in our current study . as mentioned previously ,
we introduce the finite - width effects to the intermediate resonances . to systematically include the finite - width contributions in r@xmath13 t
, one has to step into the next - to - leading order of @xmath64 computation , e.g the one - loop calculations , which is far beyond the scope of our study .
nevertheless , we will employ two different parameterizations for the energy - dependent widths for the broad resonances in order to see how important the off - shell width effects will affect the fit results .
for the narrow - width resonances , such as @xmath150 and @xmath178 , we simply use the constant widths in their propagators . for the first case , which will be referred as fit i in later discussion
, we construct the energy - dependent widths for broad resonances following the approach given in ref .
@xcite , where the authors have taken the chiral symmetry , analyticity and unitarity into account when discussing the form of the off - shell width for the @xmath22 resonance .
the result of the previous reference has been generalized to other resonances in many phenomenological discussions @xcite . in this formalism
, we construct the energy - dependent widths in the following way : @xmath179\ , , \nonumber \\
\gamma_{\rho'}(s)&=&\gamma_{\rho'}\frac{s}{m_{\rho'}^2}[\frac{\sigma_{\pi\pi}^3(s)+\frac{1}{2}\sigma_{kk}^3(s ) } { \sigma_{\pi\pi}^3(m_{\rho'}^2)+\frac{1}{2}\sigma_{kk}^3({m_{\rho'}}^2)}]\ , , \nonumber \\
\gamma_{\omega'}(s)&=&\gamma_{\omega'}\frac{s}{m_{\omega'}^2}[\frac{\sigma_{\rho\pi}^3(s ) + \frac{(2\sqrt{2}\cos\theta_v-\sin\theta_v)^2}{3(\sqrt{2}\cos\theta_v+\sin\theta_v)^2}\sigma_{k^{\ast}k}^3(s ) } { \sigma_{\rho\pi}^3({m_{{\omega'}}^2 } ) + \frac{(2\sqrt{2}\cos\theta_v-\sin\theta_v)^2}{3(\sqrt{2}\cos\theta_v+\sin\theta_v)^2 } \sigma_{k^{\ast}k}^3(m_{{\omega'}}^2)}]\ , , \nonumber \\
\gamma_{\phi'}(s)&=&\gamma_{\phi'}\frac{s}{m_{\phi'}^2}[\frac{\sigma_{k^{\ast}k}^3(s ) + \frac{4(\sqrt{2}\cos\theta_v+\sin\theta_v)^2}{3(\cos\theta_v+2\sqrt{2}\sin\theta_v)^2}\sigma_{\phi\eta}^3(s ) } { \sigma_{k^{\ast}k}^3({m_{{\phi'}}^2 } ) + \frac{4(\sqrt{2}\cos\theta_v+\sin\theta_v)^2}{3(\cos\theta_v+2\sqrt{2}\sin\theta_v)^2 } \sigma_{\phi\eta}^3(m_{{\phi'}}^2)}]\,,\end{aligned}\ ] ] where @xmath180 with @xmath181 the standard step function . in fit ii
, we take a different asymptotic behavior for the energy - dependent widths as used in fit i @xmath182 with @xmath183 defined in eq . .
in the really asymptotic region when @xmath184 , the power of the energy - dependent width should be less than one , in such a way that the phase of the amplitudes in eqs . and would approach to @xmath185 . in fit
i , we adopt the results from ref .
@xcite , where only @xmath186 and @xmath187 channels are considered when deriving the finite width for @xmath22 . in principle
infinite number of channels will be open when @xmath184 , which could finally alternate the asymptotic behaviors obtained by considering finite number of channels .
a definite answer to this problem clearly deserves an independent work . as a phenomenological study
, we will exploit both forms to perform the fits and examine to what extent different forms can affect the final outputs .
.[tablepar ] the parameters result from fit i and fit ii . [ cols="^,^,^",options="header " , ] the values for the fit parameters are summarized in table [ tablepar ] .
the results of @xmath188 , @xmath159 and @xmath189 from the fits are given in table [ tablekstark ] .
the final results corresponding to the @xmath26 processes and the @xmath6 spectral function from @xmath43 are shown in figs .
[ figkstark ] and [ figvtauomegapi ] , respectively . in these two figures ,
we distinguish different results from fit i and fit ii by solid and dashed lines , respectively .
the resulting curves for the @xmath190 form factors are plotted in fig .
[ figffomega ] , together with the results from refs . @xcite . some comments about the fitting results are given below .
firstly , we observe that fit i is slightly preferred over fit ii , according to the values of @xmath191degrees of freedom ( d.o.f ) .
nevertheless , due to the fact that the values of @xmath192 are already bigger than 2 , the preference is not really significant .
regarding to the somewhat large values of @xmath192 , we find that they are mainly contributed by the @xmath193 form factors from the na60 and lepton - g measurements , especially from the data points in the energy region around @xmath194 gev .
therefore we have tried another kind of fits to exclude the data from these two measurements .
it turns out that the resulting values of @xmath192 decrease to around 1.2 in fit i and to 1.9 in fit ii and we do not observe significant changes of the fit parameters comparing with the results in table [ tablepar ] .
later we will come back to analyze the @xmath193 form factor in detail .
secondly , regarding to the @xmath195 and @xmath88 parameters given in table [ tablepar ] , we see that the resulting values from the two fits are more or less compatible .
the situation for the masses and widths of broad resonances is different . roughly speaking , we observe that the determinations of the masses and widths for @xmath8 from the two fits are consistent , or at least not so different from each other .
this conclusion also holds roughly for the masses of @xmath104 .
nevertheless , we find the masses and widths of @xmath7 from the two fits are clearly not compatible , indicating that the mass and width of the @xmath7 are more sensitive to the forms of energy - dependent widths , compared to the @xmath8 case . focusing on the mass and width of @xmath8 , our determinations in table [ tablepar ]
are in good agreement with the corresponding parameters of @xmath196 reported in pdg @xcite .
though the mass of @xmath8 in ref .
@xcite is compatible with pdg , its width is somewhat larger .
in addition , the uncertainties in our determination of the mass and width for @xmath8 are much smaller than those obtained in ref . @xcite .
furthermore , we do not need to add the so - called background terms by hand in our amplitude , while this is not the case in the experimental analyses @xcite .
all of the terms appearing in our amplitudes are obtained from the symmetry allowed operators illustrated in sect .
[ sect.lag ] , where the contact terms and contributions from @xmath197 and @xmath198 are incorporated in a consistent way . regarding to the @xmath7 resonance
, we should mention that the spectra of the excited @xmath199 resonances are still not clear in pdg @xcite .
the value of the @xmath7 mass from the recent determination of snd collaboration @xcite is around 1490 mev , which is clearly lower than our determinations in table [ tablepar ] .
as mentioned previously , we find that the resulting @xmath7 masses from our fits vary accordingly by using different forms of energy - dependent widths . in ref .
@xcite , a constant width is used for the @xmath7 resonance , which is obviously different from ours .
we have explicitly checked that if a constant width for the @xmath7 is used our determination for its mass decreases to around 1500 mev in both fits , which becomes close to the value in ref .
the masses and widths of @xmath7 from our fits lie between @xmath200 and @xmath201 .
notice that there are roughly two peaks in the experimental isovector component of the @xmath202 cross sections , see fig .
[ figkstark](c ) , which may be due to the contributions of @xmath200 and @xmath201 .
however since the statistics are still very poor , we only include one set of the @xmath7 resonances in the current fit .
this indicates that the @xmath7 here may be regarded as a combined effect of @xmath200 and @xmath201 .
finally , we find the mixing angles of @xmath28 from the two fits are roughly consistent .
the resulting angle is around @xmath203 and it is clearly different from the ideal mixing case .
this indicates that the @xmath64 suppressed operator in eq .
is quite important in the determination of the masses of excited vector resonances , since the ideal mixing will result if only the leading order operators at large @xmath17 are included for the resonance masses . for the ratio @xmath204 ,
the @xmath10 symmetry tells us that @xmath205 @xcite , while the experimental value is @xmath206 @xcite .
the @xmath10 symmetry breaking effect in the @xmath108 decay has been discussed in many previous works @xcite . in the context of vector meson dominance ( vmd )
model @xcite , ref .
@xcite points out that there is a significant contribution from an intermediate @xmath207 pair ( @xmath178 meson ) in these decays , and if all partial @xmath207 production is suppressed by only ( 20 - 30)% relative to those involving nonstrange pairs , the ratio of the neutral to charged @xmath208 decay widths would be brought into line with the data .
including the @xmath10 symmetry breaking in the hidden local symmetry model @xcite , it is found that the rate of @xmath36 is reduced comparing with the @xmath10 symmetry prediction and especially it is pointed out that the vector mesons should be `` renormalized '' @xcite . in the framework of r@xmath13 t
, there is a special operator in eq . , namely the @xmath209{\rangle})$ ] term , which breaks the @xmath10 symmetry by the factor @xmath210 and exclusively contributes to the charged processes @xmath211
. hence the value of the coupling @xmath34 is important to decode the @xmath10 symmetry mechanism in the present work .
in fact , the @xmath34 value has been determined in ref .
@xcite . in that reference ,
the authors only included the lightest multiplet of vector resonances , so the heavier degrees of freedom , such as the excited @xmath7 state , are implicitly included in their @xmath34 .
through fitting the first 6 data points of the isovector component of @xmath212 , they get @xmath213 and through fitting the branching ratios of @xmath214 they obtain @xmath215 .
clearly both values determined in ref .
@xcite are not compatible with our results in table [ tablepar ] and the magnitudes of @xmath34 in our fits are around one order smaller than those in ref .
@xcite .
we would like to bring the attention that in ref .
@xcite an unexpected phenomenon is observed by using @xmath216 in the radiative decay of @xmath217 .
typically one would expect that for a low energy cutoff of the photon in @xmath217 , the decay rate should be dominated by the internal bremsstrahlung ( ib ) contribution .
for example , to take the cutoff for the photon energy at 50 mev , the ib part contributes more than 90% of the decay rate for the @xmath218 process , see table i in ref .
however if one looks at table ii in the same reference , the contribution from the ib part in @xmath217 dramatically decreases to 6% by taking @xmath216 and in this case the @xmath34 term dominates the whole process . in the case with @xmath219 , one can see some reasonable results .
this clearly gives us a hint that the large magnitude of @xmath34 ( compared to 0.07 ) , does not lead to reasonable results in the radiative tau decays .
therefore our current determination for @xmath34 , around one order smaller in magnitude than the values given in ref .
@xcite , seems more meaningful .
however due to the lack of experimental measurements on the radiative tau decays , we can still not make a concrete conclusion whether the @xmath34 values in table [ tablepar ] will reasonably describe the @xmath217 process .
a future measurement on this channel is definitively helpful to answer this question . in order to have a better understanding of the origin of the differences for the @xmath34 values between ours and ref .
@xcite , we next employ the same @xmath19 lagrangian as used in ref .
@xcite , indicating in the discussions below we exclude the excited vector resonances in our theory . in this case
, it is interesting to point out that after taking into account the high energy constraints given in the previous work @xcite @xmath220 we can totally predict the @xmath29 decay width , which turns out to be @xmath221 kev and agrees well with the experimental value @xmath222 kev @xcite .
for the charged process
@xmath223 , we only have one free parameter , @xmath34 , to determine its decay width .
therefore , one could use the experimental information @xmath224 kev to determine the value of @xmath34 .
we then obtain two solutions @xmath225 by setting @xmath226 in the decay amplitudes , we find the @xmath34 term then dominates the @xmath227 process , which indicates that the @xmath10 symmetry breaking effect overwhelmingly controls this process and clearly contradicts a general rule that this effect should be at most around 30% .
this tells us only the solution of @xmath228 in eq .
corresponds to the physical one .
indeed this observation also confirms the conclusion about the role of @xmath34 in radiative tau decays @xcite , where a small magnitude of @xmath34 is preferred . in the determination of eq .
, we do not explicitly include the vector resonance excitations and the corresponding results after the incorporation of the excited states can be found in table [ tablepar ] , which magnitude of @xmath34 is still one order smaller than that in ref .
@xcite . in the present work we provide another important and easy way to determine the @xmath34 value , comparing with ref .
@xcite , which can be useful for the future study of various tau decays .
in order to have a clear idea about how different parts contribute to the @xmath10 breaking , we decompose the amplitudes @xmath229 and @xmath230 in eqs . and into the two parts : the @xmath10 symmetric term and @xmath10 breaking term .
the decomposition takes the form @xmath231 \nonumber\\ & & -\frac{2 f_v}{3f_k m_{k^ * } } \bigg ( \frac{1}{m_v^2 } -\frac{3}{m_v^2}-\frac{2}{m_v^2 } \bigg)\bigg [ ( d_1 + 8d_2-d_3)m_k^2 + d_3 m_{k^*}^2 \bigg]\nonumber\\ & & -\frac{f_{v_1}}{3f_k m_{k^ * } } \bigg(\frac{1}{m_{v_1}^2 } -\frac{3}{m_{v_1}^2}-\frac{2}{m_{v_1}^2 } \bigg ) ( d_m m_k^2+d_m m_{k^*}^2 ) , \nonumber\\ & & g_{k^{\ast0}k^{0}\gamma}^{su(3 ) breaking}= -\frac{2 f_v}{3f_k m_{k^ * } } \bigg [ \bigg ( \frac{1}{m_\omega^2 } -\frac{3}{m_\rho^2}-\frac{2}{m_\phi^2 } \bigg ) \nonumber \\ & & - \bigg ( \frac{1}{m_v^2 } -\frac{3}{m_v^2}-\frac{2}{m_v^2 } \bigg ) \bigg ] \bigg [ ( d_1 + 8d_2-d_3)m_k^2 + d_3 m_{k^*}^2 \bigg ] \nonumber\\ & & -\frac{f_{v_1}}{3f_k m_{k^ * } } \bigg\ { \bigg [ \frac{(2\sqrt{2}\cos\theta_v-\sin\theta_v)\sin\theta_v}{m_{\omega'}^2 } -\frac{3}{m_{\rho'}^2 } -\frac{(\cos\theta_v+2\sqrt{2}\sin\theta_v)\cos\theta_v}{m_{\phi'}^2 } \bigg ] \nonumber\\ & & -\bigg(\frac{1}{m_{v_1}^2 } -\frac{3}{m_{v_1}^2}-\frac{2}{m_{v_1}^2 } \bigg ) \bigg\ } ( d_m m_k^2+d_m m_{k^*}^2)\ , , \label{eqkstar0ksu3symplusbreaking}\end{aligned}\ ] ] @xmath232 \nonumber\\ & & -\frac{2 f_v}{3f_k m_{k^ * } } \bigg ( \frac{1}{m_v^2 } + \frac{3}{m_v^2}-\frac{2}{m_v^2 } \bigg ) \bigg [ ( d_1 + 8d_2-d_3)m_k^2 + d_3 m_{k^*}^2\bigg]\nonumber\\ & & -\frac{f_{v_1}}{3f_k m_{k^ * } } \bigg ( \frac{1}{m_{v_1}^2 } + \frac{3}{m_{v_1}^2}-\frac{2}{m_{v_1}^2 } \bigg ) ( d_m m_k^2+d_m m_{k^*}^2 ) , \nonumber\\ & & g_{k^{\ast+}k^{+}\gamma}^{su(3 ) breaking}= -\frac{2 f_v}{3f_k m_{k^ * } } \bigg [ \bigg ( \frac{1}{m_\omega^2 } + \frac{3}{m_\rho^2}-\frac{2}{m_\phi^2 } \bigg ) \nonumber \\ & & -\bigg ( \frac{1}{m_v^2 } + \frac{3}{m_v^2}-\frac{2}{m_v^2 } \bigg ) \bigg ] [ ( d_1 + 8d_2-d_3)m_k^2 + d_3 m_{k^*}^2]\nonumber\\ & & -\frac{f_{v_1}}{3f_k m_{k^ * } } \bigg\ { \bigg [ \frac{(2\sqrt{2}\cos\theta_v-\sin\theta_v)\sin\theta_v}{m_{\omega'}^2 } + \frac{3}{m_{\rho'}^2 } -\frac{(\cos\theta_v+2\sqrt{2}\sin\theta_v)\cos\theta_v}{m_{\phi'}^2 } \bigg]\nonumber\\ & & -\bigg(\frac{1}{m_{v_1}^2 } + \frac{3}{m_{v_1}^2}-\frac{2}{m_{v_1}^2 } \bigg ) \bigg\ } ( d_m m_k^2+d_m m_{k^*}^2)+\frac{16\sqrt{2}}{3f_k m_v m_{k^*}}c_4(m_k^2-m_\pi^2)\ , . \label{eqkstarpksu3symplusbreaking}\end{aligned}\ ] ] the criteria to make the above decomposition is that in the @xmath10 symmetric terms we do not distinguish the mass differences of intermediate vectors and also assume the ideal mixing for @xmath28 .
so it is easy to check that @xmath233 , consistent with the @xmath10 symmetry requirement . while for the @xmath10 breaking parts , there are three sources : different masses of the intermediate vector resonances , the non - ideal mixing angle of the @xmath28 and the @xmath34 term in the charged process . the mass differences and the mixing angle of @xmath28 are directly related to the resonance operators in eqs . and
therefore it is crucial to include these mass splitting parameters in the lagrangian approach in order to systematically study the @xmath10 symmetry mechanism .
next let us take the results from fit i in table [ tablepar ] to make a concrete analysis on the strengths of different parts contributing to the @xmath10 symmetry breaking .
the following numbers can be straightforwardly obtained according to the parameters from fit i in table [ tablepar ] @xmath234 it is obvious that these @xmath10 breaking effects suppress the neutral @xmath235 decay width but enhance the charged @xmath235 mode , as a result the ratio between them becomes compatible with the data .
notice that among the @xmath10 breaking effects in @xmath223 , the @xmath34 term contributes at the same order as the other @xmath10 breaking effects in r@xmath13 t , although the values of the @xmath34 parameter , with the order of @xmath236 in table [ tablepar ] , seem unnaturally small .
the small magnitude of @xmath34 can be attributed to the normalization of the @xmath237 operators in eq . .
this conclusion can be easily verified by comparing the values of other @xmath237 couplings with @xmath34 .
we notice that one @xmath237 operator is considered in ref .
@xcite , namely the @xmath238 term in eq.(15 ) of this reference .
it is easy to obtain that the @xmath238 term is related to our @xmath239 term in eq . through @xmath240 , with
@xmath238 estimated to be around @xmath241 in ref .
@xcite , which leads to @xmath242 .
therefore we show that the magnitude of @xmath34 parameter given in table [ tablepar ] is at the same order as other @xmath195 couplings in the @xmath237 lagrangian estimated from other works .
recently , the new measurements on the @xmath243 form factors from the na60 collaboration @xcite have activated several updated theoretical works @xcite .
the main problem behind is that the conventional vmd approach can not well incorporate these data , so new ideas are proposed to have a deeper understanding why the successful vmd is not working here . in refs .
@xcite , a new counting scheme different from r@xmath13 t is used : the masses of both vector mesons and pseudoscalar mesons are treated as small expansion parameters .
so in their counting scheme , the lagrangian that accounts for the decay mechanism through exchanging an intermediate vector meson belongs to the leading order , while the lagrangian responsible for the contact terms in the decay amplitude belongs to the next - to - leading - order . in ref .
@xcite , all of the leading order contributions in their scheme and some incomplete parts of next - to - leading - order contributions are considered in their @xmath42 form factor . after a careful check
, we realize that the @xmath19-type of lagrangian in eq . involving the lowest vector multiplet coincides with the so - called leading order lagrangian proposed in ref .
@xcite up to some normalization factors .
for example , from eqs.(10 ) and ( 15 ) of ref .
@xcite , it is easy to see that their @xmath244 corresponds to our @xmath245 , their @xmath246 relates with our @xmath247 . and the equation of motion @xmath248 @xcite
, we have @xmath249 . the last term with two traces
can be neglected , as it is @xmath64 suppressed .
since the parameters of @xmath250 and @xmath245 always appear in the combination as @xmath251 in our case , the number of independent terms with two vector resonances in this work agrees with that used in ref .
@xcite . ] the @xmath238 term in ref .
@xcite , which belongs to the next - to - leading order in their counting , is equivalent to our @xmath239 term , one of the six @xmath237-type operators in eq . .
one should notice that the authors in ref .
@xcite mostly focus on the leading order computation and in order to make a rough quantitative estimate on the next - to - leading order effects they select a particular operator to perform the numerical discussion .
while in our case , the contact terms , i.e. the @xmath237 operators in eq . , are not suppressed by any counting rule and fully included in the analyses . also the excited vector resonances are not considered in refs .
@xcite . though we have some similar operators from the beginning , the qcd high energy behavior is not appreciated in the discussion of refs . @xcite . as we have mentioned in the introduction ,
the transition form factor with proper high energy constraint can be directly applied in other physical quantities , such as the light - by - light scattering .
so we consider it is an improvement to impose the qcd short distance constraints in the form factor .
next we make a close comparison with ref .
a similar theoretical framework is employed in the former references as ours , but different experimental data are analyzed . the focus of refs .
@xcite is the @xmath252 form factors from the @xmath150 decay process , not from the @xmath150 production .
it turns out that the form factors in refs .
@xcite can fairly well describe the @xmath253 data from na60 collaboration .
then it is interesting for us to perform another fit to only consider the same type of data as in refs .
@xcite , which will be referred as fit iii . to be more specific , we only include the form - factor data from the @xmath150 decay @xcite in fit iii and exclude these from the @xmath150 production @xcite and the @xmath254 spectral function @xcite . as shown in fig .
[ figffomega ] , the final output of our fit iii ( the orange dotted line ) tends to be quite similar with those from ref .
@xcite in the focused energy region , which is below the @xmath150 mass . however , we observe that the @xmath6 form factors in the production region are clearly incompatible with the data from snd collaboration taking the resulting parameters from fit iii , see the orange dotted lines in the region above 0.9 gev in fig .
[ figffomega ] .
similarly , we find that the form factors in refs .
@xcite , when extrapolated to the production region are not explicitly given in refs .
we have exploited the theoretical formulas presented in t references to plot the curves in fig .
[ figffomega ] .
] , are in contradiction with the snd data as well , see the magenta dot - dot - dashed and the cyan dashed lines in fig .
[ figffomega ] .
this tells us that a good reproduction of the form factors in @xmath150 decay does not necessarily guarantee a consistent description of the form factors in @xmath150 production .
therefore it is important to consider both types of data simultaneously , as done in our fit i and fit ii . after including the snd and cleo data in the analysis , i.e. the fits presented in table [ tablepar ]
, we find that our description of the na60 data becomes obviously worse , see the difference between red solid line and the orange dotted line in fig .
[ figffomega ] .
it turns out that our theoretical framework starts to be incompatible with the na60 data above 0.5 gev . in this work the excited vectors
are explicitly included , which are obviously important in the @xmath150 production processes from the snd @xcite and cleo @xcite measurements on the @xmath6 form factors . while , it is also interesting to see to what extent the excited vector resonances can affect the @xmath6 form factor in the @xmath150 decay , i.e. the energy region below @xmath255 . in fig .
[ figffomega ] , we designate the green dot - dashed line to the situation in which we simply exclude the @xmath7 contribution .
it is obvious that the @xmath7 plays an important role in describing the snd collaboration data , while its effect in the low energy region is quite small . in a short summary about the fit to @xmath6 form factor , we could describe the na60 data in our case to the same extent as in ref .
@xcite , if we had not considered the snd and cleo data .
after the inclusion of the latter two measurements , we conclude that our description is still inadequate for the @xmath256 form factors from the na60 measurements in the energy region near the kinematical boundary , i.e. @xmath255 . hence we consider this is still an open problem in our framework .
in this work , the resonance chiral theory is used to study @xmath257 , @xmath258 processes , the @xmath259 transition form factor , and the spectral function for @xmath254 . through fitting the corresponding experimental data , the values of resonance couplings , that can not be fixed through the qcd high energy constraints , are then provided . the masses and widths of our @xmath8 resonance
are quite compatible with the @xmath196 in pdg @xcite . while our @xmath7 looks more like a combined effect of the proposed @xmath200 and @xmath201 from pdg .
the mass and width of the @xmath7 resonance are found to be sensitive to the forms of the energy - dependent widths used in the propagator , while the parameters for the @xmath8 are more or less stable .
the resulting values of the @xmath28 mixing angle are found to be around @xmath203 , clearly different from the ideal mixing case as in the @xmath260 case .
the resonance coupling @xmath34 obtained here is around one order smaller in magnitude than those in literature and our results seem more meaningful when considering the radiative tau decays . with the @xmath34 value obtained here ,
we analyze the different @xmath10 breaking effects contributing to the ratio of @xmath261 in detail .
and we find that the three sources of @xmath10 symmetry breaking effects , e.g. the different masses of intermediate resonances , the non - ideal mixing angle of @xmath28 and non - vanishing value of @xmath34 , are more or less equally important at the numerical leve .
the excited vector resonances are found to be essential to reproduce the @xmath262 form - factor data from snd and the cleo @xmath6 spectral function from tau decays .
although the low energy @xmath6 form - factor data from na60 can be well reproduced in our approach , the steep rise behavior close to the upper kinematical boundary region @xmath263 , is still not fully understood in the current work , even after taking into account the excited vector resonances .
this work is supported in part by national nature science foundations of china ( nsfc ) under contract nos . 10925522 , 11021092 , 11105038 and 11035006 .
yhc acknowledges support by grants no.11261130311 ( crc110 by dfg and nsfc ) .
zhg acknowledges the grants from education department of hebei province with contract number yq2014034 , natural science foundation of hebei province with contract number a2011205093 , doctor foundation of hebei normal university with contract number l2010b04 .
r. arnaldi _ et al . _
( na60 collaboration ) , phys .
* b 677 * ( 2009 ) 260 .
( na60 collaboration ) , nucl .
phys . * a 855 * ( 2011 ) 189 .
m. n. achasov _
( snd collaboration ) , phys .
* b 486 * ( 2000 ) 29
. m. n. achasov _
( snd collaboration ) , phys . rev . *
d 88 * ( 2013 ) 054013 .
b. aubert _
( babar collaboration ) , phys . rev . *
d 77 * ( 2008 ) 092002 .
f. jegerlehner and a. nyffeler , phys
. rept .
* 477 * ( 2009 ) 1 .
j. bijnens , e. pallante and j. prades , nucl .
b * 474 * ( 1996 ) 379 .
j. bijnens , e. pallante and j. prades , nucl .
b * 626 * ( 2002 ) 410 . g. ecker , j. gasser , a. pich and e. de rafael , nucl .
b321 * ( 1989 ) 311 . s. weinberg , physica * 96a * ( 1979 ) 327 . j. gasser and h. leutwyler , annals phys .
* 158 * ( 1984 ) 142 ; j. gasser and h. leutwyler , nucl .
* b 250 * ( 1985 ) 465 .
g. ecker , j. gasser , h. leutwyler , a. pich and e. de rafael , phys .
b * 223 * ( 1989 ) 425 .
v. cirigliano _ et al .
_ , nucl . phys .
* b753 * ( 2006 ) 139 .
d. ruiz - femenia , a. pich , j. portoles , jhep * 0307 * ( 2003 ) 003 .
m. jamin , a. pich , and j. portoles , phys . lett .
* b 640 * , 176 ( 2006 ) ; * 664 * ( 2008 ) 78 .
d.gomez dumm , p.roig , a. pich , and j. portoles , phys . rev . *
d 81 * ( 2010 ) 034031 .
v.mateu and j. portoles , eur .
j. * c 52 * ( 2007 ) 325 .
z. h. guo , j. j. sanz cillero and h. q. zheng , jhep * 0706 * ( 2007 ) 030 . z. h. guo , phys . rev . *
d 78 * ( 2008 ) 033004 .
z. -h .
guo and p. roig , phys .
d * 82 * ( 2010 ) 113016 .
y. h. chen , z. h. guo , h. q. zheng , phys .
d 85 * 054018 ( 2012 ) .
p. roig and j. j. sanz - cillero , phys .
b * 733 * ( 2014 ) 158 .
m. knecht and a. nyffeler , eur
. phys .
j. * c21 * ( 2001 ) 659 .
k. kampf and j. novotny , phys .
d * 84 * ( 2011 ) 014036 .
s. peris , m. perrottet and e. de rafael , jhep * 9805 * ( 1998 ) 011 ; m. knecht , s. peris , m. perrottet and e. de rafael , phys.rev .
* 83 * ( 1999 ) 5230 ; s. peris , b. phily and e. de rafael , phys . rev . lett .
* 86 * ( 2001 ) 14 ; p. masjuan and s. peris , jhep * 0705 * ( 2007 ) 040 . r. escribano , s. gonzalez - solis and p. roig , jhep * 1310 * ( 2013 ) 039 .
j. a. oller , e. oset and j. e. palomar , phys .
d * 63 * ( 2001 ) 114009 .
r. escribano , p. masjuan and j. j. sanz - cillero , jhep * 1105 * ( 2011 ) 094 .
z. -h .
guo and j. a. oller , phys .
d * 84 * ( 2011 ) 034005 .
z. -h .
guo , j. a. oller and j. ruiz de elvira , phys .
d * 86 * ( 2012 ) 054006 .
p. j. odonnell , rev .
mod . phys .
* 53 * ( 1981 ) 673 .
j. beringer , @xmath264 , phys . rev . *
d 86 * ( 2012 ) 010001 .
j. w. durso , phys . lett .
* b 184 * ( 1987 ) 348 .
e. bagan , a.bramon , f.cornet , phys
d 31 * ( 1985 ) 2270 .
a. bramon , a. grau , g. pancheri , phys . lett .
b 344 * ( 1995 ) 240 . m. benayouna , l. delbuonoa , s. eidelmana , v. n. ivanchenkoa , h. b. o.connell , phys .
d 59 * ( 1999 ) 114027 .
sakurai , theory of strong interactions , _ ann .
( ny ) _ * 11 * , 1 - 48 ( 1960 ) ; j.j .
sakurai , _ currents and mesons _
, university of chicago press ( 1969 ) .
f. klingl , n. kaiser , w. weise , z. phys .
* a356 * ( 1996 ) 193 .
r. i. dzhelyadin _
et al . _ ,
* b 102 * ( 1981 ) 296 .
a. faessler , c. fuchs , m. i. krivoruchenko , phys .
* c 61 * ( 2000 ) 035206 .
v. cirigliano , g. ecker , h. neufeld and a. pich , jhep * 0306 * ( 2003 ) 012 .
zhi - hui guo and j. j. sanz - cillero , phys .
d * 79 * ( 2009 ) 096006 .
z. -h .
guo and j. j. sanz - cillero , phys .
d * 89 * ( 2014 ) 094024 .
w. j. marciano and a. sirlin , phys .
lett . * 61 * ( 1988 ) 1815 .
r. i. dzhelyadin _
et al . _ ,
* b 102 * ( 1981 ) 296 ; jetp lett . * 33 * , 228 ( 1981 )
. k. w. edwards _ et al . _
( cleo collaboration ) , phys .
* d 61 * ( 2000 ) 072003 .
d. gomez dumm , a. pich and j. portoles , phys .
d * 62 * ( 2000 ) 054014 .
e. arganda , m. j. herrero , and j. portoles , j. high energy phys .
* 06 * ( 2008 ) 079 . c. terschlsen , s. leupold , phys
* b 691 * ( 2010 ) 191 . c. terschlsen , s. leupold , and m.f.m .
lutz , arxiv:1204,4125 .
s. p. schneider , b. kubis , f. niecknig , phys . rev . *
d 86 * ( 2012 ) 054013 . | within the framework of chiral effective field theory , we study various electromagnetic processes with light vector resonances : @xmath0 , @xmath1 and the @xmath2 form factors . with two multiplets of vector resonances being introduced , we fit the decay widths of @xmath3 , @xmath4 , and the pertinent measurements from the @xmath5 cross sections , such as moduli and relative phases between the isoscalar and isovector components from babar collaboration , together with the @xmath6 form factors from na60 , snd and cleo collaborations .
the values of resonance couplings , masses and widths of the excited vector states @xmath7 and @xmath8 are then determined . the @xmath9-@xmath8 mixing angle is discussed and turns out to be quite different from the ideal mixing case .
three sources of @xmath10 symmetry breaking effects in the @xmath11 decays are identified and analyzed in detail . |
in circumstellar disks , the major sources of dust are thought to be minor bodies , like comets .
both the stellar radiation drag ( poynting - robertson drag ) and stellar wind drag tend to induce dust inflow toward the star .
as the dust passes by planets in its infall , it interacts with them , for example , by accumulating in planetary resonances ( liou & zook 1999 , ozernoy et al .
the inner region of the @xmath0 pictoris dusty disk is tilted by a few degrees with respect to the outer disk ( burrows et al .
1996 , mouillet et al .
1997 , heap et al . 2000
( for references , see the last paper ) .
heap et al .
( 2000 ) concluded that the size and shape of the warp favors the presence of a planet(s ) in a slightly inclined orbit .
the dynamics of interplanetary particles are determined by several effects , including stellar radiation pressure ; the poynting - robertson ( p - r ) and stellar wind drags ; resonance effects by planets ; and gravitational encounters with planets . to simulate the gravitational dynamics of comets and nonconservative dynamics of dust particles around a star , we use an implicit second - order integrator ( taidakova & gorkavyi , 1999 ) . in the absence of a planet ,
the surface density of dust between the star and the pericenter of dust sources is constant , while the number density of dust varies as @xmath2 , @xmath3 being heliocentric distance ( gorkavyi et al .
the presence of a planet induces dramatic changes in the cometary and dust distribution .
this is illustrated in fig . 1 , which shows a representative trajectory of dust particles . [ !
ht ] [ !
ht ] two algorithms for modeling the @xmath0 pictoris disk were considered .
_ algorithm i _ : \1 .
input an initial symmetric cometary disk .
input the mass and orbital radius of the planet on an orbit inclined to the initial cometary disk .
simulate the asymmetric dust distribution on a time scale of @xmath4 yrs .
inclination of particles orbits to the plane of the initial cometary disk and asymmetry of dust is the result of precession of drifting particles around the planet s orbital plane ( see fig . 1 ) .
simulate the dust - scattered light distribution . :
\1 . input an initial symmetric cometary disk .
input the mass and orbital radius of the planet on an orbit inclined to the initial cometary disk .
determine the asymmetric distribution of comets due to orbital precession and gravitational scattering by the planet ( these processes can be realized without any drifts ) after @xmath5 yrs .
simulation of the asymmetric dust distribution for short life times of particles .
asymmetry of dust is mainly a result of an asymmetry in the initial cometary distribution .
simulate the dust - scattered light distribution ( see fig . 2 ) .
both alternatives produce a similar warp , but the second alternative seems to be more preferable due to : * a more realistic ( shorter ) lifetime of particles ; * accounting for gravitational interactions between comets and the planet .
the stis images show the disk to be warped at small distances to the star in a sense that the inner disk is tilted by @xmath6 with respect to the outer disk ( heap et al .
see fig . 2 to compare the stis observations with our modeling .
asymmetric structures in the circumstellar disk indicate at least one planet embedded in the disk of @xmath0 pictoris . | our efficient numerical approach has been applied to modeling the asymmetric circumstellar dust disk around @xmath0 pictoris as observed with the hst / stis .
we present a new model on the origin of the warping of the @xmath0 pic disk .
we suggest that the observed warp is formed by the gravitational influence of a planet with a mass of about 10 masses of earth , at a distance of 70 au , and a small inclination ( @xmath1 ) of the planetary orbit to the main dust disk .
results of our modeling are compared with the stis observations . |
frustrated quantum magnets can exhibit highly unconventional ground states , in which local moments are highly correlated but nevertheless evade a conventional ordering transition and remain strongly fluctuating down to zero temperature .
these unusual states are commonly referred to as quantum spin liquids @xcite despite their rather diverse physical properties ranging from gapped states with an emergent topological order to gapless states with an emergent spinon fermi surface .
a common motif in the search for quantum spin liquids has been to look for quantum antiferromagnets on geometrically frustrated lattices , i.e. lattices where the elementary building blocks prohibit the formation of a conventional nel state .
paradigmatic examples of geometric frustration include lattices formed by corner - sharing tetraedra such as the pyrochlore lattice , or by corner - sharing triangles such as the kagome lattice in two spatial dimensions and the hyperkagome lattice in three spatial dimensions .
an alternative route to induce frustration in a quantum magnet is to look for systems , in which competing interactions can not be simultaneously satisfied .
archetypal examples of such exchange frustration are given by the quantum compass models @xcite , in which the easy - axis of an anisotropic spin exchange strongly depends on the spatial orientation of the exchange path a scenario which can prohibit even a ferromagnet on a bipartite lattice from undergoing a finite - temperature ordering transition .
the best known example in this class of compass models is the kitaev model @xcite on the honeycomb lattice , in which the easy - axis of an ising - like spin exchange points along the @xmath5 , @xmath6 , and @xmath7 directions for the three different bond types of the hexagonal lattice , which is captured by the hamiltonian @xmath8 where su(2 ) spins @xmath9 on sites @xmath10 and @xmath11 are connected via a bond in the @xmath12 direction .
the kitaev model is quintessential in that it harbors three different types of quantum spin liquids a gapped , @xmath13 topological spin liquid if one of the three exchange couplings is significantly larger than the couplings associated with the two other bond directions ( i.e. @xmath14 ) , and a gapless spin liquid in the vicinity of equal - strength exchange couplings ( @xmath15 ) .
if applying an external magnetic field along the @xmath16-direction , the latter can be gapped out into a topological spin liquid with non - abelian vortex excitations .
the kitaev model not only stands out for the unusual richness of its ground states , but the fact that it is one of the very few examples of an interacting spin model that can be rigorously solved .
it should , however , be pointed out that the kitaev model has not only attracted the imagination of phenomenologically inclined theorists , but has also stirred some excitement in the materials oriented community after it has been pointed out that the significantly enhanced spin - orbit coupling in 5d transition metal oxides and in particular certain iridates can give rise to unconventional mott insulators where the local moment is a spin - orbit entangled @xmath3 moment @xcite . the orbital contribution to these moments results in a highly anisotropic , spatially oriented exchange @xcite , which can in fact mimic those of the kitaev model . in terms of actual materials the layered iridates na@xmath1iro@xmath2 and
li@xmath1iro@xmath2 have attracted much recent interest and are intensely discussed @xcite as possible candidate materials realizing the two - dimensional honeycomb kitaev model . in this manuscript
, we turn to generalizations of the kitaev model on three - dimensional lattices a move that is prompted by the recent synthesis of @xmath0-li@xmath1iro@xmath2 @xcite , which forms a truly three - dimensional lattice structure of the ir@xmath4 ions .
this structure , which has quickly been dubbed hyperhoneycomb lattice @xcite , keeps the trivalent vertex structure of the hexagonal lattice and thereby the essential feature allowing for an analytical solution of the kitaev model .
in fact , the kitaev model on the hyperhoneycomb lattice had been identified and studied before by mandal and surendran @xcite who reported the occurrence of a gapless spin liquid with an emergent spinon fermi surface on a line in momentum space for approximately equal - strength interactions ( @xmath15 ) as well as the occurrence of a gapped topological spin liquid for anisotropic exchange strength @xcite .
more recently , extensions to a heisenberg - kitaev model @xcite have established the stability of this gapless phase in the presence of weak isotropic spin exchange @xcite
. this motivated us to ponder alternative three - dimensional lattices that keep the trivalent vertex structure and led us to consider what we call the hyperoctagon lattice @xcite illustrated in fig .
[ fig : hyperoctagonlattice ] .
the hyperoctagon lattice is closely related to the hyperkagome lattice
the hyperoctagon lattice is the premedial lattice of the hyperkagome lattice obtained by shrinking each triangle of the hyperkagome lattice to a single vertex and the new bonds indicating the original connectivity of the triangles , schematically summarized in fig .
[ fig : latticesummary ] .
the hyperoctagon lattice is a chiral lattice , which contains two elementary motifs a spiraling octagonal helix and a counter - spiraling square helix as illustrated in fig .
[ fig : hyperoctagonlattice ] .
its space group i4@xmath1732 ( no . 214 ) indicates the presence of 4- , 3- and 2-fold ( skew ) symmetries ( the details of which we will provide below ) that will turn out to play an essential role in stabilizing the gapless modes of the quantum spin liquid emerging for the kitaev model on this lattice .
the presence of these symmetries is also the key distinction to the hyperhoneycomb lattice , another somewhat less symmetric three - dimensional trivalent lattice structure which has been revealed in the recent synthesis of @xmath0-li@xmath1iro@xmath2 .
our main result is the observation of a gapless quantum spin liquid with an extended two - dimensional majorana fermi surface around the point of isotropic couplings for the kitaev model on the hyperoctagon lattice . this result is rigorously established by an exact analytical solution of the spin model , which can be cast into a free fermion system by majorana fermionization , thus employing the same powerful techniques that have already allowed the solution of the kitaev model on other trivalent lattices @xcite
. our discussion in the remainder of the paper is structured as follows . in section [ sec : lattice ]
we will discuss trivalent lattice structures in two and three spatial dimensions and in particular provide a detailed introduction of the hyperoctagon lattice .
the kitaev model on the hyperoctagon lattice is subsequently introduced and exactly solved in section [ sec : kitaev_model ] where we also provide a detailed discussion of its ground state phase diagram , in particular the gapless spin liquid with a majorana fermi surface emerging for a broad range of parameters .
possible instabilities of the majorana fermi surface are discussed in section [ sec : fermisurface ] .
we conclude with an outlook in section [ sec : outlook ] .
before we dive into the physics of the kitaev model we start our discussion with a short review of the underlying lattice structure . in its original form
the kitaev model has been discussed for the honeycomb lattice , a two - dimensional lattice with a trivalent coordination of all vertices as depicted in fig .
[ fig : honeycombs ] a ) . keeping this motif of a trivalent lattice structure the model can readily be associated with a broader class of lattices a move that not only allows a @xmath18 assignment of the three different exchange types to the bonds around the vertices , but is also key to keep the analytical tractability of the model , which we will review in the following section @xcite . in two spatial dimensions
one such generalization is the square - octagon lattice of ref . .
in three spatial dimensions such trivalent lattices are considerably less common .
one example is the so - called hyperhoneycomb lattice , which is depicted in fig .
[ fig : honeycombs ] b ) .
the elementary building blocks of the hyperhoneycomb lattice are zig - zag chains running along the crystallographic @xmath19 and @xmath20 axis , respectively , as depicted in fig .
[ fig : honeycombs ] b ) .
these zig - zag chains are coupled by bonds along the @xmath21 axis , readily implying that there is no general symmetry possibly interchanging the three crystallographic axises .
another example of a trivalent lattice in three dimensions is the so - called hyperoctagon lattice , which we describe in detail in the following .
the hyperoctagon lattice is a body - centered cubic lattice without inversion symmetry .
its symmetries correspond to space group i4@xmath1732 ( no . 214 ) . in cartesian
coordinates , the atomic positions can be constructed starting from the point @xmath22 in the unit cell and applying all symmetry transformations of the space group on it . in particular , the symmetries of space group i4@xmath1732 include the following : i ) a four - fold symmetry which is obtained by 90 degree _
screw_-rotations around the ( 1,0,0 ) , ( 0,1,0 ) , or ( 0,0,1 ) directions .
ii ) a three - fold symmetry which leave the lattice invariant under 120 degree rotations around the ( 1,1,1 ) , ( -1,1,1 ) , ( 1,-1,1 ) , or ( 1,1,-1 ) directions , and iii ) a two - fold symmetry corresponding to 180 degree rotations around the directions @xmath23 , @xmath24 , and @xmath25 . as a guide to the eye ,
[ fig : rotations ] shows the projection of the lattice onto the planes normal to the rotation axis for three examples of the above ( screw-)rotations .
note that the ( projected ) square - octagon structure in fig .
[ fig : rotations ] a ) is not planar , which is why the additional translation is needed .
@xmath26 ) .
a ) view along ( 0,1,0 ) as an example of a 90 degree screw - rotation symmetry .
b ) view along ( 1,1,-1 ) as an example of a 120 degree rotation symmetry .
c ) view along ( 0,1,1 ) as an example of a 180 degree rotation symmetry . ]
while the hyperoctagon lattice arises quite naturally as the premedial lattice of the hyperkagome lattice , which is realized for instance in the spin liquid material na@xmath27ir@xmath2o@xmath28 @xcite , there are so far no known realizations of the hyperoctagon lattice in the diverse family of recently synthesized iridates . to provide some abstract guidance
as to which chemical compositions might possibly realize magnetic hyperoctagon systems , we have made an attempt at designing possible materials candidates . with the iridium atoms assumed to occupy the sites of the hyperoctagon lattice ,
our further thinking is guided by the microscopic prerequisites that allow a dominant anisotropic kitaev - like interaction to emerge the occurrence of double ir - o - ir exchange paths that suppress the isotropic spin exchange @xcite .
the latter can be achieved by placing the iridium atoms in bond - sharing iro@xmath29 cages .
in fact , the symmetries of the hyperoctagon lattice allow to embed each iridium atom into a perfectly undistorted iro@xmath29 octahedron , when placing the oxygen atoms at position @xmath30 in the unit cell .
the resulting iro@xmath2 structure is illustrated in fig .
[ fig : o6cages ] .
such a sparse octahedron structure has indeed been observed for the subhalides , e.g. la@xmath2br@xmath2si @xcite , where the silicium atoms form the hyperoctagon lattice and the lanthanum atoms form octahedra around them .
finally , one might want to attempt to fill the remaining interstitial sites of the octahedron structure .
taking into account the space group symmetries there are several distinct ways of doing so , as described in some detail in appendix [ sec : appspacegroupmaterials ] .
in particular , one might start to add atoms to a single interstitial site ( and its space group related siblings ) as illustrated in fig .
[ fig : space_group ] a ) of the appendix . this would result in the chemical composition of the alkaline - earth iridates airo@xmath2 where a is one of the alkaline - earth elements ca , sr , or ba . the alkaline - earth iridates are known to exhibit quite distinct electronic properties for the different a - site materials , including an @xmath31 antiferromagnetic mott insulator for cairo@xmath2 @xcite , a weak ferromagnetic semiconductor for bairo@xmath2 @xcite and a non - fermi liquid metal in sriro@xmath2 @xcite . while various crystal structures have been reported for the different airo@xmath2 compounds , no crystals in space group i@xmath32
have so far been synthesized for any of the alkaline - earth iridates .
an alternative possibility to fill the interstitial sites is to add two additional atoms resulting in a chemical composition of the form a@xmath1iro@xmath2 with a being one of the alkali metals na or li .
the resulting crystal structure is illustrated in fig .
[ fig : space_group ] b ) of the appendix .
this is a particularly interesting idea to entertain as it would point to a possible existence of a third crystallization pattern for a@xmath1iro@xmath2 beyond the already known examples of quasi two - dimensional honeycomb layers and the recently synthesized three - dimensional hyperhoneycomb structure .
octahedra around the central ir atoms .
a ) view along the ( 1,0,0 ) direction .
b ) view along the ( 1,1,1 ) direction . ]
not only motivated by a possible relevance to future materials , but also driven by a curiosity to explore unusual spin liquid states we now turn to a three - dimensional variant of the kitaev model on the hyperoctagon lattice .
we proceed with an introduction and precise definition of the model and discuss some of its general properties before presenting an analytical solution of the model in terms of an exact majorana fermionization of the spin degrees of freedom . finally , we present our main result of identifying a gapless spin liquid ground state with a majorana fermi surface .
kitaev originally introduced his elementary spin model as a system of su(2 ) spin-1/2 degrees of freedom interacting on the two - dimensional honeycomb lattice .
its fundamental beauty not only arises from its exact analytical solution , but the fact that the spin model harbors a number of paradigmatic ground states besides an abelian topological phase
, it exhibits an extended gapless spin liquid ground state , which can be gapped out into a non - abelian topological phase by an external magnetic field .
this variety of different ground states arises from highly frustrated spin interactions which favor the alignment ( or anti - alignment ) of different spin components along the three principle directions of the honeycomb lattice . here
we generalize this idea to the three - dimensional hyperoctagon lattice , which because of its trivalent vertices allows the definition of an analogous spin model . to this end
, we cover the lattice with bonds that favor spin alignments along the @xmath5 , @xmath6 , and @xmath7 directions and which we call @xmath33 , @xmath34 , and @xmath35-bonds , respectively . while there are many different ways to realize such coverings on a given lattice
, there is only a single covering that is compatible with all the lattice translation symmetries of the hyperoctagon lattice .
this unique covering is illustrated in fig .
[ fig : kitaevmodel ] and will serve as our primary definition of the kitaev model on this lattice
. we will briefly discuss alternative models based on other coverings in section [ sec : fermisurface ] .
-couplings , red bonds to @xmath34-couplings , and blue bonds to @xmath35-couplings , respectively . ] in order to provide a self - contained description of the model , we start by introducing the four - site unit cell compatible with the unique covering of exchange bonds such that all lattice translation symmetries are kept .
the atomic positions in this unit cell are given by @xmath36 where @xmath37 is the unit cell position .
the offset in @xmath38 is chosen to be consistent with the conventions in section [ sec : lattice_symmetries ] and can be mostly ignored in the following discussion .
the corresponding lattice translation vectors are then given as @xmath39 which are also illustrated in fig .
[ fig : kitaevmodel ] . with these definitions in place
we can now define the kitaev hamiltonian on the hyperoctagon lattice as @xmath40 for the following discussion , it is beneficial to introduce a ` bond operator ' @xmath41 for a bond @xmath42 @xmath43 in terms of these bond operators , the hamiltonian then reduces to the compact form introduced earlier @xmath44 our first step in analyzing hamiltonian is to identify conserved quantities , which we will find to be intimately connected to closed paths ( or loops ) on the lattice .
the elementary loops of the hyperoctagon lattice have length ten . for each unit cell
there are six distinct such loops , which are visualized in fig .
[ fig : loops_hyperoctagon ] .
all other elementary loops can be obtained by lattice translations . for each loop @xmath45 we can define a corresponding loop operator @xmath46 , which measures the ` magnetic flux ' through the plaquette that is enclosed by @xmath45 .
we can define the loop operator by the product of bond operators of all the bonds contained in the loop @xmath47 because of the even length of the loops these loop operators square to the identity , thus they have eigenvalues @xmath48 .
it can further be verified that the loop operators commute with the hamiltonian as well as with each other .
each loop operator thus defines an ` integral of motion ' and a corresponding conserved quantity the extensive number of which greatly simplifies the problem . for one , we can divide the hilbert space into distinct sectors that are each labeled by the eigenvalues of all the loop operators @xmath46 and restrict the hamiltonian to a particular sector . before proceeding we note that an alternative definition of the loop operators can be formulated as a product over all sites contained in the loop : @xmath49 where @xmath50 corresponds to the spin component at each vertex , which is not included along the loop .
for instance , the loop operator of the loop in fig .
[ fig : loops_numbered ] a ) is given by @xmath51 in both the honeycomb and hyperhoneycomb model , the two definitions of the loop operator , @xmath46 in and @xmath52 in , are identical .
however , in our particular case there is a relative minus sign between the two , i.e. @xmath53 , which leaves some freedom in how to define the magnetic flux . in the following ,
we define magnetic flux by the eigenvalue of : if @xmath46 has eigenvalue @xmath54 we say that there is a magnetic flux ( vortex ) penetrating the plaquette enclosed by @xmath45 , while an eigenvalue @xmath55 corresponds to no flux @xcite .
a further important difference to the two - dimensional case is that the loop operators are not all linearly independent . as an example , we consider the two loops depicted in fig .
[ fig : loops_numbered ] . the loop operator for the loop in fig .
[ fig : loops_numbered ] a ) is given in eq .
, while the one for the loop in fig .
[ fig : loops_numbered ] b ) is given by @xmath56 note that we can define a third loop of length ten by combining the bonds that are contained in @xmath57 or @xmath58 , but not in both of them ( see the loop illustrated in panel e ) of fig .
[ fig : loops_hyperoctagon ] ) @xmath59 the product of the three loops is the identity operator @xmath60 which implies that the eigenvalue of @xmath61 is uniquely determined by the ones of @xmath57 and @xmath58 .
a direct consequence of this linear dependence of three loops is that there is no full - flux sector in this model for which all loop operators have eigenvalue -1 .
note that if we would have chosen the alternative definition of the loop operator in eq .
then we would have concluded that there is no zero - flux sector in this model .
this constraint can also be understood graphically .
each loop defines an enclosed surface as illustrated in fig .
[ fig : volumes_hyperoctagon ] . in general
, a product of loop operators is constrained if their respective surfaces form a closed ` volume ' .
for the hyperoctagon lattice , note that the three surfaces corresponding to the respective loops indeed form a closed object though the encompassed volume actually vanishes .
this situation should be contrasted to the three - dimensional hyperhoneycomb lattice , which for a complete and self - consistent presentation we discuss in appendix [ app : hyperhoneycomb ] . in the hyperhoneycomb lattice ,
four loops of length ten encompass a closed volume as depicted in fig .
[ fig : loops_hyperhoneycomb ] in the appendix and are thus linearly dependent . however , with an even number of linear dependent loops , the corresponding spin model allows for both a zero - flux and a full - flux sector , while in our case only one of the two sectors can exist .
in order to gain a better insight into the physics underlying the different magnetic flux sectors , let us note that the elementary loops of length ten can be uniquely labeled by their midpoint @xmath62 . these midpoints form a ( deformed ) hyperkagome lattice .
the constraint on the loop operator eigenvalues is then enforced on each of the triangles of the hyperkagome lattice : there are either zero or two loops per triangle that carry flux .
we can thus count the number of independent configurations by noting that there are six midpoints ( = loops ) and four triangles ( = constraints ) per unit cell .
thus , there are only @xmath63 loop eigenvalues per unit cell that can be chosen freely . as a result
there are in total @xmath64 distinct flux sectors , where @xmath65 is the number of unit cells . in order to determine in which magnetic flux sector the ground state resides
, we can not follow the same route as taken in two spatial dimensions and resort to lieb s theorem stating that the ground state always resides in the flux - free sector @xcite , but instead have to carefully consider the energetics of the different flux sectors / loops .
for the great majority of points in parameter space , we numerically observe that creating or enlarging loops costs energy and as such the ground state lies in the flux - free sector also in three spatial dimensions .
we therefore restrict our following discussion to the zero - flux sector .
let us briefly consider the effective theory for the magnetic flux excitations arising from flipping a loop operator eigenvalue from @xmath55 to @xmath54 .
we note that the midpoints of loops with loop operator eigenvalue @xmath54 form themselves closed loop configurations , which live again on the links of a hyperoctagon lattice albeit with opposite chirality to the original one . due to the constraint ,
only closed loop configurations are allowed , i.e. there are no magnetic monopoles in this @xmath66 theory .
we now proceed to discuss the exact analytic solution of hamiltonian . in analogy to the two - dimensional kitaev model , such an analytical solution is possible by recasting the original spin degrees of freedom in terms of majorana fermions a step that effectively reduces the interacting spin system to a free fermion problem , which is given by majorana fermions hopping in a static gauge field @xcite .
the fermion system can thus be diagonalized in a straight - forward way thereby also revealing the physics of the interacting spin model . as first step ,
we rewrite the original spin degrees of freedom by introducing four majorana fermion degrees of freedom @xmath67 and @xmath20 per spin @xmath9 , such that @xmath68 where @xmath69 denotes the spin component .
the four - dimensional hilbert space of the four majorana fermions can be projected back to the two - dimensional physical hilbert space of the original spin degrees of freedom by requiring @xmath70 as we need to introduce four majorana fermions per site , we have to introduce additional labels to indicate the unit cell index @xmath11 as well as the unit cell position @xmath38 , i.e. @xmath71 . the majorana fermions obey the usual anti - commutation relations @xmath72 of a clifford - algebra . in terms of the majorana fermions ,
the hamiltonian becomes @xmath73 where we introduced the link operators @xmath74 with @xmath75 being the label of the bond @xmath42 .
the link operators commute among themselves as well as with the hamiltonian , which implies that we can fix the eigenvalues of all the link operators i.e. choose a specific ` reference configuration ' and compute the spectrum of the resulting quadratic hamiltonian for any given @xmath76 sector .
when doing this , one needs to define a direction on the bonds , as @xmath77 .
we choose the convention that the @xmath33-bonds are directed along the @xmath78-direction , the @xmath34-bonds along the @xmath79-direction , and the @xmath35-bonds along the @xmath80-direction .
this convention ensures that the following discussion remains symmetric in permutations of @xmath81 , @xmath82 , and @xmath83 .
one may think of the link degrees of freedom as a static @xmath66 gauge field .
the gauge transformations are generated by the @xmath84 operators and eq .
is equivalent to demanding the physical states to be gauge invariant .
in fact , the gauge invariant objects are precisely the loop operators introduced earlier . choosing a reference configuration is equivalent to choosing a specific gauge .
the physical properties , such as the majorana excitation spectrum , are independent of the specific gauge choice , as was already pointed out in kitaev s original solution @xcite of the honeycomb model .
as we restrict our discussion to the flux - free sector , we may choose all link operators to have eigenvalues @xmath55 .
using the fourier transformation @xmath85 with @xmath65 being the number of unit cells , we can compute the majorana hamiltonian in momentum space @xmath86 where @xmath87 is defined as the coefficient of the reciprocal lattice vectors @xmath88 with @xmath89 arriving at hamiltonian has thus reduced the original problem to a four - band hamiltonian that can be easily diagonalized @xcite . from the diagonal form of the hamiltonian we can readily read off the elementary structure of the phase diagram in @xmath90-parameter space by carefully analyzing the excitation spectrum of the majorana sector . in particular , we observe that the hamiltonian allows for zero - energy solutions indicative of a gapless phase in a range of parameters , while both the excitations of the majorana sector and the magnetic flux sector remain gapped in other parts of the phase diagram .
( indicated by a point ) . around the three corners of the phase diagram where one of the couplings dominates extends a gapped spin liquid phase , which is separated from the gapless phase via a line of ( continuous ) phase transitions indicated by the yellow line . ]
the occurrence of zero - energy solutions is equivalent to requiring that @xmath91 , which becomes @xmath92&=\frac 1{16 } ( j_x^4+j_y^4+j_z^4 + 2j_y^2 j_z^2 \cos(k_x ) \nonumber\\ & + \
, 2j_x^2j_z^2\cos(k_y)+2j_x^2 j_y^2 \cos(k_z ) ) \nonumber\\ & \equiv 0\end{aligned}\ ] ] in cartesian coordinates . in order to determine whether or not the above equation has solutions , let us analyze the limiting behavior of the determinant .
expression is bounded from above by @xmath93 & \leq \frac 1{16 } ( j_x^2+j_y^2+j_z^2)^2,\end{aligned}\ ] ] when setting @xmath94 and bounded from below by @xmath95 & \geq - \frac 1{16 } ( j_x+j_y -j_z)(j_x+j_z - j_y ) \nonumber\\ & \times ( j_y+j_z - j_x)(j_x+j_y+j_z ) , \end{aligned}\ ] ] when setting @xmath96 . as the upper bound is always strictly positive
, there is a zero - energy solution iff the lower bound is negative ( or zero ) .
the latter is equivalent to requiring the triangular inequality @xmath97 . in the region around the isotropic point
@xmath98 , there are gapless modes in the majorana sector and the ground state is a gapless spin liquid . if one of the three coupling dominates the majorana spectrum remains gapped and the ground state is a gapped spin liquid .
the gapped and gapless phases are connected via lines of phase transitions which are parametrized by the equalities in the triangular inequality .
it should be noted that the fundamental shape of this phase diagram is precisely the same one as the ones found for the two - dimensional honeycomb lattice @xcite and the three - dimensional hyperhoneycomb lattice @xcite .
what sets the phase diagrams apart is the actual nature of the two principle gapped and gapless phases for the respective lattices as we will discuss in the following section .
the fundamental distinction in the phase diagram of the hyperoctagon model is the nature of the gapless phase in the vicinity of the isotropic coupling point ( @xmath98 ) .
its main feature is an extended two - dimensional majorana fermi surface of gapless modes . to see
the emergence of such a fermi surface in the majorana spectrum around the isotropic coupling point one needs to invert eq .
, which gives @xmath99 \ , .
\label{eq : surface}\end{gathered}\ ] ] in combination with the requirement that the lower bound in eq . becomes negative or zero this parametrizes an entire manifold of @xmath100-points , or more precisely two distinct , non - intersecting continuous surfaces in momentum space centered around the corners of the brillouine zone at @xmath101 , respectively , as illustrated in fig . [
fig : majoranafermisurface ] .
it is important to note that while the two surfaces are symmetry related they can not be mapped onto each other by a reciprocal lattice vector . as a direct consequence momentum conservation ensures that the zero - energy modes can not gap out in a pairwise fashion and the surfaces have to remain stable throughout the gapless region .
indeed varying the coupling constants away from the point of isotropic coupling only deforms the surfaces , but does not destroy them .
this is illustrated in the sequence of panels of fig .
[ fig : majoranafermisurface ] where we plot the evolution of the surfaces along a line in parameter space defined by @xmath81 and @xmath102 .
starting from the isotropic point and increasing @xmath81 elongates the surface along the @xmath80-direction and contracts it in the orthogonal @xmath78- and @xmath79-directions as illustrated in the upper panel of fig .
[ fig : majoranafermisurface ] . upon further increasing @xmath81 the surface contracts towards the corner of the brillouine zone .
as one approaches the phase transition to the gapped spin liquid at @xmath103 , the surfaces have reduced to the points @xmath104 at the corners of the brillouine zone . on the other hand , decreasing @xmath81 from the isotropic point flattens the surface in the @xmath80-direction as illustrated in the lower panel of fig .
[ fig : majoranafermisurface ] . as @xmath81 goes to zero the two opposite sides of the surface approach each other and touch precisely at the decoupling point @xmath105 ( and at which the notion of a two - dimensional surface ultimately breaks down as well ) . to reveal the nature of the zero - energy gapless modes we plot the dispersion of the four principal bands of hamiltonian along certain high - symmetry lines in fig .
[ fig : dispersion ] . for
the entire gapless phase the dispersion of the bands crossing zero energy is always linear along the direction normal to the surface reminiscent of the energy spectrum of a fermi liquid in the vicinity of the fermi energy .
one should however keep in mind that the four bands in our hamiltonian are not spanned by conventional fermionic degrees of freedom , but by majorana fermions .
as such the zero - energy surfaces revealing themselves in the energy spectrum should in fact be thought of as majorana fermi surfaces . as the phase transition to the gapped phase at @xmath106
is approached , the relevant majorana band moves up in energy and at @xmath103 no longer crosses the zero - energy level , but merely touches @xmath107 in a single point with a quadratic dispersion .
this scenario is in complete analogy to the phase transitions from the gapless to gapped majorana phases in both the honeycomb and hyperhoneycomb models .
for conventional fermi liquids it is well appreciated that the fermi surface is susceptible to a variety of instabilities , the most notable of which is the formation of superconductivity .
as such two questions immediately arise with regard to the majorana fermi surface in our hyperoctagon model why is the majorana fermi surface stable in the first place and what are its possible instabilities ?
we will address the first question the stability of the majorana fermi surface in the following by showing that it can be tracked to the underlying lattice symmetries .
we will then devote the remainder of this section to a discussion of possible instabilities of the majorana fermi surface focusing on instabilities arising from a reduction in lattice symmetries . to discuss the stability of the majorana fermi surface
let us first recall some basic facts about majorana fermions .
an immediate consequence of the majorana condition @xmath108 in real space is that the ` creation operator ' @xmath109 in momentum space is defined by @xmath110 .
the latter implies that for every energy state @xmath111 there is a ` particle - hole - conjugate ' partner at @xmath112 for which @xmath113 quite generally additional energy relations might exist , which depend on the underlying lattice geometry .
of particular importance in our case is the bipartite nature of the lattice . for a majorana hamiltonian on a bipartite lattice with vanishing intra - sublattice hopping amplitudes
, one can verify that @xmath114 where @xmath115 is the reciprocal lattice vector of the translation relating the two sublattices .
note that for the honeycomb and hyperhoneycomb lattice we have two - site and four - site unit cells , which allow @xmath116 and , combining the two energy relations found above , @xmath117 .
this implies that zero - energy modes always occur in pairs at a given momentum . on the other hand ,
it is important to note that for the hyperoctagon lattice @xmath115 generically does not vanish , because its elementary four - site unit cell is not consistent with a bipartite coloring of the lattice , see also fig .
[ fig : kitaevmodel ] . for the specific example discussed in the previous section , @xmath118 . as a consequence , the zero - modes are in general all separated in momentum space . in order to see what effect this has on the stability of the gapless modes , we start by revising the situation for the honeycomb model following kitaev s original arguments and afterwards extend this discussion to the three - dimensional generalizations . in the honeycomb model ,
the majorana hamiltonian in momentum space is a @xmath119 matrix of the form @xmath120 where @xmath121 is a complex - valued function .
the vanishing diagonal elements are in fact protected by time - reversal symmetry @xcite .
the eigenvalues of the hamiltonian are given by @xmath122 and zero - energy modes occur when @xmath123 .
let us start by noting that the conditions @xmath124=0 $ ] and @xmath125=0 $ ] define ( several ) closed lines in momentum space , denoted in the following by @xmath126 and @xmath127 .
the zeroes of @xmath121 are then , in general , given by the intersections of these lines . as a consequence , each pair of zero - modes at momentum @xmath128 comes with a partner , which in fact is located at @xmath112 .
changing parameters deforms the line @xmath129 , which in turn moves the zeroes .
the only way to gap out the system is by moving the lines @xmath129 sufficiently , such that they do not intersect any longer .
the phase transition corresponds to a situation , where @xmath126 and @xmath127 merely touch .
this structure of the eigenenergies readily confirms that _ separated pairs _ of zero - modes are stable in a finite parameter regime . on the other hand ,
a similar line of reasoning shows that _ lines _ of zero - modes are not stable and can generically be gapped out completely by even an infinitesimal change in parameters .
there is , however , a generic way to stabilize lines of gapless modes in two - dimensional generalizations of the kitaev model , which we will comment on below .
we now extend the above discussion to the three - dimensional models , first considering the hyperhoneycomb model .
in fact , the arguments of this section are valid for any three - dimensional kitaev - type model on a bipartite lattice , which is time - reversal invariant and where the unit cell is compatible with a bipartite coloring of the lattice . these assumptions are sufficient to determine the majorana hamiltonian to be a block matrix , where only the off - diagonal matrices are non - vanishing @xmath130 where @xmath131 is a complex matrix .
the eigenvalues of the hamiltonian are given by @xmath132 , where @xmath133 are the eigenvalues of @xmath131 .
analogously to the two - dimensional case , there are zero - energy solutions for @xmath134 . however , as the model is three dimensional the conditions re@xmath135=0 $ ] and im@xmath134 now define ( several ) surfaces in momentum space , denoted again by @xmath129 .
zero - energy modes occur at the intersection of @xmath126 with @xmath127 and thus in general form closed lines in momentum space . changing parameters in the model
deforms the surfaces @xmath129 and , thus , the corresponding line of gapless modes , but can not induce a gap in the system . the latter can only be done by changing parameters sufficiently , such that the surfaces @xmath126 with @xmath127 do not intersect any longer , during which the line shrinks to a single point and vanishes .
this shows that extended lines of gapless modes are topologically stable in these types of models
. a similar line of reasoning substantiates that this does not apply to separated points as well as surfaces of gapless modes both of which are not stable objects for these kinds of models and can only be accidental .
the important distinction of the hyperoctagon model to the ones discussed above , is that the unit cell is not compatible with a bipartite coloring of the lattice .
time - reversal invariance thus only requires the diagonal elements to vanish .
the hamiltonian can then be written as @xmath136 which is a band hamiltonian with the additional property @xmath137 due to the majorana condition @xcite .
zero - energy modes occur , when bands cross the @xmath107 line , which results in _ surfaces _ of gapless modes . in general
, the incompatibility of the unit cell and the bipartite coloring of the lattice implies that @xmath138 . as a result , there is generically only a _
single _ majorana zero - mode at a given momentum .
the surfaces are thus trivially stable changing parameters deforms the energy bands and thus the surfaces , but gapping a surface can only be done by either shrinking it to a point or by superimposing two such surfaces , in which case there are two gapless majorana modes at the same momentum .
the latter is not stable and can always be gapped out by an infinitesimal change in parameters .
likewise , one can verify that lines or separated points of gapless modes are not stable objects for these types of majorana hamiltonians .
this line of reasoning also sheds light on how to obtain two - dimensional models with a stable fermi surface . in analogy to the case
above , one needs to consider two - dimensional lattices , where the unit cell is not compatible with the bipartite coloring of the lattice .
an example would be the square - octagon lattice studied in @xcite .
considering various flux sectors , which do not enlarge the unit cell , indeed demonstrates that separated points of gapless modes are not stable , while closed lines are . following the line of reasoning in the previous subsection points to a natural way to destabilize the fermi surface by enlarging the unit cell such that the two majorana fermi surfaces are mapped onto each other .
this requirement is identical to identifying an enlarged unit cell that allows for a bipartite coloring of the lattice within that unit cell ( and thus @xmath115 vanishes ) and the disappearance of a lattice symmetry that prohibits the zero - energy modes from gapping out pairwise . below
we will see that the majorana fermi surface is indeed no longer stable when enlarging the unit cell . however , the surface does not gap out completely , but instead reduces to a line of gapless modes similar to the situation of the kitaev model on the hyperhoneycomb lattice
. a closed line of gapless modes is , on the other hand , a topologically stable object for three - dimensional hamiltonians such as .
changing parameters can only deform the line , but not gap it out . in order to elucidate this ,
let us consider an alternative covering of the hyperoctagon lattice with @xmath33 , @xmath34 , and @xmath35 bonds , which is no longer invariant under @xmath139 translations .
the enlarged unit cell thus allows for a bipartite coloring of the lattice .
a specific realization of this is shown in figure [ fig : kitaev2 ] a ) .
compared to the original model discussed in section [ sec : the_model ] , the @xmath33 and @xmath34 bonds are switched in every second unit cell along the @xmath140 direction .
this enlarges the unit cell with the new translation vectors becoming @xmath141 , @xmath142 and @xmath143 .
note that these are the simple cubic translations , i.e. we have moved from a body - centered cubic structure to a simple cubic one by doubling the unit cell along the @xmath140 direction .
and b ) @xmath144 direction .
the arrow marks the broken translation vector .
the unit cell of the model in a ) is compatible with a bipartite coloring of the lattice while the one in b ) is not . ]
the model still fulfills that no vertex is connected to the same bond type twice and is , therefore , still exactly solvable with the methods used above .
in contrast to the original hamiltonian , the new hamiltonian is no longer isotropic in the coupling constants , as can already be deduced from fig .
[ fig : kitaev2 ] a ) the resulting kitaev model is only symmetric in @xmath145 , while the @xmath35 type bond stands out .
thus , we expect qualitatively different behavior when @xmath146 compared to @xmath147 .
let us first comment on the flux sectors on the model .
it can be shown that the constraints on the loop operators are independent on the choice of the covering of the hyperoctagon lattice in @xmath33 , @xmath34 , and @xmath35 bonds , even though the loop operators themselves change .
in analogy to our original discussion in section [ sec : the_model ] we restrict ourselves to the vortex - free sector in the following and focus on the changes in the majorana sector induced by the alternation of the kitaev interactions .
in particular , we are interested in the gapless modes in the majorana sector .
the parameter region , in which gapless modes exist , is identical to the original model , i.e. determined by the triangular inequality .
the transition lines to the gapped phases are parametrized by the equalities of eq . .
in addition , the behavior of the model along the @xmath146 line in parameter space is identical to the original model .
the latter can be understood by noting that the modified majorana hamiltonian at this line ` looks ' like the original one , albeit with an enlarged unit cell .
in particular , on the line @xmath146 we still find a surface of gapless majorana modes when @xmath146 .
however , the surface is not stable against departing from that parametrization condition .
even an infinitesimal discrepancy in the coupling constants @xmath147 immediately opens up a gap on most of the surface and the two - dimensional majorana fermi surface collapses onto a line of gapless excitations , with linear dispersion in the normal directions .
this line of gapless modes then remains stable throughout the rest of the gapless phase and contracts to a point on the transition lines to the gapped phases .
the phase diagram for this model is visualized in fig .
[ fig : phasediagramline ] . a ) . along the line @xmath146
, there is a majorana fermi surface , which is identical to the original model defined in section [ sec : the_model ] .
away from this line , the majorana surface is partly gapped out and reduced to a closed line of dirac cones . ] , the model exhibits a majorana fermi surface centered around @xmath148 as shown in a ) .
away from @xmath146 , the surface reduces to a line , which lies in the @xmath149 plane .
panels b ) and c ) show the behavior of the gapless line , when b ) increasing @xmath81 and c ) decreasing @xmath81 , and setting @xmath102 . ]
this behavior should be contrasted to what happens when increasing the unit cell in the @xmath144 direction , such that the enlarged unit cell is still incompatible with a bipartite coloring of the lattice .
one possible way to achieve this is to switch the @xmath33 and @xmath34 bonds in every second unit cell along the @xmath150 direction , shown in fig .
[ fig : kitaev2 ] b ) . note that this again breaks the isotropy of the original model in the coupling constants . similarly to the two previously discussed models , the region of parameter space with gapless excitations is again defined by the triangular inequality .
for the same reasons as stated above , the model with the enlarged unit cell has the same properties as the original one along the @xmath146 line in parameter space .
in contrast to the model with enlarged unit cell along the @xmath151 direction , discussed above , the surface remains stable , even when departing from this parametric condition .
the behavior of the surface is similar to the original model . in particular
, the dispersion around the gapless surface is linear in the normal direction , except at the phase transition to the gapped phases , where the surface shrinks to a single point with quadratic dispersion .
the reason for the stability of the surfaces can again be tracked to their relative displacement in momentum space a direct consequence that the unit cell is not compatible with the bipartite coloring of the lattice .
to a certain extent one can take the perspective that our results for the kitaev model on the hyperoctagon lattice complete a family of analytically tractable spin liquids of growing complexity starting from a two - dimensional dirac spin liquid on the honeycomb lattice over the intermediate step of a three - dimensional spin liquid with a line of gapless modes on the hyperhoneycomb lattice to finally a spin liquid with a full , two - dimensional surface of gapless modes for the hyperoctagon lattice . despite this ascent in complexity , we want to point out that all three instances share certain features such as rapidly decaying dynamical spin - spin correlation functions while differing in other aspects such as the nature of the phases induced by time - reversal symmetry breaking perturbations such as a magnetic field .
starting with the similarities , it is interesting to observe that independent of the nature of the manifold of their gapless modes all three spin liquids exhibit dynamical two - spin correlation functions that are identically zero beyond nearest neighbor separation .
this extremely rapid decrease was first observed in the context of the honeycomb model @xcite , but the very same arguments employed there also hold for the hyperhoneycomb and hyperoctagon lattices .
another parallel arises when distorting the couplings in the kitaev model such that one of the coupling constants becomes dominant ( i.e. such that the triangular inequality does not hold any longer ) and the majorana sector is gapped out , see the phase diagram in fig .
[ fig : phasediagram ] for comparison . in these gapped phases
the low - lying excitations are instead given by configurations , where some of the loop operators have negative eigenvalue .
while for the two - dimensional honeycomb lattices two types of excitations can be discerned and identified with the point - like ( electric and magnetic ) excitations of a @xmath66 gauge theory , the effective low - energy theory for the three - dimensional lattice is somewhat more elaborate .
mandal and surendran argued @xcite that there is only a single loop - like excitation in the gapped phases of the hyperhoneycomb model that exhibits non - trivial ( semionic ) braiding properties .
likely , a variant of their argument with similar conclusions can be applied to the hyperoctagon model as well .
a clear distinction between the three spin liquid phases arises when considering the effect of time - reversal symmetry breaking perturbations such as a magnetic field . for the honeycomb model
such a perturbation is of utmost interest as the gapless majorana modes are protected by time - reversal and even an infinitesimal magnetic field ( applied along the 111-direction ) gaps out the spin liquid into a gapped topological phase with non - abelian vortex excitations . on a more technical level
the reason for this drastic change induced by the magnetic field can be seen in terms of the symmetry class classification @xcite of the underlying free ( majorana ) fermion problem @xcite . in the absence of time - reversal symmetry breaking the model
is in symmetry class bdi , while in the presence of a magnetic field the symmetry class changes to class d. the latter allows for a @xmath152 classification of topological phases in two spatial dimensions .
indeed the two bands split by the magnetic field are characterized by chern numbers @xmath48 indicating the non - abelian nature of the gapped phase . when considering a similar line of arguments for the three - dimensional models a different picture emerges .
as noted earlier , the zero - energy modes of the hyperhoneycomb model are protected by time - reversal symmetry , while the ones of the hyperoctagon lattice are not .
so while we expect the gapless phase of the hyperhoneycomb model to gap out immediately in the presence of a magnetic field , this is far from obvious for the zero - energy modes of the hyperoctagon model as the spectrum is robust against any two - fermion term that does not break translation symmetry .
independent of whether a gap opens in the spectrum , it still holds that the symmetry class of the underlying free ( majorana ) fermion model changes from class bdi to class d in the presence of a time - reversal symmetry breaking term . however , in contrast to its two - dimensional counterpart symmetry class d does not harbor any topological phases in three dimensions @xcite . as a consequence , the three - dimensional systems can not be driven into a ( non - abelian ) topological phase by applying a magnetic field ( or any other time - reversal symmetry breaking perturbation ) .
however , one might still be able to employ similar ideas to the ones used by ryu in ref .
@xcite to stabilize a non - trivial topological phase by introducing additional ( orbital ) degrees of freedom such that the augmented model can be reformulated as a free fermion model in symmetry class diii .
the latter does have a @xmath152 classification in three dimensions and , thus , allows for three dimensional analogs of the topological phase in the honeycomb model .
finally , an interesting perspective emerges when recasting our results in the terminology conventionally used to characterize various spin liquid states @xcite . in this language ,
we have discovered a spin liquid with a _ spinon fermi surface _ that covers an extensive two - dimensional manifold in momentum space .
the quest to identify magnetic systems harboring such spinon fermi surfaces has typically inspired theorists to consider a slave - fermion approach where the fermion interacts with a _ fluctuating _
@xmath153 gauge field a situation that is notoriously hard to track analytically and
any progress coming at the expense of compromises on the level of various decoupling / mean - field approaches .
this situation should be contrasted to the current situation where we have stumbled upon a system with a spinon fermion surface with a much simpler and analytically exact description in terms of majorana fermions interacting with a _ static _ @xmath66 gauge field .
however , it is important to note that this difference is not a mere conceptual one , but one that has direct implications for thermodynamic observables such as the specific heat coefficient @xmath154 . for a @xmath153 spin liquid
the specific heat diverges as @xmath155 i.e. the specific heat coefficient @xmath156 _ diverges _ logarithmically at low temperatures @xcite . for our case of a spinon fermi surface emerging from majorana fermions interacting with a @xmath66 gauge field
we find @xmath157 i.e. the specific heat coefficient @xmath75 goes to a _ constant _ at low temperatures .
finally , this situation should be contrasted to the spin liquid with a fermi line , as it was found for the hyperhoneycomb lattice , where the specific heat grows as @xcite @xmath158 i.e. the specific heat coefficient @xmath75 _ vanishes _ in the limit of @xmath159 .
remarkably enough , this implies that a simple thermodynamic experiment could immediately distinguish these three seemingly equally exotic spin liquids .
returning to the perspective of the free ( majorana ) fermion system underlying our gapless spin liquid , there is one obvious bouquet of questions that we have not addressed in the manuscript at hand namely the various pairing instabilities that the fermi surface of our system might exhibit
. one might be particularly interested in asking what instabilities can be induced by additional interactions such as a heisenberg exchange argued to accompany the kitaev interactions in any microscopic description of iridate compounds @xcite .
the effective description of the hyperoctagon model in terms of _ spinless _ fermions suggests @xmath160-wave pairing as the natural candidate for opening a gap .
interaction terms of this type can indeed arise in a perturbative analysis of the heisenberg exchange .
the question of whether or not these terms lead to a collapse of the majorana fermi surface or even gap out all majorana modes in the system is left for future work . +
* acknowledgements. * we thank a. akhmerov , a. altland , p. becker - bohaty , s. parameswaran , f. pollmann , and especially a. rosch for insightful discussions .
st acknowledges the hospitality of the aspen center for physics where some of the ideas underlying this manuscript were perceived .
st is indebted to l. balents for a beginner s guide to 3d printing .
we acknowledge partial support from sfb tr 12 of the dfg .
in this appendix , we want to expand our discussion of possible magnetic materials candidates in space group i4@xmath1732 ( no . 214 ) .
the guiding idea in our analysis is to put the space group symmetries of i4@xmath1732 to work and look for various possible ways to fill the interstitial sites between the network of edge - sharing iro@xmath29 cages as illustrated in fig .
[ fig : o6cages ] of the main text .
in addition , we have to take into account that the fundamental building blocks of iro@xmath2 have valency @xmath161 and as such are looking for interstitial fillings that allow to chemically compensate this valence .
let us first consider the effect of the space group symmetries .
there are in total 48 symmetry operations in the space group i4@xmath1732 .
a generic point in the cubic cell is , thus , mapped to in total 48 distinct points ; the set of these points will , in the following , often be labeled by a single representative .
however , there are several high - symmetry points , respectively lines in the cubic cell , which are mapped to far fewer points . in total
, we can distinguish 5 types of lattice point , according to the number of distinct lattice points that can be reached by the symmetry operations .
\i ) the lattice point is mapped to 8 distinct lattice points in the unit cell .
this is possible in two inequivalent ways .
the resulting points form the two chiralities of the hyperoctagon lattice .
the representatives of the two possibilities are @xmath162 and @xmath163 . in the main text
, we chose the hyperoctagon lattice generated by @xmath164 ; thus , in the following we place the iridium atoms on this set of sites .
\ii ) the lattice point is mapped to 12 distinct lattice points in the unit cell .
this is again possible in two inequivalent ways .
the representatives are given by @xmath165 and @xmath166 .
these sites form effective lattices , which are deformations of the two chiral versions of the hyperkagome .
in fact , they can be identified as the medial lattices of the two chiral hyperoctagon lattices in i ) .
\iii ) the lattice point is mapped to 16 distinct lattice points in the unit cell .
there are infinitely many such possibilities , as long as the representative is chosen on the line @xmath167 ( except the high symmetry points already listed in i ) ) .
\iv ) the lattice point is mapped to 24 distinct lattice points in the unit cell .
there are many high - symmetry lines in the cubic unit cell , which lead to this behavior .
the oxygen sites are an example for this type of lattice points represented by @xmath168 .
\v ) the lattice point is mapped to 48 distinct lattice points in the unit cell , which applies to all lattice points that do nt lie on one of the above mentioned high - symmetry lines .
structures , indicated by the grey octahedra . panel a ) shows the crystal structure by placing atoms of valency @xmath169 on the sites of type i ) ( see text ) . in panel b ) , atoms with valency @xmath55 are placed on sites of type iii ) , which are generated by the representative @xmath170 . ]
this structure , imposed by the symmetry group , severely restricts the composition of possible compounds . assuming the presence of edge - sharing iro@xmath29 cages
, we note that the above analysis implies that there are 8 iro@xmath2 in a cubic unit cell .
thus , the remaining atoms must compensate a total valency of @xmath171 .
the latter can , for instance , be achieve by placing 8 atoms with valency @xmath169 on the remaining set of sites of type i ) .
the resulting compound is of the form airo@xmath2 , where a is one of the alkaline - earth elements ca , sr or ba .
another possibility is to place 16 atoms of valency @xmath55 on the sites of type iii ) .
this results in a material of the type a@xmath1iro@xmath2 , where a is one of the alkali atoms na or li .
the resulting crystal structures for both possibilities are visualized in fig .
[ fig : space_group ] .
to complement our discussion of the kitaev model on the hyperoctagon lattice , we will present a brief , self - contained summary of the kitaev model on the hyperhoneycomb lattice , an alternative trivalent 3d lattice , in this appendix . for clarity
, we will use similar notations and conventions as in the main text , but emphasize that in doing so we also closely follow the original solution of the kitaev model on the hyperhoneycomb lattice as discussed in some detail in ref . .
the hyperhoneycomb lattice is an alternative 3d lattice with a trivalent lattice structure which is illustrated in fig .
[ fig : hyperhoneycomb_unitcell ] .
its elementary building blocks are zig - zag chains running along the crystallographic @xmath172- and @xmath173-axises , which are coupled along the @xmath174-axis .
its crystal structure can be classified as a face centered orthorhombic lattice of space group number 70 .
of particular importance for our discussion is the fact that it is not spatially isotropic as the hyperoctagon lattice , but has one ` preferred ' direction the @xmath174 direction in fig .
[ fig : hyperhoneycomb_unitcell ] .
following our discussion in the main text relating the hyperoctagon and hyperkagome lattices as ( pre)medial lattices of each other , one can establish a similar set of relations for the hyperhoneycomb lattice as well .
the medial lattice of the hyperhoneycomb is a lattice of corner - sharing triangles illustrated in fig .
[ fig:3dkagome ] , which can be considered another generalization of the kagome lattice to three spatial dimension , albeit one distinct from the hyperkagome and one which we dub _ orthorhombic - kagome _ lattice .
its main motif are sheets of two parallel triangle lines , which are staggered in a rotated way as illustrated in fig .
[ fig:3dkagome ] .
the set of relations between the hyperhoneycomb and orthorhombic - kagome lattices as well as their relation to the pyrochlore and diamond lattices are summarized in fig .
[ fig : latticesummary2 ] . in comparison .
the hyperhoneycomb lattice is the medial lattice of the so - called orthorhombic - kagome lattice depicted in fig .
[ fig:3dkagome ] . like the hyper - kagome lattice
the orthorhombic - kagome lattice can be obtained from the pyrochlore lattice via depletion of 1/4 of the triangles , see the inset on the left .
the premedial lattice of the pyrochlore is the diamond lattice , which can be depleted by 1/4 of its bonds to obtain the hyperhoneycomb lattice . ]
similar to the hyperoctagon lattice we can define a covering of @xmath33 , @xmath34 , and @xmath35-couplings on the hyperhoneycomb lattice which is commensurate with a four - site unit cell .
this unit cell and related translation vectors as indicated in fig .
[ fig : hyperhoneycomb_unitcell ] .
the kitaev hamiltonian then takes the form @xmath175 where @xmath38 denotes the unit cell position .
similar to our analysis of the kitaev model on the hyperoctagon lattice we can identify conservative quantities for this model by considering the structure of closed loops , which again have length ten for this model .
for each elementary loop one can again identify a conserved quantity via the loop operators @xmath46 , which again have eigenvalues @xmath48 .
in contrast to the hyperoctagon lattice , the smallest volume enclosed by these elementary loops is now formed by _
four _ loops as illustrated in fig .
[ fig : loops_hyperhoneycomb ] . graphically speaking this constrains each ` tetraeder ' to have an even number of loops with eigenvalue @xmath54 . a counting argument similar to the one presented in our main analysis in section [ sec : flux_sectors ]
shows that there are @xmath64 different flux sectors , where @xmath65 is the number of unit cells .
numerical simulations @xcite indicate that the ground state indeed resides in the zero flux sector , which is why we restrict the following discussion to this sector . in order to solve the hamiltonian
, we proceed as in the main text and introduce four types of majorana fermions per site to write @xmath176 . introducing bond operators @xmath74 and choosing the zero - flux sector as a reference sector , we again obtain a free fermion hamiltonian of majorana fermions hopping in a static @xmath66 gauge field @xmath177 after a fourier transformation , the hamiltonian is straightforward to diagonalize .
the principle energy bands in the majorana spectrum are thereby found to be @xmath178}}\,,\\\end{gathered}\ ] ] where @xmath179 the momenta @xmath180 are defined as the coefficients of the reciprocal lattice vectors , i.e. @xmath181 with @xmath182 . .
the gapless line is located in the plane @xmath183 , which is indicated in grey .
the other panels show the behavior of the gapless line when a ) increasing @xmath81 and b ) decreasing @xmath81 , while setting @xmath102 .
the extent of the brillouine zone in the plane @xmath183 is indicated by the hexagon . ]
zero - energy solutions are obtained by setting the second term in the root to zero , which implies @xmath184 the first line of eq .
( [ eq : hh_constraint ] ) can be inverted to yield @xmath185 the second line determines the ( unique ) value of @xmath186 given @xmath187 and @xmath188 .
the gapless majorana modes thus form a _ line _ in momentum space with linear dispersion along the normal directions . the line of gapless modes and its dependence on the coupling constant @xmath83 setting @xmath189 is shown in fig [ fig : hyperhoneycomb_gapless ] . for this choice of parameters
the gapless line always lies in the @xmath183 plane , although this is no longer true when @xmath190 .
we note that when approaching the gapped phase at @xmath191 , @xmath192 , the gapless line shrinks to a point at @xmath170 .
s. k. choi , r. coldea , a. n. kolmogorov , t. lancaster , i. i. mazin , s. j. blundell , p. g. radaelli , yogesh singh , p. gegenwart , k. r. choi , s .- w .
cheong , p. j. baker , c. stock , and j. taylor , phys .
lett . * 108 * , 127204 ( 2012 ) .
further generalizations of the kitaev models also to lattices with higher vertex coordination number , in particular the kagome , triangular , hyperkagome and pyrochlore lattices have been studied @xcite .
it should however be noted that it is precisely the higher coordination number of the vertices in these lattices which prohibits to follow the same analytical route that can be used for the trivalent ones .
the relative sign for the two alternative definitions of the loop operator , and subsequent ` freedom ' to define magnetic flux , appears also for other trivalent lattices , for instance the two - dimensional square - octagon lattice . using the operator @xmath46 , eq .
, to define the flux ensures that the notion of 0/@xmath193-flux per plaquette is consistent with the one used by lieb @xcite .
in particular , the ground state of the square - octagon lattice resides in the _ full - flux sector _ according to lieb s theorem , as the loops have lengths four respectively eight . | motivated by the recent synthesis of @xmath0-li@xmath1iro@xmath2 a spin - orbit entangled @xmath3 mott insulator with a three - dimensional lattice structure of the ir@xmath4 ions we consider generalizations of the kitaev model believed to capture some of the microscopic interactions between the iridium moments on various trivalent lattice structures in three spatial dimensions .
of particular interest is the so - called hyperoctagon lattice the premedial lattice of the hyperkagome lattice , for which the ground state is a gapless quantum spin liquid where the gapless majorana modes form an extended two - dimensional majorana fermi surface .
we demonstrate that this majorana fermi surface is inherently protected by lattice symmetries and discuss possible instabilities .
we thus provide the first example of an analytically tractable microscopic model of interacting su(2 ) spin-1/2 degrees of freedom in three spatial dimensions that harbors a spin liquid with a two - dimensional spinon fermi surface . |
there exists substantial evidence that the bulges of nearby massive galaxies contain quiescent supermassive nuclear black holes ( @xmath13 m@xmath14 : magorrian et al .
1998 ; van der marel 1999 ; kormendy et al . 1998 ) . the almost ubiquitous presence of such objects , in combination with the possibility that nuclear black hole mass scales with stellar bulge mass ( i.e. , the magorrian relation : magorrian et al .
1998 ) , indicates that supermassive black holes are an essential part of spheroid formation , and that mass accretion onto nuclear black holes during stellar bulge formation likely resulted in periods of agn activity .
there also exists growing observational evidence that core formation and agn activity may be linked with ultraluminous infrared galaxy ( ulig ) \gtrsim 10^{12}$ ] l@xmath14 . ]
mergers , lending support to the evolutionary connection between uligs and qsos first proposed by sanders et al .
( 1988a ) .
the evidence can be summarized as follows : first , observations of massive concentrations of molecular gas in the nuclear regions of local uligs relative to that observed in normal galaxies ( e.g. kormendy & sanders 1992 ; downes & solomon 1998 ; bryant & scoville 1999 ) support the highly dissipative nature of uligs .
second , spectroscopic evidence that a significant fraction of uligs , most of which have `` warm '' infrared colors ( i.e. , 25@xmath15 m to 60@xmath15 m flux density ratio , @xmath16 ) , have seyfert - like emission line
spectra ( veilleux et al . 1995 ; veilleux , kim , & sanders 1999 ) may indicate that merger events are intimately connected with agn activity .
third , recent co(@xmath3 ) observations show the association of molecular gas with the agn nucleus of several double nucleus uligs ( evans et al . 1999 ; evans et al . 2000 ) , supporting the necessity of dissipative material for fueling agn activity .
fourth , the recent evidence of similarities in the space densities of high - redshift submillimeter ( scuba ) galaxies ( e.g. , smail , ivison , & blain 1997 ; barger et al .
1998 ; hughes et al . 1998 ; barger , cowie , & sanders 1999 ) and local giant elliptical galaxies has been used to infer that uligs are connected with massive galaxy formation ( barger et al .
1999 ; trentham 2000 ) . from the above
, it seems reasonable to assume that a significant fraction of qsos may have molecular gas rich hosts and that molecular gas may play an important role in fueling not only circumnuclear starbursts , but also agn activity in galaxies experiencing dissipative collapse . as a first step towards establishing the presence of abundant reservoirs of molecular gas in qsos , a survey of co(@xmath3 ) emission in a complete sample of low - redshift
, optically - selected qsos has been initiated .
in addition to detecting the possible fueling source for both agn and star formation activity , these observations provide a major missing link in comparing the properties of low - redshifts qsos and infrared luminous galaxies with both `` cool '' ( i.e. , 25@xmath15 m to 60@xmath15 m flux density ratio , @xmath17 ) and `` warm '' , seyfert - like ( @xmath18 ) mid - to - far infrared flux density ratios .
specifically , these data make it possible to determine whether the relative molecular gas properties of pg qsos and uligs are consistent with the hypothesis that cool and warm uligs represent successive stages in the evolution toward uv - excess qsos ( sanders et al . 1988a ) .
the co data presented here are complemented by previous studies of the far - infrared properties ( sanders et al .
1989a ) and host galaxy morphologies ( bahcall et al . 1997 ; surace 1998 ; surace & sanders 2000 : hereafter ss00 ) of optical qsos , and thus facilitate a more well - rounded comparison with the wealth of data available of low - redshift ( @xmath19 ) infrared luminous galaxies .
this paper is divided into 6 sections .
section 2 is a summary of the sample selection . in 3 ,
the general observing procedure is discussed , followed by the results in 4 .
a comparison of the co and infrared properties of the pg qsos with those of local infrared luminous galaxies and high redshift qsos is provided in 5 , along with a summary of molecular gas and infrared properties of the pg qsos relative to the optical morphologies of their host galaxies . finally ,
6 summarizes the paper . throughout this paper , @xmath20 km s@xmath21 mpc@xmath21 and @xmath22 are assumed .
the qsos observed as part of this co study were selected from a complete sample of 18 qsos defined by ss00 .
the reader is referred to ss00 for a detailed discussion of the sample selection , but it is briefly summarized here for completeness .
the selection criteria consisted of all @xmath23 qsos from the palomar green bright quasar survey ( pg bgs : schmidt & green 1983 ) having infrared - to-``big blue bump '' luminosities , @xmath24 and absolute blue magnitudes , @xmath1 .
the former criteria selects qsos with infrared excesses that overlap with warm infrared luminous galaxies ( see 5.2 ) .
there are two advantages to using the ss00 sample .
first , the far - infrared flux densities of these qsos have been measured using the infrared astronomical satellite ( iras ) data ( sanders et al .
1989a ) , making direct comparisons of their infrared and molecular gas properties with those of infrared luminous galaxies possible .
second , the molecular gas properties of these qsos can be compared with pre - existing optical images of the qso host galaxies to search for correlations of @xmath25 with the morphology of the host galaxies . while qsos with infrared excesses may at first seem to be unrepresentative of optical qsos , it must be remembered that at least two - thirds of all the pg qsos in the bright quasar survey show clear evidence of infrared emission consistent with thermal dust emission , with the infrared components contributing 20 - 40% of their total bolometric luminosity ( sanders et al . 1989a ) .
thus , the sample selected for this survey is simply the optical qsos with the highest probable dust content . of the 18 qsos in the ss00 sample , two of them ( izw1 and mrk 1014 )
have previous confirmed co(@xmath3 ) detections ( e.g. sanders , scoville , & soifer 1988 ; barvainis et al . 1989 ; solomon et al .
1997 ; schinnerer , eckart , & tacconi 1998 ) and two others ( pg 0838 + 770 and pg 1613 + 658 ) have unconfirmed co detections ( alloin et al .
in addition to observing pg 0838 + 770 and pg 1613 + 658 , eight pg qsos with @xmath26 ratios in the range of 0.361.1 were selected for co observations .
the sample is listed in table 1 .
aperture synthesis maps of co(@xmath3 ) emission in the sample of pg qsos were made with the owens valley radio observatory ( ovro ) millimeter array during 21 observing periods from 1999 april to 2000 october .
the array consists of six 10.4 m telescopes , and the longest observed baseline was 103 m .
each telescope was configured with 120@xmath274 mhz digital correlators .
all but one of the observations were done in the low - resolution configuration , providing a @xmath28 ( full width at half the maximum intensity , or fwhm ) synthesized beam with natural weighting . for each qso ,
the receivers were tuned to the expected frequency of the redshifted co emission line .
these redshifts were estimated from published optical emission line redshifts compiled by the nasa extragalactic database ( ned ) ; typically , the average or the most common redshift was adopted . during the observations ,
nearby quasars were observed every 25 minutes to monitor phase and gain variations , and several brighter quasars were observed to determine the passband structure .
observations of uranus were also made for absolute flux calibration ; typical uncertainties in the flux calibration are 15 - 20% .
a journal of observations is provided in table 2 .
the data were reduced and calibrated using the standard owens valley array program mma ( scoville et al .
the data were then exported to the mapping program difmap ( shepherd , pearson , & taylor 1995 ) and to the nrao software package aips for spectral line extraction .
figure 1 shows the integrated co(@xmath3 ) intensity maps for the detected pg qsos ( integrated intensity values are listed in table 3 ) .
six out of the ten qsos observed were detected in co ; all of the detections have a signal - to - noise ratio , s / n , greater than 4.5@xmath29 .
a comparison was done between the co beam centroid positions and the optical qso positions , the latter of which were extracted from the usno - a2.0 catalog . with the exception of pg 1351 + 640 ( 1.44@xmath30 offset ) and
pg 1613 + 658 ( 1.05@xmath30 offset ) , all the co centroid and optical positions agreed to within 0.9 .
of the four non - detections , two are of the qsos with the highest redshifts in the sample ( pg 1402 + 261 and pg 1202 + 281 ) .
for all four non - detections , the 1.0 ghz continuum channel data was examined to search for co emission over a broader ( albeit less sensitive ) bandwidth than covered by the spectrometer .
such searches yielded null results . in the present sample of ten qsos ,
pg 0007 + 106 is the only one having detectable 3 mm continuum emission .
the unresolved continuum emission is observed to be at ra = 00:10:30.98 dec = + 10:58:29.8 ( j2000.0 ) , which is within 0.75@xmath31 of the usno - a2.0 optical position , and the continuum flux density is measured to be 0.1 jy .
figure 2 shows the co(@xmath3 ) spectra of the six detected qsos .
the co emission line widths are varied , ranging from fwhm of @xmath32 km s@xmath21 for pg 1415 + 451 and pg 0838 + 770 to @xmath33 km s@xmath21 for pg 1613 + 658 . all of co emission line centroids are within 300 km s@xmath21 of the adopted optical redshifts , indicating that the optical emission lines are only moderately red / blueshifted relative to the systemic velocity .
both pg 0838 + 770 and pg 1613 + 658 are detected in co ; the emission line widths and redshifts are similar to those reported by alloin et al .
( 1992 ) , however , the integrated co line intensity of pg 0838 + 770 is only 40% of that reported by alloin et al .
( 1992 ) . by comparison
, the co line intensity of pg 1613 + 658 agrees with the alloin et al .
( 1992 ) measurement to within 12% ( i.e. within the error of the ovro measurements ) .
while the co emission line of pg 1415 + 451 was detected with a signal - to - noise of 5.2@xmath29 , the narrow appearance of the emission feature may be an indication that the `` line '' is actually an artifact ; i.e. , a spurious interference spike created at the phase center of the interferometer .
we believe that the emission feature is authentic based on a combination of the following three facts : first , the emission feature spans five independent ( i.e. , unbinned ) channels .
second , while the phase center of the pg 1415 + 451 observations is within the beam of the emission feature , and the centroid position of the feature is @xmath34 closer to the optical qso position ( obtained from the usno
a2.0 catalog ) than the phase center is to the optical qso position .
third , the co observation of pg 0838 + 770 , which has a co line width less than a factor of two wider than the emission feature observed for pg 1415 + 451 , was a confirmation of prior observations reported by alloin et al .
( 1992 ) , and is thus authentic .
it is thus likely that the host galaxies of both qsos simply have molecular disks that are being observed nearly face - on
. table 3 summarizes the emission line properties of the ten pg qsos .
the luminosity distance for a source at a given redshift is , @xmath35 @xmath36 .
\eqno(1)\ ] ] given the measured flux , @xmath37 , the co luminosity of a source at redshift @xmath38 is , @xmath39 where @xmath40 is the speed of light , @xmath41 is the boltzmann constant , and @xmath42 is the observed frequency . in terms of useful units , @xmath11 can be written as , @xmath43 @xmath44 .
\eqno(3)\ ] ] to calculate the mass of molecular gas in the pg qsos , the reasonable assumption is made that the co emission is optically thick and thermalized ( see discussion section ) , and that it originates in gravitationally bound molecular clouds .
thus , the ratio of the h@xmath45 mass and the co luminosity is given by , @xmath46^{-1 } , \eqno(4)\ ] ] where @xmath47h@xmath48 and @xmath49 are the density of h@xmath45 and brightness temperature for the co(1@xmath500 ) transition ( scoville & sanders 1987 ; solomon , downes , & radford 1992 ) .
multitransition co surveys of molecular clouds in the milky way ( e.g. sanders et al .
1993 ) , and in nearby starburst galaxies ( e.g. gsten et al .
1993 ) have shown that hotter clouds tend of be denser such that the density and temperature dependencies tend to cancel each other .
the variation in the value of @xmath51 is approximately a factor of 2 for a wide range of kinetic temperatures , gas densities , and co abundance ( e.g. @xmath52 [ k km s@xmath21 pc@xmath53^{-1}$ ] : radford , solomon , & downes 1991 ) .
more recent dynamical mass estimates of a low - redshift infrared galaxy sample observed in co with the plateau de bure interferometer indicate that @xmath51 can be as low as 1 @xmath54 [ k km s@xmath21 pc@xmath53^{-1}$ ] ( downes & solomon 1998 ) .
we adopt a value of 4 @xmath54 ( k km s@xmath21 pc@xmath55 for @xmath51 , which is similar to the value determined for the bulk of the molecular gas in the disk of the milky way ( scoville & sanders 1987 ; strong et al .
1988 ) . in the four cases where only upper limits to the co flux were determined ,
the average fwhm velocity of all the pg qsos detected in co to date was adopted ( 280 km s@xmath21 ) .
conservative upper limits to the co luminosity and molecular gas mass were then calculated from the 3@xmath29 upper limits to the co flux density .
the detection of co emission in the host galaxies of these pg qsos provides direct evidence that these infrared - excess qsos live in host galaxies containing significant supplies of molecular gas .
it also supports the hypothesis that galaxies with a molecular gas - rich interstellar medium ( ism ) may be capable of igniting and sustaining both agn and star formation activity in qso host galaxies .
the measured co luminosities for the six detected qsos in table 3 translate into molecular gas masses of @xmath56 m@xmath14 , which , along with the additional two qsos izw1 and mrk 1014 , span the full range of masses derived for luminous infrared galaxies ( ligs ) = 10^{11.0 - 11.99}$ ] l@xmath14 . ] and uligs .
these six pg qsos are among the faintest co(@xmath3 ) sources yet detected at redshifts @xmath57 .
an interesting issue is whether or not the co luminosities of these sources , and perhaps the co luminosities of other pg qsos as well , could have been predicted apriori . consider the plot of co flux versus the 100@xmath15 m flux density of the pg qso sample , in addition to a comparative sample of @xmath57 warm and cool uligs observed in co to date ( figure 3 ) .
solomon et al . (
1997 ) use the tightness of this correlation , in combination with the small sizes of the co emitting regions in uligs , to argue that the dust emission at 100@xmath15 m and the co(1@xmath500 ) emission is optically thick ( see their figure 4 ) .
the @xmath58 resolution of the present co dataset is insufficient to determine how compact the co emitting regions in these pg qsos are ( e.g. , @xmath58 corresponds to @xmath59 kpc at @xmath60 ) . for our purposes
, the fact that all of the pg qsos detected in co to date also fall along this correlation indicates that the co fluxes ( and thus , the molecular gas masses ) of optical qsos may potentially be estimated from the 100@xmath15 m flux density , @xmath61 , to within a factor of a few if @xmath61 is known to high accuracy .
note , however , that the error in @xmath61 of any _ iras _ measurements exceeds 25% below 0.5 jy ; improved far - infrared measurements of qsos will be possible via observations with the space infrared telescope facility ( sirtf ) . with the detection of molecular gas in the six of the ten optical qsos listed in table 3 , plus the previous detection of izw1 and mrk 1014 , physically meaningful comparisons between qsos and infrared - selected galaxies can be better addressed .
the remainder of the discussion section will thus be focussed on comparing the infrared and co luminosities of the eight qsos detected in co to date with infrared luminous galaxies , followed by a brief discussion of recent images of the host galaxies of these pg qsos ( bahcall et al . 1997 ; ss00 ) .
the four qsos with co non - detections will be discussed where applicable .
figure 4 is a plot of co luminosity versus redshift of the 12 qsos . in order to examine how the co luminosity of these pg qsos compares with that of the most co luminous galaxies presently known at
any given redshift , the infrared luminous galaxies observed in co(@xmath3 ) by the sanders et al .
( 1989b ) , sanders , scoville , & soifer ( 1991 ) , and solomon et al .
( 1997 ) surveys have also been plotted .
all but eight of the infrared luminous galaxies at @xmath62 are uligs .
three features of figure 4 are readily apparent .
first , the co luminosity range of all twelve qsos spans more than an order of magnitude .
this range is significantly more than that of the cool uligs , all of which have @xmath63 k km s@xmath21 pc@xmath64 over the same redshift range . second , the three qsos with the highest @xmath11 ( izw1 , pg 1613 + 658 , and mrk 1014 ) fall within the scatter of the cool uligs distribution , whereas the remaining nine qsos form a distinct group with lower values of co luminosity for their given redshift .
third , seven of the eight qsos detected in co have a range in @xmath11 similar to that of the warm uligs .
thus , while the @xmath11 values of the qsos are diverse , there is also evidence of overlap in @xmath11 between these optically - selected and infrared - selected samples .
consider now the infrared luminosity of both samples relative to the co luminosity .
figure 5 shows two ways in which infrared and co data of infrared luminous galaxies are commonly plotted .
the top panel ( figure 5a ) is a plot of @xmath6 versus @xmath11 , which shows the relationship between the fuel available for star formation and agn activity ( @xmath11 ) and the luminosity of the imbedded sources ( @xmath6 ) .
the bottom panel ( figure 5b ) is a plot of @xmath10 versus @xmath6 . in starburst galaxies , @xmath10
is commonly referred to as the star formation efficiency , it is a measure of the number of high - mass stars ( as traced by @xmath12 ) created per unit molecular gas mass ( as traced by @xmath11 ) .
an examination of figure 5a shows that the quantity @xmath6 increases as @xmath11 increases , and for any given co luminosity bin , the qsos are in the upper half of the distribution . in figure 5b ,
the qsos populate the upper half of the lig - ulig distribution of @xmath65 .
note also that , in general , the warm infrared galaxies and qsos populate similar areas of both plots .
the interpretation of figure 5 depends on whether the infrared emission in qsos is due to dust heating by agn or massive stars .
first , consider the former possibility : sanders et al .
( 1988a ) proposed that uligs are dust - enshrouded qsos ; in this scenario , the optical qso is revealed after most of the molecular gas is consumed by star formation and agn activity , and after most of the dust is destroyed / blown out by winds associated with supernovae and agn . during this process , uligs pass from a cool infrared phase where vigorous star formation and black hole creation / accretion has commenced , through a warm infrared phase ( see also sanders 1988b ) .
this idea is further addressed by sanders et al .
( 1989a ) , who propose that the infrared emission from pg qsos is thermal emission from dust that has absorbed 2040% of the emission emanating from the qso ( i.e. , @xmath66 l_{\rm bol}$ ] , where @xmath67 is the bolometric luminosity ) . given this ,
the quantity @xmath68 is not a measure of the star formation efficiency in qsos , but is an indication that the agn is contributing significantly to heating the dust in qsos ( see also evans et al .
1998 ) .
if the dust - enshrouded quasar model of uligs is correct and the ratio @xmath69 for both uligs and qsos is due to covering factor and not due to orientation effects ( e.g. barthel 1989 ) , optically - selected qsos must have had a covering factor of unity in their past ulig phase .
consider now where the qsos would be on figure 5a and 5b if the qso nuclei were completely enshrouded in dust ( i.e. @xmath70 ) .
this increases @xmath71 of the qsos by @xmath720.30.5 in figure 5a , and shows to first order a possible range in locations of enshrouded qsos on this diagram . in figure 5b ,
the quantity @xmath68 of the qsos is increased to extreme values ( @xmath73 ) .
note that , in both panels , uligs with high ( @xmath74 ) @xmath65 occupy the region covered by the qsos , similar to what would be expected if these uligs were dust - enshrouded qsos . as mentioned above ,
the second possible interpretation of the infrared emission in qsos is that @xmath6 is primarily due to dust heating by massive stars .
thus , high values of @xmath6 ( figure 5a ) and @xmath10 ( figure 5b ) would result from an extremely efficient production of high - mass stars per unit molecular gas mass in qsos relative to most ligs and uligs .
the attraction of this scenario is that both stars forming in the central regions of galaxies and agn contribute significantly to the overall energy budget of their host galaxies , and that agn activity is somehow strongly dependent on an accompanying starburst .
indeed , observed correlations between the masses of quiescent supermassive nuclear black holes and both the stellar bulge masses ( e.g. , magorrian et al .
1998 ) and velocity dispersions ( ferrarese & merritt 2000 ; gebhardt et al .
2000 ) of nearby normal galaxies suggests that the accretion of mass onto nuclear black holes may accompany stellar bulge formation .
however , whether or not a significant fraction of the stellar bulge population of the present sample of pg qsos is in the process of being formed can not be addressed with the present data set .
an obvious issue to consider at this point is whether or not the diversity of @xmath11 and @xmath6 of these 12 qsos is at all correlated with the galaxy type or evolutionary state of their host galaxies .
an examination of optical and near - infrared images of all 12 qsos ( bahcall et al .
1997 ; surace et al .
1998 ; ss00 ) reveals that two of the qsos with the highest @xmath11 ( mrk 1014 and pg 1613 + 658 ) have irregular morphologies consistent with an advanced merger of two molecular gas - rich disk galaxies ( see simulations by barnes & hernquist 1996 ) , and the other ( izw1 ) has asymmetric spiral arms , suggestive of a strong interaction with a much smaller mass companion .
the remaining nine qsos have host galaxies that appear to be a mixture of barred spiral galaxies , elliptical galaxies , galaxies with indeterminant features , and at least one ongoing merger ( i.e. , pg 0007 + 106 ) .
the morphologies are suggestive of a diversity in paths that may ultimately lead to qso activity ( ss00 ) .
first , the qsos having the highest values of @xmath11 appear to be ongoing major mergers ( collisions of two or more molecular gas - rich disk galaxies where prominent tidal features are still visible ) .
thus , in terms of @xmath25 , @xmath6 , and host morphologies , these qsos are similar to uligs .
second , the less co - luminous qsos with barred spiral galaxy hosts have a less clear origin , however , the bar likely acts as the means by which molecular gas in the spiral arms is torqued and drained of enough angular momentum to agglomerate in the nuclear regions ( e.g. sakamoto et al . 1999 ) .
these barred spiral galaxies may have lower @xmath11 because only one , massive molecular gas - rich galaxy is involved ( though the bar instability may be triggered via an interaction with a satellite galaxy ) .
third , the qsos having host galaxies with an indeterminant classification may be spiral galaxies with features too faint to discern with the sensitivity of the ss00 observations , fully evolved elliptical galaxies , or well - evolved major mergers which are evolving into elliptical galaxies ( see simulations by barnes & hernquist 1996 ) .
thus , these host galaxies may have low @xmath11 because the host galaxy is intrinsically molecular gas poor , or because the molecular gas has already been consumed via both circumnuclear star formation and the qso .
finally , the qsos with elliptical host galaxies may be the end stage of the merger one two disk galaxies , or elliptical host galaxies that have cannibalized a satellite galaxy ( such as the @xmath75 galaxy iras psc 09104 + 4109 : e.g. evans et al . 1998 ) .
a significant number of qsos have been detected at high redshift ( @xmath76 : barvainis et al .
1994 , 1998 ; ohta et al . 1996
; omont et al . 1996 ; guilloteau et al . 1997 ; downes et al .
these qsos , many of which were initially selected for co observations based on their strong rest - frame infrared luminosity , are likely among the most molecular gas rich qsos at their redshift .
figure 6 is a plot of the co luminosity of the eight pg qsos detected in co to date , as well as all of the moderate ( @xmath77 ) and high - redshift co luminous qsos , versus the redshift .
several of the qsos at high redshift are known to be gravitationally lensed ; in these cases , both observed and the intrinsic co luminosities are plotted .
based on the present sample , there is no convincing evidence for evolution of molecular gas mass of the most co luminous qsos at any given redshift as a function of increasing redshift .
the ratio of @xmath11 of bri 1335 - 0415 ( @xmath78 ) to that of the most co luminous low - redshift qso ( mrk 1014 : @xmath79 ) is three .
however , given the uncertainty in the @xmath11to@xmath80h@xmath48 conversion factor , @xmath51 , the factor of three is also consistent with no evolution . the derived molecular gas masses of the most co luminous qsos at any given redshift are significantly below the average stellar mass of present - day elliptical galaxies ( @xmath72 several@xmath81 m@xmath14 ) . given that massive elliptical galaxies are a likely end stage for many qso host galaxies ( e.g. , as inferred from bahcall et al .
1997 or the magorrian relation ) , the mass discrepancy can be rectified by concluding that present - day massive galaxies are created from the merger of smaller galaxies , and/or that galaxies with molecular gas masses in excess of several@xmath82 m@xmath14 quickly consume the gas by forming high mass stars ( e.g. evans et al .
1996 , 1998 ) .
millimeter ( co ) observations of 10 infrared - excess , optically - selected palomar - green qsos with redshifts in the range @xmath0 have been presented .
these observations represent the first step towards detecting molecular gas in a complete sample of optical qsos .
the following conclusions have been reached : \1 ) six out of 10 qsos in the sample were detected in co@xmath83 .
the co detections in pg 1613 + 658 and pg 0838 + 770 confirm previously reported detections by alloin et al .
these detections , combined with previous detections of co in the pg qsos izw1 and mrk 1014 , provide strong support for the possibility that molecular gas is a likely source of fuel for both agn and star formation in qso host galaxies .
\2 ) all of the qsos have high values of @xmath5 relative to most @xmath19 infrared luminous galaxies of the same @xmath6 .
this is because qsos contribute significantly to the heating of dust in their host galaxies , and/or because the dust is heated by massive stars formed with high efficiency ( i.e. , per unit molecular gas mass ) relative to most infrared luminous galaxies .
if large values of @xmath84 are achieved primarily through dust heating by qsos , a significant fraction of uligs with similar values of @xmath85 may also contain buried agn . alternatively ,
if high @xmath10 is due primarily to star formation , then high - mass star formation may be intimately connected to the qso phenomenon .
\3 ) the warm infrared luminous galaxies , which by definition have seyfert - like mid - to - far infrared flux density ratios ( i.e. , @xmath86 ) , have similar ranges of @xmath5 and @xmath11 as the infrared - excess pg qsos .
\4 ) a comparison of the co luminosities and optical morphologies of the qso host galaxies indicates that qsos with the highest infrared and co luminosities reside in morphologically disturbed galaxies , whereas the rest have a mixture of barred spiral host galaxies and `` featureless '' host galaxies .
\5 ) there is no convincing evidence of evolution in the molecular gas content of the most co luminous qsos in the redshift range @xmath87 .
the most co luminous high - redshift qso ( bri 1335 - 0415 ) has an @xmath25 a factor of three greater than the most co luminous low - redshift qso ( mrk 1014 ) . however , given the uncertainty in the @xmath25to@xmath80h@xmath48 conversion factor , @xmath51 , the factor of three is also consistent with no evolution .
we thank the staff and postdoctoral scholars of the owens valley millimeter array for their support both during and after the observations were obtained .
ase also thanks j. carpenter and a. baker for useful discussions and assistance during the preparation of this manuscript , and the referee for comments that led to the clarification of several key points .
ase was supported by rf9736d and ast 0080881 .
jas was supported by the jet propulsion laboratory , california institute of technology , under contract with nasa .
the owens valley millimeter array is a radio telescope facility operated by the california institute of technology and is supported by nsf grant ast 9981546 .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory .
integrated intensity maps of the six pg qsos detected in co(@xmath3 ) .
a ) pg 0838 + 770 - contours are plotted as 1@xmath88 .
the peak intensity is 0.017 jy beam@xmath21 and corresponds to the position ra=08:44:45.22 dec=+76:53:08.96 ( j2000.0 ) .
b ) pg 1119 + 120 - contours are plotted as 1@xmath89 .
the peak intensity is 0.014 jy beam@xmath21 , and corresponds to the position ra=11:21:47.12 dec=+11:44:18.30 ( j2000.0 ) .
c ) pg 1351 + 640 - contours are plotted as 1@xmath90 ; the peak intensity is 0.0055 jy beam@xmath21 , and corresponds to the position ra=13:53:15.62 dec=+63:45:45.72 ( j2000.0 ) .
d ) pg 1415 + 451 - contours are plotted as 1@xmath91 ; the peak intensity is 0.051 jy beam@xmath21 , and corresponds to the position ra=14:17:00.76 dec=+44:56:06.50 ( j2000.0 ) .
e ) pg 1440 + 356 - contours are plotted as 1@xmath92 ; the peak intensity is 0.016 jy beam@xmath21 , and corresponds to the position ra=14:42:07.48 dec=+35:26:22.33 ( j2000.0 ) .
f ) pg 1613 + 658 - contours are plotted as 1@xmath93 ; the peak intensity is 0.012 jy beam@xmath21 , and corresponds to the position ra=16:13:57.15 dec=+65:43:09.62 ( j2000.0 ) . 0.1 in figure 2 .
co(@xmath3 ) spectra of the six detected pg qsos . for pg 0838 + 770 and pg 1415 + 451 ,
the data have been smoothed to a resolution of 16 mhz with a 8 mhz sampling . for pg 1119 + 120
, the data have been smoothed to a resolution of 32 mhz with a 16 mhz sampling . for pg 1351 + 640
, the data have been smoothed to a resolution of 40 mhz with a 20 mhz sampling . for pg 1440 + 356 and pg 1613 + 658 ,
the data have been smoothed to a resolution of 24 mhz with a 12 mhz sampling . 0.1 in figure 3 . a plot of the co flux vs. the 100 @xmath15 m flux density for pg qsos and uligs detected to date in co. the co data for izw1 and mrk 1014 were obtained from barvainis et al .
( 1989 ) and solomon et al .
( 1997 ) , respectively .
the uligs co data were obtained from sanders et al .
( 1989b ) , sanders , scoville , & soifer ( 1988 , 1991 ) , solomon et al .
( 1997 ) , and evans et al .
the 100@xmath15 m flux densities were obtained from sanders et al .
( 1989a , 1991 ) and solomon et al .
arrows denote 3@xmath29 upper limits on the co luminosity of pg 1126 - 041 , pg 1202 + 281 , and pg 1402 + 261 .
only upper limits on @xmath11 and @xmath61 currently exist for pg 0007 + 106 , and thus pg 0007 + 106 is not included on the plot .
adapted from figure 4 of solomon et al .
( 1997 ) . 0.1 in figure 4 .
a plot of @xmath94 versus redshift ( @xmath38 ) for the low-@xmath38 pg qso sample , a flux - limited sample ( @xmath95 jy ) of infrared luminous galaxies and a sample of ultraluminous infrared galaxies .
the vertical dashed lines represent the upper and lower @xmath38 boundaries of the pg qso sample .
the lig and uligs data have been obtained from the same sources as in figure 3 .
arrows denote 3@xmath29 upper limits on the co luminosity of pg 1126 - 041 , pg 1202 + 281 , and pg 1402 + 261 . 0.1 in figure 5 .
a ) a plot of @xmath6 vs. @xmath11 for the low-@xmath38 qso sample , a flux - limited sample ( @xmath96 jy ) of infrared luminous galaxies and a sample of ultraluminous infrared galaxies .
arrows denote 3@xmath29 upper limits on the co luminosity of pg 1126 - 041 , pg 1202 + 281 , and pg 1402 + 261 .
b ) a plot of @xmath65 vs. @xmath6 for the low-@xmath38 qso sample , a flux - limited sample ( @xmath97 jy ) of infrared luminous galaxies and a sample of ultraluminous infrared galaxies .
arrows denote 3@xmath29 lower limits on @xmath65 of pg 1126 - 041 , pg 1202 + 281 , and pg 1402 + 261 . for simplicity ,
all of the cool infrared luminous galaxies are plotted as plus signs . 0.1 in figure 6 . a plot of @xmath94 versus redshift ( @xmath38 ) for the low-@xmath38 pg qso sample and the high - redshift qsos detected in co to date .
the moderate and high - redshift qso data have been obtained from the following references : 3c 48 ( @xmath98 : scoville et al . 1993 ) , h1413 + 117 ( @xmath99 : barvainis et al .
1994 ) , mg 0414 + 0534 ( @xmath100 : barvainis et al .
1998 ) , apm 08279 + 5255 ( @xmath101 : downes et al . 1999 ) , bri 1335 - 0415 ( @xmath102 : guilloteau et al .
1997 ) , and br 1202 - 0725 ( @xmath103 : omont et al .
1996 ) . for gravitationally
lensed qsos , @xmath11 is plotted in terms of both the observed value and the intrinsic value . for mg 0414 + 0534
, the amplification is unknown , thus mg 0414 + 0534 is plotted with a solid line extending downward from the observed @xmath11 . adapted from figure 3 of frayer et al .
( 1999 ) .
lcclccc pg 0007 + 106 = iiizw2 & 00:10:30.98 & + 10:58:29.8 & 0.089 & 11.63 & 12.21 & 0.36 pg 0838 + 770 & 08:44:45.36 & + 76:53:09.2 & 0.131 & 11.60 & 12.01 & 0.63 pg 1119 + 120 = mrk 734@xmath104 & 11:21:47.15 & + 11:44:19.0 & 0.050 & 11.17 & 11.47 & 1.00 pg 1126 - 041 = mrk 1298 & 11:29:16.72 & -04:24:07.6 & 0.060 & 11.47 & 11.94 & 0.50 pg 1202 + 281 & 12:04:42.12 & + 27:54:11.8 & 0.165 & 11.78 & 12.07 & 1.05 pg 1351 + 640 & 13:53:15.83 & + 63:45:45.6 & 0.088 & 11.82 & 12.37 & 0.40 pg 1402 + 261 = ton182 & 14:05:16.21 & + 25:55:34.1 & 0.164 & 11.85 & 12.33 & 0.49 pg 1415 + 451 & 14:17:00.84 & + 44:56:06.5 & 0.114 & 11.48 & 11.83 & 0.80
pg 1440 + 356 = mrk 478 & 14:42:07.48 & + 35:26:23.1 & 0.079 & 11.62 & 11.95 & 0.87 pg 1613 + 658 = mrk 876 & 16:13:57.21 & + 65:43:10.6 & 0.129 & 11.99 & 12.20 & 1.58 lllllcl pg 0007 + 106 & 2000 oct 11 & 0106 + 013 & 01:08:38.72 & + 01:35:00.38 & 1.82 & 3c454.3,3c84 & 2000 oct 12 & & & & & pg 0838 + 770 & 2000 april 24 & 1044 + 719 & 10:48:27.63 & + 71:43:35.84 & 1.36 & 3c 273 & 2000 april 25 & & & & & pg 1119 + 120 & 2000 march 24 & 1055 + 018 & 10:58:29.57 & + 01:33:58.83 & 2.11 & 3c 273 & 2000 april 19 & & & & & pg 1126 - 041 & 1999 may 30 & 1055 + 018 & 10:58:29.57 & + 01:33:58.83 & 2.14 & 3c 273 & 1999 may 31 & & & & & pg 1202 + 281 & 1999 may 02 & 1156 + 295 & 11:59:31.80 & + 29:14:43.37 & 3.41 & 3c 273 & 1999 may 03 & & & & & 3c 84,3c 273,3c 454.3 pg 1351 + 640 & 1999 may 15 & 1044 + 719 & 10:48:27.63 & + 71:43:35.84 & 0.62 & 3c 273 & 1999 may 18 & & & & & & 1999 may 20 & & & & & pg 1402 + 261 & 2000 may 21 & 1308 + 326 & 13:10:28.62 & + 32:20:43.42 & 0.74 & 3c 273 & 2000 oct 11 & & & & 0.68 & 3c273,3c345 & 2000 oct 18 & & & & 0.68 & 3c273 pg 1415 + 451 & 2000 may 07 & 1308 + 326 & 13:10:28.62 & + 32:20:43.42 & 0.67 & 3c 273 pg 1440 + 356 & 1999 may 22 & 1611 + 343 & 16:13:40.99 & + 34:12:47.67 & 1.80 & 3c 273,3c 454.3 & 1999 june 03 & & & & & 3c 84,3c 111,3c 273 , 3c 345 pg 1613 + 658 & 1999 april 19 & 1803 + 784 & 18:00:45.48 & + 78:28:03.85 & 1.69 & 3c 273 & 1999 april 25 & & & & & lcllcrr pg 0007 + 106 & 360 & & & @xmath105 & @xmath106 & @xmath107 pg 0838 + 770 & 540 & 0.132 & 90 & 3.4@xmath108 & @xmath109 & @xmath110 pg 1119 + 120 & 200 & 0.050 & 260 & 4.5@xmath111 & @xmath112 & @xmath113 pg 1126 - 041 & 240 & & & @xmath114 & @xmath115 & @xmath116 pg
1202 + 281 & 685 & & & @xmath117 & @xmath118 & @xmath119 pg 1351 + 640 & 360 & 0.088 & 230 & 4.6@xmath120 & @xmath121 & @xmath122 pg 1402 + 261 & 680 & & & @xmath123 & @xmath124 & @xmath125 pg 1415 + 451 & 470 & 0.114 & 50 & 3.3@xmath126 & @xmath127 & @xmath128 pg 1440 + 356 & 320 & 0.078 & 370 & 8.7@xmath129 & @xmath130 & @xmath131 pg 1613 + 658 & 530 & 0.129 & 490 & 8.5@xmath120 & @xmath132 & @xmath133 | the initial results of a millimeter ( co ) survey of infrared - excess , optically - selected quasars from the palomar - green ( pg ) bright quasar survey ( bqs ) with redshifts in the range @xmath0 are presented .
these observations represent the first step towards establishing with a complete sample whether or not quasi - stellar objects ( qsos ) reside in molecular gas - rich galaxies , as well as towards determining how the infrared and molecular gas properties of qsos compare with those of ultraluminous infrared galaxies ( uligs ) , which are a possible evolutionary precursor of qsos .
the sample consists of qsos having absolute blue magnitudes , @xmath1 , and infrared excesses , @xmath2 , where the contribution to the bolometric luminosity of infrared thermal dust emission for all pg qsos is typically 2040% .
six out of 10 of the qsos observed are detected in the co(@xmath3 ) emission line ; two detections confirm previous , less sensitive detections of co(@xmath3 ) in pg 1613 + 658 and pg 0838 + 770 , and four additional qsos are detected for the first time ( pg 1119 + 120 , pg 1351 + 640 , pg 1415 + 451 , and pg 1440 + 356 ) .
these six detections , plus two previous detections of co in izw1 and mrk 1014 , bring the total number of @xmath4 infrared - excess pg qsos detected in co to date to eight , and provide possible evidence that , in addition to fueling star formation , molecular gas may also serve as a primary source of fuel for qso activity .
both the eight qsos detected in co and the four qsos with non - detections have high infrared - to - co luminosity ratios , @xmath5 , relative to most infrared luminous galaxies of the same @xmath6 .
the placement of these qsos on the @xmath7 plane may be due to significant contributions to dust heated by the qso in their host galaxies , due to dust heated by massive stars formed with high efficiency ( i.e. , per unit molecular gas mass ) relative to most infrared luminous galaxies , or a combination of both .
if the observed high values of @xmath8 are primarily due to dust heating by qsos , a significant fraction of uligs with similar values of @xmath9 may also contain buried agn . alternatively ,
if high @xmath10 is due primarily to star formation , then an enhanced star formation rate may be intimately connected to the qso phenomenon . a comparison of the infrared and co luminosities of the eight detected and four undetected qsos with the optical morphologies of their host galaxies shows that the three qsos with @xmath6 and @xmath11 similar to uligs appear to reside in morphologically disturbed galaxies ( i.e. , ongoing major mergers involving two or more gas - rich disk galaxies ) , whereas the host galaxies of the remaining eight qsos with lower @xmath12 and @xmath11 appear to be a mixture of barred spiral host galaxies , elliptical galaxies , galaxies with an indeterminant classification , and at least one ongoing major merger . |
convection near solar surface has strongly non - local and dynamical character .
hence numerical simulation provide useful information on the structure spatial scales by convection and help to construct consistent models of the physical processes underlying the observed solar phenomena .
we investigate effects of compressibility and weak of magnetic field on formation non - local structure of convection using realistic physics and conservative tvd numerical scheme of godunov type .
the previous simulations were confined by small computational domain and studied processes on scales order size of granulation [ @xcite ] . in order to investigate collective interaction of convective modes different scales and process of formation of supergranulation we conducted calculation in three dimensional computational box by size 30 mm in horizontal direction and by size 18 mm in vertical direction .
we take distribution of the main thermodynamic variables by radius due to standard solar model [ @xcite ] with parameters @xmath0 , where @xmath1 and @xmath2 are hydrogen and helium abundance by mass , and @xmath3 is the ratio of mixing length to pressure scale height in convection region .
we use opal opacities and equations of state for solar matter [ @xcite ] .
we solve fully compressible nonideal magnetohydrodynamics equations : @xmath4 @xmath5 = \rho \vec g + \nabla \cdot \tau\ ] ] @xmath6\ ] ] @xmath7 @xmath8 where @xmath9 is the total energy , @xmath10 is the energy transferred by radiation and @xmath11 is viscous stress tensor .
we assume that small scales are independent of resolved scales ( large eddy simulations ) and rate dissipation is defined from buoyancy and shear production terms [ @xcite ] .
the numerical method that we used was an explicit godunov - type conservative tvd difference scheme [ @xcite ] @xmath12 where @xmath13 and operator @xmath14 is @xmath15 flux along each direction , for example x , was defined by local - characteristic method as follows @xmath16\ ] ] where @xmath17 is matrix whose columns are right eigenvectors of @xmath18 evaluated at generalized roe average for real gases of @xmath19 and @xmath20 .
the @xmath21 is the matrix of numerical dissipation .
term @xmath22 is accounted effect of gravitation forces and radiation .
the one step of time integration is defined by runge - kutta method [ @xcite ] as @xmath23 @xmath24 @xmath25 the scheme is second order by space and time . for approximation
viscous terms we used central differences . for evaluation radiative term in energy equation we used the diffusion approximation @xmath26\ ] ] we use uniform grid in x - y directions and nonuniform grid in vertical z one .
we apply periodic boundary condition in horizontal planes and choose on the top and bottom as follows @xmath27 @xmath28 @xmath29 in initial moment magnetic field equal 50 g and has just one vertical component . for the magnetic diffusitvity
we take constant value @xmath30@xmath31sec@xmath32
on the figures 1 - 4 results of development convection after 12 solar hr mhd numerical simulation are shown .
we founded that magnetic field concentrate on the boundary of convective cells in forms magnetic flux and sheets .
diverging of convective flows from centre supergranular expell weak magnetic field on the edges of convective cell .
average size of supergranular celss is 10 - 20 mm with lifetime about 8 - 10 hours .
we aplly procedure averaging by interval time two hour and find value of velocity from centre supergranular equal to 1 - 1.5 km / sec . in places action of strong magnetic field with strength 700 - 900 g
we observe effect of suppression of convection and decreasing of fluctuation of temperature .
magnetic pressure in regions concentration of magnetic flux prevent inflow of matter .
transfer of radiation energy in these places is suppressed .
we founded from simulation that maximum of value of magnetic field in computational domain equal to 1300 g. inside of supergranular we have usual picture of evolution of convection on scale of granulation with average sizes of cells about 1 - 2 mm and lifetimes about 10 minutes .
here we see wider upflows of warm , low density , and entropy neutral matter and downflows of cold , converging into filamentary structures , dense material .
we observe continuous picture formation and destruction of granules .
granules with highest pressure grow and push matter against neighboring granules , that then shrink and disappear . ascending flow increases pressure in center of granule and upflowing fluids decelerates motion .
this process reduce heat transport to surface and allow material above the granule to cool , become denser , and by action gravity to move down .
we observe formation new cold intergranule lane splitting the original granule . from figure 4
we see distinctly existence three different regions development of convection . in near of solar surface to the depth 4 mm we founded zone of turbulent convection . in this region
cold blobs of matter move down with velocity in maximum about 4 km / sec with maximum mach number equal 1 .
the downdrafts has different and very complicate vertical structure .
some ones travel small distance from surface and become weak enough to be broken up by the surrounding fluid motion .
other ones conserve motion with high velocity and move on distance about 6 mm .
we observe that different nature such behavior is due to initial condition of formation downdrafts . in place of confluence of convective cells to one point more energy is released . in region from 5 mm to 8 mm of depth
we reveal more quiet character of convective flow than in turbulent zone . below 8 mm
we see clear separate large scale density fluctuations and streaming flow of matter similar jets with average velocity about 1km / sec .
the magnetic field in these places has values about 300 g .
distance beetween different narrow jets gives size of supergranular cells .
places of intersections of path jets with horizontal plane are vertices of huge convective cells . from distribution iso - surface of magnetic field with strength 500 g
we have discovered that pumping of magnetic field occupy more part of all computational domain up to bottom boundary .
magnetic field near solar surface in our numerical simulation has very complicate structure .
inside supergranular cells we see formation , growth and evolution with time many arising to surface loops of magnetic field .
there is places in vertices of huge cells with vorticity motion that provide quick rate magnetic helicity transport across solar photosphere(fig.5,6 ) . on the boundary supergranular cell magnetic field
has just vertical component . in these parts
we observe quick changing sign and big values of current helicity(fig.7 ) .
i would like to thank gary zank and nikolai pogorelov from riverside university for financial support for participation me in conference astronum-2006 . | three - dimensional magnetohydrodynamical large eddy simulations of solar surface convection using realistic model physics is conducted .
the effects of magnetic fields on thermal structure of convective motions into radiative layers , the range of convection cell sizes and penetration depths of convection is investigated .
we simulate a some portion of the solar photosphere and the upper layers of the convection zone , a region extending 30 x 30 mm horizontally from 0 mm down to 18 mm below the visible surface .
we solve equations of the fully compressible radiation magnetohydrodynamics with dynamical viscosity and gravity . for numerical simulation
we use : 1)realistic initial model of sun and equation of state and opacities of stellar matter , 2 ) high order conservative tvd scheme for solution magnetohydrodynamics , 3 ) diffusion approximation for solution radiative transfer 4 ) calculation dynamical viscosity from subgrid scale modelling .
simulations are conducted on horizontal uniform grid of 320 x 320 and with 144 nonuniformly spaced vertical grid points on the 128 processors of supercomputer mbc-1500 with distributed memory multiprocessors in russian academy of sciences . |
the standard model ( sm ) of strong and electroweak interactions , based on the gauge group @xmath2 , has been extremely successful phenomenologically .
it has provided the theoretical framework for the description of a very rich phenomenology spanning a wide range of energies , from the atomic scale up to the @xmath3 boson mass , @xmath4 .
it is being tested at the level of a few tenths of a percent , both at very low energies and at high energies @xcite , and has correctly predicted the range of the top quark mass from loop corrections .
however , the sm has a number of shortcomings .
in particular , it does not explain the origin of mass , the observed hierarchical pattern of fermion masses , and why there are three generations of quarks and leptons .
it is widely believed that at high energies , or in very high precision measurements , deviations from the sm will appear , signaling the presence of new physics . in this report
we discuss the prospects for precision tests of the standard model at future collider experiments , focussing on electroweak measurements .
the goal of these measurements is to confront the sm predictions with experiment , and to derive indirect information on the mass of the higgs boson .
the existence of at least one higgs boson is a direct consequence of spontaneous symmetry breaking , the mechanism which is responsible for generating mass of the @xmath0 and @xmath3 bosons , and fermions in the sm . in section
ii we identify some of the relevant parameters for precision electroweak measurements , and review the present experimental situation .
expectations from future collider experiments are discussed in section iii .
we conclude with a summary of our results .
there are three fundamental parameters measured with high precision which play an important role as input variables in electroweak physics .
the fine structure constant , @xmath5 is known with a precision of @xmath6 ppm .
the muon decay constant , @xmath7 gev@xmath8 is measured with @xmath9 ppm from muon decay @xcite .
finally , the @xmath3 boson mass , @xmath10 gev / c@xmath11 @xcite is measured with @xmath12 ppm in experiments at lep and slc .
knowing these three parameters , one can evaluate the @xmath0 mass , @xmath13 , and the weak mixing angle , @xmath14 , at tree level .
when loop corrections are taken into account , @xmath13 and @xmath14 also depend on the top quark mass , @xmath15 , and the higgs boson mass , @xmath16 .
the two parameters depend quadratically on @xmath15 , and logarithmically on @xmath16 .
if the @xmath0 mass and the top quark mass are precisely measured , information on the mass of the higgs boson can be extracted .
constraints on the higgs boson mass can also be obtained from the effective weak mixing angle and @xmath15 .
the ultimate test of the sm may lie in the comparison of these indirect determinations of @xmath16 with its direct observation at future colliders .
the mass of the top quark is presently determined by the cdf and d collaborations from @xmath17 production at the tevatron in the di - lepton , the lepton plus jets , and the all hadronic channels @xcite .
the combined value of the top quark mass from the lepton + jets channel , which yields the most precise result , is @xmath18 the @xmath0 boson mass has been measured precisely by ua2 , cdf , and d .
currently , the most accurate determination of @xmath13 comes from the tevatron cdf and d run ia analyses @xcite and a preliminary d measurement @xcite based on data taken during run ib .
the current world average is @xcite @xmath19 figure [ fig : one ] compares the results of the current @xmath13 and @xmath15 measurements in the @xmath20 plane with those from indirect measurements at lep and slc @xcite , and the sm prediction for different higgs boson masses .
the cross hatched bands show the sm prediction for the indicated higgs boson masses .
the width of the bands is due primarily to the uncertainty on the electromagnetic coupling constant at the @xmath3 mass scale , @xmath21 , which has been taken to be @xmath22 .
recent estimates give @xmath23 @xcite , which corresponds to @xmath24 . the uncertainty on @xmath21
is dominated by the error on the hadronic contribution to the qed vacuum polarization which originates from the experimental error on the cross section for @xmath25 . using dispersion relations @xcite
, the hadronic contribution to @xmath21 can be related to the cross section of the process @xmath25 via @xmath26 where @xmath27 denotes the principal value of the integral , and @xmath28 the relative contributions to @xmath29 and the uncertainty are detailed in fig .
[ fig : alpha ] @xcite .
about 60% of the uncertainty comes from the energy region between 1.05 gev and 5 gev .
more precise measurements of the total hadronic cross section in this energy region , for example at novosibirsk , dap@xmath30ne or bes may reduce the uncertainty on @xmath21 by about a factor 2 in the near future .
the @xmath0 mass can also be determined indirectly from radiative corrections to electroweak observables at lep and sld , and from @xmath31 scattering experiments .
the current indirect value of @xmath13 obtained from @xmath32 experiments , @xmath33 gev / c@xmath11 @xcite , is in excellent agreement with the result obtained from direct measurements ( see fig . [ fig : one ] ) .
the determination of @xmath13 from @xmath31 scattering will be discussed in section iii.c .
the effective weak mixing angle , @xmath34 , has been determined with high precision from measurements of the forward backward asymmetries at lep , and the left - right asymmetries at the slc @xcite . here ,
@xmath34 is defined by @xmath35 where @xmath36 and @xmath37 are the effective vector and axial vector coupling constants of the leptons to the @xmath3 boson , and is related to the weak mixing angle in the @xmath38 scheme , @xmath39 , by @xcite @xmath40 a fit to the combined lep and sld asymmetry data yields @xmath41 the experimental constraints in the @xmath42 plane are compared with the sm predictions in fig . [
fig : four ] .
the measured value of @xmath34 agrees well with the sm expectation .
the star in the lower lefthand corner of fig .
[ fig : four ] indicates the @xmath0 mass and effective weak mixing angle predicted by taking the running of @xmath43 into account only .
the arrow represents the current uncertainty on @xmath13 and the effective weak mixing angle from @xmath29 : @xmath44 the estimated theoretical error from higher orders introduces an additional uncertainty of @xcite @xmath45 while direct measurements of @xmath15 and @xmath13 presently do not impose any constraints on the higgs boson mass , indirect measurements from lep and sld seem to indicate a preference for a relatively light higgs boson .
the 68% confidence level contours in the @xmath15 and @xmath16 plane for the fits to lep data only , and to all data sets @xcite ( lep , sld , cdf and d ) , are shown in fig .
[ fig : two ] . taking the theoretical error due to missing higher order corrections into account ,
one obtains @xmath46 or @xmath47 the results of such a fit from current data , however , should be interpreted with caution .
removing one or two quantities from the fit can drastically change the predicted higgs boson mass range . excluding from the fit the hadronic width of the @xmath3 boson , which depends on @xmath48 , results in @xcite @xmath49 omitting in addition the sld data on @xmath50 which yield a somewhat low value for the effective weak mixing angle , leads to @xmath51 gev / c@xmath11 . in the future ,
only marginal improvements of the indirect measurements from lep data are expected since lep data taking at the @xmath3 peak has ceased .
however , a significant reduction of the errors on @xmath15 and @xmath13 from direct experiments at lep2 , the tevatron ( run i , run ii and tev33 ) , the lhc , and perhaps the nlc and/or a @xmath52 collider is expected , which should result in a more stable prediction for @xmath16 .
this will be discussed in more detail in the next section .
precise measurements of @xmath13 and @xmath15 , if inconsistent with the range allowed by the sm , could indicate the existence of new phenomena at or above the electroweak scale , such as supersymmetry . in the near future direct and indirect measurements of the top quark and @xmath0 boson mass
are expected to begin to yield useful constraints on the parameter space of the minimal supersymmetric standard model ( mssm ) .
this is illustrated in fig .
[ fig : three ] , where the predictions for @xmath13 as a function of @xmath15 in the sm ( shaded bands ) and in the mssm ( area between the dashed lines ) are shown , together with results from direct cdf and d measurements , and indirect measurements from lep and sld .
the mssm band has been obtained by varying the model parameters so that they are consistent with current experimental data .
in addition , it was assumed that no supersymmetric particles are found at lep2 @xcite .
the prospects of measuring the top quark mass in future collider experiments are discussed in detail in ref .
we therefore only briefly summarize the results here . for the tevatron
, the expected accuracy in @xmath15 for run ii ( @xmath53 fb@xmath54 ) and for tev33 ( @xmath55 fb@xmath54 ) can be extrapolated using current and anticipated cdf and d acceptances and efficiencies , together with theoretical predictions . using various different methods and techniques @xcite ,
one expects that @xmath15 can be determined to @xmath56 gev / c@xmath11 ( @xmath57 gev / c@xmath11 ) in run ii ( tev33 ) .
the uncertainty on the top quark mass will be dominated by systematic errors .
soft and hard gluon radiation , and the jet transverse energy scale constitute the most important sources of systematic errors in the top quark mass measurement at hadron colliders . at the lhc , one also expects a precision of about 2 gev / c@xmath11 for @xmath15 @xcite . at an @xmath32 linear collider ( nlc ) or a @xmath52 collider
, the top quark mass can be determined with very high precision from a threshold scan . for an integrated luminosity of 10 fb@xmath54 ( 50 fb@xmath54 ) ,
the expected uncertainty on @xmath15 at the nlc is @xmath58 mev / c@xmath11 ( 200 mev / c@xmath11 ) @xcite . at a @xmath52 collider , the reduced beamstrahlung and initial state radiation result in a better beam energy resolution which should make it possible to measure the top quark mass with a somewhat higher precision than at the nlc , for equal integrated luminosities .
simulations suggest @xmath59 mev / c@xmath11 for 10 fb@xmath54 @xcite .
the precision which can be achieved for @xmath15 at different colliders is summarized in table [ tab : one ] .
.expected top quark mass precision at future colliders . [ cols="<,^",options="header " , ] in our subsequent calculations we shall always assume that the top quark mass can be determined with a precision of @xmath60 presently , the single most precise determination of the effective weak mixing angle originates from the measurement of the left - right asymmetry , @xmath61 at sld . here , @xmath62 is the total production cross section for lefthanded ( righthanded ) electrons . in the sm
, the left - right asymmetry at the @xmath3 pole , ignoring photon exchange contributions , is related to the effective weak mixing angle by @xmath63 if the planned luminosity upgrade @xcite ( `` slc2000 '' ) can be realized , it will be possible to collect @xmath64 @xmath3 decays over a period of three to four years at sld .
this should result in an uncertainty of @xmath65 which is approximately a factor 2 better than the current uncertainty from the fit to the combined lep and sld asymmetry data ( see eq .
( [ eq : asym ] ) ) .
further improvements could come from measurements of the left - right forward - backward asymmetry in @xmath66 , @xmath67-[\sigma_r^f(z ) -\sigma_r^f(-z ) ] \over [ \sigma_l^f(z)+\sigma_l^f(-z)]+[\sigma_r^f(z ) + \sigma_r^f(-z ) ] } \nonumber \\ & = & { 2g_{vf}g_{af}\over g_{vf}^2+g_{af}^2}\ , { 2z\over 1+z^2}\ , , \end{aligned}\ ] ] where @xmath68 , and @xmath69 is the scattering angle .
@xmath70 directly measures the coupling of the final state fermion @xmath71 to the @xmath3 boson from which it is straightforward to determine @xmath34 .
in particular , with the self - calibrating jet - charge technique @xcite , a precise measurement of the @xmath72 coupling should be possible . at hadron colliders ,
the forward backward asymmetry , @xmath73 , in di - lepton production , @xmath74 , ( @xmath75 ) , makes it possible to measure the effective weak mixing angle .
@xmath73 is defined by @xmath76 where @xmath77 b&=&\int_{-1}^0{d\sigma\over d\cos\theta^*}\,d\cos\theta^ * , \label{eq : deffb}\end{aligned}\ ] ] and @xmath78 is the angle between the lepton and the incoming quark in the @xmath79 rest frame . in @xmath80 collisions at tevatron energies ,
the flight direction of the incoming quark to a good approximation coincides with the proton beam direction .
@xmath78 can then be related to the components of the lepton and anti - lepton four - momenta via @xcite @xmath81 with @xmath82 here , @xmath83 is the invariant mass of the lepton pair , @xmath84 is the energy , and @xmath85 is the longitudinal component of the momentum vector . in this definition of @xmath78 ,
the polar axis is defined to be the bisector of the proton beam momentum and the negative of the anti - proton beam momentum when they are boosted into the @xmath79 rest frame .
the four - momenta of the quark and anti - quark can not be determined individually .
the definition of @xmath78 in eq .
( [ eq : cstar ] ) has the advantage of minimizing the effects of the momentum ambiguity induced by the parton transverse momentum .
first measurements of the effective weak mixing angle using the forward backward asymmetry at hadron colliders have been performed by the ua1 and cdf collaborations @xcite .
figure [ fig : six]a shows the variation of @xmath73 with the @xmath32 invariant mass in @xmath86 for @xmath87 tev , assuming @xmath88 .
the error bars indicate the statistical errors for 100,000 events , corresponding to an integrated luminosity of about 2 fb@xmath54 .
the largest asymmetries occur at di - lepton invariant masses of around 70 gev / c@xmath11 and above 110 gev / c@xmath11 .
a preliminary study of the systematic errors , indicates that most sources of error are small compared with the statistical error . the main contribution to the systematic error originates from the uncertainty in the parton distribution functions .
since the vector and axial vector couplings of @xmath89 and @xmath90 quarks to the @xmath3 boson are different , the measured asymmetry depends on the ratio of @xmath89 to @xmath90 quarks in the proton .
most of the systematic errors are expected to scale with @xmath91 , where @xmath92 is the number of events .
the effect of the electromagnetic calorimeter resolution is rather moderate , as shown in fig .
[ fig : six]b .
it is found that most of the sensitivity of this measurement to @xmath93 is at @xmath94 due to the strong variation of @xmath73 with @xmath93 and the high statistics in this region . including qed radiative corrections , the @xmath86 forward backward asymmetry in the @xmath3 boson resonance region ( @xmath95 )
can be parameterized in terms of the effective weak mixing angle by @xcite @xmath96 the expected precision of @xmath93 in the electron channel ( per experiment ) versus the integrated luminosity at the tevatron is shown in fig .
[ fig : seven ] , together with the combined current uncertainty from lep and sld experiments .
a similar precision is expected in the muon channel . combining the results of the electron and the muon channel , an overall uncertainty per experiment of @xmath97
is expected for an integrated luminosity of 30 fb@xmath54 . at the lhc ,
the lowest order @xmath98 cross section is approximately 1.6 nb for each lepton flavor . for the projected yearly integrated luminosity of 100 fb@xmath54 , this results in a very large number of @xmath98 events which
, in principle , could be utilized to measure the forward backward asymmetry and thus @xmath34 with extremely high precision @xcite .
since the original quark direction is unknown in @xmath99 collisions , one has to extract the angle between the lepton and the quark in the @xmath79 rest frame from the boost direction of the di - lepton system with respect to the beam axis : @xmath100 in order to arrive at a non - zero forward - backward asymmetry .
in contrast to tevatron energies , sea quark effects dominate at the lhc . as a result , the probability , @xmath101 , that the quark direction and the boost direction of the di - lepton system coincide is significantly smaller than one .
this considerably reduces the forward backward asymmetry .
events with a large rapidity of the di - lepton system , @xmath102 , originate from collisions where at least one of the partons carries a large fraction @xmath103 of the proton momentum . since valence quarks dominate at high values of @xmath103 , a cut on the di - lepton rapidity increases @xmath101 , and thus the asymmetry @xcite and the sensitivity to the effective weak mixing angle .
imposing a @xmath104 cut and including qed corrections , the forward backward asymmetry at the lhc in the @xmath52 channel in the @xmath3 peak region ( @xmath105 ) can be parameterized by @xmath106 for an ideal detector . for an integrated luminosity of 100 fb@xmath54 ,
this then leads to an expected error of @xmath107 a similar precision should be achievable in the electron channel .
however , electrons and muons can only be detected for pseudorapidities @xmath108 in the currently planned configurations of the atlas @xcite and cms @xcite experiments at the lhc .
the finite pseudorapidity range available dramatically reduces the asymmetry . in the region around the @xmath3 pole ,
the asymmetry is again approximately a linear function of @xmath34 with ( for @xmath52 final states ) latexmath:[\[a_{fb}=0.65\,(0.2488-\sin^2\theta^{lept}_{eff})~~ { \rm for}~~ reduction of the total @xmath3 boson cross section by roughly a factor 5 . as a result ,
the uncertainty expected for @xmath34 increases by almost a factor 7 to @xmath110 in order to improve the precision beyond that expected from future slc and tevatron experiments , it will be necessary to detect electrons and muons in the very forward pseudorapidity range , @xmath111 , at the lhc
. the effective weak mixing angle can also be measured at the nlc in fixed target mller and bhabha scattering . in fixed target mller scattering one
hopes to achieve a precision of @xmath112 @xcite . in bhabha scattering , it should be possible to measure the effective weak mixing angle with a precision of a few @xmath113 @xcite , depending on the energy and polarization available .
possibilities to determine the effective weak mixing angle at a @xmath52 collider have not been investigated so far .
the potential of extracting useful information on the higgs boson mass from a fit to the sm radiative corrections and a precise measurement of @xmath34 and @xmath15 is illustrated in fig .
[ fig : eight ] .
here we have assumed @xmath114 gev / c@xmath11 , @xmath115 , and @xmath116 .
from such a measurement , one would find @xmath117 gev / c@xmath11 . the corresponding log - likelihood function
is shown in fig .
[ fig : nine ] . from fig .
[ fig : eight ] it is obvious that the extracted higgs boson mass depends very sensitively on the central value of the effective weak mixing angle .
the relative error on the higgs boson mass , @xmath118 , however , depends only on the uncertainty of higher order corrections , @xmath93 , @xmath15 , and @xmath21 .
for the precision of @xmath93 and @xmath15 assumed here , the theoretical error from higher orders , and the uncertainty in @xmath21 begin to limit the accuracy which can be achieved for the higgs boson mass .
future experiments provide a variety of opportunities to measure the mass of the @xmath0 boson with high precision . in @xmath31 scattering
, @xmath13 can be determined indirectly through a measurement of the neutral to charged current cross section ratio @xmath119 in the sm , @xmath120 can be used to directly determine the weak mixing angle via the lowest order expression @xmath121 where @xmath122 and @xmath123 is a correction factor which incorporates , among others , effects due to charm production and longitudinal structure functions .
electroweak radiative corrections modify the leading order prediction . in the on - shell scheme , where @xmath124 to all orders in perturbation theory , the ( leading ) radiative corrections to @xmath14 and @xmath120 almost perfectly cancel @xcite .
this implies that , in the sm , @xmath31 scattering directly measures the @xmath0 mass , given the very precisely determined @xmath3 boson mass .
a new ccfr measurement @xcite gives @xmath125 gev / c@xmath11 . with the data which one hopes to collect in the nutev experiment during the current fermilab fixed target run , one expects @xcite @xmath126 figure [ fig : ten ] compares the current results for @xmath13 from direct measurements at cdf , d and lep2 ( see below ) with indirect determinations from lep and sld via electroweak radiative corrections , and the @xmath0 mass obtained from ccfr , other @xmath31 experiments @xcite , and the expectation for nutev .
the @xmath0 mass can also be determined from measurements of the charged and neutral current cross sections at hera . moving the low @xmath127 quadrupoles closer to the interaction region ,
one hopes to achieve integrated luminosities of the order of 150 pb@xmath54 per year with a 70% longitudinally polarized electron beam .
the expected constraints on @xmath13 and @xmath15 , together with the sm predictions for @xmath128 gev / c@xmath11 and @xmath129 gev / c@xmath11 are shown in fig .
[ fig : eleven ] @xcite .
when combined with a measurement of the top quark mass with a precision of @xmath130 gev / c@xmath11 , the projected hera results yield a precision of @xmath131 taking @xmath132 gev / c@xmath11 instead only marginally improves the accuracy on the @xmath0 mass . in deriving the result shown in eq .
( [ eq : hera ] ) , a 1% relative systematic uncertainty of the charged and neutral current cross sections at hera was assumed . for a systematic error of 2% ,
one finds @xmath133 .
precise measurements of the @xmath0 mass at lep2 @xcite can be obtained using the enhanced statistical power of the rapidly varying total @xmath134 cross section at threshold @xcite , and the sharp ( breit - wigner ) peaking behaviour of the invariant mass distribution of the @xmath0 decay products . during the recent lep2 run at @xmath135 gev ,
the four lep experiments have each accumulated approximately 10 pb@xmath54 of data .
the total @xmath134 cross section as a function of the @xmath0 mass is shown in fig .
[ fig : twelve ] , together with the preliminary experimental result @xcite . combining the results obtained from the @xmath136 , the @xmath137 and the @xmath138 ( @xmath139 ) channel , the @xmath0 pair production cross section at @xmath135 gev is measured to be @xmath140 pb .
this translates into a @xmath0 mass of @xcite @xmath141 a much more accurate measurement of @xmath13 will be possible in the future through direct reconstruction methods when lep2 will be running at energies well above the @xmath0 pair threshold . here
, the breit - wigner resonance shape is directly reconstructed from the @xmath142 final states using kinematic fitting techniques .
the potentially most important limitation in using this method originates from color reconnection @xcite and bose - einstein correlations @xcite in the @xmath136 channel . taking common errors into account , the expected overall precision from this method at lep2 for a total integrated luminosity of 500 pb@xmath54 per experiment is anticipated to be @xcite @xmath143 the same method can in principle also be used at the nlc .
however , the beam energy spread limits the precision which one can hope to achieve at an @xmath32 linear collider .
preliminary studies indicate that one can hope for a precision of @xmath144 mev / c@xmath11 at best .
no studies for a @xmath52 collider have been performed so far . in @xmath0 events produced in a hadron collider in essence
only two quantities are measured : the lepton momentum and the transverse momentum of the recoil system .
the latter consists of the `` hard '' @xmath0-recoil and the underlying event contribution . for @xmath0-events
these two are inseparable .
the transverse momentum of the neutrino is then inferred from these two observables . since the longitudinal momentum of the neutrino
can not be determined unambiguously , the @xmath0-boson mass is usually extracted from the distribution in transverse : @xmath145 where @xmath146 is the angle between the electron and neutrino in the transverse plane .
the @xmath147 distribution sharply peaks at @xmath148 . both the transverse mass and lepton transverse momentum are , by definition , invariant under longitudinal lorentz boosts .
in determining the @xmath0 mass , the transverse mass is preferred over the lepton transverse momentum spectra because it is to first order independent of the transverse momentum of the @xmath0 . under transverse lorentz boosts along a direction @xmath149 ,
@xmath147 and @xmath150 transform as @xmath151 with @xmath152 , @xmath153 and @xmath154 .
the asterisk indicates quantities in the @xmath0 rest frame .
the disadvantage of using the transverse mass is that it uses the neutrino transverse momentum which is a derived quantity .
the neutrino transverse momentum is identified with the missing transverse energy in the event , which is given by @xmath155 where @xmath156 is the transverse momentum of the @xmath0-recoil system and @xmath157 the transverse energy flow of the underlying event , which depends on the luminosity .
it then follows that the magnitude of the missing @xmath158 vector and the true neutrino momentum are related as @xmath159 this relation can be interpreted as the definition of the neutrino momentum scale .
note that the underlying event gives rise to a bias in the _ measured _ neutrino momentum with respect to the _ true _ neutrino momentum . in case
there are more interactions per crossing , @xmath160 behaves as a two - dimensional random walk and is proportional to @xmath161 , where @xmath162 is the average number of interactions per crossing .
the shift in measured neutrino momentum is thus directly proportional to the number of interactions per crossing .
the resolution increases as @xmath161 .
the above equation for the missing transverse energy deserves some more attention .
the two components directly related to the @xmath0 decay , @xmath163 and @xmath156 , are only indirectly affected by multiple interactions through the underlying event .
it is the measurement of @xmath157 which governs the luminosity dependence . because of multiple interactions
, @xmath157 will show a dependence on luminosity following poisson statistics , with the two effects indicated above : @xmath164 a degradation of the resolution and @xmath165 a shift in the measured neutrino momentum .
this is demonstrated in fig .
[ fig : thirteen ] where we show the @xmath147 distribution for various values of @xmath162 at the tevatron . for run ii
one expects @xmath166 , and at tev33 , @xmath167 @xcite . both effects , of course ,
propagate into the measurement of the transverse mass and the uncertainty on @xmath13 will not follow the simple @xmath91 rule anymore @xcite .
in addition , however , the detector response to high luminosities needs to be folded in . in the above discussion
it was assumed that the detector response is linear to the number of multiple interactions which in general is not the case .
the effects of pile - up in the calorimeter and occupancy in the tracking detectors produce a @xmath168 shift in @xmath169 for an electron with transverse momentum of 40 gev / c at @xmath170 , which will further affect the uncertainty on the @xmath0 mass adversely @xcite .
another uncertainty that will not , and has not in the past , scaled with luminosity is the theoretical uncertainty coming from the @xmath171 model and the uncertainty on the proton structure .
parton distributions and the spectrum in @xmath171 are correlated .
the 0 experiment has addressed this correlation in the determination of its uncertainty on the @xmath0 mass @xcite .
the parton distribution functions are constrained by varying the cdf measured @xmath0 charge asymmetry within the measurement errors , while at the same time utilizing all the available data .
new parametrizations of the cteq 3 m parton distribution function were obtained that included in the fit the cdf @xmath0 asymmetry data from run ia @xcite , where all data points had been moved coherently up or down by one standard deviation . in addition one of the parameters , which describes the @xmath172-dependence of the parameterization of the non - perturbative functions describing the @xmath171 spectrum @xcite ,
was varied .
the constraint on this parameter was provided by the measurement of the @xmath173 spectrum .
the uncertainty due to parton distribution functions and the @xmath171 input spectrum was then assessed by varying simultaneously the parton distribution function , as determined by varying the measured @xmath0 charge asymmetry , and the parameter describing the non - perturbative part of the @xmath171 spectrum .
the cdf experiment uses their measurement of the @xmath0 charge asymmetry as the sole constraint on the uncertainty due to the parton distribution functions . figure [ fig : fourteen ]
compares the preliminary cdf @xmath0 charge asymmetry measurement @xcite with several recent fits to parton distribution functions .
figure [ fig : fifteen ] shows the correlation between the uncertainty on the @xmath0 mass , @xmath174 , and @xmath175 the deviation between the average measured asymmetry for run ia and ib data and various recent nlo fits @xcite .
the fitted @xmath0 mass is seen to be strongly correlated with the @xmath0 charge asymmetry .
the @xmath0 charge asymmetry , however , is mainly sensitive to the slope of the ratio of the @xmath89 and @xmath90 quark @xmath176 and does not probe the full parameter range describing the
. future measurements of the @xmath173 distribution will provide a constraint on the @xmath169 distribution of the @xmath0 boson .
moreover , the measurements of the @xmath0 charge asymmetry , together with measurements from deep inelastic scattering experiments , will provide further constraints on the parton distribution functions .
an effort needs to be made , though , to provide the experiments with parton distributions with associated uncertainties . at high luminosities alternate methods to determine the @xmath0-mass may be advantageous . because of the similarity of @xmath0 and @xmath3 production , methods based on ratios of relevant quantities , such as the charged lepton transverse momenta are particularly interesting @xcite .
the ratio of the lepton @xmath169 distributions is thought to be very promising for fitting the @xmath0 mass in the high luminosity regime since the procedure is independent of many resolution effects .
however , the shapes of the lepton transverse momentum distributions are sensitive to the differences in the @xmath0 and @xmath3 production mechanisms , which need to be better understood . here
we concentrate on a similar method which utilizes the transverse mass ratio of @xmath0 and @xmath3 bosons @xcite .
preliminary results from an analysis of the transverse mass ratio have recently been presented by the d collaboration @xcite .
only the electron channel will be discussed in the following , although the method is expected to work for muon final states as well .
the transverse mass ratio method treats the @xmath177 sample similar to the @xmath178 sample , thus cancelling many of the common systematic uncertainties .
a transverse mass for the @xmath3 boson is constructed with one of the decay electrons , while the @xmath179 is derived by adding the transverse energy of the other electron to the residual @xmath179 in the event .
hence , two such combinations can be formed for each @xmath3 event .
the @xmath3 transverse mass distribution is scaled down in finite steps and compared with the @xmath147 distribution of the @xmath0 boson .
the @xmath0 mass is then determined from the scale factor @xmath180 which gives the best agreement between the @xmath147 distributions using a kolmogorov test . since differences in the production mechanism , acceptances and resolution effects between the @xmath0 and the @xmath3 sample lead to differences in the shapes of the transverse mass distributions
, one has to correct for these effects .
the dominant systematic uncertainty arises from the uncertainty on the underlying event .
electromagnetic and hadronic resolution effects mostly cancel in the transverse mass ratio , as expected . the systematic uncertainty due to the and the transverse momentum of the @xmath0 boson is reduced by more than a factor 3 compared with that found using the conventional @xmath0 transverse mass method @xcite .
the total systematic error from the drun ia data sample is estimated to be 75 mev / c@xmath11 .
for comparison , the total systematic error obtained using the transverse mass distribution of the @xmath0 using d run ia data is 165 mev / c@xmath11 @xcite . in the analysis of the run ia data sample , electrons from @xmath0 and @xmath3 decay
are identified as in the conventional @xmath0 mass analysis .
@xmath0 candidates are selected by requiring @xmath181 gev / c and @xmath182 gev / c , while electrons from @xmath3 decays are required to have @xmath183 gev / c , since they are eventually scaled down .
electrons from @xmath0 decay and at least one electron from @xmath3 decay are required to be in the central pseudorapidity region , @xmath184 .
@xmath3 events are used twice if both electrons fall in the central region .
the shape comparison is performed in the fitting window @xmath185 .
the selected @xmath3 sample is scaled down in finite steps and , at every step , the shape of the @xmath3 and @xmath0 @xmath147 distribution is compared using the kolmogorov test .
figure [ fig : sixteen ] shows the @xmath186 distribution superimposed on the @xmath187 distribution for one of the fits .
the preliminary result for @xmath13 from run ia data is @xmath188 the limitation of the method described here comes entirely from the limited @xmath3 statistics , which is expected to scale exactly as @xmath91 in future experiments .
the power of the @xmath147 ratio method becomes apparent when one compares the uncertainty on @xmath13 expected for 1 fb@xmath54 and 10 fb@xmath54 with that expected from the traditional @xmath0 transverse mass analysis @xcite .
the results for both methods are listed in table [ tab : two ] . to calculate the projected statistical ( systematic ) errors in the transverse mass ratio method
, we have taken the errors of eq .
( [ mtresult ] ) and scaled them with @xmath91 ( @xmath189 ) , assuming @xmath190 ( @xmath191 ) for 1 fb@xmath54 ( 10 fb@xmath54 ) .
both , electron and muon channels are combined in table [ tab : two ] , assuming that the two channels yield the same precision in @xmath13 .
lcc + & @xmath192 & @xmath193 + @xmath194 & @xmath190 & @xmath191 + statistical & @xmath195 & @xmath196 + systematic & @xmath197 & @xmath198 + total & @xmath199 & @xmath195 + + & @xmath192 & @xmath193 + @xmath194 & @xmath190 & @xmath191 + statistical & @xmath195 & @xmath200 + systematic & @xmath201 & @xmath202 + total & @xmath203 & @xmath204 + the @xmath0 mass can also be determined from the transverse energy ( momentum ) distribution of the electron ( muon ) in @xmath205 ( @xmath206 ) events , which peaks at @xmath207 .
the prospects of a precise measurement of @xmath13 from the @xmath208 distribution in run ii and at tev33 have been investigated in ref .
the measurement of the lepton four - momentum vector is independent of the @xmath179 resolution , and the electron @xmath158 resolution is dominated by the intrinsic calorimeter resolution .
hence the statistical uncertainty of the @xmath0 mass measurement from the @xmath209 distribution is expected to scale approximately as @xmath91 .
simulations have shown that a sample of 30,000 events ( similar to the d run ib data sample ) gives a statistical error on the @xmath0 mass of 100 mev / c@xmath11 from the @xmath208 fit .
this is in agreement with the result of the preliminary drun ib @xmath0 mass analysis @xcite .
the systematic error from this method is expected to be about 170 mev / c@xmath11 for the same number of events . scaling the total uncertainty as @xmath91 ,
the projected uncertainty of @xmath13 from the electron @xmath158 fit is : @xmath210 in estimating the uncertainties given in eq .
( [ eq : et ] ) and table [ tab : two ] , we have assumed that the current uncertainty from and the theoretical uncertainty originating from higher order electroweak corrections can be drastically reduced in the future .
in order to measure @xmath13 with high precision , it is crucial to fully control higher order electroweak ( ew ) corrections .
so far , only the final state @xmath211 photonic corrections have been calculated @xcite , using an approximation which indirectly estimates the soft + virtual part from the inclusive @xmath212 @xmath213 width and the hard photon bremsstrahlung contribution . using this approximation ,
electroweak corrections were found to shift the @xmath0 mass by about @xmath214 mev / c@xmath11 in the electron , and @xmath215 mev / c@xmath11 in the muon channel @xcite .
currently , a more complete calculation of the @xmath211 ew corrections , which takes into account initial and final state corrections , is being carried out @xcite .
the calculation is performed using standard monte carlo phase space slicing techniques for nlo calculations . in calculating the initial state radiative corrections , mass ( collinear ) singularities
are absorbed into the parton distribution functions through factorization , in complete analogy to the qcd case .
qed corrections to the evolution of the are not taken into account .
a study of the effect of qed on the evolution indicates that the change in the scale dependence of the pdf is small @xcite . to treat the qed radiative corrections in a consistent way
, they should be incorporated in the global fitting of the pdf . the relative size and the characteristics of the various contributions to the ew corrections to @xmath0 production
is shown in fig .
[ fig : ratio ] . initial state ( photon and weak ) radiative corrections are found to be uniform and , therefore , are expected to have little effect on the @xmath0 boson mass extracted . while initial state photon radiation increases the cross section by @xmath216 , weak one - loop corrections almost completely cancel the initial state photonic corrections .
the complete @xmath211 initial state ew corrections reduce the leading order ( lo ) cross section by about 0.1% .
initial and final state photon radiation interfere very little .
the interference effects are uniform and have essentially no effect on the @xmath147 distribution .
final state photon radiation changes the shape of the transverse mass distribution and reduces the lo cross section by up to @xmath217 in the @xmath0 resonance region .
weak corrections again have no influence on the lineshape , but reduce the cross section by about 1% . the @xmath0 mass obtained from the @xmath147 distribution including the full ew one - loop corrections
is expected to be several mev / c@xmath11 smaller than that extracted employing the approximate calculation of ref .
@xcite .
since final state photon radiation introduces a significant shift in the @xmath0 mass , one also has to worry about multiple photon radiation .
a calculation of @xmath218 @xcite which includes all initial and final state radiation and finite muon mass effects shows that approximately 0.8% of all @xmath219 events contain two photons with @xmath220 gev ( the approximate tower threshold of the electromagnetic calorimeters of cdf and d ) and @xmath221 .
this suggests that the additional shift in @xmath13 from multiple photon radiation may not be negligible if one aims at a measurement with a precision of @xmath222 . at the lhc ,
the cross section for @xmath0 production is about a factor 4 larger than at the tevatron . during the first year of operation
, it is likely that the lhc will run at a reduced luminosity of approximately @xmath223 , resulting in roughly @xmath224 @xmath178 events with a central electron ( @xmath225 ) and a transverse mass in the range @xmath185 .
a similar number of @xmath219 events is expected .
both lhc detectors , atlas @xcite and cms @xcite , will be able to trigger on electrons and muons with a transverse momentum of @xmath226 gev / c ( @xmath75 ) , and should be fully efficient for @xmath227 gev / c .
they are well - optimized for electron , muon and @xmath179 detection . at @xmath223 ,
the average number of interactions per crossing at the lhc is approximately @xmath228 , which is significantly smaller than what one expects at the tevatron for the same luminosity .
a precision measurement of the @xmath0 mass at the lhc running at a reduced luminosity , using the traditional transverse mass analysis , thus seems feasible @xcite .
qcd corrections to the transverse mass distribution at the lhc enhance the cross section by 10 20% in the @xmath147 range which is normally used to determine @xmath13 .
this is illustrated in fig .
[ fig : nineteen ] , where the lo and nlo qcd transverse mass distribution is shown , together with the nlo to lo differential cross section ratio . here ,
a @xmath227 gev / c and a @xmath229 gev / c cut have been imposed , and the pseudorapidity of the lepton is required to be @xmath230 . the slight change in the shape of the @xmath147 distribution induced by the nlo qcd corrections is due to the cuts imposed .
so far , no detailed study of the precision which one might hope to achieve for @xmath13 at the lhc has been performed . for a _ crude order of magnitude _
estimate , one can use the statistical and systematic errors of the current cdf and d analyses @xcite , and scale them by @xmath189 . for an integrated luminosity of 10 fb@xmath54 , one obtains @xcite : @xmath231 in order to see whether lhc experiments can perform a measurement of @xmath13 which is significantly more precise than what one expects from tev33 or the nlc , a more detailed study which also considers other quantities such as the transverse mass ratio of @xmath0 and @xmath3 bosons @xcite has to be carried out .
the potential of extracting useful information on the higgs boson mass from a fit to the sm radiative corrections and a precise measurement of @xmath13 and @xmath15 is illustrated in fig .
[ fig : twenty ] . here
we have assumed @xmath114 gev / c@xmath11 , @xmath232 gev / c@xmath11 , and @xmath116 .
such a measurement would constrain the higgs boson mass to @xmath233 gev / c@xmath11 .
the corresponding log - likelihood function is shown in fig .
[ fig : twentyone ] .
a measurement of the @xmath0 mass with a precision of @xmath234 mev / c@xmath11 and of the top mass with an accuracy of 2 gev / c@xmath11 thus translates into an indirect determination of the higgs boson mass with a relative error of about @xmath235 from a global analysis of all electroweak precision data one might then expect @xmath236 . for the precision of @xmath15 and @xmath13 assumed here , the theoretical error from higher orders and the uncertainty in the electromagnetic coupling constant @xmath21 become limiting factors for the accuracy which can be achieved for @xmath16 .
efforts to calculate higher order corrections and to significantly improve the error on @xmath21 beyond what one can expect from measurements at novosibirsk , dap@xmath30ne , or bes , need increased emphasis from both experimentalists and theorists in order to be able to achieve an ultimate relative precision on @xmath16 better than about 15% .
in this report , we have highlighted some current high precision electroweak measurements , and explored prospects for further improvements over the next decade .
the aim of precision electroweak measurements is to test the sm at the quantum level , and to extract indirect information on the mass of the higgs boson .
the confrontation of these indirect predictions of @xmath16 with the results of direct searches for the higgs boson will be perhaps the most exciting development of the next decade in the field of particle physics . although a global fit to all available precision electroweak data yields @xmath237 gev / c@xmath11 , the higgs boson mass extracted strongly depends on the input quantities used in the fit . excluding a particular observable which displays a statistically significant deviation from the sm prediction , _
e.g. _ the sld left - right asymmetry , may easily increase the central value of @xmath16 by a factor 4 .
one therefore has to conclude that present data are not quite sufficient to obtain a stable estimate of the higgs boson mass .
results of future collider experiments are expected to drastically change this situation . in these experiments
one hopes to precisely determine three observables which are key ingredients in obtaining reliable indirect information on the higgs boson mass : * the uncertainty on the top quark mass is expected to be reduced by at least a factor 3 in tevatron and lhc experiments . at the nlc or a @xmath52 collider ,
a precision of a few hundred mev / c@xmath11 may be possible .
* it should be possible to reduce the error on @xmath93 by at least a factor two through measurements of the left - right asymmetry at a luminosity upgraded slc , and the forward backward asymmetry in the @xmath3 peak region at the tevatron and lhc . *
the most profound improvement is likely to occur for the @xmath0 mass , where a gain of a factor 5 seems to be within reach .
new strategies developed for extracting @xmath13 at hadron colliders @xcite will make it possible to fully exploit the expected increase in integrated luminosity at the tevatron .
as the electroweak measurements improve , the theoretical error from higher orders and the uncertainty in @xmath21 will gradually become more and more important limitations in the precision which can be achieved .
the determination of @xmath21 is limited by the knowledge of the photon hadron coupling at small momentum transfer .
an increased experimental and theoretical effort is needed to overcome the present limitations in determining @xmath21 , and to calculate higher order corrections to the electroweak observables .
_ m. demarteau _ , fermilab - conf-96/354 , to appear in the proceedings of the _ dpf96 conference _ , minneapolis ,
mn , august 10 15 , 1996 . _ r.m .
barnett _ et al . _
( particle data group ) _ , phys
* d54 * , 1 ( 1996 ) .
_ d. gerdes _ , these proceedings .
_ f. abe _ _ et al . _
( cdf collaboration ) , phys .
lett . * 75 * , 11 ( 1995 ) and phys . rev .
* d52 * , 4784 ( 1995 ) ; _ s. abachi _ _ et al .
_ ( d collaboration ) phys .
77 * , 3309 ( 1996 ) .
_ s. abachi _ _ et al .
_ ( d collaboration ) fermilab - conf/96 - 251-e , submitted to the _ `` 28@xmath238 international conference on high energy physics '' _ , warsaw , poland , 25 31 july 1996 .
_ h. burkhardt and b. pietrzyk _ , phys .
b356 * , 398 ( 1995 ) ; _ s. eidelmann and f. jegerlehner _ , z. phys .
* c67 * , 585 ( 1995 ) ; _ r.b .
nevzorov , a.v .
novikov and m.i .
vysotsky _ , jetp lett . * 60 * , 399 ( 1994 ) ; _ a.d .
martin and d. zeppenfeld _ , phys .
b345 * , 558 ( 1995 ) ; _ m. swartz _ , phys . rev . *
d53 * , 5268 ( 1996 ) .
_ n. cabbibo and r. gatto _ , phys . rev . *
124 * , 1577 ( 1961 ) .
_ p. gambino and a. sirlin _ , phys .
* d49 * , 1160 ( 1994 ) .
_ g. altarelli _ , cern - th/96 - 265 ( november 1996 ) , lectures given at the nato advanced study institute on techniques and concepts of high energy physics , st .
croix , u.s .
virgin islands , 10 23 july , 1996 .
[ rosner1 ] _ j. rosner _ , cern - th/96 - 245 , hep - ph/9610222 , lectures given at the cargse summer institute on particle physics , 1996 .
_ p. chankowski _ _ et al .
b417 * , 101 ( 1994 ) ; _
a. dabelstein _ _ et al .
_ , in proceedings of the _ `` ringberg workshop on perspectives for electroweak interactions in @xmath32 collisions '' _ , ringberg , germany , february 5 8 , 1995 ; _
d. pierce _ _ et al .
_ , slac - pub-7180 , preprint , june 1996 .
d. gerdes _ _ et al .
_ , these proceedings .
d. amidei _ _ et al .
`` future electroweak physics at the fermilab tevatron : report of the tev_2000 study group '' _ , eds .
d amidei and r. brock , fermilab - pub/96 - 082 ( 1996 ) , p. 13 .
_ s. kuhlmann _ _ et al .
_ , _ `` physics and technology of the next linear collider '' _ , fermilab pub-96/112 ( 1996 ) .
_ r. palmer _ _ et al .
`` @xmath52 collider : a feasibility study '' _ , fermilab - conf-96/092 ( 1996 ) .
_ m. breidenbach _ _ et al .
_ , slac - cn-409 ( 1996 ) .
_ k. abe _ _ et al . _
( sld collaboration ) , phys .
74 * , 2890 ( 1995 ) .
_ j. collins and d. soper _ , phys . rev . *
d16 * , 2219 ( 1977 ) . _ c. albajar _ _ et al . _
( ua1 collaboration ) , z. phys .
* c44 * , 15 ( 1989 ) ; _ f. abe _ _ et al . _
( cdf collaboration ) , phys .
lett . * 67 * , 1502 ( 1991 ) . _
f. abe _ _ et al . _
( cdf collaboration ) , phys .
lett . * 77 * , 2616 ( 1996 ) .
_ u. baur , s. keller and w. sakumoto _ , in preparation .
_ p. fisher , u. becker and p. kirkby _ , phys .
* b356 * , 404 ( 1995 ) .
_ m. dittmar _ , ethz - ipp pr-96 - 01 ( preprint , march 1996 ) , to appear in phys . rev . *
d. gingrich _ _ et al .
_ ( atlas collaboration ) , atlas letter of intent , cern - lhcc-92 - 4 ( october 1992 ) ; _ w. w. armstrong _ _ et al . _
( atlas collaboration ) , atlas technical design report , cern - lhcc-94 - 43 ( december 1994 ) .
_ m. della negra _
_ et al . _
( cms collaboration ) , cms letter of intent , cern - lhcc-92 - 3 ( october 1992 ) ; _
g. l. bayatian _ _ et al .
_ ( cms collaboration ) , cms technical design report , cern - lhcc-94 - 38 ( december 1994 ) .
_ k. kumar _ , these proceedings . _ f. cuypers and p. gambino _ , phys
* b388 * , 211 ( 1996 ) . _
marciano and a. sirlin _ , phys . rev .
* d22 * , 2695 ( 1980 ) ; _ r.g .
stuart _ , z. phys .
* c34 * , 445 ( 1987 ) .
_ et al . _ ( ccfr / nutev collaboration ) , fermilab - conf-96/227-e , to appear in the proceedings of the _ `` xxxist recontres de moriond , electroweak interactions and unified theories '' _ , les arcs , france , march 16 23 , 1996 .
_ j. allaby _ _ et al . _
( charm collaboration ) , z. phys .
* c36 * , 611 ( 1985 ) ; _
a. blondel _ _ et al .
_ ( cdhs collaboration ) , z. phys . *
c45 * , 361 ( 1990 ) ; _ c. arroyo _ _ et al . _
( ccfr collaboration ) , phys .
lett . * 72 * , 3452 ( 1994 ) .
[ hera1 ] _ r.j .
cashmore _ _ et al .
_ , mpi / pth/96 - 105 , to appear in the _ `` proceedings of the workshop on future physics at hera''_. _
a. ballestrero _ _ et al .
_ , in proceedings of the workshop on physics at lep2 , g. altarelli , t .
sjostrand and f. zwirner ( eds . ) , cern yellow report cern-96 - 01 ( 1996 ) , vol . 1 ,
stirling _ , nucl .
b456 * , 3 ( 1995 ) .
_ k. ackerstaff _ _ et al . _
( opal collaboration ) , cern - ppe/96 - 141 ( preprint , october 1996 ) , submitted to phys . lett . *
n. watson _
( opal collaboration ) , talk given at cern , october 1996 ; _
m. pohl _
( l3 collaboration ) , talk given at cern , october 1996 .
_ t. sjstrand and v. khoze _ , z. phys .
* c62 * , 281 ( 1994 ) ; phys . rev . lett .
* 72 * , 28 ( 1994 ) .
_ l. lnnblad and t. sjstrand _ , phys
b351 * , 293 ( 1995 ) .
_ g. jackson _ , these proceedings ; _
m. johnson _ , these proceedings .
_ u. baur , m. demarteau , and s. errede _ , in _
`` future electroweak physics at the fermilab tevatron : report of the tev_2000 study group '' _ , eds . d amidei and r. brock , fermilab - pub/96 - 082 ( 1996 ) , p. 63 .
kotwal _ , these proceedings .
_ f. abe _ _ et al . _
( cdf collaboration ) , phys .
* 74 * , 850 ( 1995 ) .
_ g. ladinsky and c .- p .
yuan _ , phys . rev .
* d50 * , 4239 ( 1994 ) .
_ a. bodek _
( cdf collaboration ) , fermilab - conf/96 - 341-e preprint , to appear in the proceedings of the _
`` 28@xmath238 international conference on high energy physics '' _ , warsaw , poland , 25 31 july 1996 . _ w. giele and s. keller _ , fermilab - conf-96/307-t , to appear in the proceedings of the _ dpf96 conference _ , minneapolis ,
mn , august 10 15 , 1996 .
_ s. rajagopalan and m. rijssenbeek _ , these proceedings .
_ m. l. kelly _ _ et al .
_ ( d collaboration ) , fermilab - conf/96 - 236-e , to appear in the proceedings of the _ dpf96 conference _ ,
minneapolis , mn , august 10 15 , 1996 .
_ m. rijssenbeek _ , fermilab - conf/96 - 365-e , to appear in the proceedings of the _
`` 28@xmath238 international conference on high energy physics '' _ , warsaw , poland , 25 31 july 1996 .
_ f. berends and r.k .
kleiss _ , z. phys . * c27 * , 365 ( 1985 ) .
_ u. baur , s. keller and d. wackeroth _ , these proceedings and paper in preparation .
_ h. spiesberger _ , phys
. rev . * d52 * , 4936 ( 1995 ) ; _ j. kripfganz and h. perlt _ , z. phys .
* c41 * , 319 ( 1988 ) .
_ u. baur , t. han , r. sobey and d. zeppenfeld _ , these proceedings .
_ s. keller and j. womersley _ , these proceedings . | we present an overview of the present status and prospects for progress in electroweak measurements at future collider experiments leading to precision tests of the standard model of electroweak interactions .
special attention is paid to the measurement of the @xmath0 mass , the effective weak mixing angle , and the determination of the top quark mass .
their constraints on the higgs boson mass are discussed .
0d # 1#1@xmath1 0.5 cm 0.7 cm fermilab - conf-96/423 + ub - het-96 - 06 * precision electroweak physics at future collider experiments * u. baur + + + * working group members * + c. balazs _ ( msu ) _ , d. errede _ ( urbana ) _ , s. errede _ ( urbana ) _ , t. han _
( davis ) _ , + s. keller _ ( fnal ) _ , y - k .
kim _ ( berkeley ) _ , a.v .
kotwal _
( columbia ) _ , f. merritt _ ( chicago ) _ , + s. rajagopalan _ ( stony brook ) _
, r. sobey _
( davis ) _ , m. swartz _ ( slac ) _ , + d. wackeroth _ ( fnal ) _ , j. womersley _ ( fnal ) _
* abstract * + |
when a polymer is placed in solution hydrophobic interactions between monomers and solvent molecules can cause it to undergo a collapse transition to a compact state@xcite .
the statistical physics of the polymer transition from the extended to the collapsed state is well understood .
however the dynamics of the transition and in particular the effect on that dynamics of the hydrodynamic properties of the solvent remains unclear .
therefore in this letter we investigate numerically the dynamical behaviour of a polymer chain collapsing in a dilute solution .
the collapse is measured with and without the presence of hydrodynamic interactions thus allowing a direct investigation of their effect .
we find that hydrodynamics accelerates the polymer collapse .
it also alters the folding pathway , allowing the folding to occur more homogeneously along the polymer chain rather than initially at the chain ends . with the introduction of zimm s model@xcite
it became apparent that hydrodynamics play a central role in the dynamics of polymers in dilute solution .
however , understanding such interactions is difficult , analytically because they present a complicated many - body problem , and numerically because they develop on time - scales long compared to the thermal fluctuations of the monomers
. theoretical work on the dynamics of polymer collapse can be divided into two approaches .
phenomenological models balance the driving and dissipative forces to give scaling laws@xcite .
they involve assumptions about how the collapsed state develops on which there is no consensus .
several authors have considered models which are based on a solution of the langevin equation@xcite .
of particular interest is work by pitard@xcite and by kuznetsov et al@xcite who find the inclusion of hydrodynamics , modelled by a preaveraged oseen tensor , speeds up the collapse .
simulations on polymer collapse@xcite , using monte carlo or langevin approaches , have not included the hydrodynamic effects of the solvent .
very recent work has shown that it is now possible to use molecular dynamics simulations with an explicit solvent to model the collapse transition if powerful computational resources are available@xcite . in an interesting recent paper chang and yethiraj@xcite
compared molecular dynamics simulations of a polymer in a solvent to brownian dynamics simulations .
they attempted to match the parameters in the two simulations and hence to compare collapse with and without hydrodynamics .
here we use a hybrid approach@xcite where the solvent is modelled by a malevanets - kapral method@xcite and the polymer by molecular dynamics in our investigation of the hydrodynamics of polymer collapse .
modelling a dilute polymer solution is a difficult task because of the existence of widely differing time scales .
the dynamical properties of polymers can be dominated by hydrodynamic interactions between different parts of the polymer chains@xcite .
in contrast with the time scale of thermal fluctuations of individual monomers , these interactions are long - ranged and evolve slowly .
therefore it is computationally too expensive to reach hydrodynamic time scales using molecular dynamics simulations for both the polymer and the solvent molecules . to overcome this problem we use a hybrid simulation approach where the equations of motion of the polymer alone are solved using a molecular dynamics algorithm .
the solvent is modelled using a mesoscale approach , developed by malevanets and kapral@xcite .
this ignores the molecular detail of the solvent but preserves its ability to transmit hydrodynamic forces .
the polymer can be thought of as moving within a `` hydrodynamic heat bath '' .
the polymer chain is modelled by beads connected via non - harmonic springs@xcite with adjacent beads along the chain backbone representing an effective kuhn length of the polymer chain .
these finitely extensible springs are represented by the fene potential @xmath0,\quad \mbox{$r$}\,\mbox{$<$}\ , r_o . \label{fene}\end{aligned}\ ] ] a lennard - jones potential@xcite which acts between all the polymer beads is used to model the excluded volume of the monomers and a long range attraction which drives polymer collapse @xmath1.\label{lj}\end{aligned}\ ] ] we take @xmath2 , @xmath3 and @xmath4 where parameters and results are quoted in reduced lennard - jones units .
newton s equations of motion for the polymer are integrated using the time reversible velocity verlet algorithm@xcite .
the molecular dynamics time step is chosen to be @xmath5 where @xmath6 is the interval between solvent collision steps , defined below .
the solvent is modelled by a large number @xmath7 of point - like particles which move in continuous space with continuous velocities but discretely in time @xcite .
the algorithm is separated into two stages . in the first of these , a free streaming step ,
the positions of the solvent particles at time @xmath8 , @xmath9 , are updated simultaneously according to @xmath10 where @xmath11 is the velocity of a particle .
the second component of the algorithm is a collision step which is executed on both solvent particles and polymer beads .
the system is coarse - grained into @xmath12 unit cells of a regular cubic lattice . in this simulation @xmath13
there is no restriction on the total number of solvent or polymer particles in each cell , although the total number of particles is conserved .
multiparticle collisions are performed within each individual cell of the coarse - grained system by rotating the velocity of each particle relative to the centre of mass velocity @xmath14 of all the particles within that cell @xmath15 @xmath16 is a rotation matrix which rotates velocities by @xmath17 around an axis generated randomly for each cell and at each time step . in the present calculations we take @xmath18 .
note that the collision step preserves the position of the solvent and polymer beads .
it transfers momentum between the particles within a given cell while conserving the total momentum of these particles . because both momentum and energy are conserved locally the thermohydrodynamic equations of motion are captured in the continuum limit@xcite
hence hydrodynamic interactions can be propagated by the solvent and , because the polymer beads are involved in the collisions , to the polymer .
note , however , that molecular details of the solvent are excluded : this allows the hydrodynamic interactions to be modelled with minimal computational expense .
the volume in phase space is invariant under both the free streaming and collision steps .
hence the system is described by a microcanonical distribution at equilibrium@xcite .
the initial solvent distribution was generated by assigning positions randomly within the system with an average density @xmath19 particles per unit cell .
the velocities were assigned from a uniform distribution which relaxed rapidly ( @xmath20 ) to the equilibrium maxwell - boltzmann form .
a particularly useful feature of the malevanets - kapral algorithm is the ease with which hydrodynamic interactions can be `` turned - off '' thus replacing the hydrodynamic heat bath by a brownian ( random ) heat bath .
this is achieved by randomly interchanging the velocities of all the solvent particles after each collision step , thus relaxing the constraint of local momentum conservation to a global one .
accordingly the velocity correlations which result in hydrodynamic interactions disappear from the fluid . running simulations with the same initial conditions and parameter values , but with hydrodynamics present or absent , greatly faciliates pinpointing the effect of the hydrodynamic interactions .
our aim is to study the dynamics of polymer collapse from an extended to a compact state@xcite .
the collapse trasnsition is driven by the attractive lennard - jones interactions between the polymer beads .
the transition is rounded and shifted by the finite chain length , and takes place at @xmath21 for chains of length @xmath22 .
initial polymer configurations were chosen at random from long runs on a polymer chain at equilibrium in a solvent with @xmath23 .
each of these extended configurations were placed in a solvent at equilibrium at @xmath24 .
this value ensures collapse but prevents the chain collapsing so quickly that hydrodynamic interactions do not have sufficient time to develop .
the rate of polymer collapse was measured by monitoring the variation of the radius of gyration @xmath25 of the chain with time @xmath26 a typical numerical result is shown in figure 1 for a chain of length @xmath22 with and without hydrodynamics . the collapse time was estimated as the time when the equilibrium radius of gyration was first attained .
results for @xmath27 are shown for chains of varying lengths in table 1 .
@xmath27 was averaged over 20 initial configurations for @xmath28 , @xmath29 and @xmath30 , and 5 for @xmath31 , and the collapse time for hydrodynamic ( @xmath32 ) and brownian ( @xmath33 ) heat baths compared .
variations in collapse times between different runs are large as expected .
however , it is strikingly apparent that hydrodynamic interactions speed up the rate of collapse by a factor@xmath34 for each polymer length . for all but 6 of the 65 runs and for all the @xmath35 and @xmath36 runs the collapse was faster with hydrodynamics switched on .
the results in table 1 also illustrate that the collapse time increases with chain length as expected ( @xmath37 , @xmath38 ) .
values predicted in the literature for the exponents relating the collapse time to chain length vary widely .
it is far from obvious that these exponents are universal : they appear to depend on quench depth and the details of the model@xcite .
moreover the final radius of gyration was recorded for both the brownian and hydrodynamic collapses and found to agree as the equilibrium polymer properties should be independent of the nature of the heat bath .
figure 2 compares typical collapse pathways with and without hydrodynamics for @xmath35 and a final temperature @xmath24 . in this and
the majority of other runs a qualitative difference was observed in the collapse mechanism . in the brownian solvent
the ends of the polymer tend to collapse first forming a dumbbell shape . with hydrodynamics
the collapse takes place more evenly along the chain .
we speculate that this occurs due to slip - streaming. the movement of beads towards each other , which is initiated by the attractive lennard - jones potential , is enhanced by hydrodynamic interactions .
once a bead starts moving in a given direction and locally drags fluid with it , it is easier for neighbouring beads to move in the same direction .
similar behaviour albeit on much larger length scales has been observed experimentally by bartlett et al@xcite for a pair of colloidal particles in solution .
we have shown that it is possible to measure the dynamics of the collapse transition of a polymer in a solvent using a hybrid mesoscale / molecular dynamics algorithm .
hence it has been possible to show that hydrodynamic interactions speed up the collapse of the polymer chain .
the hydrodynamics alters the collapse pathway , allowing the folding to occur more homogeneously along the chain , rather than initially at the chain ends .
qualitative features of the collapse pathways are in agreement with those reported by chang and yethiraj@xcite who recently compared molecular dynamics simulations of a polymer in a solvent , which include hydrodynamics , to brownian dynamics simulations , which did not .
these authors also found that the brownian simulations could become trapped in a metastable free energy minimum .
it would be interesting to address whether this is due to deeper quenches than those considered here or to the different simulation approaches employed .
the solvent is modelled using the malevanets - kapral algorithm which sustains hydrodynamic modes but does not include molecular interactions . hence we caution that we can not investigate the late stages of collapse where trapping of the water molecules in the polymer chain may be important .
moreover collapse is driven by attractive lennard - jones interactions between the polymer beads rather than repulsive monomer - solvent interactions .
it would be of interest to include the solvent near the polymer in the more realistic molecular dynamics updating and the possibility of doing this is under investigation .
finally it may be of interest to consider whether the early stages of the folding of proteins in aqueous solution might be influenced by the hydrodynamic interactions of the solvent increasing the cooperativity of the collapse .
.averaged collapse time in units of the solvent time step @xmath6 of a polymer chain of length @xmath39 with ( @xmath40 ) and without ( @xmath41 ) hydrodynamics for a final temperature @xmath24 . the radius of gyration after collapse in reduced lennard - jones units with ( @xmath42 ) and without ( @xmath43
) hydrodynamics is also listed .
[ cols="^,^,^,^,^",options="header " , ] | we investigate numerically the dynamical behaviour of a polymer chain collapsing in a dilute solution .
the rate of collapse is measured with and without the presence of hydrodynamic interactions .
we find that hydrodynamic interactions both accelerate polymer collapse and alter the folding pathway . + * pacs : 83.80 .
rs * 2 |
Subsets and Splits