article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the optical and electronic properties of low - dimensional materials have been an important subject of study in the field of condensed matter physics . in particular ,
one - dimensional ( 1-d ) materials are predicted to possess unique properties that are distinctly different from those at higher dimensions,@xcite primarily due to the enhanced coulomb interactions among the quantum confined charge carriers .
one common feature of optically - excited low - dimensional systems is the formation of strongly bound electron - hole ( @xmath0-@xmath1 ) pairs , or excitons,@xcite which dominate interband optical spectra .
1-d semiconductors are expected to show an almost complete suppression of optical absorption at the band edges , with a significant fraction of the total oscillator strength taken by the lowest excitonic state.@xcite early reports of lasing from semiconductor quantum wires ( qwrs)@xcite invoked much interest in the physics of _ high density _ 1-d excitons .
a number of studies have thus far been performed on such qwr systems during the last two decades to understand many - body phenomena ( e.g. , lasing , band - gap renormalization , biexciton formation , and the mott transition),@xcite but many aspects are still under debate and not well understood .
more recently , studies on high - density @xmath0-@xmath1 pairs have been extended to novel 1-d materials such as conjugated polymers@xcite and single - walled carbon nanotubes ( swnts).@xcite the latter are tubular materials made of @xmath2-bonded carbon atoms,@xcite attracting much recent interest from diverse research fields due to their unique properties.@xcite semiconducting swnts are known to have extremely strong quantum confinement of @xmath3 1 nm , giving rise to large exciton binding energies on the order of 0.5 - 1 ev,@xcite much larger than those of gaas qwrs ( @xmath3 20 mev)@xcite and comparable to or larger than those of conjugated polymers ( @xmath3 0.4 ev@xcite and @xmath4 0.1 ev@xcite ) . here , we report results of experimental and theoretical investigations on the properties of photoluminescence ( pl ) from excitons in swnts through nonlinear photoluminescence excitation ( ple ) spectroscopy using intense optical pulses . from the clear saturation behavior observed in the intensities of all
the pl features as a function of excitation laser intensity as well as the complete _ flattening _ of the ple spectra observed at very high laser intensities , we show the existence of an upper limit on the density of excitons that can be accommodated in swnts .
such an upper limit is considered to be caused by the diffusive motion of the excitons@xcite combined with highly rapid and efficient exciton - exciton annihilation ( eea ) in swnts.@xcite as described in section [ model ] , we have developed a theoretical model for describing diffusion - limited eea processes in 1-d , which enabled us to simulate the pl saturation curves and estimate the densities of excitons in swnts as a function of excitation intensity .
a portion of this work was described in our earlier letter .
@xcite the purpose of the present paper is to provide a complete description of both experimental and theoretical aspects of this study .
the sample was prepared by ultrasonicating comocat swnts in d@xmath5o with 1 wt% sodium cholate for 1 hour , followed by ultracentrifugation at 111,000 g for 4 hours .
this centrifugation condition is sufficient to remove swnt bundles effectively.@xcite only the upper 50 % of the supernatant was collected and used for the experiment .
the solution was put in a 1-mm - thick quartz cuvette .
the optical density of the sample around the @xmath6 resonance was below 0.2 , which helped avoid non - uniform excitation and re - absorption of the emitted pl within the sample .
the excitation source was an optical parametric amplifier ( opa ) , producing @xmath3 250 fs pulses at a repetition rate of 1 khz , tunable in the visible and near - infrared ranges , pumped by a chirped - pulse amplifier ( clark - mxr , inc .
, cpa-2010 ) .
optical filters were carefully selected and set in the beam path to thoroughly eliminate any parasitic wavelength components ( mostly in the ultraviolet and near - infrared regions ) contained in the opa beam .
the opa beam was focused onto the sample to a spot size of 300 - 400 @xmath7 m .
only the central @xmath3 2 mm portion of the opa beam profile ( @xmath3 6 mm ) was taken out by using an aperture just before the focus to enhance the spot uniformity at the sample .
the pl from the sample was focused onto the monochromator entrance and recorded with a liquid - nitrogen - cooled ingaas 1-d array detector .
the obtained pl spectra were corrected for the wavelength - dependence of the grating efficiency and detector sensitivity . for the data shown in fig .
[ pl_change](c ) , a different sample as well as a different excitation light source was used for verifying the universality of the phenomena observed . in this case , the sample was a dried film of comocat swnts embedded in @xmath8-carrageenan , formed by drying a mixed gel of @xmath8-carrageenan and the centrifuged supernatant of comocat swnts on the surface of a sapphire substrate .
the sapphire substrate served as a mechanical support as well as a heat sink of the film during the measurements .
the excitation light was 1 khz and @xmath3 250 fs optical pulses with a central wavelength of 653 nm ( fwhm = 10 nm ) , produced by filtering whitelight pulses generated by focusing the cpa beam onto a sapphire crystal . [ tbp ] o and ( c ) comocat swnts embedded in a dried @xmath8-carrageenan film : ( a ) black solid spectrum obtained with opa pulse ( 654 nm , 29 nj ) .
red dotted curve spectrum obtained with a cw laser diode ( 658 nm , 100 @xmath7w ) .
inset shows that the two spectra coincide when the opa pulse energy is very low ( 300 pj ) .
( b ) change of pl spectra with the pulse energy of opa beam ( 654 nm ) varied between 1 nj and 30 nj in the order of # 1 to # 7 .
curve # 4 corresponds to the highest fluence ( @xmath31.3 @xmath9 10@xmath10 photons/@xmath11 ) .
( c ) change of pl spectra with the pulse energy of 653 nm light ( fwhm = 10 nm ) varied between 1 nj and 20 nj in the order of # 1 to # 7.,title="fig : " ] figure [ pl_change](a ) compares two pl spectra . the black solid curve was obtained using the opa with a wavelength of 654 nm ( or 1.90 ev ) and a pulse energy of 29 nj , while the red dotted curve was obtained using a weak ( 100 @xmath7w ) cw laser with a wavelength of 658 nm ( or 1.88 ev ) .
it is seen that the relative intensities of different pl peaks are drastically different between the two curves .
the inset confirms that the two spectra coincide accurately when the opa pulse energy was kept very low ( 300 pj ) .
figure [ pl_change](b ) shows pl spectra measured with pulse energies of 1 nj ( curves # 1 and # 7 ) , 4 nj ( # 2 and # 6 ) , 10 nj ( # 3 and # 5 ) , and 30 nj ( # 4 ) .
the ( 7,5 ) peak is dominant at low fluences while the ( 6,5 ) peak becomes dominant at high fluences .
it is important to note that the different curves were taken in the order of # 1 to # 7 , demonstrating that the observed changes are reproducible and are not caused by any laser - induced permanent change in the sample .
additionally , note that the pl intensities tend to saturate at high laser fluences , while their peak positions do not change at all .
figure [ pl_change](c ) shows pl spectra measured for the dried @xmath8-carrageenan film using 653 nm optical pulses with a fwhm bandwidth of 10 nm at different pulse energies .
the pulse energies were 1 nj ( curves # 1 and # 7 ) , 3 nj ( # 2 and # 6 ) , 10 nj ( # 3 and # 5 ) , and 20 nj ( # 4 ) measured in the order of # 1 to # 7 .
figure [ pl_change](c ) exhibits the same behavior as that shown in fig .
[ pl_change](b ) , demonstrating that the observed changes shown in fig .
[ pl_change](b ) did not result from any artifacts , e.g. , caused by the fluidic nature of the sample or by the unnoticed parasitic wavelength components in the opa beam . in the following , we use the excitation pulse fluence in terms of the number of incident photons per @xmath11 per pulse to express the intensity of excitation pulses .
[ tbp ] 10@xmath12 , ( b ) 1.2 @xmath9 10@xmath13 , ( c ) 4.1 @xmath9 10@xmath13 , and ( d ) 1.2 @xmath9 10@xmath10 photons/@xmath11 .
( e - h ) : ple spectra corresponding to ( a)-(d ) at emission wavelengths of 983 nm ( circles ) , 1034 nm ( squares ) , and 1125 nm ( triangles).,title="fig : " ] figures [ nl - ple](a)-[nl - ple](d ) show ple maps taken with various pump fluences .
the step size for the pump photon energy was 20 mev .
the data taken with the lowest fluence ( 1.2 @xmath9 10@xmath12 photons/@xmath11 ) [ fig .
[ nl - ple](a ) ] is essentially the same as that taken with low - intensity cw light .
however , as the fluence is increased [ figs .
[ nl - ple](b)-[nl - ple](d ) ] , the @xmath6 excitation peaks gradually broaden and eventually become completely flat at the highest fluence ( 1.2 @xmath9 10@xmath10 photons/@xmath11 ) i.e. , _ pl intensities become independent of the excitation wavelength_. the corresponding ple spectra are shown in figs .
[ nl - ple](e)-[nl - ple](h ) for three pl wavelengths at 983 , 1034 , and 1125 nm . again , such changes in the ple spectra were reproducible over the fluence range tested here , indicating that no sample damage was induced . in order to obtain pl intensity ( @xmath14 ) versus pump intensity ( @xmath15 ) relationships for different emission peaks , we measured pl spectra at different photon fluences for various excitation wavelengths .
each pl spectrum was decomposed and fitted by multiple peaks corresponding to the swnt types / chiralities involved in the measured wavelength range .
50% gaussian + 50% lorentzian lineshape was assumed , and the decomposition was performed by optimizing the peak - width so that the decomposition gives the best fitting for the original pl spectrum .
the optimum widths at the highest fluence was larger by @xmath3 15 % than those at the lowest fluence in fig .
[ pl_change ] , and such an increment of the width is considered to be caused by the enhanced interactions among excitons or their reduced lifetime in the presence of high density excitons . throughout the decomposition analysis
performed , peak positions of all the pl features and the ratios among their widths were fixed regardless of the excitation wavelength and fluence .
figure [ saturation ] shows the obtained integrated pl peak intensities ( @xmath14 ) plotted against the incident photon fluence ( @xmath15 ) for ( 6,5 ) , ( 7,5 ) , and ( 8,3 ) swnts at excitation wavelengths of 570 , 615 , and 658 nm .
the resonance wavelengths of these swnt types at the @xmath6 levels are approximately 570 , 647 , and 673 nm , respectively .
it can be seen that the integrated pl intensity begins to saturate at a lower ( higher ) fluence when swnts are excited resonantly ( non - resonantly ) .
unexpectedly fast saturation of the pl from ( 7,5 ) with 570 nm excitation ( which is non - resonant ) can be attributed to its proximity to the phonon sideband at @xmath3 585 nm.@xcite [ tbp ] 5% .
the solid and dashed curves are fitting by eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) and eq .
( [ conventional_3 ] ) , respectively.,title="fig : " ] we interpret these observations as results of very efficient exciton - exciton annihilation,@xcite a non - radiative process that occurs at high exciton densities where two excitons are spatially close enough to interact with each other , resulting in the annihilation of the two excitons and simultaneous creation of an _ e - h _ pair in a higher energy state ( either as a bound exciton or an unbound free _ e - h _ pair ) .
we assume that the formation of @xmath16 excitons occurs in a very short time scale after an optical excitation around @xmath6 , because of much faster @xmath6-to-@xmath16 relaxation ( e.g. , @xmath340 fs @xcite ) than the duration of our opa pulse ( @xmath3250 fs ) .
thus , excitons quickly accumulate in the @xmath16 state during and right after photo - creation of @xmath0-@xmath1 pairs .
however , the number of excitons that can be _ accommodated _ in the @xmath16 state is limited by eea . as the exciton density , @xmath17 , approaches its maximum value , eea begins to prevent a further increase by _ efficiently removing excitons non - radiatively _ , which explains the pl saturation behavior .
since eea serves as a bottleneck for the exciton density , the pl intensity becomes insensitive to whether the excitons were created resonantly or non - resonantly around the @xmath6 level and independent of the pump wavelength , resulting in the flattening of ple spectra .
namely , at very high pump fluences , the pl intensity is determined not by how efficiently excitons are created but by _ how many @xmath16 excitons can be accommodated within a particular type of swnt _ as well as by the relative abundance of that type of swnt in the sample .
we also performed optical transmission spectroscopy in the @xmath6 range using opa pulses and found that the absorption spectra do not exhibit any change even at high pulse fluences .
thus , nonlinear optical effects such as phase - space filling in the @xmath6 range are not playing any role in the observed ple broadening / flattening and pl saturation .
in this section , we develop a theoretical model for explaining the experimental results , taking into account the generation , diffusion - limited eea , and spontaneous decays of 1-d excitons . under certain approximations , the model provides a direct analytical relationship between the intensity of the excitation light ( @xmath15 ) and that of the emitted pl ( @xmath14 ) for limiting cases , which allows us to estimate the density of excitons in swnts through fitting to the experimentally obtained @xmath15 vs. @xmath14 curves .
figure [ schematic_1](a ) shows a schematic energy diagram of the excitons in semiconducting swnts under consideration .
we are interested in calculating the population of excitons @xmath18 ( indicated by the dotted box ) in the lowest energy state @xmath19 .
first , excitons are created by optical excitation at an energy around the @xmath20 level , which is typically in the visible wavelength range .
the excitation intensity is denoted by @xmath15 .
as soon as excitons are created , they decay to the @xmath19 level within a very short time ( @xmath340 fs)@xcite by transferring their energies to the lattice .
recent studies have reported that the excitons created around @xmath20 levels primarily decay to the @xmath19 level with a probability close to unity.@xcite the influx of excitons to the @xmath19 level is denoted by @xmath21 . on the other hand , the spontaneous decay time @xmath22 of the @xmath19 excitons to the ground state ( g. s. ) has been reported to be 10100 ps.@xcite such a spontaneous decay consists of radiative and non - radiative processes , with respective rates @xmath23 and @xmath24 ( s@xmath25 ) , where @xmath26 .
the outflux of excitons from the @xmath19 level _ via the spontaneous decay process _ is denoted by @xmath27 .
therefore , the flux of the pl photons or the pl intensity ( @xmath14 ) is equal to @xmath28 , where @xmath29 ( @xmath30 ) denotes the branching ratio for the radiative decay from the @xmath19 level . )
is the domain of interest where @xmath18 excitons are populated .
all the symbols are defined in the text .
( b ) schematic description of @xmath31 excitons randomly distributed over a swnt with a length of @xmath32 .
the horizontal arrow with a length of @xmath33 denotes the average length traversed by one exciton during its spontaneous decay lifetime @xmath22 .
the lower part is the equivalent of the upper but emphasizes that the total length of the unoccupied region is @xmath34 where @xmath18 vertical thick bars denote the borders of the areas occupied by those excitons .
the ends of the swnt are assumed to be a cyclic boundary . ] as the density of excitons increases , the eea process becomes important .
if the _ e - h _ pair created in the higher energy state returns back to the @xmath19 exciton level with a probability of @xmath35 ( @xmath36 ) , the initial two excitons are eventually reduced to @xmath35 excitons ( as an expectation value ) through such an eea process .
we make the following two assumptions : ( 1 ) once two excitons intersect in a swnt , eea occurs instantaneously with a probability of one , and ( 2 ) the positions where excitons are created in swnts by optical excitations are random .
figure [ schematic_1](b ) schematically shows a situation in which @xmath18 excitons exist in a swnt of length of @xmath32 .
each exciton is considered to `` occupy '' a characteristic length @xmath33 in a swnt . in a static limit
, @xmath33 should simply be the exciton size .
on the other hand , when exciton diffusion is present , @xmath33 is assumed to be the average distance traveled by an exciton during the spontaneous decay lifetime @xmath22 and is @xmath37 , where @xmath38 ( @xmath39/s ) is the exciton diffusion constant .
namely , it is assumed that @xmath33 is determined by the diffusion length , and any two excitons created within @xmath33 undergo eea .
first , a 1-d space of length @xmath32 is considered where @xmath18 segments ( or excitons ) of length @xmath33 are randomly present without overlapping each other , as shown in fig .
[ schematic_1](b ) .
the probability of a new segment ( of length @xmath33 ) to enter the system _ without _ overlapping any of the @xmath18 existing segments , @xmath40 , is given by the product of the following two probabilities @xmath41 which is the probability for the center of the new segment to land on an unoccupied area [ hatched regions in fig .
[ schematic_1](b ) ] , and @xmath42 which is the probability that the occupying length ( @xmath33 ) of the new segment ( whose center has landed on an unoccupied area ) does not interfere with any of the @xmath18 existing segments [ vertical thick bars in the lower part of fig .
[ schematic_1](b ) ] .
hence , @xmath40 is written as @xmath43 the expectation value of the increment of @xmath18 due to the introduction of a new segment into the system , @xmath44 ( @xmath45 ) , depends on the type of two - particle annihilation considered .
for the general `` @xmath46 '' case , @xmath44 is expressed as @xmath47 in the following , the derivation for the case of @xmath48 is shown as a specific example , since this case is considered to represent our experimental situation .
the final results will be presented for both the @xmath48 and @xmath49 cases . here , we consider the _ steady - state _
limit , where the number of excitons @xmath18 in the system is steady , corresponding to cw excitation . in order to derive the relationship between @xmath15 and @xmath14
, we consider the relationship between @xmath21 and @xmath27 [ see fig .
[ schematic_1](a ) ] , where both @xmath21 and @xmath27 are _ rates _ , having units of s@xmath25 .
when @xmath18 is sufficiently small and eea is negligible , @xmath50 . when eea is present , however , this relationship becomes @xmath51 in the case of @xmath48 , eq .
( [ n_n ] ) and eq .
( [ g_in ] ) lead to @xmath52 we introduce two dimensionless variables @xmath53 and @xmath54 , the former of which is a dimensionless exciton population ( @xmath55 ) . using these variables , eq .
( [ i_pump_beta ] ) can then be rewritten in dimensionless form : @xmath56 expanding the second factor in the denominator of eq .
( [ psi - zeta1 ] ) and eliminating higher - order terms of @xmath57 leads to an equation of only @xmath58 and @xmath59 , expressed as @xmath60 summing up to @xmath61 in eq .
( [ psi - expansion ] ) is sufficient to reproduce eq .
( [ psi - zeta1 ] ) for @xmath62 . finally , noting that the denominator of eq .
( [ psi - expansion ] ) is equal to the taylor expansion of an exponential function , the solution for the steady - state limit for @xmath48 ( @xmath63 ) is expressed as @xmath64 here , since @xmath65 and @xmath66 , @xmath14 and @xmath15 are proportional to @xmath58 and @xmath59 , respectively , i.e. , @xmath67 and @xmath68 , where @xmath69 and @xmath70 are real constants . on the other hand
, the solution for the case of @xmath49 ( @xmath71 ) is derived through a similar procedure , yielding @xmath72 equations ( [ steady - state - eqn_(lambda=1 ) ] ) and ( [ steady - state - eqn_(lambda=0 ) ] ) , valid for cw pl experiments , are implicit equations relating the pl intensity ( @xmath14 ) and the pump intensity ( @xmath15 ) in terms of their respective dimensionless variables @xmath58 and @xmath59 .
these equations contain no fitting parameters except the two linear scaling factors @xmath69 and @xmath70 and simply become equivalent ( @xmath73 ) in the low density limit ( @xmath74 ) . here
, we consider the _ instantaneous _ limit in which creation of all the excitons by an infinitesimally short optical pulse and their internal relaxation to @xmath19 level occur instantaneously at @xmath75 , before diffusion - limited eea and spontaneous decay processes begin to occur subsequently .
this limit represents a situation where the duration of the optical pulse and the time required for intraband relaxation are much shorter than the spontaneous decay time @xmath22 , as in the case of the present experimental situation .
the pump intensity @xmath15 in this limit is directly proportional to @xmath76 , the number of excitons or @xmath0-@xmath1 pairs created at @xmath75 , while the pl intensity @xmath14 is proportional to the number of excitons @xmath18 that survived eea .
the relationship between @xmath76 and @xmath18 is described by the differential equation @xmath77 where @xmath44 is given by eq .
( [ n_n ] ) . when @xmath48 is assumed , eq .
( [ dn_0 ] ) is rewritten as @xmath78 as in the previous case ( steady - state limit ) , we introduce dimensionless variables @xmath53 and @xmath79 . again , since @xmath65 and @xmath80 , @xmath67 and @xmath68 , where @xmath69 and @xmath70 are constants . with @xmath58 and @xmath59 ,
( [ dn0_dn ] ) can be written in a dimensionless form : @xmath81 furthermore , similar to the steady - state case , an expansion is performed on the second factor in the denominator of eq .
( [ psi - zeta2 ] ) together with elimination of the higher - order terms of @xmath82 , resulting in a differential equation of only @xmath58 and @xmath59 : @xmath83 finally , by integrating eq .
( [ instantaneous - eqn - before - integ_(lambda=1 ) ] ) from 0 to @xmath58 , we obtain the solution for @xmath48 ( @xmath63 ) as @xmath84 where @xmath85 is the exponential integral . similarly , the solution for @xmath49 ( @xmath71 ) in the instantaneous limit is @xmath86 where the integral has to be solved numerically . those dimensionless equations ( [ instantaneous - eqn_(lambda=1 ) ] ) and ( [ instantaneous - eqn_(lambda=0 ) ] ) become equivalent ( @xmath73 ) in the low density limit ( @xmath74 ) .
to compare with the experimental data shown in fig .
[ saturation ] , we use eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) , which is for the instantaneous limit with @xmath48 .
the choice of this equation is because of the short duration of the optical pulses used ( @xmath3250 fs ) , the very fast ( @xmath340 fs@xcite ) and efficient@xcite internal decay of excitons created at @xmath6 levels to the lowest @xmath16 level , and the much longer spontaneous decay time from the @xmath16 level to g. s. ( 10100 ps).@xcite the solid curves shown in fig .
[ saturation ] are fitting curves using eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) , indicating that the model agrees well with the experimentally observed pl saturation curves .
table [ parameters ] is a summary of the fitting analysis performed on the data in fig .
[ saturation ] using eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) .
the first two columns show the optimum values of the scaling factors @xmath69 and @xmath87 .
these are thought to be proportional to the oscillator strength for the pl emission from @xmath16 levels and the oscillator strength for the optical absorption around @xmath6 levels , respectively , as expected from @xmath88 and @xmath89 .
the right three columns ( @xmath58 , @xmath18 , and @xmath90 ) are values at @xmath91 photons/@xmath11 ( see fig .
[ saturation ] ) .
the values of @xmath18 were obtained through the relationship @xmath92 , where @xmath33 was assumed to be 45 nm ( one half of the average exciton excursion range defined in ref . ) .
the exciton density in the highly saturated regime ( @xmath93 cm@xmath25 ) is still more than one order magnitude smaller than the expected mott density in swnts ( @xmath95 cm@xmath25 , assuming an exciton size of @xmath31.5 nm@xcite ) , as has been discussed in detail in ref . .
@xmath90 in the right - most column denotes the ratio between the number of as - created @xmath0-@xmath1 pairs and the number of excitons that survived the eea process until their spontaneous decay to g. s. , or @xmath96 .
the values of @xmath90 show that approximately 90% of the initially created @xmath0-@xmath1 pairs / excitons decay non - radiatively through the eea path when the @xmath6 levels are resonantly excited at a fluence of @xmath91 photons/@xmath11
. as expected , the values of @xmath69 are similar within the same chirality type , regardless of the excitation wavelength . however , for the particular case of ( 6,5 ) with 570 nm excitation , where the excitation wavelength exactly coincided with the @xmath6 resonance peak , the value of @xmath69 ( @xmath97 ) is appreciably smaller than those at the other excitation wavelengths ( @xmath98 ) .
such a smaller @xmath69 value implies that more excitons decayed to the g. s. through non - radiative paths compared to the other cases . this may be due to an emergence of stronger nonlinear processes that have not been taken into account in the assumptions , such as the breakdown of the @xmath48 assumption and/or the appearance of _ more - than - two_-body annihilation processes because of the very high initial @xmath0-@xmath1 pair density , achieved with the strong @xmath6 resonance in this particular case .
the values of @xmath87 also show the expected tendency toward higher ( lower ) values when the excitation wavelength is closer to ( further from ) the @xmath6 resonance wavelength .
the slightly higher @xmath87 value for the case of ( 7,5 ) with 570 nm excitation is again attributed to its vicinity to the @xmath6 phonon sideband at @xmath3585 nm.@xcite .summary of the optimal parameters to used fit the experimental data , obtained from the analysis with eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) .
@xmath58 , @xmath18 , and @xmath90 are values at the fluence of @xmath91 photons/@xmath11 , corresponding to the largest fluence in fig .
[ saturation ] . [ cols="^,^,^,^,^,^,^ " , ] ) ( black solid curve ) and the result obtained by a monte carlo simulation of the eea process .
the abscissa denotes the density of initially created excitons while the ordinate corresponds to the resultant density .
the experimental range of @xmath76 in fig .
[ saturation ] is @xmath99 cm@xmath25 . ] to further confirm the validity of the model , we performed a computational simulation based on the monte carlo method . at the beginning of the simulation , a random distribution of @xmath76 excitons along a line
is created at time @xmath75 , corresponding to the instantaneous limit . for @xmath100 ,
each exciton undergoes a random movement in each computational step @xmath101 by the distance given by the probability distribution of @xmath102 , where @xmath103 denotes the normal distribution centered at @xmath104 with variance @xmath105 and @xmath106 . upon intersection of any two excitons
, the eea of @xmath107 takes place .
in addition , each exciton is eliminated from the system with a probability of @xmath108 ( @xmath109 ) , corresponding to possible spontaneous decay during each computational step . the total density of excitons that have decayed spontaneously ( @xmath18 )
can thus be calculated for each value of the initial density @xmath76 .
figure [ simulation ] shows a comparison between the model for the instantaneous limit [ eq . ( [ instantaneous - eqn_(lambda=1 ) ] ) ] with @xmath33 = 45 nm and the result of the monte carlo simulation , plotting the relationship of @xmath76 and @xmath18 .
the simulation was performed with @xmath110 @xmath11/s and @xmath111 ps,@xcite which resulted in an average exciton displacement of 45 nm based on the same simulation performed without eea .
the comparison shows satisfactory agreement , indicating that the simple analytical form of eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) well describes the phenomenon of diffusion - limited eea , hence validating our model .
conventionally , a rate equation of the form @xmath112 has been used to explain eea processes observed for various materials from 1-d to 3-d.@xcite the terms on the right hand side , from left to right , represent the rate of excitons entering the system ( or the @xmath19 level in the present case ) , the rate of excitons spontaneously decaying from the system to the g. s. ( at @xmath113 ) , and the rate of excitons leaving the system because of eea ( at @xmath114 ) , respectively . here , the probability of finding two excitons at the same position ( or the eea rate ) is assumed to be proportional to @xmath115 , and hence , exciton diffusion as well as the finite size ( i.e. , length ) that each 1-d exciton occupies are not taken into account in eq .
( [ conventional_1 ] ) .
equation ( [ conventional_1 ] ) is solved for the pulse - wise creation of @xmath76 excitons at @xmath75 [ i.e. , @xmath116 in order to be compared with the proposed model for the instantaneous limit [ eq . ( [ instantaneous - eqn_(lambda=1 ) ] ) ] as well as with the experimental results . with this initial condition , eq .
( [ conventional_1 ] ) is readily solved to be @xmath117 \;. \label{conventional_2}\ ] ] the total number of photons emitted from the sample , @xmath14 , is obtained by integrating eq .
( [ conventional_2 ] ) from @xmath75 to @xmath118 as @xmath119 where @xmath120 as before .
equation ( [ conventional_3 ] ) describes the relationship between @xmath14 and @xmath76 in terms of two adjustable components @xmath121 and @xmath122 .
the dashed lines drawn in fig . [ saturation ] are the optimum fits by eq .
( [ conventional_3 ] ) to the experimental data . while eq .
( [ conventional_3 ] ) reproduces the behavior well for the regime of weaker pl saturation with non - resonant excitations , the deviation becomes clearer for the regime of stronger pl saturation with resonant excitations .
most importantly , however , the density predicted by eq .
( [ conventional_3 ] ) has _ no upper limit _ as recognized by the steadily increasing dashed lines shown in fig .
[ saturation ] [ or by the logarithmic form of eq .
( [ conventional_3 ] ) ] , while eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) correctly demonstrates the existence of an upper limit , as observed experimentally and in the monte carlo simulation . another important difference between eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) and eq .
( [ conventional_3 ] ) is described as follows : the solution of the conventional rate equation [ eq . ( [ conventional_3 ] ) ] contains two independent quantities : @xmath121 and @xmath122 .
the former can be used to scale @xmath14 . however , since the parameter @xmath122 simultaneously scales @xmath15 ( @xmath123 ) _ and _ changes the shape of the saturation curve , one essentially can not estimate the density of excitons from the fitting analysis based on eq .
( [ conventional_3 ] ) . on the other hand , since the shape of eq .
( [ instantaneous - eqn_(lambda=1 ) ] ) has been uniquely determined by its dimensionless representation , the only thing one can do is to _ linearly and independently _ change the two scaling constants @xmath69 and @xmath70 for @xmath14 and @xmath15 , respectively , yielding sets of @xmath58 and @xmath59 values along the scaled ( or fitted ) @xmath14 vs. @xmath15 curves .
we have investigated photoemission properties of high - density 1-d excitons in single - walled carbon nanotubes . as the excitation intensity increases , all photoluminescence emission peaks arising from different chiralities showed clear saturation in intensity .
each peak exhibited a saturation value that was independent of the excitation wavelength , indicating that there is an upper limit on the exciton density for each nanotube species .
we interpret these results in terms of diffusion - limited exciton - exciton annihilation processes through which high - density excitons decay non - radiatively . to quantitatively understand the saturation behavior observed in the experiment , we have developed a theoretical model , taking into account the generation , diffusion - limited exciton - exciton annihilation , and spontaneous decays of 1-d excitons .
the saturation curve predicted by the model under appropriate approximations reproduced the experimental saturation curves well , and the fitting analysis allowed us to estimate the density of excitons for a given diffusion constant .
we also compared our results with monte carlo calculations , confirming the validity of our model .
additionally , we examined the saturation behavior predicted by the solution of the conventional exciton - exciton annihilation rate equation that does not take into account diffusion and showed that the solution qualitatively failed to fit the experimental data at high exciton densities , showing the inappropriateness of its use for excitons in carbon nanotubes .
the approach presented in this paper should have wide applicability for predicting the intensities of photoluminescence from single - walled carbon nanotubes under various excitation conditions . an important conclusion drawn from the current study
is that a large density of electron - hole pairs is difficult to achieve in single - walled carbon nanotubes .
this imposes a serious challenge for making active optoelectronic devices based on semiconducting single - walled carbon nanotubes 1-d semiconductor lasers , in particular , which would require densities comparable to , or higher than , the mott density .
the existence of an upper limit on the density of excitons would also prevent fundamental studies of bosonic characters of 1-d excitons expected at quantum - degenerate densities .
hence , novel methods are needed for minimizing non - radiative decay processes such as exciton - exciton annihilation in single - walled carbon nanotubes , which would require ways to control the dynamic parameters of excitons such as the spontaneous decay rate and diffusion constant through , e.g. , varying the temperature as well as applying an external magnetic field .
@xcite the authors thank ajit srivastava for valuable discussions and erik einarsson for emendation .
we thank the robert a. welch foundation ( c-1509 ) and nsf ( dmr-0325474 and oise-0530220 ) for support .
one of us ( y. m. ) was financially supported by grants - in - aid for scientific research ( # 18 - 09883 ) from the japan society for the promotion of science ( jsps ) .
y. m. thanks tatsuya okubo and shigeo maruyama for support to his fulfillment of the jsps research program . | through an investigation of photoemission properties of highly - photoexcited single - walled carbon nanotubes , we demonstrate that there is an upper limit on the achievable excitonic density . as the intensity of optical excitation increases , all photoluminescence emission peaks arising from different chirality single - walled carbon nanotubes showed clear saturation in intensity .
each peak exhibited a saturation value that was independent of the excitation wavelength , indicating that there is an upper limit on the excitonic density for each nanotube species .
we propose that this saturation behavior is a result of efficient exciton - exciton annihilation through which excitons decay non - radiatively . in order to explain the experimental results and obtain excitonic densities in the saturation regime ,
we have developed a model , taking into account the generation , diffusion - limited exciton - exciton annihilation , and spontaneous decays of one - dimensional excitons . using the model
, we were able to reproduce the experimentally obtained saturation curves under certain approximations , from which the excitonic densities were estimated .
the validity of the model was confirmed through comparison with monte carlo simulations .
finally , we show that the conventional rate equation for exciton - exciton annihilation without taking into account exciton diffusion fails to fit the experimentally observed saturation behavior , especially at high excitonic densities . |
the harris - plischke - zuckermann model@xcite has long been used to study the effects of random uniaxial anisotropy on ferromagnetism .
the hamiltonian of this random anisotropy model ( ram ) is @xmath9 where each @xmath10 , the dynamical on site @xmath11 , is usually taken to be a classical three - component spin of unit length .
each @xmath12 is a time - independent unit vector .
the @xmath13 on different sites are assumed to be uncorrelated random variables .
@xmath14 is a sum over nearest neighbors on some lattice . in this work
we will use a simple cubic lattice with periodic boundary conditions , and we will study the case of two - component ( @xmath15 ) spins . as was discussed in some detail in an earlier paper,@xcite
if one chooses the @xmath10 and the @xmath12 to be two - component vectors , then the hamiltonian can be mapped onto a model of a spin - density wave ( sdw ) in an anisotropic material with an easy axis . for @xmath0 spins , _
i.e. _ @xmath15 , the hamiltonian of the model may be rewritten as @xmath16 \ , .\ ] ] each @xmath17 is a dynamical variable which takes on values between 0 and @xmath18 .
the @xmath19 indicates here a sum over nearest neighbors on a simple cubic lattice of size @xmath1 .
we choose each @xmath20 to be an independent identically distributed quenched random variable , with the probability distribution @xmath21 for @xmath22 between 0 and @xmath18 .
a constant term has also been added to the anisotropy , to make the hamiltonian well - behaved in the limit @xmath23 . in this work
we will study eqn . 2 on the simple cubic lattice over a range of @xmath24 , using monte carlo simulations .
the large increase in available computing resources over the last fifteen years makes possible significant improvements over the earlier results.@xcite by studying a range of @xmath2 , we will be able to learn about the stability of long - range order against random pinning which respects the kramers degeneracy , such as alloy disorder , and the critical behavior of a sdw in an easy - axis material with this type of pinning .
in the limit @xmath23 , often called the ising limit , both analytical@xcite and numerical@xcite calculations become substantially simplified .
this is due to the fact that in this ising limit the random anisotropy term in the hamiltonian becomes a projection operator , and each spin has only two allowed states .
it has been argued that for large @xmath24 the behavior is close to the @xmath25 limit as long as @xmath26.@xcite it has also been found , however , that for @xmath15 at low temperatures and moderately large values of @xmath24 the magnetization per spin on @xmath1 simple cubic lattices , @xmath27 , decreases@xcite as the temperature , @xmath8 , is lowered .
this effect was not seen for @xmath25 .
a similar effect is seen in the case of the random bond ising model ( rbim ) , where the nishimori gauge symmetry causes the magnetization to have a maximum at a finite @xmath8 on the nishimori line.@xcite the rbim is the natural extension@xcite of the ram to the case of ising spins , @xmath28 . thus it should be expected that the phase diagram of the @xmath15 ram has a close relation to that of the rbim . however , there are aspects of the phase diagrams which remain somewhat mysterious .
for example , chen and lubensky@xcite found that the critical exponents which describe the stability of the ferromagnet - spin glass - paramagnet multicritical point for the random bond model in @xmath29 dimensions are well - behaved for @xmath28 , but become complex for @xmath15 and @xmath30 .
one interpretation of this puzzling result is that the multicritical point itself becomes unstable in @xmath29 dimensions , so that it becomes a region of the phase diagram , rather than a single point . in this expanded multicritical region
one might expect to find quasi - long - range order ( qlro ) .
although an explicit calculation has not been done , a similar result would not be surprising for the ram .
the existence of qlro in the ram was first suggested by aharony and pytte@xcite in 1980 .
they later@xcite pointed out that higher order terms might make the correlation length , @xmath31 , finite below @xmath5 .
feldman@xcite has argued that qlro should be common in disordered magnets and similar systems .
thus there are a number of possibilities available for the topology of the phase diagram . in a cayley - tree mean - field theory , where qlro does not occur , it is known@xcite that in the limit @xmath32 the phase diagram depends on the parameter @xmath33 , where @xmath34 is the number of nearest neighbor spins .
thus it is to be expected that the phase diagram in three dimensions will also depend on the lattice type , @xmath35 and the range of the exchange interactions , as well as on @xmath36 and @xmath24 . for the simple cubic lattice ( which has @xmath37 ) it has been shown@xcite that in the limit @xmath32
the ground state is an ising spin glass when @xmath38 . for small @xmath24
, however , where one does not expect the qualitative behavior to depend on @xmath34 , feldman@xcite predicts qlro in @xmath39 for @xmath40 . in the @xmath30 case , this appears to be confirmed by monte carlo calculations.@xcite the presence of a reentrant phase is difficult to demonstrate conclusively using the type of numerical calculations we have performed here .
it was only relatively recently that reentrance was demonstrated convincingly in the @xmath41 rbim.@xcite there may also be a range of @xmath24 for which the three - dimensional @xmath15 model has a reentrant ferromagnetic phase .
one motivation for believing this is that reentrance is frequently observed in laboratory experiments .
another is the work of pelcovits , pytte and rudnick@xcite who argue that ferromagnetism should be unstable in the ram for low @xmath8 and small @xmath24 .
since magnetization can increase with increasing @xmath8 at low @xmath8 , ( which was not known at the time of their work , ) it is not correct to claim that the absence of ferromagnetism near @xmath42 precludes the existence of a ferromagnetic phase in the ram at a somewhat higher @xmath8 .
larkin@xcite studied a model for a vortex lattice in a type - ii superconductor .
his model replaces the spin - exchange term of the hamiltonian with a harmonic potential , so that each @xmath17 is no longer restricted to lie in a compact interval .
he argued that for any non - zero value of a random field this model has no long - range order on a lattice whose dimension @xmath43 is less than or equal to four .
this argument , using the harmonic potential instead of the spin - exchange , is only rigorously correct in the limit @xmath44 .
a more intuitive derivation of the result was given by imry and ma,@xcite who assumed that the increase in the energy of an @xmath45 lattice when the order parameter is twisted at a boundary scales as @xmath46 , just as it does in the nonrandom ferromagnet .
as argued by imry and ma,@xcite and later justified more carefully,@xcite within an @xmath47-expansion one finds the phenomenon of dimensional reduction " . within this perturbation theory the critical exponents of any @xmath43-dimensional @xmath48 random - field model ( rfm )
( for which the kramers degeneracy is broken by the randomness ) appear to be identical to those of an ordinary @xmath48 model of dimension @xmath49 . for the ising ( @xmath28 )
case , this dimensional reduction was shown rigorously to be incorrect.@xcite another interesting development was the calculation of mezard and young,@xcite who showed that random fields caused breaking of replica symmetry below @xmath5 for any finite value of @xmath35 .
thus there is no good reason to expect that dimensional reduction should be correct near @xmath5 for any finite value of @xmath35 .
although there is certainly a family resemblance between the rfm and the ram , the difference between breaking the kramers degeneracy at the level of hamiltonian and breaking it spontaneously has profound consequences .
one such consequence is theorem 4.4 of aizenman and wehr,@xcite which applies to the rfm , but not to the ram .
a naive but not entirely misleading analogy may be drawn between the relationship of the rfm to the ram and the relationship between applying a uniform magnetic field to a ferromagnet or to an antiferromagnet .
the field which couples linearly to the order parameter has a qualitatively stronger effect than the field which couples quadratically to the order parameter .
translation invariance of @xmath50 is broken for any non - zero value of @xmath51 , since the vectors @xmath12 are random . within a high - temperature perturbation theory , performing a configuration average over the ensemble of random lattices appears to restore translation invariance above @xmath5 .
however , the radius of convergence of this perturbation theory can not be greater for @xmath52 than it is for @xmath53 . for models described by eqn .
( 1 ) , the @xmath5 predicted by extrapolating the low orders of perturbation theory is always maximal at @xmath53 .
this implies that for @xmath52 the high - temperature perturbation theory does not converge near @xmath5 .
the inadequacy of perturbation theory to describe @xmath0 models in @xmath39 , because of the effects of vortex lines , has been discussed by halperin.@xcite while it is not clear than halperin s argument is valid for the rfm , where the kramers degeneracy is broken by the hamiltonian , it should be valid for the ram .
thus it seems quite implausible that for @xmath54 the twist energy for eqn .
( 2 ) really scales as @xmath46 when @xmath52 , even though this is correct to all orders in the configuration - averaged perturbation theory .
the argument of pelcovits@xcite for the @xmath15 ram , which is a prototype for much subsequent work,@xcite assumes that if one goes to small enough @xmath24 and low enough @xmath8 then the effects of vortex lines can be ignored .
in essence , what is done is to replace the spin variables by a noncompact `` elastic manifold '' .
these authors then claim that this does not effect the behavior one is studying
. however , this can not be true when one considers behavior on scales larger than the imry - ma length.@xcite the basic point is that imry - ma - type arguments for continuous o(_n _ ) spins ( _ i.e. _ @xmath55 ) are not self - consistent .
one begins by assuming that the random field is weak so that the twist energy scales as @xmath46 , as in the absence of the randomness .
then one shows that , if @xmath56 and @xmath57 , the effective coupling to the random field increases as @xmath2 increases .
if the effective random coupling is strong , however , then assuming that the twist energy is uniformly distributed throughout the volume is not reasonable .
the conclusion which should be drawn from this is that a deeper analysis is needed when @xmath56 . in order to understand whether the problems with perturbation theory are actually due to vortex lines , and
thus restricted to the @xmath15 case , or if similar problems can also be expected for @xmath58 , it may be helpful to reconsider the analysis of pelcovits , pytte and rudnick.@xcite these authors show that within their perturbation theory the pure o(_n _ ) mean - field theory critical fixed point remains stable against random anisotropy for @xmath59 .
this contrasts to the random - field case , where mean - field theory is only stable for @xmath60 .
then they argue that for @xmath56 and @xmath55 there is no stable critical fixed point for the ram , because under rescaling transformations the random anisotropy coupling constant runs off to @xmath4 .
however , they did not ( and within their formulation could not ) examine the possibility that there could exist another ferromagnetic critical fixed point at a large value of @xmath24 .
the reason why such an object may exist is that there exist alternative formulations of mean - field theory@xcite for the ram in the limit @xmath32 .
it is useful to consider the generalization of eqn .
( 2 ) to @xmath61-fold random fields.@xcite in the @xmath41 case@xcite it has been shown that there continues to be a kosterlitz - thouless phase as long as @xmath62 , _
i.e. _ for @xmath62 a weak @xmath61-fold random field does not destroy the kosterlitz - thouless phase .
it was claimed by aharony@xcite that ferromagnetism should be unstable for any value of @xmath61 when @xmath39 . however , a computer simulation study@xcite for @xmath63 is not consistent with this claim , which is based on the weak randomness perturbation theory around the @xmath53 model .
this @xmath39 computer simulation finds that there is a mass gap at @xmath42 , an effect which can not be reproduced within the perturbation theory .
the interpretation of this is that for @xmath64 the thickness of a domain wall remains finite in the limit @xmath65 , _
i.e. _ the domain wall becomes localized by random pinning .
removing vortex lines from the pure @xmath0 model by letting the vortex fugacity become large forces the system into a ferromagnetic state at any temperature.@xcite this result is true even in the presence of a strong @xmath66 random anisotropy,@xcite but the @xmath66 case is more complicated than @xmath64 . for @xmath66 , as we shall see , the domain walls probably have a fractal structure , rather than becoming completely localized .
in this work , we will present results obtained from heat bath monte carlo calculations .
the data were obtained from @xmath1 simple cubic lattices with @xmath67 using periodic boundary conditions .
the calculations were done for a 12-state clock model , _
i.e. _ a @xmath68 approximation@xcite to the @xmath0 model of eqn .
the computer program was an adaptation of the code used recently for the @xmath0 model in a random field,@xcite modified to replace the random field term with the random 2-fold anisotropy term of eqn .
( 2 ) . for any integer value of the quantity
@xmath3 one can use a lookup table for the boltzmann factors , because all the energies in the problem are then expressible as sums of integers and integer multiples of @xmath69 .
the values of @xmath3 for which data were obtained are 1 , 2 , 3 , 6 , and @xmath4 .
the discretization of the phase space of the model has significant effects at very low @xmath8 , but the effects at the temperatures we study here are expected to be negligible compared to our statistical errors .
the probability distributions for the local magnetization of equilibrium states which are calculated for the @xmath68 model are found to have very small contributions from the third and higher harmonics of @xmath70 and @xmath71 .
this is strong evidence that the 12-state clock model is an accurate approximation to the @xmath0 model within our range of parameters .
the @xmath68 model shows equivalent behavior for @xmath51 and @xmath72 , unlike the @xmath73 model used earlier.@xcite the program uses two independent linear congruential pseudorandom number generators , one for choosing the values of the @xmath22 , and a different one for the monte carlo spin flips , which are performed by a single - spin - flip heat - bath algorithm .
the code was checked by setting @xmath53 , and seeing that the known behavior of the pure ferromagnetic system was reproduced correctly .
each sample was started off in a random spin state at a temperature significantly above the @xmath5 for the pure model , and cooled slowly .
thermal averages for @xmath74 were obtained at a set of temperatures spanning the critical region .
angle - averaged structure factor for @xmath75 lattices with @xmath76 at @xmath77 .
the axes are scaled logarithmically.,width=326 ] angle - averaged structure factor for @xmath75 lattices with @xmath78 at @xmath79 .
the axes are scaled logarithmically.,width=326 ] angle - averaged structure factor for @xmath75 lattices with @xmath80 at @xmath81 .
the axes are scaled logarithmically.,width=326 ] angle - averaged structure factor for @xmath75 lattices with @xmath82 at @xmath83 .
the axes are scaled logarithmically.,width=326 ] angle - averaged structure factor for @xmath75 lattices with @xmath84 at @xmath85 .
the axes are scaled logarithmically.,width=326 ] the magnetic structure factor , @xmath86 , for @xmath15 spins is @xmath87 where @xmath88 is the vector on the lattice which starts at site @xmath11 and ends at site @xmath89 . here
the angle brackets denote a thermal average . for a ram with @xmath90 , unlike the rbim , the longitudinal part of the magnetic susceptibility , @xmath91 , which is given by @xmath92 where @xmath93 , and @xmath94 . for o(2 )
spins @xmath95 \ , , \ ] ] and @xmath96 thus @xmath97 is not the same as @xmath98 , even above @xmath5 .
the scalar quantity @xmath97 is a well - behaved function of the lattice size @xmath2 for finite lattices , which approaches its large @xmath2 limit smoothly as @xmath2 increases , except possibly at a phase transition .
the vector @xmath99 , on the other hand , may not be a well - behaved function of @xmath2 for an @xmath0 model in a two - fold random field . knowing the local direction in which @xmath99 is pointing , averaged over some small part of the lattice , may not give us a strong constraint on what @xmath99 for the entire lattice will be . the critical exponent @xmath100 , is defined at @xmath101 by the small @xmath102 behavior @xmath103 where @xmath104 is some constant . for each value of @xmath24
, results for four different @xmath7 configurations of the random anisotropy @xmath22 were averaged .
the same four samples of random @xmath22 were used for all values of @xmath8 , and all values of @xmath24 , in order to facilitate the comparison of results for different values of @xmath8 and @xmath51 .
all of the data shown in these figures were obtained from monte carlo runs which used hot start initial conditions , starting at at temperature well above @xmath5 .
the value of @xmath8 was then lowered in steps .
the initial part of the run at each @xmath8 was discarded to allow the system to equilibrate . for these @xmath7
runs with @xmath3 = 1 , at each @xmath8 a sequence of spin states obtained at intervals of 20,480 monte carlo steps per spin ( mcs ) was fourier transformed and averaged .
for the larger values of @xmath51 , where the relaxation times are longer , this interval was chosen to be 102,400 mcs .
the number of these selected spin states was chosen to be 16 for each of the finite values of @xmath24 , and 32 for @xmath25 .
the fourier transformed spin state data were then binned according to the values of @xmath105 = @xmath106 , to give the angle - averaged @xmath6 .
finally , a configuration average over the four random samples was performed . both equally weighted and logarithmically weighted averages were tried .
no significant differences were found between these two types of weighting , and only the equally weighted averages will be displayed here . the results for @xmath3 = 1 , 2 , 3 , 6 , and @xmath4 are shown in figs . 1 , 2 , 3 , 4 , and 5 , respectively
. the values of @xmath8 which are used in these figures are convenient binary fraction approximations to the values of @xmath5 at these values of @xmath3 .
the best estimates of @xmath5 were determined later , by the analysis of the data over a range of @xmath2 and @xmath8 .
we see from these figures that @xmath6 is only a weakly varying function of @xmath24 , at least for @xmath7 .
the values of @xmath107 , as displayed on the figures , were found by least - squares fits to the data points for @xmath108 , where the data are well - approximated by eqn .
note that @xmath100 appears to be a slowly varying monotonic function of @xmath24 , and that the extrapolation of @xmath100 down toward @xmath51 = 0 appears to be significantly different from the value of @xmath100 found for the nonrandom @xmath15 ferromagnet.@xcite it is also interesting to note that the value of @xmath100 found for @xmath84 appears to be identical to the value of @xmath100 for the nonrandom system , but the significance of this is unclear .
the fact that @xmath100 appears to vary with @xmath24 is an indication that the claim of reed@xcite is too simplistic .
he did not calculate a numerical value for @xmath100 , but he argued that the finite - size scaling ( fss ) behavior at @xmath3 = 1 was indistinguishable from that of the nonrandom system .
one should not conclude from these data that @xmath100 is varying continuously with @xmath51 , so that there is a line of critical points .
another explanation of the data is that for any @xmath51 we have a function @xmath109 , which increases very slowly as @xmath2 increases , up to a value @xmath110 then we will only find @xmath111 when @xmath2 becomes large enough so that @xmath112 . in the cayley tree mean - field
approximation,@xcite whether @xmath113 is finite or infinite depends on the value of @xmath33 .
( color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath76 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297](color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath76 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297 ] ( color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath78 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297](color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath78 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297 ] ( color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath80 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297](color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath80 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature . the @xmath114-axis is scaled logarithmically.,title="fig:",width=297 ] ( color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath82 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297](color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath82 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297 ] ( color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath25 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297](color online ) finite - size scaling near @xmath5 for @xmath1 lattices with @xmath25 .
( a ) configuration - averaged magnetization vs. temperature .
( b ) @xmath91 vs. temperature .
the @xmath114-axis is scaled logarithmically.,title="fig:",width=297 ] if we make the assumption that the usual critical exponent scaling laws for translation invariant models remain valid for the ram , we can easily obtain values of the exponent combinations @xmath115 and @xmath116 from our computed values of @xmath107 .
these combinations are exactly what we need for fss of the magnetization @xmath117 and the magnetic susceptibility@xcite @xmath118 .
thus , by making standard fss plots,@xcite we can test the validity of these scaling laws for the ram . in fig .
6(a ) we show a fss plot of the configuration average of @xmath119 on @xmath1 lattices , for @xmath2 between 16 and 64 .
the number of sample configurations used for each @xmath120 was 8 for @xmath3 = 1 , 2 , 3 and 6 , and 16 for @xmath84 . for @xmath2 = 64
, the number of samples was 4 for all @xmath51 .
6(b ) shows a similar plot for @xmath91 .
figs . 7 , 8
, 9 and 10 show the corresponding plots for @xmath3 = 2 , 3 , 6 and @xmath4 , respectively .
since the values of @xmath100 used here were taken from the fits to the small k behavior of @xmath6 , the only two adjustable fitting parameters used in these figures were the values of @xmath121 and @xmath5 , which were required to be identical for parts ( a ) and ( b ) of each figure . in these fss plots ,
the temperature coordinate scales as @xmath122 .
the reader should note that the range of @xmath8 which we cover in these plots is about an order of magnitude larger than the range which one would typically use for a problem where one is already confident about the nature of the phase transition , and one is trying to obtain high precision estimates of @xmath5 and the critical exponents by concentrating on the range of @xmath8 where @xmath123 . as a consequence of this ,
the spacings between the values of @xmath8 for which we have taken data are rather large .
thus we are unable to use histogram reweighting@xcite to obtain essentially continuous values for the thermodynamic functions . from the results given in these figures ,
we see that the estimates of @xmath121 increase monotonically and the estimates of @xmath5 decrease monotonically as @xmath3 increases .
we also see that the peak in @xmath91 is slightly above @xmath5 for finite @xmath2 , which is typical for ferromagnetic critical behavior .
the data collapse is good near this peak , which is the range of @xmath8 for which @xmath124 .
we do not give estimates of statistical errors for @xmath121 , because we believe that the variation in @xmath121 in the range @xmath3 = 1 to 6 is due to variation in the value of @xmath125 .
we will discuss this further in the next section .
the errors in the values of @xmath5 are estimated to be less than @xmath126 .
( color online ) finite - size scaling of the difference between @xmath127= 3.00 and the configuration - averaged @xmath128 vs. temperature near @xmath5 for @xmath1 lattices with @xmath76.,width=326 ] ( color online ) finite - size scaling of the difference between @xmath127= 2.65 and the configuration - averaged @xmath128 vs. temperature near @xmath5 for @xmath1 lattices with @xmath78.,width=326 ] ( color online ) finite - size scaling of the difference between @xmath127= 2.40 and the configuration - averaged @xmath128 vs. temperature near @xmath5 for @xmath1 lattices with @xmath80.,width=326 ] ( color online ) finite - size scaling of the difference between @xmath127= 1.94 and the configuration - averaged @xmath128 vs. temperature near @xmath5 for @xmath1 lattices with @xmath82.,width=326 ] ( color online ) finite - size scaling of the difference between @xmath127= 0.81 and the configuration - averaged @xmath128 vs. temperature near @xmath5 for @xmath1 lattices with @xmath84.,width=326 ] fig .
11 shows the difference between an estimate of the specific heat at @xmath5 for an infinite system , @xmath127 , and the calculated specific heat of a finite system at temperature @xmath8 , @xmath128 for @xmath3 = 1 .
the only new adjustable fitting parameter here is @xmath127 .
12 , 13 , 14 and 15 show the corresponding plots for @xmath3 = 2 , 3 , 6 and @xmath4 , respectively .
the values of @xmath127 decrease monotonically as @xmath3 increases . in all cases , the values of @xmath127 given in the figures
are estimated to be accurate to about 1% . as we also saw for @xmath129 and @xmath91 ,
the fss data collapse is not good below @xmath5 for @xmath130 .
the results for @xmath84 are in very good agreement with the earlier results@xcite obtained with the @xmath131 approximation .
according to imry and ma@xcite and pelcovits , pytte and rudnick@xcite , for small @xmath24 this model should appear ferromagnetic when @xmath2 is smaller than the `` imry - ma length '' , which is determined by balancing the domain wall energy against the energy of random pinning .
if this length exists , when @xmath2 is larger than the imry - ma length , the system will break up into domains , without long - range order . we have argued here , however , that in the presence of random pinning one should not believe that the domain wall energy scales as @xmath46 .
one can try to patch up this picture by assuming that the domain wall energy scales as @xmath132 , with @xmath133 .
if this were the case , then it would still be possible for @xmath39 to find a length scale where the domain wall energy balanced the random pinning energy
. then it would continue to be true in @xmath39 that the system would break up into imry - ma - like domains when @xmath2 became very large . what we see in our fss plots for @xmath134
, however , is that when @xmath57 the leading correction to finite - size scaling increases the magnetization as @xmath2 increases . therefore this model appears to be stable against domain formation for @xmath134 , at least for some range of @xmath8 below @xmath5 .
the natural interpretation of this result is that @xmath135 must be less than 3/2 in @xmath39 for the @xmath15 case . fig .
12 , the fss magnetization plot for @xmath3 = 3 , shows that as @xmath2 increases the data for @xmath57 seem to be converging to a scaling function which is independent of @xmath2 . for the data in fig . 15 for @xmath3 = 6
, the data appear to be in this @xmath2-independent limit .
if we were able to do the monte carlo calculations at substantially larger values of @xmath2 , we would expect to see the same type of convergence for @xmath3 = 1 and 2 .
if we had data at such large values of @xmath2 , so that the magnetization scaling function had converged to an @xmath2-independent limit , then our estimates of the critical exponents would be expected to shift somewhat .
therefore , it is likely that @xmath136 , the true value of @xmath100 in the range of @xmath3 from 1 to 6 , is actually independent of @xmath24
. the reader must also remember that the ferromagnetic phase is allowed to be reentrant .
therefore , we do not claim that the ferromagnetic behavior which we see below @xmath5 must be stable down to @xmath8 = 0 over the entire range of @xmath24 .
also , we do not claim stable ferromagnetism for very large values of @xmath24 .
it must be stated , however , that this only applies to the simple cubic lattice with nearest neighbor interactions .
we expect that it would be possible to stabilize a ferromagnetic phase at @xmath84 by adding further neighbor finite - range exchange interactions .
we point out that our earlier claim@xcite of infinite magnetic susceptibility without ferromagnetism when @xmath84 was based on results at @xmath42.@xcite since the magnetization of finite simple cubic lattices with @xmath84 seems to be a monotonically decreasing function of @xmath8,@xcite however , we consider the existence of true ferromagnetism on this lattice to be unlikely at any @xmath8 for @xmath84 .
the author sees no reason to believe that the exponent @xmath135 should be independent of @xmath35 for @xmath39 .
thus , while we claim the existence of a ferromagnetic phase for @xmath15 , we are not making any claim here about the behavior for @xmath30 .
we do expect that @xmath135 must converge to 2 in the limit @xmath44 , in agreement with the result of larkin.@xcite the reason for this is that for @xmath44 the `` elastic membrane '' approximation becomes valid
. clearly , it would be desirable to obtain a direct estimate of @xmath135 , by , for example , calculating the change in energy of a sample between periodic and antiperiodic boundary conditions along one direction .
since the energies involved are subextensive and the domain wall energy goes to zero at @xmath5 , it is difficult to do such a calculation .
in this work we have presented monte carlo results for the @xmath39 @xmath0 random anisotropy model , eqn .
( 2 ) , for several values of the anisotropy strength @xmath24 . by studying the finite - size scaling behavior of @xmath137 simple cubic lattices over the range @xmath67
, we find that , for values of @xmath24 which are not very large , there appears to be a finite - temperature critical point at which the model undergoes a transition into a ferromagnetic phase .
for this lattice at very large @xmath24 , the transition appears to be into a phase with qlro , but no true magnetization .
r. harris , m. plischke and m. j. zuckermann , phys .
31 * , 160 ( 1973 ) .
r. fisch , phys .
b * 51 * , 11507 ( 1995 ) .
b. derrida and j. vannimenus , j. phys .
c * 13 * , 3261 ( 1980 ) .
r. fisch and a. b. harris , phys .
b * 41 * , 11305 ( 1990 ) . a. b. harris , r. g. caflisch and j. r. banavar , phys .
b , * 35 * , 4929 ( 1987 ) . c. jayaprakash and s. kirkpatrick , phys . rev .
b * 21 * , 4072 ( 1980 ) .
f. p. toldin , a. pelissetto and e. vicari , j. stat .
p06002 ( 2006 ) .
f. liers , j. lukic , e. marinari , a. pelissetto and e. vicari , phys .
b * 76 * , 174423 ( 2007 ) .
h. nishimori , prog .
. phys . * 66 * , 1169 ( 1981 ) .
m. hasenbusch , f. p. toldin , a. pelissetto and e. vicari , phys .
b * 76 * , 184202 ( 2007 ) . j. h. chen and t. c. lubensky , phys . rev .
b * 16 * , 2106 ( 1977 ) .
r. fisch , phys .
b * 41 * , 11705 ( 1990 ) .
a. aharony and e. pytte , phys .
* 45 * , 1583 ( 1980 ) .
a. aharony and e. pytte , phys .
b * 27 * , 5872 ( 1983 ) .
d. e. feldman , int .
b * 15 * , 2945 ( 2001 ) .
r. fisch , phys .
b * 58 * , 5684 ( 1998 ) . c. wang , j. harrington and j. preskill , ann . phys .
( n.y . ) * 303 * , 31 ( 2003 ) .
m. hasenbusch , f. p. toldin , a. pelissetto and e. vicari , phys .
e * 77 * , 051115 ( 2008 ) .
r. a. pelcovits , e. pytte and j. rudnick , phys . rev .
lett . * 40 * , 476 ( 1978 ) .
r. a. pelcovits , phys . rev .
b * 19 * , 465 ( 1979 ) a. i. larkin , zh .
. fiz . * 58 * , 1466 ( 1970 ) [ sov .
jetp * 31 * , 784 ( 1970 ) ] .
y. imry and s .- k .
ma , phys .
35 * , 1399 ( 1975 ) .
a. aharony , y. imry and s .- k .
ma , phys .
lett . * 36 * , 1364 ( 1976 ) .
g. parisi and n. sourlas , nucl
b206 * , 321 ( 1982 ) .
j. z. imbrie , phys .
lett . * 53 * , 1747 ( 1984 ) . j. bricmont and a. kupiainen , phys .
* 59 * , 1829 ( 1987 ) .
m. mezard and a. p.
young , europhys
. lett . * 18 * , 653 , ( 1992 ) .
m. aizenman and j. wehr , commun .
phys . * 130 * , 489 ( 1990 ) .
b. i. halperin , in _ physics of defects , les houches 1980 _ , edited by r. balian , m. kleman and j. p. poirier ( north holland , amsterdam , 1981 ) pp .
837 - 844 .
e.g. d. s. fisher , phys .
b * 31 * , 7233 ( 1985 ) .
s. n. coppersmith , phys .
b * 44 * , 2887 ( 1991 ) .
a aharony , j. phys .
c * 14 * , l841 ( 1981 ) .
a. houghton , r. d. kenway and s. c. ying , phys .
b * 23 * , 298 ( 1981 ) .
j. l. cardy and s. ostlund , phys .
b * 25 * , 6899 ( 1982 ) .
r. fisch , phys .
b * 46 * , 242 ( 1992 ) .
g. kohring , r. e. shrock and p. wills , phys .
* 57 * , 1358 ( 1986 ) .
r. fisch , phys .
b * 52 * , 12512 ( 1995 ) .
r. fisch , phys .
b * 62 * , 361 ( 2000 ) .
r. fisch , phys .
b * 55 * , 8211 ( 1997 ) .
r. fisch , phys .
b * 76 * , 214435 ( 2007 ) .
j. c. le guillou and j. zinn - justin , phys . rev .
b * 21 * , 3976 ( 1980 ) .
y .- h . li and s. teitel , phys .
b * 40 * , 9122 ( 1989 ) .
p. reed , j. phys .
a * 24 * , l117 ( 1991 ) .
m. e. fisher and m. n. barber , phys .
* 28 * , 1516 ( 1972 ) .
m. e. j. newman and g. t. barkema , _ monte carlo methods in statistical physics _ , ( oxford university press , oxford , 1999 ) , pp .
232 - 239 .
a. m. ferrenberg , d. p. landau and r. h. swendsen , phys .
e * 51 * , 5092 ( 1995 ) .
r. fisch , phys .
lett . * 66 * , 2041 ( 1991 ) . | we have performed monte carlo studies of the 3d @xmath0 model with random uniaxial anisotropy , which is a model for randomly pinned spin - density waves .
we study @xmath1 simple cubic lattices , using @xmath2 values in the range 16 to 64 , and with random anisotropy strengths of @xmath3 = 1 , 2 , 3 , 6 and @xmath4 .
there is a well - defined finite temperature critical point , @xmath5 , for each these values of @xmath3 .
we present results for the angle - averaged magnetic structure factor , @xmath6 at @xmath5 for @xmath7 .
we also use finite - size scaling analysis to study scaling functions for the critical behavior of the specific heat , the magnetization and the longitudinal magnetic susceptibility .
good data collapse of the scaling functions over a wide range of @xmath8 is seen for @xmath3 = 6 and @xmath4 . for our finite values of @xmath3
the scaled magnetization function increases with @xmath2 below @xmath5 , and appears to approach an @xmath2-independent limit for large @xmath2 .
this suggests that the system is ferromagnetic below @xmath5 . |
the next generation of low - background physics experiments , including direct detection of weakly interacting massive particle ( wimp ) dark matter may provide significant insight to physics beyond the standard model and the nature of the dominant matter constituent of the universe @xcite .
the successful detection of wimps will provide valuable information about the nature of dark matter .
several current and next generation large - scale detectors designed to search for wimp dark matter make use of liquified noble gas ( ne , ar , xe ) .
descriptions of several experiments of this type may be found in refs.@xcite .
these experiments will search for the scintillation light and in some cases the ionization charges generated from the recoiling nucleus after a putative wimp - nucleus scatter . on the other hand ,
neutrons having energies less than about 10 mev may be produced in such experiments via alpha particles produced in decays of @xmath2u and @xmath3th in the detector and surrounding materials that undergo @xmath4 reactions .
one particular site of these reactions is in borosilicate photomultiplier tube ( pmt ) glass , because the @xmath5b@xmath6n cross section is considerably larger than @xmath4 cross - section data for other typical detector and shielding materials .
pmts are a common technology for detecting scintillation light in these noble - gas detectors .
neutron - induced nuclear recoils are a particularly dangerous background for these experiments , since they can mimic wimp events . the deap / clean experimental program uses large volumes of liquid argon or neon to search for wimps from the scintillation channel only @xcite .
both materials show excellent scintillation properties and make use of pulse - shape discrimination to separate electron and nuclear recoils @xcite .
the lack of elastic scattering cross - section measurements for neutrons in argon and neon poses a problem for background estimates , performed primarily via monte carlo calculations .
precise differential neutron scattering cross sections are required to determine the background contribution due to neutrons in these experiments .
both total and differential cross sections for elastic scattering are important for dark matter experiments , as the rate of multiple scattering inside the detector s sensitive volume is determined by the angular differential cross section .
neutrons that multiple scatter can be cut using event - reconstruction techniques , but calibrations and monte carlo simulations using known cross sections are required to determine the efficacy of these analysis cuts .
we have performed measurements and optical - model fits for the neutron elastic scattering cross section for neon at @xmath7 = 5.0 and 8.0 mev and for argon at @xmath8 = 6.0 mev for use as inputs to quantitative estimates of neutron backgrounds in these experiments .
these neutron energies were determined by where the @xmath2u and @xmath3th induced @xmath9neutron yields are the largest in boron @xcite .
these measurements also provide additional data to benchmark the optical model in a nuclear mass where the model is not well constrained , and will enhance nuclear databases .
the neutron time - of - flight ( ntof ) facility at the triangle universities nuclear laboratory ( tunl ) @xcite , located on the campus of duke university , in durham , nc , usa was used to measure the neutron elastic scattering cross section of argon and neon .
the 10 mv tandem van de graaff accelerator is capable of producing pulsed , mono - energetic neutron beams from about 5 - 15 mev using the @xmath10h@xmath11he reaction @xcite .
neutrons were generated by bombarding a 3.15-cm long deuterium gas cell with a deuteron beam pulsed at 2.5 mhz and beam bunches of 2 ns full width at half maximum .
typical deuteron beam current on target for this experiment was about 1.0 @xmath12a .
the deuterium gas cell pressure ( 4.0 atm for @xmath8 = 5.0 and 6.0 mev , and 7.8 atm for @xmath8 = 8.0 mev ) was chosen to minimize the energy spread in the neutron beam caused by energy loss of the incident deuterons due to ionization in the gas while producing an adequate neutron flux on target to carry out these measurements .
the choice of gas pressure allowed a neutron energy spread of about 400 kev .
the scattering samples were suspended from thin steel wires located about 10 cm from the neutron production cell and were aligned with the beam line .
the gas target cell is described fully in ref.@xcite .
the cell was a 21.0-mm diameter stainless steel sphere with 0.5-mm wall thickness .
it was filled with 99.999% natural neon or natural argon using a coupling device connected to a high - pressure filling station at tunl .
the measurements were carried out with the cell pressurized to 170 atm absolute , which corresponds to a steel - to - gas ratio of about 4-to-1 by mass and is well below the maximum pressure rating of 550 atm .
the achievable pressure was limited by the ability to condense the gas in the cell at the filling station .
the number of nuclei in the cell was determined by measuring the mass to 0.01 mg accuracy before and after filling .
the pressure in the cell was stable , with no measurable difference during the experimental runs , sometimes up to 48 hours .
the argon gas cell was filled only once at the beginning of the experiment and the same cell was used throughout .
.description of the gas targets .
the diameter is the value of the inner diameter of the sphere . [
cols="<,^,^,^,^,^ " , ] [ tab : opt ] the best fit to the cross - section data calculated from the som potential ( eqn . [ eqn : opt ] ) is shown in figs .
[ fig : cs_argon ] [ fig : cs_2d ] . for argon ,
the total elastic scattering cross section found from the optical - model fit was 2020 mb for @xmath8 = 6.0 mev and 995 mb for @xmath8 = 14.0 mev .
the cross section was found to differ significantly from endf / b - vii.0 .
this discrepancy is largely due to an over - estimation in the radii ( @xmath13,@xmath14,@xmath15 ) in the endf / b - vii.0 parameterization .
compared to the prediction using the local koning - delaroche potential for @xmath1ar , the total elastic scattering cross section was found to differ by 8% at 6.0 mev and 3% at 14.0 mev .
+ for neon , the total elastic scattering cross section found from the optical - model fits was 1160 mb for @xmath8 = 5.0 mev and 990 mb for @xmath8 = 8.0 mev .
the cross section was found to differ significantly from the extrapolation of the dave gould potential to the @xmath16 range .
the origin of this discrepancy is largely due to the difference in the real volume term ( @xmath17 ) in the optical - model potential .
compared to the prediction using the global koning - delaroche potential for @xmath18ne , the total elastic scattering cross section was found to differ by 6% at 5.0 mev and 13% at 8.0 mev .
this is the first available data for nuclear masses from _ a _ = 17 to _ a _ = 23 .
+ + many direct dark matter searches rely on monte - carlo codes such as geant4 @xcite to simulate experiments and determine background estimates . in geant4 ,
the ` g4neutronhp ` ( high precision ) class is used to simulate neutron transport below 20 mev .
it relies on g4ndl , a library of data - driven cross sections and angular distributions , mainly from endf-6 @xcite to simulate neutron processes . in the case where no data exist for
a given isotope , ` g4neutronhp ` will replace it with the cross section information from the isotope with the nearest _ z _ and _ a _ @xcite . since cross - section information does not exist for neon in g4ndl library versions up to at least g4ndl 3.14 , ` g4neutronhp
` replaces @xmath19ne with @xmath20na , which has completely different nuclear properties .
a linear interpolation is done by ` g4neutronhp ` for the data in the g4ndl libraries .
we have created data files to add to g4ndl for the cross section and angular distributions for elastic scattering of neutrons up to 20.0 mev based on our data for argon and neon . for argon , the cross section constructed from our optical - model analysis were used from 0.65 to 20.0 mev . below 650 kev , the original cross section , based on the data of winters _ et al . _
@xcite , including the resonances , from the g4ndl 3.14 data file for @xmath1ar was used .
the original g4ndl 3.14 cross sections and the modified cross sections are shown in fig . [ fig : argon_sim_ang_compare ] for the specific neutron energies analyzed in this work .
a comparison of the total elastic scattering cross sections is shown in fig .
[ fig : argon_sim_tot_compare ] . + ar
the cross section shows significant resonance behavior below 650 kev.,scaledwidth=75.0% ] for neon , for reasons discussed above , we chose to restrict our optical - model analysis to between 3.0 and 10.0 mev .
below 3.0 mev and from 10.0 to 20.0 mev , the cross section calculated from the koning delaroche global potential @xcite was used . the original g4ndl 3.14 cross section and the modified cross section
is shown in fig .
[ fig : neon_sim_ang_compare ] for the specific neutron energies analyzed in this work .
a comparison of the total elastic scattering cross section is shown in fig .
[ fig : neon_sim_tot_compare ] . at 3.0 mev ,
the transition from the cross section using the global potential and the fits from this work is smooth .
there is a kink in the cross section at 10.0 mev , where the cross section transitions back to the calculation from the global potential .
we conclude that although the cross - section calculation from the koning delaroche global potential is better than the current @xmath20na data file used by ` g4neutronhp ` , is not adequate to describe the cross section between 10.0 and 20.0 mev .
additional measurements would be necessary to constrain our model at energies above 10.0 mev based on our data . the cross section in this energy range
do not affect dark matter background estimates from @xmath2u- and @xmath3th - induced ( @xmath9,n ) reactions because the neutron spectrum cuts off sharply at about 8 mev @xcite .
+ na and modified cross section for @xmath21ne . the solid curve is the elastic scattering cross section of neutrons from @xmath21ne added to the g4ndl 3.14 library .
the dashed curve is the original g4ndl 3.14 cross section for @xmath20na.,scaledwidth=75.0% ]
the differential elastic scattering cross section was measured for 5.0- and 8.0-mev neutrons incident on neon and for 6.0-mev neutrons incident on argon .
these data , along with the 14.0-mev data for argon from beach _ et al . _
@xcite were fit using the spherical optical model and compared to global optical - model predictions .
precise neutron scattering cross sections are required to estimate neutron backgrounds from ( @xmath9,n ) reactions in direct dark matter searches .
these cross sections , which were previously unmeasured , are an important component in background estimates of these experiments because the simulation relies on available data in the energy range below 20 mev .
our data significantly improve the current cross - section libraries in the geant4 simulation code , add to the nuclear databases , and provide accurate data for benchmarking global optical models .
this work was primarily supported by doe grants # de - fg02 - 97er4104 and # de - fg02 - 97er41033 .
we thank r. reifarth for providing us with the empty gas cells and the associated filling adapters .
we also acknowledge the support of the tunl technical staff . 10 j. r. primack , d. seckel , and b. sadoulet , ann .
nucl . part .
sci . * 38 * , 751 ( 1988 ) .
p. f. smith and j. d. lewin , phys .
reports * 187 * , 203 ( 1990 ) .
m. g. boulay and a. hime , astropart .
phys . * 25 * , 179 ( 2006 ) .
e. aprile _
et al . _ ,
lett . * 107 * , 131302 ( 2011 ) .
p. benetti _
et al . _ ,
* 28 * , 495 ( 2008 ) .
m. haraczyk _
et al . _ , acta physica polonica b * 41 * , 1441 ( 2010 ) .
m. g. boulay , a. hime , and j. lidgard , arxiv preprint arxiv : nucl - ex/0410025v1 ( 2004 ) .
m. g. boulay _
_ , j. phys .
. ser . * 136 * , 042081 ( 2008 ) .
d. n. mckinsey , nucl .
* 173 * , 152 ( 2007 )
. w. h. lippincott _
c * 78 * , 035801 ( 2008 ) .
j. a. nikkel , r. hasty , w. h. lippincott , and d. n. mckinsey , astropart .
* 29 * , 161 ( 2008 ) .
mei , c. zhang , and a. hime , nucl .
methods a * 606 * , 651 ( 2009 ) , neutronyield.usd.edu . , http://www.tunl.duke.edu/facilities/ , accessed january 26 , 2012 .
a. hutcheson _
et al . _ ,
methods b * 261 * , 369 ( 2007 ) .
et al . _ ,
methods a * 608 * , 152 ( 2009 ) . , accessed november 27 , 2012 .
v. g. j. stoks _
et al . _ ,
c * 48 * , 792 ( 1993 ) . , accessed november 2 , 2011 .
g. dietze and h. klein , technical report ptb - nd-22 ( 1982 ) .
m. anghinolfi _
et al . _ ,
methods * 165 * , 217 ( 1979 ) .
r. pedroni , ph.d .
thesis , duke university , 1986 .
w. e. kinney , nucl .
methods * 83 * , 15 ( 1970 )
. m. b. chadwick _
et al . _ ,
data sheets * 107 * , 2931 ( 2006 ) , http://www-nds.iaea.org/exfor/endf.htm .
d. e. trotter _
methods a * 599 * , 234 ( 2009 ) . g. c. wick , atti reale .
fis . mat .
* 13 * , 1203 ( 1943 )
. g. c. wick , phys
. rev . * 75 * , 1459 ( 1949 ) .
f. s. dietrich _
et al . _ , phys .
c * 68 * , 064608 ( 2003 ) .
_ , phys . rev . * 111 * , 250 ( 1958 ) .
f. vaughn _
et al . _ ,
phys . rev . * 118 * , 683 ( 1960 ) .
et al . _ ,
phys . rev . *
156 * , 1201 ( 1967 ) .
p. e. hodgson , the optical model of elastic scattering , oxford university press ( 1963 ) .
f. perey , computer code genoa - a searching procedure for global optical model parameters , oak ridge national laboratory , unpublished , 1967 .
f. guimaraes and c. fu , tng - genoa user s manual , ornl / tm-2000/252 , 2000 .
a. j. koning and j. p. delaroche , nucl .
a * 713 * , 231 ( 2003 ) .
a. koning , s. hilaire , and m. c. duijvestijn , aip conf . proc . * 769 * , 1154 ( 2005 ) , proceedings of the international conference on nuclear data for science and technology . p. e. hodgson , rep .
* 50 * , 1171 ( 1987 ) .
p. a. moldauer , phys .
c * 14 * , 764 ( 1976 ) .
p. a. moldauer , nucl .
a * 344 * , 185 ( 1980 ) .
j. h. dave and c. r. gould , phys .
c * 28 * , 2212 ( 1983 ) .
s. agostinelli _
et al . _ ,
methods a * 506 * , 250 ( 2003 ) , http://geant4.web.cern.ch / geant4/. j. allison _ et al . _ , ieee trans .
53 * , 270 ( 2006 ) .
m. herman and a. trkov , brookhaven national lab report bnl-90365 - 2009 ( 2009 ) .
e. mendoza _
et al . _ ,
iaea technical report indc(nds)-0612 ( 2012 ) , http://www-nds.iaea.org/geant4 .
r. winters _ et al .
c * 43 * , 492 ( 1991 ) .
l l l l l + + + @xmath22 & ( d@xmath23/d@xmath24)@xmath25 & @xmath26(d@xmath23/d@xmath24)@xmath25 & ( d@xmath23/d@xmath24)@xmath27 & @xmath28 + ( degrees ) & ( mb / sr ) & ( mb / sr ) & ( mb / sr ) & + 0 & 535 & 50 & 549 & 0.08 + 23 & 373 & 76 & 409 & 0.2 + 31 & 308 & 37 & 320 & 0.09 + 39 & 228 & 21 & 232 & 0.03 + 47 & 201 & 18 & 157 & 5.7 + 52 & 109 & 18 & 120 & 0.3 + 63 & 71 & 5 & 71 & 0.005 + 73 & 56 & 4 & 60 & 0.9 + 83 & 80 & 6 & 67 & 3.7 + 93 & 72 & 6 & 75 & 0.2 + 103 & 83 & 7 & 75 & 1.3 + 113 & 71 & 5 & 63 & 2.2 + 122 & 38 & 3 & 47 & 6.0 + 134 & 30 & 4 & 27 & 0.7 + 142 & 27 & 3 & 20 & 4.4 + 149 & 16 & 4 & 19 & 0.5 + 156 & 20 & 4 & 23 & 0.2 + l l l l l + + + @xmath22 & ( d@xmath23/d@xmath24)@xmath25 & @xmath26(d@xmath23/d@xmath24)@xmath25 & ( d@xmath23/d@xmath24)@xmath27 & @xmath28 + ( degrees ) & ( mb / sr ) & ( mb / sr ) & ( mb / sr ) & + 0 & 670 & 70 & 706 & 0.3 + 29 & 329 & 33 & 315 & 0.2 + 34 & 233 & 18 & 225 & 0.2 + 38 & 153 & 13 & 162 & 0.4 + 42 & 122 & 10 & 110 & 1.3 + 46 & 61 & 6 & 68 & 1.4 + 50 & 41 & 4 & 39 & 0.3 + 54 & 27 & 4 & 20 & 2.7 + 60 & 7 & 2 & 8 & 0.1 + 65 & 12 & 2 & 10 & 0.9 + 69 & 14 & 1 & 17 & 3.2 + 73 & 35 & 2 & 25 & 12.1 + 83 & 43 & 4 & 45 & 0.2 + 93 & 58 & 5 & 55 & 0.4 + 103 & 49 & 3 & 53 & 0.9 + 113 & 44 & 3 & 41 & 1.2 + 123 & 25 & 1 & 25 & 0.01 + 134 & 14 & 1 & 13 & 1.0 + 145 & 13 & 1 & 14 & 0.04 + 155 & 29 & 2 & 28 & 0.4 + l l l l l + + + @xmath29 & @xmath30 & @xmath31 & @xmath32 & @xmath33 + ( degrees ) & ( mb / sr ) & ( mb / sr ) & ( mb / sr ) & + 0 & 1650 & 165 & 1842 & 1.4 + 23 & 1354 & 75 & 1268 & 1.3 + 33 & 893 & 70 & 796 & 1.9 + 38 & 567 & 33 & 575 & 0.05 + 43 & 380 & 23 & 386 & 0.05 + 48 & 218 & 14 & 241 & 2.5 + 53 & 148 & 10 & 139 & 0.8 + 61 & 56 & 4 & 51 & 1.7 + 71 & 13 & 1 & 14 & 0.04 + 81 & 6 & 1 & 6 & 0.5 + 91 & 9 & 1 & 9 & 0.08 + 101 & 20 & 1 & 20 & 0.003 + 111 & 28 & 1 & 27 & 0.6 + 122 & 25 & 2 & 27 & 0.2 + 133 & 28 & 2 & 27 & 0.3 + 144 & 30 & 2 & 30 & 0.001 + 155 & 39 & 3 & 39 & 0.02 + l l l l l + + + @xmath29 & @xmath30 & @xmath31 & @xmath32 & @xmath33 + ( degrees ) & ( mb / sr ) & ( mb / sr ) & ( mb / sr ) & + 0 & 1880 & 190 & 1604 & 2.1 + 10 & 1378 & 66 & 1290 & 1.8 + 15 & 987 & 72 & 976 & 0.02 + 20 & 645 & 24 & 648 & 0.02 + 26 & 353 & 14 & 367 & 1.07 + 31 & 180 & 7 & 179 & 0.02 + 36 & 80 & 4 & 79 & 0.08 + 41 & 46 & 3 & 47 & 0.06 + 46 & 69 & 3 & 51 & 29.0 + 51 & 51 & 3 & 63 & 24.0 + 56 & 77 & 4 & 68 & 6.0 + 61 & 61 & 3 & 62 & 0.1 + 66 & 61 & 3 & 48 & 14.5 + 71 & 34 & 2 & 34 & 0.05 + 76 & 21 & 2 & 23 & 1.09 + 81 & 19 & 2 & 17 & 1.2 + 91 & 22 & 1 & 21 & 0.1 + 101 & 30 & 3 & 29 & 0.1 + 111 & 24 & 2 & 24 & 0.04 + 121 & 12 & 1 & 12 & 0.001 + 131 & 9.4 & 1.1 & 8 & 1.6 + 141 & 11 & 1 & 11 & 0.03 + 151 & 12 & 1 & 11 & 0.6 + 159 & 13 & 2 & 13 & 0.05 + | background : : : the most significant source of background in direct dark matter searches are neutrons that scatter elastically from nuclei in the detector s sensitive volume .
experimental data for the elastic scattering cross section of neutrons from argon and neon , which are target materials of interest to the dark matter community , were previously unavailable .
purpose : : : measure the differential cross section for elastic scattering of neutrons from argon and neon in the energy range relevant to backgrounds from ( @xmath0 ) reactions in direct dark matter searches .
method : : : cross - section data were taken at the triangle universities nuclear laboratory ( tunl ) using the neutron time - of - flight technique .
these data were fit using the spherical optical model .
results : : : the differential cross section for elastic scatting of neutrons from neon at 5.0 and 8.0 mev and argon at 6.0 mev was measured . optical - model parameters for the elastic scattering reactions were determined from the best fit to these data . the total elastic scattering cross section for neon was found to differ by 6% at 5.0 mev and 13% at 8.0 mev from global optical - model predictions . compared to a local optical - model for @xmath1ar ,
the elastic scattering cross section was found to differ from the data by 8% at 6.0 mev .
conclusions : : : these new data are important for improving monte - carlo simulations and background estimates for direct dark matter searches and for benchmarking optical models of neutron elastic scattering from these nuclei . |
a key ingredient to successful analysis of images produced with the aid of an adaptive optics ( ao ) system is the determination of the point - spread function ( psf ) .
the application dictates how precisely the psf must be determined .
for example , limiting the uncertainty in psf - fitting photometry in a crowded star field to only a few percent will demand very high accuracy in knowledge of the strehl ratio .
another example , but a case where a good reference star is less likely to be found , is the subtraction of the point - like core from the image of a quasar - host or radio galaxy @xcite . in this case
strehl ratio may be low , and the ao psf approximately gaussian . for a fixed volume ,
the peak of a 2-dimensional circular gaussian is inversely proportional to the square of its full - width at half maximum ( fwhm ) .
therefore , reducing the fwhm by , say , 25% will drive the peak up by 78% .
a 25% underestimate of the qso core fwhm will have a detrimental effect on the quality of fit . in practice ,
accuracy of only 10% to 20% in psf fwhm is sufficient for detection of the host galaxy .
the task of determining the psf is made difficult by three factors .
first , the performance of the ao system depends on the brightness of the star which it uses to measure correction - the `` guide star . ''
second , the delivered correction depends directly on quickly changing seeing conditions .
third , the ao psf is strongly field dependent ( anisoplanatic ) .
the first factor requires that scientific observations and calibration measurements employ either the same guide star or ones of similar brightness . choosing a fainter guide star for calibration may require , to preserve @xmath0 , that the ao system obtain longer integration times per measurement .
phase corrugations change continuously as the wind moves turbulence across the telescope pupil , and therefore the lower bandwidth will result in poorer correction .
the latter two factors are related to the degree of turbulence along the line of sight .
the seeing scales linearly with the inverse of the coherence length of phase distortions , @xmath1 ( fried s parameter ) , which for observations at zenith angle @xmath2 , are related to the atmospheric refractive - index structure `` constant '' , @xmath3 , and wavelength , @xmath4 , by @xmath5 where @xmath6 is height in the atmosphere ( for review papers see roddier 1981 , 1999 ; beckers 1983 ) .
thus a measurement of the psf needs to be simultaneous with the target observation or at least sample the same variation in @xmath1 .
differences in phase distortion along different lines of sight , exacerbated by a @xmath7 profile skewed to high altitude , will dramatically degrade both strehl ratio and fwhm with increasing telescope offset from the guide star .
a separation of as little as 10 is sufficient to render images of the guide star unusable as a psf estimate for the target .
the effect of tilt correlation also causes the off - axis psf to be anisotropic . for any significant
offset the amount of beam overlap between guide and psf star in the tangential direction ( perpendicular to the offset ) will be greater than for the sagittal direction ( along the offset ) .
the lower correlation of aberrations in the sagittal direction causes the psf to be elongated towards the guide star ( see e.g. mcclure et al . 1991 , voitsekhovich & bara 1999 ) .
limited ao system bandwidth may also elongate the psf in the direction of the prevailing wind .
thus the psf can depend on position angle in the frame , and this requires that the psf measurement be made near the target position as well .
figure [ figure_star_onaxis_offaxis ] illustrates the dramatic spatial dependence of the psf .
it shows natural - guide - star ao observations of two well - separated stars , taken simultaneously ( these data are discussed later , in section [ cfht_pueo ] ) .
the right - hand panel is an image of the guide star and the left - hand panel is an image of a star at an offset of 20 .
both images have a field of view ( fov ) of roughly 2 @xmath8 2 , and the direction to the guide star is indicated by the arrow in the left - hand panel
. the on - axis image has a high strehl ratio , greater than 30% , but the off - axis strehl ratio is closer to 10% .
note the degradation of the off - axis psf and its elongation towards the guide star .
if a sufficiently bright star were within a few arcseconds of the target , the task of psf measurement would be greatly simplified .
the observer could use this `` psf star '' as a nearby and simultaneous measurement of the psf during the target observation .
unfortunately , the probability that a suitably bright psf reference will be available is very small .
furthermore , some ao systems operate with imagers originally designed for uncompensated seeing , and their new optics have sacrificed field of view in order to obtain nyquist sampling .
for example , the small fov such as the 4 @xmath8 4 provided by the @xmath9 pixel keck ao near - infrared spectrometer ( nirspec ) imager makes a fortuitous observation of both target and psf star even more unlikely .
when faint extended objects are observed with natural guide - star ao , they will always be offset from the guide star .
for a large fov imager such as the canada - france - hawaii telescope ( cfht ) @xmath10 pixel infrared camera ( kir ) , the guide star is often within the 36 @xmath8 36 fov of the detector .
this large fov has the advantage of ensuring that a bright star will be available to characterize the on - axis psf .
unfortunately , target integrations are typically long enough to leave a saturated image of the star .
still , several very short calibration exposures could be interleaved with those observations .
an average of these would provide an unsaturated image of the star to characterize the on - axis psf .
but even without images of the guide star , one can still determine the on - axis psf .
consider a wave - front with corrugations consistent with a kolmogorov spectrum .
now , consider perfect correction of the wave - front phase over an unobstructed pupil .
an analytic expression can be found that gives the improvement in strehl ratio due to successive removal of an increasing number of zernike polynomials from the aberrated wave - front @xcite .
for example , removing the first 3 zernike modes ( excluding piston ) corresponds to a reduction in the residual mean square wave - front distortion by a factor of 5 . in a real ao system , these residual errors can in principle be determined from wave - front sensor measurements and thus used to predict the delivered on - axis psf .
the wavelengths of sensing and correction may also be different .
the light from the guide star is usually split ; optical light is sent to the wave - front sensor while correction is applied to the instrument light path in the near - infrared .
@xcite have exploited this and developed software that uses wave - front information to predict the on - axis psf in the near - infrared for the cfht ao system . knowing the on - axis psf
will not give the psf at the target position .
however , an estimate of the change in the psf with increasing offset can be determined theoretically . for the case of a single thin turbulence layer at height @xmath6 , infinite telescope aperture , and perfect correction of an infinite number of modes , the off - axis strehl ratio , @xmath11 ,
falls as @xcite @xmath12 \eqno ( 2)\ ] ] where @xmath13 is the offset angle on the sky , and @xmath14 is the isoplanatic angle . for @xmath15 cm and @xmath16 km
this gives @xmath17 for @xmath18-band observations at zenith .
that is , the psf at an offset of 19 will have a strehl ratio degraded by a factor of @xmath19 from the on - axis value . determining the real ao off - axis psf is more complicated .
a real @xmath7 profile will contain significant power at low altitude ( dome seeing , ground - layer ) , but the more important contribution to anisoplanicity occurs at higher altitude .
for example , measurements using scintillation detection and ranging ( scidar ) show that a single thin high - altitude layer typically dominates the nighttime free - atmosphere @xmath7 profile above mauna kea @xcite .
these observations predate the use of so - called generalized scidar , however , and thus do not give a true picture of the conditions at the ground layer .
but , evidence from generalized scidar measurements at other observatories suggest that dominance by at most a few high layers may occur elsewhere as well ( cf . klueckers et al .
equation 3 suggests that the isoplanatic angle is inversely proportional to the height of the dominant layer . in reality
, the off - axis strehl ratio will depend on the mean height of the atmospheric turbulence , @xmath20 , and on the equivalent number of corrected zernike modes , @xmath21 .
higher order modes will decorrelate more quickly than low order ones , resulting in a smaller isoplanatic angle ( see e.g. roddier 1999 ) .
that is , equation 2 should be replaced by @xcite @xmath22 , \eqno ( 4)\ ] ] with the exponent tending towards 2 for very partial compensation @xcite .
the `` instrumental '' isoplanatic angle is then given by @xmath23 where @xcite @xmath24 now appropriately modified , equations 4 and 5 will give an estimate of the isoplanatic angle and layer height from measurements of the off - axis strehl ratio . in practice , these strehl ratio measurements could be made with a calibration observation of a suitably dense star field .
unfortunately , the strehl ratio , both on - axis and off - axis , will vary with the seeing .
in fact , since the off - axis psf depends on not just the quickly varying @xmath1 , but also on an evolving @xmath3 profile , it will vary even more rapidly , and with greater amplitude @xcite .
thus , if the image is to be obtained as a mosaic , it must be completed quickly , before changes in @xmath1 or @xmath20 can introduce objectionable inhomogeneities .
this calibration measurement , plus knowledge of the on - axis psf , could be used to predict the off - axis psf as long as there is no significant variation in the height of the dominant turbulence layer(s ) or change in the equivalent number of compensated modes .
it should be stressed that the off - axis calibration image contains degenerate information about @xmath20 and @xmath21 .
furthermore , if one does not know @xmath1 during the on - axis psf measurement , its value of strehl ratio is also degenerate , but in @xmath1 and @xmath21 .
that is , as @xmath1 changes so does @xmath21 , which in turn affects the isoplanatic angle .
although the value of @xmath1 can be determined from the ao system wavefront - sensor signals , thus relieving the on - axis psf ambiguity , an independent measure of @xmath20 is more difficult to obtain .
taking scidar measurements during the ao observations could provide this information , but at substantial cost and effort .
however , even with the degeneracy one might obtain useful results .
if a uniform and low value of @xmath21 were maintained , the off - axis psf would change very gradually with offset .
figure [ figure_star_onaxis_offaxis ] illustrates the case where strehl ratios are high , and small changes in @xmath1 will dramatically affect not only the on - axis strehl ratio , but also the spatial variation in the psf .
a `` first - order '' method of determining the off - axis psf variation - one that does not account for the variation in @xmath21 - may be less successful here . in response to a need to obtain reliable off - axis psfs for the processing of faint galaxies , one of us ( es ) developed a means of providing psf information at various field positions using calibration observations of dense star fields @xcite . in this method
, the observer obtains images of a dense star field once or at most a few times each night ; the core of a globular cluster works best .
the ao system is guided on a star of similar brightness to the one used in the target observations .
frequently , the required psf is at an off - axis position outside the fov of the camera , so a mosaic of the star field is generated by repositioning the telescope at successively greater off - axis distances .
the goal is to obtain an unsaturated image of the guide star as well as a roughly uniform distribution of stars at many off - axis positions .
now , although the guide star in the star - field calibration image is chosen to be similar in brightness to the one used in the target observation , it is very unlikely that @xmath1 and , thus , the delivered psf is identical
. therefore , the full psf reconstruction technique must fold in information about the on - axis psf _ during _ the scientific observation with the star - field mosaic data .
our method to deal with time dependence is to deconvolve the image of the dense star field with the image of the guide star in that field .
we assume that the effects of field - dependent aberrations in the camera optics are small compared to anisoplanicity .
we would expect , if the full image is deconvolved with a sub - image of just the guide star until the entire flux is contained within one pixel , that re - convolution with that same sub - image will reconstitute the original image .
that is , the psf at each of the off - axis star positions will be restored .
therefore , we should be able to take the deconvolved image of the star field and convolve it _ with the image of the guide star from a target observation_. the result would be the image of the star field as if it were observed under the natural seeing conditions at the time of the target observation . in this way , the image of the deconvolved star field contains the differences between the on - axis and off - axis psfs .
it is a map of the convolution kernel necessary to restore the off - axis psfs over the field and will be referred to here as the psf kernel map .
equations 4 and 5 suggest that the accuracy of predictions with the kernel map will decrease if the values of @xmath20 and @xmath21 during construction of the mosaic differ significantly from those during the target observation .
to restate this another way , the off - axis psf can not be exactly separated into an on - axis psf which depends only on @xmath1 , and an off - axis kernel which depends only on @xmath20 .
even so , we propose that the psf kernel map is a useful approximation and that it provides a plausible method for predicting the off - axis psf .
it is not a mathematically exact method , and thus the results may depend on the deconvolution technique , number of iterations , convergence crition , among other things .
for example , deconvolution of the guide - star to a spike of one pixel in width may lead to a spatial - frequency finite - support problem .
one way to remedy this would be to convolve with the updated on - axis psf first , and then deconvolve with the guide star from the calibration field .
however , we will show that variation in the atmospheric conditions during observation of the star field can account for most of the uncertainty in the results , and thus for our demonstration the details of the deconvolution are not as important .
observation of the calibration field is difficult and time - consuming , and must be a consideration in the application of the method .
if only a small number of targets are observed , i.e. at only a few offset angles , a more practical method may be to measure the kernel with binaries of the appropriate separations .
this may also allow for the interleaving of calibration and scientific observations , which the scarcity of globular clusters makes more difficult .
we will show , however , that even a single calibration observation taken during the night is helpful .
it may be that the problem is sufficiently well constrained that if one knows the on - axis strehl and the strehl degradation at one point , a model dependent on @xmath21 and @xmath20 will be sufficient for reconstruction of the kernel map , but we leave that work for a later paper . a theoretical expression for psf anisoplanatism has been derived by @xcite , but needs an independent measurement of @xmath3 ( in their example case , from thermosonde data ) to provide the kernel . in this paper
we focus on what results can be obtained from just a calibration image .
a benefit of a semi - empirical approach is that it will apply to all ao systems .
conjugation of the ao correction to high altitude layers , such as employed in gemini altair , will reduce variation in the off - axis psf , and should improve the results from our method .
multi - conjugate ao ( mcao ) should lessen psf variation even more .
in mcao , psf uniformity is more dependent on system geometry than atmospheric conditions , and a direct measurement of the psf kernel map may prove to be a very successful tool .
this may allow field dependent deconvolution , for example .
broad - band ao observations of dense star fields are frequently made by other observers , but almost always for scientific observations of the stars in those fields and not for calibration purposes .
they are generally unusable for our analysis for one or more of the following reasons : 1 .
the image of the guide star is saturated .
the fov of the mosaic is too small or contains too few stars .
3 . typically ,
a long time elapses between the observation of the guide star and the other parts of the mosaic .
variations in seeing thus render any measurement of anisoplanicity unreliable .
the work discussed here involves the observation and analysis of dense star fields for the sole purpose of calibrating the ao off - axis psf .
the observations and data reductions are outlined in section [ observations ] .
computer simulations were used to further characterize the data and the psf prediction method was applied to them .
these analyses are found in section [ analysis ] along with a discussion of the uncertainties in the predictions .
our conclusions follow in section [ conclusions ] .
observations of dense star fields were made with two ao systems : pueo on the 3.6 m cfht on mauna kea , and the lawrence livermore national labs ( llnl ) ao system on the 3.0 m shane telescope of the lick observatory .
both are compact systems capable only of low - order correction - providing fewer than 100 actuators in their deformable mirrors .
both are mounted at the cassegrain focus .
however , these systems represent two very different ao designs .
the pueo is a curvature wavefront - sensing system using avalanche photo - diodes ( apds ) in its sensor and employing a 19-element bimorph mirror for correction @xcite .
the llnl ao system uses a ccd - based shack hartmann sensor and a piezostack mirror with 61 active elements @xcite .
the pueo and lick ao systems are also located at very different observing sites .
the cfht is at an elevation of 4200 m and benefits from the already excellent seeing of mauna kea . also , the cfht observing floor is refrigerated to obtain good dome seeing by minimizing the temperature difference between the telescope and outside air .
the shane telescope depends on passive radiation and is at an elevation of only 1350 m. as a result , the cfht provides @xmath25-band median - seeing conditions of fwhm@xmath26 , whereas at the shane telescope the median seeing is over 1 .
the image quality delivered by these systems depends on both atmospheric conditions and system performance .
guide stars must be selected to have sufficient brightness to operate the wave - front sensors . in practice ,
the llnl system can obtain diffraction - limited performance with a star as faint as @xmath27 , while the low - noise apds of the cfht system permit diffraction - limited performance for guide stars as faint as @xmath28 .
even the selection of a sufficiently bright guide star , however , does not guarantee consistent performance .
for example , the encroachment of thin cirrus clouds has been seen to temporarily attenuate guide - star signals and degrade phase measurement .
this causes poorer delivered correction .
furthermore , small changes in the ao systems themselves also have an effect . two examples
are variation in the setup and calibration of the optical bench in the afternoon before observing and flexure of the system due to gravity during the night
. it would therefore be a very difficult task to model _ in detail _ the performance of either system with a computer simulation in order to compare with observations .
a journal of the observations is given in table [ table_journal_of_observations ] .
observations were made through @xmath29 or @xmath30 filters of globular cluster fields where archival @xmath25 , @xmath31 , or @xmath32-band hubble space telescope ( hst ) wide - field planetary camera 2 ( wfpc2 ) images were already available .
the requirement of previous wfpc2 imaging was adopted for two reasons . first , to ensure that potential guide stars were not doubles or otherwise too crowded for use with the ao system .
this was impossible to determine from previous ground - based images .
second , the large 2.6 @xmath8 2.6 fov of wfpc2 helped to provide astrometry for registration of the mosaiced ao images .
the cfht observations were made during the commissioning of pueo in june 1996 .
observations of the globular cluster were made ; two observations using the same @xmath33 guide star were made during one night .
the generation of a wide - field mosaic was made difficult by the small fov ( 9 @xmath8 9 ) of the montreal near infrared camera ( monica ) @xcite .
a mosaic pattern was used to obtain a roughly 12 wide by 30 long strip with the guide - star at one end .
the mosaic was composed of 6 individual pointings of the telescope , with a total exposure time of 30 seconds at each .
these images overlapped their neighbours by a few arcseconds , which allowed us to monitor the seeing by comparing multiple observations of the same star .
each of the complete mosaics was constructed in under 10 minutes .
figure [ figure_m5 ] is an image of the resulting field .
the star used for guiding was bright , 4 magnitudes brighter than the faint limit for pueo , and sufficient for optimal performance of the system .
however , there was some cirrus present during the night , and this may have caused the delivered performance to fluctuate independent of seeing variation .
no independent measurements of the seeing , such as those provided by a dedicated seeing monitor instrument , were available at cfht during this run .
however , wave - front sensor telemetry was recorded automatically by pueo .
values of @xmath34 were estimated using the technique of @xcite by means of software provided on the pueo website .
values ranged from 20 to 30 cm , consistent with median or better seeing on mauna kea .
since only two observations were made with the cfht , they provided an insufficient dataset for measuring the variation in anisoplanicity . to measure this we needed many observations of the same cluster taken over the course of an entire night .
such observations were obtained with the lick ao system on the globular cluster , as part of a program funded by the center for adaptive optics to study ao psfs .
lick ao observations were obtained through a @xmath30 filter with the infrared camera for adaptive optics at lick ( ircal ) @xcite in september 2000 .
the same guide star was used for all star - field observations .
the small ircal fov ( 19 @xmath8 19 ) required the construction of oblong mosaics much like that obtained at cfht . in this case , three individual pointings of the telescope produced strips roughly 20 wide by 60 long .
figure [ figure_m15 ] is an image of the resulting field .
the exposure time at each pointing was 100 seconds , and each strip took approximately 9 minutes to complete .
a total of 10 strips was obtained over two nights .
the guide star was selected because it was in a particularily dense part of m 15 , but it was fainter than ideal .
it was at the faint limit for lick ao guide stars .
thus , a nearby but brighter isolated star was also observed from time to time to track optimal performance .
we interleaved the mosaics of with observations of this bright star and obtained total integrations of 50 seconds with both the ao system providing full correction and tip - tilt correction only .
no seeing monitor was available at lick .
in an effort to obtain independent measures of the seeing during our run , we obtained simultaneous images of stars with a different telescope nearby on the mountain .
a measurement obtained from within the same dome would have been ideal , but was impractical .
we utilized the nickel 1 m telescope , located roughly 0.5 km west of the shane .
unfortunately , changes in telescope focus made these data unusable .
we were still able , however , to determine @xmath1 during the run .
wavefront - sensor telemetry was recorded at intervals by the ao system and used to determined @xmath34 in a manner similar to @xcite .
the mean values were @xmath35 cm for 14 september 2000 and @xmath36 cm for the following night .
all the globular - cluster data were compared to hst wfpc2 observations of the same fields in order to determine the plate scale and orientation of the cameras .
we obtained good relative astrometry of stars , with sub - pixel accuracy .
the data were flat - fielded , and aperture photometry was carried out on all the sky - subtracted images .
strehl ratios were calculated for all the stars in the following manner : the diffraction pattern for a centrally obstructed circular aperture was scaled to the pixel sampling of the camera and to a flux of unity .
this image was multiplied by the total flux of each measured star , shifted to that star s sub - pixel coordinates , and divided by the original sky - subtracted image .
the ratio of the peaks gives the strehl ratio .
the shifting of the diffraction pattern is easily done ( in this case , with imshift in iraf ) to sub - pixel accuracy , which will limit the error due to different pixellation in the two images
. a better method may be to resample the original image by a factor of 2 or more to reduce the intra - pixel averaging , but we are mostly interested in psfs much poorer than the diffraction limit , where this effect is less severe . a further , yet systematic error in this measurement arises due to the finite size of the photometric aperture . for our photometry
this corresponds to roughly a 10% overestimate of strehl ratio , which is a consistent bias for all measurements .
the dense star fields selected provided good sampling of the spatial variation in strehl ratio .
however , photometric accuracy , and therefore , accuracy of the measurement , dropped for fainter stars and for smaller separations .
we overcame this problem by selecting from each field a sub - sample of the brightest and most isolated stars .
we maintained this selection throughout the analysis and obtained an uncertainty in strehl ratio for any given star in a single frame that was not much worse than its random photometric error , or on the order of few percent .
the mosaics ideally give an instantaneous measure of the psf over the field .
since a time lapse of several minutes occured during the observation of each mosaic , it instead also samples a temporally varying psf .
the overlapping regions provide an estimate of this variation . the strehl ratio of a particular star in both frames typically varied by 15% for cfht and 20% for lick , but some discontinuities as large as 50% occured for lick .
the mean strehl ratios in the cfht and lick data are approximately 25% and 5% respectively , which suggests the mean _ absolute _ uncertainty in the strehl ratio at any position in a mosaic is about 3% .
the correction delivered by both ao systems when using bright stars was similar . the mean on - axis strehl ratio for the cfht observations of m 5 was 0.24 .
this is comparable to the average value of 0.25 obtained for the lick observations of the bright isolated star in .
note , of course , that the cfht images were obtained in @xmath29 band while the lick images were obtained through a @xmath30 filter .
since , to first order @xmath37 , observing at redder wavelengths with lick compensated somewhat for its poorer seeing conditions .
the value of @xmath34 at cfht was roughly 25 cm , while at lick it was approximately 9 cm , or almost 3 times worse .
correcting these values of @xmath1 to the observing wavelengths gives @xmath38 cm and @xmath39 cm at cfht ( 1.6 @xmath40 ) and lick ( 2.2 @xmath40 ) respectively , indicating that the effective seeing conditions are comparable , differing by less than a factor of 2 .
we obtained an estimate of the instrumental isoplanatic angle , @xmath41 , by a least - squares fit of equation 4 to the measurements of strehl ratios for stars in the mosaics .
the values are given in table [ table_journal_of_observations ] without correction for zenith angle .
the ( cfht ) observations have a mean @xmath41(1.6 @xmath40 ) of 33 . for ( lick ) it was approximately 50 ( 2.2 @xmath40 ) on average . combining these values of @xmath41 ( after correcting for zenith angle ) with measurements of @xmath1 should yield estimates of @xmath20 via equation 5 . for the observations in the mean value of @xmath1 was approximately 25 cm , which suggests that @xmath20 was approximately 5 km at cfht .
similarily , @xmath42 cm for the observations of gives @xmath43 km at lick .
since the observations of the mosaics and the bright isolated star in the lick data were interleaved , we can get a further diagnostic of the lick ao system performance by comparing the on - axis strehl ratios of the two observations .
the difference between that obtained with full correction during the mosaics and during observations with tip - tilt only correction of the bright isolated star is a measure of how many more zernike modes were corrected in the mosaic observation .
we found that the strehl ratio improved , on average , from 0.01 to 0.05 , a factor of 5 over tip - tilt correction alone .
this suggests that we achieved correction of only a few modes beyond tip and tilt @xcite .
this is plausible because the guide star magnitude was near the useful limit for the lick system .
a goal of the observations was to relate gross changes in , say , @xmath1 or the number of corrected modes to the accuracy of the semi - empirical psf prediction .
we therefore want to , first , model the performance of the ao systems and determine what most affects their delivered isoplanatic angles .
next , we will generate the psf kernel maps .
finally , we will relate the accuracy of the predictions based on these maps to the stability of the atmospheric conditions and performance of the ao system .
a computer simulation was developed to characterize the data .
it is a simple model that includes a single atmospheric layer , a circular unobstructed pupil , and ideal wavefront sensing and compensation .
the simulation codes were written by one of us ( sh ) and are based on a simulation by @xcite . like that code ,
the wavefront corrugations are due to kolmogorov turbulence and are contained in fixed phase screens ; but in our case the model of the ao system is much less sophisticated and only a single phase screen is employed .
this phase - screen image can be scaled according to fried s parameter , @xmath1 , and the height of the atmospheric layer , @xmath6 .
the telescope pupil is then superimposed at a random position in the phase screen , and the aberrations are decomposed into zernike modes via the prescription of @xcite .
the code simulates compensation of the phase error by summing only a small subset , @xmath21 , of the measured modes , re - binning this result at lower resolution to mimic the true number of actuators in the cfht or lick system , and subtracting the result from the phase screen .
the delivered strehl ratio , @xmath11 , is calculated at several off - axis positions using the extended marechal approximation on the residual phase error in the phase screen .
the correction is done without time delay , and thus the code simulates open - loop performance of infinite bandwidth .
all the simulation steps are repeated for hundreds of iterations , and the output is the average change in strehl ratio with offset .
the values of @xmath1 , @xmath6 , and @xmath21 can then be determined by a best fit optimization to these curves . because only @xmath1 is directly specified by the observations ( see section [ combined_dataset ] ) , we kept @xmath6 and @xmath21 as free parameters in the model .
we set @xmath44 , which models perfect correction of low - order modes such as tip , tilt , defocus , astigmatism , coma , and trefoil .
figure [ figure_m5_plots ] shows plots of @xmath11 versus offset for the two observations at cfht .
the data are shown as filled circles , and the results of the models are overplotted as smooth curves .
the best - fit models have @xmath45 km .
poorer fits were obtained with higher and lower altitudes ; the results of @xmath46 km and @xmath47 km are shown .
correction was good for the first mosaic , suggesting @xmath48 , while for the second mosaic @xmath21 is closer to 9 .
for these choices of parameters , the simulations suggest a value of @xmath49 for both observations .
figure [ figure_m15_plots ] shows the results for the observations at lick .
the data are shown as filled circles .
the probable cause of discontinuties in the data for observations 3 and 10 is variation in @xmath1 during the construction of those mosaics .
several simulations are shown as smooth curves ; solid curves represent a simulation with @xmath50 , dotted curves : @xmath51 , and dot - dashed curves : @xmath52 .
three choices of layer height are plotted for each choice of number of modes corrected .
curves corresponding to @xmath6 of 1 km , 2 km , and 4 km are progressively steeper , with @xmath53 km being the steepest .
the models suggest that , for lick , a low number of modes were corrected by the system and the dominant layer was at an altitude between 1 km and 4 km .
however , the best fit model has @xmath50 for mosaic 1 , 2 , 5 , and 9 .
a value of @xmath51 is a better fit for mosaics 3 , 4 , 6 and 7 ; and mosaic 8 suggests @xmath52 . to within the errors of the data the best model - fits all have similar values of @xmath41 , tending towards 70 arcseconds .
each of the mosaiced images from both the cfht and lick datasets was deconvolved with the image of its guide star .
the lucy - richardson algorithm was employed .
other deconvolution methods such as maximum entropy were tried , with similar results .
linear deconvolution methods may prove to be faster , but were not studied .
iterations were allowed to continue until almost all the flux in the deconvolved image of the guide star was contained within a single pixel .
the following convergence criterion was adopted : iterations ceased when less than 5% of the guide star flux remained outside the one - pixel spike .
this was generally achieved after a few hundred iterations . reducing the residual to , say
, 1% was much more computationally expensive , required thousands of iterations , and did not significantly improve the final results .
it should be pointed out that no number of iterations will reduce the residual to zero .
for example , one can imagine that if the centroid of the star is not centered on a pixel , deconvolution can not leave a spike that is restricted to just one pixel . to help remedy this
, we applied sub - pixel shifts to the images so that the guide - star was centered on a pixel .
the residuals for stars at off - axis positions are all larger than that for the guide - star , and less sensitive to positional errors .
this deconvolution process generates the psf kernel maps discussed in section [ introduction ] .
the kernel changes with increasing offset from the guide star , and the degree of change indicates the severity of degradation in the off - axis psf . within a few arcseconds of the guide - star ,
the kernel can be approximated by a delta - function , but at larger offset it is more gaussian in shape .
offsets between 20 and 40 are of particular interest . despite anisoplanicity , ao still provides useful correction at these offsets .
however , these offsets are larger than the typical fov of imagers currently available with ao on large telescopes .
thus , simultaneous observation of the guide star and target is not possible , and some means of calibrating the off - axis psf is needed .
figure [ figure_m5_m15_kernel ] shows radial plots of kernels for an offset of 25 .
the kernels for the m 5 ( cfht ) data are shown on the left , and two of the ten kernels for the m 15 ( lick ) data are shown on the right . for the cfht data ,
the filled circles represent the kernel from the first mosaic , while the open circles represent the second mosaic . for the lick data ,
the filled circles represent the second mosaic and the open circles represent the eighth mosaic .
these mosaics were chosen because they represent the worst - case largest possible difference in on - axis strehl ratio .
each datum represents a pixel , and no azimuthal averaging has been done .
one - dimensional gaussian fits to the data are overplotted for comparison . for cfht ,
the fhwm of the fit is @xmath54 , while for lick it is @xmath55 .
the overlap of the filled and open circles indicates that , for both the cfht and lick kernels , the separate kernel estimates roughly agree .
the scatter in the data is due to variation in @xmath1 during the observation and to the elongation of the kernel towards the guide - star , but the true shape of the kernel is somewhat complicated .
it is not described by , say , a simple two - dimensional gaussian with a fixed ellipticity .
for example , the wings of the cfht kernel can be fit by a gaussian but the same profile is too broad for the core .
the wings of the lick kernel fall off faster than a gaussian profile . to help illustrate this ,
2-dimensional representations of the cfht and lick kernels are inset in figure [ figure_m5_m15_kernel ] .
each image is an average of the same data displayed in the plot .
note the complex asymmetric shapes of both .
we are in the process of parameterizing the variation of the kernel with field position and have collected more data with lick ao to assist in that work .
we leave that analysis for a future paper
. this paper deals with the kernel maps in an empirical way and will use them directly to predict the off - axis psf .
a goal of the observations was to determine if the reconstruction technique outlined in section [ introduction ] could measure the ao psf over a large fov and predict it under different seeing conditions .
our sample of mosaics consists of long thin strips , and therefore we can not use them to measure a change in the psf with varying position angle about the guide star .
however , a simple test can check whether a kernel map generated with different on - axis seeing conditions can be used to find the first - order radial variation in the off - axis psf .
the test is this : reproduce the star field based on a psf variation map generated from a different observation of that same star field .
the best means of demonstrating this is to reconstruct an off - axis psf with the use of only an image of the on - axis guide - star and the psf kernel map .
then we can compare this reconstructed psf with the real image of that star .
for high strehl ratios , above roughly 30% , comparison of the peak brightnesses is the proper figure of merit , because the fwhm will change little . for lower strehl ratios ,
fwhm becomes more relevant . except for the first observation ,
the strehl ratio was 12% or less for all of the strips .
thus , for the most part fwhm and not strehl ratio is a better measure of image quality for our data .
we therefore normalized the predicted psf to have the same peak intensity as the data and compared their fwhm .
our more careful discussion of errors is left for the data , where strehl ratio is 10% or lower .
figure [ figure_m5_profiles ] shows the azimuthally averaged psf of the guide star in ( cfht ) and similarily averaged psfs of stars at radial distances of approximately 15 and 25 .
the filled circles represent actual data from the second mosaic while the dashed curves represent the normalized , average psf profiles from the first mosaic .
the solid line is the psf for that location reconstructed from the psf kernel using the first mosaic data .
the dashed curves are what we would predict the psf to be at a later point in the night if all that was available was the mosaic from earlier in the night .
the model is clearly a better fit to the data than the earlier psfs even though it underpredicts the fwhm . at an offset of 25
the model predicts a fwhm 31% too small , but without the model the prediction would be in error by 120% .
figure [ figure_m15_profiles ] is similar to figure [ figure_m5_profiles ] but shows the azimuthally averaged psfs for ( lick ) . here
a kernel map generated with the eighth mosaic was used to predict the psfs during the second mosaic observation .
these mosaics were chosen because they represent the worst - case largest possible difference in on - axis strehl ratio .
the results are qualitatively similar to those for cfht , but here the predicted fwhm is broader than the data by 44% . without the model , the prediction would be an underestimate and in error by 69% .
the availability of several mosaics allows us to estimate the uncertainty in predicting the psf for the lick data .
first , each of the guide - star images was used to predict the psf at an offset of 25 , i.e. without accounting for anisoplanatism at all .
the fwhm was found to be in error by 60% on average .
the model was then applied .
each of the images was generated using each of the other kernel maps .
figure [ figure_m15_errors ] gives results of comparing the observed and predicted fwhm for a star at an offset of 25 .
the error in predicted fwhm is plotted against the difference in number of corrected modes in the mosaic used for the kernel map versus the predicted mosaic .
the number of corrected modes was assigned according to the determination in section [ computer_modeling ] .
this plot shows a clear correlation .
the trend is towards over - estimating the fwhm when a mosaic with too few corrected modes is used to predict the psf .
the mosaics with highest on - axis strehl ratios , or equivalently , the highest number of corrected modes were obtained at the lowest airmasses .
this suggests that calibration observations should be taken at equivalent or lower airmasses than target observations . following this prescription and assuming that an under - estimate and over - estimate of fwhm are equally undesirable
, we can determine the mean error in predicting the psf .
if we had simply used a star at an offset of 25 to predict the psf , the mean absolute error would be 33% .
for the model psfs it is 19% ; an improvement by a factor of 1.7 .
note that an error of 19% in predicting the psf is also consistent with the 25% uncertainty in strehl ratio for the lick mosaics .
thus the major source of error in prediction of the psf can be attributed to variation in ao correction , and thus plausibly to variation in @xmath1 , during construction of the mosaics .
but variation in @xmath1 is not the only way to account for the error .
variation in @xmath21 and not @xmath1 could arise from poorer correction due to the presence of cirrus , for example .
this highlights how difficult it is to make a meaningful empirical prediction , and that the best method would provide independent measurements of @xmath1 , @xmath20 , and @xmath21 .
the trend in figure [ figure_m15_errors ] suggests one path to decreasing the error .
the second - order effect of lower angular decorrelation with smaller @xmath21 could be measured , and an appropriate model for increasing the off - axis kernel size developed .
we have discussed observations of dense star fields for the purpose of calibrating ao off - axis psfs at cfht and lick observatory .
the results of our simulations suggest that a simple model with a single thin turbulence layer is a useful model for our data . for both the and observations , a dominant layer at between 2 and 5 km above the telescope and a low number of corrected modes
reproduces the anisoplanicity seen in the data .
a semi - empirical technique was applied that utilizes these findings to improve the prediction of ao off - axis psfs .
the results show this simple method reduces error in the prediction of the basic radial variation in the psf .
we have insufficient data to determine if it can predict the suspected variation in anisoplanicity due to position angle about the guide star .
the observations of suggest that a single calibration observation of a dense stellar field is a useful prediction of anisoplanicity over the course of a night , or even for a different night , at least for the atmospheric conditions present at lick observatory .
errors in prediction of fwhm with this method are approximately 19% compared to 33% without any on - axis information , and 60% without any anisoplanatic correction at all . for maximum accuracy
the number of corrected modes should be similar , or larger , in the calibration image .
the accuracy of determining anisoplanicity from a calibration image is limited because the effects of @xmath1 , @xmath20 , and @xmath21 are intertwined .
better results could be realized by separating these effects through independent measurement of @xmath3 .
the simulations help us understand why the method is successful , if only modestly .
the dominant - layer height evidently does not change during a night or from night to night significantly enough to dramatically alter the isoplanatic patch delivered by the system .
obtaining a calibration image that measures this anisoplanicity is , however , a difficult task .
a mosaiced image must be completed quickly , before variations in delivered strehl ratio ruin the measurement .
each of our mosaics was constructed in under 10 minutes , but variation in strehl ratio by 15% to 50% during that time ultimately limited the accuracy of the results . in september 2001
we again used the lick ao system but this time employed its laser beacon to acquire more calibration observations of dense star fields . instead of narrow strips , coverage of a large box - shaped region of sky was obtained .
this will allow us to characterize anisoplanatic behavior both radially away from and azimuthally about the guide star .
future observations would also benefit from confirmatory scidar measurements of the @xmath3 profile .
we would like to thank the staffs of the cfht and lick observatories , and for the helpful comments of the referee , olivier lai .
es acknowlegdges support from the university of victoria department of physics and astronomy .
the cfht observations and some preliminary analyses were carried out while he was a graduate student there .
portions of this work were performed under the auspices of the u.s .
department of energy by the university of california , lawrence livermore national laboratory under contract no . w-7405-eng-48 .
this work was supported by the national science foundation science and technology center for adaptive optics , managed by the university of california at santa cruz under cooperative agreement no .
ast-9876783 .
bauman , b.j . , gavel , d.t .
, waltjen , k.e . ,
freeze , g.j . , keahi , k.a . ,
kuklo , t.c . ,
lopes , s.k . , newman , m.j . , & olivier , s.s .
1999 , , 3762 , 194 beckers , j.m .
1993 , annu .
rev . astron .
astrophys . , 31 , 13 flicker , r.c . , & rigaut , f. 2002 , in preparation fried , d.l .
1982 , josa , 72 , 52 fusco , t. , conan , j .-
m . , mugnier , l.m . ,
michau , v. , & rousset , g. 2000 , , 142 , 149 hutchings , j.b . , crampton , d. , morris , s.l .
, & steinbring , e. 1998 , , 110 , 374 hutchings , j.b . , crampton , d. , morris , s.l . ,
durand , d. , & steinbring , e. 1999 , , 117 , 1109 klueckers , v.a . , wooder , n.j . , nicholls , t.w . ,
adcock , m.j . , munro , i. , & dainty , j.c .
1998 , , 130 , 141 lloyd , j.p . ,
liu , m.c . ,
macintosh , b.a . ,
severson , s.a . ,
deich , w.t . , & graham , j.r .
2000 , , 4008 , 814 mcclure , r.d . ,
arnaud , j. , fletcher , j.m . , nieto , j .- l . , racine , r. 1991 , , 103 , 570 nadeau , d. , murphy , d.c . , doyon , r. , rowlands , n. 1994 , , 106 , 909 noll , r.j . 1976 , josa , 66 , 3 , 207 racine , r. & ellerbroek , b.l .
1995 , , 2534 , 248 rigaut , f. , arsenault , r. , kerr , j.m . ,
salmon , d.a . ,
northcott , m.j . ,
dutil , y. , & boyer , c. 1994 , , 2201 , 149 rigaut , f. , salmon , d. , arsenault , r. , thomas , j. , lai , o. , rouan , d. , vran , j.p .
, gigan , p. , crampton , d. , fletcher , j.m . , stilburn , j. , boyer , c. , & jagourel , p. 1998
, , 110 , 152 rigaut , f. , & sarazin , m. 1999 , in eso / osa proc .
_ astronomy with ao : present results and future programs _ , ed .
d. bonaccini , 56 , 383 roddier , f. 1981 , in progress in optics , 19 , north - holland , ed .
e. wolf , 281 roddier , f. , northcott , m. , graves , j. , mckenna , d. , & roddier , d. 1993 , josa , 10 , 957 roddier , f. 1999 , in adaptive optics in astronomy , cambridge university press , new york , ed . f. roddier steinbring , e. 2001 , phd dissertation , univ . of victoria steinbring , e. , crampton , d. , & hutchings , j.b .
2002 , , 569 , 611 steinbring , e. , crampton , d. , & hutchings , j.b .
1998 , , 30(4 ) , 1280 vran , j .- p . , rigaut , f. , matre , h. , & rouan , d. 1997 , josa a , 14(11 ) , 3057 voitsekhovich , v.v . , & bara , s. 1999 , , 137 , 385 lrccccclc , 1 & 25/06/1996 06:06 & 10 & 30 & 1.12 & @xmath56 & 0.36 & @xmath57 & + , 2 & 07:26 & 10 & 30 & 1.05 & @xmath58 & 0.12 & @xmath59 & + psf & 14/09/2000 04:07 & 8 & 50 & 1.26 & @xmath35 & 0.19 & & 0.01 + , 1 & 04:55 & 12 & 100 & 1.15 & & & @xmath60 & + , 2 & 05:47 & 12 & 100 & 1.11 & & 0.02 & @xmath61 & + psf & 06:28 & 8 & 50 & 1.11 & & 0.22 & & 0.01 + , 3 & 07:11 & 12 & 100 & 1.14 & & 0.03 & @xmath62 & + , 4 & 08:03 & 12 & 100 & 1.24 & & 0.05 & @xmath63 & + psf & 08:42 & 8 & 50 & 1.37 & & 0.18 & & 0.01 + , 5 & 09:55 & 12 & 100 & 1.83 & & 0.04 & @xmath64 & + psf & 15/09/2000 04:18 & 8 & 50 & 1.20 & @xmath36 & 0.12 & & 0.01 + , 6 & 05:00 & 12 & 100 & 1.14 & & 0.07 & @xmath65 & + , 7 & 05:51 & 12 & 100 & 1.11 & & 0.06 & @xmath66 & + psf & 06:25 & 8 & 50 & 1.11 & & 0.31 & & 0.01 + , 8 & 07:10 & 12 & 100 & 1.14 & & 0.08 & @xmath67 & + , 9 & 08:05 & 12 & 100 & 1.26 & & 0.03 & @xmath68 & + psf & 08:43 & 8 & 50 & 1.40 & & 0.50 & & 0.03 + , 10 & 09:53 & 12 & 100 & 1.81 & & 0.06 & @xmath69 & + | even though the technology of adaptive optics ( ao ) is rapidly maturing , calibration of the resulting images remains a major challenge . the ao point - spread function ( psf ) changes quickly both in time and position on the sky . in a typical observation
the star used for guiding will be separated from the scientific target by 10 to 30 .
this is sufficient separation to render images of the guide star by themselves nearly useless in characterizing the psf at the off - axis target position .
a semi - empirical technique is described that improves the determination of the ao off - axis psf .
the method uses calibration images of dense star fields to determine the change in psf with field position .
it then uses this information to correct contemporaneous images of the guide star to produce a psf that is more accurate for both the target position and the time of a scientific observation .
we report on tests of the method using natural - guide - star ao systems on the canada - france - hawaii telescope and lick observatory shane telescope , augmented by simple atmospheric computer simulations . at 25 off - axis , predicting the psf full width at half maximum using only information about the guide star results in an error of 60% . using an image of a dense star field
lowers this error to 33% , and our method , which also folds in information about the on - axis psf , further decreases the error to 19% . |
the rvs is a slitless spectrograph whose spectral domain is 847 - 874 nm and resolving power @xmath2 .
the expected accuracy is 1 km s@xmath1 for f0 to k0 stars brighter than v@xmath3 , and for k1 to k4 stars brighter than v@xmath4 .
the main scientific objectives of rvs are the chemistry and dynamics of the milky way , the detection and characterisation of multiple systems and variable stars ( for more details , see * ? ? ?
those objectives will be achieved from a spectroscopic survey of : * radial velocities ( @xmath5 objects , v @xmath6 ) * rotational velocities ( @xmath7 objects , v @xmath8 ) * atmospheric parameters ( @xmath7 objects , v @xmath8 ) * abundances ( @xmath9 objects , v @xmath10 ) each star will be observed @xmath11 times on average by rvs over the 5 years of the mission .
because the rvs has no calibration module on board , the zero point of its radial velocities has to be determined from reference sources .
ground - based observations of a large sample of well - known , stable reference stars as well as of asteroids are thus critical for the calibration of the rvs .
a sample of 1420 candidate standard stars has been established @xcite and has to be validated by high spectral resolution observations .
two measurements per candidate star are being made before gaia is launched ( or one , depending on already available archived data ) .
another measurement will occur during the mission .
the measurements will allow us to check the temporal stability of radial velocities , and to reject any targets with significant rv variation .
the ongoing observations are performed with three high spectral resolution spectrographs : * sophie on the 1.93-m telescope at observatoire de haute - provence , * narval on the tlescope bernard lyot at observatoire pic - du - midi , * coralie on the euler swiss telescope at la silla . as of june 2011
we have observed 995 distinct candidates with sophie , coralie and narval .
the detailed observations per instrument are : * 691 stars ( 1165 velocities ) with sophie * 669 stars ( 945 velocities ) with coralie * 93 stars ( 98 velocities ) with narval figure [ chemin : fig1 ] ( left - hand panel ) represents the spatial distribution in the equatorial frame of the sample and the number of measurements per object we have done so far with the three instruments .
in addition to those new observations , we use radial velocity measurements available from the spectroscopic archives of two other high - resolution instruments : elodie , which is a former ohp spectrograph , and harps which is currently observing at the eso la silla 3.6-m telescope .
the archived data allow us to recover 1057 radial velocities for 292 stars ( elodie ) and 1289 velocities for 113 stars ( harps ) . figure [ chemin : fig1 ] ( right - hand panel ) summarizes the status of the total number of measurements for the sample of 1420 candidate stars performed with all five instruments .
we have derived the variation of radial velocity of each star for which we have at least two velocity measurements separated by an elapsed time of at least 100 days .
these stars represent a subsample of 1044 among 1420 targets .
the variation is defined as the difference between the maximum and minimum velocities , as reported in the frame of the sophie spectrograph .
its distribution is displayed in figure [ chemin : fig2 ] .
a candidate is considered as a reference star for the rvs calibration when its radial velocity does not vary by more than an adopted threshold of 300 m s@xmath1 .
such a threshold has been defined to satisfy the condition that the variation of the rv of a candidate must be well smaller than the expected rvs accuracy ( 1 km s@xmath1 at best for the brightest stars ) . as a result , we find @xmath12 of the 1044 stars exhibiting a variation larger than 300 m s@xmath1 , as derived from available measurements performed to date .
those variable stars will have to be rejected from the list of standard stars .
note that about 75% of the 1044 stars have very stable rv , at a level of variation smaller than 100 m s@xmath1 .
stability threshold.,width=377 ]
observations of asteroids are very important for the radial velocity calibration .
indeed they will be used to determine the zero - points of the rvs measured with sophie , coralie and narval ( as well as the gaia - rvs zero - point ) .
those goals will be achieved by comparing the spectroscopic rvs of asteroids from ground - based measurements with theoretical kinematical rvs from celestial mechanics .
the theoretical rvs are provided by imcce and are known with an accuracy better than 1 m s@xmath1 .
about 280 measurements of 90 asteroids have been done so far . as an illustration , figure
[ chemin : fig3 ] ( left - hand panel ) displays the residual velocity ( observed minus computed rvs ) of asteroids observed by the sophie instrument as a function of the observed rvs .
the average residual of asteroids observed with sophie is 30 m s@xmath1 and the scatter is 38 m s@xmath1 . in figure [ chemin : fig3 ]
( right - hand panel ) we also show the variation of the residual rvs with time . it nicely shows how stable the rvs are as a function of time .
the residual rvs are relatively constant within the quoted errors .
the error - bars represent the dispersion of all measurements performed at each observational run .
their amplitude is mainly related to the conditions of observations that differ from one session to another ( in particular the moonlight contamination ) .
though significant ( between 10 and 50 m s@xmath1 ) those error - bars remain smaller than our target stability criterion of 300 m s@xmath1 , which will enable us to determine correctly the rv zero point of each instrument . .
a horizontal dashed line represents the mean residual velocity of 30 m s@xmath1 .
* right panel : * comparison of residual velocities of asteroids for sophie , coralie and narval as a function of time for the various observing runs.,title="fig:",width=302 ] .
a horizontal dashed line represents the mean residual velocity of 30 m s@xmath1 .
* right panel : * comparison of residual velocities of asteroids for sophie , coralie and narval as a function of time for the various observing runs.,title="fig:",width=302 ] note also we have verified that intrinsic properties of asteroids ( e.g. their size , shape , rotation velocity , albedo , etc ... ) have negligible systematic impacts on the determination of rvs zero points for the spectrographs . from now
, observations of asteroids shall be performed with reduced moonlight contamination .
crifo f. , jasniewicz g. , soubiran c. , et al .
, 2009 , sf2a 2009 conference proceedings , 267 crifo f. , jasniewicz g. , soubiran c. , et al . , 2010 ,
a&a , 524 , a10 wilkinson m. , et al . , 2005 , mnras , 359 , 1306 | the radial velocity spectrometer ( rvs ) on board of gaia will perform a large spectroscopic survey to determine the radial velocities of some @xmath0 stars .
we present the status of ground - based observations of a sample of 1420 candidate standard stars designed to calibrate the rvs .
each candidate star has to be observed several times before gaia launch ( and at least once during the mission ) to ensure that its radial velocity remains stable during the whole mission .
observations are performed with the high - resolution spectrographs sophie , narval and coralie , completed with archival data of the elodie and harps instruments .
the analysis shows that about 7% of the current catalogue exhibits variations larger than the adopted threshold of 300 m s@xmath1 .
consequently , those stars should be rejected as reference targets , due to the expected accuracy of the gaia rvs .
emphasis is also put here on our observations of bright asteroids to calibrate the ground - based velocities by a direct comparison with celestial mechanics .
it is shown that the radial velocity zero points of sophie , narval and coralie are consistent with each other , within the uncertainties . despite some scatter ,
their temporal variations remain small with respect to our adopted stability criterion .
galaxy : kinematics and dynamics galaxy : structure stars : kinematics and dynamics minor planets , asteroids : general surveys techniques : radial velocities |
magnetic fields ( mf ) in single evolved stars are still poorly studied .
most of the g
k m giants presently been known as active are fast rotators .
( fekel&balachandran , 1993 ; huensch et al . ,
angular momentum dredge - up has been suggested to provide the fast rotation , driven by the convective zone reaching near the stellar core ( simon&drake , 1989 ) . in this way
, mfs could be generated by a new dynamo .
the new generation spectro - polarimeters like narval at tbl , pic du midi , france ( aurire , 2003 ) and the lsd technique ( donati et al . , 1997 )
are very suitable for precision detections of mfs in giants ( konstantinova - antova et al . , 2008 ; aurire et al . ,
we observed 7 fast rotating giants with narval in the period of november 2006 to april 2008 . a precise computation of the longitudinal mf @xmath2 using the first moment method ( donati et al .
, 1997 ; rees&semel , 1979 ) was carried out , and the time - behavior of the activity indicator caii k ( s index ) was studied .
mf structures , indicative of a dynamo , were detected for 6 of the 7 giants ( exception : hd233517 ) , see ( incl .
s index and @xmath2 ) table 1 .
we determined the evolutionary status and masses of these 6 giants , using hipparcos parallaxes , @xmath3 from the wright catalogue ( 2003 ) , and matching evolutionary tracks from schrder et al .
( 1997 ) , see figure [ fig : hr ] .
all 6 giants with mf - detection have masses @xmath41.5 @xmath5 and their convective zones are currently deepening while the stars are evolving .
their evolutionary stages reach from the hertzsprung gap to the agb .
while the particular reasons for the fast rotation could be different , depending on mass and evolutionary history , an @xmath0@xmath1 dynamo presents a likely reason for the detected magnetic activity . | red giants offer a good opportunity to study the interplay of magnetic fields and stellar evolution . using the spectro - polarimeter narval of the telescope bernard lyot ( tbl ) , pic du midi , france and the lsd technique we began a survey of magnetic fields in single g - k - m giants .
early results include 6 mf - detections with fast rotating giants , and for the first time a magnetic field was detected directly in an evolved m - giant : ek boo .
our results could be explained in the terms of @xmath0@xmath1 dynamo operating in these giants . |
an approximate census of massive galaxies at redshift @xmath1 provides a morphological classification into 40% disks , 15% extended spheroids , 25% compact spheroids and 20% peculiar objects ( e.g. * ? ? ?
more than 9 gyrs later , in the present - day universe , a similar census of massive galaxies gives 30% disks , 65% extended spheroids and 5% peculiar morphologies .
strikingly , compact spheroids are virtually absent in the present - day universe ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . at @xmath1 , the majority ( @xmath280 % ) of massive compact spheroids
were already quiescent , while the rest were actively forming stars .
the massive high - redshift quiescent compact galaxies have been termed _
red nuggets _
recent studies on the number density evolution of red nuggets point to the star forming compact massive galaxies ( scmg ) as their probable progenitors ( e.g. * ? ? ?
* ; * ? ? ?
* ) . at redshift @xmath3 ,
the number density of scmgs matched that of red nuggets .
by @xmath1 , 1.5 gyr later , the number density of red nuggets has increased by an order of magnitude while that of scmgs has decreased by a factor of two .
the straightforward interpretation is that the red nuggets are the descendants of rapidly quenched scmgs .
but unlike this relative consensus on the origin of red nuggets , there is not the same agreement regarding their fate after @xmath1 .
red nuggets were discovered almost a decade ago ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and more than a hundred studies have been carried out on their evolution . since their discovery , red nuggets have been linked to present - day elliptical galaxies ; the two classes share a similar spheroidal morphology , and the fraction of elliptical galaxies ( extended spheroids ) has increased by a factor of four in the last 9 gyrs @xcite .
but for the former population to have evolved into the latter , there must be a mechanism by which an ultra - compact spheroid with half - light radius of @xmath21 kpc can become a fully - fledged elliptical , three to six times larger .
several physical processes , including major and minor mergers ( e.g. * ? ? ?
* ; * ? ? ?
* ) and puffing - up by agn activity @xcite , have been proposed to explain the considerable growth in size .
the prevailing paradigm , put forward by @xcite and @xcite , postulates an appealing scenario in which red nuggets ended up in the centres of present - day giant ellipticals .
this proposal is consistent with the inside - out growth scenario ( e.g. * ? ? ?
* ) in which compact cores are formed at high redshift ( @xmath4 ) through highly dissipative processes ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and then grow an extended stellar halo through dissipation - less minor mergers ( e.g. * ? ? ?
* ; * ? ? ?
in addition to this traditional evolutionary scenario , there is another potentially important channel leading from red nuggets to disk galaxy bulges .
the main observational evidences for this scenario have been provided by @xcite , @xcite and @xcite .
@xcite sample includes the set of 19 red nuggets observed by @xcite and a catalogue of few hundred bulges compiled by @xcite . in two subsequent studies , @xcite and @xcite carried out detailed bulge+disk ( b+d ) decompositions , but only for a small sample of present - day s0 galaxies , finding that s0 bulges are structurally similar to red nuggets .
figure 1 of @xcite presents an instructive sketch of mass - size and mass - density diagrams for all the spheroidal families , from giant elliptical galaxies to globular clusters . in the accompanying text
, the author emphasises the considerable overlap of the structural properties of red nuggets with those of present - day bulges .
based on cosmological simulations , @xcite and @xcite show that extended star - forming discs develop around red nuggets after compaction . the aim of the present study is to carry out a comprehensive analysis of the connection between red nuggets and present - day compact cores without any restriction on the morphology of the host galaxies .
specifically , the hypothesis we want to test is that a significant fraction of red nuggets have _ survived _ , masked inside spheroidal or disc galaxies in the present - day universe . to support this hypothesis
, we would need to find a sufficient population of compact cores embedded inside present - day galaxies , where : * _ core _ is a generalization of the bulge component in b+d decomposition of galaxies with all morphologies . * _ compact core _ is a core that fulfills the same compactness criterion used for red nuggets .
we test the robustness of our results using three different definition of compactness , from the literature . *
_ sufficient _ means that the abundance of present - day red nugget - like cores should be approaching 20 % of the red nuggets to have been consumed or destroyed in merger events , so the abundances should not necessarily match exactly . ] that of the red nuggets at @xmath1 .
the present study takes advantage of sdss - based catalogues containing b+d decomposition of hundreds of thousands of galaxies ( e.g. * ? ? ?
* ; * ? ? ?
in particular , we have used the @xcite catalogue , itself a derivation of @xcite , which provides stellar masses for the b and d structural components .
we impose the b+d decomposition on the entire present - day galaxy sample , regardless of their morphology , designating as _ cores _ both the central components of ellipticals and the bulges of disc galaxies .
after b+d decomposition , cores are treated as independent structures and evaluated for their compactness , according to standard criteria ( see section 2 ) .
the layout of the paper is the following : section 2 is devoted to the data description , including the selection of the z@xmath20 compact core sample , reference elliptical galaxies and the compilation of red nuggets from the literature .
section 3 is devoted to the results of the present study , in particular the check of the working hypothesis , i.e. that red nuggets have survived as the compact cores of present - day galaxies .
discussion and conclusions follow in sections 4 and 5 .
decimal logarithms are used along the study . to keep consistency with other studies
, we assume a @xmath5cdm cosmology with @xmath6 = 0.7 , @xmath7 = 0.3 and @xmath8 = 70 km s@xmath9 mpc@xmath9 .
several samples are extracted from the parent catalogue @xcite .
first we introduce the compactness criteria , which allow the selection of structurally similar cores and red nuggets .
second , low - redshift samples , namely the compact cores and reference ellipticals , are introduced and third , the high - redshift red nugget sample is presented . our goal is to measure the abundance of [email protected] galaxy cores structurally similar to red nuggets , and compare it to that of red nuggets at [email protected] .
therefore , information on the red nugget number density as a function of redshift becomes crucial for our work .
three different studies have addressed this issue in the literature , based on different criteria to define the red nuggets .
as emphasised by @xcite , the galaxy number densities and the structural parameters are rather sensitive to the choice of the compactness criterion , as well as to the stellar mass range probed .
consequently , we have to assure that each comparison is carried out under the same selection criteria . in the present study ,
circularised half - light radii @xmath10 have been used , with @xmath11 where @xmath12 is the half - light radius along the major - axis and ( b / a ) is the axis ratio of the galaxy . in order to test the robustness of our results ,
we compute the number density of red nuggets using three alternative definitions of compactness : @xmath13 the criterion proposed by @xcite ( hereinafter barro13 ) , used by several authors ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , @xmath14 this is generally associated with a mass limit @xmath15 .
@xmath13 a more restrictive compactness criterion , proposed by @xcite ( hereinafter vdokkum15 ) , @xmath16 with a @xmath17 limit .
@xmath13 finally , the criterion proposed by @xcite ( hereinafter vdwel14 ) @xmath18 with a @xmath19 limit .
it is worth mentioning that this criterion uses @xmath12 , the half - light radius measured along the galaxy major axis , instead of the more common circularized radius , @xmath10 .
while we test our results on the abundances of red nuggets and compact cores using all three different sets of selection criteria ( [ sec : numberdens ] ) , for the first part of this paper ( [ sec : photo ] - [ sec : dyn ] ) we adopt the vdokkum15 definition .
in fact , the definition of vdwel14 would not allow us to perform a direct comparison to other spheroidal systems ( [ sec : photo ] ) as it is based on major - axis effective radii .
on the other hand , the lower mass cutoff of barro13 would imply a significantly higher contamination from star - forming galaxies in our sample of red nuggets , with respect to the vdokkum15 higher mass cutoff ( see * ? ? ? * ) .
the parent sample used in our study is the @xcite catalogue ( hereafter mendel14 ) , an upgraded section of the @xcite catalogue ( simard11 ) , which contains the b+d decomposition of more than one million sdss - dr7 galaxies .
the main improvement to the original catalogue is the stellar mass calculation of the bulge and disk components .
stellar masses are computed through spectral energy distribution ( sed)-fitting to the five sdss-@xmath20 bands , using dusty synthetic templates from @xcite with a @xcite stellar initial mass function ( imf ) .
the mendel14 catalogue includes 657,996 galaxies from the sdss - dr7 @xcite with spectroscopic redshifts and successful @xmath20 b+d decomposition available .
the original 2d decomposition in simard11 used three alternative galaxy fitting models : 1 .
de vaucouleurs bulge ( fixed sersic index @xmath21 = 4 ) + exponential disk ( sersic index @xmath22 = 1 ) , 2 .
exponential disk + free-@xmath23 , with @xmath23 allowed to vary from 0.5 to 8 and 3 .
a single component free @xmath24-sersic model to fit the whole galaxy ( @xmath24-sersic ) .
it is important to notice that mendel14 stellar masses were computed using only two of those fitting models : ( ii ) and ( iii ) .
the bias introduced by the use of ( i ) instead of ( ii ) is discussed in appendix [ app : bias ] .
two low - redshift subsamples are extracted from the mendel14 catalogue .
the main one is the _ compact core _
subsample ( [ sec : compl ] ) , including galaxy cores that fulfill the vdokkum15 compactness criterion .
the second subsample consists of bona - fide ellipticals ( [ sec : ellipt ] ) , and serves as a reference point for the study of galaxy structural properties ( [ sec : photo ] ) .
bulge - disk decomposition is a delicate procedure demanding extra care to guarantee reliable outcomes .
we have applied a rigorous screening , with eight steps , generally following the recommendations of the catalogue authors ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we select compact cores from the mendel14 catalogue based on different criteria .
this is summarized in table [ tab : subs ] , where we also report the number of objects surviving each selection step .
first , we adopt a surface brightness limit of @xmath25 mag arcsec@xmath26 to assure a complete sample . although the formal surface brightness limit of the sdss spectroscopic sample is @xmath27 mag arcsec@xmath26 @xcite , the sample in the range @xmath28 23.0 - 24.5 mag arcsec@xmath26 is not complete because galaxies were targeted only under some particular local sky conditions .
second , we exclude objects with uncertain decomposition , using the parameter _ dbd _ , which is the offset between two alternative galaxy mass calculations , @xmath29 + @xmath30 ( sum of independently derived bulge and disk masses ) and @xmath31 ( mass derived from the total bulge+disk photometry ) in units of the standard error .
significant inequality between these two masses is an indication of inconsistent decomposition , so systems with the cutoff @xmath32 , proposed by mendel14 , are excluded .
these two cuts removed 9.5 % of the objects , as seen in table [ tab : subs ] ) .
three compactness criteria , barro13 , vdokkum15 and vdwel14 ( described in the text ) , are overplotted and compact cores are located below the dashed lines , rendering 65,637 systems under the barro13 criterion .
the more restrictive vdokkum15 and vdwel14 criteria render 10,566 and 12,721 respectively .
note that the vdwel14 line is only approximate , because the criterion requires major - axis sizes instead of the circularized radii used in the plot .
the isolated white dot marks the approximate location of the milky way s bulge ( e.g. di matteo et al . 2015 ; licquia & newman , 2015 ) . ] the third cutoff , limiting the redshift range , is set to achieve a significant sample completeness in terms of the core stellar mass , @xmath33 , the main reference parameter of this study .
the choice of redshift range , @xmath34 , yields a significant completeness in our target mass range , @xmath35 , as discussed in appendix [ app : compl ] .
notice that in the number density estimations ( [ sec : numberdens ] ) , the v / vmax statistic is also used to correct for any residual incompleteness . in the fourth step ,
we exclude galaxies reported to show anomalous b+d decomposition , tagged ` type-4 ' in the mendel14 classification .
they comprise galaxy / star superposition and inverted b+d profiles , in which a disk dominates at the center and a bulge at the external parts .
following @xcite , fifth and sixth cutoffs were applied to promote face - on inclinations and to avoid strong bars .
face - on galaxies ( disk inclination @xmath36 60 degrees ) are favoured in order to minimize internal disk extinction on the bulge light , while high bulge ellipticities are avoided to prevent strong bars being mistaken for bulges ( e@xmath37 0.6 ) . finally , the potential impact of poorly resolved galaxy images on the b+d decomposition is considered .
we follow the study by @xcite , which showed that the structural properties of cores can be reliably retrieved provided that their effective radii are larger than @xmath2 80 per cent of the psf half - width - at - half - maximum ( hwhm ) , i.e. ( @xmath38/psf ) @xmath39 0.8 . so far , by imposing these seven selection criteria , the parent sample has been reduced to 203,369 objects ( 31% of the original sample ; see tab . [
tab : subs ] ) .
the last selection is on the cores that fulfill the vdokkum15 compactness criterion ( equation [ eq : vd15 ] ) .
additionally , table [ tab : subs ] shows the corresponding results for the other criteria ( equations [ eq : barro13 ] and [ eq : vdw14 ] ) .
figure [ fig : compactness ] provides a visual display of the compact core selection criteria for both barro13 and vdokkum15 on top of a 2d histogram including the 203,369 objects , which got through the first seven screening steps .
the _ compact core _ definition is essential to the present study and deserves further clarification .
the b+d decomposition is straightforward for disk galaxies , where a compact core is simply a bulge that has satisfied the compactness criterion . a decomposed elliptical galaxy
should not be interpreted literally as having physically meaningful bulge and a disk .
in addition to the central _ core _ , the external component can be a disk , a halo or any significant departure from a single fit .
distinct inner components have been extensively observed in ellipticals ( e.g. * ? ? ?
* ; * ? ? ?
more recently , @xcite have shown that more than 75 % of their sample of 94 nearby elliptical galaxies were better described when additional subcomponents were added to the conventional single sersic function .
.subsample selection , listing the steps leading from the parent sample ( mendel14 ) to our _ compact core _
steps are described in the main text . [ cols="<,>",options="header " , ]
our goal here is to assess the morphological variety of the fully - fledged galaxies that host compact cores ( rather than analyzing the compact core structure itself as in
[ sec : photo ] ) .
figure [ fig : morphology ] and table [ tab : morphotype ] show these results , with mass estimates , @xmath40 , also being for the the total host galaxy ( not the core mass ) .
figure [ fig : morphology ] shows the variation along @xmath40 of the four morphology class fractions . aside from the scd class , which contributes a negligible amount ,
the other classes are identified by coloured bands . inside each colour band ,
lines and symbols identify the results of each compactness criterion .
only the barro13 results span the mass range down to the @xmath41 level .
the histograms , in the lower part , show the mass range covered by each compactness criterion .
note that very few galaxies , hosting compact cores , populate the @xmath42 bin ( 19 galaxies for vdokkum15 and 40 for vdwel14 ) and their presence is imperceptible in the histogram .
nevertheless , the morphology fraction is calculated for those galaxies .
table [ tab : morphotype ] gives for morphology fractions integrated across all masses ( hence less detailed than the mass - dependant results shown in figure [ fig : morphology ] ) .
using the vdokkum15 sample in table [ tab : subs ] , all host galaxies with compact cores of log([email protected] ( 10,566 ) are integrated into four morphology fraction values .
host galaxy morphology essentially divides into a 50:50 mixture of elliptical and disk morphology ( note that the majority of the sample is in early - type systems : e + s0 ) .
an alternative option to figure [ fig : morphology ] , using m@xmath44 masses is considered in appendix [ app : morpho ] .
according to figure [ fig : morphology ] , s0 and sab classes dominate the morphologies of galaxies hosting compact cores , for log([email protected] . at the high - mass end ,
elliptical galaxies predominate over their disk counterparts .
integrated values in table [ tab : morphotype ] give more balanced results with comparable fractions of elliptical and s0 classes and minor contributions from later - type disk classes .
a further simple computation can provide an estimation of the average extra stellar mass incorporated on top of the naked red nuggets to become fully - fledged present - day galaxies .
assuming that the red nuggets survive intact since redshift @xmath46 , the extra mass acquired during the last @xmath2 9.2 gyr equates to the difference between the total and the core masses , roughly equal to the b+d decomposed disk - mass ( md ) in mendel14 .
the average extra mass for the vdokkum15 selected sample is @xmath47 = 10.53 , which means that an average [email protected] % of the total galaxy mass at [email protected] has been accreted since [email protected] at a growth rate @xmath49 .
these results are consistent with those of @xcite , who estimate that @xmath50 of the total stellar mass at @xmath51 would have been added through mergers since @xmath52 under this hypothesis .
there is even better agreement with @xcite , who report a 33 % mass increase from [email protected] to [email protected] .
a few recent studies have offered interesting explanations for the early red nugget density evolution at redshifts larger than @xmath52 .
@xcite offer a convincing explanation for the number density increase of red nuggets earlier than @xmath52 , built on the simultaneous decline of the number density of star forming compact massive galaxies ( scmg ) .
their result has been corroborated by @xcite .
this interpretation suggests that the scmg are being continuously converted into red nuggets by processes quenching their star formation , while new scmgs are replenished by cold flows , until that process becomes inefficient by @xmath53 .
oddly , red nugget evolution below @xmath1 is characterised by observational disagreement and the lack of unanimous interpretation .
for instance , our relatively high number density of @xmath0 compact cores , @xmath54 , ( see table [ tab : numberdens ] and figure [ fig : numberdensity ] ) come into apparent conflict with the majority of the results from the literature , which range from lower ( @xmath55 mpc@xmath56 , * ? ? ? * ; * ? ? ? * ) to higher number densities ( @xmath57 mpc@xmath56 , * ? ? ? * ; * ? ? ? * ) .
this conflict is only _ apparent _ because the discrepancy results from the adoption of different search strategies , looking for either compact full galaxies ( literature ) or compact cores ( this study ) .
contrary to compact cores , which are relatively abundant , very few full galaxies satisfy the demanding compactness criteria in their own right . in order to check this statement
we have used our own data to calculate the number density of full galaxies satisfying the vdokkum15 compactness criterion . by using the single - sersic fit version of the mass - size parameters from the mendel14 catalogue , we obtain a number density of @xmath58 , an order of magnitude lower than the corresponding compact core number density ( as shown in figure [ fig : numberdensity ] ) . according to the hypothesis we are testing here , red nuggets - once formed - do not significantly change their internal structure or composition , but simply aggregate 2d or 3d envelopes to become the compact massive cores of present - day galaxies . under this hypothesis ,
the number density of @xmath0 compact cores should approximately match , or be slightly less than , the maximum value attained by the red nuggets .
we have now shown that this is the case ( see figure [ fig : numberdensity ] and table [ tab : numberdens ] ) .
this result is in agreement with a recent study by @xcite , their sample of 21 structurally decomposed lenticular galaxies also having a relatively high number density , comparable to our own lower - limits .
comparison with full galaxies at @xmath0 , rather than with compact cores , leads to a discrepancy , as shown here and in the literature .
this is in no conflict with the above hypothesis which asserts only that the cores are sufficient not necessary building blocks for @xmath59 galaxies .
in this study , we have tested the hypothesis that high - redshift compact , quiescent and massive galaxies , nicknamed _
red nuggets _ , have survived as the compact cores of massive present - day elliptical or disk galaxies .
two main questions are posed to test this hypothesis : 1 .
do compact cores , structurally similar to red nuggets , exist at @xmath0 ?
2 . if so , is their abundance comparable to that of red nuggets at @xmath1 ?
we carry out this test using the recently published sdss - based large catalogues of bulge+disk ( b+d ) galaxy decompositions , in particular from @xcite , which includes stellar masses of the b and d structural components .
we approach the test from a comprehensive perspective , imposing the b+d decomposition to the whole present - day galaxy sample , regardless of their morphology types .
we adopt the generic name _ core _ to include the central component of ellipticals and bulge of disks , and study these as if they were independent structures , defined based on their compactness , according to same criteria adopted in the literature to identify the red nuggets themselves ( e.g. * ? ? ?
? * ; * ? ? ?
it is worth mentioning that our results depend heavily on the reliability of the photometric bulge+disk decomposition of sdss galaxy data .
nevertheless , the huge size of the parent catalogue has allowed us to carry out a very strict quality screening , only passed by @xmath60 % of the objects . in summary
, the study concludes that the answer to both these questions above is _
yes_. the details of this conclusion and other key results are as follows : @xmath13 we confirm that a small but significant fraction ( @xmath615% ) of cores in present - day galaxies ( regardless of morphology ) is structurally similar to [email protected] red nuggets . in the mass - size and mass - density relations ,
the red nuggets and compact cores are significantly segregated from the normal elliptical galaxies , with average sizes @xmath24 times smaller and densities @xmath220 times larger than normal ellipticals at a given mass ( e.g. @xmath62 ) ( figure [ fig : spheroids ] ) .
@xmath13 not only do such compact cores exist , but their abundance matches approximately that of the red nuggets at @xmath1 ( e.g. * ? ? ?
this result is quite consistent with previous findings on the present - day abundance of compact galaxies , as we use of b+d decomposed cores ( rather than to full galaxies ) .
@xmath13 that these number densities are , if anything , slightly lower than the highest values for red nuggets , further agrees with the hypothesis , leaving room for the some @xmath220% of red nuggets to have been destroyed in major merger events ( as found in some cosmological simulations ( e.g. * ? ? ?
* ; * ? ? ?
@xmath13 dynamical measurements of the samples of red nuggets , compact core and ellipticals corroborates the photometric analysis .
velocity dispersion , @xmath63 , and dynamical mass density in red nuggets and compact cores are systematically larger than in normal ellipticals .
for instance at a given @xmath64 , @xmath63 is 1.5 - 1.7 times larger and dynamical - mass density 11 - 35 times higher than in the normal ellipticals ( figures [ fig : veldisp ] and [ fig : mdyn ] ) .
@xmath13 a simple calculation allows us to estimate the extra mass accreted by the red nuggets to become fully - fledged galaxies , under this hypothesis . assuming that red nuggets start ` naked ' or with only rudimentary disks at @xmath1 , the extra stellar mass acquired over the last @xmath65gyr is the difference between the mass of the galaxy and that of the compact core .
we find that galaxies hosting a compact core would typically have accreted one quarter of their mass over the last @xmath65gyr , at a growth rate of @xmath66 .
@xmath13 galaxies hosting compact cores ( according to the vdokkum15 criterion ) show varied morphological classes extending from ellipticals to spirals ( scd ) .
the integrated fractions of morphology classes with @xmath6710.5 are : 50.9% ( e ) , 43.4% ( s0 ) , 5.4% ( sab ) and 0.3% ( scd ) .
results for alternative mass ranges and compactness criteria are given in figure [ fig : morphology ] and appendix [ app : morpho ] .
these findings are consistent with red nuggets becoming cores of both present - day elliptical and disk galaxies .
we thank the referee for a very constructive report , which improved the paper .
we also thank ignacio trujillo , mara cebrin and the galaxy group at the iac for enlightening discussions .
we acknowledge the use of sdss data ( http://www.sdss.org/collaboration/credits.html ) , and support from grants aya2013 - 48226-c3 - 1-p and aya2013 - 47742-c04 - 02-p from the spanish ministry of economy and competitiveness ( mineco ) . if thanks the iac for hospitality under the severo ochoa visitor programme . ms and cdv acknowledge financial support from the mineco under the 2011 severo ochoa program sev-2011 - 0187 .
cdv , ms and im - v acknowledge financial support from mineco under grants aya2013 - 46886-p and aya2014 - 58308-p . 99 abazajian , k. n. , adelman - mccarthy , j. k. , ageros , m. a. et al . , 2009 , apjs , 182 , 543 baldry , i. k. , glazebrook , k. , 2003 , apj , 593 , 258 baldwin , j. a. , phillips , m. m. , terlevich , r. , 1981 , pasp , 93 , 5 barro , g. , faber , s.m . ,
p ' erez - gonz ' alez , p.g .
2013 , apj , 765 , 104 barro , g. , faber , s.m . , dekel , a. et al .
2015 , apj submitted ( arxiv.1503.07164 ) belli , s. , newman , a. b. , ellis , r. s. , konidaris , n. p. , 2014a
, apj , 788 , l29 belli , s. , newman , a. b. , ellis , r. s. , 2014b , apj , 783 , 117 berg , t. a. m. , simard , l. , mendel t. j. , ellison , s. l. , 2014 , mnras , 440 , l66 bernardi , m. , shankar , f. , hyde , j. b. , mei , s. , marulli , f. , sheth , r. k. , 2010 , mnras , 404 , 2087 bezanson , r. , van dokkum , p. g. , tal , t. , marchesini , d. , kriek , m. , franx , m. , coppi , p. , 2009
, apj , 697 , 1290 bezanson , r. , van dokkum , p. , van de sande , j. , franx , m. , kriek , m. , 2013 , apj , 764 , l8 bezanson , r. , franx , m. , van dokkum , p. , 2015 ,
apj , 799 , 148 bluck , a.f.l . , mendel , j.t . ,
ellison , s.l .
2014 , mnras , 441 , 599 buitrago , f. , trujillo , i. , conselice , c. j. , hussler , b. , 2013 , mnras , 428 , 1460 cappellari , m. , bacon , r. , bureau , m. et al . , 2006 , mnras , 366 , 1126 cappellari , m. , di serego alighieri , s. , cimatti , a. et al . 2009 , apj , 704 , l34 cardelli , j.a . ,
clayton , g.c . ,
mathis , j.s .
, 1989 , apj , 345 , 245 carrasco , e. r. , conselice , c. j. , trujillo , i. , 2010 , mnras , 405 , 2253 cassata , p. , giavalisco , m. , guo , y. et al .
2010 , apj , 714 , l79 chabrier , g. , 2003 , apj , 586 , l133 cid fernandes , r. , mateus , a. , sodr ' e , l. , stasinska , g. , gomes , j.m . , 2005 , mnras , 358 , 363 cimatti , a. , cassata , p. , pozzetti , l. et al . , 2008 , , 482 , 21 ciotti , l. , 1999 , , 249 , 99 conroy , c. , gunn , j. e. , white , m. , 2009 , apj , 699 , 486 ct , p. , piatek , s. , ferrarese , l. et al . , 2006 , apjs , 165 , 57 daddi , e. , renzini , a. , pirzkal , n. et al . , 2005 , apj , 626 , 680 damjanov , i. , mccarthy , p.j . , abraham , r.g .
et al . , 2009 , apj , 695 , 101 damjanov , i. , abraham , r.g . ,
glazebrook , k. et al . , 2011 ,
apj , 739 , l44 damjanov , i. , geller , m.j .
, zahid , h.j . , hwang , h.s .
, 2015 , apj , 806 , 158 dekel , a. , birnboim , y. , engel , g. et al .
2009 , nature , 457 , 451 di matteo , p. , g ' omez , a. , haywood , m. et al .
2015 , , 577 , 1 dullo , b. , graham , a.w .
, 2013 , apj , 768 , 36 fan , l. , lapi , a. , de zotti , g. , danese , l. , 2008 , apj , 689 , l101 ferreras , i. , trujillo , i. , m ' armol - queralt ' o , e. et al . , 2014 , mnras , 444 , 906 furlong , m. , bower , r.g .
, crain , r.a .
et al . , 2015 ,
preprint ( arxiv.1510.05645 ) gadotti , d.a .
, 2008 , mnras , 384 , 420 graham , a.w .
, 2013 , planets , stars and stellar systems . vol . 6 : extragalactic astronomy and cosmology , ed .
t. d. oswalt & w. c. keel ( dordrecht : springer ) , 91 graham , a. w. , dullo , b. t. , savorgnan , g. a. d. , 2015 , apj , 804 , 32 graham , a.w . ,
worley , c.c .
, 2008 , mnras , 388 , 1708 hopkins , p. f. , bundy , k. , murray , n. , quataert , e. , lauer , t. r. , ma , c .- p .
, 2009 , mnras , 398 , 898 huang , s. , ho , l. c. , peng , c. y. , li , z .- y . , barth , a. j. , 2013 , apj , 766 , 47 huertas - company , m. , aguerri , j.a.l .
, bernardi , m. , mei , s. , sanchez almeida , j. , 2011 , , 525 , 157 jrgensen , i. , franx , m. & kjaergaard , p. , 1995 ,
mnras , 276 , 1341 johansson , p.h . , naab , t. , ostriker , j.p
. , 2012 , apj 754 115 kewley l. j. , dopita m. a. , sutherland r. s. , heisler c. a. , trevena j. , 2001 , apj , 556 , 121 krogager , j .- k . , zirm , a. w. , toft , s. , man , a. , brammer , g. , 2014 , apj , 797 , 17 kroupa , p. , 2001 ,
mnras , 322 , 231 la barbera , f. , de carvalho , r.r . , de la rosa , i.g .
, 2010 , mnras , 408 , 1313 lackner , c.n . ,
gunn , j.e .
, 2012 , mnras , 421 , l2277 lauer , t.r . ,
gebhardt , k. , faber , s.m .
et al . , 2007 , apj , 664 , 226 licquia , t.c . ,
newman , j.a .
, 2015 , apj , 806 , 96 loeb , a. , peebles , p. j. e. , 2003 , apj , 589 , 29 longhetti , m. , saracco , p. , severgnini , p. et al . , 2007 , mnras , 374 , 614 mcgrath , e. j. , stockton , a. , canalizo , g. , 2007 , apj , 669 , 241 meert , a. , vikram , v. , bernardi , m. , 2015 , mnras , 446 , 3943 mendel , j. t. , simard , l. , palmer , m. , ellison , s. l. , patton , d. r. , 2014 , apjs , 210 , 3 moresco , m. , pozzetti , l. , cimatti , a. et al . , 2013 , , 558 , 61 morishita , t. , ichikawa , t. , 2016 , apj , 816 , 87 mosleh , m. , williams , r.j . , franx , m. , 2013 , apj , 777 , 117 naab , t. , johansson , p. h. , ostriker , j. p. , efstathiou , g. , 2007 , apj , 658 , 710 naab , t. , johansson , p. h. , ostriker , j. , 2009 , apj , 699 , l178 newman , a. b. , ellis , r. s. , treu , t. , bundy , k. , 2010 , apj , 717 , l103 newman , a. b. , ellis , r. s. , bundy , k. , treu , t. , 2012 , apj , 746 , 162 nair , p.b . ,
abraham , r. g. , 2010 , apjs , 186 , 427 onodera , m. et al .
, 2012 , apj , 755 , 260 oser , l. , ostriker , j. p. , naab , t. , johansson , p. h. , burkert , a. , 2010 , apj , 725 , 2312 patel , s.g . ,
van dokkum , p.g . ,
franx , m. et al . , 2013 ,
apj , 766 , 15 peth , m. a. , lotz , j. m. , freeman , p. e. et al . , 2015 , mnras submitted ( arxiv.1504.01751 ) poggianti , b. m. , moretti , a. , calvi , r. , donofrio m. , valentinuzzi t. , fritz j. , renzini a. , 2013 , apj , 777 , 125 rettura , a. , rosati , p. , nonino , m. et al . 2010 , apj , 709 , 512 ryan , r. e. , jr . , mccarthy , p. j. , cohen , s. h. et al . , 2012 ,
apj , 749 , 53 saglia , r. p. , s ' anchez - bl ' azquez , p. , bender , r. et al .
2010 , , 524 , a6 s ' anchez - bl ' azquez , p. , 2016
, in laurikainen e. , peletier r. , gadotti d. , eds . , astrophysics and space science library , vol .
418 , galactic bulges .
springer international publishing switzerland , p. 127
saracco , p. , longhetti , m. , gargiulo , a. , 2011 , mnras , 412 , 2707 saulder , c. , van den bosch , r.c.e . ,
mieske , s. , 2015 , , 578 , 134 schade , d. , lilly , s.j . , crampton , d. et al .
1999 , apj , 525 , 31 shen , s. , mo , h.j . , simon , d.m .
2003 , mnras , 343 , 978 simard , l. , mendel , j. t. , patton , d. r. , ellison , s. l. , mcconnachie , a. w. , 2011 , apjs , 196 , 11 strauss , m. a. , weinberg , d. h. , lupton , r. h. et al .
, 2002 , aj , 124 , 1810 tacchella , s. et al . , 2015 , mnras submitted ( arxiv:1509.00017 ) taylor , e. n. , franx , m. , glazebrook , k. et al . , 2010 , apj , 720 , 723 thomas , d. , steele , o. , maraston , c. et al . , 2013 , mnras , 431 , 1383 toft , s. , gallazzi , a. , zirm , a. , wold , m. , zibetti , s. , grillo , c. , man , a. , 2012 , apj , 754 , 3 treu t. , ellis r.s . , liao t.x . et al . 2005 , apj , 633 , 174 trujillo , i. , feulner , g. , goranova , y. et al .
, 2006a , mnras , 373 , l36 trujillo , i. , frster schreiber , n.m . , rudnick , g. et al . , 2006b , apj , 650 , 18 trujillo , i. , conselice , c. j. , bundy , k. et al .
, 2007 , mnras , 382 , 109 trujillo , i. , cenarro , a. j. , de lorenzo - c ' aceres , a. et al . , 2009 , apj , 692 , l118 trujillo , i. , ferreras , i. , de la rosa , i.g . , 2011 , mnras , 415 , 3903 valentinuzzi , t. , fritz , j. , poggianti , b. m. et al . 2010 , apj , 712 , 226 van de sande , j. , kriek , m. , franx , m. et al . 2013 , apj , 771 , 85 van der wel , a. , franx , m. , van dokkum , p. g. et al . 2014 , apj , 788 , 28 van der wel , a. , holden , b. p. , zirm , a. w. et al .
2008 , apj , 688 , 48 van dokkum , p. g. , franx , m. , kriek , m. et al . , 2008 ,
apj , 677 , l5 van dokkum , p. g. , kriek , m. , franx , m. , 2009 , nature , 460 , 717 van dokkum , p. g. , whitaker , k.e . ,
brammer , g. , et al . , 2010 , apj , 709 , 1018 van dokkum , p. g. , nelson , e. j. , franx , m. et al . , 2015 ,
apj , 813 , 23 vazdekis , a. , s ' anchez - bl ' azquez , p. , falc ' on - barroso , j. , cenarro , a. j. , beasley , m. a. , cardiel , n. , gorgas , j. , peletier , r. f. , 2010 , mnras , 404 , 1639 wellons , s. , torrey , p. , ma , ch - p .
et al . , 2016 , mnras , 456 , 1030 wolf , j. , martinez , g.d . , bullock , j.s .
et al . , 2010 ,
mnras , 406 , 1220 zahid , h. j. , damjanov , i. , geller , m. , chilingarian , i. , 2015 , apj , 806 , 122 zolotov , a. et al . , 2015 , mnras , 450 , 2327
the mendel14 parent sample spans a redshift range 0.005 @xmath68 0.4 and covers magnitudes 14.0 @xmath36 mag@xmath69 17.77 , where the bright - end magnitude limit is set to avoid problematic deblending of luminous galaxies and the faint - end limit marks the completeness flux - limit of the spectroscopic sdss sample @xcite . rather than a flux - limited sample , we need a complete mass - limited one , given that the most relevant reference parameter of the present study is the core stellar mass ( @xmath33 ) .
although we use the v / vmax statistics to correct for residual incompleteness in number density estimations ( [ sec : numberdens ] ) , it is advisable to rely on a sample with the highest possible degree of completeness throughout our target mass range , @xmath70 . to assess the sample completeness ,
we have worked out the core mass function ( cmf ) for redshift slices of @xmath71 .
( see figure [ fig : complete ] ) . to construct the cmf we have used the mendel14 decomposition parameter _ logmb _
, corresponding to the logarithm of @xmath33 , instead of the more common total stellar mass of the galaxy .
the area coverage of the sdss - dr7 legacy survey @xcite is 8,032deg@xmath72 . as a first approximation , the upper envelope of all the curves
is taken as our reference cmf . our strategy to obtain a nearly complete sample for @xmath70
is based on avoiding redshift ranges which are poorly populated due to incompleteness .
the redshift range @xmath73 attains the highest completeness level for the @xmath70 mass - limit .
the limiting cmfs for @xmath34 and @xmath74 are highlighted in figure [ fig : complete ] to enhance their contrast with the excluded neighbour redshift slices .
the low redshift slice @xmath75 is rather incomplete at its high - mass end for @xmath70 , and the high redshift slice @xmath76 is also very incomplete at its low - mass end . = 0.025 in the parent mendel14 sample .
our strategy to obtain a nearly complete sample for @xmath77 is based on avoiding redshift ranges which are poorly populated due to incompleteness .
for instance , the z @xmath78 0.025 range is excluded due to high incompleteness at @xmath77 .
the highest redshift bin is equally excluded and the selected redshift range for the sample is [email protected] . ]
two possible bias sources have been checked for their impact on the photometric b+d decompositions , namely the agn emission and the choice of a fixed de vaucouleurs shape for the compact cores .
fit ( open purple triangles ) ; core effective radii calculated with @xmath80 ( blue triangles ) and @xmath81free ( red circles ) . ] in particular , the presence of strong emission from active galactic nuclei ( agn ) can bias our measurements of galaxy sizes towards small values .
following @xcite , we have extracted emission line fluxes to construct the bpt diagram @xcite .
we designate an object as _ emission _ when the four diagnostic lines have been detected with signal - to - noise ratio above 2 .
the agn identification in the bpt diagram follows the empirical separation from @xcite and includes both the seyfert and liner classes .
galaxies designated as _ emission _ ( agn ) make the 14 ( 5)% of the compact core subsample , while corresponding figures for the non - core elliptical subsample are 10(4)% .
the mass - size and mass - density relations ( figure [ fig : spheroids ] ) , represented by the median values in mass bins , are unaffected by the exclusion of agns or even the exclusion of _ emission _ objects , proving that agn contamination is not an issue for our results . concerning the choice of the bulge sersic index , mendel14 catalogue uses a fixed de vaucouleurs bulge fit ( @xmath23=4 ) , instead of a free-@xmath23 fit .
according to the authors , the vast majority of sdss images have insufficient spatial resolution and/or signal - to - noise ratio to provide an adequate n - bulge determination and there is no significant advantage on using free - n@xmath82 over fixed n@xmath82=4 fits .
as explained by @xcite , at the sdss resolution , the bulges have their profile tail subsumed in the disk , while the central peak is washed out by the psf , making a bulge with say @xmath23=4 indistinguishable from one with 3@xmath83 . while mendel14 only derive masses for the @xmath84 b+d
fits , simard11 provide photometric parameters for both free - n and fixed n=4 fits .
we have used our compact core sample ( 10,566 objects ) to assess the potential bias introduced by the sersic index choice on the structural parameters .
alternative galaxy / core size calculations are compared in figure [ fig : appb_1 ] . as a common base
, we have used the total galaxy mass @xmath85 obtained from a single fit with a free - n sersic function . for @xmath86 ,
free - n cores are , on average , @xmath87 dex larger than n=4 cores of similar mass .
the use of free - n would decrease the reported separation of @xmath88 dex between the reference ellipticals and compact cores ( see section [ fig : spheroids ] ) .
as already mentioned , we avoided the use of free - n cores in the present study , due to the lack of @xmath33 information in the catalogues . in conclusion , the choice of sersic index does not drastically affect our results .
the moderate sensitivity to the sersic index choice is an expected outcome of the distribution of the sersic indexes in the free - n fits , which has 78 % of the values in the interval @xmath89 free-@xmath90 , close to the fixed n=4 value . in figure
[ fig : appb_1 ] , the behavior of the lines representing the full - galaxy ( green ) and disk ( purple ) sizes can be explained in terms of the host galaxy morphology .
the slight predominance of disks below @xmath91 is seemingly related to the larger fraction of disk galaxies in the lower end of our mass interval ( see figure [ fig : morphology ] ) .
the prevalence of ellipticals at the high mass end generates a drop in the disk sizes .
1 galaxies .
the galaxies hosting compact cores show that the disk / halo accounts for any mass excess above a typical core mass of @xmath92 figure [ fig : appb_2 ] represents alternative fits to the galaxies of the reference elliptical sample . as expected , the disk size is visibly smaller than the full galaxy and even the cores .
note that the cores of these elliptical galaxies are predominantly extended , only 4.5 % of them being compact . for the extended cores ,
the discrepancy between core sizes measured with free - n and n=4 amounts to @xmath93dex , less than half the offset measured for compact cores ( see figure [ fig : appb_1 ] ) finally , figure [ fig : appb_3 ] shows the variation with total mass of the mass fraction contained in the disks and bulges of our two sdss samples : compact cores and reference ellipticals .
as expected from the selection rules , the ellipticals have ( b / t ) @xmath2 1 and nearly all their mass is packed in the extended cores .
galaxies hosting a compact core show a distinct behavior in which the more massive the galaxy is , the more relevant the disk / halo becomes . in the frame of our hypothesis , this behavior is compatible with galaxies having a central , prototypical compact core of @xmath94 , embedded in an accreted disk / halo ( whose mass accounts for any excess above the core mass ) .
in face - on galaxies , contaminant disk light entering the fiber aperture does not contribute to the velocity dispersion broadening of the core ( i.e. dynamical mass ) , but contributes to the core stellar mass .
therefore , disk contaminated cores tend to show an _ unphysical _ @xmath95 . in principle , to minimise disk contamination , we could restrict @xmath63 measurements to cores in which the sdss fiber radius ( 1.5 arcsec ) were smaller than the core effective radius .
this would , however , be a deficient strategy because the core light - profile frequently gets submersed below the disk before attaining its half - light radius .
for instance , cores with @xmath96 1.5 arcsec could still be contaminated . our approach to avoid light contamination is based on the simple fact that disk light originates in much younger stellar populations than those from the core ( e.g. * ? ? ? * and references therein ) .
light from a recent star formation event on top of an essentially old core star formation history ( sfh ) is an indication of disk light contamination . instead of using colours
, we adopt the luminosity - weighted age @xmath97 as a more sensitive parameter to detect disk light contamination .
we have carried out a stellar population study in which average ages ( e.g. @xmath97 ) and metallicities are obtained through spectral fitting , using the _ starlight _ synthesis code @xcite to find the optimal mixture of simple stellar populations ( ssps ) that describes a sdss spectrum .
for the present study , the basis ssps correspond to 114 solar - scaled miles models @xcite .
these models have a kroupa universal imf @xcite and span a range of six metallicities , from z / z@xmath98= 1/50 to 1.6 , and 19 different ages , from 0.5 to 12.6 gyr .
the fitting interval spans from 4000 to 5500 , with emission lines and bad pixels being masked out .
the extinction due to foreground dust is modelled with the ccm law @xcite .
since the sdss pipeline does not compute velocity dispersions for spectra with significant emission lines ( as for some disk galaxies ) , we have also used the _ starlight _
spectral fitting to compute our own velocity dispersion values . , by demanding @xmath99 .
density contours replace the individual points , while connected solid dots represent the median values in @xmath100 bins , with dashed lines outlining the one-@xmath63 rms dispersion around the median . at @xmath100 = 1.0 ,
a rather conservative @xmath101 = 9.05 gyr cutoff value has been obtained . ]
figure [ fig : contamina ] shows the relation between the @xmath102 and the @xmath103 ratio for a mixed sample of our compact cores and reference ellipticals ( 29,482 ) .
individual points have been replaced by density contours and the connected solid dots represent the median values in bins of @xmath103 , with dashed black lines outlining the one-@xmath63 rms dispersion around the median .
the relation supports the idea that unphysical @xmath104 values are connected to low @xmath105 values , i.e. disk contamination .
a rather conservative cut has been worked out to prevent core contamination , by selecting the @xmath106gyr at which the @xmath103 ratio becomes larger than unity .
figure [ fig : morphology ] displays one of the most important results of the present study .
however , an alternative perspective to that result is presented here .
using @xmath33 instead of the @xmath107 . in this case ,
figure [ fig : morpho_mcore ] , the plot shape does not change substantially , but it is roughly shifted with respect to those using @xmath107 .
the mean @xmath108 is 0.25 dex , so morphology fractions of host galaxies with cores in the @xmath109 bin should be compared with those in the @xmath110 bin .
it is worth emphasising that the integrated results of figure [ fig : morpho_mcore ] are identical to those in table [ tab : morphotype ] , because the integration mixes all the galaxies with either @xmath111 or @xmath11210.5 . | massive high - redshift quiescent compact galaxies ( nicknamed _ red nuggets _ ) have been traditionally connected to present - day elliptical galaxies , often overlooking the relationships that they may have with other galaxy types .
we use large bulge - disk decomposition catalogues based on the sloan digital sky survey ( sdss ) to check the hypothesis that red nuggets have _ survived _ as compact cores embedded inside the haloes or disks of present - day massive galaxies . in this study , we designate a _ compact core _ as the bulge component that satisfies a prescribed compactness criterion .
photometric and dynamic mass - size and mass - density relations are used to show that , in the inner regions of galaxies at @xmath0 , there are _ abundant _ compact cores matching the peculiar properties of the red nuggets , an abundance comparable to that of red nuggets at @xmath1 .
furthermore , the morphology distribution of the present - day galaxies hosting compact cores is used to demonstrate that , in addition to the standard channel connecting red nuggets with elliptical galaxies , a comparable fraction of red nuggets might have ended up embedded in disks .
this result generalises the inside - out formation scenario ; present - day massive galaxies can begin as dense spheroidal cores ( red nuggets ) , around which either a spheroidal halo or a disk are formed later .
[ firstpage ] galaxies : evolution galaxies : formation galaxies : stellar content galaxies : structure galaxies : bulges . |
stars are formed in various environments with different density , temperature , rotation , turbulence , magnetic fields , metallicities , and so on . to take account of these effects , computational simulations including many physical processes have been extensively performed and played leading roles in understanding star formation processes . among these physical processes , metallicities are one of the important parameters which control thermal evolution through radiation transfer .
thermodynamics is of critical importance in dynamics of collapsing clouds ; for instance , the effective adiabatic index @xmath1 is an important criterion whether a gas sphere collapses dynamically by self - gravity .
it is well accepted that the early phase of star formation , from molecular clouds cores to protostellar cores , proceeds through two distinctive stages @xcite .
initially a dense cloud core collapses almost isothermally because radiation cooling is highly efficient .
when the gas gets dense enough and radiation cooling becomes inefficient , the gas pressure balances with gravity and a quasi hydrostatic object , a so - called first core , is formed .
this first core evolves by gas accretion from envelope .
when the gas temperature around the center of the cloud reaches about 2,000k where molecular hydrogen starts to dissociate , the first core becomes unstable and collapses again .
after this second collapse , finally a second core , or a protostellar core , is formed when the dissociation almost completes and the gas becomes stiff again .
modern multi - dimensional simulations with many physical processes including radiation transfer confirmed this scenario @xcite and now it is well established . by the nature of its formation , it is obviously expected that the evolution and properties of first cores depend on radiation transfer and therefore on metallicities . however , it is not very clear how the formed protostars are affected by the different evolution in the early phase .
the effects of metallicities have been studied quite well , especially in the context of formation of the first and second generation stars .
using one - zone approximation or 1d hydrodynamic calculations , the effects of different metallicities on thermal evolution has been investigated elaborately @xcite .
multi - dimensional hydrodynamic simulations with the barotropic approximation using the evolution track obtained in those calculations have been also performed , and conditions and outcomes of fragmentation have been investigated @xcite . on the other hand ,
in the context of present - day star formation , @xcite performed radiation hydrodynamic simulations of cluster - scale star formation , and found that the initial mass function is insensitive to the metallicities .
while a statistical study on the large scale was done in their work , their resolution is limited ( @xmath2 ) and optically - thick objects ( i.e. first cores and circumstellar disks ) are not well resolved , where the effects of different metallicities are expected to be more significant . in this paper ,
i report the results of 3d nested - grid rhd and rmhd simulations of protostellar collapse from molecular cloud cores to protostellar cores with different metallicities . because of the limited physical processes involved in the code , i focus on the @xmath3 models and compare them with the solar abundance models in the context of present - day low - mass star formation .
the main goal of this work is to clarify the effects of the metallicities especially on the structures and evolution of first cores and protostellar cores .
this paper is organized as follows . in
2 the numerical methods and setups used in this paper are explained .
the results of the simulations and analyses are shown in
finally conclusions and discussions are presented in 4 .
i calculate collapse of unstable molecular cloud cores until protostellar cores are formed .
the simulations are performed using the @xmath4 code which includes 3d nested - grid , resistive mhd , self - gravity , radiation transfer and a realistic equation - of - state ( eos ) . in this work ,
i do not use the resistive mhd part and i perform ideal mhd simulations and non - magnetized hydrodynamic simulations .
the detailed description of the code is given in @xcite ( hereafter paper i ) .
the mhd part is solved by the hlld approximate riemann solver @xcite with second - order linear reconstruction .
the solenoidal constraint is numerically enforced using the mixed correction proposed by @xcite . to solve the poisson equation for self - gravity , i use the multi - grid solver @xcite .
the gray flux limited diffusion approximation ( fld ; @xcite ) is adopted for radiation transfer and the radiation subsystem is integrated implicitly .
finally , i use the tabulated equation - of - state which includes the effects of chemical reactions between seven species ( @xmath5 and @xmath6 ) and internal degrees of freedom ( rotation and vibration ) .
the ortho : para ratio of molecular hydrogen is fixed to be 3:1 . to simulate low - metallicity clouds ,
the opacity tables are modified from paper i. although dust properties such as the structure , composition and size distribution can depend on the metallicities and other environmental parameters ( e.g. * ? ? ?
* ; * ? ? ?
* ) , i adopt the dust opacity tables provided by @xcite simply assuming that the dust opacities are proportional to the metallicities . for gas opacities ,
i combine the tables of @xcite and @xcite with adequate abundances . in order to isolate the effects of different metallicities , other model parameters , the eos in particular , are kept unchanged . because the contribution from heavy elements is small , the effects of the different metallicities on the eos are small .
note that clouds with @xmath7 can be reasonably simulated with the current code but it is not adequate if the metallicity is below @xmath8 . in clouds with such low metallicities ,
radiation cooling by molecular lines is significant in the relatively low - density region @xcite , but it is not implemented in the code as the dust continuum is assumed to be the dominant opacity source in the low - temperature region
. therefore , here i only study @xmath9 cases and compare them with the solar abundance models .
i use @xmath10 , @xmath11 and @xmath12 for the solar abundance models , while the low metallicity models are calculated with @xmath10 , @xmath13 .
i calculate six models ; a spherical model , a rotating model without magnetic fields and a magnetized rotating model for each metallicity .
the model parameters are summarized in table .
[ tb : models ] .
the first character of the model corresponds to the initial conditions and the second represents the metallicities . in all the models ,
the bonnor - ebert spheres @xcite with @xmath14 density enhancement are used as the initial conditions .
the central density is @xmath15 , the mass is @xmath16 and the radius is @xmath17 .
rigid - body rotation with @xmath18 is imposed for the rotating models as well as regularlized @xmath19 m=2 perturbation .
finally uniform mangetic fields of @xmath20 aligned to the rotational axis are introduced in the magnetized models .
these initial conditions are similar to the slow rotating models in paper i. the simulations are stopped when the central temperature reaches @xmath21 in the spherical and magnetized models , but the non - magnetized models are stopped at @xmath22 because they take much longer and i could not calculate the evolution after the second collapse long enough . [ cols="^,^,^,^,^,^,^,^,^",options="header " , ] the properties of the first cores , protostellar cores and outflows in the simulations . from left to right , the masses , radii and lifetimes of the first cores , the masses , radii of the protostellar cores , and the sizes of the outflows .
the largest radii are shown when the first and protostellar cores are not spherical .
note that the first and protostellar cores are still evolving and their properties are transient.[tb : restable ] in this paper , i performed 3d rhd and rmhd simulations of protostellar collapse in low metallicity environments ( @xmath23 ) and compared them with the models with the solar metallicity .
note that the evolutions after dust evaporation in the low metallicity models are largely changed from the presentation at protostars & planets vi because there were errors in the low - metallicity gas opacities . in order to simulate thermodynamics and dynamics in star formation processes , multi - dimensional simulations including radiation transfer are crucial because the thermal evolution is not a simple function of the local gas properties but highly non - linear , non - local and time - dependent .
in particular , the barotropic approximation used in many previous works can not properly reproduce the thermal properties .
the properties of the first cores , protostellar cores and outflows obtained in the simulations are summarized in table .
[ tb : restable ] . in short , the early evolution and structures of the protostellar cores
weakly depend on the metallicities while the first cores are significantly more sensitive to the metallicities .
although i simulated only the earliest evolution of the protostellar cores , these results are useful as input for further stellar evolution studies .
the detailed findings in this work are summarized as follows : \1 .
the first cores in the low - metallicity environments are colder because of more efficient radiation cooling . as a result ,
the low - metal first cores have smaller radii , masses and shorter lifetimes .
therefore , first cores are less likely to be found in low - metallicity environments .
the first cores and protostellar cores in the ideal mhd models are very similar to those in the spherical models , except for the outflows .
this is because of the efficient angular momentum transport by magnetic fields .
the molecular outflows in the mhd models are very similar in their speeds and morphologies because the thermodynamics does not affect the driving mechanism of the outflows .
the different sizes of the outflows at the end of the first core phase are simply due to their lifetimes . \4 . in the spherical and the ideal mhd models ,
the properties of the formed protostellar cores are not very sensitive to the metallicities .
this is solely a consequence of the dissociation and ionization of hydrogen and therefore quite general ( see appendix for the theoretical explanation ) .
since typical magnetic fields observed in star forming clouds are strong enough ( e.g. @xcite and references therein ) , the metallicity - dependencies of the protostellar cores are expected to be small at least when they are very young , while the first cores are more sensitive to the metallicities .
\5 . when there is rotation but no or very weak magnetic fields , the first cores form disks supported by rotation .
as they evolve more slowly , the disks become significantly colder than the first cores in the spherical models due to radiation cooling .
the effects of radiation cooling is more prominent in the low metallicity environments . as a result
, considerable differences may appear in the protostellar cores compared to the spherical and magnetized cases .
although the first cores in this work have not fragmented , these results indicate that the disks would be more susceptible to the gravitational instability and fragmentation in the low - metallicity environments .
naively speaking , this implies a higher binary rate and probability of planetary - mass companions , which is qualitatively consistent with the previous simulations using the barotropic approximation @xcite . note that
, however , this is not trivial because the initial conditions can depend on the metallicities . in order to predict the outcomes of star formation processes such as the initial mass function and the binary or multiple rates , it is crucial to convolve the initial distributions and the outcome of each formation process .
as these results are just case studies on the early phases of the specific models , more broad parameter survey and long - term calculations are obviously required . for example
, the initial rotation and magnetic fields are very important because the angular momentum and its transport are crucial in the evolution of the disks .
besides , the mass of the initial molecular cloud also has significant impact .
the radiation cooling effects would be more significant in the low mass clouds because the evolution is slower due to the lower accretion rate @xcite .
although current observations of low metallicity star forming regions are limited , future observations of star forming clouds in the outer regions of our galaxy or nearby satellites like smc and lmc will be useful to understand the effects of metallicities . as a future work ,
the effects of metallicities on non - ideal mhd effects are important and interesting . while the ohmic dissipation extends the first core lifetimes slightly by suppressing the angular momentum transport ( paper i ) , the ionization degree and resulting resistivity themselves depend on the metallicities . in the relatively low - density region where cosmic rays predominantly contribute to ionization
, the ionization degree would be higher and the resistivity would be lower than in the solar metallicity environments . on the other hand , in the relatively high density region , typically the interior of first cores , decay of radioactive nuclei is the dominant ionization source .
therefore , the ionization degree could decrease and the non - ideal mhd could be more effective .
also , as the abundance of potassium is smaller , the re - coupling due to the thermal ionization of potassium ( k ) would occur at a slightly higher density .
thus , the non - ideal mhd effects are not trivial and detailed studies are required .
+ the author thanks prof .
eve ostriker , prof .
james stone , prof .
kazuyuki omukai , prof .
masahiro machida , dr .
takashi hosokawa and dr .
shintaro takayoshi for fruitful discussions and comments .
numerical computations were partly performed on nec sx-9 at center for computational astrophysics of national astronomical observatory of japan and at japan aerospace exploration agency .
kt is supported by japan society for the promotion of science ( jsps ) postdoctoral fellowship for research abroad .
in this appendix , i explain why the adiabatic curves with different entropies converge after the dissociation of molecular hydrogen ( fig .
[ lm_rhot ] ) . for simplicity ,
i only discuss hydrogen gas as contribution of helium is small .
suppose two adiabatic tracks with the same number of hydrogen atoms but different entropies contracting from fully molecular gas to fully ionized gas .
track a starts from the initial state @xmath24 ( @xmath25 ) and ends at the final state @xmath26 ( @xmath27 ) , while track b starts from @xmath28 ( @xmath29 ) and ends at @xmath30 ( @xmath31 ) .
the entropy difference between the tracks can be estimated by considering isochoric processes @xmath32 and @xmath33 : @xmath34 for simplicity , let us assume the heat capacity is constant during these processes and neglect non - ideal effects such as chemical reactions and interaction between particles . then , the entropy difference is estimated to be @xmath35 this indicates that the temperature difference is smaller when @xmath36 is larger , and it is most significant during the second collapse because dissociation of molecular hydrogen involves a huge latent heat and increases @xmath36 . at @xmath24 and @xmath28 where the temperatures are so low ( @xmath37 ) that rotational and vibrational degrees of freedom are not excited , the heat capacity is @xmath38 , where n is the number of particles . at @xmath26 and @xmath30 where
hydrogen is completely dissociated and ionized ( @xmath39 ) , the heat capacity is also @xmath40 . because the two tracks are adiabatic ,
the entropy difference is constant everywhere .
therefore a simple relation between the temperature ratios is derived : @xmath41 as the gas is fully molecular in the initial states and fully ionized in the final states , @xmath42 ( it is possible to improve this estimate by considering contribution from helium ) .
finally , @xmath43 therefore , the temperature ratio after the second collapse is much smaller than that before the second collapse . the temperature ratio between the two tracks in fig . [ lm_rhot ]
starting from @xmath44 and @xmath45 is @xmath46 at @xmath47 , while it is @xmath48 at @xmath49 .
this is in good agreement with the theoretical prediction , @xmath50 , but slightly deviates from the prediction probably because of the neglected effects including contribution from helium and its ionization .
because this significant effect is solely a consequence of thermodynamics and chemical reactions , it leads to a quite general prediction ; as long as radiative loss during the second collapse is small ( although this might not be the case in the non - magnetized rotating models discussed in the text ) , formed protostellar cores only weakly depend on the initial conditions . | among many physical processes involved in star formation , radiation transfer is one of the key processes since it dominantly controls the thermodynamics . because metallicities control opacities , they are one of the important environmental parameters which affect star formation processes . in this work ,
i investigate protostellar collapse in solar - metallicity and low - metallicity ( @xmath0 ) environments using 3d radiation hydrodynamic and magnetohydrodynamic simulations . because radiation cooling is more effective in the low - metallicity environments , first cores are colder and have lower entropies . as a result ,
first cores are smaller , less massive and have shorter lifetimes in the low - metallicity clouds .
therefore , first cores would be less likely to be found in low - metallicity star forming clouds .
this also implies that first cores tend to be more gravitationally unstable and susceptible to fragmentation .
the evolution and structure of protostellar cores formed after the second collapse weakly depend on metallicities in the spherical and magnetized models despite the large difference in the metallicities . because this is due to the change of the heat capacity by dissociation and ionization of hydrogen
, it is a general consequence of the second collapse as long as the effects of radiation cooling are not very large during the second collapse . on the other hand ,
the effects of different metallicities are more significant in the rotating models without magnetic fields , because they evolve slower than other models and therefore more affected by radiation cooling . |
the science of self - assembly " , as introduced by george whitesides and coworkers about fifteen years ago @xcite as a promising research direction , is a large and rapidly growing field of tremendous technological potential and fundamental interest .
its motivating factor is simple : direct micro- and nano - fabrication of circuitry ( electronic , optoelectronic , or others ) becomes prohibitively time - consuming and/or expensive below a certain length scale , and as a result , scientists and engineers are searching for particles on the mesoscopic scale that by themselves assemble into potentially useful structures by virtue of their mutual interactions ; hence the term ` self - assembly ' .
there is a large number of examples of theoretical , experimental , and computational studies on this topic @xcite .
colloidal systems command particular interest , as interaction potentials in these systems have become increasingly tailorable with the advent of new ways to functionalize the colloidal surface .
these systems have tremendous capacity to assemble exotic lattices and hence to yield photonic bandgap structures .
in particular , the diamond lattice is known to have a pronounced photonic bandgap @xcite . in this paper
, we obtain the fundamentally and technologically important result that the three - dimensional diamond and wurtzite structures can self - assemble with isotropic interactions possessing a stronly repulsive core , and present the corresponding potential functions , which have been derived using optimization techniques developed previously @xcite .
specifically , we show that an @xmath0-body classical system with particles interacting via one of our derived isotropic potentials has as its ground state the corresponding lattice , in a specific volume ( or density ) range . unlike previous attempts to solve this problem , we come to this conclusion only after satisfying several important criteria : ( 1 ) lattice sums show that there is a positive pressure range in which the given lattice is stable ; ( 2 ) all crystal normal mode frequencies are real at specific volume @xmath1 ; ( 3 ) defects ( vacancies and interstitials ) are shown to cost the system energy ; and ( 4 ) the system self - assembles in a molecular dynamics simulation that starts above the freezing point and is slowly cooled . in ( 4 ) , the system may start from an entirely random configuration or with a layer of fixed particles to promote epitaxial growth .
hence we make the important distinction here between homogeneous and heterogenous nucleation in self - assembly .
it is of course a more stringent requirement that the desired lattice self - assemble from a random configuration ( homogeneous ) .
we find that the diamond crystal assembles from a completely random configuration , but the wurtzite crystal requires an epitaxial layer for it to do so ( heterogeneous ) .
the importance of criterion ( 4 ) to establish a ground state structure has been generally overlooked in previous work . currently , most studies rely exclusively on criteria ( 1 ) and/or ( 2 ) to determine lattice stability .
the only result that is guaranteed by satisfying ( 1 ) and ( 2 ) is that the structure is a local minimum , not a global one .
computationally , the only practical way that one can ensure that the desired target structure is the ground state is by doing the cooling and annealing step , ( 4 ) .
however , if the presumed ground state is not achieved by an annealing , it does not necessarily imply that it is not the ground state if the resulting energy of the structure produced exceeds that of the target structure .
there may simply be a kinetic bottleneck that prevents the annealing process from finding the ground state in a reasonable amount of computer time .
the approach by theorists and experimentalists towards finding complex self - assembling materials has primarily been _ edisonian _
( i.e. , trial - and - error ) in nature rather than being systematic and deductive .
motivated by the increasing control that experimentalists have over interparticle interactions , in colloids in particular , we have previously derived a systematic procedure to optimize these interactions to produce desired structures .
this is an example of `` inverse statistical mechanics '' : instead of using the interactions in a system to find the resulting behavior , we target a particular behavior ( e.g. structure ) , and derive optimal interactions to yield such a result .
we have recently applied these schemes to two - dimensional monodisperse , isotropic systems @xcite , and in the present paper , we use one of these methods , the so - called `` zero - temperature scheme '' , to derive isotropic interaction potentials that permit the self - assembly of the diamond and wurtzite lattices respectively .
the self - assembly of any low - coordinated target structure in three dimensions is considerably more challenging than in two dimensions because the number of possible competing structures grows dramatically with increasing dimension . with the exception of the so - called patchy particles @xcite , colloidal interactions are isotropic , and
thus we believe it is a technologically relevant as well as theoretically intriguing question to ask : what are the limits of isotropic interaction potentials for self - assembly ?
our discovery adds to fundamental understanding of the nature of the solid state by showing that isotropic interactions enable the self - assembly of crystal structures with a much broader range of coordination numbers than previously thought .
the self - assembly of a diamond lattice of dielectric spheres has been something of a holy grail in colloid science , since it was found that in the diamond lattice , unlike in the face - centered cubic ( fcc ) and body - centered cubic ( bcc ) lattices , for example , there exists a bandgap in the photonic spectrum @xcite .
this is the central necessary condition for and enabler of photonic circuitry , since this property allows control over passage vs. blockage of light , in analogy to how transistors control electric current @xcite .
it should be noted that there has been recent work on icosahedral quasicrystals as desirable photonic bandgap structures @xcite . while this is a promising avenue of research
, its central stumbling block is that self - assembly of quasicrystals in three dimensions is extremely difficult .
indeed , a single defect destroys strict quasicrystallinity .
furthermore , there is an ongoing controversy over whether a quasicrystal can be assembled using only local information .
the diamond - structure potential we find is radially complex , and seemingly difficult to fabricate in the lab directly .
that said , technology that allows experimentalists to control colloidal and nanoparticle interaction has been advancing significantly in recent years , with dna functionalization @xcite , for example , becoming a very useful method .
furthermore , radial complexity in the interaction potential is found in potentials of mean force @xcite .
recently , experimentalists have observed diamond - like behavior in mixed silver and gold nanoparticle systems @xcite , however the precise form of the interparticle interaction potential is currently unknown .
watzlawek , likos , and lwen ( wls ) have devised an isotropic interaction potential that models star polymer systems , with good agreement with experiment @xcite . in particular , the wls group has found , for certain of their parameters , a region of stability for the diamond lattice , and thus an isotropic potential that favors it . the type of potential used in the wls study has a soft core ; in particular , for small @xmath2 , @xmath3 .
while their potential is structurally simpler than the one we present here , it is not a potential that can be considered analogous to those in colloidal systems with present technology because colloidal particles have too hard a core .
we have discovered , in the course of performing optimization runs , that assembling open lattices with hard core potentials requires the potential to have more features than that of the wls potential .
we believe that tetrahedrally coordinated self - assembly via an isotropic potential is a counterintuitive result because conventional wisdom , based on atomic and molecular chemistry , has biased thinking towards the belief that highly anisotropic interactions are necessary for such locally directional structure as in diamond or wurtzite , as in the case of the carbon - carbon covalent bonds in the true diamond crystal .
the present work constitutes a clear demonstration that this chemical intuition is too restrictive .
we have previously found an isotropic potential that causes spontaneous assembly of the 6-fold coordinated simple cubic lattice @xcite , but the tetrahedral lattices present a significantly greater challenge .
finding such potentials is difficult , as the diamond and wurtzite lattices are close structural competitors with one another ( and therefore close in energy for short - ranged potentials ) . in order to understand how the diamond lattice is energetically distinguished from other lattices ,
consider wurtzite as a competitor to diamond in a system with an isotropic interaction potential . whereas the diamond lattice can be considered as two interlaced fcc lattices ,
the wurtzite is two interlaced hcp lattices .
both are tetrahedrally coordinated .
for the sake of argument , call the nearest neighbor distance @xmath4 in both lattices .
both have 12-fold coordinated neighbors at @xmath5 , but the wurtzite has a single extra neighbor at @xmath6 .
both have neighbors at @xmath7 , but the diamond is 12-fold coordinated there where the wurtzite is 9-fold .
it is evident that energetically differentiating between these two lattices ( and causing one to self - assemble in a simulation rather than the other ) using an isotropic , short ranged potential is challenging .
table 1 is gives the diamond and wurtzite coordination numbers for their first few coordination shells .
+ .coordination structure of the diamond and wurtzite lattices . here
, we set @xmath4 to be the distance to the nearest neighbor . if both lattices have the same specific volume , @xmath4 for diamond is equal to @xmath4 for wurtzite . [ cols="<,<,<",options="header " , ] in section ii , we briefly discuss the optimization methodology , and then proceed in section iii to a discussion of the results for the diamond and wurtzite cases . in this section
, we report on all necessary criteria : lattice sums , phonons , defect energies , and md self - assembly .
we also include a discussion of the melting behavior of each crystal .
we conclude in section iv with some final comments about this approach .
the inverse method used to obtain the interaction potentials , @xmath8 and @xmath9 , is discussed extensively in our previous work @xcite , and so will not be discussed here at any length .
it is , however , important to note that this `` zero - temperature scheme '' optimizes the energetic stability of a given target lattice ( in this case , diamond and wurtzite ) , in comparison to competitor lattices ( chosen previously ) , subject to the condition that the target lattice is linearly mechanically stable at an appropriately chosen specific volume ( inverse density ) .
linear stability is tested while the optimization is being carried out by calculating the phonon spectrum , and thus making sure that every normal mode of the crystal , at any given wave vector in the brillouin zone ( bz ) is real .
the competitor lattices used were the fcc , bcc , hexagonal close - packed ( hcp ) , simple hexagonal ( shex ) , and simple cubic ( sc ) structures . of course the optimization scheme for the isotropic interaction can not search over the entire space of functions of a single variable , therefore we choose a suitably normalized , well - motivated family of potentials defined by a finite number of parameters , @xmath10 , and find the set of @xmath11 s that maximizes the energetic difference between the target and competitor lattices subject to the constraint of linear mechanical stability at a particular specific volume .
_ a. diamond _ + in this case , a parameterization was chosen such that the potential has three local minima , with ratios of the positions of the second to the first , and of the third to the first , being @xmath12 and @xmath13 respectively .
these are the exact distance ratios of the neighbors in the diamond lattice .
the target coordination numbers for the first three neighbors are 4 , 12 , and 12 , and thus the first local minimum was constrained to be positive in the parameterization , in order to discriminate against closed packed and the body - centered cubic lattices , which have coordination numbers 12 and 8 respectively .
it should be noted that a number of different parameterizations were attempted , many with simpler functional forms ( i.e. with fewer local minima and maxima ) , but the potential we report here is by all criteria the most suitable .
it is @xmath14\\ -0.362477\exp\left[-150(r-1.702)^2\right ] - 0.263218\exp\left[-166(r-1.998)^2\right].\ ] ] , given in eq .
1 . given in reduced units.,width=377 ]
this potential is shown in fig .
[ fig : potential ] .
it consists of a lennard - jones - type term ( a 12 - 10 lennard - jones ) , an exponential decay , and two gaussians that serve to localize the second and third neighbors for the sake of mechanical stability of the lattice .
it is important to note here that this potential defines the energy and length units that we use for this study .
thus , derived quantities , such as pressure and volume , have units determined by the length and energy units of this potential . in the optimization scheme ,
the specific volume is chosen for us : it is that which places the nearest neighbor of the diamond lattice exactly at the first local minimum
. this occurs at @xmath15 , which sets the specific volume to be @xmath16 .
= 1 ) . given in reduced units .
, width=377 ] an essential ingredient of the optimization scheme is the lattice sums , which give the madelung energies ( crystal energies ) of a number of lattices , including diamond , over a range of specific volume .
these are shown in fig .
[ fig : latticesums ] .
the lattice sums not only show that the diamond lattice has the lowest energy of all of its competitors at the chosen @xmath1 , but also the double - tangent construction yields the zero - temperature range of pressures in which the diamond lattice is stable .
that range is 0.12 through 1.89 , in terms of the units defined by the potential ( eq .
fifteen molecular dynamics simulations ( at constant number of particles , volume , and energy - microcanonically ) were run in which 216 particles were started from a random configuration at @xmath17 , at a temperature well above the freezing point ( homogeneous case ) . upon slow cooling , all simulations resulted in diamond lattices with differing numbers and types of defects , with 1 of 15 exhibiting no defects whatsoever .
this configuration is shown in fig .
[ fig : config ] .
although it may not be immediately clear from the two - dimensional projections , this is indeed the 4-coordinated diamond lattice , with the first neighbor at distance @xmath18 .
the next neighbors are at @xmath19 ( 12-coordinated ) , @xmath20 ( 12-coordinated ) , and @xmath21 ( 6-coordinated ) , and so on .
these are the correct distances and coordination numbers of the diamond lattice .
heterogeneous nucleation also produced self - assembly of the diamond lattice in a 252-particle system , with some point defects . in order to calculate the energy of defects ( vacancies and interstitials )
, molecular dynamics simulations were run in which a particle was either removed ( n=215 ) or inserted ( n=217 ) .
the diamond lattice is a sub - lattice of bcc , and the interstitial was initially set on a non - occupied bcc site ( and it did not move significantly during the relaxation ) .
the specific volume was set to be @xmath17 in both cases .
the vacancy energy was found to be 2.866 and the interstitial energy was 3.794 . in order to study the melting behavior of the crystal , gibbs - ensemble monte carlo simulations were run ( at constant number of particles , pressure , and temperature ) . the pressure was set to be 1.0 , well within the range of stability of the lattice .
the temperature in the simulation was slowly raised , starting near absolute zero , giving the system sufficient time to equilibrate , and the resulting volume and configurational energy recorded .
these two quantities , plotted in fig .
[ fig : meltingcurve ] , show dramatically how strongly first - order the melting transition is .
another interesting property shown in this figure is that the system expands upon melting the crystal ( by roughly 45% ) .
this sort of qualitative behavior is what we would expect for a dense crystal , but perhaps not for such an open lattice as diamond . indeed , solid water ( ice ) has a tetrahedral structure ( as do real diamond and silicon ) and contracts on melting , by approximately 8.3% at atmospheric pressure @xcite .
while there is no intrinsic reason to believe that our model system should act similarly , it is interesting to observe how qualitatively differently its melting behavior is . the phonon spectra for the potential @xmath8 with particles arranged in a diamond lattice are shown in fig .
[ fig : dia - pho ] . for the wave vectors included in the plot , all modes have real frequencies ( i.e. are propagating ) .
we mention that although only certain special wave vectors are shown in the plot , at the given specific volume @xmath17 , all frequencies in the three dimensional bz are real , and thus the crystal is linearly mechanically stable .
+ , at specific volume @xmath17 . only certain trajectories between points of high symmetry in the brillouin zone are shown , although at every wave vector , @xmath22 .
this implies that the crystal is linearly mechanically stable .
given in reduced units.,width=377 ] _ b. wurtzite _
+ the wurtzite potential , @xmath9 , is given by @xmath23 - 0.80\exp\left[-100(r-\sqrt{8/3})^2\right]- 0.40\exp\left[-100(r-\sqrt{11/3})^2\right]- 0.40\exp\left[-100(r-\sqrt{16/3})^2\right].\label{equ : wur - pot}\ ] ] this potential is shown in fig .
[ fig : wur - pot ] .
the rationale for its shape is similar to that for diamond . the most important feature , perhaps , is the negative gaussian centered at @xmath24 , because this is comparatively close to the next neighbor at @xmath25 , which has six - fold coordination .
the analogous neighbor in diamond is relatively separate from its next neighbor . for wurtzite , the specific volume is @xmath26 . + , given in eq .
[ equ : wur - pot ] . given in reduced units.,width=377 ] the lattice sums for @xmath9
are given in fig .
[ fig : wur - lat ] .
they show that there is a region of stability for the wurtzite lattice , and using the double - tangent construction , they give the pressure range of stability as from 0.0 through 0.94 in our reduced units .
it is clear from the lattice sums that the hcp lattice is an important competitor to the wurtzite ( at lower specific volume ) .
this is natural : the wurtzite is nothing but two interlaced hcp lattices , just as the diamond is two interlaced fcc lattices . , given in eq .
[ equ : wur - pot ] .
the double - tangent construction determines that at zero temperature , the pressure range of stability for the wurtzite lattice is 0.0 through 0.94 . given in reduced units .
, width=377 ] the wurtzite crystal assembled upon heterogeneous , epitaxial growth from a fixed layer of particles within a 288-particle md simulation , upon slow cooling .
the layer consisted of two triangular lattice planes , one from each of the two interlaced hcp lattices , such that growth proceeded in the c - direction ( out of plane ) .
all other particles in the system were inserted randomly .
the self - assembly was not perfect , i.e. there were a number of ( positive - energy ) point defects remaining in the system that could not be annealed out , but this is to be expected for kinetic reasons .
defect calculations were performed as in section iii.a .
, the diamond case .
the vacancy energy was calculated by simply removing an atom from the perfect wurtzite crystal and allowing the system to relax in a microcanonical md simulation .
we report the vacancy energy as @xmath27 , in reduced units . for the interstitial case ,
consider one of the two interlaced hcp lattices that together form wurtzite .
this is an ` abab ... ' stacking of planar triangular lattices , contrasting the ` abcabc ... ' stacking of planar triangular lattices in the fcc structure . the wurtzite interstitial particle placed in a triangular plane directly below the would - be `` c '' site .
its energy is reported as @xmath28 in reduced units .
the melting behavior of the crystal was studied in the same way as that of the diamond crystal in the previous section was studied , namely using gibbs - ensemble monte carlo simulations .
the pressure was set to be 0.5 , well within the range of stability of the lattice . as in the diamond case ,
the temperature in the simulation was slowly raised , giving the system sufficient time to equilibrate , and the resulting volume and configurational energy recorded .
these two quantities , plotted in fig .
[ fig : wur - meltingcurve ] , show dramatically how strongly first - order the melting transition is for the wurtzite case ( expansion of roughly 50% ) .
the melting behavior is similar to the diamond case .
it should be noted that this is qualitatively different from the behavior of ice - ih , also a wurtzite structure , which contracts upon melting . the wurtzite phonon spectrum is shown in fig .
[ fig : wur - pho ] .
trajectories of high symmetry are shown in the figure , but indeed all normal mode frequencies in the brillouin zone are real , indicating that the cyrstal is linearly mechanically stable . at specific volume @xmath29
. only certain trajectories between points of high symmetry in the brillouin zone are shown , although at every wave vector , @xmath22 .
this implies that the crystal is linearly mechanically stable . given in reduced units.,width=377 ]
in this paper , we have reported two isotropic interaction potentials such that in an n - body classical system with particles interacting via these potentials , the diamond and wurtzite lattices self - assemble as the system is slowly cooled .
the functional forms of @xmath8 and @xmath9 were derived using an inverse statistical mechanical methodology developed by the authors , to be employed in tailoring intercomponent interactions to achieve self - assembly of targeted structures . admittedly , these potentials have substantial radial complexity , but this is reminiscent of the complexity one sees in potentials of mean force in many - body systems .
furthermore , the possibilities of engineering multi - feature potentials are growing significantly as the techniques of dna and polymer functionalization of colloids and nanoparticles are refined . in future work
, we plan on further exploring the self - assembly possibilities for isotropic potentials , including other carbon structures such as graphite , fullerenes , and nanotubes .
we would like to kindly thank christos likos for very helpful and productive discussions during his visit to princeton .
this work was supported by the office of basic energy sciences , doe , under grant no .
de - fg02 - 04er46108 . | using inverse statistical - mechanical optimization techniques , we have discovered isotropic pair interaction potentials with strongly repulsive cores that cause the tetrahedrally - coordinated diamond and wurtzite lattices to stabilize , as evidenced by lattice sums , phonon spectra , positive - energy defects , and self - assembly in classical molecular dynamics simulations .
these results challenge conventional thinking that such open lattices can only be created via directional covalent interactions observed in nature .
thus , our discovery adds to fundamental understanding of the nature of the solid state by showing that isotropic interactions enable the self - assembly of open crystal structures with a broader range of coordination number than previously thought .
our work is important technologically because of its direct relevance generally to the science of self - assembly and specifically to photonic crystal fabrication . |
this paper continues the project , begun in @xcite , of using heegaard splittings to construct closed essential surfaces in finite covers of 3-manifolds .
the idea , based on the work of casson and gordon @xcite , is to lift a heegaard splitting of a 3-manifold to a finite cover in which there are disjoint compressing disks on each side . by compressing the lifted heegaard surface along an appropriate choice of such disks ,
we hope to arrive at an essential surface . by a _ knot manifold _
we mean a connected , compact , orientable 3-manifold whose boundary is a single torus .
a _ tunnel system _ for a knot manifold @xmath0 is a collection @xmath4 where the @xmath5 s are disjoint , properly embedded arcs in @xmath0 , such that @xmath6 is homeomorphic to a handlebody .
the _ tunnel number _ of @xmath0 , denoted @xmath7 , is the minimal cardinality of a tunnel system for @xmath0 .
we focus attention on tunnel number one , non - fibered knot manifolds .
these are obtained by attaching a single 2-handle to a genus two handlebody .
we shall give a condition on the 2-handle which , if satisfied , ensures that in all large enough cyclic covers , the lifted heegaard splitting can be compressed to obtain an essential surface .
there is also a statement about incompressibility after dehn surgery .
freedman and freedman ( @xcite ) have already proved that , in _ any _ non - fibered knot manifold , all but finitely many cyclic covers are large ( i.e. contain closed essential surfaces ) .
cooper and long @xcite then proved a result about virtually haken dehn surgery for these manifolds , and also obtained a bound on the number of excluded covers in terms of the genus of the knot .
however , our results provide a computational benefit .
we computed , for all of the 453 non - fibered , 2-generator knot manifolds in the snappea census , a covering degree past which all cyclic covers are large , and the bounds obtained are typically improvements over known bounds . for other connections between heegaard splittings and virtually haken 3-manifolds ,
see @xcite , @xcite , @xcite , @xcite .
let @xmath8 be a connected , closed , orientable surface of positive genus . recall that a _ compression body _
@xmath9 is a 3-manifold obtained from @xmath10 by first attaching a collection of 2-handles along disjoint curves in one component of @xmath11 , and then capping off all resulting 2-sphere boundary components with 3-balls .
one of the boundary components of @xmath9 is homeomorphic to @xmath8 , and is called the _ outer boundary _ of @xmath9 , denoted @xmath12 .
the other components of @xmath13 form the _ inner boundary _ , denoted @xmath14 .
if @xmath15 is a 3-manifold with boundary , and @xmath16 is a collection of disjoint compression disks for @xmath17 , we let @xmath18 . if @xmath15 is a 3-dimensional submanifold of a 3-manifold @xmath19 , and
if @xmath20 is a collection of disjoint compression disks for @xmath17 , then we let @xmath21 $ ] denote @xmath22 , where @xmath23 is a regular neighborhood of @xmath24 in @xmath25 .
a _ disk system _ @xmath26 for a compression body @xmath9 is a set of disjoint compressing disks for @xmath13 of minimal cardinality such that @xmath27 has incompressible boundary .
we shall use some basic facts about compression bodies , which can be found in @xcite .
heegaard splitting _ of a compact 3-manifold is a decomposition @xmath28 , where the @xmath29 s are compression bodies with outer boundary homeomorphic to @xmath8 .
the _ heegaard genus _ of @xmath0 , denoted @xmath30 , is the minimal genus of @xmath8 for all such decompositions .
if @xmath0 has boundary , then a _ tunnel system _ for @xmath0 is a collection of properly embedded arcs in @xmath0 , whose exterior is a handlebody .
the _ tunnel number _ of @xmath0 , denoted @xmath7 , is the minimal cardinality among all tunnel systems for @xmath0 .
it is an elementary fact that , if @xmath0 is a knot manifold , then @xmath31 . for the remainder of the paper
, @xmath0 will be a fixed knot manifold with @xmath32 and @xmath33 .
thus there is a heegaard splitting @xmath34 , where @xmath35 is a genus 2 handlebody , and @xmath9 is a genus 2 compression body . let @xmath36 be a disk system for @xmath35 , and let @xmath37 be the unique ( up to isotopy ) non - separating compression disk for @xmath9 .
we assume that @xmath37 has been isotoped so that every component of @xmath38 represents an essential arc in @xmath39 . since @xmath33
, there is a unique surjective homomorphism @xmath40 , where @xmath41 is the free factor of @xmath42 .
let @xmath43 denote the corresponding n - fold cyclic cover , with @xmath44 denoting the infinite cyclic cover .
let @xmath45 and @xmath46 be the pre - images in @xmath43 of @xmath47 and @xmath8 , respectively . then @xmath48 is a heegaard splitting of @xmath43 of genus @xmath49 .
let @xmath50 be simple closed curves transverse to @xmath51 such that @xmath52 ( the kronecker delta function ) .
we also assume that @xmath53 and @xmath54 intersect in a single point @xmath55 , which will be the base point for @xmath56 , and we assign orientations to @xmath57 and @xmath58 so that the algebraic intersection numbers @xmath59 are both @xmath60 .
we call such pair of curves @xmath61 _ dual curves _ for the disk system @xmath62 of @xmath35
. [ system ] we may choose a disk system @xmath63 so that @xmath64 and @xmath65 .
suppose @xmath66 , and that @xmath67 .
let @xmath68 be an oriented embedded arc in @xmath69 such that @xmath70 , and that its orientation agrees with the orientation on @xmath54 , but disagrees with the orientation on @xmath53 .
let @xmath71 be a properly embedded disk in @xmath35 obtained by band sum of @xmath72 and @xmath73 along the arc @xmath68 , and let @xmath74 .
then obviously we may assume that @xmath75 and @xmath76 are disjoint , and see that they form a disk system for @xmath35 .
choose @xmath77 , @xmath78 . then up to an obvious homotopy of @xmath79 in @xmath80 , we may consider @xmath79 as a simple closed curve , and see that @xmath81 form a dual curve pair for the disk system @xmath82 ( with a suitable choice of orientation for @xmath83 ) .
further we have that @xmath84 , and @xmath85 .
now we replace the disk system @xmath63 with the disk system @xmath86 , and repeat the above procedure . applying the euclidean algorithm
, we may continue until we have a disk system @xmath63 for which @xmath87 ( say ) and @xmath88 . since @xmath89 , @xmath90 and @xmath91
are relatively prime , so the resulting disk system satisfies the requirements of lemma [ system ] . for the remainder of the paper
, we shall assume that the disk system @xmath63 has been chosen as in lemma [ system ] .
let @xmath92 be a base point for @xmath0 , and let @xmath93 be the lifts to @xmath43 ( where @xmath94 ) , with the natural indexing
. let @xmath95 in @xmath96 be arcs connecting @xmath55 with @xmath97 .
let @xmath98 denote the lift of @xmath99 to @xmath43 corresponding to the lift of @xmath100 with base point @xmath101 ( in our notation , we have suppressed the dependence of @xmath102 on @xmath3 , trusting the meaning to be clear from context ) .
see figure 1 .
let @xmath103 be the geometric intersection pairing , and for a loop @xmath104 in @xmath105 , define the _ width _ of @xmath104 to be : @xmath106 if @xmath104 is a loop in @xmath8 which lifts to @xmath105 , and @xmath107 and @xmath108 are any two lifts to @xmath105 , then @xmath109 .
thus for any such loop , we define @xmath110 , where @xmath107 is any lift of @xmath104 to @xmath105 .
set ( for the remainder of the paper ) @xmath111 we shall restrict attention to the case where @xmath37 intersects @xmath72 and @xmath73 non - trivially . if @xmath112 , let @xmath113 denote the lift of @xmath37 to @xmath43 which intersects @xmath114 , but is disjoint from @xmath115
. then @xmath116 forms a disk system for @xmath117 and @xmath118 forms a disk system for @xmath119 ( c.f . fig . 1 for an example with @xmath120 and @xmath121 ) .
set @xmath122 , set @xmath123 , and @xmath124 , @xmath125 .
recall that a collection @xmath126 of disjoint simple closed curves in the boundary of a handlebody @xmath15 is _ disk busting _ if @xmath127 is incompressible in @xmath15 .
[ m - lift ] let @xmath128 be an integer .
we say that @xmath37 satisfies the _ m - lift condition _ if : + 1 .
the curve @xmath129 is disk - busting in @xmath35 .
the set @xmath130 is disk - busting in @xmath131 .
the set @xmath132 is not disk - busting in @xmath131 , for all @xmath133 .
+ 4 . for each @xmath133
, there is a compression disk @xmath134 such that @xmath135 $ ] is linearly independent from @xmath136 , ... , [ \partial e^m ] \}$ ] in @xmath137 . _
if @xmath138 $ ] is fibered , then it follows from lemma [ deg2 ] below that condition 2 fails , so @xmath37 does not satisfy the @xmath139-lift condition for any @xmath139 .
+ recall that a 3-manifold @xmath0 is _ large _ if it contains a closed essential surface , i.e. an incompressible surface which is not parallel to a component of @xmath140 .
we prove : [ cyclic ] if @xmath37 satisfies the @xmath139-lift condition , then @xmath43 is irreducible and large for any @xmath141
. let @xmath142 be a longitude for @xmath0 ( i.e. a simple , closed , essential curve in @xmath140 which lifts to a loop in @xmath44 ) .
fix a meridian @xmath143 for @xmath0 ( i.e. a simple closed curve in @xmath144 which intersects @xmath145 exactly once ) .
a slope @xmath146 in @xmath144 means the pair of homology class @xmath147+q[\l])$ ] and @xmath148 .
we use @xmath149 to denote the closed manifold obtained by dehn filling of @xmath0 with slope @xmath146 .
let @xmath150 .
then @xmath151 is a finite integer .
then we have : [ surgery ] if @xmath37 satisfies the @xmath139-lift condition , then @xmath152 is virtually haken for any @xmath153 , @xmath154 , and q with @xmath155 . given a 2-generator , 1-relator presentation of a 3-manifold group , there is an algorithm to decide if this presentation corresponds to a genus 2 heegaard splitting ( conjecturally it always does , see @xcite ) . from the data of such a geometric presentation , it is possible to check if the @xmath139-lift condition holds for a given @xmath139 . using the computer program gap , we have shown : every 2-generator , 1-relator 3-manifold @xmath0 in the snappea census of 1-cusped hyperbolic 3-manifolds has a genus 2 heegaard splitting .
moreover , if @xmath33 and @xmath0 is non - fibered , then @xmath0 has a genus 2 heegaard splitting whose 2-handle satisfies the @xmath139-lift condition for some @xmath139 .
a complete table of the values of @xmath139 is available at www.math.buffalo.edu/@xmath156jdmaster .
the first few values are given in table 1 .
to prove these theorems , consider the surface @xmath46 ( recall this is the pre - image of the heegaard surface @xmath8 in @xmath43 ) .
then @xmath46 is a heegaard surface for @xmath43 of genus @xmath49 , which we shall compress to both sides .
on the handlebody side , we compress @xmath46 along a single lift , @xmath157 , of @xmath73 ; on the compression body side we compress @xmath46 along all the lifts of @xmath37 which are disjoint from @xmath157
. we shall show that if @xmath37 satisfies the @xmath139-lift condition and @xmath158 , then the resulting surface is incompressible .
let @xmath159,\\ y_n & = & ( h_n / d_2 ^ 1)[\mathcal{e}_{n - k+1}],\textrm { and}\\ s_n & = & \partial y_n.\end{aligned}\ ] ] note that @xmath160 . [ nonparallel ] the surface @xmath161 is connected , has @xmath162 , and is not parallel to @xmath163 . to prove that @xmath161 is connected , it is enough to show that @xmath164 , [ \partial e^1 ] , ... , [ \partial e^n]$ ] are linearly independent in @xmath165 .
let @xmath166 denote the algebraic intersection pairing on the first homology group of a surface .
recall from lemma [ system ] that @xmath167 is a generator of @xmath168 .
it follows that @xmath169 , [ \partial d_2 ] ) = 0 $ ] and that @xmath170 , [ \p d_2]\}$ ] are linearly independent .
we may complete @xmath171 , [ \partial d_2 ] \}$ ] to a symplectic basis @xmath171 , [ \partial e]^ * , [ \partial d_2 ] , [ \partial d_2]^ * \}$ ] , i.e. a basis satisfying @xmath172 , [ \partial e]^ * ) = i([\partial d_2 ] , [ \partial d_2]^ * ) = 1 $ ] , and @xmath172^ * , [ \partial d_2 ] ) = i([\partial e]^ * , [ \partial d_2]^ * ) = i([\p e],[\p d_2])=i([\p e],[\p d_2]^*)=0 $ ] .
let @xmath173 be an embedded loop representing @xmath174^*$ ] , and intersecting @xmath129 geometrically exactly once ( such representative always exists ) .
then @xmath175 lifts homeomorphically to loops @xmath176 ( since @xmath177,[\p d_2])=0 $ ] ) .
as @xmath178 also lifts to @xmath46 as @xmath179 , it is easy to see that @xmath180 .
also @xmath181 , [ \partial d_2 ^ 1 ] ) & = & i([\tilde{\alpha}_i ] , [ \bigcup_j \partial d_2^j]/n ) ( \textrm { since all lifts of } \p d_2 \textrm { are homologous in } f_n ) \\ & = & i([\alpha ] , [ \partial d_2])/n\\ & = & 0.\end{aligned}\ ] ] recall that we have a map @xmath182 ; let @xmath183 be a loop with @xmath184 , and let @xmath185 be the pre - image in @xmath43 .
then @xmath186 , [ \tilde{\alpha}_1 ] , ... , [ \tilde{\alpha}_n]$ ] are dual classes for @xmath164 , [ \partial e^1 ] , ... , [ \partial e^n]$ ] in @xmath187 , which proves the linear independence , and completes the proof that @xmath161 is connected .
the linear independence of @xmath164 , [ \partial e^1 ] , ... , [ \partial e^n]$ ] also allows us to compute : @xmath188 finally , note that @xmath161 is not parallel into @xmath163 , since every loop in @xmath161 projects to an element in @xmath189 ( because @xmath24 is disjoint from @xmath157 ) , but each component of @xmath163 contains a loop whose projection is not in @xmath190 . to prove theorems [ cyclic ] and theorem [ surgery ]
, we shall show that when @xmath191 , @xmath161 is incompressible in both @xmath192 and @xmath193 , and that when @xmath194 , @xmath161 remains incompressible in an equivariant dehn filling of @xmath43 along @xmath195 ( which may have several components ) which is a free cyclic cover of @xmath152 .
we will require the following result of magnus ( see @xcite ) .
the statement given here is easily seen to be equivalent to the standard statement .
[ frei ] ( freiheitsatz for 1-relator groups ) let @xmath196 be a 1-relator group , where @xmath197 is a freely reduced word .
let @xmath198 , let @xmath199 , and suppose that some @xmath200 appears in @xmath197 .
then @xmath201 freely generates a free subgroup of @xmath202 .
[ staggered ] let @xmath203 be a word in which @xmath204 and @xmath205 appear non - trivially , and consider the group @xmath206 then for any @xmath207 , the set @xmath208 freely generates a free subgroup of @xmath202 .
let @xmath209 , and let @xmath210 . by repeated applications of theorem [ frei ] , the group @xmath202 has the structure of the following iterated amalgamated free product over free subgroups : @xmath211 . by theorem [ frei ]
, each @xmath212 injects into @xmath213 and @xmath214 .
each @xmath213 thus injects into @xmath202 , and so we obtain the corollary .
suppose @xmath215 is a 1-relator group ( where @xmath197 is a cyclically reduced word ) which admits a surjective homomorphism @xmath216 , such that @xmath217 .
let @xmath218 .
then @xmath219 is generated by the @xmath220 s , and the relation @xmath197 lifts to a relation @xmath221 on the @xmath220 s , so that @xmath219 has a presentation as in the statement of corollary [ staggered ] .
write @xmath222 , and consider the finite integer sequence @xmath223
. then we have : [ 2gen](brown ) if @xmath223 has a repeated minimum ( or maximum ) ( i.e. it assumes its minimum ( or maximum ) value more than once ) , then @xmath224 is not finitely generated .
the case of a repeated minimum is a special instance of theorem 4.2 in @xcite , and the case of a repeated maximum follows from a trivial modification of the proof ( which is an application of the freiheitsatz ) .
we also need the following , a special case of corollary 2.2 of @xcite , which is in turn a slight modification of a theorem proved by jaco in @xcite .
( handle addition lemma ) [ handle ] let @xmath0 be an irreducible 3-manifold with compressible boundary of genus at least 2 , and suppose @xmath225 is a simple closed curve , such that @xmath226 is incompressible in @xmath0 . then the 3-manifold obtained by adding a 2-handle to @xmath0 along @xmath175 is irreducible , and has incompressible boundary .
the results presented in this section are essentially combinations of results due to brown ( @xcite ) and bieri - nuemann - strebel ( @xcite ) .
if @xmath227 is a compressing disk for @xmath228 , and @xmath175 is a simple closed curve in @xmath228 , then let @xmath229 be the geometric intersection number of @xmath175 and @xmath227 ; in other words , @xmath229 is the minimal cardinality of @xmath230 over all curves @xmath231 which is isotopic to @xmath175 in @xmath228 .
if @xmath63 is a disk system for @xmath35 , then there are dual simple loops which form a free basis for @xmath232 .
if @xmath233 corresponds to the generator @xmath234 , and if @xmath175 is a simple loop in @xmath35 , then @xmath229 is the number of times the generator @xmath234 appears in a cyclically reduced representative for the conjugacy class of @xmath235 $ ] in @xmath232 .
[ deg2 ] suppose @xmath0 is a knot manifold with @xmath32 , and suppose that @xmath236 and @xmath102 are as defined in section [ defs ] .
( a ) . if @xmath0 is fibered , then @xmath237 in @xmath228.(b ) .
if @xmath0 is non - fibered , then @xmath238 and @xmath239 in @xmath228 .
suppose @xmath238 or @xmath239 .
then by theorem [ 2gen ] ( together with the note in the proceeding paragraph of the present lemma ) , @xmath189 is not finitely generated , and so by @xcite , @xmath0 is not fibered .
suppose @xmath240 intersects one of the disks , say @xmath241 , exactly once .
we shall show that in this case @xmath0 is fibered .
dual to each @xmath242 is a generator @xmath220 for the fundamental group of @xmath44 .
the boundary of the disk @xmath240 gives a relation among these generators which involves @xmath205 only once ; therefore @xmath243 . similarly , using the relation corresponding to the disk @xmath244
, we get that @xmath245 continuing in this way , we see that all of the generators @xmath220 with @xmath246 can be expressed in terms of @xmath247 .
let @xmath248 be the component of @xmath249 containing @xmath250 , and let @xmath251 $ ] , which is a submanifold of @xmath44 .
the argument we just gave shows that @xmath252 is finitely generated .
note that is a non - separating incompressible surface @xmath24 in @xmath0 with boundary slope @xmath145 such that @xmath253 is the infinite cover dual to @xmath24 .
let @xmath254 be a lift of @xmath24 to @xmath44 which is disjoint from @xmath255 , and let @xmath256 be the component of @xmath257 which contains @xmath255 .
then @xmath258 is compact , and since @xmath259 is finitely generated , @xmath260 is finitely generated as well .
let @xmath261 , let @xmath262 and @xmath263 be the two pre - images of @xmath24 in @xmath264 , let @xmath265 be the pre - images of @xmath24 in @xmath44 , and let @xmath266 be the submanifold of @xmath255 bounded by @xmath267 and @xmath265 . since @xmath265 is incompressible , @xmath266 is @xmath268-injective in @xmath255 for each @xmath207 .
if neither of the maps @xmath269 is onto , then @xmath270 forms an an infinite sequence of subgroups of @xmath259 , with @xmath271 properly containing @xmath272 for each @xmath207 , which is a contradiction , since @xmath259 is finitely generated .
therefore , one of the induced maps @xmath273 is onto , and so , as in the proof of theorem 2 in @xcite , @xmath0 is fibered .
[ fib ] if @xmath0 is fibered , then @xmath37 does not satisfy the @xmath139-lift condition for any @xmath139 .
suppose @xmath0 is fibered , let @xmath128 be an integer , and consider the cover @xmath274 of @xmath8 . in @xmath274 , we have that @xmath275 is disjoint from @xmath276 for all @xmath277 , and by lemma [ deg2 ] a , @xmath278 .
therefore there is a compressing disk @xmath279 in @xmath280 ( whose boundary is equal to @xmath281 ) with @xmath282 , and so @xmath37 fails condition 2 .
we shall now begin the proof of theorem [ cyclic ] , which will occupy the next three sections .
[ irred ] if @xmath283 has incompressible boundary , then @xmath43 is irreducible for all @xmath3 . by theorem
[ handle ] , @xmath284 is irreducible . by @xcite ( or @xcite ) , the cover of an irreducible manifold is irreducible , so @xmath285 is irreducible for all @xmath286 .
we remark that conditions 2 - 4 of the @xmath139-lift condition are not needed in this case , and so our proof implies that @xmath192 has incompressible boundary whenever @xmath140 is incompressible .
let @xmath287 . by theorem [ handle ]
it is enough to prove : [ bdy ] if @xmath288 , then @xmath289 has incompressible boundary .
first we need : [ mer ] for each @xmath3 there is a loop @xmath290 such that @xmath291 . by the exact sequence of the pair , there is a loop @xmath292 such that @xmath293 = i(\a , d_2 ^ 1 ) \neq 0 $ ] . letting @xmath294
be the pre - image of @xmath295 in @xmath195 , we have @xmath296 . since @xmath297 = [ d_2^j ] \in h_1(m_n)$ ] for all @xmath298
, then we have @xmath299 .
( of lemma [ bdy ] ) suppose otherwise that there is a compressing disk in @xmath300 .
first , if there is a compressing disk , we claim that there must be a non - separating one . to see this ,
suppose that @xmath301 is a separating compressing disk . if there are no non - separating compressing disks , then one of the components of @xmath302 is homeomorphic to a surface cross an interval , and the other component is a handlebody containing @xmath303 .
every curve in @xmath195 lies on the surface cross interval side , but by lemma [ mer ] , there is a curve in @xmath195 which has non - trivial intersection with @xmath157 , yielding a contradiction .
so we may assume that there is a non - separating compressing disk @xmath301 in @xmath300 .
consider the heegaard surface @xmath46 and the curves @xmath304 and @xmath305 , @xmath306 , in @xmath46 .
note that since @xmath288 , @xmath307 are all disjoint from @xmath308 and @xmath309 ( by considering the definition of @xmath310 ) .
the two simple closed curves @xmath308 and @xmath309 cut @xmath46 into two components , @xmath311 and @xmath312 , one of which , say @xmath311 , is disjoint from all @xmath313 . the curves @xmath314 may intersect @xmath308 and @xmath309 .
but by the property of the width @xmath315 again , any arc component of @xmath316 has either both endpoints in @xmath308 or both endpoints in @xmath309 .
further , every arc component of @xmath316 is disjoint from @xmath303 .
let @xmath317 be the subsurface of @xmath46 which is the union of @xmath311 and a small regular neighborhood of the arcs @xmath316 in @xmath312 ( in other words , @xmath317 is @xmath311 with some bands attached , one for each arc component in @xmath316 ) .
then @xmath317 is a connected subsurface of @xmath46 which contains all @xmath318 but is disjoint from all @xmath319 and @xmath303 .
let @xmath320 be the surface obtained from @xmath317 by surgery along the curves @xmath314 ( i.e. cut @xmath317 open along @xmath321 and fill each of new boundary circles with a disk ) , which may not be connected .
note that @xmath322 is a disk system for the compression body @xmath323 and @xmath324 is an @xmath325-bundle over a surface .
as @xmath320 is disjoint from @xmath307 , @xmath320 is contained in the horizontal boundary of the @xmath325-bundle .
so @xmath326 are vertical annuli of this @xmath325-bundle . by standard
cut - and - paste operations along arcs and circles of @xmath327 , we get a non - separating compressing disk , still denoted @xmath301 , which is disjoint from the annuli @xmath328 .
it is easy to see that @xmath301 can not be contained in @xmath329 , so it follows that @xmath330 is disjoint from all @xmath303 , @xmath331 , ... , @xmath308 .
hence the width of @xmath301 is strictly less than @xmath315 .
let @xmath332 , and let @xmath333 be a lift of @xmath68 in @xmath44 , the group @xmath334 has the following presentation : @xmath335 where the relations correspond to the lifts of @xmath37 . since @xmath336 , the loop @xmath333 represents an element in the subgroup of @xmath334 generated by the elements @xmath337 , for some @xmath207 . by corollary
[ staggered ] , these elements are a basis for a free subgroup ; since @xmath333 represents a trivial element in @xmath334 , we see that @xmath333 represents the trivial word in @xmath337 .
thus @xmath333 bounds a disk in @xmath228 by dehn s lemma , and thus @xmath68 bounds a disk in @xmath117 .
thus there is a non - separating sphere in @xmath43 , contradicting lemma [ irred ] .
[ x_n ] suppose @xmath288
. then @xmath192 has incompressible boundary .
this is a consequence of lemma [ bdy ] and theorem [ handle ] .
in this section , we are under the assumption that @xmath338 satisfies the @xmath139-lift condition .
recall that @xmath339 .
[ nohomotopy ] the curve @xmath340 can not be isotoped in @xmath341 $ ] to intersect @xmath342 fewer than two times .
we have @xmath343 & = & < x_1 , ... , x_n| w_1 , ... , w_{n - k}>\\ & \cong & < x_1 , ... , x_{n-1}| w_1 , ... , w_{n - k } > * < x_n>\end{aligned}\ ] ] where @xmath344 is dual to @xmath114 , and the word @xmath345 corresponds to @xmath346 .
the word @xmath347 can be cyclically permuted to have the form @xmath348 where the @xmath349 s are freely reduced words involving only @xmath344 s with @xmath350 , and each @xmath351 represents a nontrivial element in the group @xmath352 .
suppose @xmath340 can be isotoped to be disjoint from @xmath342 in @xmath353 $ ] .
then , using the relations @xmath345 , @xmath354 , the word @xmath347 can be re - written entirely in terms of @xmath344 s , @xmath355 , and this would imply that one of the @xmath351 s must be trivial in @xmath356 $ ] .
however , @xmath351 is a freely reduced word on @xmath357 , which by theorem [ staggered ] freely generate a free subgroup , for a contradiction .
suppose @xmath340 can be isotoped to intersect @xmath342 exactly once .
then , as in the proof of lemma [ deg2 ] , @xmath0 is fibered , and so by corollary [ fib ] , @xmath338 does not satisfy the @xmath139-lift condition , for a contradiction .
[ y_n ] if @xmath158 , then @xmath193 has incompressible boundary .
let @xmath358 .
we first prove the result in the case where @xmath359 .
+ + _ claim : _ the manifold @xmath360 has incompressible boundary . + + _
proof of claim : _ recall @xmath361 . by condition 2 of definition [ m - lift ] , the manifold @xmath362 has compressible boundary , and by condition 1 of definition [ m - lift ] , @xmath363 has incompressible boundary ; therefore , by theorem [ handle ] , the manifold @xmath364 $ ] has incompressible boundary .
suppose for some @xmath365 $ ] , that @xmath366- \partial \mathcal{e}_i$ ] has incompressible boundary , where @xmath367 . by condition 3 of definition [ m - lift ] , there is a compression disk for @xmath368 which is also a compression disk for @xmath366- \partial \mathcal{e}_{i-1}$ ] .
therefore , by theorem [ handle ] , @xmath369 - \p \mathcal{e}_{i-1}$ ] has incompressible boundary . by induction on @xmath207 , it follows that @xmath370 $ ] has incompressible boundary .
+ now , suppose @xmath371 . and proceed by induction on @xmath3 .
suppose @xmath193 has incompressible boundary .
the manifold @xmath372 is obtained from @xmath193 by adding a 1-handle @xmath373 to @xmath374 , and then attaching a 2-handle @xmath375 .
we claim that @xmath376 has incompressible boundary .
suppose otherwise , and let @xmath279 be a compressing disk . since @xmath193 has incompressible boundary , the maximal compression body for @xmath377 has a unique disk system consisting only of @xmath378 .
therefore if @xmath279 is non - separating , it is isotopic to @xmath378 ; then @xmath379 can be isotoped off of @xmath380 in @xmath381 , contradicting lemma [ nohomotopy ] .
suppose @xmath279 is separating , so it separates off a solid torus @xmath382 , with @xmath383 . since @xmath384 , we have @xmath385 .
so @xmath372 contains the punctured lens space @xmath386 $ ] . by lemma [ nohomotopy ] , this lens space can not be @xmath387 .
so @xmath193 is not irreducible , contradicting lemma [ irred ] .
this completes the proof that @xmath376 has incompressible boundary .
then by theorem [ handle ] , @xmath372 has incompressible boundary .
( of theorem [ cyclic ] ) assume @xmath0 satisfies the hypotheses of theorem [ cyclic ] .
then by lemmas [ x_n ] and [ y_n ] , @xmath43 contains an incompressible closed surface @xmath161 , which is not parallel into @xmath163 by lemma [ nonparallel ] .
( of theorem [ surgery ] ) suppose the hypotheses of theorem [ surgery ] are satisfied .
let @xmath388 be the number of boundary components of @xmath43 .
then it is easy to see that @xmath388 is equal to the largest common divisor of @xmath3 and @xmath389 ( recall from the proceeding paragraph of theorem [ surgery ] that @xmath150 is a finite integer ) .
let each boundary component of @xmath195 have the coordinate basis induced from the basis @xmath390 of @xmath144 .
let @xmath391 denote the closed manifold obtained by dehn filling each component of @xmath195 with slope @xmath392 .
then it is easy to check that @xmath152 is cyclically covered by @xmath393 .
we have shown that @xmath43 contains an incompressible surface @xmath161 . when @xmath394 as well , there are @xmath389 successive lifts of @xmath142 contained in @xmath395 , which implies that @xmath161 has an annular compression to each component of @xmath195 , with slope @xmath396
. since @xmath397 , then by repeatedly applying theorem 2.4.3 of @xcite @xmath388 times , we see that @xmath161 remains incompressible in @xmath393 .
also , by @xcite , @xmath393 is irreducible .
hence @xmath152 is virtually haken .
a. checking that snappea presentations are geometric + given an algebraic word on @xmath398 and @xmath399 , we need to know if it can be represented by a simple closed curve in the boundary of a genus 2 handlebody . to do this , we attempt to draw a heegaard diagram .
so we start with four disks in the plane , corresponding to @xmath400 and @xmath401 .
the word @xmath197 is represented by a collection of edges connecting these disks , and we need to find a representation which embeds in the plane . to program this on the computer , we start placing edges on the graph one by one , as indicated by the word @xmath197 .
when placing an edge , the intial point is determined from the previous step , but there may be a choice of terminal point .
however , it is possible to keep track of the choices which are made , and if an impossible situation is arrived at , we retreat to the previous choice , and change it . in this way
, the computer found for every given word , a geometric representation .
+ + b. checking that the @xmath139-lift condition holds + the only non - elementary step in checking the @xmath139-lift condition algorithmically is to find , for a given collection of loops in the fundamental group of a handlebody @xmath35 , a specific compressing disk for @xmath35 which is disjoint from the loops .
an algorithm for this was given by whitehead ( @xcite , or see @xcite ) , which we implemented on gap .
note that whitehead s algorithm allows one to determine the existence of a compressing disk in polynomial time in the length of the word ; however , to construct the disk explicitly requires exponential time . for our application
, we are saved this difficulty , since for each compressing disk @xmath279 we only need to compute @xmath402 \in h_1(\partial h)$ ] .
this allows the algorithm to run in polynomial time .
w. magnus , a. karrass and d. solitar , _ combinatorial group theory : presentations of groups in terms of generators and relations _ , interscience publishers [ john wiley & sons , inc ] , new york - london - sidney , 1966 . | we use heegaard splittings to give a criterion for a tunnel number one knot manifold to be non - fibered and to have large cyclic covers .
we also show that such a knot manifold ( satisfying the criterion ) admits infinitely many virtually haken dehn fillings . using a computer
, we apply this criterion to the 2 generator , non - fibered knot manifolds in the cusped snappea census . for each such manifold @xmath0 ,
we compute a number @xmath1 , such that , for any @xmath2 , the @xmath3-fold cyclic cover of @xmath0 is large . |
direct detection of gravitational waves appears likely in the near future .
existing terrestrial long - baseline interferometers , such as the laser interferometer gravitational - wave observatory ( ligo ) and virgo , have achieved their design sensitivity @xcite .
next - generation interferometers now under construction are expected to detect tens of events per year , if contemporary estimates of compact binary coalescence rates are correct @xcite .
searches for periodic sources have the advantage of integrating over long observation times to increase the signal - to - noise ratio .
the most likely periodic sources detectable by terrestrial interferometers are rapidly rotating , slightly deformed neutron stars @xcite .
they emit at the spin frequency @xmath4 and its first harmonic @xmath5 @xcite .
x - ray timing measurements find 0.3 khz @xmath6 0.6 khz for low - mass x - ray binaries @xcite , placing these sources squarely in the ligo - virgo band . directed searches for electromagnetically observed targets @xcite and blind , all - sky searches for unknown sources @xcite have both been reported .
targeted searches for known pulsars use the radio ephemeris to guide the search and reduce computational expense by assuming the electromagnetic and gravitational wave phases track each other closely .
one search for the crab pulsar allowed for a mismatch of up to one part in @xmath7 between twice the radio pulse frequency and the gravitational wave frequency @xcite . targeted searches for the crab and vela pulsars have set direct upper limits on the wave strain of @xmath8 and @xmath9 respectively , beating the indirect spin - down limits inferred from radio observations @xcite .
a search of 78 pulsars using data from the third and fourth science runs of the ligo and geo 600 detectors set upper limits on the wave strain @xmath10 and ellipticity @xmath11 , the tightest of which are @xmath12 for psr j1603 - 7202 and @xmath13 for psr j2124 - 3358 @xcite .
blind , all - sky searches address more sources than targeted searches but are expensive computationally as they cover a larger parameter domain to keep track of the unknown frequency evolution .
a number of ligo all - sky searches for periodic sources have been conducted @xcite , some of which leverage the distributed computing power of the einstein@home project @xcite . frequently , they are based on a maximum - likelihood detection statistic , called the @xmath0-statistic@xcite . the @xmath0-statisticis computed by coherently integrating over the observation time @xmath14 , assuming a biaxial neutron star at a specific spin frequency and sky position . @xcite generalized the @xmath0-statisticto apply to multiple detectors and sources .
the computational expense of a fully coherent search for unknown sources becomes prohibitive as @xmath14 increases .
semi - coherent methods address this problem , dividing the observation time into intervals , which are individually searched coherently but combined incoherently .
semi - coherent methods are more sensitive than a fully coherent search for the same computational cost if the parameter space is large @xcite .
@xcite reported results of a semi - coherent all - sky search for periodic sources .
they described and compared three semi - coherent methods : stackslide , powerflux , and hough number count @xcite . for binary sources , @xcite derived the @xmath1-statistic , which incoherently combines individual @xmath0-statisticvalues with a comb template that matches the orbital sideband structure . in this paper
, we propose a new method for enhancing gravitational wave searches using a second - pass method known as higher criticism .
higher criticism was suggested originally by john tukey and developed further by @xcite , who proved that it has some mathematical optimality properties in the case of a specific noise distribution .
the ultimate goal of any gravitational wave search is to identify an individual object as a definite source .
higher criticism is not suited for this purpose .
it is designed to search against a known background for a group of sparse signals too weak to be detected individually .
higher criticism detects the presence of the group but can not reliably estimate the waveform .
we discuss applications of higher criticism to gravitational wave detection , focusing on the @xmath0-statistic@xcite as a case study . in general , we find that higher criticism is significantly more robust than other methods ; it ignores some information that other detection statistics incorporate , so may be less sensitive , but makes fewer assumptions about the form of the signal , especially important where the signal evolves during the search .
the paper is structured as follows . in section [ sec : hc ] , we define the higher criticism statistic , compare the detectability of signals with different amplitudes and sparsities , and compute detection thresholds . in section
[ sec : fstat ] , we briefly review the @xmath0-statistic . in section [ sec :
binary ] , we apply a form of higher criticism to targeted searches for a binary pulsar and compare its performance with the @xmath1-statistic .
we discuss how higher criticism can accommodate phase wandering in section [ sec : wandering ] .
while we apply higher criticism specifically to periodic gravitational wave searches in this paper , it is a general method with other applications , including detecting non - gaussianity in wilkinson microwave anisotropy probe ( wmap ) data @xcite .
the higher criticism statistic was introduced by @xcite to test the hypothesis that @xmath15 independent and identically distributed ( i.i.d . )
samples @xmath16 come from the same zero - mean gaussian distribution @xmath17 , against the alternative that a small fraction of them have a nonzero mean @xmath18 , that is , to test : @xmath19 for example , @xmath20 could correspond to a situation where all samples are just background noise with no signal ( @xmath21 ) , whereas in @xmath22 a fraction @xmath23 of samples contain signal ( @xmath24 ) .
higher criticism is particularly designed for the very challenging situation where the signal is sparse ( @xmath25 ) and weak ( @xmath26 is so small that the largest extreme values under @xmath22 are essentially the same as those under @xmath20 ) .
although we introduce the statistic in the context of gaussian noise , it can be generalized easily to other distributions ( see section [ sec : generalhc ] ) .
the case of correlated , or non - white , noise has been treated by @xcite .
let @xmath27 denote a random variable with distribution @xmath28 .
the higher criticism statistic @xmath29 is computed from the @xmath15 p - values @xmath30 $ ] generated by the @xmath15 tests @xmath31 against @xmath32 , @xmath18 , @xmath33 .
it is defined by @xcite , @xmath34}{\sqrt{p_{(i)}[1 - p_{(i)}]}}~,\ ] ] where @xmath35 are the @xmath36 s sorted into increasing order .
the statistic @xmath29 rejects @xmath20 when @xmath37 , where @xmath38 is a threshold chosen so that @xmath39=p_{h_0}(\textrm{reject $ h_0$})\leq \alpha$ ] , where @xmath40 denotes the probability when @xmath20 is true .
[ our notation differs from @xcite in that we use the symbol @xmath41 for the threshold instead of @xmath10 to avoid confusion with the gravitational wave strain @xmath10 . ] in other words , the higher criticism test detects a signal when @xmath42 .
@xcite showed that , when @xmath20 is true , one has @xmath43 asymptotically as @xmath44 for all @xmath45 , and that if @xmath46 and @xmath15 is large enough , one can use the threshold @xmath47 .
however for more general @xmath48 and @xmath15 , this does not work well in practice , as we illustrate in section [ sec : detectthresh ] , where we discuss how to choose @xmath38 accurately as a function of @xmath15 and @xmath48 for finite @xmath15 .
next , we quantify more precisely how small @xmath23 and @xmath26 can be for the test based on @xmath29 to be able to distinguish @xmath20 from @xmath22 .
let @xmath49 and @xmath50 be two positive parameters .
@xcite study the properties of the @xmath29 statistic for @xmath26 and @xmath23 of the form @xmath51 these represent very difficult situations since , with @xmath49 small , @xmath26 of this form is smaller than the mean of the upper extreme statistics , and with @xmath50 large , the proportion of samples with non zero mean is very close to zero .
clearly , a detection is not possible if @xmath49 is very small and @xmath50 is very large at the same time .
@xcite showed that , as @xmath44 , there exists a function @xmath52 defined by @xmath53 such that @xmath20 and @xmath22 are distinguishable only if @xmath54 .
if @xmath55 , there does not exist a test that can distinguish @xmath20 from @xmath22 .
@xcite proved that , under some conditions , the @xmath29 statistic is able to distinguish @xmath20 from @xmath22 throughout the detectable region @xmath54 , and that it has full asymptotic power , that is , @xmath56 as @xmath44 ( here @xmath57 denotes the conditional probability when @xmath22 is true ) .
another test has also been considered in the literature , namely the neyman - pearson likelihood ratio test , but it is less attractive since it requires @xmath49 and @xmath50 to be known @xcite , whereas the test based on the higher criticism statistic does not .
interestingly , @xcite also show that , in addition to the detection boundary , there is a second region @xmath58 called the estimable region , where @xmath20 and @xmath22 can be distinguished and the mean @xmath26 can also be estimated consistently .
figure [ fig : detectregions ] , which can be found in @xcite , illustrates the detectable and estimable regions on the @xmath49-@xmath50 plane . in this section ,
we present briefly some results from monte - carlo simulations to illustrate the behavior of @xmath29 as a function of signal amplitude @xmath26 and sparsity @xmath23 for @xmath59 sample values drawn randomly from the @xmath22 distribution given by equation ( [ eq : h1 ] ) .
figure [ fig : hcdist - gaussian - vary_mean ] displays a histogram of @xmath29 values for @xmath60 trials with @xmath61 and @xmath62 .
figure [ fig : hcdist - gaussian - vary_sparsity ] displays a similar histogram with @xmath63 and @xmath64 .
the top left panel in both figures consists of samples drawn from the null distribution @xmath20 .
as @xmath26 or @xmath23 increases , so too does the average value of @xmath29 . in section
[ sec : hctheory ] , we defined @xmath29 in the context of gaussian noise .
as shown in @xcite , higher criticism can be generalized to other noise and signal distributions , @xmath65 and @xmath66 say , where @xmath26 describes the amplitude of the signal ( e.g. mean of the gaussian distribution in section [ sec : hctheory ] ) . in this more general setting we test to discriminate between the hypotheses @xmath67 in this context , the higher criticism statistic is computed as in equation ( [ eq : hcdef ] ) , but replacing the @xmath68 s by the ordered values of @xmath30 $ ] , @xmath33 , where @xmath27 is a random variable that has the distribution @xmath65 ; see also @xcite .
the detection boundary varies with the noise distribution . for example , when @xmath65 is the @xmath69 distribution that we use later , the proportion @xmath23 is defined as in equation ( [ eq : parameps ] ) , but the intensity of the signal is defined in terms of the noncentrality parameter @xmath70 of the @xmath69 distribution , through @xmath71 instead of the mean @xmath26 in equation ( [ eq : mu ] ) . with these definitions of @xmath49 and @xmath50 ,
the detection boundary is given by equation ( [ eq : detection_boundary ] ) , as in the normal case .
in this section , we discuss the choice of the detection threshold @xmath38 for @xmath29 .
as already noted earlier , the threshold satisfies @xmath72=p_{h_0}(\textrm{reject } h_0 ) \le \alpha$ ] , and under somewhat restrictive conditions , it can be approximated asymptotically by @xmath73 .
however , in finite samples and under more general conditions , this theoretical value @xmath73 is not a good approximation to @xmath38 . to illustrate this we conduct monte carlo simulations to determine the threshold @xmath38 for finite @xmath15 .
note that , when @xmath20 is true , the p - values are always independent and identically distributed according to the uniform distribution @xmath74 , and thus @xmath38 is independent of the specific noise distribution @xmath65 .
therefore it suffices to generate p - values from the @xmath74 distribution to compute the threshold .
we run @xmath75 noise - only simulations with @xmath76 to determine monte - carlo thresholds . for each , we generate @xmath15 p - values from @xmath74 and compute @xmath29 .
figure [ fig : cdf ] displays the cumulative distribution function of @xmath29 for the simulated noise samples .
threshold values obtained from the simulations by solving an empirical version of @xmath77 are listed in table [ tab : thresholds ] for @xmath78 and @xmath79 . for comparison ,
the theoretical value @xmath73 is also quoted . from this and figure
[ fig : cdf ] , we can see that this theoretical value generally does not work well in practice . in each case , a large fraction ( @xmath80 50% ) of noise - only trials fall above this threshold . in the remainder of this paper
, we commonly use the threshold @xmath81 as determined from monte - carlo simulations with @xmath82 ( 1% false alarm rate ) .
ccccc false alarm rate & + @xmath48 & @xmath83 & @xmath84 & @xmath85 & @xmath59 + 0.5 & 2.10 & 2.26 & 2.37 & 2.46 + 0.1 & 3.62 & 3.66 & 3.70 & 3.73 + 0.05 & 4.72 & 4.73 & 4.73 & 4.73 + 0.01 & 10.0 & 10.1 & 10.1 & 10.0 + theoretical threshold & & & & + @xmath86 & & & +
the @xmath0-statisticis an efficient detection statistic for periodic gravitational waves based on maximum likelihood .
a number of blind and targeted searches for periodic sources have been carried out using it and its relatives @xcite .
we use the notation @xmath87 to represent the noncentral ( central if @xmath88 ) @xmath69 distribution with @xmath89 degrees of freedom and noncentrality parameter @xmath90 . in the absence of a signal , and assuming stationary , gaussian noise , @xmath91 is distributed according to a @xmath92 distribution , while in the presence of a signal , @xmath91 obeys a @xmath93 distribution @xcite . in terms of the waveform
, one writes @xmath94 where @xmath95 is the spectral noise density of the interferometer , @xmath14 is the total observation time , and @xmath96 is the gravitational wave strain at the detector , which depends on the beam pattern functions @xmath97 and @xmath98 . for
a biaxial star , the strains @xmath99 and @xmath100 depend on the wobble angle @xmath101 between the rotation and symmetry axes , the inclination angle @xmath102 between the rotation axis and line of sight , right ascension @xmath103 , declination @xmath104 , polarization angle @xmath105 , and the characteristic strain amplitude @xmath106 where @xmath11 is the ellipticity , @xmath107 is the moment of inertia , and @xmath108 is the distance to the star @xcite .
we obtain an approximate relation between @xmath70 and @xmath109 by averaging equation ( [ eq : noncentrality_paramater ] ) over all relevant angles @xcite , @xmath110 with the assumption that the spectral noise density is achromatic , viz .
@xmath111 . equation ( [ eq : rho2h0relation ] ) allows us to translate detection limits on @xmath70 into limits on @xmath109 throughout the rest of the paper . for a known source
, the amplitude of the average signal that can be detected coherently with a 1% false alarm rate and a 10% false dismissal rate is @xcite @xmath112 equation ( [ eq : ligothresh_averaged ] ) is often seen with a factor of 11.4 instead of 15.6 ( e.g. , * ? ? ?
the 11.4 factor is obtained assuming the wobble angle @xmath113 , whereas we have averaged over @xmath101 to arrive at equation ( [ eq : ligothresh_averaged ] ) .
for an unknown source , we search across a large number of bins and @xmath109 must be larger than the value in equation ( [ eq : ligothresh_averaged ] ) to rise above the background .
@xcite derived the @xmath0-statisticfor multiple detectors or multiple sources and found that combining the @xmath0-statisticvalues from multiple sources increases their detectability as long as the sources have a squared signal - to - noise ratio greater than one fifth of the brightest source .
for example , of the known millisecond pulsars , it is advantageous to combine the five strongest ones but no more .
semi - coherent searches divide the observation time into @xmath114 intervals of length @xmath115 .
the @xmath0-statisticis computed for each time interval @xmath116 and combined to obtain @xmath117 as each @xmath118 follows a @xmath119 distribution , @xmath120 follows a @xmath121 distribution .
searches for periodic sources require detailed knowledge of the frequency evolution . for pulsars ,
the radio ephemeris is used to guide the search .
any difference between the radio and gravitational wave phases causes significant problems . for unknown sources
, we have even less information and are powerless to correct for phase wandering .
recently , @xcite proposed a `` phase - relaxed '' @xmath0-statisticfor an all - sky search , modifying the semi - coherent detection statistic to accommodate a phase offset @xmath122 for each coherent time interval @xmath116 , while preserving a monochromatic phase model overall .
in this paper we outline a new approach : applying higher criticism on a second pass to reanalyze the results of gravitational wave searches .
higher criticism relies on detailed knowledge of the background noise distribution .
the detection statistics discussed so far assume stationary , gaussian noise .
realistic detector noise is neither stationary nor gaussian .
we discuss this key point further in section [ sec : discussion ] . for now , we persevere with the common , simplifying assumption of stationary , gaussian detector noise in order to assess the overall viability of higher criticism , noting that modified forms of higher criticism have been constructed for correlated noise @xcite or when the noise distribution is imperfectly known @xcite .
low - mass x - ray binaries ( lmxbs ) are accreting neutron stars in a binary orbit with a low mass companion .
x - ray pulsations and burst oscillations place the spin frequencies of discovered objects in the range 270 hz @xmath123 619 hz @xcite , well below the centrifugal break - up frequency @xcite .
gravitational radiation is one effective means to balance the accretion torque and stall spin up @xcite . in this section ,
we compare a second - pass search with higher criticism to the side - band @xmath1-statisticalgorithm proposed by @xcite .
the gravitational wave signal from a neutron star in a binary has its frequency modulated into sidebands with spacing @xmath124 , where @xmath125 is the orbital period @xcite .
therefore , an @xmath0-statisticsearch sees a large number of relatively weak signals across many frequency bands . for a neutron star with intrinsic gravitational wave frequency @xmath126
, @xmath127 obeys a @xmath128 $ ] distribution , where the frequency dependent noncentrality parameter can be written @xcite @xmath129 with @xmath130 , where @xmath131 represents the largest integer less than @xmath132 , @xmath133 is the @xmath15-th order bessel function of the first kind , @xmath134 is the light crossing time of the orbital radius projected onto the line of sight , @xmath135 is the frequency of the @xmath15th sideband around a source with frequency @xmath126 , @xmath136 is the source signal - to - noise ratio squared [ see equation ( [ eq : noncentrality_paramater ] ) ] , and @xmath137 is the fourier transform of the window function @xmath138 , which equals 1 or 0 when the detector is on or off respectively . assuming no gaps in the data
, one finds @xmath139 with the window function defined as in @xcite , an extra factor of @xmath140 is included in equation ( [ eq : windowfunction ] ) so that @xmath1 is dimensionless .
the @xmath1-statisticsums incoherently the @xmath0-statisticpower in the orbital sidebands @xcite as follows : @xmath141 for noise only , @xmath91 is distributed according to a @xmath119 distribution .
hence , @xmath1 , which is the sum of @xmath142 @xmath127 values , obeys a @xmath143 distribution , where @xmath144 is the total number of sidebands @xcite . in the presence of a signal
, @xmath145 is distributed according to a noncentral @xmath128 $ ] distribution with a frequency dependent noncentrality parameter @xmath146 and @xmath1 follows a noncentral @xmath147 distribution with noncentrality parameter @xcite @xmath148
how does a @xmath1-statisticsearch , which assumes that @xmath125 ( and hence the sideband locations ) are known , perform in comparison with a second - pass higher criticism search over @xmath0-statisticor @xmath1-statisticvalues ?
we run monte - carlo simulations to answer this question .
we simulate a targeted search for sco x-1 , similar to the one described in sammut et al .
( 2013 , in preparation ) .
the search parameters are @xmath149 s , @xmath150 s , @xmath151 days , and source frequency @xmath152 hz ( assumed ) .
we search over the range of frequencies @xmath153 hz with frequency bin spacing @xmath154 hz , corresponding to @xmath155 frequency bins in total .
the signal is modulated into @xmath156 sidebands separated by @xmath124 .
@xmath156 depends on the signal frequency , with @xmath157 for 100 hz @xmath158 1000 hz .
the width in frequency space of the entire sideband structure ( the comb ) is 0.03 hz @xmath159 0.27 hz .
let us apply higher criticism to the problem .
we take advantage of the relatively narrow sideband structure and divide the frequency domain @xmath160 $ ] into @xmath161 windows of equal width @xmath162 , each containing @xmath163 frequency bins , viz .
@xmath164 $ ] , @xmath165 $ ] , @xmath166 $ ] , ... ,
@xmath167 $ ] , @xmath168 $ ] . the width @xmath162 is chosen to be twice the maximum width of the sidebands , @xmath169 , and windows overlap by 50% to ensure that the entire signal is contained entirely within a single window .
a more sophisticated method would vary the width with frequency as the sideband width increases from 0.03 hz for @xmath170 hz to 0.27 hz for @xmath171 hz , however , for simplicity , we construct our search windows based on the maximum sideband width at @xmath172 hz , obtaining @xmath173 hz , @xmath174 , and @xmath175 .
the data are synthesized as follows .
first , the noncentrality parameter is computed for each frequency bin .
the signal is modulated into sidebands and the frequency dependent noncentrality can be written ( sammut et al .
2013 , in preparation ) , @xmath176 we simplify this equation assuming @xmath177 ( true in the limit @xmath178 ) and assume the @xmath179 all have similar amplitude , which we absorb into the constant @xmath136 .
additionally , we adjust the frequency of each sideband @xmath180 to be equal to the corresponding closest frequency bin , @xmath181 , using @xmath182 to represent the frequency bin closest to frequency @xmath183 , neglecting the potential loss of amplitude caused by the mismatch between @xmath180 and @xmath181 . we denote the frequency on the @xmath116-th frequency bin @xmath184 , using the ` prime ' notation to differentiate between frequency bins and the exact frequency of the signal @xmath126 or the sidebands @xmath180 .
the simplified noncentrality parameter used to generate the @xmath0-statisticfor each frequency bin @xmath184 is @xmath185~,\ ] ] where @xmath186 is the standard kronecker delta rebranded so that we can see the subscripts clearly .
equation ( [ eq : simplerho2 ] ) corresponds to @xmath187 equal to @xmath136 for the frequency bin closest to each of the true sideband locations and zero elsewhere .
we generate synthetic @xmath127 values drawn from the @xmath188 $ ] distribution , for each frequency bin @xmath184 . to compute @xmath1-statisticvalues , we construct a comb centered at each frequency bin @xmath184 assuming @xmath189 , @xmath190~,\ ] ] with @xmath191 and @xmath192 .
the @xmath1-statisticis computed by convolving the @xmath193 with @xmath194 @xcite @xmath195 we compute the higher criticism statistic as described in section [ sec : generalhc ] for the @xmath127 and the @xmath1 values contained in a given search window and call the results @xmath196 and @xmath197 respectively . the noise distribution for @xmath127 is @xmath92 .
it is more complicated for @xmath1 , as the number of sidebands @xmath198 depends on the frequency @xmath184 .
hence , the p - values for each @xmath199 are calculated with respect to the @xmath200 noise distribution ( four degrees of freedom for each sideband summed ) . to compare the performance of the three detection statistics @xmath1 , @xmath196 and @xmath197 , we compute thresholds with a false alarm rate @xmath201 .
we have @xmath15 @xmath127 and @xmath1 values , one for each frequency bin , and @xmath161 @xmath196 and @xmath197 values , one for each search window .
thresholds are found by insisting that noise - only data fall below the threshold in every bin or window with probability at least @xmath202 [ e.g. , @xcite ] , @xmath203\right\}^{n } \ge 1 - \alpha~,\ ] ] @xmath204\right\}^{n_\mathrm{w } } \ge 1 - \alpha~,\ ] ] where @xmath205 is the @xmath1-statisticthreshold .
the @xmath29 threshold @xmath206 is determined from monte - carlo simulations , as described in section [ sec : detectthresh ] .
we integrate the @xmath1-statisticprobability distribution to find @xmath205 and use the result to compute the threshold noncentrality parameter @xmath207 for false dismissal rate @xmath208 from @xmath209 @xmath210 = \delta~,\ ] ] where @xmath211 is the probability density function of the @xmath87 distribution and @xmath212 is the maximum number of sidebands for all @xmath184 in the window . to simulate the gravitational wave search , we choose @xmath213 and draw @xmath127 values from the @xmath188 $ ] distribution for each frequency bin @xmath184 .
the search windows are constructed to ensure the entire sideband structure lies completely within a single window .
sidebands are also present in adjacent windows , however we focus only on a single window , assuming ( conservatively ) that all other windows contain only noise .
equation ( [ eq : hcthresh ] ) makes the same assumption when computing the detection thresholds against which @xmath196 and @xmath197 are compared .
we conduct a 100 monte - carlo simulations for each @xmath213 and find the fraction where the computed statistic lies above its threshold value , which we call the detection rate .
figure [ fig : binary ] displays detection rates for @xmath1 , @xmath196 , and @xmath197 as functions of increasing wave strain @xmath214 , where @xmath215 is defined as the wave strain corresponding to @xmath216 . for @xmath208 ( detection rate of 90% ) ,
@xmath197 detects wave strains 6% smaller than @xmath1 alone and @xmath196 detects wave strains @xmath217 times greater than @xmath1 .
additionally , higher criticism on @xmath127 compared to @xmath127 provides a advantage comparable to @xmath197 over @xmath1 ( not displayed in figure [ fig : binary ] ) .
the results in figure [ fig : binary ] assume perfect knowledge of the binary orbital period @xmath125 . in summary
, therefore , second - pass higher criticism on @xmath1 slightly boosts the sensitivity of the @xmath1-statistic , while second - pass higher criticism on @xmath127 is not competitive .
both higher criticism statistics , in particular @xmath196 , are more robust than @xmath1 , as demonstrated in the next section .
one significant advantage of higher criticism is its robustness when applied to signals that differ from the assumed form , e.g. due to phase wandering . in this section ,
we explore the performance of the statistics @xmath1 , @xmath218 , and @xmath196 when the form of the signal remains unaltered but the true signal parameters differ from those assumed in the search . * * for a comb search targeting a binary source , the orbital period @xmath125 determines the sideband spacing of the @xmath0-statisticin the frequency domain .
an error @xmath219 between the assumed and true value of @xmath125 produces a comb that does not coincide with the actual sidebands . the @xmath1-statisticsums noise at each sideband whose frequency does not match a true sideband location , lowering the sensitivity .
in contrast , @xmath196 only assumes that all sidebands are located within a particular frequency window .
it does not rely on knowledge of the precise location of each sideband and consequently is unaffected by an error in @xmath125 . both @xmath197 and the @xmath1-statisticlose sensitivity as @xmath219 increases
, however one would expect the relative advantage of @xmath197 compared to @xmath1 is expected to remain .
observations of the orbital period of sco x-1 have reported conflicting values for @xmath125 .
@xcite measured @xmath220 s from archival optical observations .
recently , @xcite measured @xmath221 s with the rossi x - ray timing explorer , but did not observe any signification periodicity at or near the previously reported value of 68023.8 s. other lmxbs have larger uncertainty in their orbital parameters [ e.g. , see tables 24 in @xcite ] . in light of the potential uncertainty in @xmath125 ,
the robustness of higher criticism is valuable .
we repeat the simulations described in section [ sec : binaryhc ] with the addition of an error @xmath219 between the assumed value of @xmath125 used to construct the comb @xmath194 , and the true value used to generate the noncentrality parameters @xmath146 for the @xmath0-statisticat every sideband .
figure [ fig : binaryrobustness ] displays detection rates with @xmath1 , @xmath196 , and @xmath197 as @xmath219 increases . as expected , @xmath196 is unaffected by @xmath219 because higher criticism does not use any information about the sideband locations .
@xmath1 and @xmath197 both lose sensitivity as @xmath219 increases , but the relative advantage of @xmath197 over @xmath1 increases from 6% when @xmath222
to 15% for @xmath223 s , 21% for @xmath224 s , and 29% for @xmath225 s.
the difference between @xmath1 and @xmath196 decreases from a factor of 7 when @xmath222 to 5 for @xmath223 s , 3 for @xmath224 s , and 2 for @xmath225 s. when @xmath222 , the @xmath1-statisticpeaks at the signal frequency @xmath126 . for @xmath226 ,
the error between the assumed and actual location of the sidebands accumulates to the point where the outermost sidebands no longer coincide with the assumed comb .
this reduces the total number of signal sidebands summed by the @xmath1-statistic , reducing its maximum amplitude ( at @xmath126 ) .
additionally , the @xmath1-statisticpeak broadens compared with @xmath222 ; this maximum amplitude requires less signal bins summed and can now be reached at a number of the central sidebands .
this situation moves closer to multiple , equal - strength signals , the problem for which higher criticism is designed , explaining the increasing advantage of @xmath197 over @xmath1 . in contrast , the @xmath1-statisticrequires a single source above threshold for detection .
interestingly , the sensitivity decrease of @xmath1 and @xmath197 stalls for 9 s @xmath227 1000 s , when the accumulated error at the outermost sidebands is at least the sideband spacing . at this point
, sidebands begin to overlap with other parts of the comb , shifted by integer multiples of the sideband spacing . as @xmath219 increases beyond 9 s
, the number of sidebands correctly located by the template continues to decrease but is compensated for by a corresponding increase in the number of integer overlap sidebands .
an accreting neutron star whose phase wanders in response to a variable accretion torque emits gravitational wave power in many frequency bins .
resampling is difficult as the phase model is usually unknown , e.g. there may be an offset between the radio / x - ray ephemeris and gravitational wave signal .
higher criticism can handle phase wandering robustly . in its simplest form , a semi - coherent @xmath0-statisticcombines @xmath114 coherent time intervals of equal length according to , @xmath228 where @xmath229 is the usual @xmath0-statisticfor the @xmath116-th interval .
this method assumes a monochromatic source or , with some modification , a source whose phase evolves in a known way .
@xcite described three semi - coherent methods to search for periodic gravitational waves in ligo data .
the methods can not be applied to a source whose phase wanders unpredictably , e.g. , due to accretion torque .
we apply higher criticism to reanalyze detection statistics from multiple time intervals with this situation in mind .
we define @xmath230 to be the number of frequency bins and @xmath231 to be the number of bins containing a gravitational wave signal , assuming for simplicity that all signal bins have equal amplitude . in practical terms
, this situation corresponds to @xmath231 distinct sources of similar strength , or @xmath231 orbitally modulated sidebands of a source in a binary system .
we also assume that @xmath230 and @xmath231 are the same in each time interval , therefore @xmath232 is constant .
we allow for the signal to jump arbitrarily between bins ( not necessarily adjacent ones ) from one time interval to the next , e.g. , due to phase wandering . for each interval
, we obtain a @xmath127 value for each of the @xmath230 frequency bins .
the @xmath231 signal bins follow a @xmath93 distribution , the remainder obey a @xmath92 distribution . combining all @xmath127 values for the @xmath114 intervals gives a total of @xmath233 values . converting from @xmath23 and @xmath15 to @xmath49 and @xmath50 through equations ( [ eq : parameps ] ) and
( [ eq : paramchi2 ] ) , we use the theoretical detection boundary [ equation ( [ eq : detection_boundary ] ) ] to find @xmath234 , the minimum noncentrality detectable with @xmath29 over @xmath114 time intervals , if @xmath231 of the @xmath230 bins contain signal .
this can be converted to a wave strain using equation ( [ eq : rho2h0relation ] ) .
the @xmath0-statisticthreshold noncentrality @xmath235 for a single interval is calculated similar to the @xmath1-statistic[equations ( [ eq : cstatthresh ] ) and ( [ eq : cstatfalsealarm ] ) ] @xcite , @xmath236 @xmath237 the semi - coherent @xmath0-statisticthreshold @xmath238 can also be calculated from equations ( [ eq : fstatthresh ] ) and ( [ eq : fstatfalsealarm ] ) , substituting @xmath239 , rather than 4 , degrees of freedom and assuming a monochromatic source .
the semi - coherent @xmath0-statisticcannot be applied to source whose phase wanders unpredictably and the threshold reverts to @xmath235 .
comparing detection thresholds for @xmath29 and @xmath0 is difficult because the detection boundary for @xmath29 , derived in the limit @xmath44 , is not equivalent to the @xmath0-statisticthresholds , computed for a specific false alarm and false dismissal probability .
furthermore , the @xmath29 detection boundary underestimates the threshold for finite @xmath15 ( see section [ sec : detectthresh ] ) .
we therefore recalibrate @xmath240 by a constant such that @xmath241 , thereby arranging that both @xmath29 and @xmath0 have equal power when @xmath242 . to test this recalibration ,
we conduct monte - carlo simulations to determine @xmath240 for @xmath201 , @xmath208 when @xmath243 for @xmath244 . we find that the monte - carlo @xmath240 agrees with @xmath235 computed from equations ( [ eq : fstatthresh ] ) and ( [ eq : fstatfalsealarm ] ) to within 1% . however , it is important to mind the difference between each threshold when interpreting the following results . figures [ fig : hcvsf_ngw ] and [ fig : hcvsf_n ] display @xmath70 required for detection with @xmath29 ( thick curves ) and @xmath0 ( thin solid and dashed curves ) for different combinations of @xmath114 , @xmath231 , @xmath230 , and @xmath201 , @xmath245 . figure
[ fig : hcvsf_ngw ] displays the thresholds @xmath240 and @xmath235 as functions of @xmath231 for @xmath246 = @xmath247 , @xmath248 , @xmath249 , @xmath250 .
figure [ fig : hcvsf_n ] displays @xmath70 as a function of @xmath114 for @xmath251 = @xmath247 , @xmath252 , @xmath249 , @xmath253 . for example , the search described in section [ sec : binary ] has @xmath254 and @xmath255 ( the number of orbital sidebands ) .
the semi - coherent @xmath256 is , in general , the most sensitive but as described above we are most interested in sources with unknown phase wandering , for which semi - coherent searches are not suitable . in all cases shown
, @xmath29 outperforms @xmath0 . in figure
[ fig : hcvsf_ngw ] , we observe reduction in @xmath240 by a factor of 1.5 to 2.9 for @xmath257 and 2.3 to 7.0 for @xmath258 , compared to @xmath240 when @xmath259 . in figure
[ fig : hcvsf_n ] , we observe reduction in @xmath240 by a factor of 1.4 to 2.2 for @xmath260 and between 2.1 and 3.0 for @xmath261 , compared to @xmath262 .
the robustness of @xmath29 described in section [ sec : wandering ] allows one to combine data from multiple time intervals to increased the sensitivity , even for a source whose phase wanders unpredictably . in principle
, one might improve detectability while remaining robust by including some simple information about the source frequency evolution , e.g. limiting wandering to some physically motivated range . in figures [ fig : hcvsf_ngw ] and [ fig : hcvsf_n ] , we assume no correlation , allowing the unlikely possibility that the frequency wanders across the entire range covered by the search . one simple method is to divide the frequency range into intervals wide enough to contain the source throughout the observation and calculate @xmath29 for each interval .
higher criticism is a recently formulated statistical method designed to detect the presence of a sparse collection of signals too weak to be detected individually . in this paper , we apply higher criticism as a second pass to reanalyze gravitational wave search statistics and explore the feasibility of applications in two contexts : a targeted binary search ( e.g. lmxb ) and a phase - wandering source ( e.g. , glitching pulsar ) .
one advantage of higher criticism is its robust nature ; it accommodates deviation from the expected phase evolution that hamper other search statistics . however , there is a trade - off .
higher criticism neglects some of the additional information used by other search methods which increase their performance under ideal conditions when the phase model is known . in section [ sec :
binary ] , we compare the performance of higher criticism and the @xmath1-statisticfor a targeted binary search .
the @xmath1-statisticis more sensitive than higher criticism applied to @xmath0-statisticvalues by a factor @xmath217 , however , higher criticism applied to @xmath1-statisticvalues gives a second - pass improvement of 6% over the @xmath1-statistic .
furthermore , higher criticism is more robust to an error @xmath219 between the true and assumed ( from observation ) binary orbital period . while the absolute sensitivity of both @xmath1 and @xmath197 decrease as @xmath219 increases , the performance of @xmath197 relative to @xmath1 increases to a sensitivity improvement of @xmath263 for @xmath223 s and @xmath264 for @xmath225 s. in section [ sec : wandering ] , we consider a phase wandering source .
the robustness of @xmath29 allows one to combine data from multiple time intervals and boost the sensitivity compared to a single interval , even for a unpredictable source . the noncentrality threshold decreases by a factor @xmath265 over 10 intervals and a factor @xmath266 over 100 intervals .
another candidate for higher criticism is an all - sky search for unknown periodic sources .
( we tested this in appendix [ sec : all_sky ] ) .
the results indicate that higher criticism provides no advantage over the @xmath0-statistic , the statistics perform similarly .
this stems from the distribution of source amplitudes : if neutron stars are uniformly distributed across the galaxy , the closest , strongest source dominates the detectability .
there is no advantage using higher criticism , which is designed to detect a group of signals , when there is effectively only one signal present . while there is reason to be optimistic about possible applications of higher criticism , we draw attention to a number of significant concerns .
chief among these , higher criticism relies on knowing the background noise distribution . in this paper
we adopted the common , simplifying assumption of stationary , gaussian detector noise .
in reality , detector noise is more complicated .
however , it is well studied and modified forms of higher criticism have been developed for correlated noise @xcite or when the noise distribution is imperfectly known @xcite .
additionally , higher criticism only detects the presence of a group of signals .
it can not directly identify an individual source , which remains the ultimate goal of gravitational wave detection .
however , given the cheap computational cost of applying higher criticism as a second pass to reanalyze already computed search statistics , and the advantages related to its robustness , it has the potential to complement and enhance existing and future searches , especially as a guide to where to look harder in parameter space for just - too - weak sources .
we thank l. sammut , v. dergachev , and members of the continuous wave search group of the ligo scientific collaboration for helpful discussions .
mfb was supported by an australian postgraduate award .
am , ad and ph were supported by grants and fellowships from the australian research council . 40 natexlab#1#1 , j. , abbott , b. p. , abbott , r. , et al .
2010 , , 722 , 1504 . 2010 , classical and quantum gravity , 27 , 173001 . 2011 , , 737 , 93 , b. , abbott , r. , adhikari , r. , et al .
2004 , , 69 , 082004 . 2005 , , 72 , 102004 . 2007 , , 76 , 082001 . 2007 , , 76 , 042001 . 2008 , , 77 , 022001 . 2008 , , 683 , l45 . 2009 , , 79 , 022001 , b. p. , abbott , r. , adhikari , r. , et al . 2009 , physical review letters , 102 , 111102 . 2009 , , 80 , 042003 . 2009 , reports on progress in physics , 72 , 076901 . 2012 , journal of instrumentation , 7 , 3012 , w. d. , schramm , d. n. , & truran , j. w. 1989 , , 339 , l25 , l. 1998 , , 501 , l89 + , p. r. , & creighton , t. 2000 , , 61 , 082001 , l. , jin , j. , & treaster , a. 2005 , , 362 , 826 , d. , morgan , e. h. , muno , m. p. , et al .
2003 , , 424 , 42 , g. b. , shapiro , s. l. , & teukolsky , s. a. 1994 , , 424 , 823 , c. 2011 , arxiv:1104.2938 , c. , gholami , i. , & krishnan , b. 2005 , , 72 , 042004 , c. , & schutz , b. f. 2005 , , 72 , 063006 delaigle , a. , & hall , p. 2009 , in perspectives in mathematical sciences i : probability and statistics , ed . b. r. n. sastry , m. delampady & t. rao ( world scientific publishing ) , 109138 delaigle , a. , hall , p. , & jin , j. 2011 , journal of the royal statistical society : series b ( statistical methodology ) , 73 , 283 , d. , & jin , j. 2004 , annals of statistics , 32 , 962 , e. w. , wright , e. l. , & liller , w. 1975 , , 195 , l33 , p. , & jin , j. 2010 , annals of statistics , 38 , 1686 , p. , krlak , a. , & schutz , b. f. 1998 , , 58 , 063001 , b. , sintes , a. m. , papa , m. a. , et al .
2004 , , 70 , 082001 , r. n. , hobbs , g. b. , teoh , a. , & hobbs , m. 2005 , , 129 , 1993 , a. , & melatos , a. 2012 , , 421 , 760 , a. , & payne , d. j. b. 2005 , , 623 , 1044 , c. , & woan , g. 2007 , classical and quantum gravity , 24 , 469 , g. , cutler , c. , & bildsten , l. 2000 , , 319 , 902 , k. w. , levine , a. m. , & rappaport , s. a. 2003 , , 115 , 739 , m. , & melatos , a. 2009 , , 395 , 1972 , a. l. , krishnan , b. , bildsten , l. , & schutz , b. f. 2008 , , 389 , 839 , k. 2012 , , 85 , 042003
one might ask whether higher criticism can be applied profitably to all - sky searches .
after all , from an estimated galactic population of @xmath267 neutron stars @xcite , only @xmath268 pulsars have been discovered as radio sources , of which @xmath269% have frequencies in the range where current interferometers are most sensitive , @xmath270 hz @xcite .
theories of neutron star quadrupoles suggest that only a handful of the observed pulsars have any chance of being detectable with current sensitivity limits @xcite , but many of the radio - quiet objects are much closer to the earth than their radio - loud brethren .
all - sky searches cover a range of frequencies at many sky locations , hoping to detect one of these unknown sources .
a number of all - sky searches have been carried out @xcite , but no detection has been announced .
the searches typically use the @xmath0-statistic , or related statistics .
hence they are candidates for reanalysis with higher criticism .
in contrast to the other applications in this paper , the sources targeted by an all - sky search have a power law ( non - uniform ) wave strain distribution .
we follow @xcite and assume neutron stars are spread uniformly throughout the galaxy .
then the distance distribution separates into two parts : a local uniform three - dimensional distribution up to the thickness of the galactic disk , and a uniform two - dimensional distribution beyond . taking the disk to be 600 pc thick and 10 kpc in diameter , the 3d and 2d
distributions correspond to @xmath271 pc and 300 pc @xmath272 5 kpc respectively , where @xmath108 is the distance from earth .
we further assume that all sources have the same intrinsic amplitude , so that @xmath109 at earth depends on distance alone . changing variables from distance @xmath108 to noncentrality parameter @xmath70 ,
[ see equation ( [ eq : noncentrality_paramater ] ) ; also @xcite for details ] we arrive at the distribution , @xmath273^{-5/2 } , & \textrm{(3d)}~ , \\
n_{\mathrm{2d } } [ \rho^2]^{-2 } , & \textrm{(2d)}~,\end{array}\ ] ] where @xmath274 and @xmath275 are normalization constants which depend on @xmath276 , @xmath95 and @xmath14 through equation ( [ eq : noncentrality_paramater ] ) . to illustrate
, we consider the same search parameters as a completed ligo all - sky search for periodic sources .
@xcite searched the 10 hr of ligo data with the best sensitivity from the second science run and computed the @xmath0-statisticfor @xmath277 frequency bins at each of @xmath278 sky locations , a total of @xmath279 @xmath0-statisticvalues .
we estimate the number of neutron stars with frequencies that fall in the range of the search ( 160 - 728.8 hz ) to be @xmath280 ( of an estimated galactic population of @xmath267 ) . under these assumptions ,
each of the @xmath281 frequency bins is equally likely to contain some signal , with @xmath282 .
we compute the monte - carlo higher criticism detection rate as a function of @xmath283 , where @xmath284 is the noncentrality threshold for a @xmath0-statisticdetection with 1% false alarm rate and 10% false dismissal rate [ see equations ( [ eq : fstatthresh ] ) and ( [ eq : fstatfalsealarm ] ) ] .
we can then compare the noncentrality of the brightest source , @xmath285 , required for detection with higher criticism with the equivalent quantity for a @xmath0-statisticsearch .
to save computation , we scale simulations from @xmath286 to @xmath287 by keeping parameters @xmath49 and @xmath50 constant in equations ( [ eq : parameps ] ) and ( [ eq : paramchi2 ] ) and converting from @xmath23 and @xmath70 for @xmath288 to @xmath289 and @xmath290 for @xmath291 .
signal values are drawn from a @xmath292 distribution with noncentrality parameter @xmath290 generated according to equation ( [ eq : rho2dist ] ) . to fix @xmath274 and @xmath275 we specify @xmath285 for a source at a distance of 10 pc .
the remaining @xmath0 values , containing only noise , are drawn from the @xmath293 distribution , and we compute @xmath29 for the @xmath0-statisticvalues .
for each @xmath70 , we run 100 simulations and the detection rate equals the fraction of simulations with @xmath29 or @xmath0 above their corresponding thresholds for detection . finally , we convert noncentrality @xmath70 to wave strain @xmath10 through equation ( [ eq : rho2h0relation ] ) to plot a more meaningful quantity .
figure [ fig : allsky ] displays the monte - carlo detection rate as a function of @xmath294 , where @xmath10 is the wave strain for a source at 10 pc , corresponding to @xmath285 , normalized by @xmath295 , the wave strain required for detection with the @xmath0-statisticat 10 pc . figure [ fig : allsky ] shows that higher criticism provides no advantage over the @xmath0-statistic ; they perform almost identically .
the fluctuations in figure [ fig : allsky ] relate to the randomly sampled source distribution . despite the large number of sources , @xmath296 in equation ( [ eq : rho2dist ] ) falls off too fast with @xmath70 and effectively only the brightest source , rather than the collection of sources ,
is detected . to test this we repeated the above simulations using only the single strongest source ( replacing the rest with noise ) and found the same result .
the similar performance of @xmath0 and @xmath297 agrees with section [ sec : wandering ] , where monte - carlo simulations found @xmath29 and @xmath0 had equal thresholds when only one source was present .
therefore , for an all sky search , higher criticism does not provide any benefit over a first pass @xmath0-statisticsearch . | we propose a new method of gravitational wave detection using a modified form of higher criticism , a statistical technique introduced by @xcite .
higher criticism is designed to detect a group of sparse , weak sources , none of which are strong enough to be reliably estimated or detected individually .
we apply higher criticism as a second - pass method to synthetic @xmath0-statisticand @xmath1-statisticdata for a monochromatic periodic source in a binary system and quantify the improvement relative to the first - pass methods .
we find that higher criticism on @xmath1-statisticdata is more sensitive by @xmath2% than the @xmath1-statisticalone under optimal conditions ( i.e. binary orbit known exactly ) and the relative advantage increases as the error in the orbital parameters increases .
higher criticism is robust even when the source is not monochromatic ( e.g. phase wandering in an accreting system ) .
applying higher criticism to a phase - wandering source over multiple time intervals gives a @xmath3% increase in detectability with few assumptions about the frequency evolution .
by contrast , in all - sky searches for unknown periodic sources , which are dominated by the brightest source , second - pass higher criticism does not provide any benefits over a first pass search . |
for an integer @xmath1 and a positive integer @xmath2 , let @xmath3 denote the set of integers @xmath4 . a finite collection of congruence classes @xmath5
is called a covering system in @xmath0 if each integer belongs to at least one congruence class .
this concept was first introduced by erds @xcite in 1950 , who constructed an infinite arithmetic sequence of odd integers not representable as @xmath6 ( @xmath7 a prime ) using a covering system .
if moreover , each integer belongs to exact one congruence class , we say it is an exact covering system .
given a number field @xmath8 , denoting the ring of algebraic integers of @xmath8 by @xmath9 .
for @xmath10 and a nonzero ideal @xmath11 of @xmath9 , @xmath12 , let @xmath13 denote the set of algebraic integers @xmath14 .
a set of congruences @xmath15 is called a covering system in @xmath8 if each element in @xmath9 belongs to at least one congruence class .
of course we call it an exact covering system if each element in @xmath9 belongs to exact one congruence class .
let @xmath5 be an exact covering system in @xmath0 .
we always assume @xmath16 . in 1971
, m. newman @xcite proved that @xmath17 where for an integer @xmath18 , @xmath19 in @xcite , m. a. berger , a. felzenbaum and a. s. fraenkel called a modulus @xmath20 division maximal if for @xmath21 , @xmath22 implies @xmath23 , then each division maximal @xmath20 must be repeated at least @xmath24 times . in @xcite ,
the same authors gave improved result that each modulus @xmath20 must be repeated at least @xmath25 times , where for an integer @xmath18 , @xmath26 this result is better as for a division maximal @xmath20 , @xmath27 implies @xmath28 is a proper divisor of @xmath20 , so @xmath29 .
for exact covering systems in number field @xmath8 , very recently , in @xcite , s. kim proved that any modulus of the largest norm must be repeated when @xmath8 is a certain quadratic field with additional conditions on the modulus ideals @xmath30 .
we are going to prove the analogous results in @xcite and @xcite for exact covering systems in any number field @xmath8 .
this paper is organized as follows . in section 2 ,
we introduce some notations about covering systems in number fields and s. kim s result @xcite . in section 3
, we ll introduce some definitions in @xcite , and then prove the analogous result of @xcite for any number field .
we ll prove the analogous result of @xcite for any number field in section 4 .
both results greatly generalize the result of s. kim in @xcite .
we always use the following notations and decompositions . let @xmath8 be a number field , @xmath9 its ring of integers and @xmath31 an exact covering system of @xmath9 .
let @xmath32 has a prime ideal decomposition @xmath33 where @xmath34 are prime @xmath9-ideals and @xmath35 .
each @xmath30 has decomposition @xmath36 with @xmath37 . for each @xmath34 , assume its norm @xmath38 .
here we define the norm @xmath39 of a nonzero ideal @xmath40 of @xmath9 as the cardinality of the finite ring @xmath41 .
now we introduce kim s results in @xcite .
the first two are about imaginary quadratic fields .
let @xmath42 be an imaginary quadratic field , @xmath15 an exact covering system of @xmath9 with @xmath43 .
if @xmath44 is principal , then it must be repeated .
let @xmath42 be an imaginary quadratic field with class number two and @xmath45 .
if @xmath15 is an exact covering system of @xmath9 , then the moduli can not be distinct .
kim also get a result about real quadratic fields .
let @xmath46 be a positive integer .
if @xmath15 is an exact covering system of @xmath9 with @xmath47 , where all the moduli are principal , then any modulus of the largest norm must be repeated .
in this section , we will prove the following result .
let @xmath15 be an exact covering of @xmath9 for any number field @xmath8 , and if @xmath30 is division maximal ( see the following definition 3 ) , then @xmath30 must repeat at least @xmath48 times . to prove the theorem 1
, we first introduce the important concept of parallelotope from @xcite . for @xmath49 with @xmath50 ,
define the lattice parallelotope or simply parallelotope @xmath51 where @xmath52 for @xmath53 .
if @xmath54 , then @xmath55 is also called the cube @xmath56 . given a parallelotope @xmath57 ,
let @xmath58 , an @xmath11-cell or simply cell @xmath8 of @xmath59 is a set of the form @xmath60 where @xmath61 for @xmath62 , @xmath63 for @xmath64 , @xmath65 is an arbitrary point in @xmath59 .
the set @xmath11 is called the index of @xmath8 , and denoted by @xmath66 . a partition @xmath67 of a parallelotope @xmath59 into cells
is called a cell partition of @xmath59 .
a cell @xmath68 is said to be subset minimal if @xmath69 , @xmath70 implies @xmath71 .
we also need the following lemma in @xcite .
let @xmath67 be a cell partition of @xmath72 into at least two cells and let @xmath73 be subset minimal .
put @xmath74 , then @xmath67 contains at least @xmath75 @xmath76-cells .
now we consider exact covering systems in number fields .
the main ideal is giving a map from an exact covering system to a cell partition of certain parallelotope .
in berger et al.s article , they used the fact that @xmath77 is cyclic and defined an addition operation on the corresponding parallelotope , but for an ideal @xmath11 , @xmath78 is not always cyclic . fortunately , we can use chinese remainder theorem to get what we need .
let @xmath79 be a complete representatives of @xmath80 .
we can define a bijective map @xmath81 by @xmath82 find some @xmath83 , then for any @xmath84 , @xmath85 is a complete representatives of @xmath86 .
we naturally extend @xmath81 to a map @xmath87 @xmath88 of course @xmath87 is also a bijective map .
now we can define a map @xmath89 for each @xmath13 to a point in @xmath72 , here @xmath90 and @xmath91 , i.e.@xmath92 assume for each @xmath93 , @xmath94 then define @xmath95 let @xmath96 be a set of @xmath9-ideals , an ideal @xmath97 is called division maximal if @xmath98 and @xmath99 implies @xmath100 .
the map @xmath101 is bijective .
taking @xmath102 as a union of some @xmath13 , then @xmath103 is a cell of @xmath55 and division maximal ideals correspond to subset minimal cells .
moreover if @xmath15 is an exact covering of @xmath9 , then @xmath104 is a cell partition of @xmath55 . by chinese remainder theorem ,
each @xmath13 one - to - one corresponds to , for @xmath105 ,
@xmath106 so @xmath89 is bijective .
we now prove @xmath103 is a cell of @xmath55 and the last claim is its trivial consequence . without loss of generality , we prove it for the case @xmath107 . according to the definition of @xmath108
, @xmath109 is a complete representatives of @xmath110 , for @xmath111 . if @xmath112 then @xmath113 so @xmath114where
@xmath115 then @xmath116 is a cell of @xmath55 . according to the above discussion , for @xmath117
, @xmath118 has index @xmath119 as @xmath120 so @xmath30 is division maximal @xmath121 is subset minimal .
now we can prove the analogous result .
it is described in the theorem 1 .
_ proof of theorem 1_.since @xmath122 , if @xmath123 , then @xmath124 . as @xmath125 , by lemma 1
, @xmath30 must repeat @xmath48 times .
then we complete the proof .
* remark*. from the theorem 1 , as each ideal @xmath30 with maximal norm is division maximal , so it must repeat at least @xmath48 times .
our result includes kim s @xcite .
in this section , we ll prove an improvement of the above result .
it is easy to see this bound is better than the above . for an ideal @xmath30 ,
if @xmath30 is not division maximal , then there exists @xmath128 such that @xmath129 , so @xmath130 , this bound is trivial .
if @xmath30 is division maximal , then for each @xmath131 with @xmath132 we have @xmath133 , so this bound is better . we need a new map to prove this result .
let @xmath134 , @xmath81 , @xmath135 just as the last section , now we define a map
@xmath136 @xmath137 @xmath136 maps an element of @xmath135 to an integer and the coefficient of @xmath138 corresponds to the coefficient of @xmath139 .
obviously @xmath136 is bijective .
now we can define a map @xmath140 for each @xmath13 to a point in @xmath141 , here @xmath142 .
assume for each @xmath93 , @xmath94 then define @xmath143 again @xmath140 is bijective .
it maps each @xmath144 to a special subset of @xmath145 . taking @xmath146 for example , just as the above section
, @xmath109 is a complete representatives of @xmath147 , for @xmath93 . if @xmath148 then @xmath149 then we have @xmath150 according to the definition of @xmath151 , @xmath152 where @xmath153 is some nonnegative integer with @xmath154 .
now we begin to prove the above theorem .
we first prove the special case that there exists some @xmath30 such that @xmath155 for all @xmath156 . without loss of generality
, we may assume @xmath157 is such an ideal , i.e. @xmath158 , for each @xmath93 .
@xmath160 is a partition of @xmath145 , here @xmath161 is defined as above .
@xmath162 is a single point , and each @xmath163 is a special subset . just as the proof in @xcite , we know that there are at least @xmath159 single points , so @xmath157 must repeat the same times .
let @xmath31 be an exact covering system in @xmath8 , we are going to get an exact covering with the above property from it . as @xmath134 is a complete representatives of @xmath80 , we let the representative of @xmath164 be just 0 , i.e. @xmath165 for all @xmath105 .
assume @xmath166 is a complete representatives of @xmath167 .
if @xmath168 is any ideal with @xmath169 , and @xmath168 has the following decomposition @xmath170 so @xmath171 . we define a subset @xmath172 of @xmath166 , @xmath173 @xmath172 includes these elements of @xmath166 which mod @xmath174 with the last @xmath175 coefficients 0 .
it is easy to see that @xmath172 is a complete representatives of @xmath176 . as @xmath177 ,
choose @xmath180 , then @xmath181 every element in @xmath144 is of the form @xmath182 with @xmath183 , and assume @xmath184 @xmath185 choose the element @xmath186 with @xmath187 and let @xmath188 , then @xmath189 so @xmath190 .
as @xmath191 , we need to prove @xmath192 . since @xmath193 for @xmath105 , if @xmath194 , as the last item of each side in above equation belongs to @xmath195 , so @xmath196 if @xmath197 , we just have @xmath198 so we always have @xmath199 for all @xmath105 , then @xmath192 , and we complete the proof . if @xmath207 , then @xmath208 with @xmath183 , @xmath209 . from the above lemma
, we have @xmath210 with @xmath211 , @xmath212 .
so we may assume @xmath213 if moreover , @xmath214 , then @xmath215 then @xmath216 , but each element of @xmath200 is in different class mod@xmath131 , so @xmath217 .
then @xmath218 and also @xmath219 , as @xmath15 is an exact covering system in @xmath8 , so @xmath220 .
therefore , @xmath206 is an exact covering system in @xmath8 .
_ proof of theorem 2_. now we prove theorem 2 .
let @xmath15 be an exact covering system .
then for a division maximal ideal @xmath131 , @xmath206 is an exact covering with each ideal is a divisor of @xmath131 . by lemma 3
, @xmath131 must repeat at least @xmath221 times . as @xmath222 is just @xmath223 , so @xmath224 then we complete the proof . | it is well known that in an exact covering system in @xmath0 , the biggest modulus must be repeated .
very recently , s. kim proved an analogous result for certain quadratic fields . in this paper , we prove that s. kim s result holds for any algebraic number field .
* keywords : * exact covering systems , lattice parallelotopes , chinese remainder theorem |
since we are exploiting the @xmath3 and @xmath4 appearance channels , the observables that we use in our numerical analysis are the number of expected electron neutrino and antineutrino events . for the central values of the already measured oscillation parameters ,
we have thus computed the expected number of electron and positron events @xmath38 and @xmath39 at the far detector located @xmath40 km off - axis at @xmath41 km , assuming positive or negative hierarchies , which are given by : @xmath42 where @xmath43 and @xmath44 are taken as perfectly known , @xmath45 denote the neutrino fluxes , and @xmath46 the cross sections .
the neutrino ( antineutrino ) flux , which peaks at @xmath47 , is integrated over a narrow @xmath48 energy window ( @xmath49 and @xmath50 ) .
the @xmath3 ( @xmath51 ) appearance probabilities in long baseline neutrino oscillation experiments , assuming the normal mass hierarchy , read @xcite : @xmath52 in the last expressions , @xmath53 and the coefficients @xmath54 and @xmath55 are determined by @xmath56 where @xmath57 , and @xmath58 denotes the index of refraction in matter , @xmath59 being the fermi constant and @xmath60 is a constant electron number density in the earth .
we denote the first , second and third terms in eqs .
( [ eqn : e_appear ] ) as the atmospheric , interference and solar terms , respectively . when @xmath10 is relatively large , the probability is dominated by the atmospheric term .
conversely , when @xmath10 is very small , the solar term dominates .
the interference term is the only one which contains the cp phase @xmath11 , and it is clear from eqs .
( [ eqn : e_appear ] ) that it is also the only one which differs for neutrinos and antineutrinos besides matter effects
. we can ask ourselves whether it is possible to unambiguously determine @xmath10 and @xmath11 by measuring the transition probabilities @xmath3 and @xmath61 at fixed neutrino energy and at just one baseline .
the answer is no . at fixed neutrino energy @xmath62 and baseline @xmath63 ,
if ( @xmath10 , @xmath11 ) are the values chosen by nature , the conditions : can be generically satisfied by another set ( @xmath65 , @xmath66 ) , known as the _ intrinsic degeneracy _
it has also been pointed out that other fake solutions might appear from unresolved degeneracies in two other oscillation parameters : 1 .
the sign of the atmospheric mass difference @xmath7 may remain unknown . in this particular case , @xmath67 @xcite . more specifically ,
the @xmath3 ( @xmath51 ) appearance probabilities for the inverted hierarchy are @xmath68 2 .
disappearance experiments only give us information on @xmath69 : is @xmath13 in the first octant , @xmath70 , or is it in the second one , @xmath71 ? . in terms of oscillation probabilities , @xmath72 @xcite .
q. r. ahmad _ et al .
_ [ sno collaboration ] , phys . rev .
lett . *
89 * , 011301 ( 2002 ) ; q. r. ahmad _ et al . _
[ sno collaboration ] , phys .
* 89 * , 011302 ( 2002 ) ; s. n. ahmed _ et al . _
[ sno collaboration ] , phys .
lett . * 92 * , 181301 ( 2004 ) .
s. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys . lett .
b * 539 * , 179 ( 2002 ) .
y. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys . rev .
* 81 * , 1562 ( 1998 ) .
k. eguchi _ et al . _
[ kamland collaboration ] , phys .
* 90 * , 021802 ( 2003 ) .
m. h. ahn _
[ k2k collaboration ] , phys .
lett . * 90 * , 041801 ( 2003 ) .
g. barenboim , l. borissov , j. lykken and a. y. smirnov , jhep * 0210 * , 001 ( 2002 ) .
a. aguilar _ et al . _
[ lsnd collaboration ] , phys .
d * 64 * , 112007 ( 2001 ) .
b. aharmim _ et al . _
[ sno collaboration ] , nucl - ex/0502021 .
t. araki _ et al .
_ [ kamland collaboration ] , phys .
* 94 * , 081801 ( 2005 )
. o. mena and s. j. parke , phys .
d * 69 * , 117301 ( 2004 ) .
y. ashie _ et al . _
[ super - kamiokande collaboration ] , hep - ex/0501064 .
e. aliu _ et al .
_ [ k2k collaboration ] , phys . rev .
lett . * 94 * , 081802 ( 2005 ) .
m. apollonio _ et al . _
[ chooz collaboration ] , phys .
b * 466 * , 415 ( 1999 ) .
k. anderson _
et al . _ , hep - ex/0402041 .
i. ambats _ et al . _
[ nova collaboration ] , fermilab - proposal-0929 .
d. s. ayres _
[ nova collaboration ] , hep - ex/0503053 .
l. bartoszek _ et al .
_ , hep - ex/0408121
. j. burguet - castell , m. b. gavela , j. j. gomez - cadenas , p. hernandez and o. mena , nucl .
b * 608 * , 301 ( 2001 ) .
o. mena and s. parke , phys . rev .
d * 70 * 093011 ( 2004 ) .
h. minakata , h. nunokawa and s. j. parke , phys .
b * 537 * , 249 ( 2002 ) .
h. minakata and h. nunokawa , jhep * 0110 * , 001 ( 2001 ) . v. d. barger , s. geer , r. raja and k. whisnant , phys .
d * 62 * , 013004 ( 2000 ) .
additional ( fake ) solutions should arise if @xmath73 , see g. l. fogli and e. lisi , phys .
d * 54 * , 3667 ( 1996 ) .
h. minakata , h. nunokawa and s. j. parke , phys .
d * 68 * , 013010 ( 2003 ) . in the following list of references we list some of the previous studies of the parameter degeneracies : j. burguet - castell , m. b. gavela , j. j. gomez - cadenas , p. hernandez and o. mena , nucl .
b * 646 * , 301 ( 2002 ) ; a. donini , d. meloni and p. migliozzi , nucl .
b * 646 * , 321 ( 2002 ) ; v. barger , d. marfatia and k. whisnant , phys .
d * 66 * , 053007 ( 2002 ) ; v. barger , d. marfatia and k. whisnant , phys .
b * 560 * , 75 ( 2003 ) ; p. huber , m. lindner and w. winter , nucl .
b * 654 * , 3 ( 2003 ) ; o. mena , hep - ph/0305146 ; d. autiero _ et al .
_ , eur .
j. c * 33 * , 243 ( 2004 ) ; j. burguet - castell , d. casper , j. j. gomez - cadenas , p. hernandez and f. sanchez , nucl .
b * 695 * , 217 ( 2004 ) ; a. donini , d. meloni and s. rigolin , jhep * 0406 * , 011 ( 2004 ) ; p. huber , m. lindner , m. rolinec , t. schwetz and w. winter , phys .
d * 70 * , 073014 ( 2004 ) ; o. yasuda , new j. phys .
* 6 * 83 ( 2004 ) ; a. donini , e. fernandez - martinez , p. migliozzi , s. rigolin and l. scotto lavina , nucl . phys .
b * 710 * , 402 ( 2005 ) ; a. donini , e. fernandez - martinez and s. rigolin , hep - ph/0411402 ; p. huber , m. lindner and w. winter , hep - ph/0412199 ; o. mena , mod .
a * 20 * 1 ( 2005 ) ; p. huber , m. maltoni and t. schwetz , phys .
d * 71 * , 053006 ( 2005 ) ; j. burguet - castell , d. casper , e. couce , j. j. gomez - cadenas and p. hernandez , hep - ph/0503021 ; m. ishitsuka , t. kajita , h. minakata and h. nunokawa , hep - ph/0504026 ; k. hagiwara , n. okamura and k. i. senda , hep - ph/0504061 . y. hayato _ et al .
_ , letter of intent , available at http://neutrino.kek.jp/jhfnu/ a. de gouvea , j. jenkins and b. kayser , hep - ph/0503079 .
h. nunokawa , s. parke and r. z. funchal , hep - ph/0503283 .
o. mena requejo , s. palomares - ruiz and s. pascoli , hep - ph/0504015 . m. albert _ et al .
_ , `` the fermilab computing farms in 2001 - 2002 , '' fermilab - tm-2209l , available at http://library.fnal.gov/archive/test-tm/2000/fermilab-tm-2209.pdf . | we explore the physics potential of the numi beamline with a detector located 10 km off - axis at a distant site ( @xmath0 km ) .
we study the sensitivity to @xmath1 and to the cp - violating parameter @xmath2 as well as the determination of the neutrino mass hierarchy by exploiting the @xmath3 and @xmath4 appearance channels .
the results are illustrated for three different experimental setups to quantify the benefits of increased detector sizes , proton luminosities and @xmath5 detection efficiencies .
july 25 , 2005 1.4 cm neutrino oscillations have been observed and robustly established by the data from solar @xcite , atmospheric @xcite , reactor @xcite and long - baseline neutrino experiments @xcite .
these results indicate the existence of non - zero neutrino masses and mixings .
the new parameters can be accommodated via the three neutrino pmns mixing matrix , the leptonic analogue to the ckm matrix in the quark sector .
neutrino oscillations within this scenario are described by six parameters : two mass squared differences throughout the paper . ]
( @xmath6 and @xmath7 ) , three euler angles ( @xmath8 , @xmath9 and @xmath10 ) and one dirac cp phase @xmath11 .
the standard way to connect the solar , atmospheric , reactor and accelerator data with the six oscillation parameters listed above is to identify the two mass splittings and the two mixing angles which drive the solar and atmospheric transitions with ( @xmath6 , @xmath8 ) and ( @xmath12 , @xmath13 ) , respectively .
the sign of the atmospheric mass splitting @xmath7 with respect to the solar doublet is one of the unknowns within the neutrino sector , i.e. we do not know if the neutrino mass spectrum is normal ( @xmath14 ) or inverted ( @xmath15 ) . the best fit point for the combined analysis of solar neutrino data @xcite together with kamland reactor data @xcite is at @xmath16 ev@xmath17 and @xmath18 . the 90% c.l .
allowed ranges of the atmospheric neutrino oscillation parameters obtained by the super - kamiokande experiment are @xcite , which lies within the best fit values for the super - kamiokande @xcite and k2k @xcite experiments . ] : @xmath19 the mixing angle @xmath10 ( which connects the solar and atmospheric neutrino realms ) and the amount of cp violation in the leptonic sector are undetermined . at present , the upper bound on the angle @xmath10 coming from chooz reactor neutrino data @xcite is : @xmath20 at the 90 % confidence level at @xmath21 .
this constraint depends on the precise value of @xmath22 , with a stronger ( weaker ) constraint at higher ( lower ) allowed values of @xmath23 .
future reactor neutrino oscillation experiments could measure the value of @xmath24 , as explored in detail in ref .
@xcite .
current neutrino oscillation experiments do not have any sensitivity to the cp - phase @xmath11 .
the experimental discovery of the existence of cp violation in the leptonic sector , together with the discovery of the majorana neutrino character , would point to leptogenesis as the source for the baryon asymmetry of the universe , provided that accidental cancellations are not present . the main aim of this paper is a careful study of the sensitivity to the currently unknown parameters mentioned above , that is to the small mixing angle @xmath10 , to the ordering of the neutrino mass spectrum , and to the amount of cp violation in the leptonic sector , which could be achieved by exploiting the numi neutrino beamline .
we thus concentrate on the numi beam potential exploited in an off - axis configuration as proposed by the no@xmath25a experiment @xcite .
the location of the far detector is at 10 km off - axis with a baseline of 810 km .
the mean neutrino energy is 2.3 gev .
we have considered two possible @xmath5 ( @xmath26 ) detection techniques .
first , the possibility of a 30 kton totally active low z tracking calorimeter detector , as the one considered in the revised no@xmath25a proposal @xcite .
the efficiencies of such a detector for @xmath5 ( @xmath26 ) identification is approximately @xmath27 and the background is typically two - thirds from electron ( anti)neutrinos in the beam produced from muon and kaon decays and one third from neutral current events faking electron neutrinos .
an alternative detection method explored here is the one provided by a liquid argon tpc detector , as the technique described in the flare letter of intent @xcite .
the efficiency of such a detector to identify the @xmath5 ( @xmath26 ) cc interaction is @xmath28 and the background is dominated by the intrinsic @xmath5 and @xmath29 components of the beam @xcite .
the statistics at the far detector is governed by the product of three parameters : the total number of protons on target ( @xmath30 ) , the detector mass ( @xmath31 ) , and the detector efficiencies to @xmath5 ( @xmath26 ) identification ( @xmath32 ) .
here we have studied the physics potential of three different experimental scenarios : * * small : * as a first step , we consider a * `` small '' * experimental setup without a proton driver .
the number of protons on target without a proton driver is @xmath33 per year .
we have considered three and a half years of running in each polarity ( i.e. three and a half years of neutrino and three and a half years of antineutrino data taking ) .
consequently , in this first scenario , the statistical figure of merit @xmath34 is equal to a total of @xmath35 in units of number of protons times kton .
this setup could be achieved , for instance , with the 30 kton totally active low z tracking calorimeter detector at @xmath27 efficiency described above or with a 9 kton liquid argon tpc detector at @xmath28 efficiency . *
* medium : * we study a possible upgrade of the * small * configuration , referred to as the * `` medium '' * experimental setup by increasing the statistical figure of merit by a * factor of five*. this statistics factor could be accomplished if the mass of the liquid argon detector is upgraded to 45 kton without a proton driver in three and a half years running in each polarity . or , equivalently , the same statistics could be achieved in the @xmath36a experiment running for four and a half years with a proton driver in each polarity ( i.e. four and a half years of neutrino and four and a half years of antineutrino data taking ) . with a proton driver
, the number of protons on target is assumed to be @xmath37 per year .
thus the statistics running with a proton driver for four and a half years is five times the statistics for running without a proton driver for three and a half years . ] . * * large : * the third experimental setup explored is a * `` large '' * experiment , in which * both * the initial detector mass and the total number of protons on target are increased by * a factor of five*. the * large * scenario could be obtained , for instance , by the combination of a 45 kt liquid argon detector with a proton driver running for four and a half years in both the neutrino and antineutrino modes . the ratio of the statistics in the three different experimental scenarios * small : medium : large * is therefore * 1:5:25*. in the next section we review the neutrino oscillation formalism , while numerical results are presented in the following sections . |
multiple protostellar systems ( mpss ) are common among every stage of protostellar evolution , but are tricky to decipher .
fragmentation is believed to be the main mechanism for forming mpss @xcite , but how and when fragmentation occurs is still a major question . to constrain the fragmentation process , information about the system s multiplicity , coevality , environment and outflows are needed . since older systems have suffered dynamical evolution @xcite , young early stage mpss , having many of the initial conditions almost intact , are the best targets for the task . in this letter we present the results of our work on
the much debated vla1623 , located in @xmath0 ophiuchus and generally considered the prototypical class 0 source ( @xcite , hereafter awb93 ) .
first detected through red- and blueshifted overlapped outflows @xcite , vla1623 s multiplicity has long been suspected from observation of multiple continuum sources ( @xcite ; @xcite ; @xcite , hereafter lmw00 ) , outflow misalingment ( @xcite ; @xcite , hereafter yc97 ) and two discreet precessing jets originating from the same region @xcite .
in contrast , the multiple continuum sources are argued to constitute a monopolar jet driven by vla1623a ( @xcite , hereafter ba97 ; @xcite ; @xcite , hereafter moa12 ) .
while vla1623a is agreed to be a real protostellar source , and was used by abw93 as the prime example to establish a new class of ysos - class 0 " , there is still debate on the nature of the other sources , mainly vla1623b ( @xcite , hereafter wkga11 ) .
we present the results of our work to determine the multiplicity of vla1623 and disentangle its outflows . our own submillimeter array
( sma ; @xcite ) observations and archival data were used , suplemented with data from literature .
our results show vla1623 to be a triple non - coeval protostellar system and find vla1623b to be a bona - fide protostar with an extremely young age .
we observed vla1623 with sma at 345ghz in the compact configuration for 8 hours on 2007 july 25 , with 2ghz correlator mode configured to observe @xmath1co ( 3 - 2 ) , c@xmath2o ( 3 - 2 ) and so ( 9 - 8 ) simultaneously with 870@xmath3 m continuum and polarization measurements .
all windows were set to have 128 channels , providing a velocity resolution of 0.704kms@xmath4 .
quasars 3c273 and 1733 - 130 were observed as the bandpass and gain calibrators .
a planet was not observed for unknown operational reasons .
this track s original purpose was polarization measurements , where lack of flux calibration is not a critical issue .
we performed standard data reduction for polarization data using miriad .
no significant polarization detection was made .
flux calibration is done with 1733 - 130 scaled to 1.31jy , value taken from the sma observer center , with measured uncertainty of 0.035jy .
imaging with natural weighting resulted in a beamsize of 236@xmath5141 .
we further observed vla1623 at 345ghz in the extended configuration with a 3-hr filler track on 2012 february 14 using 4ghz correlator mode . only the continuum emission is considered .
calibration was performed on idl mir package following the standard calibration pipeline .
3c84 , 1733 - 130 and callisto were used for bandpass , gain and flux calibration . calibrated
flux of 1733 - 130 is [email protected] , in agreement with the flux listed on sma observer center .
data imaging and analysis were performed with miriad using natural weighting resulting in a beamsize of 151@xmath5076 .
we compared the two datasets by binning vectorially averaged amplitudes , and when the tracks overlap , the flux level was comparable .
uncertainty in flux calibration is typically @xmath710% .
vla1623 was observed at 230ghz in the very extended configuration on 2007 june 11 and 2009 july 3 .
the 2009 data were presented by moa12 . here
we redo the data reduction and analysis .
the data quality of the 2007 data appears to be better than the 2009 data ; the system temperatures are 92@xmath624k and 130@xmath639k , and the atmospheric opacities are 0.06 and 0.1 for the 2007 and 2009 data , respectively .
both observations used the 2ghz correlator mode configured for @xmath1co ( 2 - 1 ) , c@xmath8o ( 2 - 1 ) , @xmath9co ( 2 - 1 ) and continuum .
the three windows with line emission in both datasets were set to have 512 channels each , giving a velocity resolution of 0.3kms@xmath4 .
the remaining windows were configured to have 32 channels each . for the 2007 data ,
the quasar 3c454.3 for bandpass , 1514 - 241 and 1622 - 297 for gain , and callisto for flux were used in calibration . for the 2009 data , bandpass and gain calibration
was done with 1924 - 292 and 1625 - 254 .
flux calibration using callisto caused the flux of vla1623b to rise 30mjy compared to the 2007 track .
we suspect a glitch occured during callisto s observation , so 1625 - 254 was used as a flux calibrator instead with 0.79jy taken from the sma observer center around the observing date , with measured uncertainty of 0.025jy . after scaling
, the flux from the two datasets are consistent within 11% .
miriad was used to separately calibrate each dataset and then combine during imaging assuming natural weighting in order to increase the signal - to - noise ratio in continuum and @xmath1co .
the resulting beamsize is of 061@xmath5043 .
only in the 2007 data c@xmath8o showed detection above 3@xmath10 and neither
dataset detected @xmath9co .
compact 870@xmath3 m data shows two continuum peaks ( figure [ figcont ] ) .
the unresolved stronger peak coincides with lmw00 s vla1623a & b. a weaker peak located 105 to the west , matches with ba97 s knot b. we will refer to this west peak with yc97 s notation vla1623w . hereafter we refer to vla1623a , vla1623b and vla1623w as a , b and w , respectively . extended 870@xmath3 m observations resolve a & b , but w is too weak to be detected in such short observations .
b appears brighter than a possibly due to b being more compact than a and that much of the emission from a , present in the compact data , has been resolved out ( figure [ figcont]c ) .
the extended 1.3 mm continuum observations resolved three peaks coinciding with the positions of a , b and w. a presents an elongated flattened structure perpendicular to the outflow axis , while b is compact and w is much dimmer ( figure [ figcont]d ) .
we identify the infrared counterparts by comparing 870@xmath3 m and 1.3 mm continuum to spitzer data ( figure [ figcont]a , b ) . unresolved a & b
are only detected in the mips1 band ( [email protected] ) with the emission centered at a. since b s sed starts to drop at 870@xmath3 m ( figure [ figsed ] ) we assume the mips1 emission is from a. lack of detection of a & b in all irac bands indicates that both components are still deeply embedded and very young . surprisingly we find that w coincides with a bright infrared source detected in the four irac bands and mips1 band
. resolved fluxes and uncertainties for each component are listed in table [ tball ] .
we construct each component s sed using all available resolved fluxes from ba97 , lmw00 , wkga11 and our work ( table [ tball ] ) .
the source size in all literature are in the 1@xmath72@xmath11 range except in mips1 band ; therefore , the flux we used originates from similar spatial scales .
the mips1 detections are point sources , so their large source size is due to poor resolution .
fluxes between 24@xmath3m@xmath12@xmath13@xmath127 mm were fit by a single - temperature greybody fit with optical depth @xmath14@xmath15@xmath16@xmath17 .
the best fit curves are shown in figure [ figsed ] and fitting results are listed in table [ tball ] .
the uncertainty in the absolute flux calibration may affect our fitting results .
if we adopt the most conservative estimate for sma flux uncertainty ( @xmath730% , @xcite ) , the derived parameters ( table [ tball ] ) will only be altered by @xmath1810% . using the dust temperature , t@xmath19 , obtained from the sed fitting , assuming @xmath20@xmath21=0.01@xmath22g@xmath4 and a distance of [email protected] @xcite
, we derive the total ( gas and dust ) circumstellar mass m@xmath23 , column and number density ( n@xmath24 & n@xmath24 ) for each component from 1.3 mm continuum ( table [ tball ] , uncertainties were obtained through error propagation accounting for errors in distance , flux and t@xmath19 ) . using the fluxes listed in table [ tball ]
we derive the bolometric and submillimeter luminositites ( l@xmath25 & l@xmath26 ) and their ratio for each component .
we find l@xmath26/l@xmath25=1.17% and 0.18% for a and w , ratios characteristic of class 0 and class i sources , respectively @xcite . for b
we find l@xmath25=8.9@xmath510@xmath27l@xmath28 , considerably lower than vellos .
@xcite showed that incomplete seds lead to underestimated derived luminosities . to examine whether this affects a & w s classification ,
we calculate the luminosities from the best fit sed curve including the area under the extrapolated curve , l@xmath29 , and then compare with the above derived luminosities ( table [ tball ] ) .
we find that l@xmath29 are higher than the derived luminosities , as predicted , however the ratios are consistent with the evolutionary classification previously obtained . in figure [ figout ]
, we show vla1623 s outflows in @xmath1co ( 3 - 2 ) and ( 2 - 1 ) with a velocity range of -62kms@xmath4 and 415kms@xmath4 for the blueshifted and redshifted emission , respectively ( v@xmath30 = 3.8kms@xmath4 , yc97 ) . overlapped red- and blueshifted morphologies are observed in both transitions , as expected ( @xcite ; @xcite ; yc97 ) , however we observe several additional features not previously reported . in @xmath1co ( 3 - 2 ) the se lobe is collimated with overlapped emission throughout , whereas the nw lobe presents a forked structure with overlapped red- and blueshifted emission only in the southernmost branch . in @xmath1co ( 2 - 1 ) a slow blueshifted feature is located at w while a dim redshifted clumpy ouflow coincides with the northernmost branch of the nw lobe . in both transitions
we observe a compact blueshifted knot elongated in the ne to sw direction and centered on b. the blueshifted knot presents a spatially overlapped redshifted counterpart only in @xmath1co ( 2 - 1 ) ( figure [ figout ] ) .
so spatially overlapped blueshifted ( -0.52.3kms@xmath4 ) and redshifted ( 57.2kms@xmath4 ) emission as well as a systemic velocity component ( 34.4kms@xmath4 ) and a southward blueshifted ridge are detected around a & b ( figure [ figout ] ) . the so emission is probably tracing circumstellar material shocked by a & b s outflows
. observed c@xmath2o and c@xmath8o emission centered on a present an extended flattened structure perpendicular to the outflow axis ( figure [ figenv ] ) , similar to a s shape in continuum .
both emission show similar velocity gradients characteristic of rotation along the major axis , seen in both velocity gradient maps and position - velocity ( pv ) diagrams ( figure [ figenv ] ) .
no velocity gradient is observed along the outflow axis , suggesting contamination from outflows is neglible .
c@xmath2o s dimensions ( fwhm 370au@xmath5240au ) are somewhat larger than a s 1.3 mm continuum dimensions ( 240au@xmath5120au ) .
assuming centrally dominated motion , we overlay keplerian rotation curves ( v@xmath15r@xmath31 ) for central masses of 0.1 and 0.2m@xmath28 on the c@xmath2o and c@xmath8o pv diagrams . for comparison
we overlay an infall curve v@xmath15r@xmath4 .
although the infall curve passes through the peak of c@xmath8o , both emission are better traced by the keplerian rotation curves .
this suggests that both molecules trace the same rotating disk - like structure , with c@xmath8o tracing the innermost region and c@xmath2o the outermost region of the structure due to different resolutions .
though smaller , the characteristics of a s keplerian disk are similar to @xcite s class i keplerian disk , hinting at the beginnings of disk evolution .
neither emission appears associated with b and no circumstellar material was detected towards w.
from our sed fit of a we obtained t@xmath19=36k , higher than previously reported ( t@xmath19=1520k , awb93 ) . this discrepancy may be due to our fitting using only resolved fluxes tracing the inner hotter regions of the core instead of considering the surrounding colder regions , as was done in awb93 . despite the higher dust temperature , we find a to be a class 0 object based on lack of detection beyond 24@xmath3 m and l@xmath26/l@[email protected]% , in agreement with previous work .
we do not consider the circumstellar - to - protostar - mass ratio as a criterion for evolutionary stage since we can not account for all of a s envelope mass from our observations .
a presents a flattened elongated envelope structure perpendicular to the outflow direction traced in c@xmath2o and c@xmath8o ( figure [ figenv ] ) as well as in continuum , most noticeable at 1.3 mm ( figure [ figcont ] ) .
both c@xmath2o and c@xmath8o present signatures of rotation that suggest pure keplerian rotation ( figure [ figenv ] ) and provide an estimate for the central object mass of a to be 0.1 - 0.2m@xmath28 .
both the envelope traced by c@xmath2o and c@xmath8o and the 1.3 mm continuum emission may be tracing a rotating disk about a , a possibility also raised by wkga11 .
the presence of this rotating disk about a would be one of the first evidence of keplerian disks around class 0 protostars , however further analysis must be carried out to confirm whether this is indeed a keplerian disk .
we find w to be a true protostar , based on detection in all irac and mips1 bands , as well as in the ( sub)millimeter and centimeter regimes .
we are unable to analyze w s environment due to lack of detection of circumstellar material .
given the lack of circumstellar material emission towards w and the small circumstellar mass m@xmath23 derived from the 1.3 mm continuum ( table [ tball ] ) , we assume that w s m@xmath23 is less than the central object mass m@xmath33 .
additionally , awb93 showed l@xmath34/l@[email protected]% to be equivalent to m@xmath36@xmath35m@xmath33 .
hence , based on an infrared spectral index @xmath37@[email protected] ( c2d catalogue ) , l@xmath26/l@xmath25=0.18% , m@xmath36@xmath35m@xmath39 and the class - i - shaped sed , we classify w as a class i source .
@xcite mistook w as a+b due to their source identification criteria ; however , from the listed source position and @xmath37@xmath38 the source is w , even more so since a lacks infrared emission at @xmath13@xmath3524@xmath3 m . the fact that w is a protostellar source confirms the suspicions of @xcite and yc97 that the outflow and jet misalignments are caused by a second source .
we find b to be an extremely cold ( t@xmath405k ) object with extremely low luminosity ( @xmath710@xmath27l@xmath28 ) .
no associated c@xmath2o and c@xmath8o emission suggest strong co depletion , consistent with the derived extremely cold dust temperature ( co sublimation temperature @xmath720k ) .
we further suspect that b may be driving a pole - on outflow ( see section [ secout ] and figure [ figout ] ) .
while we can not determine b s nature unambiguously , some possibilities can be ruled out .
@xcite show vla1623 s jet knots have gas temperatures of 17004200k and velocities between 60100kms@xmath4 .
the strong co depletion at b indicates that the gas temperature is @xmath3520k , far below that expected for jet knots .
observations of monopolar continuum jet knots in t tauri stars @xcite show obvious shifts in the knot positions in short periods of time . assuming simple ballistic motion and velocities of 60100kms@xmath4
, b should have shifted from ba97 s initially reported position by @xmath723@xmath11 in a period of 15 years .
however , comparison of ba97 s and our 2012 observations show no such shift .
vla high resolution ( @xmath705 ) centimeter continuum observations of jet - driving class 0 and i objects @xcite found most detected continuum peaks to be protostellar sources with continuum jets seen as extended emission in the direction of the source s outflow / jet .
b is observed as a separate compact component and is coincident with counterparts in the ( sub)millimeter regimes , making it more like a yso and less like a jet knot .
although vla1623 s alignment and distribution might resemble a jet , several mpss are observed to have their components either aligned or within their companion s outflow ( e.g. ngc1333 svs13 : @xcite ; l1448 c : @xcite ) , possibly caused by filament fragmentation . thus vla1623 s configuration is not highly unlikely .
furthermore , a quick jeans mass calculation shows that given b s mass , dust temperature and density , the core is unstable and most likely collapsing .
hence even if it were product of outflow - environment interaction it is on its way to becoming a yso .
thus , b can not be a jet / outflow feature as proposed by moa12 .
other possibilities for the nature of b are a starless core or proto - brown dwarf . while the co depletion and extremely low temperature and luminosity are characteristics of starless cores @xcite , the mass , density and presence of outflow rule out the possibility of b being a starless core .
if we assume a star formation efficiency of 30% the resulting object from b would be a @xmath70.1m@xmath28 star , well above the brown dwarf limit . with current data and b s physical and chemical characteristics
, we can not determine the eventual outcome but we can conclude that b is an extremely young object between the starless core and class 0 stages .
comparison of both @xmath1co maps and the presence of three sources suggest that vla1623 s outflow is not product of a single source .
we propose that the nw lobe is not a cavity but instead two outflows . a drives the northernmost branch while w drives the southermost branch based on its position and the @xmath1co ( 2 - 1 ) blueshifted feature .
the se lobe is difficult to disentangle from the observations presented here , requiring comparison of @xmath1co and jet observations . for b , the observed spatially overlapped @xmath1co ( 2 - 1 ) emission suggests a compact pole - on outflow .
the red- and blueshifted spatially overlapped so emission and absence of a redshifted lobe in @xmath1co ( 3 - 2 ) further support this morphology , since the redshifted lobe of b would be located directly behind the blueshifted lobe which is stronger and more easily detected by the compact configuration observations .
additionally , the @xmath1co ( 3 - 2 ) redshifted lobe of b is easily lost among the much stronger redshifted emission of the outflows from a & w. further analysis on vla1623 s kinematics , outflows and whether it is gravitationally bound , as well as determining b s nature , are required to fully disentangle vla1623 .
however , from its alignment , physical characteristics , environment and differing evolutionary stages we can conclude that vla1623 is a triple non - coeval system probably formed through the delayed fragmentation of a filament .
these characteristics make vla1623 a good target for constraining and studying fragmentation in low - mass protostars .
aikawa , y. , herbst , e. , roberts , h. & caselli , p. 2005
, , 620 , 330 andr , p. , martin - pintado , j. , depois , d. & montmerle , t. 1990 , , 236 , 180 andr , p. , ward - thompson , d. & barsony , m. 1993 , , 406 , 122 bontemps & andre 1997 , poster proc .
182 caratti o garatti , a. , giannini , t. , nisini , b. & lorenzetti , d. 2006 , , 449 , 1077 chen , x. , launhardt , r. & henning , t. 2009 , , 691 , 1729 dent , w. r. f. , matthews , h. e. & walther , d. m. 1995 , , 277,193 dunham , m.m . ,
crapsi , a. , evans , n.j .
ii , bourke , t.l .
, huard , t.l . ,
myers , p.c .
& kauffmann , j. 2008 , , 179 , 249 enoch , m. l. , evans , n. j. , ii , sargent , a. i. & glenn , j. 2009 , , 692 , 973 froebrich , d. 2005 , , 156 , 169 goodwin , s. , kroupa , p. , goodman , a. & burkert , a. 2007 , ppv , 133 hirano , n. , ho , p. p. t. , liu , s .- y . ,
shang , h. , lee , c .- f .
& bourke , t. l. 2010 , , 717 , 58 ho , p. p. t. , moran , j. m. & lo , k. y. 2004 , , 616 , l1 leous , j.a .
, feigelson , e.d . ,
andre , p. & montmerle , t. 1991 , , 379 , 683 loinard , l. , torres , r. m. , mioduszewski , a. j. & rodrguez , l. f. , 2008 , , 675 , l32 looney , l.w .
, mundy , l.g . & welch , w.j .
2000 , , 529 , 477 maury , a.j . , andr , p. , hennebelle , p. , motte , f. , stamatellos , d. , bate , m. , belloche , a. , duchne , g. & whitworth , a. , 2010 , , 512 , 40 maury , a.j . , ohashi , n. & andre , p. 2012
, , 539 , 130 reipurth , b. , rodrguez , l.f . , anglada , g. & bally , j. 2004 , , 127 , 1736 rodrguez , l. f. , gonzlez , r. f. , raga , a. c. , cant , j. , riera , a. , loinard , l. , dzib , s.a .
& zapata , l.a .
2012 , , 537 , a123 strom , k. , kepner , j. & strom , s. 1995 , , 438 , 813 tafalla , m. , myers , p. c. , caselli , p. , walmsley , c. m. & comito , c. 2002 , , 569 , 815 takakuwa , s. , saito , m. , lim , j. , saigo , k. , sridharan , t. k. , & patel , n. a. 2012 , , 754 , 52 tohline , j. e. 2002 , araa , 40 , 349 ward - thompson , d. , kirk , j. m. , greaves , j. s. & andr , p. 2011
, , 415 , 2812 young , c. h. & evans , n. j. , ii 2005 , , 627 , 293 yu , t. & chernin , l.m .
1997 , , 479 , l63 ) .
the solid black line shows the best single - temperature graybody fit for fluxes at 7mm@xmath41@xmath13@xmath4124@xmath3 m , with the best fit dust temperature and dust emissivity index @xmath42 written in the upper right corner . a simple interpolation ( dashed line ) is made for points outside the fitting range . ]
ra & 16:26:26.397 & 16:26:26.308 & 16:26:25.628 & 1 + dec & -24:24:30.71 & -24:24:30.57 & -24:24:29.60 & 1 + separation ( ) & 0 & 1.18 & 10.51 & 1 + 3.6@xmath3 m & & & 1.9 @xmath6 0.09 & 2 + 4.5@xmath3 m & & & 6.6 @xmath6 0.2 & 2 + 5.8@xmath3 m & & & 0.5 @xmath6 0.7 & 2 + 8@xmath3 m & & & 7.2 @xmath6 0.2 & 2
+ 24@xmath3 m & 40.4 @xmath6 4.2 & & 272.2 @xmath6 15.1 & 2 + 870@xmath3 m & 1153 @xmath6 100 & 87.2 @xmath6 12 & 128 @xmath6 16 & 1 + 1.3 mm & 152 @xmath6 8 & 106 @xmath6 8 & 25.7 @xmath6 8 & 1 + 2.7 mm & 25.5 @xmath6 6.3 & 25.8 @xmath6 3.5 & 3.1 @xmath6 1.5 & 3 + 7 mm & 1.2 @xmath6 0.2 & 2.0 @xmath6 0.2 & & 4 + 3.6 cm & 0.1 @xmath6 0.02 & 0.4 @xmath6 0.02 & 0.2 @xmath6 0.02 & 5 + 4.4 cm & @xmath35 0.2 & @xmath35 0.2 & & 4 + 6 cm & @xmath35 0.4 @xmath6 0.02 & @xmath35 0.4 @xmath6 0.02 & [email protected] & 4 + t@xmath19 ( k ) & 36 @xmath6 5 & 5 @xmath6 1 & 41 @xmath6 5 & 1 + @xmath42 & 1.2 @xmath6 0.6 & 1.7 @xmath6 0.5 & 1.6 @xmath6 0.4 & 1 + reduced @xmath43@xmath44 & 21.7 & 5.8 & 0.7 & 1 + l@xmath25 ( l@xmath28 ) & 0.03 @xmath6 0.003 & ( 8.9 @xmath6 0.5 ) @xmath5 10@xmath27 & 0.03 @xmath6 0.002 & 1 + l@xmath26 ( l@xmath28 ) & ( 3.9 @xmath6
0.3)@xmath5 10@xmath45 & & ( 4.8 @xmath6 0.5 ) @xmath5 10@xmath27 & 1 + l@xmath26/l@xmath25 ( @xmath46 ) & 1.2 @xmath6 0.1 & & 0.2 @xmath6 0.02 & 1 + l@xmath47 ( l@xmath28 ) & 1.1 @xmath6 0.2 & ( 2.9 @xmath6 0.1 ) @xmath5 10@xmath45 & 1.4 @xmath6 0.2 & 1 + l@xmath48 ( l@xmath28 ) & ( 1 @xmath6 0.6)@xmath5 10@xmath49 & & ( 5 @xmath6 2 ) @xmath5 10@xmath50 & 1 + l@xmath48/l@xmath47 ( @xmath46 ) & 0.9 @xmath6 0.5 & & 0.4 @xmath6 0.2 & 1 + m@xmath23 ( m@xmath28 ) & 0.02 @xmath6 0.004 & 0.3 @xmath6 0.2 & 0.003 @xmath6 0.001 & 1 + n@xmath24 ( cm@xmath49 ) & ( 4.4 @xmath6 0.8 ) @xmath5 10@xmath52 & ( 9.9 @xmath6 5.0 ) @xmath5 10@xmath53 & ( 3.6 @xmath6 1.2 ) @xmath5 10@xmath52 & 1 + n@xmath24 ( cm@xmath50 ) & ( 2.2 @xmath6 0.4 ) @xmath5 10@xmath55 & ( 5.5 @xmath6 2.8 ) @xmath5 10@xmath56 & ( 4.1 @xmath6 1.4 ) @xmath5 10@xmath55 & 1 + | commonplace at every evolutionary stage , multiple protostellar systems ( mpss ) are thought to be formed through fragmentation , but it is unclear when and how .
the youngest mpss , which have not yet undergone much evolution , provide important constraints to this question .
it is then of interest to disentangle early stage mpss . in this letter we present the results of our work on vla1623 using our observations and archival data from the submillimeter array ( sma ) .
our continuum and line observations trace vla1623 s components , outflow and envelope , revealing unexpected characteristics .
we construct the sed for each component using the results of our work and data from literature , as well as derive physical parameters from continuum and perform a simple kinematical analysis of the circumstellar material .
our results show vla1623 to be a triple non - coeval system composed of vla1623a , b & w , with each source driving its own outflow and unevenly distributed circumstellar material . from the sed , physical parameters and ir emission
we conclude that vla1623a & w are class 0 and i protostars , respectively , and together drive the bulk of the observed outflow .
furthermore , we find two surprising results , first the presence of a rotating disk - like structure about vla1623a with indications of pure keplerian rotation , which , if real , would make it one of the first evidence of keplerian disk structures around class 0 protostars .
second , we find vla1623b to be a bonafide extremely young protostellar object between the starless core and class 0 stages . |
many imaging techniques reconstruct a signal from its frequency or fourier measurements .
but in practice the phase information of the data in the frequency domain may not be available to the detecting and sensing devices , e.g. , in x - ray crystallography @xcite .
it therefore arises the problem of recovering a signal from its fourier magnitude only measurements , known as phase retrieval .
other applications of phase retrieval include optics @xcite , diffraction imaging @xcite , astronomical imaging @xcite and magnetic resonance imaging @xcite , to name just a few .
since more than one signal can result in the same fourier magnitude measurements , the phase retrieval problem is ill - posed .
for example , a solution to the problem can be subject to so - called global phase ambiguity including a constant global phase , spatial shift and conjugate inversion .
though such ambiguities are acceptable in practical applications , there may exist infinitely many solutions beyond these trivial associates . to resolve the problem , it is shown in @xcite that twofold oversampling in each dimension in the frequency domain
almost always specifies a unique solution ( up to global phase ) for finitely supported , real - valued and nonnegative signals while it does not work for 1d signals . due to the nonlinear relationship between the magnitude measurements and the signal of interest , the algorithm design remains a major problem in phase retrieval .
current approach to signal reconstruction with the oversampling technique include the popular iterative projection algorithms @xcite and recent convex optimization method @xcite .
it is of great interest to reduce the required number of measurements in phase retrieval since every single one of the measurements may have a potential time and/or power cost .
moreover , oversampling can be inconvenient or impossible in applications , e.g. , bragg sampling from periodic crystalline structures @xcite . compressed sensing ( cs ) @xcite is an emerging technique which aims at reconstructing a high dimensional signal from its low dimensional linear measurements under a sparsity prior .
a prosperous direction is to combine cs with phase retrieval , known as compressive phase retrieval ( cpr ) @xcite , which reduces the required number of measurements by exploiting the signal sparsity .
it is shown in @xcite that a sparse 2d image can be uniquely determined ( up to global phase ) with significantly reduced , undersampled fourier magnitude measurements .
modified iterative projection algorithms are proposed in @xcite for cpr in the noise free case .
convex optimization method is developed in @xcite inspired by @xcite with slow computing speed .
the state of the art is generalized approximate message passing ( gamp ) based algorithm which is introduced in @xcite and shown to have good performance in both recovery accuracy and computing speed . in this paper ,
nonconvex @xmath0 minimization problems are formulated for noiseless and noisy cpr problems inspired by existing cs results .
a desirable property is shown in the widely studied scenario of real - valued , nonnegative image reconstruction that the image of interest is an optimal solution of the formulated @xmath0 minimization problem in the noise free case .
first - order iterative algorithms are proposed to solve the formulated problems based on the alternating direction method ( adm ) @xcite in convex optimization .
numerical simulations are provided for simulated sparse images and nonnegative image to demonstrate the fast , accurate and robust performance of the proposed approach .
the rest of the paper is organized as follows .
section [ sec : theory ] presents our problem formulations for the noiseless and noisy cpr problems .
section [ sec : algorithm ] presents the adm - based algorithms for the proposed @xmath0 minimization problems .
section [ sec : simulation ] provides the simulation results and section [ sec : conclusion ] concludes the paper .
we study the recovery of sparse signals from their fourier magnitude measurements , known as compressive phase retrieval . inspired by the recent cs technique , we propose to solve the following lasso - like optimization problem where @xmath1 is the signal of interest , @xmath2 denotes the discrete fourier transform ( dft ) constrained on the index set @xmath3 , @xmath4 denotes the observed ( generally noisy ) fourier magnitude measurements on @xmath3 and @xmath5 is a regularization parameter .
the second term in ( [ formu : lasso ] ) fits the observed magnitude data while the @xmath0 norm is used to promote the signal sparsity
. an equivalent formulation of ( [ formu : lasso ] ) is where @xmath6 controls the fidelity of the reconstruction to the measured data .
it is known that ( [ formu : lasso ] ) and ( [ formu : bpdn ] ) are equivalent with appropriate choices of @xmath7 and @xmath6 . as @xmath8 and @xmath9 , both ( [ formu : lasso ] ) and ( [ formu : bpdn ] ) reduces to in the noise free case .
we note that the @xmath0 minimization problem ( [ formu : bp ] ) has been studied in @xcite .
however , there have been few theoretical results on its performance analysis . the following theorem deals with a practical and widely studied scenario where the signal of interest is nonnegative ( nonnegative means real - valued as well in this paper ) .
although far from satisfactory , this result nevertheless provides some assurance on the reliability of the proposed @xmath0 minimization approach .
if @xmath10 and its fourier magnitudes @xmath11 is observed , where @xmath3 contains the zero frequency , then @xmath12 is an optimal solution to ( [ formu : bp ] ) .
[ thm : optimality ] _ proof : _ we need to show that the inequality @xmath13 holds for any @xmath1 satisfying @xmath14 . to do this , consider the dc ( zero frequency ) term of the frequency data which is proportional to the sum of all elements of the signal vector .
so the dc term must be nonnegative since @xmath10 .
it follows from @xmath14 that @xmath15 . as a result , where the last equality follows from the positivity of @xmath12 . ' '' '' theorem [ thm : optimality ] states that a nonnegative signal has the least @xmath0 norm among all possible candidates which result in the same fourier magnitudes . though our focus is on the recovery of sparse signals , theorem [ thm : optimality ] does hold for general nonnegative signals and hence brings new insights to the applicable scope of the @xmath0 optimization .
it is noted that the proposed @xmath0 optimization methods do not result in a unique solution .
in fact , they suffer from at least the same ambiguities as existing approaches to conventional phase retrieval .
for example , an optimal solution to either one of ( [ formu : lasso])-([formu : bp ] ) after a modification of constant global phase , spatial shift and/or conjugate inversion remains optimal .
the augmented lagrangian alternating direction method ( adm ) solves the structured optimization problem where @xmath16 and @xmath17 are convex functions of @xmath1 and @xmath18 , respectively .
the augmented lagrangian function of the problem is given by where @xmath19 is a lagrangian multiplier and @xmath20 is a penalty parameter . starting with @xmath21 and @xmath22 ,
the adm iterates as follows : where @xmath23 guarantees the convergence under some technical assumptions @xcite .
the adm is very efficient when explicit solutions are available for ( [ formu : solvex ] ) and ( [ formu : solvez ] ) .
the adm has been a popular approach to solutions of large scale problems , since it can typically produce a modestly accurate solution within a few tens of iterations though its convergence to high accuracy may be slow @xcite .
problems ( [ formu : lasso])-([formu : bp ] ) are nonconvex due to the presence of the magnitude operator . in this section
we solve ( [ formu : lasso])-([formu : bp ] ) using the adm .
problem ( [ formu : lasso ] ) can be formulated into according to ( [ formu : admproblem ] ) , we see in this case that @xmath24 is convex while @xmath25 is nonconvex . by recognizing that the dft is unitary
, @xmath1 can be explicitly updated using a soft thresholding operator ( later given in ( [ formu : xupdate_lasso ] ) ) .
let @xmath26 .
then the augmented lagrangian involving @xmath18 is expressed as where the ` @xmath27 ' in the inequality holds if @xmath28 for @xmath29 and @xmath30 for @xmath31 , and @xmath32 is a constant independent of @xmath18 . as a result , the augmented lagrangian obtaines the minimum at the @xmath18 such that @xmath33 for @xmath29 and @xmath30 for @xmath31 .
so the alternating direction algorithm for ( [ formu : lasso ] ) is summarized as follows : where @xmath34 , @xmath35 is a soft thresholding operator with @xmath36 .
next , ( [ formu : bpdn ] ) can be formulated as where @xmath37 , and @xmath38 is an indicator function with @xmath39 if @xmath40 , and @xmath41 otherwise . as a result , we have @xmath24 and @xmath42 is nonconvex since the set @xmath43 is nonconvex .
denote @xmath44 the projection ( in euclidean norm ) onto @xmath43 . according to the adm we obtain the algorithm for ( [ formu : bpdn ] ) which is the same as that for ( [ formu : lasso ] ) except for the update rule of @xmath18 , @xmath45 .
so , provided @xmath46 we have where @xmath47 and @xmath48 is defined as before .
a detailed derivation of ( [ formu : zupdate_bpdn ] ) is omitted .
as @xmath8 and @xmath9 , both ( [ formu : zupdate_lasso ] ) and ( [ formu : zupdate_bpdn ] ) reduce to the following update rule of @xmath18 for solving ( [ formu : bp ] ) : while again the updates for @xmath1 and @xmath19 are the same as ( [ formu : xupdate_lasso ] ) and ( [ formu : yupdate_lasso ] ) , respectively .
unfortunately , the proposed adm algorithms can not be guaranteed to converge .
in fact , the convergence issue remains a difficult problem in the nonconvex phase retrieval problem and a mathematically rigorous analysis has not been found for many existing algorithms for conventional phase retrieval including hio , hpr and raar @xcite .
the performance of the proposed algorithms will be numerically studied in section [ sec : simulation ] , where we observe that the proposed algorithms can either converge to a good result or not converge at all .
we provide numerical simulations to demonstrate the performance of the proposed adm algorithms in this section .
2d random sparse images and a nonnegative satellite image shown in fig . [ fig : satellite ] are considered . due to the page limit , a thorough comparison of our method with existing approaches will be reported in a future publication .
* algorithm implementation : * the algorithms start with @xmath49 and a random @xmath22 such that @xmath50 and @xmath51 for @xmath31 .
@xmath52 is set such that a fixed portion , @xmath53 , of the elements of @xmath54 are kept nonzero with @xmath55 .
we set @xmath56 for random sparse images and @xmath57 for the satellite image .
the algorithm is terminated if @xmath58 or a maximum number of iterations , 500 , is reached . in the first simulation ,
we consider random images in the noiseless case and study the success rate of recovery versus the sparsity level @xmath59 .
three types of images are considered including complex , real and nonnegative ones to test whether they have different recovery performance since theorem [ thm : optimality ] holds only for nonnegative signals .
we consider images of dimension @xmath60 ( number of image pixels @xmath61 ) with image sparsity level @xmath59 varying from @xmath62 to @xmath63 and acquire a number of @xmath64 random fourier magnitude measurements .
an image is claimed to be correctly recovered if the relative root mean squared error ( rmse ) @xmath65 , where @xmath66 denotes the recovered image after removing the possible ambiguities of constant global phase , spatial shift and conjugate inversion . for each combination of @xmath59 and the image type
, @xmath67 trials are repeated .
the success rates of recovery are presented in fig .
[ fig : successrate_vs_sparsity ] . in the case of a moderate or low sparsity level ,
the proposed algorithm is observed to converge to the correct solution in most trials starting from a random point .
moreover , it is shown that nonnegative images have the highest success rate of recovery followed by real and then complex ones .
the inherent reason will be explored in future studies . .
the number of image pixels and fourier magnitude measurements are @xmath68 and @xmath69 , respectively.,width=316 ] the second simulation studies the variation of the recovery error with respect to the noise energy .
we fix the sparsity level @xmath70 and set @xmath71 and @xmath72 as before .
complex random images are scaled to unit frobenius norm . after obtaining the noiseless fourier magnitude measurements ,
a gaussian random noise is added with the noise energy @xmath73 varying from @xmath74 to @xmath75 with a step size of @xmath76 .
the averaged relative rmse is presented in fig .
[ fig : rrmse_vs_noise ] , where it is shown that the recovery error grows approximately linearly with the noise level @xmath6 . with @xmath68 , @xmath69 and @xmath70.,width=316 ] the last simulation studies the reconstruction of a nonnegative @xmath77 satellite image ( @xmath78 ) which is shown in fig .
[ fig : satellite ] .
the satellite image has a sparsity ratio @xmath79 .
the number of fourier magnitude measurements , @xmath72 , is set such that @xmath80 and @xmath62 , respectively . in each setting , a gaussian random noise is added to the noiseless magnitude measurements such that the signal to noise ratio ( snr ) is @xmath81 db . a fixed number of iterations , 200 , is used to reconstruct the images .
our simulation results are presented in fig .
[ fig : result_satellite ] , where , remarkably , a faithful reconstruction is obtained with only @xmath82 of the fourier magnitude measurements . without accounting for the effects of the possible ambiguities mentioned before , the three reconstructed images have snrs of @xmath83 , @xmath84 and @xmath85 db , respectively . moreover ,
the proposed algorithm is very fast and takes @xmath86 , @xmath87 and @xmath88 s , respectively , to obtain the reconstructed images using matlab v7.7.0 on a pc .
the noiseless and noisy compressive phase retrieval problem was studied in this paper while the @xmath0 norm is exploited to promote the signal sparsity inspired by compressive sensing .
the optimality of the formulated @xmath0 minimization problem was proven for nonnegative signals in the sense that the signal of interest is an optimal solution of the @xmath0 minimization problem in the noise free case .
efficient alternating direction algorithms were proposed for the problem solving and promising results were presented to demonstrate their performance .
o. bunk , a. diaz , f. pfeiffer , c. david , b. schmitt , d. satapathy , and j. veen , `` diffractive imaging for periodic samples : retrieving one - dimensional concentration profiles across microfluidic channels , '' _ acta crystallographica section a : foundations of crystallography _ , vol .
63 , no . 4 , pp . 306314 , 2007 .
m. hayes , `` the reconstruction of a multidimensional sequence from the phase or magnitude of its fourier transform , '' _ ieee transactions on acoustics , speech and signal processing _ , vol .
30 , no . 2 ,
pp . 140154 , 1982 .
s. mukherjee and c. seelamantula , `` an iterative algorithm for phase retrieval with sparsity constraints : application to frequency domain optical coherence tomography , '' in _ acoustics , speech and signal processing ( icassp ) , 2012 ieee international conference on_.1em plus 0.5em minus 0.4em ieee , 2012 , pp .
553556 .
s. boyd , n. parikh , e. chu , b. peleato , and j. eckstein , `` distributed optimization and statistical learning via the alternating direction method of multipliers , '' _ foundations and trends in machine learning _ , vol . 3 , no . 1 ,
1122 , 2011 . | phase retrieval refers to a classical nonconvex problem of recovering a signal from its fourier magnitude measurements .
inspired by the compressed sensing technique , signal sparsity is exploited in recent studies of phase retrieval to reduce the required number of measurements , known as compressive phase retrieval ( cpr ) . in this paper , @xmath0
minimization problems are formulated for cpr to exploit the signal sparsity and alternating direction algorithms are presented for problem solving . for
real - valued , nonnegative image reconstruction , the image of interest is shown to be an optimal solution of the formulated @xmath0 minimization in the noise free case .
numerical simulations demonstrate that the proposed approach is fast , accurate and robust to measurements noises . |
according to the _ recycling _ model ( e.g. @xcite ) millisecond - or recycled - pulsars ( msps ) are formed in binary systems where the companion star transfers mass , and hence angular momentum , onto the neutron star ( ns ) .
extended mass transfer from low mass companions can cause the ns to be spun up to rotational periods of at least 1.5 milliseconds .
higher mass systems , transferring mass over shorter time scales and mostly via a wind @xcite , can not push the recycling process down to periods below @xmath0 20 milliseconds . at the same time , due to still unclear mechanisms , possibly linked to the accreted mass itself , the magnetic field b is decreased down to @xmath1 g. about 75% of all known msps in the galactic field ( i.e. outside globular clusters , where exchange interactions can change the final outcome of the evolution of an msp ) are found in binary systems ( data from the atnf pulsar catalogue[multiblock footnote omitted ] ; @xcite ) , supporting the above scenario . the formation path of isolated msps ,
though , is still a matter of debate : those with spin periods of few tens of ms or more could derive from double neutron star systems disrupted during the second supernova explosion ( disrupted recycled pulsars , drps ; @xcite ) while the shorter period pulsars could be the result of a binary system where the companion was so light that the interaction with the pulsar wind was able to completely destroy it @xcite
. there are now a few examples of the so called _ black widow _
( bw ) pulsar systems @xcite , in which the companion star is very light and in which the radio signal is eclipsed by matter ablated from the companion for significant parts of the orbit .
these objects have been suggested as possible progenitors of the shorter period isolated msps , although the ablation time scales seem to be generally too long for this process @xcite .
untill recently very few bw pulsars were known in the galactic field .
the situation has changed in the last few years thanks to deep radio follow - ups of gamma - ray point sources detected by the lat @xcite instrument on board the _ fermi _ satellite ( see e.g. @xcite and references therein ) , through which also binary pulsars whose radio signal was eclipsed for a large fraction of the orbit were discovered .
we note that , in average black widows detected through this channel are more energetic , show longer eclipses and have lighter companions than those previously known , hence it is likely that the mass loss rate is higher and the time for complete ablation of their companions shorter .
other possible explanations of the formation of isolated msps invoke different paths : one example is the accretion - induced collapse of a white dwarf @xcite , another is white dwarf mergers @xcite .
finding intrinsic differences between the two msp populations might point to a different origin and/or evolutionary path for the isolated msps .
msps are more stable clocks than the non - recycled pulsars ( psrs ) .
when in binary systems , their stable rotation allows precise determination of the keplerian orbital parameters and , in some cases , also the post - keplerian parameters , making these pulsars laboratories for testing relativistic gravity in the strong field regime ( e.g. @xcite ) .
if the rotational stability and timing precision are high enough , msps can be used in timing arrays for gravitational wave detection @xcite . the high time resolution universe survey ( htru ; @xcite ) , whose observations started at the parkes observatory ( nsw , australia ) in 2008 , was primarily designed to discover msps missed in previous surveys .
thanks to the new digital backends adopted in this experiment , the time and frequency resolutions have improved by a factor of 4 and 8 respectively .
this allows the survey to sample a much larger volume of the galaxy for short period pulsars as the deleterious effects of dispersion are greatly diminished .
the suitability of the chosen parameters of the survey to discover msps has been already demonstrated in @xcite where we presented the first five new objects of this class found in the intermediate latitude ( @xmath2 ) part of the survey . in this paper
we present the discovery and timing parameters of five further such objects , confirming the very high potential of our acquisition system . in section [ sec : time ] we briefly describe the follow - up observations and report the spin , orbital and astrometric parameters resulting from the timing campaign on the five discovered msps , highlighting their potential as ` test particles ' in the timing array experiments . given the increased sample of recycled pulsars , in [ sec : lum ] we revise the luminosity distribution of isolated versus binary msps and find them to be very different . in [ sec : mw ] we briefly report on the search for gamma - ray counterparts to our discoveries . finally , in
[ sec : pol ] we describe the polarimetric characteristics for four of the new discoveries .
of the five newly discovered msps , four were found in the data of the intermediate latitude part of the htru survey , consisting of 540s pointings covering the area enclosed by galactic coordinates @xmath3 , @xmath2 .
the fifth , j2236@xmath45527 , was discovered in the high latitude part of the survey : 270s pointings covering the entire southern sky below declination @xmath5 ( except the intermediate latitude part mentioned above ) . when a pulsar is discovered only its approximate spin period , dispersion measure ( dm
, i.e. the free electron column density along the line of sight ) and position are known . to ensure that the full potential of these discoveries is exploited , a precise determination of the pulsars position is needed ( e.g. to look for counterparts at other wavelengths ; see [ sec : mw ] ) and
accurate measurements of the spin , astrometric , and if they belong to a binary system , orbital parameters are necessary . to obtain these parameters
one has to start a follow - up timing campaign covering a time span of at least one year . for the htru msps , confirmations ,
gridding observations ( aimed at determining a more precise position for the pulsar with a grid of 5 adjacent pointings ; @xcite ) and the first timing observations , are done with the same instrumentation used for the survey pointings ( see @xcite ) .
this allows the raw data to be retained with all the time samples and the frequency channels , in such a way as to be able to recover the periodic signal even if the apparent spin period is significantly different from the discovery one because of orbital motion .
the rest of the timing observations , done with an approximately monthly cadence , are performed either with the parkes 64-m telescope or , for the objects visible from the northern hemisphere , with the lovell telescope ( jodrell bank , manchester , uk ) . at both sites an atnf digital filterbank ( dfb )
is used as backend and most of the observations are performed at a wavelength close to 20 cm . in particular , at the parkes telescope the data are acquired over 256 mhz of bandwidth centered at a frequency of 1369 mhz and split into 1024 frequency channels , while the lovell data cover a frequency range of 384 mhz centred at 1532 mhz and split into 1024 frequency channels , more than 1/4 of which , untill september 2012 , were removed trough a bandpass filter because of radio frequency interferences ( rfi ) . after that date
the lovell data have been manuallly cleaned from rfi and a larger part of the band has been used .
dfb data are folded on - line using the pulsar ephemeris obtained from the first stages of timing and refined as the follow - up observations proceed .
observations of psrj1832@xmath40836 at jodrell bank were performed also using the _ roach _ pulsar backend .
this backend uses a roach fpga board[multiblock footnote omitted ] to perform either baseband recording or coherent dedispersion of two 8-bit nyquist sampled orthogonal polarisations over a total bandwidth of 512mhz .
after sampling and digitization , the roach board channelizes the signal into 32 sub - bands of 16mhz , each of which is coherently dedispersed and folded using an 8-core intel xeon computer running the ` dspsr ` software package @xcite .
as the total usable bandwidth of the l - band receiver is 400mhz , the data from 25 subbands is combined and analyzed with the ` psrchive ` package[multiblock footnote omitted ] @xcite .
also in this case about 1/4 of the bandwidth was removed because of rfi untill september 2012 .
after about one year of timing observations the covariance in the timing parameters is broken and a precise timing solution is obtained .
times of arrival are obtained using from through cross - correlation with an analytic profile template created by fitting von mises functions to profile components .
table [ tab1 ] reports the main detection characteristics of the five newly discovered pulsars : for each object the timing position in galactic coordinates , the beam in which the pulsar was detected , the radial distance of the pulsar from the discovery beam centre in units of the beam radius , the signal - to - noise ratio ( s / n ) of the discovery observation , the mean flux density @xmath6 and the pulse widths at 50% and 10% of the peak of the mean pulse profile are listed .
for the three pulsars timed at the parkes observatory , the values of the fluxes are properly calibrated on a reference source ( see [ sec : pol ] )
. for polarisation measurement purposes ( see [ sec : pol ] ) j1832@xmath40836 , timed over more than one year with the lovell telescope , was also observed on one occasion for @xmath7 hrs with the parkes radio telescope and hence as also properly flux calibrated .
the flux density value for the remaining pulsar has been obtained using the radiometer equation ( see e.g. @xcite ) scaled using the calibrated flux densities of the other four msps . for j1545@xmath44550 , for which we have a flux measurement also at 10 cm and 40 cm , we could calculate a spectral index @xmath8 ( with @xmath9 ) of @xmath10 for the main peak , and a steeper , though less constrained , spectral index of @xmath11 for the second peak . [
cols="<,>,>,^,^,^,^,^,^ " , ] polarisation studies are fundamental to investigate the geometry of pulsars as well as their magnetic fields and radio beam structures . in the ideal case ,
the interpretation of the results of this kind of analysis could be carried out using the rotating vector model ( rvm ) by @xcite .
this model assumes that the position angle of the linear polarisation ( @xmath12 ) is tied to that of the pulsar magnetic field lines . in this way ,
its prediction is to observe a swing of the polarisation angle ( pa ) curve across the pulse ( that should produce a characteristic `` s - shape '' ) , due to the sweeping of the observer s line of sight through the radio beam .
the fit of the model should permit the inclination between the magnetic and the rotational axes and the impact parameter between the line of sight and the magnetic axis to be obtained .
this however is not always possible , for example because the polarisation signal - to - noise ratio ( snr ) is not high enough to allow a good determination of the pa , or the pulse is so narrow that there is not enough information to constrain the fit , or also because the predicted pa swing is not displayed ( for a thorough discussion of the difficulties and limitation of the rvm see @xcite ) .
msp pas , in particular , often display jumps or discontinuities in the pa profile : this is the case of our msps , as described below , for which hence a fit using the rvm has not been possible . the pulse profile for j1431@xmath45740 ( fig .
[ fig : j1431 ] ) shows a single slightly asymmetric peak .
the linear polarisation , however , displays two peaks , whereof the leading is fainter and wider in amplitude than the trailing .
the minimum of the linear polarisation occurs roughly at the phase of the peak in the total intensity profile .
the pas in correspondence with the leading peak show a slightly decreasing slope , while the pa curve of the trailing component is flat .
the two series of pas exhibit an orthogonal ( within the uncertainties ) jump in between .
the circular polarisation profile shows a single component , whose peak appears in phase with that of the total intensity .
5740 at a centre frequency of 1369 mhz .
the lower panel shows , in black , the total intensity , in light grey ( red in the electronic version ) the linearly polarised intensity @xmath13 and in dark grey ( blue in the electronic version ) the circularly polarised intensity @xmath14 .
the top panel gives the pa of the linearly polarised emission . ] this pulse profile at 20 cm ( central panel of figure [ fig : j1546 ] ) shows three well defined components : two of them are contiguous , while the third one is 0.22 in rotational phase away from the closest component .
a fourth component is barely visible about 0.4 in phase before the main peak .
it acquires statistical significance when comparing the 20-cm profile to the 40-cm one ( bottom panel of fig .
[ fig : j1546 ] ) where this fourth component emerges clearly .
the linear polarisation profile follows the total intensity one , and the percentage of @xmath13 under every component is very high , especially beneath the second peak .
the pa curve of the component with higher snr shows a short , steep rise and a smooth decline toward the separation between the first and the second component . a similar smooth decline with almost the same inclination
is shown by the pa curve of the third peak , while that of the second one is flatter and exhibits a jump of @xmath15 with respect to the first .
the circular polarisation appears mainly under the first component where it is fully negative , while there are just hints of circular polarisation beneath the other two peaks , with opposite signs . as mentioned in
[ sec : bin ] , for this pulsar a few observations at 10 and 40 cm are also available . given the smaller number of data points , the resulting profiles , especially the one at 40 cm , whose band is often corrupted by rfi ,
are not as high s / n as the 20-cm one . at 10 cm ( top panel of figure [
fig : j1546 ] ) the first peak appears @xmath16 per cent narrower than at 20 cm and the third peak is statistically undetectable .
the linear and circular polarisation profiles of the main peak behave as at 20 cm , while the second peak appears to be unpolarised .
the 40-cm profile ( bottom panel of figure [ fig : j1546 ] ) shows four components : besides the three clearly present at 20 cm , whose peaks are closer together at this lower frequency , a fourth component precedes in phase the main peak by 1/3 . the percentage of linear polarisation of the main peak drops by about 30 per cent with respect to the 20-cm value , while the second and third peaks appear 100 per cent linearly polarised ( even though for the latter , as for the fourth peak , the statistic is quite poor ) .
the circular polarisation is negative below the main peak and positive below the second one .
in contrast with what is generally found @xcite , for this object the percentage of linear polarisation ( at least in the main peak where a trend through all three frequencies can be easily seen ) decreases with decreasing frequency while the percentage of circular polarisation increases ( in absolute value ) .
the pattern of the pa curves are very similar at the three frequencies and the rm values calculated on the three data sets separately are always compatible with zero at 2-@xmath17 .
4550 at centre frequencies of 3100 mhz ( top ) , 1369 mhz ( centre ) and 732 mhz ( bottom ) .
lines and panels as in figure [ fig : j1431 ] . the pulses at the different frequencies are aligned so that the centroid of the main peak falls at phase 0.,title="fig : " ] 4550 at centre frequencies of 3100 mhz ( top ) , 1369 mhz ( centre ) and 732 mhz ( bottom ) .
lines and panels as in figure [ fig : j1431 ] .
the pulses at the different frequencies are aligned so that the centroid of the main peak falls at phase 0.,title="fig : " ] 4550 at centre frequencies of 3100 mhz ( top ) , 1369 mhz ( centre ) and 732 mhz ( bottom ) .
lines and panels as in figure [ fig : j1431 ] .
the pulses at the different frequencies are aligned so that the centroid of the main peak falls at phase 0.,title="fig : " ] this msp shows a complex pulse profile displaying six components ( see fig .
[ fig : j1832 ] ) , two of which are blended with the adjacent one .
the polarisation values for the two peaks with the two blended components are computed for both components together in table [ tab : pol ] , listing hence a total of four peaks ( centred approximately around phases @xmath40.05 , 0.23 , 0.3 and 0.57 ) .
unfortunately the polarisation is quite small and the pas do not allow us to clearly understand the beam structure leading to such a peculiar profile .
the pas in correspondence of the two components around phase 0.0 are separated by a jump of almost @xmath18 , with the second component showing a slightly negative slope .
in contrast the peaks at phase @xmath19 and @xmath20 have pas that appear flatter , with a slightly positive slope for the latter .
the linear polarisation profile shows five peaks , with only the second component of the peak around phase 0.3 apparently unpolarised .
we computed an rm value for every linearly polarised peak separately , however , due to the poor signal - to - noise ratio of the polarised emission , the errors are often very high and most of the rms are compatible with zero at 2-@xmath17 .
the rm measured on the peak centred at phase 0.0 , the one showing the highest level of polarisation and giving the smallest error is @xmath21 rad / m@xmath22 . to confirm this value we also measured the rm with the methods described in @xcite and in @xcite , yielding respectively values of @xmath23 rad / m@xmath22 and @xmath24 rad / m@xmath22
testing these results with a matrix template matching technique , we found out that a value close to 25 rad / m@xmath22 better describes the data .
the circular polarised emission is visible , with changing signs , in correspondence with the peaks at phase @xmath25 , @xmath26 and @xmath20 .
0836 at a centre frequency of 1369 mhz .
lines and panels as in figure [ fig : j1431 ] . ] the total intensity profile of this msp ( fig .
[ fig : j2236 ] ) shows two blended peaks and two additional much fainter components separated in phase from the main one by 0.42 and 0.22 respectively .
we considered just the main components , since there is no evidence of polarised signal for the other two .
the linear polarisation profile shows three peaks .
we computed the rm for each of the peaks separately : the rms of the first and the third peak seem to be consistent with zero although with a very high uncertainty , while the central peak gives a rm value of @xmath27 rad / m@xmath22 .
the pas of the first component draw a flat curve , and are separated from the pas related to the second peak by a jump of almost @xmath28 . for the central component
the pa curve shows a steep decrease , whereas it increases in correspondence with the third component .
circularly polarised emission is present mainly at the same pulse longitude of the second and the third component of the linear polarisation , with a @xmath14 profile sign change occurring at the same time of the pa s slope change .
5527 at a centre frequency of 1369 mhz .
lines and panels as in figure [ fig : j1431 ] . ]
0319 at a centre frequency of 1369 mhz .
lines and panels as in figure [ fig : j1431 ] . ]
in this paper we presented the discovery and timing results of five msps detected in the htru survey data .
four of them belong to binary systems with he - wd companions , while the fifth is one of the only two isolated msps found so far in this survey .
a revision of the luminosity distribution of known isolated and binary msps is presented ; with the sample of isolated msps increased by more than 60 per cent and the total sample of msps tripled with respect to the previous works addressing this issue ( e.g. @xcite ) , the intrinsic luminosity difference between these two populations appears to be confirmed .
the small percentage of isolated objects among msp discoveries in the htru survey , sampling a deeper volume of the galaxy with respect to previous experiments , also supports the hypothesis that the intrinsic luminosity difference is indeed real .
this effect could be the consequence of a different evolutionary history experienced by isolated and binary msps . for four of our targets ,
those for which we have follow - up observation carried out at the parkes radio telescope , we also reported the results of a polarimetric analysis . the possibility to calibrate their polarisation
accurately also allowed us to obtain an improved timing solution for two of our msps .
thanks to the precise ephemerides obtained through radio timing , we could fold in phase the gamma - ray photons collected by the _ fermi - lat _ instrument since its launch from the direction of the msps .
unfortunately no pulsating high - energy counterpart to any of our msps was found , nor were point sources corresponding to the position of our targets present in the 2-yr _ fermi _ point source catalog @xcite .
one of the main aims of the high time resolution universe pulsar survey @xcite is that of discovering stable millisecond pulsars suitable to be included in pulsar timing array experiments for the detection of gravitational waves @xcite .
two of the newly discovered msps show quite small timing residuals , which makes them good candidates to be used in pulsar timing arrays : j1545@xmath44550 is indeed the second htru msp to be included in the parkes pta and j1832@xmath40836 is being considered for the european and north american ptas .
the parkes radio telescope is part of the australia telescope which is funded by the commonwealth of australia for operation as a national facility managed by csiro .
j. w. t. , et al .
, eds , american institute of physics conference series vol . 1357 of american institute of physics conference
series , a 350-mhz gbt survey of 50 faint fermi @xmath29-ray sources for radio millisecond pulsars . | this paper presents the discovery and timing parameters for five millisecond pulsars ( msps ) , four in binary systems with probable white dwarf companions and one isolated , found in ongoing processing of the high time resolution universe pulsar survey ( htru ) .
we also present high quality polarimetric data on four of them .
these further discoveries confirm the high potential of our survey in finding pulsars with very short spin periods .
at least two of these five msps are excellent candidates to be included in the pulsar timing array projects .
thanks to the wealth of msp discoveries in the htru survey , we revisit the question of whether the luminosity distributions of isolated and binary msps are different . using the cordes and lazio distance model and our new and catalogue flux density measurements , we find that 41 of the 42 most luminous msps in the galactic disk are in binaries and a statistical analysis suggests that the luminosity functions differ with 99.9% significance .
we conclude that the formation process that leads to solitary msps affects their luminosities , despite their period and period derivatives being similar to those of pulsars in binary systems .
[ firstpage ] stars : pulsars : individual : psrj14315740 , psrj15454550 , psrj18250319 , psrj18320836 , psrj22365527 |
let @xmath0 be a rational function of degree @xmath1 on the riemann sphere and let @xmath2 be its julia set .
our goal is to study the spectrum of lyapunov exponents of @xmath3 . given @xmath4 we denote by @xmath5 and @xmath6 the _ lower _ and _ upper lyapunov exponent _ at @xmath7 , respectively , where @xmath8 if both values coincide then we call the common value the _ lyapunov exponent _ at @xmath7 and denote it by @xmath9 .
for given numbers @xmath10 we consider also the following level sets @xmath11 we denote by @xmath12 the set of _ lyapunov regular points _ with exponent @xmath13 . if @xmath14 then @xmath15 is contained in the set of so - called _ irregular points _ @xmath16 recall that it follows from the birkhoff ergodic theorem that @xmath17 for any @xmath18-invariant probability measure @xmath19 .
while the first results on the multifractal formalism go already back to besicovitch @xcite , its systematic study has been initiated by work of collet , lebowitz and porzio @xcite .
the case of spectra of lyapunov exponents for conformal uniformly expanding repellers has been covered for the first time in @xcite building also on work by weiss @xcite ( see @xcite for more details and references ) . to our best knowledge
, the first results on irregular parts of a spectrum were obtained in @xcite . its first complete description ( for digit expansions )
was given by barreira , saussol , and schmeling @xcite . in this work
we will formulate our results on the spectrum of lyapunov exponents in terms of the topological pressure @xmath20 . for
any @xmath21 let us first define @xmath22 and @xmath23 let @xmath24 $ ] be the interval on which @xmath25 ( a more formal and equivalent definition is given in below ) . before stating our main results
, we recall what is already known in the case that @xmath26 is a uniformly expanding repeller with respect to @xmath18 , that is , @xmath26 is a compact @xmath18-forward invariant ( i.e. , @xmath27 ) isolated set such that @xmath3 is uniformly expanding . recall that @xmath18 is said to be _ uniformly expanding _ or _ uniformly hyperbolic _ on a set @xmath28 if there exist @xmath29 and @xmath30 such that for every @xmath31 and every @xmath32 we have @xmath33 .
recall that a set @xmath28 is said to be _ isolated _ if there exists an open neighborhood @xmath34 of @xmath28 such that @xmath35 for every @xmath36 implies @xmath32 . in our setting the julia set @xmath26 is a uniformly expanding repeller if it does not contain any critical point nor parabolic point . here
a point @xmath7 is said to be _ critical _ if @xmath37 and to be _ parabolic _ if @xmath7 is periodic and its multiplier @xmath38 is a root of @xmath39 .
if @xmath26 is a uniformly expanding repeller then for every @xmath40 $ ] we have @xmath41 ( see @xcite ) and @xmath42 if and only if @xmath43 $ ] @xcite .
this gives the full description of the regular part of the lyapunov spectrum .
moreover , in this setting of a uniformly expanding repeller @xmath26 , the interval @xmath24 $ ] coincides with the closure of the range of the function @xmath44 and the spectrum can be written as @xmath45 where @xmath46 is the unique number satisfying @xmath47 and where @xmath48 is the unique equilibrium state corresponding to the potential @xmath49 .
if @xmath50 is not cohomologous to a constant then we have @xmath51 , and @xmath52 and @xmath53 are real analytic strictly convex functions that form a legendre pair .
we now state our first main result .
[ main1 ] let @xmath18 be a rational function of degree @xmath54 with no critical points in its julia set @xmath26 .
for any @xmath55 , @xmath56 , we have @xmath57 in particular , for any @xmath58\setminus\{0\}$ ] we have @xmath59 if there exists a parabolic point in @xmath26 ( and hence @xmath60 ) then @xmath61 moreover , @xmath62 we denote by @xmath63 the set of all critical points of @xmath18 . following makarov and smirnov (
* section 1.3 ) , we will say that @xmath18 is _ exceptional _ if there exists a finite , nonempty set @xmath64 such that @xmath65 this set need not be unique
. we will further denote by @xmath66 the largest of such sets ( notice that it has no more than 4 points ) .
[ main2 ] let @xmath18 be a rational function of degree @xmath54 .
assume that @xmath18 is non - exceptional or that @xmath18 is exceptional but @xmath67 .
for any @xmath68 we have @xmath69 in particular , for any @xmath58 \setminus \{0\}$ ] we have @xmath70 and @xmath71 moreover , @xmath72 and @xmath73 if @xmath18 is exceptional and @xmath74 ( this happens , for example , for chebyshev polynomials and some latts maps ) then the situation can be much different from the above - mentioned cases .
for example , the map @xmath75 possesses countably many points with lyapunov exponent @xmath76 , two points with lyapunov exponent @xmath77 , a set of dimension 1 of points with lyapunov exponent @xmath78 , and no other lyapunov regular points .
hence , for this map the lyapunov spectrum is not complete in the interval @xmath79=[\log 2 , 2\log 2]$ ] .
the present paper does not provide a complete description of the irregular part of the lyapunov spectrum even in the case @xmath80 .
we do not know how big the set @xmath81 is except in the case when @xmath18 has only one critical point in @xmath26 ( in which case @xmath81 consists only of the backward orbit of this critical point ) .
moreover , we do not know whether the set @xmath82 contains any points other than the backward orbits of critical points contained in @xmath26 and we only have some estimation for the hausdorff dimension of the set @xmath15 even for values @xmath13 , @xmath83 $ ] .
the paper is organized as follows . in section [ sec : tools ] we introduce the tools we are going to use in this paper . in particular , we construct a family of uniformly expanding cantor repellers with pressures pointwise converging to the pressure on @xmath26 ( proposition [ decaf ] ) .
section [ sec:3.5 ] discusses general properties of the spectrum of exponents . in section
[ sec : up ] we obtain upper bounds for the hausdorff dimension .
here we use conformal measures to deal with conical points ( proposition [ prop:1 ] ) and we prove that the set of non - conical points with positive upper lyapunov exponent is very small using the pullback construction ( proposition [ prop : wander ] ) . in section [ sec:4 ]
we derive lower bounds for the dimension .
to do so , we first consider the interior of the spectrum and we will use the sequence of cantor repellers from section [ sec : tools ] to obtain for any @xmath84 a big uniformly expanding subset of points with lyapunov exponent @xmath13 from which we derive our estimates .
we finally study the boundary of the spectrum and the irregular part of the spectrum using a construction , that generalizes the w - measure construction from @xcite .
given a compact @xmath18-invariant set @xmath85 , we denote by @xmath86 the family of @xmath18-invariant borel probability measures supported on @xmath28 .
we denote by @xmath87 the subset of ergodic measures . given @xmath88 ,
we denote by @xmath89 the lyapunov exponent of @xmath19 . notice that we have @xmath90 for any @xmath91 @xcite .
given @xmath92 , we define the function @xmath93 by @xmath94 given a compact @xmath18-invariant uniformly expanding set @xmath85 , the _ topological pressure _ of @xmath95 ( with respect to @xmath96 ) is defined by @xmath97 where @xmath98 denotes the entropy of @xmath18 with respect to @xmath19 .
we simply write @xmath99 and @xmath100 if we consider the full julia set @xmath26 and if there is no confusion about the system .
a measure @xmath101 is called _ equilibrium state _ for the potential @xmath95 ( with respect to @xmath3 ) if @xmath102 for every @xmath92 we have the following equivalent characterizations of the pressure function ( see @xcite , where further equivalences are shown ) .
we have @xmath103 here in the first equality the supremum is taken over the set @xmath104 of all ergodic @xmath18-invariant borel probability measures on @xmath26 that have a positive lyapunov exponent and are supported on some @xmath18-invariant uniformly expanding subset of @xmath26 . in the second equality the supremum
is taken over all uniformly expanding repellers @xmath85 .
in fact , in the second equality it suffices to take the supremum over all uniformly expanding cantor repellers , that is , uniformly expanding repellers that are limit sets of finite graph directed systems satisfying the strong separation condition with respect to @xmath18 , see section [ sec : hypsub ] .
let us introduce some further notation .
let @xmath105 where the given characterizations follow easily from the variational principle .
recall that , given @xmath106 , we define @xmath107 and @xmath108 note that @xmath109 we describe a construction of connecting two given hyperbolic subsets of the julia set by building bridges " between the sets .
we call a point @xmath110 _ non - immediately post - critical _ if there exists some preimage branch @xmath111 , @xmath112 , @xmath113 that is dense in @xmath26 and disjoint from @xmath63 . if @xmath18 is non - exceptional or if it is exceptional but @xmath67 then for every hyperbolic set all except possibly finitely many points ( in particular , all periodic points ) are non - immediately post - critical
. we will now consider a set @xmath28 that is an @xmath18-uniformly expanding cantor repeller ( ecr for short ) , that is , a uniformly expanding repeller and a limit set of a finite graph directed system ( gds ) satisfying the strong separation condition ( ssc ) with respect to @xmath18 .
recall that by a gds satisfying the ssc with respect to @xmath18 we mean a family of domains and maps satisfying the following conditions ( compare @xcite ) : * there exists a finite family @xmath114 of open connected ( not necessarily simply connected ) domains in the riemann sphere with pairwise disjoint closures .
* there exists a family @xmath115 of branches of @xmath116 mapping @xmath117 into @xmath118 with bounded distortion ( not all pairs @xmath119 must appear here ) .
+ note that a general definition of gds allows many maps @xmath120 from each @xmath121 to each @xmath118 .
here however there can be at most one , since we assume that @xmath18-critical points are far away from @xmath28 and that the maps @xmath120 are branches of @xmath116 .
* we have @xmath122 we assume that we have @xmath123 and hence that for each @xmath124 there exists @xmath125 and for each @xmath125 there exists @xmath124 such that @xmath126 .
we can view @xmath127 , @xmath128 , @xmath129 as vertices and @xmath130 as edges from @xmath125 to @xmath124 of a directed graph @xmath131 .
this definition easily implies that @xmath18 is uniformly expanding on the limit set @xmath28 of such a gds , and that @xmath28 is a repeller for @xmath18 .
clearly @xmath28 is a cantor set .
in fact a sort of converse is true ( though we shall not use this fact in this paper , but it clears up the definitions ) .
namely we observe the following fact .
[ lem:1111 ] if @xmath85 is an @xmath18-invariant compact uniformly expanding set that is a cantor set , then @xmath28 is contained in the limit set of a gds satisfying the ssc ( this limit set can be chosen to be contained in an arbitrarily small neighborhood of @xmath28 ) .
hence , @xmath28 is contained in an @xmath18-ecr set .
we can multiply the standard sphere riemann metric by a positive smooth function such that with respect to this new metric @xmath132 we have @xmath133 on @xmath28 .
it is easy to show that one can find an arbitrarily small number @xmath134 such that the neighborhood @xmath135 consists of a finite number of connected open domains @xmath118 with pairwise disjoint closures .
we account for our gds the branches of @xmath136 on the sets @xmath118 such that each @xmath137 intersects @xmath28 .
then @xmath120 maps each @xmath138 into some @xmath139 because it is a contraction ( by the factor @xmath140 ) .
hence the family of maps @xmath141 satisfies the assumptions of a gds with the ssc . in the proof of the following lemma we will `` build bridges '' between two ecr s . [ bridge ] for any two disjoint @xmath18-ecr sets @xmath142 , @xmath143 that both contain non - immediately postcritical points there exists an @xmath18-ecr set @xmath85 containing the set @xmath144 . if @xmath18 is topologically transitive on each @xmath145 , @xmath146 , @xmath147 , then @xmath18 is topologically transitive on @xmath28 .
let @xmath142 , @xmath143 be two sets satisfying the assumptions of the lemma and let @xmath148 and @xmath149 be two non - immediately postcritical points .
consider a family @xmath150 of open connected domains and a family @xmath151 of branches of @xmath116 mapping @xmath152 to @xmath153 that define the dgs s satisfying the ssc that have @xmath145 as their limit sets , for @xmath146 , @xmath147 , respectively .
let @xmath154 we can assume that each @xmath155 is an arbitrarily small neighborhood of @xmath145 , by replacing @xmath156 by @xmath157 , where by @xmath158 we denote the family of all compositions @xmath159 for each @xmath146 , @xmath147 let us choose a backward trajectory @xmath160 of the point @xmath161 ( the `` bridge '' ) such that @xmath162 @xmath163 for all @xmath164 , @xmath113 , @xmath165 and @xmath166 , @xmath167 . let us denote by @xmath168 the branch of @xmath169 that maps @xmath161 to @xmath160 , that is , let @xmath170 let @xmath171 be an open disc centered at @xmath161 that is contained in @xmath155 and satisfies @xmath172 for all @xmath164 , @xmath113 , @xmath165 ( note that this is possible provided we choose @xmath171 small enough ) and @xmath173 let us consider an integer @xmath174 such that the component of @xmath175 containing @xmath161 is contained in @xmath171 , that is , that @xmath176 for @xmath146 , @xmath147 .
now let us replace @xmath156 by @xmath177 , let us replace each map @xmath178 by the family of its restrictions to @xmath179 contained in @xmath139 and and let us denote by @xmath180 the union of those families .
this defines a gds with graph @xmath181 , for @xmath146 , @xmath147 .
now we restrict each bridge @xmath168 to the element @xmath182 of @xmath183 that contains @xmath161 . as the next step we consider @xmath184 where for @xmath185 we choose an arbitrary prolongation of the bridge @xmath160 by maps @xmath186 .
finally , we consecutively thicken slightly @xmath187 along the bridges such that @xmath188 .
for each @xmath189 let us denote by @xmath190 the branch of @xmath116 from @xmath187 to @xmath191 for @xmath192 , @xmath128 , @xmath193 .
let us denote by @xmath194 the family of all these branches . by construction
the family of domains @xmath195 and the family of maps @xmath196 form our desired gds with a graph @xmath197 satisfying the ssc and hence defines an @xmath18-ecr set @xmath85 that containes @xmath198 .
finally notice that this system has topologically transitive limit set @xmath28 since its transition graph @xmath199 is transitive .
this follows from the assumption that due to topological transitivity of @xmath200 the graphs @xmath201 are transitive and from the construction of the bridges .
our approach is to `` exhaust '' the julia set @xmath26 by some family of subsets @xmath202 and to show that the corresponding pressure functions converge towards the pressure of @xmath3 . in particular , in order to be able to conclude convergence of associated spectral quantities , it is crucial that each such @xmath203 is an invariant uniformly expanding and topologically transitive set .
we start by stating a classical result from pesin - katok theory . in follows for example from ( * ? ? ?
* theorems 10.6.1 and 11.2.3 ) . recall that an iterated function system ( ifs ) is a gds that is given by a complete graph .
[ l0 ] for every ergodic @xmath18-invariant measure @xmath19 that is supported on @xmath26 and has a positive lyapunov exponent , for every continuous function @xmath204 and for every @xmath205 , there exist an integer @xmath206 and an @xmath207-ecr set @xmath208 that is topologically transitive and a limit set of an ifs , such that @xmath209 @xmath210 where we use the notation @xmath211 , and in particular @xmath212 our aim is to apply lemma [ l0 ] to potentials @xmath213 and to use the resulting ecr sets to construct a sequence of @xmath214-ecr sets @xmath203 on which the pressure function @xmath215 converges pointwise to @xmath216 .
if the ecr sets generated by lemma [ l0 ] are pairwise disjoint , we can simply build bridges between such sets applying lemma [ bridge ] . in the general situation
we start with the following lemma .
[ bridge2 ] let @xmath28 be a topologically transitive @xmath207-ecr set .
let @xmath217 be a finite number of continuous functions .
then for any open disc @xmath218 intersecting @xmath28 and for any @xmath205 one can find a set @xmath219 and a natural number @xmath220 such that @xmath221 is an @xmath222-ecr set and for every @xmath223 satisfies @xmath224 consider in @xmath28 a clopen set @xmath225 contained in @xmath218 .
let @xmath226 .
let us choose @xmath227 large enough such that in the case that @xmath228 and @xmath229 the pullback satisfies @xmath230 .
note that it is enough to take @xmath231 where @xmath232 is the expanding constant on @xmath28 in the metric @xmath132 defined as in the proof of lemma [ lem:1111 ] .
note that the topological transitivity of @xmath233 implies that every pullback of @xmath225 can be continued by a bounded number of consecutive pullbacks until it hits @xmath225 , say this number is bounded by a constant @xmath234 . this way we obtain an ifs for @xmath222 , where @xmath235 , with its limit set @xmath221 contained in @xmath225 . recall the equivalent definition of tree pressure established in ( * ? ? ?
* theorems a , a.4 ) . due to topological transitivity in the definition of pressure
we can consider separated sets that are contained in the set of preimages @xmath236 .
therefore , for any @xmath223 the pressures with respect to @xmath237 and with respect to @xmath238 differ by at most @xmath239 from each other .
as @xmath227 can be chosen arbitrarily big , this proves the lemma .
the following approximation results are fundamental for our approach . [ decaf ]
assume that @xmath18 is non - exceptional or that @xmath18 is exceptional but @xmath80 .
then there exists a sequence @xmath240 of positive integers and a sequence @xmath202 of @xmath214-invariant uniformly expanding topologically transitive sets such that for every @xmath92 , we have @xmath241 for every @xmath242 we have @xmath243 and @xmath244 where @xmath245 and @xmath246 are defined as in and but with @xmath247 instead of @xmath248 . to prove it is enough to construct an @xmath214-ecr set @xmath202 such that @xmath249 for all @xmath250 $ ] . recall that we have . as @xmath251 is uniformly lipschitz continuous ,
we only need to check for a finite number of potentials @xmath252 . given @xmath253
, we apply lemma [ l0 ] to potentials @xmath223 , obtaining a family of @xmath254-ecr sets @xmath255 on which the pressure @xmath256 .
we then apply lemma [ bridge2 ] to construct a family of pairwise disjoint @xmath257-ecr sets @xmath258 that satisfy latexmath:[$\frac 1 { \ell_i } p_{f^{\ell_i } @xmath258 are also disjoint @xmath214-ecr sets for @xmath260 and by our assumption contain non - immediately post - critical points .
hence we can consecutively apply lemma [ bridge ] to them .
we obtain an @xmath214-ecr set satisfying for all @xmath223 .
this proves . as @xmath251 and
@xmath261 form a legendre pair , and follow from by a result of wijsman @xcite .
it is enough for us to work with hyperbolic sets for some iterations of @xmath18 instead of hyperbolic sets for @xmath18 ( in other words , to use instead of ) .
however , notice that we could instead also extend the set @xmath262 to an @xmath18-ecr set using lemma [ lem:1111 ] .
the dynamical properties of any measure @xmath263 with respect to @xmath3 are captured through its jacobian .
the _ jacobian _ of @xmath263 with respect to @xmath3 is the ( essentially ) unique function @xmath264 determined through @xmath265 for every borel subset @xmath266 of @xmath26 such that @xmath267 is injective . in particular , its existence yields the absolute continuity @xmath268 . a probability measure @xmath263 that satisfies @xmath269 is called _
@xmath270-conformal measure_. if @xmath271 then one can always find a @xmath270-conformal measure @xmath272 that is positive on each open set intersecting @xmath26 , see @xcite .
when @xmath273 such a measure can always be found if @xmath18 is not an exceptional map or if @xmath18 is exceptional but @xmath80 ( see ( * ? ? ?
* appendix a.2 ) ) .
when @xmath3 is not uniformly expanding , we can still observe a slightly weaker form of non - uniform hyperbolicity .
we recall two concepts that have been introduced .
a number @xmath274 is called a _ hyperbolic time _ for a point @xmath7 with exponent @xmath275 if @xmath276 it is an immediate consequence of the pliss lemma ( see , for example , @xcite ) that for a given point @xmath4 , for any @xmath277 there exist infinitely many hyperbolic times for @xmath7 with exponent @xmath275 .
we denote by @xmath278 the maximal distortion of a map @xmath120 on a set @xmath279 .
after @xcite , we will call a point @xmath4 _ conical _ if there exist a number @xmath134 , a sequence of numbers @xmath280 and a sequence @xmath281 of neighborhoods of @xmath7 such that @xmath282 and that @xmath283 is bounded uniformly in @xmath284 .
in the following two lemmas we will investigate which numbers can occur at all as upper / lower lyapunov exponents .
[ lem : finit1 ] we have @xmath285 if @xmath26 does not contain any critical point of @xmath18 then we have @xmath286 consider an arbitrary @xmath4 and a sequence @xmath287 such that @xmath288 and @xmath289 in the weak@xmath290 topology .
the limit measure @xmath19 is @xmath18-invariant , ( * ? ? ?
* theorem 6.9 ) . define @xmath291 notice that @xmath292 is a monotonically decreasing sequence of continuous functions that converge pointwise to @xmath293 .
hence we obtain @xmath294 where the equality follows from the lebesgue monotone convergence theorem .
this proves the first statement .
the second statement follows simply from the fact that @xmath295 is continuous on @xmath26 if @xmath18 has no critical points in @xmath26 .
we remark that the same method of proof gives a slightly stronger result than the fact that @xmath296 for every @xmath7
. namely we have @xmath297 see ( * ? ? ?
* proposition 2.3 .
item 2 ) . recall that on a set of total probability we have @xmath298 @xcite .
in fact , a better estimate can be given for any lyapunov regular point @xmath7 ( compare the proof of ( * ? ? ?
* proposition 4.1 ) ) .
[ lem : finit2 ] if @xmath4 has a finite lyapunov exponent @xmath9 then @xmath299 , that is , we have @xmath300 if there are no critical points in @xmath26 then @xmath81 is empty .
if there is only one critical point in @xmath26 then @xmath81 consists only of this critical point and its preimages .
let @xmath4 be a lyapunov regular point with exponent @xmath9 and assume that @xmath301 .
it is enough for us to prove that there exists a periodic point with lyapunov exponent arbitrarily close to @xmath9 , the contradiction will then follow from the definition of @xmath302 .
note first that if @xmath9 exists and is finite then @xmath303 must be a cauchy sequence and hence satisfies @xmath304 by ( * ? ?
* corollary to lemma 6 ) , we then can conclude that @xmath298 .
choose now a small number @xmath305 .
let @xmath306 be a hyperbolic time for @xmath7 with exponent @xmath307 ( recall that @xmath7 has infinitely many hyperbolic times with that exponent , and hence that @xmath306 can be chosen arbitrarily big ) .
because of , there exists an integer @xmath308 such that for all @xmath309 we have @xmath310 and we can assume that @xmath311 .
we start with a construction that is standard in pesin theory . for each @xmath312 , @xmath113 , @xmath313
we define @xmath314 where @xmath315 is the greatest degree of a critical point of @xmath18 . this way we define a sequence of balls that are centered at points of the backward trajectory of @xmath316 and that have diameters shrinking slower than the derivative of @xmath317 along this branch and at the same time have diameters much smaller than their distance from any critical point . as we have @xmath318
, implies that for @xmath306 big enough for any @xmath319 the set @xmath320 does not contain any critical point and that @xmath321 where @xmath322 is some constant that depends only on @xmath18 .
+ * claim : * if @xmath306 is sufficiently big then for any @xmath323 , @xmath113 , @xmath306 the map @xmath324 is univalent and has bounded distortion @xmath325 on the set @xmath326 and satisfies @xmath327 .
+ to prove the above claim , note that the ball @xmath328 shrinks as @xmath306 increases .
hence , it is enough to prove the statement for @xmath329 .
the statement for the initial finitely many steps @xmath330 , @xmath113 , @xmath306 is then automatically provided @xmath306 is big enough .
let us assume that @xmath306 is sufficiently big such that also @xmath331 by construction of the family @xmath332 and by , for each @xmath333 we have @xmath334 hence , if for @xmath335 we have @xmath327 for every @xmath127 , @xmath113 , @xmath336 then implies @xmath337 on the other hand , recall that @xmath306 is a hyperbolic time for @xmath7 with exponent @xmath307 .
hence , if @xmath338 then @xmath339 .
the above claim now follows by induction over @xmath125 .
let us consider the set @xmath340 notice that @xmath341 is nonempty whenever the hyperbolic time @xmath306 is big enough .
for such @xmath306 , let @xmath342 . from our distortion estimations for @xmath343 in the claim
we obtain @xmath344 in the remaining proof we will follow closely techniques in ( * ? ? ? * section 3 ) .
let us fix some arbitrary @xmath18-invariant uniformly expanding set @xmath85 that has positive hausdorff dimension ( the existence of such a set follows for example from lemma [ l0 ] ) .
as @xmath63 is a finite set , we have @xmath345 and we can find a point @xmath346 .
in particular @xmath347 for @xmath306 large .
note that in particular for every @xmath306 large enough we have @xmath348 and hence on the disk @xmath349 any pull - back @xmath350 is univalent . by (
* lemma 3.1 ) , there exists a number @xmath351 depending only on @xmath18 and a sequence of disks @xmath352 such that @xmath353 is connected , @xmath354 is the center of disk @xmath355 , @xmath356 is the center of disk @xmath357 , @xmath358 and the number of disks is bounded by @xmath359 . by the koebe distortion lemma , for each branch of @xmath350 and for every @xmath155 we have @xmath360 where @xmath361 is some constant .
hence , in particular @xmath362 for some @xmath363 . together with , this implies that @xmath364 whenever @xmath306 has been chosen large enough . by hyperbolicity of @xmath96 , there exist positive constants @xmath365 and @xmath366 such that for all @xmath367 the map @xmath368 is univalent and has bounded distortion @xmath369 on @xmath370 .
recall that the julia set @xmath26 has the property that there exists @xmath371 such that the image @xmath372 is equal to @xmath26 .
let @xmath125 be the smallest positive integer such that @xmath373 .
hence @xmath374 and it is easy to show that @xmath375 for some constants @xmath376 , @xmath377 .
so in particular , one of the preimages @xmath378 is in @xmath370 and the corresponding pull - back @xmath379 satisfies @xmath380 .
it follows from that @xmath381 is univalent and hence that there is a repelling fixed point @xmath382 in @xmath383 . using the above
, the lyapunov exponent of @xmath384 can be estimated by @xmath385 now recall that the hyperbolic time @xmath306 can be chosen arbitrarily large and that @xmath386 was chosen arbitrarily small .
this proves the first claim of the lemma .
if there are no critical points in @xmath26 then @xmath295 ( and hence @xmath387 ) is bounded from below .
the remaining claim of the lemma follows from ( * ? ? ?
* lemmas 2.1 and 2.3 ) .
in this section we will derive upper bounds for the hausdorff dimension of the level sets .
we first start with a particular case .
we study the particular case that there is no critical point in the julia set @xmath26 ( though , parabolic points in @xmath26 are allowed ) .
we start by taking a more general point of view and investigate those points that have some least degree of hyperbolicity . in terms of lyapunov exponents ,
this concerns points @xmath7 with @xmath388 .
we begin our analysis by presenting a simple lemma , that will be useful shortly after .
[ lem : sequ ] let @xmath389 be a sequence of real numbers such that @xmath390 converges to zero but @xmath389 does not have a limit .
then for any natural number @xmath391 and for any number @xmath392 $ ] there exists a subsequence @xmath393 such that @xmath394 and for every @xmath124 we have @xmath395 we will restrict our hypothesis to the case that @xmath396 .
the general case then follows from considering the subsequence @xmath397 . first note that every number @xmath398 $ ] is the limit of some subsequence of @xmath389 .
hence , if @xmath399 then for every @xmath400 there must exist infinitely many numbers @xmath401 such that @xmath402 similarly , if @xmath403 then for every @xmath400 there must exist infinitely many numbers @xmath401 such that @xmath404 since we assume that the sequence @xmath389 does not have a limit , @xmath405 must satisfy one of the abovementioned properties . hence , if we choose a decreasing family @xmath406 and for each @xmath407 one of the corresponding numbers @xmath408 , we obtain @xmath409 the second part of the assertion is immediately satisfied . lemma [ lem : sequ ] will help us to establish some bounded distortion properties .
the following result implies in particular that every point @xmath7 with @xmath388 is conical ( recall the definition of a conical point in section [ sec : hypinst ] ) .
[ lem : bdp ] assume that @xmath26 does not contain any critical points of @xmath18 .
let @xmath4 be a point with @xmath410 .
then there exists a number @xmath411 such that for every @xmath412\setminus \{0\}$ ] there exists a number @xmath305 and a sequence @xmath413 such that * @xmath414 , + * @xmath415 here @xmath129 is a universal constant , while @xmath386 depends on the number @xmath405 but not on the point @xmath7 . as @xmath26 does not contain any critical point , the only accumulation points in @xmath26 of the orbit of some critical point can be parabolic points .
let @xmath391 be the least common multiplier of the periods of all parabolic points in @xmath26
. given a number @xmath412\setminus \{0\}$ ] , there exists a number @xmath416 such that if @xmath417 is @xmath418-close to some parabolic point then @xmath419 further , there exists a number @xmath420 such that if @xmath421 is @xmath418-away from any parabolic point then no orbit of a critical point passes through @xmath422 . to prove the claimed property
, it will suffice to find a sequence @xmath423 for which i ) is satisfied and for which @xmath424 is in distance at least @xmath418 from any parabolic point .
indeed , in such a situation the backward branch of @xmath425 will not catch any critical point and the distortion estimations ii ) will follow from the koebe distortion lemma .
if @xmath7 is a lyapunov regular point and @xmath426 is its lyapunov exponent ( which , by our assumptions , must be positive ) then the claim i ) is automatically satisfied . in this case
we just need to choose @xmath427 such that @xmath424 is far away from any parabolic point .
note that there must be infinitely many such times @xmath427 because otherwise the lyapunov exponent at @xmath7 would be no greater than @xmath428 .
if @xmath7 is not a lyapunov regular point , then we apply lemma [ lem : sequ ] to the sequence @xmath429 notice that @xmath430 is satisfied , because there are no critical points in @xmath26 and hence @xmath431 is uniformly bounded .
hence , from the first assertion of lemma [ lem : sequ ] we obtain a sequence @xmath423 that satisfies i ) .
notice that we have @xmath432 whenever @xmath424 is @xmath418-close to some parabolic point .
this inequality can not be true for big @xmath427 ( when @xmath433 is already close to @xmath405 ) because of the second part of assertion of lemma [ lem : sequ ] .
this proves that for any time @xmath427 large enough the point @xmath424 is @xmath418-away from any parabolic point . we are now prepared to prove an upper bound for the hausdorff dimension of the level sets under consideration . to start with the most general approach that will be needed in the subsequent analysis
, we first study a set of points @xmath7 for which @xmath388 and for which additionally the lyapunov exponent ( possibly with respect to some subsequence of times ) is guaranteed to be within a given interval @xmath434 $ ] .
let us first introduce some notation .
given @xmath435 , @xmath56 , let @xmath436 [ prop:1 ] assume that @xmath26 does not contain any critical points of @xmath18 . for every @xmath56 and @xmath437 we have @xmath438 if @xmath439 or @xmath440 then @xmath441 .
by lemma [ lem : bdp ] , for every point @xmath442 there exist a number @xmath443\setminus \{0\}$ ] , a number @xmath305 , and a sequence @xmath423 of numbers such that @xmath444 and @xmath445 recall that for every @xmath92 there exists a @xmath446-conformal measure @xmath272 that gives positive measure to any open set ( see section [ sec : conf ] ) .
hence there exists @xmath447 such that for every @xmath427 we have @xmath448 . using again the distortion estimates , we can conclude that @xmath449 which implies that @xmath450 and in particular that this limit exists .
hence , there exists @xmath451 such that for every @xmath452 we have @xmath453 here we used lemma [ lem : bdp ] to obtain the last inequality . applying and we yield @xmath454 for any @xmath455 $ ] ,
@xmath456 , and @xmath205 there exist a small interval @xmath457 , @xmath458 , containing @xmath459 and a number @xmath92 such that for all @xmath460 @xmath461 we can choose a countable family of intervals @xmath462 covering @xmath434\setminus \{0\}$ ] and a sequence of corresponding numbers @xmath463 . defining the measure
@xmath464 we obtain @xmath465 where the second inequality follows from . applying the frostman lemma
we obtain that @xmath466 since @xmath467 can be chosen arbitrarily small , this finishes the proof of the first claim .
the second claim was already proved in lemma [ lem : finit1 ] .
note that for every @xmath468 $ ] , @xmath469 , we have @xmath470 , which readily proves the following result . [ cor:1 ] under the hypotheses of proposition [ prop:1 ]
, we have @xmath471 in particular , for every @xmath106 , we have @xmath472 we now consider the general case that there are critical points inside the julia set .
we need two technical results from the literature .
the first one is the following telescope lemma from @xcite .
recall the definition of hyperbolic times given in section [ sec : hypinst ] .
[ telescope ] given @xmath205 and @xmath473 , there exist constants @xmath474 and @xmath475 such that the following is true .
given @xmath4 with upper lyapunov exponent @xmath476 , for every number @xmath477 , for every @xmath478 being a hyperbolic time for @xmath7 with exponent @xmath275 , and for every @xmath479 we have @xmath480 to formulate our second preliminary technical result we need the following construction , see @xcite .
+ * pullback construction : * fix some @xmath206 and let @xmath481 . fix some @xmath482 and let @xmath483 be some backward trajectory of @xmath354 , i.e. @xmath484 and @xmath485 for every @xmath146 , @xmath113 , @xmath486 .
let @xmath487 be the smallest integer for which @xmath488 contains a critical point .
for every @xmath333 let then @xmath489 be the smallest integer greater than @xmath490 such that @xmath491 contains a critical point and so on . in this way , for each backward branch @xmath492 we construct a sequence @xmath493 that must have a maximal element not greater than @xmath306 .
let this element be @xmath124 and consider the set @xmath279 of all pairs @xmath494 built from all the backward branches of @xmath18 that start from @xmath354 .
let @xmath495 .
we have the following estimate , see ( * ? ? ?
* lemma 3.7 ) and ( * ? ? ? * appendix a ) .
[ prs ] given @xmath496 , there exist @xmath351 and @xmath497 such that for all @xmath498 we have @xmath499 uniformly in @xmath354 and @xmath306 .
recall the definition of conical points in section [ sec : hypinst ] . using the above two lemmas we can now show the following result .
[ prop : wander ] the set of points @xmath4 that are not conical and satisfy @xmath388 has hausdorff dimension zero .
let us choose some numbers @xmath500 and @xmath205 .
let @xmath501 where @xmath322 and @xmath502 are constants given by lemma [ telescope ] and where @xmath503 is given by lemma [ prs ] .
we can choose a finite family of balls @xmath504 of radius @xmath505 such that any ball of radius @xmath506 intersecting @xmath26 must be contained in one of the balls @xmath507 .
in the case that we can prove existence of a sequence @xmath508 such that @xmath509 does not contain critical points for any @xmath510 , the koebe distortion lemma will imply that @xmath7 is a conical point for @xmath391 , @xmath508 , and @xmath511 .
let @xmath512 be the set of points @xmath4 with upper lyapunov exponent greater than @xmath275 for which with @xmath391 chosen above for all @xmath513 the backward branch of @xmath350 from ball @xmath514 onto a neighborhood of @xmath7 will necessarily meet a critical point , that is , for some @xmath479 we have @xmath515 first we claim that @xmath516 .
let us denote by @xmath517 the subset of @xmath512 for which @xmath513 is a hyperbolic time with exponent @xmath275 .
recall that @xmath518 implies that there exist infinitely many hyperbolic times for @xmath7 with exponent @xmath275 .
hence , we have @xmath519 let @xmath520 .
let @xmath521 be the ball that contains @xmath514 .
we will apply the pullback construction " for lemma [ prs ] to the point @xmath354 , the numbers @xmath306 , @xmath522 , and to the backward branch @xmath523 . and let @xmath524 and @xmath525 be given by the pullback construction ( compare figure [ fig.reia ] ) .
.,width=264 ] we first note that lemma [ telescope ] and @xmath526 imply that @xmath527 this implies @xmath528 and hence @xmath529 because @xmath530 . thus , with fixed @xmath306 and @xmath531 , by lemma [ prs ] the point @xmath7 must belong to one of at most @xmath532 preimages of @xmath531 . as @xmath533 and @xmath534 , by lemma [ telescope ] this pre - image of @xmath531 has diameter not greater than @xmath535 .
hence we showed that every point in @xmath536 belongs to the @xmath306th pre - image of some ball @xmath531 along a backward branch for that @xmath537 ( where @xmath524 is as in the pullback construction ) .
thus , by lemma [ telescope ] the set @xmath538 is contained in a union of at most @xmath539 sets of diameter not greater than @xmath540 . using those sets to cover @xmath541 and applying , we obtain @xmath542 as @xmath467 can be chosen arbitrarily small , the claim follows .
finally note that the set of points that we want to estimate in the proposition is contained in the union @xmath543 since for any set in this union its hausdorff dimension is zero , the assertion follows . note that in the above proof we were able to show something more . namely notice that the choice of @xmath391 depended on @xmath275 alone ( and not directly on the point @xmath7 ) . we are now prepared to prove the following estimate .
[ prop : kjof ] let @xmath544 .
we have @xmath545 let @xmath13 and @xmath546 be like in the assumptions and let @xmath547
. by proposition [ prop : wander ] we can restrict our considerations to the case that @xmath7 is a conical point with corresponding number @xmath134 , sequence @xmath423 , and family of neighborhoods @xmath548 .
hence , there exist numbers @xmath305 and @xmath549 such that @xmath550 choosing , if necessary , a subsequence of @xmath423 , we can find @xmath551 $ ] for which we have @xmath552 recall that for any @xmath92 there exists a @xmath446-conformal measure @xmath272 that gives positive measure to any open set that intersects @xmath26 ( see section [ sec : conf ] ) . hence there exists a number @xmath553 for which for every @xmath427 we have @xmath554 .
using again distortion estimates , we can conclude that @xmath555 the rest of the proof is similar to the proof of proposition [ prop:1 ] .
first we obtain that for every @xmath92 we have @xmath556 choosing the right number @xmath315 , we then conclude that @xmath557 .
we finish the proof by constructing a measure such that for an arbitrarily small chosen number @xmath467 and some number @xmath558 $ ] the lower pointwise dimension of that measure at every @xmath559 is not greater than @xmath560 .
notice that in the general case , in which we do have critical points in @xmath26 , for a point @xmath4 with @xmath561 we can not apply the same techniques as in section [ sec : nocrit ] . in particular , in the above proof for each point @xmath562 we only know that there exists some number @xmath563 $ ] to which the lyapunov exponents over some subsequence of times @xmath423 will converge , while in proposition [ prop:1 ] we were able to take an arbitrary number @xmath405 in that interval .
hence , to show the following result we can only consider the set @xmath564 and not @xmath565 for an arbitrary number @xmath468 $ ] , and the following corollary is weaker than corollary [ cor:1 ] .
however , the implications for the regular part of the spectrum remain the same . proposition [ prop : kjof ] and @xmath564 readily imply the following result . for @xmath566\setminus\{0\}$
] we have @xmath567
in this section we will derive lower bounds for the hausdorff dimension .
we will either assume that @xmath18 is a non - exceptional map or that @xmath18 is exceptional but @xmath80 . recall that under those assumptions proposition [ decaf ] is valid ,
so we can approximate the pressure with respect to @xmath3 with pressures that are defined with respect to a sequence of cantor repellers @xmath568 that were constructed in section [ sec : hypsubpress ] .
one more case we would like to exclude is @xmath569 .
it is not very interesting because in this case any measure supported on any hyperbolic set @xmath202 has lyapunov exponent @xmath13 .
hence we automatically have @xmath570 and the supremum on the right hand side is in this case equal to @xmath571 .
therefore , in the following considerations we will assume that @xmath51 .
we use the sequence of cantor repellers to obtain for any exponent from the interior of the spectrum a big uniformly expanding subset of points with lyapunov exponent @xmath13 that provides us with an estimate from below .
let us consider the sequence of cantor repellers @xmath568 from proposition [ decaf ] .
by , for each number
@xmath574 there exists @xmath575 such that @xmath576 for every @xmath577 .
obviously we have @xmath578 since @xmath203 is a uniformly expanding repeller with respect to @xmath214 , for any exponent @xmath579 there exists a unique number @xmath580 such that @xmath581 and an equilibrium state @xmath582 for the potential @xmath583 ( with respect to @xmath568 ) such that the lyapunov exponent of @xmath582 with respect to @xmath214 is equal to @xmath584 ( compare the classical results in the introduction ) . for the measure
@xmath585 we have @xmath586 .
hence , the variational principle implies @xmath587 we obtain that @xmath588 whenever @xmath263 is an @xmath18-invariant ergodic borel probability measure with positive lyapunov exponent @xcite .
this implies that for every @xmath577 @xmath589 from here the we can conclude that @xmath590 . together with proposition [ decaf ]
the statement is proved .
unfortunately , the above approach does not suffice to analyze the level sets for exponents from the boundary of the spectrum .
our main goal in this section is to prove the following result .
it will not only enable us to describe the boundary of the spectrum but also provide us with dimension lower bounds for level sets of irregular points .
[ closing ] let @xmath591 be a sequence of subsets of @xmath26 .
we assume that each @xmath145 is a uniformly expanding repeller for some iteration @xmath592 and contains non - immediately postcritical points .
let @xmath593 be a sequence of hlder continuous potentials and let @xmath594 be a sequence of equilibrium states for @xmath223 with respect to @xmath595 . then @xmath596 and @xmath597 since we assume @xmath51 , there exists a sequence @xmath591 of uniformly expanding repellers and a sequence @xmath594 of equilibrium states for the potential @xmath605 with respect to @xmath200 such that @xmath606 and @xmath607 and @xmath608 .
theorem [ closing ] then implies that @xmath609 since @xmath610 , this proves the claimed estimate .
proposition [ coco ] implies that @xmath614 . as @xmath615 ,
the pressure function @xmath616 is nonnegative and @xmath617 .
since @xmath26 does not contain any recurrent critical points of @xmath18 , we can apply ( * ? ? ?
* theorem 4.5 ) and conclude that @xmath618 . to prove theorem [ closing
] , we will construct a sufficiently large set of points that have precisely the given lower / upper lyapunov exponent .
note that such a set will not be seen " by any invariant measure in the non - trivial case that lower and upper exponents do not coincide . also points with lyapunov exponent zero , though lyapunov regular , will not be seen " by any interesting invariant measure .
our approach is to show that such a set is large with respect to some necessarily non - invariant probability measure .
it is generalizing the construction of so - called w - measures introduced in @xcite for which we strongly made use of the markov structure that is not available to us here .
[ shadowing ] let @xmath619 be a conformal map and @xmath620 be a compact @xmath120-invariant hyperbolic topologically transitive set , @xmath19 a @xmath120-invariant ergodic measure on @xmath28 with lyapunov exponent @xmath621 , entropy @xmath622 , and hausdorff dimension @xmath623 .
let @xmath624 be an open set of positive measure @xmath19 . if @xmath625 is small enough then for any @xmath205 for @xmath19-almost every point @xmath626 there exist a number @xmath411 and a sequence @xmath508 such that for each @xmath627 there is a set @xmath628 such that for all @xmath629 * @xmath630 , * @xmath631 for all @xmath632 , * the branch @xmath633 mapping @xmath634 onto @xmath635 extends to all @xmath636 and the distortion of the resulting map is bounded by @xmath129 , * we have @xmath637 * for @xmath638 we have @xmath639 * for any @xmath640 and @xmath134 we have @xmath641
* we have @xmath642 as @xmath643 is uniformly expanding , there exist numbers @xmath644 and @xmath645 such that for every @xmath646 , @xmath647 , @xmath648 , and a backward branch @xmath649 mapping @xmath634 onto @xmath354 for every @xmath650 , @xmath651 we have @xmath652 and in particular the mapping @xmath649 extends to all of @xmath636 .
moreover , there is some constant @xmath653 not depending on @xmath306 or @xmath354 such that @xmath654 we assume that @xmath625 is so small that for any two points @xmath7 , @xmath354 with distance from @xmath28 and mutual distance @xmath655 , for any point @xmath656 there is at most one point @xmath657 such that @xmath658 .
this is true for any small enough number @xmath625 because @xmath28 is in positive distance from any critical point of @xmath120 .
let @xmath659 be such that @xmath660 and that @xmath661 is nonempty and of positive measure , which is possible whenever @xmath625 is small enough .
let @xmath205 .
there is a set of points @xmath665 with @xmath666 ( actually , it can be chosen to have arbitrarily large measure ) and a number @xmath667 such that : for every @xmath668 and for every @xmath669 we have we choose now a family @xmath676 of subsets of @xmath221 such that for every @xmath668 the set @xmath677 is a maximal @xmath678-separated subset of @xmath221 .
given @xmath668 , let @xmath679 .
for each @xmath680 let @xmath681 for every @xmath682 let @xmath683 .
obviously , all the sets @xmath684 are pairwise disjoint . by ( c1 ) the trajectory of every point from @xmath685 visits @xmath686 at some time @xmath668 ( in fact , at infinitely many times ) .
let @xmath7 be such a point and @xmath306 be such a time , that is , @xmath687 . because @xmath677 is maximal
, @xmath7 must be @xmath688-close to some point @xmath689 .
hence we have @xmath690 and @xmath691 this shows that @xmath682 .
this together with ( c1 ) implies that @xmath692 and hence , as it is satisfied for all @xmath693 , we obtain @xmath694 for infinitely many @xmath695 .
this gives us a sequence of times @xmath696 and points @xmath697 for which the assertion of the lemma is true .
indeed , @xmath698 and @xmath699 for @xmath700 are disjoint because we are in a sufficiently large distance from @xmath63 .
we also know that those maps @xmath701 have uniformly bounded distortion .
recall that @xmath702 , which implies that @xmath703 and @xmath704 for @xmath700 are not only disjoint but in distance that is comparable to the sum of their diameters and v ) follows .
further properties i ) and iii ) are checked directly from the construction .
moreover , vi ) follows from ( c4 ) , iv ) follows from our choices of @xmath125 , ii ) from ( c2 ) and vii ) from iv ) and ( c3 ) . as a first step
, we are going to construct `` bridges '' between the repellers @xmath145 ( compare figure [ fig.heute ] ) .
this is very similar to the proof of lemma [ bridge ] though here we will not necessarily require that the repellers are disjoint . to fix some notation ,
let @xmath705 be a collection of _ bridges _ , that is , let @xmath706 where the numbers @xmath707 , @xmath708 , @xmath709 and the points @xmath710 are appropriately chosen such that @xmath711 is a homeomorphism and that @xmath712 and @xmath713 the particular choice of the numbers and points will be specified in the following so that we are able to apply lemma [ shadowing ] to the each of the sets @xmath145 and the maps @xmath714 .
let us outline the following cantor set construction .
lemma [ shadowing ] enables us to select sufficiently many preimages for each of the disks @xmath507 ( as we choose @xmath715 and then select preimages @xmath716 using the lemma ) . the construction of this cantor set is easily described in terms of backward branches : at level @xmath284 we start with the disk @xmath507 , apply lemma [ shadowing ] and find a large number of components of @xmath717 in @xmath718 , then we go backwards through the bridge obtaining the components in @xmath719 .
we repeat the procedure in @xmath719 , @xmath113 , @xmath720 . given the sequence of uniformly expanding repellers @xmath595 with equilibrium states @xmath721 with lyapunov exponent @xmath722 and entropy @xmath723 ( both with respect to the map @xmath592 ) , let us denote @xmath724 and @xmath725 let @xmath726 the constants @xmath709 can be chosen to be arbitrarily small ( at the cost of decreasing @xmath707 and increasing @xmath708 , thus changing @xmath727 and @xmath728 accordingly ) .
in particular we can choose each @xmath709 sufficiently small such that lemma [ shadowing ] applies to the map @xmath714 and the set @xmath145 .
let @xmath729 we choose a fast decreasing sequence @xmath730 .
we will denote by @xmath731 the numbers @xmath732 as given by lemma [ shadowing ] ( for @xmath714 ) where we choose each point @xmath733 such that @xmath715 .
let @xmath734 for one point @xmath635 and @xmath735 as provided by lemma [ shadowing ] .
we have @xmath736 and for every @xmath737 and @xmath738 we have @xmath739 ( note that @xmath735 can be chosen to be arbitrarily big , as guaranteed by lemma [ shadowing ] ) .
we apply lemma [ shadowing ] now to @xmath740 , @xmath741 , @xmath742 , @xmath743 , @xmath744 , @xmath745 , which provides us with a family @xmath746 .
let @xmath747 for @xmath748 .
those sets are contained in the set @xmath749 .
we also have @xmath750 and for every @xmath751 and @xmath752 we have @xmath753 such sets will be used later to distribute the w - measure accordingly .
it provides us also with a family of sets @xmath754 such that @xmath755 .
such sets will be used later to define our cantor set on which the w - measure is supported .
we repeat this procedure for every @xmath284 , using lemma [ shadowing ] repeatedly : for every @xmath756 we find a family @xmath757 and the corresponding components @xmath758 of @xmath759 contained in @xmath171 that satisfy @xmath760 and for every @xmath761 and @xmath632 @xmath762 and @xmath763 in addition , by lemma [ shadowing ] iv ) we have @xmath764 and by lemma [ shadowing ] vi ) for any @xmath765 and @xmath134 we have @xmath766 note here that for every @xmath767 we have @xmath768 let @xmath769 let @xmath770 be the component of @xmath771 for which @xmath772 .
similarly , let @xmath773 be the component of @xmath774 for which @xmath775 .
+ the sets @xmath776 form a decreasing sequence of unions of topological balls .
moreover , the pair @xmath777 is an image of @xmath778 under a branch of the map @xmath779 , the distortion of that branch is bounded by @xmath780 and the absolute value of its derivative is between @xmath781 ( @xmath782 depend on @xmath284 only ) . *
( lyapunov exponents on the islands ) for @xmath784 and @xmath785 we have @xmath786 * ( lyapunov exponents on the bridges ) for @xmath784 and @xmath787 we have @xmath788 * we have @xmath789 * there are at least @xmath790 sets @xmath791 contained in every set @xmath792 , * for any @xmath793 and any @xmath794 we have @xmath795 * we have @xmath796 * for any @xmath765 and @xmath134 we have @xmath797 * we have @xmath798 additional assumptions on @xmath508 and properties guaranteed by lemma [ shadowing ] will enable us to estimate the hausdorff and packing dimensions of the constructed set and describe the upper and lower lyapunov exponents at each point .
this will be done in the following .
we continue to consider the cantor set @xmath266 constructed in .
let @xmath19 be the probability measure which on each level @xmath284 is distributed on the cylinder sets @xmath799 of level @xmath284 in the following way @xmath800 we extend the measure @xmath19 arbitrarily to the borel @xmath275-algebra of @xmath266 .
we call the probability measure a _ w - measure _ with respect to the sequence @xmath801 . + after these preparations
, we are now able to prove theorem [ closing ]
. we will continue to use the notations in the above construction of the set @xmath266 .
in the course of the following proof we will choose some sequence @xmath406 and then construct a sequence of positive integers @xmath508 . here
each of those numbers @xmath627 has to satisfy several conditions that depend on @xmath406 , the parameters of the hyperbolic sets @xmath802 and the measures @xmath803 , and the previously chosen numbers @xmath804 , @xmath805 , @xmath147 , @xmath113 , @xmath806 . naturally , it is always possible to satisfy all those conditions at the same time .
we will first check that the cantor set defined in satisfies @xmath807 ( under some appropriate assumptions about @xmath508 ) .
then we will estimate the hausdorff and packing dimensions of @xmath266 using the w - measure @xmath19 defined in .
let us first consider the lyapunov exponent at a point in the set @xmath266 .
let @xmath808 for @xmath809 we have @xmath810 we know that @xmath811 stays in distance at least @xmath812 from the critical points for every @xmath813 .
hence , for @xmath814 we have @xmath815 and for @xmath735 big enough ( in comparison with @xmath816 and @xmath817 ) the right hand side of is between @xmath818 and @xmath819 .
for @xmath820 we have @xmath821 while for @xmath822 we have @xmath823 estimating the second summand using lemma [ aproposition ] i ) and the third one using lemma [ aproposition ] ii ) and assuming that @xmath627 is big enough ( in comparison with @xmath824 , @xmath825 , @xmath826 , and @xmath827 ) , we can first prove that @xmath828 ( by induction ) and then prove that for all @xmath829 we have @xmath830 as the upper ( the lower ) lyapunov exponent of @xmath7 equals the upper ( the lower ) limit of @xmath831 as @xmath832 , we have shown that every point @xmath833 satisfies @xmath834 let us now estimate the hausdorff and packing dimensions of the set @xmath266 . to do so
, we will apply the frostman lemma .
let us calculate the pointwise dimension of the measure @xmath19 defined in at an arbitrary point @xmath833 .
notice that we can write @xmath835 for some appropriate symbolic sequence @xmath836 . by lemma [ aproposition ]
v ) , the ball @xmath837 does not intersect any of the sets @xmath838 if @xmath839 whenever we have @xmath840 consider @xmath841 .
we have @xmath842 let @xmath843 .
we continue with @xmath844 using lemma [ aproposition ] vi ) and vii ) we can estimate @xmath845 let @xmath846 using lemma [ aproposition ] vi ) and viii ) , we can now estimate the first factor in and we obtain @xmath847 provided that we assume that @xmath627 has been chosen big enough ( in comparison with @xmath848 , @xmath427 , @xmath849 , @xmath850 ) . for the second factor in we yield @xmath851
provided that we assume that @xmath627 has been chosen big enough ( in comparison with @xmath848 , @xmath427 , @xmath849 , and @xmath852 , @xmath850 ) .
this implies that @xmath853 further , using and lemma [ aproposition ] iii ) we obtain @xmath854 and hence @xmath855 in addition , by lemma [ aproposition ] iii ) we have @xmath856 and @xmath857 assuming that @xmath627 has been chosen big enough . to estimate
, we consider now the two cases : a ) that @xmath858 and b ) that @xmath859 . in case
a ) we have that @xmath860 note that the the right hand side of this estimate no longer depends on @xmath391 and hence , @xmath861 in case b ) we have @xmath862 and hence @xmath863 we again need to distinguish two cases : if @xmath864 , then @xmath865 while in the case @xmath866 we have @xmath867 the estimations , , and prove that if the sequence @xmath508 increases fast enough then for every @xmath833 we have @xmath868 hence , applying the frostman lemma , we obtain @xmath869 and hence the assertion of theorem [ closing ] follows .
l. barreira , ya .
pesin , and j. schmeling , _ on a general concept of multifractality .
multifractal spectra for dimensions , entropies , and lyapunov exponents .
multifractal rigidity _ ,
chaos * 7 * ( 1997 ) , 2738 .
f. przytycki , j. rivera - letelier , and s. smirnov , _ equivalence and topological invariance of conditions for non - uniform hyperbolicity in the iteration of rational maps _ , invent .
* 151 * ( 2003 ) , 2963 . | we study the dimension spectrum of lyapunov exponents for rational maps on the riemann sphere .
this research was supported by the eu fp6 marie curie programmes spade2 and cody .
the research of f.p and m.r . were supported by the polish mnisw grant nn201 0222 33 ` chaos , fraktale i dynamika konforemna ' .
k.g .
was partially supported by the deutsche forschungsgemeinschaft and by the humboldt foundation .
k.g . and m.r .
thank for the hospitality of mpi pks dresden and i m pan warsaw where part of this research was done . |
partial or full ring - like structures , which apparently are projections of shells , are often found in the interstellar medium ( ism ) .
these dense shells could possibly be driven by one of the numerous sources like , ionising radiation from young star - clusters , early type massive stars , blast - waves from supernovae , or energetic stellar winds . detailed observations of these shells in various bands of the infrared wavelength have also revealed isolated sites of massive star formation ( e.g. deharveng , zavagno & caplan 2005 ) . in the recent past a number of such sites have been reported , for instance in the hii region rcw79 ( zavagno _ et al .
_ 2006 ) , and rcw120 ( anderson _ et al .
a catalogue of 600 such shells in the galactic disk was drawn up by churchwell _
( 2006 ) as part of the glimpse survey .
the survey showed that a large proportion ( @xmath290 % ) of these shells are thin , i.e. the shell thickness is less than a third of the outer shell radius , and driven primarily by massive stars ( @xmath3 shells ) , see also zavagno _ et al .
_ 2010 .
it is well - known that relatively high - mass stars emit powerful radiation that ionises gas in the local neighbourhood , and heats it to temperatures typically of the order of 10@xmath4 kelvin .
this hot plasma propagates in the interstellar medium ( ism ) at a highly supersonic speed ( typical velocity of expansion in the initial phases is @xmath5 km / s ) , whence it gradually equilibrates to a significantly lower temperature , of order a few thousand kelvin .
the expanding , roughly spherical volume of hot plasma , the so called stroemgren sphere , cools primarily via collisional excitation of heavier elements such as carbon , nitrogen , and oxygen , and sputtering of dust - grains while sweeping up a dense shell of gas in the ism ; the familiar snow - plough phase .
this shell is confined by two shocks , first , due to the wind driving it , and second , due to the reverse shock resulting from the propagation of the shell in the ism .
the stability of such shells has been discussed by numerous authors , e.g. elmegreen & elmegreen ( 1978 ) , larson ( 1985 ) , elmegreen ( 1989 ) , vishniac ( 1983 , 1994 ) , whitworth ( 1994 ) , wuensch & palou@xmath6 ( 2001 ) and anathpindika ( 2010 ) .
it has been demonstrated by these and several other authors that , a shock - confined shell is unstable to instabilities arising out of shock - induced turbulence within layers of the shell .
turbulence leads to enhanced transfer of momentum in different regions of the shell that makes it unstable to the so called thin shell instability ( tsi ) ( vishniac 1983 , 1994 ) , and raises the effective local sound - speed .
the stability of the shell depends on the critical interplay between the gravitational instability ( gi ) and the tsi .
a full fledged analytic treatment of the stability of a shock - confined slab is rather complex as has been demonstrated by vishniac ( 1994 ) , for instance , who showed that such slabs were likely to be unstable to the so called non - linear thin shell instability ( ntsi ) .
traditionally , stability analysis of shocked shells and/or slabs , for all practical purposes , have been simplified by excising shock dynamics in favour of a simple high - pressure approximation thereby eliminating perturbative effects of the tsi , and its non - linear excursion ( e.g. whitworth _ et al . _
1994b ) . anathpindika ( 2009 , 2010 ) numerically showed that a shocked slab and/or shell is unstable to the tsi , which , soon after its formation , develops wiggles on its surface and grows non - linearly . ehlerov@xmath7 & palou@xmath6 ( 2002 ) derived the critical density for a shell to become gravitationally unstable . under a sinusoidal approximation for these perturbations , anathpindika ( 2010 ) deduced an expression for the wavenumber of the fastest growing mode .
below , we propose to test the validity of this expression for hii shells driven by typical candidate stars listed in table 5.3 of spitzer ( 1978 ) , and deduce a mass function for clumps condensing out of these shells .
the case of a shell driven by a supernova blast wave will be considered separately .
we shall demonstrate that the mass function so derived is consistent with that reported by fukui _
et al._(1999 ) , yamaguchi _ et al . _
( 2001 ) , and roslowsky ( 2005 ) for massive clouds . in
2 we shall deduce our set of equations and demonstrative calculations , including the calculation of a mass spectrum for fragments , will be undertaken in 3 .
we conclude in 4 .
let us consider a typical source of ionising radiation that emits n@xmath8 number of photons per second . if @xmath9 , @xmath10 are the respective number of protons and @xmath11 per unit volume , then @xmath12 is the number of electrons captured per @xmath13 in the ground state , and the flux of photons flowing through a shell of radius , @xmath14 , is- @xmath15 where @xmath16 is the recombination coefficient that excludes electron captures to the ground state , and defined as @xmath17 at a temperature @xmath18 ; @xmath19 is the ionic charge and @xmath20 is the recombination coefficient function corresponding to @xmath16 ( see table 5.2 , spitzer 1978 ) .
the ionisation fraction of hydrogen , @xmath21 , has been set equal to unity so that the stroemgren radius is defined as - @xmath22 where the condition of approximate charge neutrality forces @xmath23 .
the temporal evolution of the ionised shell can be obtained by rewriting eqn .
( 2 ) as @xmath24,\ ] ] integration of which yields , @xmath25.\ ] ] maximising this equation gives the timescale , @xmath26 , over which the shell of ionising radiation makes a transition from the initial rarefied phase , to the dense phase , which is @xmath27 the radius of the expanding shell in the dense phase is given by @xmath28 ( e.g. shore 2007 ) .
for an approximately spherical shell , its average volume density , @xmath29 , is @xmath30 after a time @xmath31 , the shell cools down to an equilibrium temperature , @xmath32 , defined by eqn .
( 8) below . to estimate the temperature , t@xmath33 , we shall first account for the likely heating and cooling mechanisms .
a crucial contributor towards heating the shell is the photoionisation of h@xmath34 , the corresponding rate of heating is @xmath35 ( tielens 2005 ) , where @xmath36 @xmath37s@xmath38 , is the recombination cooling coefficient
. the cooling rate due to collisional excitation of carbon is @xmath39 ( tielens 2005 ) , where @xmath40 , is the abundance of carbon in the ism .
the equilibrium temperature of the gas , @xmath32 , within the shell can be estimated using the condition for thermal equilibrium , @xmath41 , and excess energy will be radiated away so that @xmath42 , i.e. @xmath43 which is roughly 10 k. the temperature , @xmath44 , of the ionised gas is @xmath45 , where @xmath46 k , is the average temperature of the preshock ism and @xmath47 is the mach number of the propagating shock , @xmath48 , @xmath49 .
the density of the shocked gas , @xmath50 , @xmath51 is the average mass of atomic hydrogen .
anathpindika ( 2010 ) , by performing a perturbative analysis , derived an expression for the fastest growing unstable mode in a shocked shell , and the wavenumber , @xmath52 , of this mode is @xmath53}.\ ] ] here @xmath54 and @xmath55 are respectively the outer , and the inner radii of the shell , and @xmath56 .
( 3 ) it follows that , @xmath57 for a shell of thickness , @xmath58 , @xmath59 , so that @xmath60 , and since @xmath61 ; @xmath62 . for the wavenumber , @xmath52 , calculated using eqn . ( 9 ) above , the corresponding wavelength , @xmath63 so that mass of a clump , @xmath64 , and the number of fragments @xmath65
. @xmath66 @xmath67^{2}\frac{1}{g^{2}\sigma_{s}},\ ] ] and since @xmath68 , @xmath69 the quantity @xmath70 , as will be shown in 3 below , so that @xmath71 calculations for typical o - b stars are deferred for 3 below .
we now discuss the stability of a shell driven by blast waves originating from a supernova that injects energy , @xmath72 , into the ism . for a shell having average density @xmath29 , and radius , @xmath73 , @xmath74 integrating eqn .
( 12 ) yields the well - known sedov solution for the shell radius , @xmath75 where @xmath76 is a numerical constant of order unity . ] the post - shock temperature , @xmath44 , can be calculated using the usual hugoniot shock conditions for an ism at a preshock temperature , @xmath46 k , as before .
the pressure due to the blast - wave , @xmath77 , under the assumption of approximate isothermality ; all symbols have their usual meaning
. the average density of the shell , @xmath78 . using eqn .
( 13 ) above , we can estimate the timescale , @xmath79 , over which the shell is likely to acquire its equilibrium temperature , @xmath32 , defined by eqn .
( 8) above , whence it may also slow down substantially , to a sonic or possibly , even sub - sonic speed , @xmath80 , is the local sound - speed . then at @xmath81 ,
@xmath82 @xmath83 the surface density of the shell , @xmath84 , is calculated as before and the mass of the shell , @xmath85 , @xmath86 @xmath87 , is the radius of the shell at that epoch ; this expression is also used to calculate the mass of the hii shell .
the fastest growing unstable mode , as before , is calculated as before , using eqn .
( 9 ) above .
the calculations for a typical supernova remnant ( snr ) are demonstrated in the following section .
_ the collect and collapse model : the hii shell _ + for illustrative purposes we consider a typical particle density , @xmath88 . the minimum timescale for radiative cooling of the shell , @xmath26 , calculated using eqn .
( 4 ) is approximately 0.1 myrs for a typical o5 star ; we have adopted physical parameters defined in table 5.3 of spitzer ( 1978 ) .
the shell , at this epoch , enters in to the dense phase whence it acquires substantial mass during the snow - plough phase , and undergoes substantial deceleration as can be seen from the characteristics plotted in figs .
1 and 2 . relatively cooler stars of intermediate mass drive weaker ionisation fronts and collect considerably lesser mass . the uppermost characteristic in fig .
1 shows that a typical o5 star may drive a shell that has a typical mass of a few times @xmath89 m@xmath1 .
an important parameter associated with the stability of an hii shell is its thickness . by a thin shell
we imply , @xmath90 .
it can in fact , be easily demonstrated that the ratio of the shell - thickness to radius , may rapidly asymptote to a value less than unity .
the average surface density of the shell , @xmath91 which following a little manipulation , leads us to @xmath92 for an hii shell , @xmath93 , and @xmath94 , then for a shell of constant surface density , @xmath95 . for a typical o5 star , this ratio is @xmath96 . the green curves in fig .
3 shows time evolution of the hii shell thickness - to - radius ratio for the set of stars used for demonstrative calculations above .
these plots suggest , the driven shell is generally thin irrespective of the spectral type of driving star .
this is consistent with that reported by the glimpse survey comprising 600 hii shells ( churchwell _ et al .
_ 2006 ) . both , theoretical ( e.g vishniac 1983 , 1994 ) , and numerical ( e.g. anathpindika 2009 , 2010 ) work has demonstrated the susceptibility of thin shells to various shearing instabilities , and particularly , to the thin shell instability ( tsi ) .
thin , shock confined shells tend to show a greater proclivity towards the tsi that apparently dominates the classical jeans instability .
the thermal jeans mass , @xmath97 , is @xmath98 which is roughly an order of magnitude smaller than the typical mass of a clump , @xmath99 , defined by eqn .
( 11 ) above which suggests , perturbations triggered by dynamical instabilities in driven shells and/or shocked slabs facilitate concentration of material in perturbed regions , the local maximas or minimas .
this was explicitly demonstrated by wuensch & palou@xmath6 ( 2001 ) .
the clump on gaining sufficient mass condenses out , which is the essence of the collect and collapse model .
+ _ a supernova shell _ for a demonstrative calculation ,
let us consider a typical supernova that releases energy , @xmath100 ergs , into the ism that as in the previous case , is assumed to have @xmath101 @xmath13 .
the timescale , @xmath79 , defined by eqn .
( 14 ) above , for a shell of average density @xmath102 g @xmath13 is @xmath103 myrs , and the velocity of the shell , @xmath104 , @xmath105 which is @xmath106km / s . the corresponding radius of the shell is then , + @xmath107 kpc . the mass of the shell swept up immediately follows from eqn .
( 15 ) , and @xmath108 m@xmath1 .
similarly , the mass of a typical fragment calculated using eqn . ( 11 ) is , @xmath109 m@xmath1 .
it might be interesting to calculate the efficiency , @xmath110 with which the initial energy , @xmath72 , is converted into mechanical energy whence matter in the ism is swept up .
this efficiency is defined as , @xmath111 which suggests that a large proportion of the initially injected energy is lost in heating the ism . the attenuated sound speed , @xmath112 , defined by eqn . ( 19 ) below accounts for the effects of turbulence generated by various hydrodynamic instabilities
which were not directly included in the perturbative analysis that led to eqn .
( 9 ) , and @xmath113}.\ ] ] the glimpse survey of supernovae shells produced a catalogue of 95 snrs ( reach _ et al .
_ 2006 ) , none of which have shown unambiguous evidence in favour of the collect and collapse model , although a sample of snrs have shown association with oh masers .
however , the source of these masers is unclear and could perhaps be due to sporadic star - formation triggered in molecular clouds over - run by snrs .
despite this being the case , one may assert that possible condensation in snrs may take a while , as suggested by the magnitude of @xmath79 , which is roughly two orders of magnitude larger than the timescale over which the hii shell produced fragments .
the resulting fragments are therefore likely to be at least as massive as those forming in the former case .
the number of clumps , @xmath114 , condensing out of a shell is @xmath115 .
then @xmath116 or equivalently , in terms of the wavenumber , @xmath52 , @xmath117 this equation defines the number of clumps in an interval @xmath118 , the integration of which , over the wave number domain yields the mass spectrum for putative clumps .
figure 4 shows a typical mass spectrum for clumps condensing out of hii shells , obtained via a monte - carlo integration of eqn .
the integration was performed for 10,000 realisations of shell fragmentation , each producing @xmath114 number of clumps .
thus , we have @xmath2180,000 clumps for the o5 star and fewer , @xmath119 , for an o9 star .
the relatively smaller shell driven by the latter star not only produces fewer fragments but also , those that are comparatively less massive than those in the former case .
however , we note that extremely large clumps , @xmath120 , may only be sparingly produced as is evident from the mass spectrum in fig .
the mass spectrum for stars of either spectral type is similar and a power - law , @xmath121 , appears to fit the derived spectrum reasonably well in either cases ; @xmath122 , and 1.5 respectively .
the spectrum derived here is consistent with the one obtained for large clouds by fukui _
( 1999 ) , and roslowsky ( 2005 ) .
similarly , we also obtain the mass spectrum for fragments condensing out of the snr driven shells .
the spectrum for this case is plotted in fig . 5 , and a power - law similar to the one in earlier two cases , fits the spectrum reasonably well . in fact , this latter spectrum is similar , @xmath123 , to that for the fragments resulting from the hii shell driven by an o5 star ; spectrum for the snr was derived for @xmath124 realisations of shell fragmentation , and shows considerable shift towards a higher mass in comparison to the distributions shown in fig .
this is probably because the velocity of an hii shell in its dense phase decays more rapidly than that of an snr , which pushes up the attenuated sound - speed , @xmath112 , for the latter that in turn raises the mass , @xmath99 , as seen in fig .
star forming clumps as massive as @xmath22500 m@xmath1 have been reported in the hii shell n49 and rcw34 ( e.g. zavagno _ et al .
_ 2010 , bik _ et al .
a similar treatment of the problem was presented by whitworth _
( 1994a , b ) who arrived at a minimum clump mass of only a few tens of m@xmath1 , which in the light of present findings , appears somewhat conservative . our claim more massive clumps
is supported by recent observations described above .
it is well known , the evolution of a cloud is governed by the dynamical effects associated with the complex interplay between self - gravity , and other contributing factors such as turbulence and the magnetic field that support a clump against the former .
it is therefore crucial to predict clump masses with reasonably good accuracy ; under - estimation of the mass of putative clumps , for a given radius , which in the present case is roughly equal to the wavelength of the unstable mode , will lower the average density and thus raise the clump - lifetime .
an increase in the longevity of clumps will also possibly make events such as clump - clump collisions more probable , than they are known to be .
inter - clump collisions , according to simulations , could lead to bursts of star - formation with fewer low - mass stars ( e.g. chapman _ et al .
_ 1992 , anathpindika 2009 ) , that will tend to shift the stellar initial mass function towards a top - heavy distribution , away from the widely reported lognormal form .
we have examined the stability of thin , dense shells driven by powerful ionising radiation originating from massive , early type stars and/or blast waves from a supernova .
our work here shows that fragmentation of these shells is likely to produce large clumps , with masses typically @xmath0 m@xmath1 .
this calculated range of fragment masses is consistent with that reported via observations of hii shells , and the fragments so formed could spawn a second generation of stars , which may in turn trigger the next generation of stars in the surrounding ism .
the simple calculations discussed above appear to suggest that the ratio of thickness - to - radius for a shell evolves only weakly with time , and that calculated here ( see fig . 3 ; @xmath125 ) , is consistent with the values reported in the glimpse survey for hii shells .
similar results are obtained for an snr , which , however , appears to be much thinner compared to the shells driven by ob stars .
this could perhaps be the reason why snrs often appear filamentary in the irac bands ( e.g. reach _ et al .
simulations studying the evolution of thin shells such those by anathpindika ( 2010 ) have demonstrated the dominance of tsi in shocked shells ; the surface of the shell was also shown to develop ripples , similar to breathing modes on fluid surfaces ( fig .
3 in anathpindika ( 2010 ) ) . while wiggles on the shell surface associated with the tsi though coplanar , are generally orthogonal to its surface , and so in case of a shell in the plane of the sky , can not alone account for the reported filamentary nature
however , breathing modes coupled with local magnetic field could perchance explain the occurrence of filaments reported by reach _
( 2006 ) . _
* acknowledgements * _ the author is supported by a post doctoral fellowship at the indian institute of astrophysics , bangalore , india , and wishes to thank an anonymous referee for a critical review of the original manuscript .
useful suggestions from prof .
harish bhatt are gratefully acknowledged .
anathpindika , s. , 2009 , a&a , 504 , 437 anathpindika , s. , 2010 , mnras , 405 , 1431 anderson , l. , zavagno , a. , rod ' on , j. , _ et al . _ , 2010 , a&a , 518 , l99 bik , a. , puga , e. , waters , l. , _ et al . _ , 2010 , apj , 713 , 883 chapman , s. , pongracic , h. , disney , m. , nelson , a. , turner , j & whitworth , a. , 1992 , nature , 359 , 207 churchwell , e. , povich , m. , allen , d. , _ et al .
_ , 2006 , apj , 649 , 759 deharveng , l . , zavagno , a . , caplan , j. , 2005 , a&a , 433 , 565 ehlerov@xmath7 , s & palou@xmath6 , j. , 2002 , mnras , 330 , 1026 elmegreen , b & elmegreen , d. , 1978 , apj , 220 , 1051 elmegreen , b. , 1989 , apj , 344 , 306 fukui , y. , onishi , t. , abe , r. , kawamura , a. , tachihara , k. , yamaguchi , r. , mizuno , a & ogawa , h. , 1999 , pasj , 51 , 751 larson , r. , 1985 , mnras , 214 , 379 reach , w. , rho , j. _ et al . _ , 2006 , apj , 131 , 1479 roslowsky , e. , 2005 , pasj , 117 , 1403 spitzer , l. , 1978 , in _ physical processes in interstellar medium _
, wiley - intersceince . pub .
shore , s. , _ astrophysical hydrodynamics _
, wiley - vch , weinheim , 2007 ; p. 141
- 43 tielens , a. , g. , 2005 , _ the physics and chemistry of the ism _ , cambridge university press , uk , p. 54 - 6
vishniac , e. , 1983 , apj , 274 , 152 vishniac , e. , 1994 , apj , 428 , 186 whitworth , a. , bhattal , a. , chapman , s. , disney , m & turner , j. , 1994a , mnras , 268 , 291 whitworth , a. , bhattal , a. , chapman , s. , disney , m & turner , j. , 1994b , a&a , 290 , 421 wuench , r & palou@xmath6 , j. , 2001 , a&a , 374 , 746 yamaguchi , r. , norikazu , m. , onishi , t. , mizuno , a & fukui , y. , 2001 , apj , 553 , l185 zavagno , a. , deharveng , l. , comer ' on , f. , brand , j. , massy , f. , caplan , j & russeil , d. , 2006 , a&a , 171 , 184 zavagno , a. , anderson , l. , russeil , d. , _ et al . _ , 2010 , a&a , 518 , l101 | early type massive stars drive thin , dense shells whose edges often show evidence of star - formation . the possibility of fragmentation of these shells , leading to the formation of putative star - forming clumps
is examined with the aid of semi - analytic arguments .
we also derive a mass - spectrum for clumps condensing out of these shells by performing monte - carlo simulations of the problem . by extending on results from our previous work on the stability of thin , dense shells
, we argue that clump - mass estimated by other authors in the past , under a set of simplifying assumptions , are several orders of magnitude smaller than those calculated here . using the expression for the fastest growing unstable mode in a shock - confined shell ,
we show that fragmentation of a typical shell can produce clumps with a typical mass @xmath0 m@xmath1 .
it is likely that such clumps could spawn a second generation of massive and/or intermediate - mass stars which could in turn , trigger the next cycle of star - formation .
we suggest that the ratio of shell thickness - to - radius evolves only weakly with time .
calculations have been performed for stars of seven spectral types , ranging from b1 to o5 .
we separately consider the stability of supernova remnants .
star formation , hiiregions , supernova remnants , instabilities |
it is important to understand theoretically the mechanisms by which quark - gluon plasmas can locally equilibrate in heavy ion collisions such as those at rhic . since the path - breaking work of baier , mueller , schiff and son @xcite on `` bottom - up '' thermalization
, one modest theoretical goal has been to understand the process of equilibration in the simplifying theoretical limit of arbitrarily high - energy collisions , where the running strong coupling @xmath4 can be treated as small .
typically , theorists expect to be able to solve weakly - coupled problems , but equilibration of weakly - coupled non - abelian plasmas has proven to be a very rich and challenging problem . in particular , the original attempt by baier did not account for the physics of plasma instabilities , which are now believed to play a crucial role during some of the early stages of thermalization @xcite .
early in a heavy - ion collision , particles have an anisotropic distribution of momenta , as measured in local frames moving with the expanding plasma @xcite .
later , they scatter and equilibrate to locally isotropic distributions , giving rise to hydrodynamic behavior . generically , anisotropic distributions of particles produce collective plasma instabilities known as weibel or filamentary instabilities .
these instabilities are associated with filamentation of particle currents and initially exponential growth of long - wavelength magnetic fields . since these magnetic fields in turn deflect the particles in the plasma , and so perhaps drive the particles towards isotropization and equilibration , it has been important to understand just how large these unstable magnetic fields grow @xcite . in this paper , we will follow the nomenclature of baier and refer to the initial particles of the plasma ( _ e.g. _ small @xmath5 gluons in the saturation picture ) as `` hard particles , '' and we will refer to their momentum scale ( _ e.g. _ the saturation scale ) as @xmath6
. we will refer to the momentum scale of magnetic plasma instabilities as @xmath7 .
once the system has expanded enough that the hard partons have perturbative density ( @xmath8 ) , one generically finds that @xmath9 @xcite .
we will assume and exploit this hierarchy of scales in what follows .
we will refer to any scale that is parametrically small compared to @xmath6 as `` soft . ''
[ fig : benergy ] shows a cartoon of the results of numerical studies in 3 + 1 dimensions of non - abelian plasma instability growth @xcite in a non - expanding system .
the energy in soft magnetic fields is plotted versus time for simulations initiated with a very tiny magnetic seed field .
the soft magnetic field energy initially grows exponentially , but the behavior changes once the fields become large enough that their non - abelian self - interactions become non - perturbative ( @xmath10 ) .
the energy growth then becomes linear with time . for a non - expanding system
, this linear growth would continue until some parametrically long time later ( not addressed by the simulations ) when the soft fields eventually have a non - perturbatively large effect on the original , anisotropic hard particles that created them . a qualitative depiction of soft magnetic field energy vs. time for the growth of non - abelian magnetic instabilities .
] recent numerical work @xcite indicates that the linear growth in energy is associated with a cascade of energy from the soft , perturbatively unstable modes toward the ultraviolet . a cartoon of the development of the spectrum is shown in fig .
[ fig : spec ] .
( a ) the system starts with the initial , anisotropic hard particles of momentum @xmath11 plus some small initial fluctuations in softer modes .
( b ) plasma instabilities initially lead to rapid , exponential growth of unstable soft modes of momenta @xmath12 .
( c ) the unstable modes stop growing in amplitude once they become non - perturbatively large , but total soft energy continues to grow ( linearly ) as interactions feed power from @xmath7 into increasingly higher momentum modes .
the spectrum of this cascade falls with some characteristic power @xmath13 , and the upper momentum limit @xmath14 of the cascade increases with time .
the goal of the present paper is to theoretically determine the exponent @xmath15 by considering the processes responsible for this cascade . a cartoon description of the energy spectrum of excitations at various times during the development of non - abelian plasma instabilities . ]
the basic plan of our analysis will be somewhat similar to investigations of thermalization during reheating after inflation in the early universe @xcite .
the details of our problem , and the result for the spectral index @xmath15 , are different , however .
let @xmath1 be the average occupancy of each mode @xmath2 . in analyzing the cascade , we shall take @xmath16 for @xmath17 . in this range of @xmath18 ,
the physics is perturbative ( @xmath19 ) and can be equally well described by a kinetic theory of particles or by a classical theory of fields ( @xmath20 ) @xcite .
are necessary but not sufficient for a kinetic theory description of relativistic systems . for quantitatively accurate ( or qualitative order - of - magnitude )
applications of kinetic theory , one must check that debroglie wavelengths and the duration of scattering events are @xmath21 ( or @xmath22 ) mean free paths , for whatever processes dominate the kinetic description . ]
the perturbative nature of the cascade for @xmath23 follows from ( i ) the non - perturbative occupation number @xmath24 at the instability scale @xmath7 and ( ii ) the decrease of @xmath18 as some power @xmath13 to be determined .
this gives @xmath25 where the superscript `` @xmath26 '' stands for `` steady - state . ''
perturbative cascades are known in the plasma physics literature as `` weak plasma turbulence '' @xcite .
we will restrict attention to the case where @xmath27 in order that ( [ eq : fk ] ) satisfy the other condition @xmath20 .
this classical field theory limit is implicit to the numerical simulations which discovered the cascade @xcite , since these simulations treated all soft fields as classical . in the next section ,
we warm up by discussing @xmath28 scattering of particles at the uv end of the cascade .
we show that this process alone would reproduce a result for the spectral index @xmath15 in ( [ eq : fk ] ) known from analogous scalar theory applications ( @xmath29 ) . in sec .
[ sec : dominant ] , we show that there is a more important process : multiple scattering from the non - perturbative infrared background produces a steeper distribution , @xmath30 .
this result is consistent with numerical simulations @xcite .
[ sec : consistent ] discusses the back reaction the cascade has on the infrared physics of the instability and poses some open questions concerning logarithmic effects in our analysis . in sec .
[ sec : bottom - up ] , we return to the question of how instabilities affect the bottom - up scenario and discuss what remains to be understood about instability development before this question can be answered .
finally , we offer our conclusions .
non - perturbative numerical studies of instability development in 3 + 1 dimensions have only studied hard particle distributions which are moderately , but not extremely , anisotropic @xcite . in the case of moderate , @xmath31 anisotropy
, there is a single scale which characterizes @xmath32 , given parametrically by @xmath33 where the integral is over the distribution of _ hard _ particle momenta @xmath34 . for _
extremely _ anisotropic distributions , the situation is more complicated . in most of this paper
, we will implicitly assume that there is only a single relevant soft scale .
we will only discuss the unresolved issues surrounding extremely anisotropic distributions in our discussion of bottom - up thermalization in sec .
[ sec : bottom - up ] .
the standard picture of driven turbulence is that something pumps energy into the cascade from the infrared ( ir ) end at a constant rate . in the absence of dissipation , this leads to linear growth in the total energy . at the same time , energy is flowing out the ultra - violet ( uv ) end by increasing @xmath36 .
the spectral index @xmath15 of the cascade can be determined by balancing ( i ) the rate at which energy flows out the uv end of the cascade by scattering processes that extend the uv cascade by increasing @xmath36 ( _ e.g. _ scattering momentum @xmath36 particles to produce momentum @xmath37 particles ) and ( ii ) the constant rate energy enters the ir end of the cascade . as a review and
warm - up example , let us for the moment ( incorrectly ) assume that the dominant process for increasing the cut - off was large - angle @xmath35 scattering with all momenta roughly of order @xmath36 .
[ fig:4momenta ] shows an example of how two particles with momenta @xmath38 and @xmath39 less than @xmath36 can produce a particle with momentum greater than @xmath36 in @xmath35 scattering .
now consider what the time scale would be for bringing @xmath40 up to its steady - state value ( [ eq : fk ] ) .
for a statistically homogeneous system , the boltzmann equation for @xmath35 scattering has the form @xmath41\ ] ] for bosons such as the soft gluons that make up our cascade . in the @xmath42 classical field limit of ( [ eq : fbounds ] ) , this becomes @xmath43 .
\label { eq : boltz22}\ ] ] even if @xmath44 were initially small for @xmath45 as in fig .
[ fig:4momenta ] , there is a term in this equation with @xmath46 .
take the square - amplitude @xmath47 for large - angle scattering as order @xmath48 times normalizations and the energy - momentum conserving delta function .
the characteristic rate @xmath49 implied by ( [ eq : boltz22 ] ) for @xmath50 to grow to its steady state value @xmath51 is then given by @xmath52 where the factor of @xmath36 on the right - hand side follows from dimensional analysis .
so @xmath53 from our assumed power - law form ( [ eq : fk ] ) for @xmath54 , this gives @xmath55 since the cutoff @xmath36 has to change on the time scale of the age @xmath56 of the system , we must have @xmath57 . on the other hand ,
let us estimate the rate from the linear growth of the total energy density in the cascade . using ( [ eq : fk ] ) , the total energy density is @xmath58 ( provided @xmath59 , which we will find ) .
taking @xmath60 , the corresponding rate is @xmath61 comparing the exponent here to ( [ eq:22exp1 ] ) , we would obtain @xmath62 , in agreement with previous analysis in the literature of certain cascades in field theory where large - angle @xmath35 scattering is the dominant process @xcite .
agrees with eq .
( 66 ) of ref .
@xcite . ]
momenta in a @xmath63 process that creates a particle with momentum larger than @xmath36 . ]
the more efficiently the system can scatter energy into the uv , the larger will be the exponent @xmath15 .
the goal is to identify the most efficient process .
we propose that the dominant process is multiple scattering of quanta of momentum @xmath36 off of the non - perturbative background field of wave - number @xmath7 . in this section
, we analyze this process and show that it is more important than the large - angle @xmath35 scattering considered in the last section .
consider a relatively high - energy particle ( @xmath64 ) propagating through a soft , random , non - perturbatively large background field @xmath65 having wave - numbers of order @xmath7 .
the particle can scatter from the background as in fig .
[ fig:1to1 ] .
non - perturbatively large means @xmath66 , and so the only parameters in the problem are @xmath67 and @xmath7 .
the probability of interaction is independent of @xmath67 in the @xmath23 ( eikonal ) limit . by dimensional analysis ,
the rate is therefore @xmath68 to change momentum by @xmath68 .
each such scattering will have a perturbative effect on the particle , randomly changing its momentum @xmath67 by @xmath69 .
@xmath70 successive collisions will change @xmath67 by order @xmath71 in a time of order @xmath72 .
taking @xmath73 , we see that the time taken to change particle momenta from @xmath36 to @xmath37 at the uv end of the cascade would be @xmath74 corresponding to a net rate that depends on @xmath36 as @xmath75 if we equate this to the rate ( [ eq : ir ] ) determined by linear energy growth , we obtain the spectral index @xmath76 the fact that this @xmath15 is larger than the @xmath29 of sec . [
sec : warmup ] indicates that the large - angle @xmath35 processes of fig .
[ fig:4momenta ] are unimportant compared to multiple scattering by fig .
[ fig:1to1 ] . and @xmath77 : eq .
( [ eq : gamma22 ] ) gives @xmath78 for large - angle @xmath35 , and eq .
( [ eq : t ] ) gives in both cases a parametrically larger rate @xmath79 for multiple scattering off of the non - perturbative background . ] a particle with momentum @xmath23 scattering off of the non - perturbative field modes with wavenumber @xmath7 . ]
we note in passing that interactions with a non - perturbative background field have also been important in the study of cascades during reheating of scalar field theories in the background of an oscillating inflation field @xcite .
however , processes analogous to fig .
[ fig:1to1 ] were not relevant in that case because the background field carried zero momentum and so could not directly transfer momentum to the particles . in that case , the dominant process was fig .
[ fig:2to1]a , which is not as efficient , giving a smaller value of @xmath15 in those applications . for our application
, we consider in the appendix the related process of fig .
[ fig:2to1]b : @xmath80 inverse bremsstrahlung of particles with momenta @xmath23 , catalyzed by the non - perturbative background field .
we find that , due to the landau - pomeranchuk - migdal ( lpm ) effect , this process is not competitive with multiple scattering via fig .
[ fig:1to1 ] .
@xmath80 processes catalyzed by a background field in ( a ) scalar @xmath81 theory and ( b ) gauge theory ( inverse bremsstrahlung ) . ] failing to find a more efficient process than multiple scattering off of the non - perturbative background , our provisional conclusion is that the spectral index @xmath15 should be 2 .
this is consistent with results from simulations @xcite , which extract @xmath82 from coulomb - gauge spectra of the cascade . from ( [ eq : ir ] ) , we also get that the uv end of the cascade should grow as @xmath83 using ( [ eq : fk ] ) , this means that our classical treatment ( @xmath84 for @xmath85 ) is valid for times @xmath86 .
having proposed a cascade with @xmath87 we should now ask whether the development of the cascade might have a back - reaction on the dynamics of the infra - red ( ir ) modes with wave - numbers of order @xmath7 .
these ir modes are the starting point for energy to enter the cascade , and
one could imagine that a back - reaction might conceivably change the rate at which energy is taken into the cascade from the original anisotropic hard particles . if such a change to the rate depended on @xmath14 and so on time , it would be harder to understand why the energy in the cascade grows linearly with time , as indicated by numerical simulations .
let us investigate how the cascade affects the soft scale @xmath7 itself , given by ( [ eq : msoft ] ) .
this scale is initially created by the hard particle distribution , as we have indicated explicitly in ( [ eq : msoft ] ) . however , once the cascade develops , its excitations could also contribute : @xmath88 where @xmath2 is integrated over momenta in the cascade .
how big is this contribution , and is its dependence on time through @xmath14 significant ? for @xmath89 , the integral ( [ eq : dm ] ) would be sensitive to large @xmath14 ; for @xmath90 , it would not .
the spectral index @xmath30 found in this paper is the borderline case , and ( [ eq : fk2 ] ) produces @xmath91 if we ignore the logarithm , then @xmath92 , and the picture can be self - consistent .
the creation of the cascade would make an @xmath31 relative change to the ir physics driving the instability , and so could make an @xmath31 relative change to the rate of linear energy growth . that still allows the system to settle down into steady growth as @xmath93 , consistent with numerical results .
the formal logarithmic dependence of ( [ eq : dm2 ] ) on @xmath14 , however , suggests that the energy growth rate should see logarithmic in time corrections .
current simulations @xcite do not support such logarithmic corrections , perhaps because the coefficient happens to be small , and perhaps because of the limited statistical power of existing simulations .
in addition , the above analysis was rather crude : the details of exactly how @xmath94 affects instability development depend , for instance , on the degree of anisotropy of the cascade particles as a function of time . to an anisotropic hard particle distribution
@xmath95 will not change the momentum scale of the instability .
it will , however , decrease the perturbative growth rate of such instabilities .
( see , for example , the discussion in ref .
@xcite . ) roughly speaking , this is because changing magnetic fields require electric fields , the ability of a medium to support electric fields depends on its conductivity , and the conductivity is affected by the number of isotropic as well as anisotropic particles . on the other hand , if @xmath96 is significantly anisotropic , it could increase the momentum scale of the instability , which would tend to increase growth rates .
numerical simulations suggest that the excitations in the cascade are somewhat but not perfectly isotropic . a simulation in ref .
@xcite , for a particular hard particle distribution , found that the magnetic fields making up the excitations in the cascade had @xmath97 equal to roughly 80% of @xmath98 and @xmath99 . ] thoroughly understanding the physics of the cascade at the level of logarithms remains an open problem .
we see that cascade particles can have an @xmath31 effect on the infrared dynamics ( up to logarithms ) . in previous work @xcite
, we argued that particles can make two important contributions to the behavior of the plasma .
one is the effect on infrared fields ( `` screening , '' in the parlance of ref .
the other is the contribution to the scattering rate for hard particles .
that is , when a hard particle ( or another cascade particle ) scatters , what fraction of the time is it scattering from a cascade particle ?
this is determined by the relative contributions to the integral @xmath100 where @xmath101 is a final state bose stimulation factor and where we will focus on @xmath67 s in the cascade . for our @xmath102 cascade ( with @xmath42 ) , the integral is of form @xmath103 which is dominated by the smallest momenta and insensitive to the moving cutoff @xmath14 .
that is , the cascade particles are not important as scatterers .
most scattering is off the @xmath7 nonperturbative background , not off the higher - momentum fluctuations .
this is another way of seeing why it is scatterings from the soft background , rather than between cascade particles , which determine the progress of the cascade .
we introduced this paper with a discussion of how nonabelian plasma instabilities play a significant role in the bottom - up scenario for quark - gluon plasma equilibration in the high energy limit .
we now have a reasonable picture of the development of non - abelian plasma instabilities in a non - expanding system .
once we understand , at least parametrically , the basic processes and their rates , we should be able to apply this knowledge to an analysis of expanding plasmas .
unfortunately , there is an obstacle . for reasons of computational practicality
, simulations have been performed for moderately , but not _ extremely _ , anisotropic hard particle distributions . for moderate anisotropy
, there is a single parametric scale @xmath7 for ir physics .
non - perturbative ir magnetic fields are therefore @xmath104 , the linear growth of soft energy density is @xmath105 , and so forth .
however , as we shall review , the first stages of the bottom - up scenario involve _ extremely _ anisotropic hard particle distributions , where ( in local fluid frames ) the angle @xmath106 between typical particle velocities and the transverse plane is parametrically small .
one must therefore understand the parametric dependence of instability development on this additional small parameter , @xmath106 .
we leave such understanding for future work . in this section ,
we simply wish to emphasize the need by showing how the current state of uncertainty can have a significant impact on the first stage of bottom - up thermalization . in particular , we will consider a possible alternative to the recent analysis of the effects of instabilities by bdeker @xcite . in their original analysis of the stages of thermalization for arbitrarily high energy heavy - ion collisions ,
baier @xcite considered an initial , non - perturbatively dense state of saturated low-@xmath5 gluons produced in the collision . in their analysis ,
these were labeled the `` hard '' partons in the plasma , with momenta @xmath6 of order the saturation scale @xmath107 .
their analysis is restricted to following the process of thermalization for @xmath108 , where @xmath109 is proper time since the collision .
they assume initial conditions at time @xmath110 with number density @xmath111 of such hard particles .
this number density dilutes as @xmath112 due to the one dimensional expansion of the system .
the scale ( [ eq : msoft ] ) characterizing infrared physics ( which they interpreted as a debye screening scale ) then evolves with time as @xmath113 our notation `` @xmath114 '' for this scale is in preparation for our later discussion of plasma instabilities , for notational consistency with previous work . ''
is that , in addition to its role in the physics of instabilities , @xmath114 turns out to be the mass scale of transverse plasmons in the large momentum limit ( _ i.e. _ @xmath115 ) @xcite . ] in the first phase of their analysis , the gluons dilute to perturbative densities as the system expands , and their _ local _ distributions of velocity temporarily become highly anisotropic due to momentum selection : over time , approximately free - streaming particles with the same @xmath116 would move to approximately the same displacement @xmath117 , along the beam axis , from the origin of the collision .
this drive towards local anisotropy is moderated only by processes which scatter the hard particles and so try to isotropize them .
baier considered random , individual , small - angle @xmath35 scattering of the hard particles and found that , early in their scenario , the balance between expansion and scattering effects produces local distributions with anisotropy @xmath118 in local fluid rest frames .
this result indicates extreme anisotropy for @xmath119 .
later , however , once the system dilutes enough that @xmath120 , baier found that other processes come into play which eventually locally isotropize and thermalize the system . here
, we modestly focus our attention on the first phase , described by ( [ eq : pzbmss ] ) in the original scenario . in the original scenario of ref .
@xcite . ] arnold , lenaghan , and moore @xcite pointed out that the first phase will be modified by plasma instabilities .
recently , bdeker @xcite has attempted to deduce how scattering of hard particles from soft fields created by plasma instabilities will modify ( [ eq : pzbmss ] ) .
bdeker assumed that soft fields consist only of a non - perturbative component ( @xmath121 ) with momenta @xmath122 given by ( [ eq : msoft ] ) .
for hard particle distributions with @xmath31 anisotropy , this is parametrically the only characteristic scale of unstable modes . with these assumptions
, bdeker found @xmath123 which is parametrically larger than ( [ eq : pzbmss ] ) .
this estimate is in keeping with the discussion of this paper ; the dominant way in which high momentum particles ( whether in the cascade , as we have discussed , or even higher momentum , as is relevant here ) undergo scattering , is by interaction with the nonperturbative fields at the scale @xmath7 .
however , this estimate is not secure , because the treatment of the unstable scale @xmath7 has been made assuming @xmath31 anisotropy , yet it predicts a parametrically large anisotropy for times @xmath108 .
therefore we must reconsider the behavior of the instability for highly anisotropic systems , a point also emphasized by bdeker @xcite .
as we shall review , there is another characteristic scale @xmath124 for such extremely anisotropic distributions .
we believe that bdeker s estimate of taking @xmath7 of order the lower scale ( @xmath114 ) in ( [ eq : pzb ] ) probably provides a lower bound on the result for @xmath125 . we will not attempt to make a complete treatment here , but will provide what we believe is a sensible parametric upper bound . for highly anisotropic , oblate distributions of hard particles ,
the spread of angles of hard particle momenta @xmath34 about the transverse plane is characterized by @xmath126 the perturbatively unstable modes are described by two scales .
one is @xmath114 , defined before , whose inverse represents the time scale that a gauge field must act on the particles before the back - reaction of particles on the field becomes important .
@xmath114 characterizes the growth rate of instabilities .
the other scale is the largest wave vector @xmath127 of any unstable soft gauge field mode , which is @xmath128 for a detailed explanation of how these scales arise , see ref .
@xcite ; here we give a physical argument . for a gauge field mode to be unstable
, the hard particles must fly for a time scale @xmath129 in the presence of that mode , and be deflected in the same direction that whole time .
that means that the magnetic field must be roughly the same all along the trajectory of the particle s path .
therefore , the gauge field must take roughly the same value over distances @xmath130 in the two transverse directions , and over a distance @xmath131 in the longitudinal direction ( since this is how far a typical particle moves in the longitudinal direction in that time scale ) .
that means that modes with @xmath132 and @xmath133 are unstable .
this agrees with a more detailed analysis @xcite .
how large can the soft fields grow ?
we can place an upper bound as follows .
gauge fields in different color directions are all growing simultaneously .
when these grow large enough that they can randomize the color of a hard excitation in a time scale @xmath134 , then the hard modes will no longer be contributing to the time development of the unstable modes on the time scale required to make the instability work .
the rotation of color charge as a hard particle moves along a trajectory is a matrix given by the wilson line @xmath135 , \label { eq : rotate}\ ] ] where @xmath136 represents path ordering . in the analysis of a particular unstable mode , think of all the other unstable modes as background .
the color rotation ( [ eq : rotate ] ) due to a non - commuting background field will therefore be significant , and so stop instability growth , if @xmath137 . to avoid color randomization of generic hard particles in our extremely anisotropic distribution
, we therefore need , for example , , one should assume that one is in a gauge where @xmath138 is relatively smooth . here , we imagine that we study the state of the system at a particular time @xmath56 by ( i ) working in @xmath139 gauge , and then ( ii ) making a spatial gauge transformation to make @xmath140 as smooth as possible at that particular time.]@xmath141 ) is to consider the propagator @xmath142 of particle fluctuations in non - abelian vlasov equations . in perturbative formulas for instabilities ( _ e.g.
_ refs . ) , such propagators appear in the perturbative form @xmath143 .
one can then ask when the @xmath144 in @xmath145 can be treated as a perturbation to @xmath146 and @xmath147 , both of which are order @xmath114 . ]
@xmath148 since the dominant instabilities have @xmath149 nearly in the @xmath117 direction , with @xmath150 , this implies that an _ upper _ bound on how large the corresponding magnetic field strength can grow is .
similar reasoning to ( [ eq : aperp ] ) could allow @xmath151 , which gives an @xmath152 contribution to @xmath153 that is the same size as ( [ eq : b ] ) . ]
@xmath154 coherent on transverse length scales of @xmath155 .
such fields will give momentum kicks to the hard modes due to the @xmath156 term in the lorentz force law .
the force is coherent over the time scale @xmath157 ( since this is both how often @xmath153 changes as one moves in the transverse direction , and how often the particle s charge gets rotated ) .
the individual momentum kicks are of size @xmath158 , and they lead to momentum diffusion in the @xmath159 direction of size @xmath160 therefore , @xmath161 [ compare to ( [ eq : pzb ] ) . ] substituting into the defining relation ( [ eq : deltatheta ] ) for @xmath162 , and using the relation ( [ eq : m_qs ] ) for the time development of @xmath114 , gives @xmath163 the above estimate is correct if our picture of how plasma instabilities are cut off in a highly anisotropic setting proves correct . while we believe that our estimates represent an upper bound on such a cutoff , we are not confident that they are correct .
the moral of this story is that better theoretical understanding is needed of the non - perturbative dynamics of instabilities in the case of extremely anisotropic distributions . until then
, it will be difficult to analyze even the first phase of bottom - up thermalization with complete certainty .
we have argued that , in the presence of non - abelian instabilities that have grown non - perturbatively large , the spectrum of the cascade of soft field energy from unstable modes into the ultra - violet should have the form @xmath164 with @xmath30 .
this is consistent with results from numerical simulations .
we should emphasize that this turbulent cascade is only a transitory phenomenon appearing during the local thermalization of the quark - gluon plasma in the high energy limit .
it is not directly related to the spectrum of particles that much later leave the collision after hadronization .
also , though the word `` turbulence '' is used , the existence of this cascade does not mean that the plasma is described by ideal hydrodynamics at the time the cascade is formed . at the early times considered in this paper ,
most of the energy that will become the quark - gluon plasma is still in hard particles , @xmath165 , which have not yet thermalized and which are locally extremely anisotropic
. it would be delightful to be able to apply the current understanding of plasma instability development to produce a complete picture of thermalization in the high energy limit , finally completing the program initiated by the original bottom - up scenario .
a lot of progress has been made over the last year in understanding the qualitative behavior and parametric dependence of the non - perturbative physics of instabilities for the case of moderately anisotropic hard particles .
unfortunately , figuring out the detailed role of instabilities in bottom - up thermalization will likely require at least a similar level of understanding of the case of _ extremely _ anisotropic hard particles , which have not yet been investigated non - perturbatively .
we would like to thank larry yaffe and berndt mueller for useful conversations .
this work was supported , in part , by the u.s .
department of energy under grant nos .
de - fg02 - 97er41027 , by the national sciences and engineering research council of canada , and by le fonds nature et technologies du qubec .
in this appendix , we consider the @xmath80 process of inverse bremsstrahlung , catalyzed by the soft , non - perturbative field , as depicted in fig .
[ fig:2to1 ] .
we will treat a case slightly more general than that depicted in the figure , namely the case of one particle with momentum @xmath166 and another particle with a momentum @xmath167 larger than @xmath7 , up to and including the scale @xmath36 .
we will find that the coherence time of this process is longer than the mean free time between collisions of the particles with the soft background ( via fig .
[ fig:1to1 ] ) , and so we will have to account for the landau - pomeranchuk migdal ( lpm ) effect @xcite in estimating the rate of inverse bremsstrahlung . for an initial estimate , however , let us momentarily ignore the lpm effect and consider the diagram of fig .
[ fig:2to1 ] in isolation .
bremsstrahlung in a soft background is a small - angle , nearly - collinear process .
parametrically , the rate for bremsstrahlung or inverse bremsstrahlung may be estimated as ( i ) the rate for small - angle scattering from the background field ( fig .
[ fig:1to1 ] ) , times ( ii ) a factor of @xmath168 for absorbing or emitting the additional gluon in fig .
[ fig:2to1 ] , times ( iii ) an initial or final state factor of @xmath18 or @xmath169 for that gluon , and ( iv ) a momentum integral @xmath170 ( responsible for the logarithmically large rate of soft bremsstrahlung emission in vacuum ) . as discussed in sec .
[ sec : dominant ] , the rate for small - angle scattering from the non - perturbative background is @xmath68 .
this gives @xmath171 for ( inverse ) bremsstrahlung .
to find the rate for a particle with momentum @xmath172 to substantially change its energy , we must correct for two things .
first , absorbing a momentum @xmath173 makes only an order @xmath174 relative change to @xmath2 .
second , it is the difference between the rates of absorption and emission of a particle of momentum @xmath167 which is relevant .
this involves the difference @xmath175 .
therefore we must introduce two powers of @xmath176 into the estimate above , to get @xmath177 for doubling energy through inverse bremsstrahlung . for @xmath178 ,
this is dominated by @xmath179 , and gives @xmath180 ( up to logarithms when @xmath30 ) . if we equated the power of @xmath36 in ( [ eq : nolpm2 ] ) with that in ( [ eq : ir ] ) , we would obtain @xmath30 , which would imply that this process was parametrically just as important as the multiple soft scattering treated in sec .
[ sec : dominant ] .
however , the rate ( [ eq : nolpm ] ) is an overestimate ; we must account for the lpm effect . in abelian theories
, lpm suppression occurs because ( i ) when a charged particle scatters , the probability that a photon will be bremsstrahlung emitted is not very sensitive to the scattering angle , and ( ii ) an emitted photon of a given wavelength can not resolve the difference between a single collision and @xmath70 collisions if the @xmath70 collisions are close enough together .
the largest such @xmath70 can be determined parametrically by equating ( i ) the time @xmath181 for @xmath70 consecutive scatterings of the form of fig .
[ fig:1to1 ] , where @xmath109 is the mean free time between scatterings , with ( ii ) the formation time @xmath182 for the complete @xmath183 bremsstrahlung ( or @xmath184 inverse bremsstrahlung ) process , where @xmath185 is the off - shellness of the energy during the process , which is @xmath186 for a @xmath187 process with @xmath188 . for the reasons of resolution already mentioned ,
each @xmath70 collisions are equivalent to a single collision in terms of the probability to induce bremsstrahlung , and so the rate @xmath49 for ( inverse ) bremsstrahlung processes is a factor of @xmath70 smaller than the naive estimate ( [ eq : nolpm ] ) which treats every scattering as independent . the same sort of estimate works for gluon bremsstrahlung in non - abelian gauge theories .
since bremsstrahlung is an almost collinear process , we can take @xmath189 , where @xmath190 is relative to the collinear axis , and approximate ( [ eq : deltae ] ) for @xmath23 as @xmath191 the main difference between photon bremsstrahlung in qed and gluon bremsstrahlung in qcd is that , in qcd , any of the three particles ( including the bremsstrahlung gluon ) can undergo the scattering .
therefore we may take each @xmath192 , in which case @xmath185 is dominated by the smallest energy particle , which is @xmath167 in our case .
further , @xmath193 can be dropped , giving @xmath194 plugging into ( [ eq : deltae2 ] ) , and equating the scattering time @xmath195 to @xmath181 as outlined previously , we have @xmath196 solving for @xmath70 , and using the fact that @xmath197 for the individual scattering processes of fig .
[ fig:1to1 ] , determines @xmath198 therefore , the rate is reduced by a factor of @xmath199 , which is a large suppression for @xmath200 .
this modifies eq .
( [ eq : nolpm2 ] ) to , @xmath201 this expression is dominated by small @xmath167 if @xmath202 .
the @xmath203 edge of the range of integration provides @xmath204 , which combined with ( [ eq : ir ] ) would now yield @xmath205 .
therefore , _ hard _ bremsstrahlung is not a competitive process .
however , the @xmath206 edge of the integration range gives @xmath207 , which combines with eq .
( [ eq : ir ] ) to give @xmath30 . absorption of an excitation with energy of order @xmath7 is competitive with ( and possibly indistinguishable from ) scattering from such excitations , but inverse bremsstrahlung of all harder scales is subdominant .
r. baier , a. h. mueller , d. schiff and d. t. son , `` ` bottom - up ' thermalization in heavy ion collisions , '' phys .
b * 502 * , 51 ( 2001 ) [ hep - ph/0009237 ] .
p. arnold , j. lenaghan and g. d. moore , `` qcd plasma instabilities and bottom - up thermalization , '' jhep 08 ( 2003 ) 002 [ hep - ph/0307325 ] .
a. h. mueller , a. i. shoshi and s. m. h. wong , `` a possible modified ` bottom - up ' thermalization in heavy ion collisions , '' hep - ph/0505164 .
d. bodeker , `` the impact of qcd plasma instabilities on bottom - up thermalization , '' hep - ph/0508223 .
e. s. weibel , `` spontaneously growing transverse waves in a plasma due to an anisotropic velocity distribution , '' phys
. rev .
* 2 * , 83 ( 1959 )
. u. w. heinz , `` quark - gluon transport theory , '' nucl .
a * 418 * , 603c ( 1984 ) .
y. e. pokrovsky and a. v. selikhov , `` filamentation in a quark - gluon plasma , '' jetp lett . *
47 * , 12 ( 1988 ) [ pisma zh .
fiz . * 47 * , 11 ( 1988 ) ] ; `` filamentation in quark plasma at finite temperatures , '' sov .
j. nucl .
phys . * 52 * , 146 ( 1990 ) [ yad .
fiz . * 52 * , 229 ( 1990 ) ] ; `` filamentation in the quark - gluon plasma at finite temperatures , '' sov .
j. nucl .
phys . * 52 * , 385 ( 1990 ) [ yad .
* 52 * , 605 ( 1990 ) ] .
o. p. pavlenko , `` filamentation instability of hot quark - gluon plasma with hard jet , '' sov .
j. nucl .
phys . *
55 * , 1243 ( 1992 ) [ yad .
55 * , 2239 ( 1992 ) ] .
s. , `` stream instabilities of the quark - gluon plasma , '' phys .
b * 214 * , 587 ( 1988 ) ; `` plasma instability at the initial stage of ultrarelativistic heavy ion collisions , '' * 314 * , 118 ( 1993 ) ; `` color collective effects at the early stage of ultrarelativistic heavy ion collisions , '' phys .
c * 49 * , 2191 ( 1994 ) ; `` color filamentation in ultrarelativistic heavy - ion collisions , '' phys .
b * 393 * , 26 ( 1997 ) [ hep - ph/9606442 ] ; j. randrup and s. , `` chromodynamic weibel instabilities in relativistic nuclear collisions , '' phys . rev .
c * 68 * , 034909 ( 2003 ) [ nucl - th/0303021 ] .
s. and m. h. thoma , `` hard loop approach to anisotropic systems , '' phys .
d * 62 * , 036011 ( 2000 ) [ hep - ph/0001164 ] ; p.
romatschke and m. strickland , `` collective modes of an anisotropic quark gluon plasma , '' phys .
d * 68 * , 036004 ( 2003 ) [ hep - ph/0304092 ] .
p. arnold and j. lenaghan , `` the abelianization of qcd plasma instabilities , '' phys .
d * 70 * , 114007 ( 2004 ) [ hep - ph/0408052 ] .
arnold , j. lenaghan , g. d. moore and l. g. yaffe , `` apparent thermalization due to plasma instabilities in quark gluon plasma , '' phys .
lett . * 94 * , 072302 ( 2005 ) [ nucl - th/0409068 ] .
a. rebhan , p. romatschke and m. strickland , `` hard - loop dynamics of non - abelian plasma instabilities , '' phys .
* 94 * , 102303 ( 2005 ) [ hep - ph/0412016 ] .
a. dumitru and y. nara , `` qcd plasma instabilities and isotropization , '' [ hep - ph/0503121 ] .
arnold , g. d. moore and l. g. yaffe , `` the fate of non - abelian plasma instabilities in 3 + 1 dimensions , '' phys .
d * 72 * , 054003 ( 2005 ) [ hep - ph/0505212 ] .
a. rebhan , p. romatschke and m. strickland , `` dynamics of quark - gluon plasma instabilities in discretized hard - loop approximation , '' hep - ph/0505261 .
p. arnold and g. d. moore , `` qcd plasma instabilities : the non - abelian cascade , '' hep - ph/0509206 .
d. t. son , `` reheating and thermalization in a simple scalar model , '' phys . rev .
d * 54 * , 3745 ( 1996 ) [ hep - ph/9604340 ] .
r. micha and i. i. tkachev , `` relativistic turbulence : a long way from preheating to equilibrium , '' phys .
lett . *
90 * , 121301 ( 2003 ) [ hep - ph/0210202 ] .
r. micha and i. i. tkachev , `` turbulent thermalization , '' phys . rev .
d * 70 * , 043538 ( 2004 ) [ hep - ph/0403101 ] .
a. h. mueller and d. t. son , `` on the equivalence between the boltzmann equation and classical field theory at large occupation numbers , '' phys .
b * 582 * , 279 ( 2004 ) [ hep - ph/0212198 ] .
v. n. tsytovich , _ lectures on non - linear plasma kinetics _ ( springer - verlag , 1995 ) .
a. b. migdal , `` bremsstrahlung and pair production in condensed media at high - energies , '' phys .
rev . * 103 * , 1811 ( 1956 ) ; doklady akad .
nauk s. s. s. r. * 105 * , 77 ( 1955 ) ; l. d. landau and i. pomeranchuk , dokl .
nauk ser .
* 92 * ( 1953 ) 535 ; `` electron cascade process at very high - energies , '' dokl . akad . nauk ser . fiz .
* 92 * ( 1953 ) 735 ; [ the last two papers are also available in english in l. landau , _ the collected papers of l.d .
( pergamon press , new york , 1965 ) . ] s. klein , `` suppression of bremsstrahlung and pair production due to environmental factors , '' rev .
* 71 * , 1501 ( 1999 ) [ hep - ph/9802442 ] .
p. arnold , g. d. moore and l. g. yaffe , `` photon and gluon emission in relativistic plasmas , '' jhep * 0206 * , 030 ( 2002 ) [ hep - ph/0204343 ] .
b. g. zakharov , `` fully quantum treatment of the landau - pomeranchuk - migdal effect in qed and qcd , '' jetp lett .
* 63 * , 952 ( 1996 ) [ hep - ph/9607440 ] ; `` light - cone path integral approach to the landau - pomeranchuk - migdal effect , '' phys .
* 61 * ( 1998 ) 838 [ yad
. fiz .
* 61 * ( 1998 ) 924 ] [ hep - ph/9807540 ] ; r. baier , y. l. dokshitzer , s. peigne and d. schiff , `` induced gluon radiation in a qcd medium , '' phys .
b * 345 * , 277 ( 1995 ) [ hep - ph/9411409 ] ; r. baier , y. l. dokshitzer , a. h. mueller , s. peigne and d. schiff , `` radiative energy loss of high energy quarks and gluons in a finite - volume quark - gluon plasma , '' nucl . phys .
b * 483 * , 291 ( 1997 ) [ hep - ph/9607355 ] ; r. baier , y. l. dokshitzer , a. h. mueller , s. peigne and d. schiff , `` radiative energy loss and p(t)-broadening of high energy partons in nuclei , '' * 484 * , 265 ( 1997 ) [ hep - ph/9608322 ] ; r. baier , d. schiff and b. g. zakharov , `` energy loss in perturbative qcd , '' ann .
nucl . part .
sci . * 50 * , 37 ( 2000 ) [ hep - ph/0002198 ] . | recent numerical work on the fate of plasma instabilities in weakly - coupled non - abelian gauge theory has shown the development of a cascade of energy from long to short wavelengths .
this cascade has a steady - state spectrum , analogous to the kolmogorov spectrum for turbulence in hydrodynamics or for energy cascades in other systems . in this paper
, we theoretically analyze processes responsible for this cascade and find a steady - state spectrum @xmath0 , where @xmath1 is the phase - space density of particles with momentum @xmath2 .
the exponent @xmath3 is consistent with results from numerical simulations .
we also discuss implications of the emerging picture of instability development on the `` bottom - up '' thermalization scenario for ( extremely high energy ) heavy ion collisions , emphasizing fundamental questions that remain to be answered .
l ss v c |
one of the predicted manifestations of lorentz invariance violation ( liv ) is a dependence of the speed of light in vacuo on its energy ( see refs . and references therein ) .
according to postulated liv effects , two photons of energies @xmath0 emitted simultaneously from a distant astrophysical source at redshift @xmath1 will travel with different velocities and will arrive with a time delay @xmath2 equal to@xcite : @xmath3 where @xmath4 is the ` quantum - gravity ( qg ) mass , ' a parameter that sets the energy scale at which the qg effects that cause liv start to become important .
its value is assumed to be near the planck mass ( @xmath5 gev/@xmath6 ) and most likely smaller than it .
the model - dependent parameter @xmath7 is assumed to be one or two , corresponding to linear ( @xmath8 , with @xmath9 ) and quadratic ( @xmath10 ) liv respectively .
the model - dependent parameter @xmath11 is equal to plus or minus one , corresponding to a speed retardation or acceleration with an increasing photon energy respectively . in ref . and using the above parametrization for @xmath2
, we have placed constraints on liv - induced dispersion in the form of lower limits on the quantum - gravity mass . in this proceeding
, we are also constraining liv using the parametrization of the standard - model extension ( sme)@xcite . because of their short duration , rapid variability , and cosmological distances ,
grbs are well - suited for constraining liv .
we used measurements on the bright and short grb 090510 ( @xmath12 ) , which triggered both the lat@xcite and gbm@xcite detectors on board _ fermi _ gamma - ray space telescope .
the gbm and lat lightcurves are shown in panels b f of fig .
[ fig : et ] . the emission detected by the lat extended to an energy of about 31 gev ( specifically @xmath13 gev with 1@xmath14 errors ) ( panel a of fig .
[ fig : et ] ) .
the fact that this 31 gev photon was detected shortly after the beginning of the burst ( @xmath150.8 s ) , and that the lat - detected emission exhibited a series of very narrow spikes that extended to high ( hundreds of mev ) energies , allowed us to set stringent limits on any liv - induced dispersion effects as described below . [
to constrain any liv - induced time delays we associated the 31 gev photon with a lower - energy ( le ) emission episode , during which we assumed that the 31 gev photon was emitted .
such an association set an upper limit on any propagation time delay of the 31 gev photon that was equal to the difference between the 31-gev - photon s detection time ( @xmath16=0.829 s ) and the time of the le - episode s start : @xmath17 .
this upper limit was then converted to a lower limit on @xmath4 using eq .
( [ eqdt ] ) . to be conservative in our calculations we used values for the 31 gev photon s energy and the grb s redshift reduced by 1@xmath14 .
the most conservative and of a very - high - confidence assumption that can be made regarding the possible emission time of the 31 gev photon is that it was not emitted before the beginning of the grb ( 30 ms before the trigger ) .
for such an assumption @xmath18 s , which corresponds to @xmath19 and @xmath20 . by associating the 31 gev photon with other le - emission episodes , as illustrated by the vertical lines in panels b f of fig .
[ fig : et ] , we also produced more stringent yet with less confidence upper limits , reported in table [ table_old ] .
our limits in the context of the sme and using the very high degree of confidence association with the e@xmath21100 mev emission ( @xmath22 s ) are shown in table [ table_sme ] .
it should be noted that all limits produced by this method correspond to the case of @xmath23 ( light speed retardation with an increasing energy ) .
[ table_old ] [ table_sme ] we also used an alternative and independent method to constrain any linear - in - energy dispersion effects ( discan method@xcite ) .
this method extracts dispersion information from all the lat - detected photons ( @xmath1530 mev@xmath1530 gev ) and is based on the fact that any qg - induced time delays would smear the spiky structure of the lightcurve .
it applies trial spectral lags to the detection time of each detected photon to find the spectral lag that maximizes the sharpness of the lightcurve .
the spectral lag that accomplishes that is equal and opposite in sign to the sum of any liv - induced and intrinsic - to - the - grb spectral lags .
figure [ fig : discan1 ] shows the value of a measure of the sharpness of the lat lightcurve versus the trial spectral - lag value .
the minimum of the curve , which denotes the most probable spectral lag , was at a value of @xmath24 ms / gev .
the errors here correspond to the trial spectral - lag values that were 100 times less probable than the best value of 0 ms / gev and are shown with the two vertical dashed lines in the same figure . to estimate the inherent uncertainties due to our choice of time interval and energy range , and due to the limited statistics of the dataset , a bootstrap analysis was performed , in which the discan method was applied to multiple data sets produced by randomizing the association between the energies and times of the actually - detected events .
the spectral lags of the randomized datasets , as measured by our method , lied within a value of @xmath2530 ms / gev in 99% of the cases .
therefore , our final combined result is a _
symmetric _ upper limit on the energy dispersion equal to @xmath2630 ms / gev which corresponds to an upper limit of @xmath27 for linear energy dispersion of either sign @xmath28 at the 99% c.l .
[ fig : discan1 ]
using the observations of grb 090510 by the gbm and lat , we constrained tiny variations on the speed of light in vacuo that are linear or quadratic in its energy .
we used two independent methods to obtain conservative and unprecedented upper limits on the magnitude of such speed variations .
our results ( @xmath29 few@xmath30 ) strongly disfavor any models predicting linear - in - energy variations in the speed of light .
the @xmath31 lat collaboration acknowledges support from a number of agencies and institutes for both development and the operation of the lat as well as scientific data analysis .
these include nasa and doe in the united states , cea / irfu and in2p3/cnrs in france , asi and infn in italy , mext , kek , and jaxa in japan , and the k.a .
wallenberg foundation , the swedish research council and the national space board in sweden .
additional support from inaf in italy and cnes in france for science analysis during the operations phase is also gratefully acknowledged .
the fermi gbm collaboration acknowledges the support of nasa in the united states and drl in germany .
9 g. amelino - camelia , nature * 393 * , 763 ( 1998 ) .
stecker , iap conf .
proc . * 1223 * , 192 ( 2010 ) , arxiv:0912.0500 .
u. jacob and t. piran , jcap * 0801*:031 , ( 2008 ) .
a. abdo , nature * 462 * , 331a ( 2009 ) .
kosteleck and m. mewes , phys . rev .
d * 80 * , 015020 ( 2009 ) .
atwood , ap . j. * 796 * , 1071 ( 2009 ) . c. meegan ,
ap . j. * 702 * , 791 m ( 2009 ) .
scargle , j.p .
norris , j.t .
bonnell , ap . j. * 673 * , 972 ( 2008 ) . | one of the cornerstones of special relativity is the postulate that all observers measure exactly the same photon speeds independently of the photon energies .
however , a hypothesized structure of spacetime may alter this conclusion at ultra - small length scales .
even a tiny energy - dependent variation in the speed of light may be revealed , when accumulated over cosmological light - travel times , by high temporal - resolution measurements of sharp features in gamma - ray burst ( grb ) lightcurves .
we report the results of a study of the emission from grb 090510 as detected by _ fermi _
s lat and gbm instruments that set unprecedented limits on the dependence of the speed of light on its energy . |
lattice quantum chromodynamic calculations predict a new state of matter of deconfined quark and gluons ( quark gluon plasma , qgp ) at an energy density exceeding @xmath4 1 gev/@xmath5 @xcite
. nuclear matter at high energy density has been extensively studied through ultra - relativistic heavy ion collisions ( for recent developements , see reference @xcite ) . within the experimental programme with heavy - ion beams at cern sps
, na57 is a dedicated experiment for the study of the production of strange and multi - strange particles in pb - pb collisions at mid - rapidity @xcite .
the measurement of strange particle production provides one of the most powerful tools to study the dynamics of the reaction .
in particular , an enhanced production of strange particles in nucleus nucleus collisions with respect to proton
induced reactions was suggested long ago as a possible signature of the phase transition from colour confined hadronic matter to a qgp @xcite .
the enhancement is expected to increase with the strangeness content of the hyperon .
these features were first observed by the wa97 experiment @xcite and subsequently confirmed and studied in more detail by the na57 experiment @xcite .
other insights into the reaction dynamics have been inferred from the @xmath6 distributions of the strange particles : the study of the transverse expansion of the collision and the @xmath6 dependence of the nuclear modification factors have been presented , respectively , in reference @xcite and @xcite .
rapidity distributions provide a tool to study the longitudinal dynamics ; e.g. , differences between protons and anti - protons have been interpreted as a consequence of the nuclear stopping @xcite . if hyperons , like protons , keep a ` memory ' of the initial baryon density , then the _ relative _ pattern for the rapidity distribution of hyperons and anti - hyperons should resemble that of protons and anti - protons @xcite .
hydrodynamical properties of the expanding matter created in heavy ion reactions have been discussed by landau @xcite and bjorken @xcite in theoretical pictures using different initial conditions . in both scenarios , thermal equilibrium
is quickly achieved and the subsequent isentropic expansion is governed by hydrodynamics .
the na57 experiment has been described in detail elsewhere @xcite .
the results presented in this paper are based on the analysis of the pb - pb data sample collected at 158 @xmath2 gev/@xmath3 ( 460 m events ) ; the sample analyzed here corresponds to the most central 53% of the inelastic pb - pb cross - section .
the data have been divided into five centrality classes , labelled as 0,1,2,3 and 4 from the most peripheral to the most central , according to the value of the charged particle multiplicity sampled at central rapidity by a dedicated silicon microstrip detector .
the procedure to measure the multiplicity distribution and to determine the centrality of the collision for each class is described in reference @xcite .
the fractions of the inelastic cross - section for each of the five classes , determined assuming an inelastic pb - pb cross - section of 7.26 barn ( as calculated with the glauber model ) , are given in table [ tab : centrality ] .
.centrality ranges for the five classes .
[ tab : centrality ] [ cols="<,<,<,<,<,<",options="header " , ] in landau hydrodynamics , the amount of entropy ( @xmath7 ) contained within a ( fluid ) rapidity @xmath8 is given by @xcite @xmath9 \left [ i_0(q ) - \frac{\beta\omega_f}{q}i_1(q ) \right ] \label{eq : landau}\ ] ] where @xmath10 , @xmath11 , @xmath12 is the velocity of sound in the medium , @xmath13 is the initial temperature , @xmath14 is the rapidity , @xmath15 is the radius of the nuclei , @xmath16 is the initial length , @xmath17 is the initial entropy density , @xmath18 and @xmath19 , @xmath20 are bessel functions .
the quantity @xmath21 is used to normalize the spectra at mid - rapidity .
the particle rapidity distribution is obtained , as for the bjorken case , as a superpositon of the multiplicity density in rapidity space ( @xmath22 ) with a thermal distribution of the fluid elements , @xmath23 in the landau model the width of the rapidity distribution is sensitive to the velocity of sound and to the ratio of the freeze - out temparature to the initial temperature .
while integrating over @xmath14 in equation [ eq : bjorken2 ] , the range of @xmath14 is treated as a parameter in case of bjorken hydrodynamics ; moreover in the bjorken case ( equation [ eq : bjorken2 ] ) the factor @xmath24 , which appears in equation [ eq : landau2 ] , has been included in the overall normalization factor @xmath2 since the entropy density @xmath25 is independent of the rapidity , in accordance with the assumption of boost invariance along the longitudinal direction @xcite . in the case of landau hydrodynamics , the integration limit is fixed by @xmath26 and the multiplicity density in @xmath14 space ( @xmath24 ) is written explicitly in the @xmath14 integration ( equation [ eq : landau2 ] ) .
landau hydrodynamics can also reproduce simultaneously the distributions for all the strange particles considered ( @xmath27 28/32 ) but we are not able to put stringent constraints on both the velocity of sound and the ratio @xmath28 .
the confidence level contours in the @xmath29 vs. @xmath30 parameter space are shown in figure [ fig : landau ] .
for instance , the hypothesys of a perfect gas ( i.e. @xmath31 ) would result ( at the 3@xmath32 confidence level ) in either @xmath33 or @xmath34 .
in fact , two physical regions are constrained at the 3@xmath32 confidence level , the first located at small values of @xmath28 and the second between 0.5 and 0.8 ; on the other hand , the region at @xmath35 is unphysical .
both physical regions span over the full range @xmath36 .
we have measured the @xmath37 distributions of high purity samples of , , @xmath0 and @xmath1 particles produced at central rapidity in pb - pb collisions at 158 @xmath2 gev/@xmath3 over a wide centrality range of collision ( i.e. the most central 53% of the pb - pb inelastic cross - section ) . in the unit of rapidity around mid - rapidity covered by na57 , we have performed fits to the @xmath37 distributions of and using a gaussian parameterization : the resulting widths are compatible with each other and constant as a function of centrality .
contrary to , the spectra are flat to good accurancy in the range of rapidity and centrality considered ; this would indicate that the hyperon conserves `` memory '' of the initial baryon density .
the rapidity distributions of the @xmath1 particle are found to be flat within the errors in one unit of rapidity for central ( 0 - 11% ) and peripheral ( 23 - 53% ) collisions .
boost - invariant bjorken hydrodynamics can describe simultaneously the rapidity spectra of all the strange particles under study with @xmath38 , yielding an average longitudinal flow velocity @xmath39 , sligthly larger than the measured transverse flow .
isotropic _ collective expansion of the system suggests large nuclear stopping . a fairly good description is also provided by landau hydrodynamics , which allows us to put constraints in the parameter space of the velocity of sound in the medium and the ratio of the freeze - out temperature to the initial temperature .
we are grateful to jan - e alam for useful comments and fruitful discussions .
33 karsch f 2002 _ lect .
notes phys . _ * 583 * 209 quark matter conference proceedings 2004 s633-s1430 caliandro r
_ et al . _ , na57 proposal , 1996 _ cern / spslc 96 - 40 , spslc / p300 _ rafelski j and mller b 1982 1066 rafelski j and mller b 1986 2334 andersen e _ et al .
_ 1999 b * 449 * 401 antinori f _ et al .
_ 1999 a * 661 * 130c bruno g e _ et al . _
2004 s717-s724 antinori f _ et al .
_ 2004 823 - 840 antinori f _ et al .
_ 2005 b * 623 * 17 - 25 busza w and goldhaber a 1984 b * 139 * 235 appelshuser h _
_ 1999 2471 landau l d 1953 _ izv .
_ sssr * 17 * 51 belenkij s and landau l d 1955 _ usp .
nauk . _ * 56 * 309 belenkij s and landau l d 1956 ( suppl . ) * 3 * 15 bjorken j d 1983 d * 27 * 140 manzari v _ et al .
_ 1999 473 manzari v _ et al .
_ 1999 a * 661 * 761c carrer n _ et al .
_ 2001 391 antinori f _ et al .
_ 2005 321 - 335 geant , cern program library long writeup w5013 fanebust k _ et al .
_ 2002 160 schnedermann e , sollfrank j and heinz u 1993 c * 48 * 2462 - 2475 antinori f _ et al .
_ 2004 b * 595 * 68 - 74 blume c
_ 2005 s685-s691 anticic t _
_ 2004 022302 afanasiev s v _ et al .
_ 2002 c * 66 * 054902 afanasiev s v _ et al .
_ 2002 b * 538 * 275 - 281 alt c _ et al .
_ alt c _ et al .
_ 2005 192301 elia d _ et al . _
2004 s1329-s1332 mohanty b and alam j 2003 c * 68 * 064903 srivastava d k , alam j , chakrabarty s , raha s and sinha b 1993 _ ann .
phys . _ * 228 * 104 | the production at central rapidity of , , @xmath0 and @xmath1 particles in pb - pb collisions at 158 @xmath2 gev/@xmath3 has been measured by the na57 experiment over a centrality range corresponding to the most central 53% of the inelastic pb - pb cross section . in this paper
we present the rapidity distribution of each particle in the central rapidity unit as a function of the event centrality .
the distributions are analyzed based on hydrodynamical models of the collisions . |
many complex systems can be expressed as networks in which a node represents an object ( e.g. , person , web page , protein ) and an edge represents the relationship between two objects ( e.g. , friendship , hyperlink , physical interaction )
. a network can be characterised by microscale , mesoscale and macroscale structural patterns such as the degree ( i.e. , the number of edges that a node has ) , clustering coefficient , and diameter @xcite . among various structural properties of networks ,
community structure is a representative mesoscale structure of networks @xcite .
a community is a group of densely interconnected nodes while different communities are sparsely interconnected .
a community often corresponds to a group of nodes sharing a role , and identifying communities aids classification of nodes and visualisation of networks @xcite .
core - periphery structure is another mesoscale structure of networks , with which we view a network as being composed of two groups of nodes called the core and periphery .
although the definition varies , a core is often defined as a group of densely interconnected nodes , and a periphery as a group of nodes that are densely connected ( i.e. , adjacent ) to the core nodes but not to other peripheral nodes @xcite .
although a core and community are both a group of densely interconnected nodes , they are different ; a core connects densely to its periphery , whereas a community is not densely connected to other nodes outside it .
core - periphery structure has been found in various networks including brain networks @xcite , metabolic networks @xcite , protein interaction networks @xcite , social networks @xcite , international trade networks @xcite , financial networks @xcite and transportation networks @xcite .
for example , in a co - authorship network among researchers , notable researchers often publish papers with other notable researchers , forming a core , while other researchers tend to publish papers with particular notable researchers such as those in the same research group , forming a periphery @xcite .
borgatti and everett introduced the first quantitative formulation of core - periphery structure @xcite . in the discrete version of core - periphery structure @xcite , which we will focus on in the present study , borgatti and
everett introduced an idealised core - periphery structure in which core nodes are adjacent to all other nodes and peripheral nodes are adjacent to all core nodes but not to any peripheral node .
although it is also realistic to assume that the core - periphery connectivity is sparser than the core - core connectivity @xcite , we use the idealised core - periphery structure in the present study .
then , they sought for the assignment of all nodes in a given network to a core or periphery such that the network is as close as possible to an idealised core - periphery structure . since then
, many core - periphery detection algorithms have been developed @xcite .
these algorithms aim to identify a single core - periphery pair embedded in a network . however , a network may be better regarded as a collection of multiple core - periphery pairs @xcite , which is the focus of the present study .
previous studies in this direction have not provided a tailored scalable algorithm to this end .
a study focussed on a related but different type of multiple core - periphery structure @xcite .
other algorithms aim to detect multiple cores but do not assume that peripheral nodes are sparsely connected to each other @xcite .
a network can also have multiple disjoint cores in the form of @xmath0-cores @xcite , @xmath0-trusses @xcite or dense subgraphs @xcite .
however , the corresponding algorithms do not tell how densely peripheral nodes are connected to each other or to which core a peripheral node belongs . another algorithm to find various mesoscale structure of networks including multiple core - periphery pairs @xcite is computationally costly and only feasible for small networks ( supplementary note 1 ) .
we present a scalable algorithm to detect multiple non - overlapping core - periphery pairs in networks , each of which is as close as possible to an idealised core - periphery structure .
our algorithm automatically determines the number and the size of the core - periphery pairs .
various algorithms to detect core - periphery structure in networks are classified to density - based and transport - based algorithms @xcite .
densely - based algorithms posit that a core is a densely connected group of nodes , whereas transport - based algorithms posit that a core is a group of nodes that can be reached to other nodes along short paths . in the present study ,
we focus on density - based algorithms .
we extend the idealised core - periphery structure introduced by borgatti and everett @xcite to the case of multiple pairs of a core and a periphery . in the borgatti
everett ( be ) algorithm , one considers a graph ( i.e. , network ) composed of @xmath1 nodes and @xmath2 edges .
let @xmath3 be the adjacency matrix , i.e. , @xmath4 if node @xmath5 and @xmath6 are adjacent by an edge , and @xmath7 otherwise .
we assume an undirected and unweighted network without self - loops , i.e. , @xmath8 and @xmath9 for all @xmath5 and @xmath6 .
let @xmath10 be a vector of length @xmath1 , where @xmath11 if node @xmath5 is a peripheral node , and @xmath12 if node @xmath5 is a core node .
we define the idealised core - periphery structure as the network where each core node is adjacent to all core and all peripheral nodes , and each periphery node is adjacent to all core nodes but no peripheral node . the corresponding adjacency matrix , @xmath13 , is given by @xmath14 the discrete version of the borgatti - everett ( be ) algorithm , which we consider in the present study , seeks @xmath15 that maximises similarity between @xmath16 and @xmath17 .
we will explain the similarity measure in section [ sec : stest ] .
we extend the idealised core - periphery structure to the case of multiple core - periphery pairs .
let @xmath18 be the number of core - periphery pairs and @xmath19 be a vector of length @xmath1 , where @xmath20 is the index of the core - periphery pair to which node @xmath5 belongs .
we exclude overlaps between core - periphery pairs , and between the core and periphery in each core - periphery pair . the corresponding adjacency matrix , @xmath21 , is given by @xmath22 where @xmath23 is kronecker delta .
we seek @xmath24 that makes @xmath21 the closest to @xmath16 by maximising @xmath25 where @xmath26 $ ] is the density of edges in the network .
the term @xmath27 represents the number of edges that are present in both the given network and the idealised core - periphery structure .
the null - model term @xmath28 is the expected number of edges that are present in both the idealised core - periphery structure and an random graph @xcite , in which each pair of nodes is adjacent with probability @xmath29 .
a large @xmath30 value indicates that the given network and the idealised core - periphery structure share more edges than by chance .
we use a label switching heuristic @xcite to maximise @xmath30 .
first , we assign each node to a different core by setting @xmath31 ( @xmath32 ) . then
, we scan all nodes in a random order . for each scanned node @xmath5 , we calculate the increment in @xmath30 when we tentatively update @xmath33 to the core of the core - periphery pair that a neighbour of node @xmath5 , denoted by @xmath6 , belongs to , i.e. , @xmath34 .
we also calculate the increment in @xmath30 when we tentatively update @xmath33 to @xmath35 .
note that we experiment on these two cases regardless of whether @xmath36 or @xmath37 .
we carry out this procedure for all neighbours of @xmath5 to measure the increment in @xmath30 in each case .
finally , we update @xmath33 to the label that has yielded the largest tentative increment in @xmath30 ( i.e. , @xmath35 or @xmath34 for a neighbour @xmath6 ) .
if any relabelling does not increase @xmath30 , we do not update @xmath38 .
when we have scanned all nodes , we stop the entire procedure if no node has changed its label in the present round .
otherwise , we draw a new random order of nodes and scan all nodes again according to the new random order . the increment in @xmath30 by changing node @xmath5 s label from @xmath39 to @xmath40
is given by @xmath41 \nonumber \\ & - [ \tilde{d}_{i,(c , 1 ) } + x \tilde{d}_{i,(c , 0 ) } - p ( \tilde{n}_{(c , 1 ) } + x \tilde{n}_{(c , 0)}-x ) ] , \end{aligned}\ ] ] where @xmath42 is the number of neighbours of node @xmath5 that have label @xmath39 , and @xmath43 is the number of nodes with label @xmath39 .
for each scanned node @xmath5 , we calculate equation ( [ eq : dq ] ) at most @xmath44 times , where @xmath45 is the degree of node @xmath5 . therefore , the time complexity for scanning all nodes in one round is @xmath46 , and that of the entire algorithm is @xmath47 , where @xmath48 is the number of rounds .
we run this algorithm 20 times starting from the same initial condition stated above and adopt the node labelling that produces the largest value of @xmath30 .
a detected core - periphery structure may be statistically insignificant @xcite .
therefore , we adapt a statistical test in the case of a single core - periphery pair @xcite to the case of multiple core - periphery pairs .
we first describe the statistical test for a single core - periphery pair @xcite .
we measure the significance of a core - periphery pair by a quality function based on the pearson correlation coefficient @xcite , which is defined by @xmath49 where @xmath50 $ ] . a core - periphery pair detected for the given network
is deemed to be significant if @xmath51 is statistically larger than @xmath51 values calculated for a null model .
one uses the random graph as the null model , in which the number of edges is the same as that of the original network .
one generates many networks using the random graph and maximises @xmath51 for each network .
the kernighan - lin ( kl ) algorithm @xcite is used for maximising @xmath51 .
the core - periphery pair detected for the original network is significant at a significance level of @xmath52 $ ] if the @xmath51 value for the original network is larger than a fraction @xmath53 of the @xmath51 values for the randomised networks . in case of multiple core - periphery pairs , we apply essentially the same statistical test to each core - periphery pair detected in the original network . for each detected core - periphery pair
, we first calculate @xmath51 .
second , we generate 3,000 networks using the random graph , which have the same number of nodes and edges as those of the core - periphery pair . in counting the number of edges , we only consider the edges connecting nodes within the core - periphery pair .
third , we detect a single core - periphery pair in each randomised network by maximising @xmath51 using the kl algorithm .
fourth , we compare the obtained @xmath51 values between the original and randomised networks . if a core - periphery pair is judged to be insignificant , we call the corresponding nodes the residual nodes , i.e. , those not belonging to any significant core - periphery pair .
if we test @xmath18 core - periphery pairs at a significance level of @xmath54 , the probability of making at least one false positive ( i.e. , an insignificant core - periphery pair is judged to be significant ) is @xmath55 , which increases as @xmath18 increases . to remedy this multiple comparison problem
, we adopt the idk correction , with which we test each core - periphery pair at a significance level of @xmath56 , which is equivalent to @xmath57 @xcite .
we set @xmath58 .
we have decided to use @xmath51 maximised by the kl algorithm as the test statistic to compare the original and randomised networks .
however , we can also use different algorithms to maximise @xmath51 .
we can also use a different test statistic including @xmath30 restricted to the case of the one core - periphery pair ( i.e. , @xmath59 ) .
for the synthetic networks with planted core - periphery structure , we measure the difference between the true core - periphery structure @xmath24 and the inferred core - periphery structure @xmath60 by the variation of information ( vi ) @xcite . the vi is given by @xmath61 ^ 2 } { \left[\sum_{(\hat c',\hat x')}p(c , x;\hat c',\hat x')\right ] \times \left[\sum_{(c',x')}p(c',x';\hat c,\hat x)\right ] } , \end{aligned}\ ] ] where @xmath62 is the fraction of nodes that have the true label @xmath39 and inferred label @xmath63 .
the vi value is equal to zero if and only if the inferred core - periphery structure is the same as the true one .
we measure the performance of an algorithm by averaging vi values over the 100 generated networks .
we compare the proposed algorithm with a previous algorithm , which detects a single core - periphery pair by maximising @xmath51 using the kl algorithm @xcite .
we refer to the latter algorithm as the be
kl algorithm .
we also compare our algorithm with another ad - hoc two - step in which we divide a network into a core and a periphery , and also divide the same network into non - overlapping communities by maximising modularity using the louvain algorithm @xcite .
then , we regard the core and peripheral nodes in each detected community as a core - periphery pair . to divide the network into a core and a periphery
, we use the be kl algorithm .
we apply the statistical test ( section [ sec : stest ] ) to the core - periphery pairs detected by the three algorithms .
we do not compare these algorithms with the algorithm introduced by and verma @xcite because of a low speed and insufficient performance of their algorithm on model networks with planted core - periphery structure ( supplementary note 1 ) .
we compare the performance of the three algorithms on four different types of synthetic networks with a planted core - periphery structure schematically shown in fig .
[ fig : sb ] .
we generate the synthetic networks using stochastic block models @xcite .
we draw label @xmath38 for the @xmath5th node @xmath64 from a distribution @xmath65 ( @xmath66 and @xmath67 ) , where @xmath68 is the probability that @xmath69 and satisfies @xmath70 . then
, we place edges between each pair of nodes with label @xmath39 and @xmath71 with probability @xmath72 . for each type of the stochastic block model , we generate 100 networks and detect core - periphery pairs by the three algorithms . as a first example , we consider a network composed of a single core - periphery pair ( fig .
[ fig : sb]a ) .
we set @xmath73 , @xmath74 , @xmath75 and @xmath76 , where @xmath77 .
we note that the core - periphery structure is evident for a large @xmath78 value .
for this network model , the vi value , quantifying the discrepancy between the true and inferred core - periphery structure , is compared for the three algorithms in fig .
[ fig : exp]a .
the vi values for the be kl algorithm are approximately equal to zero for most of @xmath78 values ( @xmath79 ) .
the vi values for the two - step algorithm are large even for a large @xmath78 value because the two - step algorithm divides the single core - periphery pair into communities .
in contrast , the vi values for the proposed algorithm are close to zero when @xmath80 .
therefore , the performance of the proposed algorithm on this network model is comparable to that of the be kl algorithm when @xmath78 is not too small .
as a second example , we examine networks composed of two core - periphery pairs ( fig . [
fig : sb]b ) .
we set @xmath81 , @xmath82 , @xmath83 , and @xmath84 for @xmath85 and @xmath86 .
the vi values for this network are shown in fig .
[ fig : exp]b .
the vi for the be
kl algorithm is large for all @xmath78 values because the be
kl algorithm is not designed for multiple core - periphery pairs .
the vi values for the two - step algorithm and those for the proposed algorithm are similar and close to zero for @xmath87 . in empirical networks
, there may be nodes that are better regarded not to belong to any core or periphery .
therefore , as a third example , we consider a network composed of a single core - periphery pair and residual nodes ( fig .
[ fig : sb]c ) .
we regard the block of the residual nodes as a single group of nodes , like a core or periphery , when calculating the vi value .
let @xmath88 be the index for the block of the residual nodes .
we set @xmath89 , @xmath90 , @xmath75 and @xmath91 for @xmath92 .
the vi values for this model are shown in fig .
[ fig : exp]c .
the vi for the be
kl algorithm is large because the residual nodes are classified as peripheral nodes .
the vi for the two - step algorithm is large for the entire range of @xmath78 .
in contrast , the vi for our algorithm is close to zero for @xmath93 . as a fourth example , we consider networks composed of two core - periphery pairs and residual nodes ( fig . [ fig : sb]d ) .
we set @xmath94 , @xmath95 , @xmath83 for @xmath96 and all other @xmath97 values to 0.01 .
the vi values for this network model are shown in fig .
[ fig : exp]d .
the vi for the be
kl algorithm is large for all @xmath78 .
the vi for the two - step and proposed algorithm decreases as @xmath78 increases .
the proposed algorithm attains smaller values of vi than the two - step algorithm in almost the entire range of @xmath78 .
we apply the three algorithms to three empirical networks . for
directed and weighted networks , we disregard the direction and the weight of edges .
consider the karate club network @xcite , which has @xmath98 nodes and @xmath99 edges ( edge density @xmath100 ) .
a node represents a member of a karate club at a university .
two members are adjacent if they have socially interacted outside club activities during the observation period . during the study ,
a conflict occurred between the instructor ( node 1 ) and the president ( node 34 ) , which fissured the club .
based on self - reports , each member was labelled the instructor side ( 15 members ) , the president side ( 16 members ) or neutral ( 3 members ) @xcite .
the core - periphery structure detected by the three algorithms is shown in fig .
[ fig : karate ] .
the be kl algorithm detects a single core - periphery pair such that both the instructor and president are core nodes ( fig .
[ fig : karate]a ) , neglecting the fissure of the club .
the two - step algorithm detects two core - periphery pairs , each of which consists mostly of the members with the same leanings ( fig .
[ fig : karate]b ) . in particular ,
the instructor and the president belong to the core of the different core - periphery pairs .
two neutral members , nodes 10 and 19 , are assigned to the president s core - periphery pair , which does not agree with the self - reports by the members .
the residual nodes consist of the members on the instructor side , those on the president side and a neutral member .
our algorithm detects almost the same two core - periphery pairs as that detected by the two - step algorithm ( fig .
[ fig : karate]c ) .
the second example is a political blog network @xcite , which has @xmath101 nodes and @xmath102 edges ( edge density @xmath103 ) .
a node is a blog on the united states president election in 2004 , and two blogs are adjacent if one blog cites the other blog on its front page .
each blog was labelled with one of the political leanings , liberal ( 586 blogs ) or conservative ( 636 blogs ) , determined by automated categorisations by several weblog directories @xcite . if a blog was uncategorised or classified to conflicting categories , the authors of ref .
@xcite manually judged the political leaning . the core - periphery structure detected by the three algorithms
is shown in fig .
[ fig : poliblog ] .
the unique core detected by the be
kl algorithm is a mixture of liberal and conservative blogs ( fig .
[ fig : poliblog]a ) .
the peripheral blogs are mostly adjacent to blogs with the same political leaning
. however , the structure detected by the be
kl algorithm alone does not tell this unless we refer to the political learning of the individual blogs . a different algorithm for a single core - periphery structure yielded similar results for the same network @xcite .
the two - step algorithm detects three core - periphery pairs , each of which mostly comprises the blogs with the same political leanings ( fig .
[ fig : poliblog]b ) .
two core - periphery pairs are much larger than the third one and have the opposite political leanings . the third small core - periphery pair is mainly composed of liberal blogs . in each core - periphery pair , a majority of the peripheral nodes is densely interconnected , which is against the idealised core - periphery structure .
this is due to the community detection step that partitions a network into communities with dense intra - community edges .
our algorithm detects two core - periphery pairs , each of which mostly comprises the blogs with the same political leaning ( fig . [
fig : poliblog]c ) .
the detected two core - periphery pairs are smaller than those detected by the two - step algorithm .
more nodes are classified as residual nodes than by the two - step algorithm . with our algorithm ,
the edges between peripheral nodes within each core - periphery pair are sparser than those detected by the two - step algorithm , respecting the notion of the idealised core - periphery structure .
our third example is a network of airports , which has @xmath104 nodes and @xmath105 edges ( edge density @xmath106 ) @xcite .
a node represents an airport .
two airports are adjacent if there is a direct commercial flight between them .
figure [ fig : airport ] shows the core - periphery structure detected by the three algorithms .
kl algorithm detects a dense core composed of 89 airports scattered in different geographical regions ( fig .
[ fig : airport]a ) .
the peripheral airports are rarely adjacent to the core airports in other regions .
furthermore , the peripheral airports tend to be adjacent to other peripheral airports in the same region , which is inconsistent with the notion of the periphery .
the two - step algorithm detects 16 geographically concentrated core - periphery pairs ( fig .
[ fig : airport]b ) .
our algorithm detects ten geographically concentrated core - periphery pairs ( fig .
[ fig : airport]c ) .
the partition of the world - wide airport network into geographically distinct groups of airports found here is consistent with the previous results derived with community detection algorithms @xcite .
compared to the two - step algorithm , the peripheral airports detected by our algorithm are not densely interconnected .
we further analyse the core - periphery structure obtained by our algorithm .
figure [ fig : map_world ] maps the locations of the core and peripheral airports .
the three largest core - periphery pairs labelled 1 , 2 and 3 are mainly based in europe , east asia and the united states , respectively .
the core - periphery pairs 1 , 2 and 3 consist of the airports in 125 , 35 and 47 countries , respectively .
each of the other core - periphery pairs labelled 410 consists of the airports in one country .
the location of the airports and metropolises in europe , east asia , the united states and their surroundings are shown in fig .
[ fig : map ] .
here the metropolis is defined as the capital city of all countries , the provincial capitals of china and the state capitals of the united states because china and the united states have many airports .
core - periphery pair 1 contains 333 core airports and 378 peripheral airports , of which 405 ( 57% ) airports are located in europe ( fig .
[ fig : map]a ) .
however , this core - periphery pair excludes most airports in the nordic countries ( 84 airports ; 68% ) .
there are 89 airports within 20 miles from a metropolis in europe , among which there are 51 core airports ( 57% ) , 28 peripheral airports ( 31% ) and ten residual airports ( 11% ) . as a comparison ,
if we select the same number of the european airports with the largest degrees as that of the european core airports , 64 airports ( 72% ) are contained in the set of 89 airports within 20 miles from a metropolis , which is more than the number of the core airports ( 51 airports ; see above ) contained in the same set of airports .
this result indicates that hub metropolitan airports , which are common , are not necessarily core airports .
the second core - periphery pair contains 161 core airports and 240 peripheral airports , among which 217 ( 54% ) airports are located in east asia ( fig .
[ fig : map]b ) . in this core - periphery pair ,
31 airports are located within 20 miles from a metropolis in east asia , among which there are 23 core airports ( 74% ) , eight peripheral airports and no residual airport ( fig .
[ fig : map]b ) .
the third core - periphery pair contains 150 core airports and 312 peripheral airports , among which 210 ( 45% ) airports are located in the united states ( fig .
[ fig : map]c ) . in this core - periphery pair , 71 airports
are located within 20 miles from a metropolis in the united states , among which there are 29 core airports ( 41% ) , 30 peripheral airports ( 42% ) and 12 residual airports ( 17% ) ( fig .
[ fig : map]c ) . we have not found the partitioning of airports into core - periphery pairs corresponding to different major airline groups ( e.g. , american airlines , delta airlines , southwest airlines and united airlines in the united states ) .
table [ ta : profile ] lists the size of core - periphery pairs and the fractions of different types of edges .
the airports in a large core are not densely interconnected compared to those in small core - periphery pairs , probably due to the limited capacity of the airports ( e.g. , a small number of runways ) .
core - periphery pairs 1 , 2 and 3 contain hub airports in each region .
the other small core - periphery pairs consist of a small number of core airports , i.e. , at most 20% of the airports in each core - periphery pair . in these core - periphery pairs ,
most of the peripheral airports are adjacent to the core airports but not to other peripheral airports in the same core - periphery pair .
this observation suggests that a small number of core airports relays most of the flights into these regions as gateway airports .
for example , the representative core airport ( i.e. , the core airport that has the largest number of neighbours in the core - periphery pair ) in pair 4 , mnl ( philippines ) , and that in pair 8 , los ( nigeria ) , serve most of the domestic airports in the respective countries . such a structure is evident in small core - periphery pairs such as core - periphery pairs 410 .
the subnetwork within the philippines is shown in fig .
[ fig : philippines_thailand]a ; see supplementary table 1 for properties of all airports in the philippines .
most of the airports ( 34 airports ; 92% ) in core - periphery pair 4 ( shown in orange in figs .
[ fig : map_world ] , [ fig : map]b and [ fig : philippines_thailand]a ) only serve domestic flights . core airport 1 ( labelled in fig .
[ fig : philippines_thailand]a ) has most of the edges ( 41 edges ; 84% ) between core - periphery pair 4 and the rest of the network .
therefore , core airport 1 functions as a gateway airport in the philippines . core airport 2 also functions as a gateway airport , but to a lesser extent than core airport 1 does .
core - periphery pairs located in alaska ( core - periphery pair 6 in table [ ta : profile ] ) , russia ( pair 7 ) and ecuador ( pair 9 ) also contain a few core airports serving as gateway airports in the respective regions ( supplementary note 2 ) .
core airports 8 and 21 in the philippines ( fig .
[ fig : philippines_thailand]a ) have a small degree , which is counterintuitive .
core nodes having degree one or two are also found in core - periphery pair 6 ( supplementary fig .
3c ) . the airports 8 and 21 in the philippines are adjacent to one peripheral airport 7 and 4 , respectively .
if we assign airport 8 to the periphery , two peripheral airports 7 and 8 would be adjacent .
similarly , if we assign airport 21 to the periphery , two peripheral airports 4 and 21 would be adjacent . to avoid edges between peripheral nodes ,
our algorithm has identified airports 8 and 21 as core nodes .
however , airports 8 and 21 may be better regarded as peripheral airports given that they are not densely interconnected to other core airports .
previous studies provided remedies for this issue @xcite ( see section [ sec : discussion ] for further discussion ) .
the subnetwork within thailand is shown in fig .
[ fig : philippines_thailand]b ; see supplementary table 2 for properties of all airports in thailand .
two major airports 1 and 14 are located in the capital city , bangkok , and belong to different core - periphery pairs . all international airports in thailand belong to core - periphery pair 2 ( shown in blue in figs .
[ fig : map_world ] , [ fig : map]b and [ fig : philippines_thailand]b ) , including core airport 14 .
most of the domestic airports ( 13 airports ; 59% ) belong to core - periphery pair 10 ( shown in magenta ) , including core airport 1 .
the subnetwork composed of core - periphery pair 10 is largely separated from the other airports in thailand , which belong to core - periphery pair 2 , and the rest of the world .
the separation of the domestic and international airports and their respective subnetworks is also observed in the philippines ( fig .
[ fig : philippines_thailand]a ) , iran and nigeria ( supplementary note 2 ) .
we compare the speed of the three algorithms on the synthetic and empirical networks .
we implement all the algorithms in matlab and run simulations on a computer with intel 2.6ghz sandy bridge processors and 4 gb of memory .
we measure the speed of an algorithm by averaging the cpu time over 100 runs .
we do not run the statistical test because it is common to all the algorithms .
the average cpu time of the three algorithms is compared in table [ ta : cputime ] .
kl algorithm is the fastest on all the synthetic and karate club networks .
however , it is slower than our algorithm on the political blog and airport networks .
the two - step algorithm is the slowest on all the networks .
our algorithm is approximately two times slower than the be
kl algorithm on the synthetic and karate club networks .
however , on the political blog and airport networks , it runs much faster than the be
kl algorithm .
our algorithm runs in @xmath47 time , where @xmath48 is the number of rounds ( section [ sec : algorithm ] ) .
therefore , our algorithm is expected to be fast on large sparse networks .
we proposed a scalable algorithm to detect multiple core - periphery pairs in networks by maximising a novel quality function @xmath30 .
the quality function @xmath30 compares the number of edges of different types in the given network with the expected number for an random graph .
in the airport network , we have found several core nodes having degree one or two ( e.g. , airport 8 and 21 in fig .
[ fig : philippines_thailand]a ) , which contradicts the intuition that core nodes would have a large degree .
our algorithm assigned these nodes to a core to suppress the edges between peripheral nodes . however , these nodes may be better regarded as peripheral nodes because they are not adjacent to all but one core node .
one remedy is to weaken the suppression of the edges between peripheral nodes @xcite .
adapting this idea to the case of multiple core - periphery structure warrants future research .
we used the random graph as the null model to define @xmath30 .
the random graph is also used as a null model for detecting communities in networks . in the inference of a stochastic block model without degree corrections (
i.e. , without assuming a heterogeneous degree distribution ) , one compares the distribution of edges within and across different blocks , between the given network and an random graph @xcite . in a community detection algorithm based on the potts model ,
one compares the number of edges within proposed communities in the given network and that for an random graph @xcite .
our @xmath30 is equivalent to the quality function used in the latter algorithm if we enforce that all nodes are core nodes , i.e. , @xmath107 .
the configuration model , with which we randomly rewire edges conserving the degree of each node , is a more commonly used null model for community detection @xcite .
if we adopt the configuration model as the null model in our algorithm , the quality function will be given by @xmath108 .
we have used the random graph because the maximisation of @xmath109 does not yield the true core - periphery structure for synthetic networks having a single core - periphery pair .
this is because @xmath109 involves @xmath110 , which makes @xmath109 small when node @xmath5 and @xmath6 have a large degree . because core nodes tend to have a large degree , maximisation of @xmath109 tends to assign core nodes that should belong to the same core - periphery pair to different core - periphery pairs ( i.e. , @xmath111 ) or a periphery ( i.e. , @xmath112 ) . to support this argument , we maximise @xmath109 using a label switching heuristic ( section [ sec : algorithm ] ) for the synthetic networks with a single core - periphery pair ( fig . [ fig : sb]a ) .
the obtained vi values are larger than 0.5 in the entire range of @xmath78 ( fig .
[ fig : conf ] ) .
this result contrasts to that obtained from our algorithm ( with the random graph null model ) , which is shown by the rectangles in fig .
[ fig : exp ] . previous studies provided algorithms to detect multiple core - periphery pairs based on community detection algorithms .
yang and leskovec used an algorithm for detecting overlapping communities in networks @xcite .
they regarded the nodes belonging to many communities as core nodes and nodes belonging to few communities as peripheral nodes .
these algorithms may detect densely interconnected peripheral nodes because the detected peripheral nodes in a single core - periphery pair belong to the same community .
in addition , a periphery may belong to multiple cores in these algorithms .
these properties are shared by the algorithms proposed in refs .
@xcite . in contrast
, our algorithm detects disjoint core - periphery pairs such that peripheral nodes are interconnected sparsely within each core - periphery pair and across different core - periphery pairs .
yan and luo focussed on a different type of structure consisting of multiple cores and a single periphery @xcite .
in contrast , a core detected by our algorithm owns its exclusive periphery , including the case of an empty periphery . 0.45 and zero otherwise .
the empty blocks correspond to the entries that are equal to one with probability 0.01 and zero otherwise .
the diagonal entries are always set to zero and shown as empty entries in the figure for the sake of simplicity .
the dashed lines indicate the borders separating different blocks .
the labels @xmath113 are also indicated at the top and left of the adjacency matrices .
label r represents a block of residual nodes .
the networks are composed of ( * a * ) a single core - periphery pair , ( * b * ) two core - periphery pairs , ( * c * ) a single core - periphery pair and residual nodes , and ( * d * ) two core - periphery pairs and residual nodes . in all cases , we set @xmath114 .
, title="fig : " ] 0.45 and zero otherwise .
the empty blocks correspond to the entries that are equal to one with probability 0.01 and zero otherwise .
the diagonal entries are always set to zero and shown as empty entries in the figure for the sake of simplicity .
the dashed lines indicate the borders separating different blocks .
the labels @xmath113 are also indicated at the top and left of the adjacency matrices .
label r represents a block of residual nodes .
the networks are composed of ( * a * ) a single core - periphery pair , ( * b * ) two core - periphery pairs , ( * c * ) a single core - periphery pair and residual nodes , and ( * d * ) two core - periphery pairs and residual nodes . in all cases ,
we set @xmath114 . ,
title="fig : " ] 0.45 and zero otherwise .
the empty blocks correspond to the entries that are equal to one with probability 0.01 and zero otherwise .
the diagonal entries are always set to zero and shown as empty entries in the figure for the sake of simplicity .
the dashed lines indicate the borders separating different blocks .
the labels @xmath113 are also indicated at the top and left of the adjacency matrices .
label r represents a block of residual nodes .
the networks are composed of ( * a * ) a single core - periphery pair , ( * b * ) two core - periphery pairs , ( * c * ) a single core - periphery pair and residual nodes , and ( * d * ) two core - periphery pairs and residual nodes . in all cases ,
we set @xmath114 . ,
title="fig : " ] 0.45 and zero otherwise .
the empty blocks correspond to the entries that are equal to one with probability 0.01 and zero otherwise .
the diagonal entries are always set to zero and shown as empty entries in the figure for the sake of simplicity .
the dashed lines indicate the borders separating different blocks .
the labels @xmath113 are also indicated at the top and left of the adjacency matrices .
label r represents a block of residual nodes .
the networks are composed of ( * a * ) a single core - periphery pair , ( * b * ) two core - periphery pairs , ( * c * ) a single core - periphery pair and residual nodes , and ( * d * ) two core - periphery pairs and residual nodes . in all cases , we set @xmath114 .
, title="fig : " ] 0.5 and @xmath7 , respectively .
the solid lines indicate the partition into core - periphery pairs .
the dashed lines indicate the partition into the core and periphery within a core - periphery pair .
the colour indicates the leaning of the members .
the instructor ( i.e. , node 1 ) and president ( i.e. , node 34 ) are indicated by the arrows .
, title="fig : " ] 0.5 and @xmath7 , respectively .
the solid lines indicate the partition into core - periphery pairs .
the dashed lines indicate the partition into the core and periphery within a core - periphery pair . the colour indicates the leaning of the members .
the instructor ( i.e. , node 1 ) and president ( i.e. , node 34 ) are indicated by the arrows .
, title="fig : " ] 0.5 and @xmath7 , respectively .
the solid lines indicate the partition into core - periphery pairs .
the dashed lines indicate the partition into the core and periphery within a core - periphery pair .
the colour indicates the leaning of the members .
the instructor ( i.e. , node 1 ) and president ( i.e. , node 34 ) are indicated by the arrows .
, title="fig : " ] 0.5 and @xmath7 , respectively .
the solid lines indicate the partition into core - periphery pairs .
the dashed lines indicate the partition into the core and periphery within a core - periphery pair .
the colour indicates the leaning of the members .
the instructor ( i.e. , node 1 ) and president ( i.e. , node 34 ) are indicated by the arrows .
, title="fig : " ] .
the average cpu time of the three algorithms on different networks .
we generate synthetic networks 14 using the stochastic block models schematically shown in figs .
[ fig : sb]a , [ fig : sb]b , [ fig : sb]c and [ fig : sb]d , respectively . for each of them , we set @xmath117 and measure the cpu time for one generated network .
[ cols="<,^,^,^,^,^ " , ] | with a core - periphery structure of networks , core nodes are densely interconnected , peripheral nodes are connected to core nodes to different extents , and peripheral nodes are sparsely interconnected .
core - periphery structure composed of a single core and periphery has been identified for various networks .
however , analogous to the observation that many empirical networks are composed of densely interconnected groups of nodes , i.e. , communities , a network may be better regarded as a collection of multiple cores and peripheries .
we propose a scalable algorithm to detect multiple non - overlapping groups of core - periphery structure in a network .
we illustrate our algorithm using synthesised and empirical networks .
for example , we find distinct core - periphery pairs with different political leanings in a network of political blogs and separation between international and domestic subnetworks of airports in some single countries in a world - wide airport network . |
the _ teichmller flow _ @xmath2 acts on components of strata in the moduli space of area one abelian or quadratic differentials for a non - exceptional surface @xmath1 of finite type .
this flow has many properties which resemble the properties of an anosov flow .
for example , there is a pair of transverse invariant foliations , and there is an invariant mixing borel probability measure @xmath3 in the lebesgue measure class which is absolutely continuous with respect to these foliations , with conditional measures which are uniformly expanded and contracted by the flow @xcite .
this measure is even exponentially mixing , i.e. exponential decay of correlations for hlder observables holds true @xcite .
the entropy @xmath4 of the lebesgue measure @xmath3 is the supremum of the topological entropies of the restriction of @xmath2 to compact invariant sets @xcite . for strata of abelian differentials
, @xmath3 is the unique invariant measure of maximal entropy @xcite .
the goal of this note is to extend further the analogy between the teichmller flow on components of strata and anosov flows .
an anosov flow @xmath5 on a compact manifold @xmath6 admits a unique borel probability measure @xmath7 of maximal entropy .
this measure can be obtained as follows @xcite .
every periodic orbit @xmath8 of @xmath5 of prime period @xmath9 supports a unique @xmath5-invariant borel measure @xmath10 of total mass @xmath11 .
if @xmath12 is the topological entropy of @xmath5 then @xmath7 is the ( unique ) weak limit of the sequence of measures @xmath13 as @xmath14 . in particular
, the number of periodic orbits of period at most @xmath15 is asymptotic to @xmath16 as @xmath14 . for any connected component @xmath0 of a stratum of abelian or quadratic differentials the @xmath2-invariant lebesgue measure @xmath3 on @xmath0
can be obtained in the same way . for a precise formulation ,
we say that a family @xmath17 of finite borel measures on the moduli space @xmath18 of area one abelian differentials or on the moduli space @xmath19 of area one quadratic differentials _ converges weakly _ to @xmath3 if for every continuous function @xmath20 on @xmath18 or on @xmath19 with compact support we have @xmath21 let @xmath22 be the set of all periodic orbits for @xmath2 contained in @xmath0 .
for @xmath23 let @xmath9 be the prime period of @xmath8 and denote by @xmath10 the @xmath2-invariant lebesgue measure on @xmath8 of total mass @xmath11 .
we show for every component @xmath0 of a stratum in the moduli space of abelian or quadratic differentials the measures @xmath24 converge as @xmath14 weakly to the lebesgue measure on @xmath0 .
the theorem implies that as @xmath14 , the number of periodic orbits in @xmath0 of period at most @xmath15 is asymptotically not smaller than @xmath16 .
however , since the closure in @xmath19 of a component @xmath0 of a stratum is non - compact , we do not obtain a precise asymptotic growth rate for all periodic orbits in @xmath0 .
namely , there may be a set of periodic orbits in @xmath0 whose growth rate exceeds @xmath4 and which eventually exit every compact subset of @xmath19 .
for periodic orbits in the open principal stratum , eskin and mirzakhani @xcite showed that the asymptotic growth rate of periodic orbits for the teichmller flow which lie deeply in the cusp of moduli space is strictly smaller than the entropy @xmath4 , and they calculate the asymptotic growth rate of all periodic orbits .
eskin , mirzakhani and rafi @xcite also announced the analogous result for any component of any stratum .
the proof of the above theorem uses ideas which were developed by margulis for hyperbolic flows ( see @xcite for an account with comments ) .
this strategy is by now standard , and the main task is to overcome the difficulty of absence of hyperbolicity for the teichmller flow in the thin part of moduli space and the absence of nice product coordinates near a boundary point of a stratum .
absence of hyperbolicity in the thin part of moduli space is dealt with using the curve graph similar to the strategy developed in @xcite .
integration of the hodge norm as discussed in @xcite and some standard ergodic theory is also used .
relative homology coordinates @xcite define local product structures for strata .
these coordinates do no extend in a straightforward way to points in the boundary of the stratum . in the case of the principal stratum
, however , product coordinates about boundary points can be obtained by simply writing a quadratic differential as a pair of its vertical and horizontal measured geodesic lamination .
our approach is to show that there is a similar picture for strata . to this end
, we use coordinates for strata based on train tracks which will be used in other contexts as well .
the construction of these coordinates is carried out in sections 3 and 4 .
the tools developed in sections 3 and 4 are used in section 5 to show that a weak limit @xmath7 of the measures @xmath25 is absolutely continuous with respect to the lebesgue measure , with radon nikodym derivative bounded from above by one . in section 6 the proof of the theorem
is completed .
section 2 summarizes some properties of the curve graph and geodesic laminations used throughout the paper .
let @xmath1 be an oriented surface of finite type , i.e. @xmath1 is a closed surface of genus @xmath26 from which @xmath27 points , so - called _ punctures _ , have been deleted .
we assume that @xmath28 , i.e. that @xmath1 is not a sphere with at most four punctures or a torus with at most one puncture .
the _ teichmller space _ @xmath29 of @xmath1 is the quotient of the space of all complete finite volume hyperbolic metrics on @xmath1 under the action of the group of diffeomorphisms of @xmath1 which are isotopic to the identity .
the fibre bundle @xmath30 over @xmath29 of all _ marked holomorphic quadratic differentials _ of area one can be viewed as the unit cotangent bundle of @xmath29 for the _ teichmller metric _ @xmath31 .
we assume that each quadratic differential @xmath32 has a pole of first order at each of the punctures , i.e. we include the information on the number of poles of the differential in the number of punctures of @xmath1 . the _ teichmller flow _
@xmath2 on @xmath30 commutes with the action of the _ mapping class group _ @xmath33 of all isotopy classes of orientation preserving self - homeomorphisms of @xmath1 .
therefore this flow descends to a flow on the quotient orbifold @xmath34 , again denoted by @xmath2 .
a _ geodesic lamination _ for a complete hyperbolic structure on @xmath1 of finite volume is a _ compact _ subset of @xmath1 which is foliated into simple geodesics .
a geodesic lamination @xmath35 is called _ minimal _ if each of its half - leaves is dense in @xmath35 .
thus a simple closed geodesic is a minimal geodesic lamination .
a minimal geodesic lamination with more than one leaf has uncountably many leaves and is called _ minimal arational_. every geodesic lamination @xmath35 consists of a disjoint union of finitely many minimal components and a finite number of isolated leaves .
each of the isolated leaves of @xmath35 either is an isolated closed geodesic and hence a minimal component , or it _ spirals _ about one or two minimal components .
a geodesic lamination @xmath35 _ fills up _
@xmath1 if its complementary components are topological discs or once punctured monogons , i.e. once punctured discs bounded by a single leaf of @xmath35
. the set @xmath36 of all geodesic laminations on @xmath1 can be equipped with the restriction of the _ hausdorff topology _ for compact subsets of @xmath1 . with respect to this topology ,
the space @xmath36 is compact .
the projectivized tangent bundle @xmath37 of a geodesic lamination @xmath35 is a compact subset of the projectivized tangent bundle @xmath38 of @xmath1 .
the geodesic lamination @xmath35 is _ orientable _ if there is an continuous orientation of the tangent bundle of @xmath35 .
this is equivalent to stating that there is a continuous section @xmath39 where @xmath40 denotes the unit tangent bundle of @xmath1 .
[ largelam ] a _ large geodesic lamination _ is a geodesic lamination @xmath35 which fills up @xmath1 and can be approximated in the hausdorff topology by simple closed geodesics .
note that a minimal geodesic lamination @xmath35 can be approximated in the hausdorff topology by simple closed geodesics and hence if @xmath35 fills up @xmath1 then @xmath35 is large .
moreover , the set of all large geodesic laminations is closed with respect to the hausdorff topology and hence it is compact .
the _ topological type _ of a large geodesic lamination @xmath35 is a tuple @xmath41 such that the complementary components of @xmath35 which are topological discs are @xmath42-gons .
let @xmath43 be the space of all large geodesic laminations of type @xmath44 equipped with the restriction of the hausdorff topology for compact subsets of @xmath1 .
a geodesic lamination is called _ complete _ if it is large of type @xmath45 .
the complementary components of a complete geodesic lamination are all trigons or once punctured monogons .
a _ measured geodesic lamination _ is a geodesic lamination @xmath35 equipped with a translation invariant transverse measure @xmath46 such that the @xmath46-weight of every compact arc in @xmath1 with endpoints in @xmath47 which intersects @xmath35 nontrivially and transversely is positive .
we say that @xmath35 is the _ support _ of the measured geodesic lamination .
the geodesic lamination @xmath35 is _ uniquely ergodic _ if @xmath46 is the only transverse measure with support @xmath35 up to scale .
the space @xmath48 of measured geodesic laminations equipped with the weak@xmath49-topology admits a natural continuous action of the multiplicative group @xmath50 .
the quotient under this action is the space @xmath51 of _ projective measured geodesic laminations _ which is homeomorphic to the sphere @xmath52 .
every simple closed geodesic @xmath53 on @xmath1 defines a measured geodesic lamination .
the geometric intersection number between simple closed curves on @xmath1 extends to a continuous function @xmath54 on @xmath55 , the _ intersection form_. we say that a pair @xmath56 of measured geodesic laminations _
jointly fills up _
@xmath1 if for every measured geodesic lamination @xmath57 we have @xmath58 . this is equivalent to stating that every complete simple ( possibly infinite ) geodesic on @xmath1 intersects either the support of @xmath46 or the support of @xmath7 transversely .
the _ curve graph _
@xmath59 of @xmath1 is the locally infinite metric graph whose vertices are the free homotopy classes of essential simple closed curves on @xmath1 , i.e. curves which are neither contractible nor freely homotopic into a puncture .
two such curves are connected by an edge of length one if and only if they can be realized disjointly .
the mapping class group @xmath33 of @xmath1 acts on @xmath59 as a group of simplicial isometries .
the curve graph @xmath59 is a hyperbolic geodesic metric space @xcite and hence it admits a _ gromov boundary _ @xmath60 .
for @xmath61 there is a complete distance function @xmath62 on @xmath63 of uniformly bounded diameter , and there is a number @xmath64 such that @xmath65 the group @xmath33 acts on @xmath60 as a group of homeomorphisms .
let @xmath66 be a _ bers constant _ for @xmath1 , i.e. @xmath67 is such that for every complete hyperbolic metric on @xmath1 of finite volume there is a pants decomposition of @xmath1 consisting of pants curves of length at most @xmath67 . define a map @xmath68 by associating to @xmath69 a simple closed curve of @xmath70-length at most @xmath67 .
then there is a number @xmath71 such that @xmath72 for all @xmath73 ( see the discussion in @xcite ) . for a number @xmath74 , a map @xmath75 @xmath76)$ ]
is an _ @xmath77-quasi - geodesic _ if for all @xmath78 we have @xmath79 a map @xmath80 is called an _
unparametrized @xmath77-quasi - geodesic _ if there is an increasing homeomorphism @xmath81 @xmath76)$ ] such that @xmath82 is an @xmath77-quasi - geodesic .
we say that an unparametrized quasi - geodesic is _ infinite _ if its image set has infinite diameter . there is a number @xmath83 such that the image under @xmath84 of every teichmller geodesic is an unparametrized @xmath85-quasi - geodesic @xcite .
choose a smooth function @xmath86 $ ] with @xmath87\equiv 1 $ ] and @xmath88 .
for each @xmath69 , the number of essential simple closed curves @xmath53 on @xmath1 whose @xmath70-length @xmath89 ( i.e. the length of a geodesic representative in its free homotopy class ) does not exceed @xmath90 is bounded from above by a constant not depending on @xmath70 , and the diameter of the subset of @xmath59 containing these curves is uniformly bounded as well .
thus we obtain for every @xmath69 a finite borel measure @xmath91 on @xmath59 by defining @xmath92 where @xmath93 denotes the dirac mass at @xmath53 .
the total mass of @xmath91 is bounded from above and below by a universal positive constant , and the diameter of the support of @xmath91 in @xmath59 is uniformly bounded as well . moreover , the measures @xmath91 depend continuously on @xmath69 in the weak@xmath49-topology .
this means that for every bounded function @xmath94 the function @xmath95 is continuous .
for @xmath69 define a distance @xmath96 on @xmath60 by @xmath97 the distances @xmath96 are equivariant with respect to the action of @xmath33 on @xmath29 and @xmath63 .
moreover , there is a constant @xmath98 such that @xmath99 for all @xmath73 ( see p.230 and p.231 of @xcite ) .
an area one quadratic differential @xmath100 is determined by a pair @xmath101 of measured geodesic laminations which jointly fill up @xmath1 and such that @xmath102 .
the laminations @xmath103 are called _ vertical _ and _ horizontal _ , respectively . for @xmath100
let @xmath104 be the set of all quadratic differentials whose horizontal projective measured geodesic lamination coincides with the horizontal projective measured geodesic lamination of @xmath105 .
the space @xmath106 is called the _ unstable _ manifold of @xmath105 , and these unstable manifolds define the _ unstable foliation _ @xmath107 of @xmath30 .
the _ strong unstable manifold _
@xmath108 is the set of all quadratic differentials whose horizontal measured geodesic lamination coincides with the horizontal measured geodesic lamination of @xmath105 .
these sets define the _ strong unstable _
foliation @xmath109 of @xmath30 .
the image of the unstable ( or the strong unstable ) foliation of @xmath30 under the flip @xmath110 is the _ stable _
foliation @xmath111 ( or the _ strong stable foliation _
@xmath112 ) . by the hubbard - masur theorem , for each @xmath100 the restriction to @xmath113 of the canonical projection
@xmath114 is a homeomorphism .
thus the teichmller metric lifts to a complete distance function @xmath115 on @xmath106 .
denote by @xmath116 the restriction of this distance function to @xmath117 . then @xmath118 are distance functions on the leaves of the stable and strong stable foliation , respectively .
for @xmath100 and @xmath119 let moreover @xmath120 be the closed ball of radius @xmath121 about @xmath105 with respect to @xmath122 @xmath123 .
let @xmath124 be the set of all marked quadratic differentials @xmath125 such that the unparametrized quasi - geodesic @xmath126 @xmath127 is infinite .
then @xmath128 is the set of all quadratic differentials whose vertical measured geodesic lamination fills up @xmath1 ( i.e. its support fills up @xmath1 , see @xcite for a comprehensive discussion of this result of klarreich @xcite ) .
there is a natural @xmath33-equivariant surjective map @xmath129 which associates to a point @xmath130 the endpoint of the infinite unparametrized quasi - geodesic @xmath126 @xmath127 .
call a marked quadratic differential @xmath100 _ uniquely ergodic _ if the support of its vertical measured geodesic lamination is uniquely ergodic and fills up @xmath1 . a uniquely ergodic quadratic differential
is contained in the set @xmath131 @xcite .
we have ( section 3 of @xcite ) [ closed ] 1 .
the map @xmath132 is continuous and closed .
if @xmath100 is uniquely ergodic then the sets @xmath133 @xmath134 form a neighborhood basis for @xmath135 in @xmath63 .
for @xmath136 and @xmath119 let @xmath137 be the closed ball of radius @xmath121 about @xmath135 with respect to the distance function @xmath138 .
as a consequence of lemma [ closed ] , if @xmath139 is uniquely ergodic then for every @xmath119 there are numbers @xmath140 and @xmath141 such that @xmath142
in this section we establish some properties of train tracks on an oriented surface @xmath1 of genus @xmath26 with @xmath27 punctures and @xmath28 which will be used in section 4 to construct coordinates near boundary points of strata . a _ train track _ on @xmath1
is an embedded 1-complex @xmath143 whose edges ( called _ branches _ ) are smooth arcs with well - defined tangent vectors at the endpoints . at any vertex ( called a
_ switch _ ) the incident edges are mutually tangent . through each switch
there is a path of class @xmath144 which is embedded in @xmath145 and contains the switch in its interior .
a simple closed curve component of @xmath145 contains a unique bivalent switch , and all other switches are at least trivalent .
the complementary regions of the train track have negative euler characteristic , which means that they are different from discs with @xmath146 or @xmath147 cusps at the boundary and different from annuli and once - punctured discs with no cusps at the boundary .
we always identify train tracks which are isotopic . throughout
we use the book @xcite as the main reference for train tracks .
a train track is called _ generic _ if all switches are at most trivalent . for each switch @xmath148 of a generic train track @xmath145
which is not contained in a simple closed curve component , there is a unique half - branch @xmath149 of @xmath145 which is incident on @xmath148 and which is _ large _ at @xmath148 .
this means that every germ of an arc of class @xmath144 on @xmath145 which passes through @xmath148 also passes through the interior of @xmath149 .
a half - branch which is not large is called _
small_. a branch @xmath149 of @xmath145 is called _ large _ ( or _ small _ ) if each of its two half - branches is large ( or small ) .
a branch which is neither large nor small is called _
mixed_. as in @xcite , all train tracks are assumed to be generic .
unfortunately this leads to a small inconsistency of our terminology with the terminology found in the literature .
a _ trainpath _ on a train track @xmath145 is a @xmath144-immersion @xmath150\to \tau$ ] such that for every @xmath151 the restriction of @xmath152 to @xmath153 $ ] is a homeomorphism onto a branch of @xmath145 .
more generally , we call a @xmath144-immersion @xmath154\to \tau$ ] a _ generalized trainpath_. a trainpath @xmath150\to \tau$ ] is _ closed _
if @xmath155 and if the extension @xmath156 defined by @xmath157 @xmath158)$ ] and @xmath159 @xmath160)$ ] is a trainpath . a generic train track @xmath145 is _ orientable _ if there is a consistent orientation of the branches of @xmath145 such that at any switch @xmath161 of @xmath145 , the orientation of the large half - branch incident on @xmath161 extends to the orientation of the two small half - branches incident on @xmath161 .
if @xmath162 is a complementary polygon of an oriented train track then the number of sides of @xmath162 is even . in particular , a train track which contains a once punctured monogon component , i.e. a once punctured disc with one cusp at the boundary , is not orientable ( see p.31 of @xcite for a more detailed discussion ) .
a train track or a geodesic lamination @xmath163 is _ carried _ by a train track @xmath145 if there is a map @xmath164 of class @xmath144 which is homotopic to the identity and maps @xmath163 into @xmath145 in such a way that the restriction of the differential of @xmath165 to the tangent space of @xmath163 vanishes nowhere ; note that this makes sense since a train track has a tangent line everywhere .
we call the restriction of @xmath165 to @xmath163 a _ carrying map _ for @xmath163 . write @xmath166 if the train track @xmath163 is carried by the train track @xmath145
. then every geodesic lamination @xmath35 which is carried by @xmath163 is also carried by @xmath145 .
a train track _ fills up _
@xmath1 if its complementary components are topological discs or once punctured monogons .
note that such a train track @xmath145 is connected .
let @xmath167 be the number of those complementary components of @xmath145 which are topological discs .
each of these discs is an @xmath42-gon for some @xmath168 @xmath169 .
the _ topological type _ of @xmath145 is defined to be the ordered tuple @xmath44 where @xmath170 ; then @xmath171 . if @xmath145 is orientable then @xmath172 and @xmath173 is even for all @xmath174 . a train track of topological type @xmath45 is called _
maximal_. the complementary components of a maximal train track are all trigons , i.e. topological discs with three cusps at the boundary , or once punctured monogons .
a _ transverse measure _ on a generic train track @xmath145 is a nonnegative weight function @xmath7 on the branches of @xmath145 satisfying the _ switch condition _ :
for every trivalent switch @xmath161 of @xmath145 , the sum of the weights of the two small half - branches incident on @xmath161 equals the weight of the large half - branch .
the space @xmath175 of all transverse measures on @xmath145 has the structure of a cone in a finite dimensional real vector space , and it is naturally homeomorphic to the space of all measured geodesic laminations whose support is carried by @xmath145 .
the train track is called _ recurrent _ if it admits a transverse measure which is positive on every branch .
we call such a transverse measure @xmath7 _ positive _ , and we write @xmath176 ( see @xcite for more details )
. a _ subtrack _ @xmath177 of a train track @xmath145 is a subset of @xmath145 which is itself a train track .
then @xmath177 is obtained from @xmath145 by removing some of the branches , and we write @xmath178 . if @xmath149 is a small branch of @xmath145 which is incident on two distinct switches of @xmath145 then the graph @xmath177 obtained from @xmath145 by removing @xmath149 is a subtrack of @xmath145 .
we then call @xmath145 a _ simple extension _ of @xmath177 . note that formally to obtain the subtrack @xmath179 from @xmath180 we may have to delete the switches on which the branch @xmath149 is incident .
[ simpleex ] 1 . a simple extension @xmath145 of a recurrent non - orientable connected train track @xmath177 is recurrent .
moreover , @xmath181 2 .
an orientable simple extension @xmath145 of a recurrent orientable connected train track @xmath177 is recurrent .
moreover , @xmath181 if @xmath145 is a simple extension of a train track @xmath177 then @xmath177 can be obtained from @xmath145 by the removal of a small branch @xmath149 which is incident on two distinct switches @xmath182 .
then @xmath183 is an interior point of a branch @xmath184 of @xmath177 @xmath185 .
if @xmath177 is connected , non - orientable and recurrent then there is a trainpath @xmath186\to \tau - b$ ] which begins at @xmath187 , ends at @xmath188 and such that the half - branch @xmath189 $ ] is small at @xmath190 and that the half - branch @xmath191 $ ] is small at @xmath192 .
extend @xmath193 to a closed trainpath @xmath152 on @xmath194 which begins and ends at @xmath187 .
this is possible since @xmath177 is non - orientable , connected and recurrent .
there is a closed trainpath @xmath195\to \tau$ ] which can be obtained from @xmath152 by replacing the trainpath @xmath193 by the branch @xmath149 traveled through from @xmath187 to @xmath188 .
the counting measure of @xmath156 on @xmath145 satisfies the switch condition and hence it defines a transverse measure on @xmath145 which is positive on @xmath149 . on the other hand ,
every transverse measure on @xmath177 defines a transverse measure on @xmath145 .
thus since @xmath177 is recurrent and since the sum of two transverse measures on @xmath145 is again a transverse measure , the train track @xmath145 is recurrent as well .
moreover , we have @xmath196 . let @xmath85 be the number of branches of @xmath145 .
label the branches of @xmath145 with the numbers @xmath197 so that the number @xmath85 is assigned to @xmath149 .
let @xmath198 be the standard basis of @xmath199 and define a linear map @xmath200 by @xmath201 for @xmath202 and @xmath203 where @xmath35 is the weight function on @xmath197 defined by the trainpath @xmath193 .
the map @xmath204 is a surjection onto a linear subspace of @xmath199 of codimension one , moreover @xmath204 preserves the linear subspace @xmath205 of @xmath199 defined by the switch conditions for @xmath145 . in particular ,
the corank of @xmath206 is at most one .
the image under @xmath204 of the cone of all nonnegative weights on the branches of @xmath145 satisfying the switch conditions is contained in the cone of all nonnegative weights on @xmath207 satisfying the switch conditions for @xmath177 .
therefore the dimension of the space of transverse measures on @xmath177 equals the space of transverse measures on @xmath145 minus one .
this implies @xmath208 and completes the proof of the first part of the lemma .
the second part follows in exactly the same way . as a consequence
we obtain [ dimensioncount ] 1 .
@xmath209 for every non - orientable recurrent train track @xmath145 of topological type @xmath44 .
2 . @xmath210 for every orientable recurrent train track @xmath145 of topological type @xmath211 .
the disc components of a non - orientable recurrent train track @xmath145 of topological type @xmath44 can be subdivided in @xmath212 steps into trigons by successively adding small branches .
a successive application of lemma [ simpleex ] shows that the resulting train track @xmath163 is maximal and recurrent .
since for every maximal recurrent train track @xmath163 we have @xmath213 ( see @xcite ) , the first part of the corollary follows . to show the second part ,
let @xmath145 be an orientable recurrent train track of type @xmath211 .
then @xmath173 is even for all @xmath174 .
add a branch @xmath214 to @xmath145 which cuts some complementary component of @xmath145 into a trigon and a second polygon with an odd number of sides .
the resulting train track @xmath215 is not recurrent since a trainpath on @xmath215 can only pass through @xmath214 at most once .
however , we can add to @xmath215 another small branch @xmath216 which cuts some complementary component of @xmath215 with at least 4 sides into a trigon and a second polygon such that the resulting train track @xmath163 is non - orientable and recurrent .
the inward pointing tangent of @xmath216 is chosen in such a way that there is a trainpath traveling both through @xmath214 and @xmath216 .
the counting measure of any simple closed curve which is carried by @xmath163 gives equal weight to the branches @xmath214 and @xmath216 .
but this just means that @xmath217 ( see the proof of lemma [ simpleex ] for a detailed argument ) . by the first part of the corollary
, we have @xmath218 which completes the proof .
[ deffullyrec ] a train track @xmath145 of topological type @xmath44 is _ fully recurrent _ if @xmath145 carries a large minimal geodesic lamination @xmath219 .
note that by definition , a fully recurrent train track is connected and fills up @xmath1 .
the next lemma gives some first property of a fully recurrent train track @xmath145 .
for its proof , recall that there is a natural homeomorphism of @xmath175 onto the subspace of @xmath48 of all measured geodesic laminations carried by @xmath145 .
[ fullyrec ] a fully recurrent train track @xmath145 of topological type @xmath44 is recurrent .
a fully recurrent train track @xmath145 of type @xmath44 carries a large geodesic lamination @xmath219 .
the carrying map @xmath220 induces a bijection between the complementary components of @xmath145 and the complementary components of @xmath35 .
in particular , a carrying map @xmath220 is necessarily surjective .
the third paragraph in the proof of lemma 2.3 of @xcite now shows that @xmath145 is recurrent .
there are two simple ways to modify a fully recurrent train track @xmath145 to another fully recurrent train track .
namely , if @xmath149 is a mixed branch of @xmath145 then we can _ shift _
@xmath145 along @xmath149 to a new train track @xmath221 .
this new train track carries @xmath145 and hence it is fully recurrent since it carries every geodesic lamination which is carried by @xmath145 @xcite . similarly , if @xmath222 is a large branch of @xmath145 then we can perform a right or left _ split _ of @xmath145 at @xmath222 as shown in figure a below .
a ( right or left ) split @xmath221 of a train track @xmath145 is carried by @xmath145 .
if @xmath145 is of topological type @xmath44 , if @xmath219 is minimal and is carried by @xmath145 and if @xmath222 is a large branch of @xmath145 , then there is a unique choice of a right or left split of @xmath145 at @xmath222 such that the split track @xmath163 carries @xmath35 . in particular
, @xmath163 is fully recurrent .
note however that there may be a split of @xmath145 at @xmath222 such that the split track is not fully recurrent any more ( see section 2 of @xcite for details ) .
the following simple observation is used to identify fully recurrent train tracks . [ nosplitor ] 1 .
let @xmath222 be a large branch of a fully recurrent non - orientable train track @xmath145 .
then no component of the train track @xmath177 obtained from @xmath145 by splitting @xmath145 at @xmath222 and removing the diagonal of the split is orientable .
2 . let @xmath222 be a large branch of a fully recurrent orientable train track @xmath145 .
then the train track @xmath177 obtained from @xmath145 by splitting @xmath145 at @xmath222 and removing the diagonal of the split is connected .
let @xmath145 be a fully recurrent non - orientable train track of topological type @xmath44 .
let @xmath222 be a large branch of @xmath145 and let @xmath148 be a switch on which the branch @xmath222 is incident .
let @xmath177 be the train track obtained from @xmath145 by splitting @xmath145 at @xmath222 and removing the diagonal branch of the split .
then the train tracks @xmath223 obtained from @xmath145 by a right and left split at @xmath222 , respectively , are simple extensions of @xmath177 .
if @xmath177 is connected and orientable then the train tracks @xmath223 are not recurrent since no transverse measure can give positive weight to the diagonal of the split ( compare the discussion in the proof of lemma [ simpleex ] ) . however , since @xmath145 is fully recurrent , it can be split at @xmath222 to a fully recurrent and hence recurrent train track .
this is a contradiction .
now assume that @xmath177 is disconnected and contains an orientable connected component @xmath224 .
let @xmath225 be a diagonal of the split connecting @xmath145 to @xmath226 @xmath185 .
if @xmath227\to \tau_i$ ] is a trainpath with @xmath228=b_i$ ] and @xmath229\in \sigma_1 $ ] then @xmath230\subset \sigma_1 $ ] and hence once again , @xmath226 is not recurrent . as above , this contradicts the assumption that @xmath145 is fully recurrent .
the first part of the corollary is proven .
the second part follows from the same argument since a split of an orientable train track is orientable .
* example : * 1 ) figure b below shows a non - orientable recurrent train track @xmath145 of type @xmath231 on a closed surface of genus two .
the train track obtained from @xmath145 by a split at the large branch @xmath222 and removal of the diagonal of the split track is orientable and hence @xmath145 is not fully recurrent .
this corresponds to the fact established by masur and smillie @xcite that every quadratic differential with a single zero and no pole on a surface of genus @xmath147 is the square of a holomorphic one - form ( see section 4 for more information ) .
\2 ) to construct an orientable recurrent train track of type @xmath211 which is not fully recurrent let @xmath232 be a surface of genus @xmath233 and let @xmath234 be an orientable fully recurrent train track on @xmath232 with @xmath235 complementary components . choose a complementary component @xmath236 of @xmath234 in @xmath232 , remove from @xmath236 a disc @xmath237 and glue two copies of @xmath238 along the boundary of @xmath237 to a surface @xmath1 of genus @xmath239 .
the two copies of @xmath234 define a recurrent disconnected oriented train track @xmath145 on @xmath1 which has an annulus complementary component @xmath162 . choose a branch @xmath216 of @xmath145 in the boundary of @xmath162 .
there is a corresponding branch @xmath240 in the second boundary component of @xmath162 .
glue a compact subarc of @xmath216 contained in the interior of @xmath216 to a compact subarc of @xmath240 contained in the interior of @xmath240 so that the images of the two arcs under the glueing form a large branch @xmath222 in the resulting train track @xmath163 .
the train track @xmath163 is recurrent and orientable , and its complementary components are topological discs .
however , by lemma [ nosplitor ] it is not fully recurrent . to each train track @xmath145 which fills up
@xmath1 one can associate a _
dual bigon track _
@xmath241 ( section 3.4 of @xcite ) .
there is a bijection between the complementary components of @xmath145 and those complementary components of @xmath241 which are not _ bigons _ , i.e. discs with two cusps at the boundary .
this bijection maps a component @xmath162 of @xmath145 which is an @xmath242-gon for some @xmath243 to an @xmath242-gon component of @xmath241 contained in @xmath162 , and it maps a once punctured monogon @xmath162 to a once punctured monogon contained in @xmath162 . if @xmath145 is orientable then the orientation of @xmath1 and an orientation of @xmath145 induce an orientation on @xmath241 , i.e. @xmath241 is orientable .
measured geodesic laminations which are carried by @xmath241 can be described as follows .
a _ tangential measure _ on a train track @xmath145 of type @xmath44 assigns to a branch @xmath149 of @xmath145 a weight @xmath244 such that for every complementary @xmath245-gon of @xmath145 with consecutive sides @xmath246 and total mass @xmath247 ( counted with multiplicities ) the following holds true .
1 . @xmath248 .
2 . @xmath249 , @xmath250 .
( the complementary once punctured monogons define no constraint on tangential measures ) .
the space of all tangential measures on @xmath145 has the structure of a convex cone in a finite dimensional real vector space . by the results from section 3.4 of @xcite ,
every tangential measure on @xmath145 determines a simplex of measured geodesic laminations which _ hit @xmath145 efficiently_. the supports of these measured geodesic laminations are carried by the bigon track @xmath241 , and every measured geodesic lamination which is carried by @xmath241 can be obtained in this way .
the dimension of this simplex equals the number of complementary components of @xmath145 with an even number of sides .
the train track @xmath145 is called _ transversely recurrent _ if it admits a tangential measure which is positive on every branch . in general ,
there are many tangential measures which correspond to a fixed measured geodesic lamination @xmath35 which hits @xmath145 efficiently .
namely , let @xmath161 be a switch of @xmath145 and let @xmath251 be the half - branches of @xmath145 incident on @xmath161 and such that the half - branch @xmath252 is large
. if @xmath253 is a tangential measure on @xmath145 which determines the measured geodesic lamination @xmath35 then it may be possible to drag the switch @xmath161 across some of the leaves of @xmath35 and modify the tangential measure @xmath253 on @xmath145 to a tangential measure @xmath254 .
then @xmath255 is a multiple of a vector of the form @xmath256 where @xmath257 denotes the function on the branches of @xmath145 defined by @xmath258 and @xmath259 for @xmath260 .
[ large ] a train track @xmath145 of topological type @xmath44 is called _ fully transversely recurrent _ if its dual bigon track @xmath241 carries a large minimal geodesic lamination @xmath219 .
a train track @xmath145 of topological type @xmath44 is called _ large _ if @xmath145 is fully recurrent and fully transversely recurrent .
a large train track of type @xmath45 is called _
complete_. for a large train track @xmath145 let @xmath261 be the set of all measured geodesic laminations whose support is carried by @xmath241 .
each of these measured geodesic laminations corresponds to a tangential measure on @xmath145 . with this identification ,
the pairing @xmath262 is just the restriction of the intersection form on measured lamination space ( section 3.4 of @xcite ) .
moreover , @xmath263 is naturally homeomorphic to a convex cone in a real vector space .
the dimension of this cone coincides with the dimension of @xmath175 .
denote by @xmath264 the set of all isotopy classes of large train tracks on @xmath1 of type @xmath44 .
* remark : * in @xcite , masur and minsky define a large train track to be a train track @xmath145 whose complementary components are topological discs or once punctured monogons , without the requirement that @xmath145 is generic , transversely recurrent or recurrent .
we hope that this inconsistency of terminology does not lead to any confusion .
as in section 2 , for a closed oriented surface @xmath1 of genus @xmath26 with @xmath27 punctures let @xmath30 be the bundle of marked area one holomorphic quadratic differentials with a simple pole at each puncture over the teichmller space @xmath29 of marked complex structures on @xmath1 . for a complete hyperbolic metric on @xmath1 of finite area , an area one quadratic differential @xmath32
is determined by a pair @xmath265 of measured geodesic laminations which jointly fill up @xmath1 and such that @xmath266 .
the _ vertical _ measured geodesic lamination @xmath267 for @xmath125 corresponds to the equivalence class of the vertical measured foliation of @xmath125 .
the _ horizontal _ measured geodesic lamination @xmath268 for @xmath125 corresponds to the equivalence class of the horizontal measured foliation of @xmath125 .
a tuple @xmath269 of positive integers @xmath170 with @xmath171 defines a _ stratum _ @xmath270 in @xmath30 .
this stratum consists of all marked area one quadratic differentials with @xmath271 simple poles and @xmath272 zeros of order @xmath273 which are not squares of holomorphic one - forms .
the stratum is a real hypersurface in a complex manifold of dimension @xmath274 the closure in @xmath30 of a stratum is a union of components of strata .
strata are invariant under the action of the mapping class group @xmath33 of @xmath1 and hence they project to strata in the moduli space @xmath275 of quadratic differentials on @xmath1 with a simple pole at each puncture .
we denote the projection of the stratum @xmath270 by @xmath276 .
the strata in moduli space need not be connected , but their connected components have been identified by lanneau @xcite .
a stratum in @xmath19 has at most two connected components .
similarly , if @xmath172 then we let @xmath277 be the bundle of marked area one holomorphic one - forms over teichmller space @xmath29 of @xmath1 . for a tuple @xmath278 of positive integers with @xmath279 ,
the stratum @xmath280 of marked area one holomorphic one - forms on @xmath1 with @xmath272 zeros of order @xmath281 @xmath169 is a real hypersurface in a complex manifold of dimension @xmath282 it projects to a stratum @xmath283 in the moduli space @xmath18 of area one holomorphic one - forms on @xmath1 .
strata of holomorphic one - forms in moduli space need not be connected , but the number of connected components of a stratum is at most three @xcite .
recall from section 2 the definition of the strong stable , the stable , the unstable and the strong unstable foliation @xmath284 of @xmath30 .
let @xmath285 be a component of a stratum @xmath270 of marked quadratic differentials or of a stratum @xmath286 of marked abelian differentials .
using period coordinates , one sees that every @xmath287 has a connected neighborhood @xmath288 in @xmath289 with the following properties @xcite . for @xmath290 let @xmath291
$ ] ( or @xmath292 $ ] ) be the vertical ( or the horizontal ) projective measured geodesic lamination of @xmath293 .
then @xmath294\mid u\in u\}$ ] is homeomorphic to an open ball in @xmath295 ( where @xmath12 is as in equation ( [ h],[h2 ] ) ) .
moreover , for @xmath296 the set @xmath297=[q^v]\}=w^s_{\tilde{{\mathcal{q}}},{\rm loc}}(q)\subset w^s(q)\ ] ] is a smooth connected local submanifold of @xmath288 of ( real ) dimension @xmath4 which is called the _ local stable manifold _ of @xmath125 in @xmath289 ( see @xcite ) .
similarly we define the _ local unstable manifold _
@xmath298 of @xmath125 in @xmath289 .
if two such local stable ( or unstable ) manifolds intersect then their union is again a local stable ( or unstable ) manifold .
the maximal connected set containing @xmath125 which is a union of intersecting local stable ( or unstable ) manifolds is the _ stable manifold _ @xmath299 ( or the _ unstable manifold _
@xmath300 ) of @xmath125 in @xmath289 . note that @xmath301 @xmath302 .
a stable ( or unstable ) manifold is invariant under the action of the teichmller flow @xmath2 .
* remark : * there may be a component @xmath289 of a stratum and some @xmath303 such that @xmath304 has infinitely many components .
the ( strong ) stable and ( strong ) unstable manifolds define smooth foliations @xmath305 of @xmath285 which are called the _
stable _ and _ unstable _ foliations of @xmath289 , respectively .
define the _
strong stable _
foliation @xmath306 ( or the _ strong unstable _ foliation @xmath307 ) of @xmath289 by requiring that locally the leaf @xmath308 ( or @xmath309 ) through @xmath125 is the subset of @xmath310 ( or of @xmath300 ) of all marked quadratic differentials whose vertical ( or horizontal ) measured geodesic lamination equals the vertical ( or horizontal ) measured geodesic lamination of @xmath125 .
the strong stable foliation of @xmath289 is transverse to the unstable foliation of @xmath289 .
the foliations @xmath311 @xmath312
are invariant under the action of the stabilizer @xmath313 of @xmath289 in @xmath33 , and they project to @xmath2-invariant singular foliations @xmath314 of @xmath315 . in this technical subsection
we describe for every component @xmath0 of a stratum in the moduli space of quadratic differentials and for every point @xmath316 a basis of neighborhoods of @xmath125 in @xmath0 with local product structures . the material is well known to the experts but a bit difficult to find in the literature . in the course of the discussion we introduce some notations which will be used throughout . for @xmath317 and @xmath318
there is a neighborhood @xmath205 of @xmath319 in @xmath320 and there is a homeomorphism @xmath321 with @xmath322 which is determined by the requirement that @xmath323 .
we call @xmath324 a _ holonomy map _ for the strong unstable foliation along the stable foliation .
similarly , for @xmath317 and @xmath325 there is a neighborhood @xmath326 of @xmath319 in @xmath327 and there is a homeomorphism @xmath328 with @xmath329 which is determined by the requirement that @xmath330 .
we call @xmath331 a _ holonomy map _ for the strong stable foliation along the unstable foliation .
the holonomy maps are equivariant under the action of the mapping class group and hence they project to locally defined holonomy maps in @xmath19 which are denoted by the same symbols .
recall from section 2 the definition of the intrinsic path - metrics @xmath122 on the leaves of the foliation @xmath332 @xmath302 .
these path metrics are invariant under the action of the mapping class group and hence they project to path metrics on the leaves of @xmath332 in @xmath19 which we denote by the same symbols . for @xmath333 and any preimage @xmath319 of @xmath125 in @xmath30 , the distance @xmath334 is the shortest length of a path in @xmath335 connecting @xmath319 to a preimage of @xmath105 .
let moreover @xmath336 be the restrictions of @xmath337 to distances on the leaves of the strong stable and strong unstable foliation of @xmath30 and @xmath19 .
let @xmath338 be the canonical projection .
for @xmath339 and @xmath119 let @xmath340 be the closed ball of radius @xmath121 about @xmath125 in @xmath341 @xmath342 with respect to the metric @xmath122 .
call such a ball @xmath343 a _ metric orbifold ball _ centered at @xmath125 if there is a lift @xmath317 of @xmath125 with the following properties . 1 .
the ball @xmath344 about @xmath319 of the same radius is contractible and precisely invariant under the stabilizer @xmath345 of @xmath319 in @xmath33 .
@xmath346 which means that the restriction of the map @xmath347 to @xmath348 factors through a homeomorphism @xmath349 .
we also say that @xmath350 is an _
orbifold quotient _ of @xmath348 .
note that every metric orbifold ball @xmath351 is contractible .
there is also an obvious notion of an orbifold ball which is not necessarily metric .
for every point @xmath339 there is a number @xmath352 such that the balls @xmath353 are metric orbifold balls @xmath354 and that for any preimage @xmath319 of @xmath125 in @xmath30 and any @xmath355 ( or @xmath356 ) the holonomy map @xmath324 ( or @xmath331 ) is defined on @xmath357 ( or on @xmath358 ) .
now let @xmath359 be borel sets and let @xmath360 , @xmath361 be the preimages of @xmath362 in @xmath363
. then @xmath364 are precisely invariant under @xmath345 .
define @xmath365 note that the map @xmath366 defined by @xmath367 is a homeomorphism , and @xmath368 is homeomorphic to the quotient of @xmath369 under the diagonal action of @xmath345 .
in particular , if @xmath362 are connected then @xmath368 is connected as well .
similarly , define @xmath370 then there is a continuous function @xmath371 which vanishes on @xmath372 and such that @xmath373 in particular , for every number @xmath374 there is a number @xmath375 such that the restriction of the function @xmath177 to @xmath376 assumes values in @xmath377 $ ] . for @xmath378
define @xmath379 then for sufficiently small @xmath380 , say for all @xmath381 , the following properties are satisfied . 1
. @xmath382 is homeomorphic to @xmath383 $ ] .
every connected component of the intersection of an orbit of @xmath2 with @xmath382 is an arc of length @xmath384 .
we call a set @xmath382 as in ( [ localproduct ] ) which satisfies the assumptions a),b ) a set with a _ local product structure_. note that every point @xmath339 has a neighborhood in @xmath19 with a local product structure , e.g. the set @xmath385 for @xmath386 and @xmath387
. moreover , the neighborhoods of @xmath125 with a local product structure form a basis of neighborhoods .
the above discussion can be applied to strata as follows . a connected component @xmath0 of a stratum @xmath276 or of a stratum @xmath388 is locally closed in @xmath19 ( here we identify an abelian differential with its square ) .
this means that for every @xmath316 there exists an open neighborhood @xmath205 of @xmath125 in @xmath19 such that @xmath389 is a closed subset of @xmath205 .
using period coordinates @xcite , one obtains that for every point @xmath316 there is a number @xmath390 and a number @xmath391 with the following property .
for @xmath392 let @xmath393 be the component containing @xmath125 of the intersection @xmath394 ( note that the intersection @xmath395 may not be closed and may have infinitely many components ) .
then @xmath396 is a neighborhood of @xmath125 in @xmath0 @xcite .
we say that this neighborhood has a _ local product structure_. we say that a borel set @xmath397 has a _ local product structure _ if there is some @xmath398 and if there are borel sets @xmath399 and a number @xmath400 such that @xmath401 .
the @xmath2-invariant borel probability measure @xmath3 on @xmath0 in the lebesgue measure class admits a natural family of conditional measures @xmath402 on strong stable and strong unstable manifolds .
the conditional measures @xmath403 are well defined up to a universal constant , and they transform under the teichmller geodesic flow @xmath2 via @xmath404 let @xmath405 be the flip @xmath406 and let @xmath407 be the lebesgue measure on the flow lines of the teichmller flow .
the conditional measures @xmath402 are uniquely determined by the additional requirements that @xmath408 and that with respect to a local product structure , @xmath3 can be written in the form @xmath409 the measures @xmath410 on unstable manifolds defined by @xmath411 are invariant under holonomy along strong stable manifolds . to summarize , we obtain the following .
the natural homeomorphism @xmath412\notag\\ \to v(b^{ss}_{{\mathcal{q}}}(q , a_{{\mathcal{q}}}(q ) ) , b^{su}_{{\mathcal{q}}}(q , a_{{\mathcal{q}}}(q)),t_{{\mathcal{q}}}(q))=v\notag\end{aligned}\ ] ] maps the measure @xmath413 on @xmath205 defined by @xmath414 to a measure of the form @xmath415 where @xmath416 is a continuous function on @xmath205 which vanishes on @xmath417 } \phi^t b^{ss}_{{\mathcal{q}}}(q , a_{{\mathcal{q}}}(q))$ ] ( see @xcite ) .
the goal of this subsection is to relate components of strata in @xmath19 to large train tracks .
this will be used to define product coordinates near points in the boundary of a stratum .
note that the natural product coordinates on strata are period coordinates .
for a point @xmath125 in the boundary of a stratum , some of the relative periods vanish and there is no canonical choice of a relative cycle near @xmath125 which can be used for period coordinates in a neighborhood of @xmath125 .
we chose to construct product coordinates near boundary points of a stratum using train tracks even though similar coordinates can be obtained using the usual period coordinate construction .
these train track coordinates will be used in other contexts as well .
we continue to use the assumptions and notations from section 2 and section 3 . for
a large train track @xmath418 let @xmath419 be the set of all measured geodesic laminations @xmath420 whose support is carried by @xmath145 and such that the total weight of the transverse measure on @xmath145 defined by @xmath35 equals one .
let @xmath421 be the set of all area one marked quadratic differentials whose vertical measured geodesic lamination is contained in @xmath422 and whose horizontal measured geodesic lamination is carried by the dual bigon track @xmath241 of @xmath145 . by definition of a large train track
, we have @xmath423 .
the next proposition relates @xmath424 to components of strata .
[ structure ] 1 .
for every large non - orientable train track + @xmath418 there is a component @xmath285 of the stratum + @xmath270 such that for every @xmath425 the set @xmath426\}$ ] is the closure in @xmath30 of an open subset of @xmath285 .
2 . for every large orientable train track @xmath427
there is a component @xmath285 of the stratum @xmath286 such that for every @xmath425 the set @xmath426\}$ ] is the closure in @xmath277 of an open subset of @xmath285 . by @xcite
, the support @xmath46 of the vertical measured geodesic lamination of a marked quadratic differential @xmath100 can be obtained from the vertical foliation of @xmath105 by cutting @xmath1 open along each vertical separatrix and straightening the remaining leaves with respect to the hyperbolic structure @xmath428 . in particular , up to homotopy a vertical saddle connection @xmath161 of @xmath105 is contained in the interior of a complementary component @xmath162 of @xmath46 which is uniquely determined by @xmath161 .
let @xmath418 .
assume first that @xmath145 is non - orientable .
let @xmath429 be such that the support of @xmath7 is contained in @xmath430 and let @xmath431 .
then @xmath7 is non - orientable since otherwise @xmath145 inherits an orientation from @xmath7 .
the measured geodesic laminations @xmath103 jointly fill up @xmath1 ( since the support of @xmath35 is different from the support of @xmath7 and the support of @xmath7 fills up @xmath1 ) and hence if @xmath35 is normalized in such a way that @xmath102 then the pair @xmath101 defines a point @xmath432 .
our first goal is to show that @xmath433 .
the support of the geodesic lamination @xmath7 is contained in @xmath430 and therefore the orders of the zeros of the quadratic differential @xmath125 are obtained from the orders @xmath273 by subdivision .
there is a non - trivial subdivision , say of the form @xmath434 , if and only if there is at least one vertical saddle connection for @xmath125 .
assume to the contrary that there is a vertical saddle connection @xmath161 for @xmath125 .
let @xmath319 be the lift of @xmath125 to a quadratic differential on the universal covering @xmath435 of @xmath1 and let @xmath436 be a preimage of @xmath161 .
let @xmath437 be the preimage of @xmath7 . as discussed in the first paragraph of this proof , the saddle connection @xmath438 is contained in a complementary component @xmath439 of the support of @xmath440 .
this component is an ideal polygon with finitely many sides .
a biinfinite geodesic line for the singular euclidean metric defined by @xmath319 is a quasi - geodesic in the hyperbolic plane @xmath435 and hence it has well defined endpoints in the ideal boundary @xmath441 of @xmath435 .
there are two vertical geodesic lines @xmath442 for @xmath319 which contain the saddle connection @xmath438 as a subarc and which are contained in a bounded neighborhood of a side @xmath443 of @xmath439 . for a fixed orientation of @xmath438 ,
the geodesics @xmath442 are determined by the requirement that their orientation coincides with the given orientation of @xmath438 and that moreover at every singular point @xmath70 , the angle at @xmath70 to the left of @xmath444 ( or to the right of @xmath445 ) for the orientation of the geodesic and the orientation of @xmath435 equals @xmath446 .
the ideal boundary of the closed half - plane of @xmath435 which is bounded by @xmath447 ( or @xmath253 ) and which is disjoint from the interior of @xmath439 is a compact subarc @xmath252 ( or @xmath149 ) of @xmath441 .
the arcs @xmath448 are disjoint ( or , equivalently , the sides @xmath443 of @xmath439 are not adjacent ) . a horizontal geodesic line for @xmath319 which intersects the interior of the saddle connection @xmath438 is a quasi - geodesic in @xmath435 with one endpoint in the interior of the arc @xmath252 and the second endpoint in the interior of the arc @xmath149 .
now a carrying map @xmath164 for @xmath7 with @xmath449 maps the support of @xmath7 onto @xmath145 and hence it induces a bijection between the complementary components of the support of @xmath7 and the complementary components of @xmath145 .
in particular , the projections of the geodesics @xmath443 to @xmath1 determine two opposite sides of the complementary component @xmath450 of @xmath145 corresponding to the projection of @xmath439 to @xmath1 .
on the other hand , by construction of the dual bigon track @xmath241 of @xmath145 ( see @xcite , if @xmath451 is any trainpath which intersects the complementary component @xmath450 of @xmath145 then every component of @xmath452 is a compact arc with endpoints on adjacent sides of @xmath450 .
in particular , a lift to @xmath435 of such a trainpath is a quasi - geodesic in @xmath435 whose endpoints meet at most one of the two arcs @xmath453 .
since the support of the horizontal measured geodesic lamination @xmath35 of @xmath125 is carried by @xmath241 by assumption , every leaf of the support of @xmath35 corresponds to a biinfinite trainpath on @xmath241 and hence a lift to @xmath435 of such a leaf does not connect the arcs @xmath454 .
this contradicts the assumption that @xmath125 has a vertical saddle connection and hence we indeed have @xmath433 .
let @xmath455 be the open set of all projective measured geodesic laminations whose support is distinct from the support of @xmath7 .
then the assignment @xmath456 which associates to a projective measured geodesic lamination @xmath457\in { { \mathcal{p}}}(\mu)$ ] the area one quadratic differential @xmath458)$ ] with vertical measured geodesic lamination @xmath7 and horizontal projective measured geodesic lamination @xmath457 $ ] is a homeomorphism of @xmath459 onto a strong stable manifold in @xmath30 . the projectivization @xmath460 of @xmath263 is homeomorphic to a ball in a real vector space of dimension @xmath461 , and this is just the dimension of a strong stable manifold in a component of @xmath270 .
therefore by the above discussion and invariance of domain , there is a component @xmath289 of the stratum @xmath270 such that the restriction of the map @xmath456 to @xmath460 is a homeomorphism of @xmath460 onto the closure of an open subset of a strong stable manifold @xmath462 . similarly , if @xmath432 is defined by @xmath463 and if the support of @xmath35 is contained in @xmath430 then @xmath433 by the above argument .
moreover , for every @xmath464\in p{{\mathcal{v}}}(\tau)$ ] the pair @xmath465,\nu)$ ] defines a quadratic differential which is contained in a strong unstable manifold @xmath309 of a component @xmath289 of the stratum @xmath270 , and the set of these quadratic differentials equals the closure of an open subset of @xmath309 .
the set of quadratic differentials @xmath125 with the property that the support of the vertical ( or of the horizontal ) measured geodesic lamination of @xmath125 is minimal and of type @xmath44 is dense and of full lebesgue measure in @xmath270 @xcite .
moreover , this set is saturated for the stable ( or for the unstable ) foliation .
thus by the above discussion , the set of all measured geodesic laminations which are carried by @xmath145 ( or @xmath241 ) and whose support is minimal of type @xmath44 is dense in @xmath175 ( or in @xmath263 ) . as a consequence ,
the set of all pairs @xmath466 with @xmath102 which correspond to a quadratic differential @xmath433 is dense in the set of all pairs @xmath467 with @xmath102 .
thus the set @xmath424 is contained in the closure of a component @xmath289 of the stratum @xmath270 .
moreover , by reasons of dimension , @xmath468\}$ ] contains an open subset of this component .
this shows the first part of the proposition .
now if @xmath418 is orientable and if @xmath7 is a geodesic lamination which is carried by @xmath145 , then @xmath7 inherits an orientation from an orientation of @xmath145 .
the orientation of @xmath145 together with the orientation of @xmath1 determines an orientation of the dual bigon track @xmath241 ( see @xcite , and these two orientations determine the orientation of @xmath1 .
this implies that any geodesic lamination carried by @xmath241 admits an orientation , and if @xmath101 jointly fill up @xmath1 and if @xmath7 is carried by @xmath145 , @xmath35 is carried by @xmath241 then the orienations of @xmath103 determine the orientation of @xmath1 . as a consequence ,
the singular euclidean metric on @xmath1 defined by the quadratic differential @xmath125 of @xmath101 is the square of a holomorphic one - form .
the proposition follows .
if @xmath289 is a component of a stratum @xmath270 and if the large train track @xmath418 is such that @xmath469 then we say that @xmath145 _ belongs _ to @xmath289 , and we write @xmath470 .
the next proposition is a converse to proposition [ structure ] and shows that train tracks can be used to define coordinates on strata .
[ all ] 1 . for every @xmath433
there is a large non - orientable train track @xmath418 and a number @xmath471 so that @xmath472 is an interior point of @xmath424 .
2 . for every @xmath473
there is a large orientable train track @xmath474 and a number @xmath471 so that @xmath472 is an interior point of @xmath424 .
fix a complete hyperbolic metric on @xmath1 of finite volume .
define the _ straightening _ of a train track @xmath145 to be the immersed graph in @xmath1 whose vertices are the switches of @xmath145 and whose edges are the geodesic arcs which are homotopic to the branches of @xmath145 with fixed endpoints .
the hyperbolic metric induces a distance function on the projectivized tangent bundle of @xmath1 . as in section 3 of @xcite
, we say that for some @xmath475 a train track @xmath145 _ @xmath476-follows _ a geodesic lamination @xmath7 if the tangent lines of the straightening of @xmath145 are contained in the @xmath476-neighborhood of the tangent lines of @xmath7 in the projectivized tangent bundle of @xmath1 and if moreover the straightening of any trainpath on @xmath145 is a piecewise geodesic whose exterior angles at the breakpoints are not bigger than @xmath476 . by lemma 3.2 of @xcite , for every geodesic lamination @xmath7 and every @xmath477
there is a transversely recurrent train track which carries @xmath7 and @xmath476-follows @xmath7 .
let @xmath433 .
assume first that the support @xmath7 of the vertical measured geodesic lamination of @xmath125 is large of type @xmath44 .
this is equivalent to stating that @xmath125 does not have vertical saddle connections . for @xmath475 let @xmath478 be a train track which carries @xmath7 and @xmath476-follows @xmath7 .
if @xmath475 is sufficiently small then a carrying map @xmath479 defines a bijection of the complementary components of @xmath7 onto the complementary components of @xmath478 .
the transverse measure on @xmath478 defined by the vertical measured geodesic lamination of @xmath125 is positive .
let @xmath480 be a complementary component of the preimage of @xmath7 in the hyperbolic plane @xmath435 .
then @xmath439 is an ideal polygon whose vertices decompose the ideal boundary @xmath441 into finitely many arcs @xmath481 ordered counter - clockwise in consecutive order . since @xmath125 does not have vertical saddle connections , the discussion in the proof of proposition [ structure ] shows the following . let @xmath272 be a leaf of the preimage in @xmath435 of the support @xmath35 of the horizontal measured geodesic lamination of @xmath125 .
then the two endpoints of @xmath272 in @xmath435 either are both contained in the interior of the same arc @xmath482 or in the interior of two adjacent arcs @xmath483 . as a consequence , for sufficiently small @xmath476 the geodesic lamination @xmath35 is carried by the dual bigon track @xmath484 of @xmath478 ( see the characterization of the set of measured geodesic laminations carried by @xmath484 in @xcite ) . moreover , for any two adjacent subarcs @xmath483 of @xmath441 cut out by @xmath439 , the transverse measure of the set of all leaves of the preimage of @xmath35 connecting these sides is positive .
therefore for sufficiently small @xmath476 , the horizontal measured geodesic lamination @xmath35 of @xmath125 defines an interior point of @xmath485 .
now the set of quadratic differentials @xmath105 so that the support of the horizontal measured geodesic lamination of @xmath105 is large of type @xmath44 is dense in the strong stable manifold @xmath486 of @xmath125 .
the above reasoning shows that for such a quadratic differential @xmath105 and for sufficiently small @xmath476 , the horizontal measured geodesic lamination of @xmath105 is carried by @xmath484 .
but this just means that @xmath487
. moreover , if @xmath119 is the total weight which the vertical measured geodesic lamination puts on @xmath478 then @xmath488 is an interior point of @xmath489 .
thus @xmath478 satisfies the requirement in the proposition .
note that @xmath478 is necessarily non - orientable .
if @xmath473 is such that the support of the vertical measured geodesic lamination of @xmath125 is large of type @xmath490 then the above reasoning also applies and yields an oriented large train track with the required property . consider next the case that the support @xmath7 of the vertical measured geodesic lamination of @xmath125 fills up @xmath1 but is not of type @xmath44 .
then @xmath125 has a vertical saddle connection .
the set of all vertical saddle connections of @xmath125 is a finite disjoint union @xmath491 of finite trees .
the number of edges of this union of trees is uniformly bounded . for @xmath477
let @xmath478 be a train track which @xmath476-follows @xmath7 and carries @xmath7 .
if @xmath476 is sufficiently small then a carrying map @xmath492 defines a bijection between the complementary components of @xmath7 and the complementary components of @xmath478 which induces a bijection between their sides as well .
modify @xmath478 as follows . up to isotopy , a vertical saddle connection @xmath161 of @xmath125 is contained in a complementary component @xmath493 of @xmath478 which corresponds to the complementary component of @xmath7 determined by @xmath161 ( see the proof of proposition [ structure ] ) . since a carrying map @xmath494 determines a bijection between the sides of the complementary components of @xmath7 and the sides of the complementary components of @xmath145 , the horizontal lines crossing through @xmath161 determine two non - adjacent sides @xmath495 of @xmath493 ( see once more the discussion in the proof of proposition [ structure ] ) .
choose an embedded rectangle @xmath496 whose boundary intersects the boundary of @xmath493 in two opposite sides contained in the interior of the sides @xmath495 of @xmath493 .
up to an isotopy we may assume that these rectangles @xmath497 where @xmath161 runs through the vertical saddle connections of @xmath125 are pairwise disjoint .
collapse each of the rectangles @xmath497 to a single segment in such a way that the two sides of @xmath497 which are contained in @xmath478 are identified and form a single large branch @xmath498 as shown in figure c. the branch @xmath498 can be isotoped to the saddle connection @xmath161 .
let @xmath163 be the train track constructed in this way .
then @xmath163 is of topological type @xmath44 .
the train track @xmath478 can be obtained from @xmath163 by splitting @xmath163 at each of the large branches @xmath498 and removing the diagonal of the split . in particular
, @xmath163 carries @xmath478 and hence @xmath7 .
the transverse measure on @xmath163 defined by the vertical measured geodesic lamination of @xmath125 is positive and consequently @xmath163 is recurrent .
moreover , for sufficiently small @xmath476 , the horizontal measured geodesic lamination of @xmath125 is carried by @xmath499 .
as above , we conclude that if @xmath477 is sufficiently small then @xmath163 is fully transversely recurrent and in fact large .
there is a tangential measure on @xmath163 which is defined by the horizontal measured geodesic lamination of @xmath125 and which gives positive weight to each of the branches @xmath498 .
thus by possibly decreasing once more the size of @xmath476 , we can guarantee that for some @xmath471 the quadratic differential @xmath472 is an interior point of @xmath500 . as a consequence
, @xmath163 satisfies the requirements in the proposition .
if the support @xmath7 of the vertical measured geodesic lamination of @xmath125 is arbitrary then we proceed in the same way .
let @xmath475 be sufficiently small that there is a bijection between the complementary components of the train track @xmath478 and the complementary components of the support of @xmath7 .
as before , we use the horizontal measured foliation of @xmath125 to construct for every vertical saddle connection @xmath161 of @xmath125 an embedded rectangle @xmath497 in @xmath1 whose interior is contained in a complementary component of @xmath478 and with two opposite sides on @xmath478 in such a way that the rectangles @xmath497 are pairwise disjoint .
collapse each of the rectangles to a single arc .
the resulting train track has the required properties .
we discuss in detail the case that the support of @xmath7 contains a simple closed curve component @xmath447 .
then @xmath478 contains @xmath447 as a simple closed curve component as well .
there is a vertical flat cylinder @xmath162 for @xmath125 foliated by smooth circles freely homotopic to @xmath447 .
the boundary @xmath501 of @xmath162 is a finite union of vertical saddle connections .
some of these saddle connections may occur twice on the boundary of @xmath162 ( if @xmath502 then this holds true for each of these saddle connections ) .
assume without loss of generality ( i.e. perform a suitable isotopy ) that @xmath447 is a closed vertical geodesic contained in the interior of @xmath162 . for each saddle connection @xmath161 in the boundary of @xmath162 choose a compact arc @xmath503 contained in the interior of @xmath161
choose moreover a foliation @xmath504 of @xmath162 by compact arcs with endpoints on the boundary of @xmath162 which is transverse to the foliation of @xmath162 by the vertical closed geodesics and such that the following holds true .
if @xmath505 are two distinct half - leaves of @xmath504 with one endpoint in the arc @xmath503 and the second endpoint on @xmath447 then the endpoints on @xmath447 of @xmath505 are distinct .
in particular , each arc @xmath503 which occurs twice in the boundary of the cylinder @xmath162 determines an embedded rectangle @xmath497 in @xmath1 .
two opposite sides of @xmath497 are disjoint subarcs of @xmath447 ; we call these sides the vertical sides .
each of the other two opposite sides consists of two half - leaves of the foliation @xmath504 which begin at a boundary point of @xmath503 and end in a point of @xmath447 .
the interior of the arc @xmath503 is contained in the interior of @xmath497 .
the rectangles @xmath497 are pairwise disjoint .
therefore each of the rectangles @xmath497 can be collapsed in @xmath1 to the arc @xmath503 .
the resulting graph is a train track which carries @xmath447 and contains for every saddle connection @xmath161 which occurs twice in the boundary of @xmath162 a large branch @xmath498 . if @xmath161 is a saddle connection on the boundary of @xmath162 which separates @xmath162 from @xmath506 then the arc @xmath503 is contained in the interior of a rectangle @xmath497 with one side contained in @xmath447 and the second side contained in the interior of a branch of the component of @xmath478 different from @xmath447 .
this branch is determined by the horizontal geodesics which cross through @xmath161 .
as before , the rectangle @xmath497 is collapsed to a single branch . to summarize
, the train track @xmath478 can be modified in finitely many steps to a train track @xmath163 with the required properties by collapsing for every vertical saddle connection of @xmath125 a rectangle with two sides on @xmath478 to a single large branch .
this completes the construction and finishes the proof of the proposition .
* remark : * in the proof of lemma [ all ] , we constructed explicitly for every quadratic differential @xmath339 a train track @xmath507 belonging to the stratum of @xmath125 .
if @xmath125 is a one - cylinder strebel differential then the train track @xmath507 is uniquely determined by the combinatorics of its vertical saddle connections on the boundary of the cylinder .
this fact in turn can be used to obtain a purely combinatorial proof of the classification results of kontsevich - zorich @xcite and of lanneau @xcite .
let again @xmath418 .
then @xmath470 for a component @xmath289 of @xmath270 .
for every @xmath429 and every @xmath431 so that the pair @xmath101 jointly fills up @xmath1 there is a unique @xmath432 with vertical measured geodesic lamination @xmath7 and horizontal measured geodesic lamination @xmath508 . thus if @xmath460 denotes the projectivization of the cone
@xmath263 then for all @xmath509 there is a natural homeomorphism @xmath456 from the subset of @xmath510 $ ] corresponding to pairs @xmath511)$ ] which jointly fill up @xmath1 onto @xmath512}\phi^t{{\mathcal{q}}}(\tau)$ ] .
the set @xmath162 is the closure in @xmath30 of an open subset of @xmath513 .
we say that the map @xmath456 defines on @xmath162 a _ train track product structure_. if @xmath514 are borel sets then we also say that the image of @xmath515 $ ] under the map @xmath456 has a train track product structure . if @xmath432 and if @xmath162 is a neighborhood of @xmath125 with a train track product structure which is precisely invariant under the stabilizer of @xmath125 in @xmath33 then we say that the projection of @xmath162 to @xmath19 has a train track product structure .
the following proposition establishes product coordinates near boundary points of strata . for
this let again @xmath0 be a component of the stratum @xmath276 , with closure @xmath516 .
let @xmath3 be the lebesgue measure on @xmath0 .
[ forwardnondiv ] for every @xmath517 and every closed neighborhood @xmath204 of @xmath125 in @xmath516 there is a closed neighborhood @xmath518 of @xmath125 in @xmath516 with the following properties .
1 . @xmath519 for some @xmath520 , and @xmath521 for @xmath522 .
2 . for each @xmath174 ,
the set @xmath523 contains @xmath125 and has a train track product structure .
our goal is to show that every point @xmath517 has a closed neighborhood @xmath524 in @xmath516 with the following property .
let @xmath525 be a connected component of the preimage of @xmath0 and let @xmath319 be a lift of @xmath125 contained in the closure of @xmath285
. then @xmath524 lifts to a contractible neighborhood @xmath526 of @xmath319 in the closure of @xmath285 which is precisely invariant under @xmath345 .
moreover , @xmath526 is contained in @xmath527}\phi^t{{\mathcal{q}}}(\eta_j)\ ] ] for some @xmath528 where @xmath529 and where @xmath319 is contained in the boundary of @xmath530 for some @xmath531 @xmath532 .
for @xmath522 we have @xmath533}\phi^t{{\mathcal{q}}}(\eta_i)\cap \bigcup_{t\in [ a_j , b_j]}\phi^t{{\mathcal{q}}}(\eta_j)\bigr)=0.\ ] ] for this assume that @xmath534 for some @xmath535 .
assume moreover for the moment that @xmath125 does not have vertical saddle connections .
let @xmath536 be a sequence converging to @xmath125 .
since the subset of @xmath0 of quadratic differentials without vertical saddle connection is dense in @xmath0 , we may assume that for each @xmath174 , @xmath537 does not have a vertical saddle connection .
let @xmath538 be a preimage of @xmath537 such that @xmath539 .
for each @xmath174 the support @xmath540 of the vertical measured geodesic lamination of @xmath541 is large of type @xmath44 .
we claim that up to passing to a subsequence , the geodesic laminations @xmath540 converge in the hausdorff topology to a large geodesic lamination @xmath46 of topological type @xmath44 .
the lamination @xmath46 then contains the support @xmath35 of the vertical measured geodesic lamination of @xmath319 as a sublamination .
since @xmath125 does not have vertical saddle connections , @xmath35 fills up @xmath1 and @xmath46 can be obtained from @xmath35 by adding finitely many isolated leaves .
these isolated leaves subdivide some of the complementary components of @xmath35 .
the number of such limit laminations is uniformly bounded . to see that this claim indeed holds true
it is enough to assume that @xmath542 and that @xmath543 for some @xmath544 and some @xmath545 @xcite- the purpose of this assumption for our argument is to simplify the notations .
then for each sufficiently large @xmath174 the quadratic differential @xmath541 has a saddle connection @xmath183 connecting a zero @xmath546 of order @xmath547 to a zero @xmath548 of order @xmath549 whose length ( measured in the singular euclidean metric defined by @xmath541 ) tends to zero as @xmath550
. more precisely , the saddle connections @xmath183 converge to a zero @xmath551 of @xmath319 of order @xmath552 .
the length of any other saddle connection of @xmath541 is bounded from below by a universal positive constant . since @xmath541 does not have vertical saddle connections , locally near @xmath546 the interior of the saddle connection @xmath183
is contained in the interior of an euclidean sector based at @xmath546 of angle @xmath446 bounded by two vertical separatrices @xmath553 of @xmath541 which issue from @xmath546 .
the union @xmath554 is a smooth vertical geodesic line passing through @xmath546 , i.e. a geodesic which is a limit in the compact open topology of geodesic segments not passing through a singular point .
there are two vertical separatrices @xmath555 issuing from @xmath548 so that the sum of the angles at @xmath556 of the ( local ) strip bounded by @xmath557 equals @xmath446 and that the same holds true for the angle sum of the ( local ) strip bounded by @xmath558 .
the vertical length of @xmath183 is positive .
the union @xmath559 is a smooth vertical geodesic line passing through @xmath548 .
equip @xmath1 with the marked hyperbolic metric @xmath560 . for
each @xmath174 lift the singular euclidean metric on @xmath1 defined by @xmath541 to a @xmath561-invariant singular euclidean metric on the universal covering @xmath435 of @xmath1 .
let @xmath562 be a lift of the saddle connection @xmath183 .
since @xmath562 is not vertical , the leaves of the vertical foliation of @xmath541 which pass through @xmath562 define a strip of positive transverse measure in @xmath435 .
this strip is bounded by the two lifts @xmath563 of the smooth vertical geodesics @xmath564 which pass through the endpoints of @xmath562 . as @xmath550 ,
up to normalization and by perhaps passing to a subsequence , the vertical geodesics @xmath565 converge in the compact open topology to vertical geodesics @xmath566 for the singular euclidean metric defined by @xmath319 which pass through a preimage @xmath567 of the zero @xmath551 of @xmath319 of order @xmath568 . by construction ,
the geodesics @xmath569 coincide in a neighborhood of @xmath567 . since there are no vertical saddle connections for @xmath319 , we necessarily have @xmath570 .
let @xmath571 be the hyperbolic geodesic with the same endpoints as @xmath572 in the ideal boundary of @xmath435 ( see @xcite and the proof of propositsion [ structure ] ) .
the projection of @xmath573 to @xmath1 subdivides the complementary component of @xmath35 containing @xmath551 into two ideal polygons with @xmath574 and @xmath575 sides , respectively .
the union of @xmath35 with this geodesic is a large geodesic lamination @xmath46 of type @xmath44 .
this lamination is the limit in the hausdorff topology of the laminations @xmath540 .
let @xmath576 be the ( finitely many ) large geodesic laminations obtained in this way .
each of the laminations @xmath577 contains @xmath35 as a sublamination , and it is determined by a decomposition of a complementary @xmath578-gon of @xmath35 into an ideal @xmath574-gon and an ideal @xmath575-gon .
the set @xmath579 is invariant under the action of @xmath345 . for sufficiently small @xmath477 ,
a train track @xmath580 which carries @xmath581 and @xmath476-follows @xmath581 ( for the hyperbolic metric @xmath582 ) is a simple extension of a train track @xmath145 which carries @xmath35 and @xmath476-follows @xmath35 .
the added branch is a diagonal of the complementary @xmath583-gon of @xmath145 defined by the zero @xmath551 of @xmath319 of order @xmath584 .
it decomposes this component into an @xmath574-gon and an @xmath575-gon in a combinatorial pattern determined by @xmath581 .
the vertical measured geodesic lamination @xmath35 of @xmath319 defines a transverse measure on @xmath580 which gives full mass to the subtrack @xmath145 and hence it is contained in the boundary of the cone @xmath585 .
we also may assume that the horizontal measured geodesic lamination of @xmath319 is carried by the dual bigon track @xmath586 ( compare the proof of lemma [ all ] ) and that the set @xmath587 is invariant under the action of @xmath345 .
since the set of geodesic laminations carried by a train track is open and closed in the hausdorff topology @xcite , for each @xmath588 the train track @xmath580 carries a minimal large geodesic lamination of type @xmath44 ( namely , the support of the vertical measured geodesic lamination of a quadratic differential @xmath589 sufficiently close to @xmath319 from the sequence which determines @xmath580 ) and hence it follows as in the proof of proposition [ structure ] that @xmath529 .
moreover , if @xmath590 is such that @xmath591 then for every @xmath475 the set @xmath592}\phi^t{{\mathcal{q}}}(\eta_j)$ ] is a closed neighborhood of @xmath319 in the closure of @xmath289 . now if @xmath522 then @xmath593 and hence this intersection is contained in an affine subspace of codimension one . since the measure class of the conditional measures @xmath410 of @xmath3 coincides with the lebesgue measure class defined by the linear coordinates for the cone @xmath585 , the equation ( [ measureestimate ] ) holds true . as a consequence , for suitable numbers @xmath528 , the set @xmath594}\phi^t{{\mathcal{q}}}(\eta_j)\ ] ] is a @xmath345-invariant closed neighborhood of @xmath319 in the closure of @xmath285 . in other words , there is a @xmath345-invariant finite collection of closed sets with train track product structures which cover a neighborhood of @xmath319 in the closure of @xmath285 and contain @xmath319 in their boundary .
this completes the proof of the proposition in the case that the support of the vertical measured geodesic lamination of @xmath319 is large of type @xmath595 .
if the support of the vertical measured geodesic lamination of @xmath319 is not large of type @xmath595 then we argue in the same way . in this case
@xmath319 has a vertical saddle connection whose horizontal length is positive .
consider the action of the group @xmath596 on the space of quadratic differentials by rotation .
there is a sequence @xmath597 with @xmath598 such that the quadratic differential @xmath599 does not have any vertical or horizontal saddle connection .
then the supports of the horizontal and the vertical measured geodesic laminations of @xmath599 are large of type @xmath595 .
let @xmath600 be a train track as in lemma [ all ] so that for some @xmath601 , @xmath602 is an interior point of @xmath424 .
for sufficiently small @xmath603 , say whenever @xmath604 , we have @xmath605}\phi^s{{\mathcal{q}}}(\tau)$ ] where @xmath606 is a fixed number .
if @xmath607 is such that @xmath608 does not have any vertical saddle connection then the argument in the beginning of this proof shows that up to passing to a subsequence , for sufficiently large @xmath588 the vertical measured geodesic lamination of @xmath609 is carried by a simple extension of @xmath145 which is large of type @xmath44 . as before , there are only finitely many such simple extensions , and these simple extensions define train track coordinates on a neighborhood of @xmath319 in the closure of @xmath285 as before . from this
the proposition follows . as an immediate consequence , we obtain the following .
let @xmath517 and let @xmath519 be as in proposition [ forwardnondiv ] .
then for each @xmath610 there is an open subset @xmath611 of a strong unstable submanifold of @xmath0 whose closure @xmath612 contains @xmath125 .
the set @xmath613 is a compact subset of @xmath614 which contains the intersection with @xmath614 of every sufficiently small neighborhood of @xmath125 in @xmath516 .
moreover , @xmath615 for @xmath522 .
let again @xmath0 be a connected component of a stratum in @xmath19 .
then @xmath0 is invariant under the teichmller flow @xmath2 . for a periodic orbit @xmath616 for @xmath2 ,
the lebesgue measure supported in @xmath8 is a @xmath2-invariant borel measure @xmath617 on @xmath0 whose total mass equals the prime period @xmath11 of @xmath8 . if we denote for @xmath618 by @xmath619 the set of all periodic orbits for @xmath2 of period at most @xmath15 which are contained in @xmath0 then we obtain a finite @xmath2-invariant borel measure @xmath25 on @xmath0 by defining @xmath620 let @xmath7 be any weak limit of the measures @xmath25 as @xmath14 . then @xmath7 is a @xmath2-invariant borel measure on @xmath19 supported in the closure @xmath516 of @xmath0 ( which may a priori be zero or locally infinite ) .
the purpose of this section is to show [ absolute ] the measure @xmath7 on @xmath516 satisfies @xmath621 .
this means that @xmath622 for every open relative compact subset @xmath288 of @xmath516 . in particular , the measure @xmath7 is finite and absolutely continuous with respect to the lebesgue measure , and it gives full mass to @xmath0 .
a point @xmath316 is called _ forward recurrent _ ( or _ backward recurrent _ ) if it is contained in its own @xmath623-limit set ( or in its own @xmath447-limit set ) under the action of @xmath2 .
a point @xmath316 is _ recurrent _ if it is forward and backward recurrent .
the set @xmath624 of recurrent points is a @xmath2-invariant borel subset of @xmath0 .
it follows from the work of masur @xcite that a forward recurrent point @xmath339 has uniquely ergodic vertical and horizontal measured geodesic laminations whose supports fill up @xmath1 . as a consequence ,
the preimage @xmath625 of @xmath626 in @xmath30 is contained in the set @xmath131 defined in ( [ tildea ] ) of section 2 . using the notations from section 2
, there is a number @xmath83 such that for every @xmath32 the map @xmath126 is an unparametrized @xmath85-quasi - geodesic in the curve graph @xmath59 .
if @xmath125 is a lift of a recurrent point in @xmath19 then this unparametrized quasi - geodesic is of infinite diameter .
recall from ( [ distance ] ) of section 2 the definition of the distances @xmath96 @xmath627 on @xmath63 and of the sets @xmath628 @xmath629 .
the following lemma is a version of lemma 2.1 of @xcite .
[ expansion ] there are numbers @xmath630 with the following property .
let @xmath631 and for @xmath632 write @xmath633 ; then @xmath634 the map @xmath635 defined in section 2 is equivariant under the action of the mapping class group on @xmath636 and on @xmath60 . in particular , for @xmath637 and @xmath119 the set @xmath638 is invariant under @xmath639 , and the same holds true for @xmath640 .
let @xmath641 be a component of the preimage of @xmath0 and let @xmath642 be the stabilizer of @xmath289 in @xmath33 .
the @xmath2-invariant borel probability measure @xmath3 on @xmath0 in the lebesgue measure class lifts to a @xmath313-invariant locally finite measure on @xmath289 which we denote again by @xmath3 .
the conditional measures @xmath643 of @xmath3 on the leaves of the strong stable and strong unstable foliation of @xmath0 lift to a family of conditional measures on the leaves of the strong stable and strong unstable foliation @xmath644 of @xmath289 , respectively , which we denote again by @xmath402 ( see the discussion in section 4 ) .
[ borelregular ] for every @xmath645 and for all compact neighborhoods @xmath646 of @xmath319 in @xmath647 there are compact neighborhoods @xmath648 of @xmath319 in @xmath649 with the following properties . 1
. @xmath650 are precisely invariant under @xmath345 .
there are numbers @xmath651 such that @xmath652 3 .
4 . if @xmath654 and if @xmath655 then @xmath656 . let @xmath316 be a recurrent point and let @xmath303 be a lift of @xmath125 .
let @xmath657 be compact neighborhoods of @xmath319 and let @xmath119 be such that @xmath658 is precisely invariant under @xmath345 and projects to a metric orbifold ball in @xmath659 . by lemma
[ closed ] , the map @xmath660 is continuous and closed , and the sets @xmath661 @xmath662 form a neighborhood basis of @xmath663 in @xmath63 .
thus there is a number @xmath664 such that @xmath665 for @xmath666 let @xmath667 be the closure of the set @xmath668 then @xmath669 is a closed neighborhood of @xmath319 in @xmath670 which is precisely invariant under @xmath345 .
moreover , @xmath671 for @xmath672 , and lemma [ closed ] shows that @xmath673 . since the conditional measure @xmath674 on @xmath647 is borel regular , for every @xmath475 there are numbers @xmath675 so that @xmath676 for these number @xmath677 , all requirements in the lemma hold true
this shows the lemma .
* remark : * since @xmath128 is dense in @xmath30 and the map @xmath660 is continuous and closed , the sets @xmath678 have dense interior .
moreover , we may assume that their boundaries have vanishing lebesgue measures .
let again @xmath641 be a component of the preimage of @xmath0 .
for @xmath316 let @xmath319 be a preimage of @xmath125 in @xmath289 and let @xmath679 be the cardinality of the quotient of @xmath345 by the normal subgroup of all elements of @xmath345 which fix @xmath289 pointwise ( for example , the hyperelliptic involution of a closed surface of genus @xmath147 acts trivially on the entire bundle @xmath30 ) .
we note [ cals ] the set @xmath680 is an open dense @xmath2-invariant submanifold of @xmath0 .
the mapping class group preserves the teichmller metric on @xmath29 and hence an element @xmath681 which stabilizes a quadratic differential @xmath317 fixes pointwise the teichmller geodesic with initial cotangent @xmath319 .
therefore the set @xmath682 is @xmath2-invariant , moreover it is clearly open . since the teichmller flow on @xmath0 has dense orbits , either @xmath682 is empty or dense .
however , @xmath33 acts properly discontinuously on @xmath29 and consequently the first possibility is ruled out by the fact that the conjugacy class of an element of @xmath33 which fixes an entire component of the preimage of @xmath0 does not contribute towards @xmath679 . for a control of the measure
@xmath7 we use a variant of an argument of margulis @xcite .
namely , for numbers @xmath683 let @xmath684 be the set of all periodic orbits of @xmath2 which are contained in @xmath0 , with prime periods in the interval @xmath685 .
for an open or closed subset @xmath205 of @xmath516 and numbers @xmath683 define @xmath686 where @xmath687 is the characteristic function of @xmath205 . to obtain control on the quantities @xmath688 we use a tool from @xcite .
namely , every leaf @xmath689 of the strong stable foliation of @xmath19 can be equipped with the _ hodge distance _
@xmath690 ( or , rather , the modified hodge distance , @xcite ) .
this hodge distance is defined by a norm on the tangent space of @xmath689 ( with a suitable interpretation ) . in particular ,
closed @xmath690-balls of sufficiently small finite radius are compact , and balls about a given point @xmath125 define a neighborhood basis of @xmath125 in @xmath689 .
we also obtain a hodge distance on the leaves of the strong unstable foliation as the image under the flip @xmath504 of the hodge distance on the leaves of the strong stable foliation .
these hodge distances restrict to hodge distances on the leaves of the foliations @xmath691 which we denote by the same symbol @xmath690 .
the following result is theorem 8.12 of @xcite , [ hodgenorm ] there is a number @xmath692 such that @xmath693 for all @xmath694 and all @xmath695 .
the next lemma provides some first volume control for the measure @xmath7 .
[ upperestimate ] for every recurrent point @xmath316 with @xmath696 , for every neighborhood @xmath205 of @xmath125 in @xmath0 and for every @xmath475 there is a number @xmath378 and there is an open neighborhood @xmath697 of @xmath125 such that @xmath698 we use the strategy of the proof of lemma 6.1 of @xcite . the idea is to find for every recurrent point @xmath316 with @xmath696 , for every neighborhood @xmath205 of @xmath125 in @xmath0 and for every @xmath699 some number @xmath378 and closed neighborhoods @xmath700 of @xmath125 in @xmath0 with dense interior such that for all sufficiently large @xmath618 the following properties hold . 1 . @xmath701 is connected and has a local product structure .
2 . @xmath702 .
3 . let @xmath703 and assume that @xmath704 for some @xmath705 .
let @xmath706 be the component containing @xmath105 of the intersection @xmath707 and let @xmath708
. then @xmath709,\ ] ] and the length of the connected orbit subsegment of @xmath710 containing @xmath105 equals @xmath384 .
4 . there is at most one periodic orbit for @xmath2 of prime period @xmath711 which intersects @xmath712 , and the intersection of this orbit with @xmath712 is connected .
the construction is as follows .
let @xmath316 be recurrent with @xmath713 and let @xmath205 be a neighborhood of @xmath125 in @xmath0 . using the notations from subsection 4.1 , for @xmath475 there are numbers @xmath714 such that @xmath715 is a set with a local product structure .
let @xmath317 be a preimage of @xmath125 . by construction
( see the discussion in section 4.1 ) , the set @xmath716 is precisely invariant under @xmath345 . in particular , since @xmath696 , the set @xmath717 is mapped homeomorphically onto @xmath701 by the projection @xmath718 .
since periodic orbits for @xmath2 are in bijection with conjugacy classes of pseudo - anosov elements of @xmath33 , up to making @xmath719 smaller we may assume that the following holds true .
for every @xmath720 , every component of the intersection @xmath721 is intersected by at most one periodic orbit for the teichmller flow with prime period contained in the interval @xmath722 $ ] , and if such an orbit exists then its intersection with @xmath721 is connected . as in ( [ theta ] ) of section 4 , for @xmath723 let @xmath724 be defined by the requirement that @xmath725 for all @xmath293 . similarly , as in ( [ zeta ] ) of section 4 , let @xmath726 be defined by @xmath727 .
we claim that for sufficiently small @xmath728 and for every @xmath729 the following holds true . 1 .
the jacobian of the embedding @xmath730 and of the embedding @xmath731 with respect to the measures @xmath732 and @xmath674 , respectively , is contained in the interval @xmath733 $ ] .
the restriction to @xmath734 of the function @xmath177 defined in ( [ sigma ] ) takes values in the interval @xmath735 $ ] .
3 . if @xmath736 and if @xmath737 is such that @xmath738 then @xmath739 and @xmath740 . here and in the sequel , for @xmath741
we denote by @xmath742 the connected component containing @xmath105 of the intersection @xmath743 .
to verify the claim , note first that property b ) can be fulfilled since @xmath177 is continuous and @xmath2-invariant and equals one at @xmath125 .
property a ) is fulfilled for sufficiently small @xmath744 since the measures @xmath745 ( or @xmath410 ) are invariant under holonomy along the strong unstable ( or the strong unstable ) foliation and since @xmath746 and @xmath411 and hence jacobians of the maps @xmath747 are controlled by the function @xmath177 . by property
b ) above and by theorem [ hodgenorm ] , the last property is fulfilled if we choose @xmath748 small enough so that for some @xmath119 and every @xmath741 the following is satisfied . for
every @xmath749 the diameter of @xmath750 with respect to the hodge distance does not exceed @xmath121 , and the hodge distance between @xmath750 and the boundary of @xmath751 is not smaller than @xmath752 . since @xmath753 , property b ) implies the following .
for all closed sets @xmath754 @xmath354 and for every @xmath755 we have @xmath756 moreover , we have @xmath757.\ ] ] by the estimate ( [ deltacomparison ] ) in section 2 , there is a number @xmath374 such that for any two points @xmath758 with @xmath759 the distances @xmath760 on @xmath60 are @xmath761-bilipschitz equivalent .
let @xmath285 be a component of the preimage of @xmath0 in @xmath30 .
let @xmath762 be a lift of @xmath125 .
choose closed neighborhoods @xmath763 of @xmath319 whose images under the flip @xmath504 satisfy the properties in lemma [ borelregular ] for some numbers @xmath764 where @xmath765 is as in lemma [ expansion ] .
choose also closed neighborhoods @xmath766 of @xmath319 with the properties in lemma [ borelregular ] for some numbers @xmath767 . by the choice of the set @xmath701 , for any two points @xmath768 the distances @xmath769 and @xmath138
are @xmath761-bilipschitz equivalent . as a consequence , for all @xmath770 the @xmath769-diameter of @xmath771 and @xmath772 does not exceed @xmath773 .
let @xmath774 by assumption , @xmath125 is recurrent and hence by lemma [ expansion ] , applied to both @xmath319 and @xmath775 , there is a number @xmath776 so that for every @xmath777 and for every @xmath778 with @xmath779 we have @xmath780 moreover , there is a mapping class @xmath781 and a number @xmath782 such that @xmath783 is an interior point of @xmath784 . by equivariance under the action of the mapping class group , for every @xmath785
the @xmath769-diameter of @xmath786 is smaller than @xmath773 . in particular ,
the @xmath787-diameter of @xmath788 is smaller than @xmath773 .
the second part of inequality ( [ lemmaupper ] ) then implies that the @xmath789-diameter of @xmath790 does not exceed @xmath193 .
thus by property c ) above , by the choice of @xmath193 and by lemma [ borelregular ] , we have @xmath791 define @xmath792 then @xmath319 is an interior point of @xmath793 ( as a subset of @xmath794 ) , and @xmath795 are precisely invariant under @xmath345 ( since a non - trivial element of @xmath345 fixes @xmath289 pointwise ) .
the conditional measures @xmath674 are invariant under holonomy along the strong stable foliation and transform under the teichmller flow by @xmath796 .
moreover , @xmath797 and hence properties a ) and b ) above and the definition of the function @xmath177 imply that @xmath798 define @xmath799 and let @xmath800 be the projection of @xmath801 to @xmath0 .
note that we have @xmath802 and @xmath803 by the choice of @xmath804 , by the estimate in a ) above , by invariance of @xmath3 under the flow @xmath2 ( which implies that @xmath805 ) and by the fact that @xmath801 is mapped homeomorphically onto @xmath800 for @xmath806 .
moreover , each of the sets @xmath800 is closed with dense interior .
let @xmath807 and let @xmath703 be a periodic point for @xmath2 of period @xmath808 $ ] . since every orbit of @xmath2 which intersects @xmath809 also intersects @xmath810 we may assume that @xmath811 .
let @xmath706 be the component containing @xmath105 of the intersection @xmath721 and let @xmath812 we claim that @xmath813.\ ] ] to see that this is indeed the case , let @xmath814 be a lift of @xmath105 . by the choice of the set @xmath815 and by the first part of the estimate ( [ lemmaupper ] ) , the @xmath816-diameter of the set @xmath817 does not exceed @xmath193 . in particular , since @xmath703 and property c ) above holds true , we have @xmath818 and similarly @xmath819 let @xmath820 be such that @xmath821 then by the estimate ( [ exactcontrol ] ) and by ( [ contain ] ) , we have @xmath822 now by the estimate ( [ productestimate ] and the fact that @xmath823 preserves the stable foliation and contracts the measures @xmath824 by the factor @xmath825 , we conclude that @xmath826 and similarly @xmath827 together with the estimate ( [ productestimate ] ) this implies the estimate ( [ lambdaestimate ] ) .
since @xmath808 $ ] and @xmath828 we conclude that the measure of the set @xmath712 is at least @xmath829 moreover , the lebesgue measure of the component containing @xmath105 of the orbit segment @xmath830 equals @xmath384 .
on the other hand , if @xmath831 are periodic points of prime periods @xmath832 $ ] then by our choice of @xmath701 the components containing @xmath833 of the intersection @xmath721 are disjoint
. thus there are at most @xmath834 such intersection arcs which are subarcs of periodic orbits of prime period in @xmath835 $ ] . however , since the lebesgue measure @xmath3 is mixing for the teichmller flow @xcite , for sufficiently large @xmath15 we have @xmath836 from this we deduce that @xmath837 for all sufficiently large @xmath618 .
this shows the lemma .
now we are ready for the proof of proposition [ absolute ] .
_ proof of proposition [ absolute ] .
_ let @xmath7 be a weak limit of the measures @xmath25 as @xmath14 .
then @xmath7 is a ( a priori locally infinite ) @xmath2-invariant borel measure supported in the closure @xmath516 of @xmath0 .
this measure is moreover invariant under the flip @xmath838 . by lemma [ upperestimate ] it suffices to show the following .
let @xmath839 be a closed @xmath2-invariant set of vanishing lebesgue measure .
then for all @xmath477 , every @xmath840 has a neighborhood @xmath288 in @xmath516 such that @xmath841 .
first let @xmath842 .
choose compact balls @xmath843 about @xmath125 for the hodge distance of radius @xmath844 @xmath354 and numbers @xmath845 such that @xmath846 is a set with a local product structure .
in particular , for every preimage @xmath319 of @xmath125 in @xmath30 the component of the preimage of @xmath847 containing @xmath319 is precisely invariant under @xmath345 .
then @xmath848 are closed neighborhoods of @xmath125 in @xmath0 .
let moreover @xmath849 we may assume that for one ( and hence every ) component @xmath850 of the preimage of @xmath847 in @xmath30 the diameter of the projection @xmath851 of @xmath850 to @xmath29 does not exceed one . as in the proof of lemma [ upperestimate ] we require that moreover the following holds true .
@xmath852 if @xmath853 and if @xmath737 is such that @xmath854 then @xmath855 and moreover @xmath856 .
that this requirement can be met follows from theorem [ hodgenorm ] and the discussion in the proof of lemma [ upperestimate ] . if @xmath517 then we choose closed neighborhoods @xmath857 of @xmath125 in @xmath516 as in proposition [ forwardnondiv ] such that @xmath858 and @xmath859 are the intersections with @xmath860 of closed balls for the hodge norm .
we require that property @xmath852 above holds true ( with a slight abuse of notation ) .
let @xmath749 and let @xmath119 be such that @xmath861 .
let @xmath326 be the connected component containing @xmath293 of the intersection @xmath862 . by the property @xmath852 , we have @xmath863 .
moreover , the connected component containing @xmath293 of the intersection @xmath864 contains the component containing @xmath293 of the intersection @xmath865 .
thus as in the proof of lemma [ upperestimate ] , we observe that for any point @xmath866 and every @xmath119 such that @xmath861 the lebesgue measure of the intersection @xmath867 is bounded from below by @xmath868 where @xmath869 is a fixed constant which only depends on @xmath870 .
moreover , the number of periodic points @xmath741 of period @xmath871 $ ] such that the intersection components @xmath872 containing @xmath873 are _ not _ disjoint is bounded from above by the cardinality of @xmath345 where @xmath319 is a preimage of @xmath125 in @xmath30 .. for @xmath874 and @xmath875 write @xmath876 if there are lifts @xmath877 of @xmath878 to @xmath30 such that @xmath879 write moreover @xmath880 if there are lifts @xmath877 of @xmath878 to @xmath30 such that @xmath881 note that if @xmath882 then also @xmath883 . for a subset @xmath884 of @xmath516
define @xmath885 then @xmath886 and @xmath887 are open neighborhoods of @xmath884 . for @xmath888 define @xmath889 then for all @xmath890 each @xmath888 , @xmath891 is an open neighborhood of @xmath892 in @xmath734 , and @xmath893 is an open neighborhood of @xmath891 in @xmath892 .
moreover , we have @xmath894 for all @xmath588 and @xmath895 . if @xmath896 then there is some @xmath897 and there are lifts @xmath898 of @xmath899 to @xmath30 such that @xmath900 for all @xmath695 .
however , up to removing from @xmath901 a set of vanishing lebesgue measure , this implies that @xmath902 @xcite .
but @xmath903 and therefore @xmath904 by absolute continuity .
since @xmath3 is borel regular , the lebesgue measures of the sets @xmath905 tend to zero as @xmath906 .
similarly , we infer that @xmath907 .
thus for every @xmath98 there are numbers @xmath908 and @xmath909 such that we have @xmath910 for all @xmath911 .
now let @xmath912 and let @xmath913 be a periodic point for @xmath2 of prime period @xmath914 $ ] .
let @xmath915 be the component of @xmath867 containing @xmath916 .
then every point in @xmath915 is contained in @xmath917 . by lemma
[ upperestimate ] and its proof , the lebesgue measure of this intersection component is bounded from below by @xmath918 where @xmath869 is as above . moreover ,
the number of periodic points @xmath919 for which these intersection components are not disjoint is uniformly bounded . in particular , there is a number @xmath920 not depending on @xmath921 such that the number of these intersection components is bounded from above by @xmath922 times the lebesgue measure of @xmath917 , i.e. by @xmath923 .
this implies that we have @xmath924 .
since @xmath374 was arbitrary we conclude that @xmath925 .
proposition [ absolute ] follows .
in this section we complete the proof of the theorem from the introduction .
we continue to use the assumptions and notations from sections 2 - 5 . as before ,
let @xmath926 be a component of a stratum , equipped with the @xmath2-invariant lebesgue measure @xmath3 .
let @xmath927 be the open dense @xmath2-invariant subset of full lebesgue measure of all points @xmath125 with @xmath928 .
then @xmath682 is a manifold .
let @xmath929 and let @xmath930 be an open relative compact contractible neighborhood of @xmath125 . for @xmath931 define a _ periodic @xmath932-pseudo - orbit _ for the teichmller flow @xmath2 on @xmath0 to consist of a point @xmath933 and a number @xmath934 such that @xmath935 .
we denote such a periodic pseudo - orbit by @xmath936 .
a periodic @xmath932-pseudo - orbit @xmath936 determines up to homotopy a closed curve beginning and ending at @xmath70 which we call a _ characteristic curve _
( compare section 4 of @xcite ) .
this characteristic curve is the concatenation of the orbit segment @xmath937 with a smooth arc in @xmath288 which is parametrized on @xmath938 $ ] and connects the endpoint @xmath939 of the orbit segment with the starting point @xmath70 .
[ quasigeodesic ] there is a number @xmath940 and for every recurrent point @xmath929 there is an open relative compact contractible neighborhood @xmath205 of @xmath125 in @xmath682 and there is a number @xmath941 depending on @xmath205 with the following property .
let @xmath942 be a periodic @xmath943-pseudo - orbit and let @xmath8 be a lift to @xmath30 of a characteristic curve of the pseudo - orbit .
then the curve @xmath944 is an infinite unparametrized @xmath77-quasi - geodesic in @xmath59 .
* remark : * lemma 4.4 of @xcite is formulated for @xmath19 rather than for a component of a stratum .
however , the statement and its proof immediately carry over to the result formulated in lemma [ quasigeodesic ] .
for a point @xmath316 choose a preimage @xmath317 of @xmath125 and for @xmath875 define @xmath945 note that @xmath946 depends on the choice of the map @xmath84 ( and on the choice of the lift @xmath319 ) .
however , by lemma 3.3 of @xcite , there is a _ continuous _ function @xmath947 and a number @xmath948 such that @xmath949 for all @xmath950 . in particular , the values @xmath951 and @xmath952 are independent of any choices made and coincide with the corresponding values for @xmath953 .
we use this observation to show by the choice of @xmath948 and by the triangle inequality , we have @xmath955 for all @xmath956 . therefore the subadditive ergodic theorem shows that for @xmath3-almost all @xmath316 the limit @xmath957 exists and is independent of @xmath125 .
we are left with showing that this limit is positive . by lemma 2.4 of @xcite
, there is a number @xmath119 such that for every @xmath100 and all @xmath958 we have @xmath959 let @xmath316 be a periodic point for @xmath2 .
then there is a number @xmath606 such that for every lift @xmath319 of @xmath125 to @xmath30 the map @xmath960 is a biinfinite @xmath149-quasi - geodesic in @xmath59 @xcite .
thus by inequality ( [ distortion ] ) and continuity of @xmath2 we can find an open neighborhood @xmath961 of @xmath125 and a number @xmath962 such that @xmath963 now if @xmath964 and if @xmath965 are such that the cardinality of the set of all numbers @xmath966 with @xmath967 is not smaller than @xmath245 then @xmath968 .
the measure @xmath3 is @xmath969-invariant and ergodic , and @xmath970 .
thus by the birkhoff ergodic theorem , the proportion of time a typical orbit for the map @xmath969 spends in @xmath288 is positive . the lemma follows .
[ lowerestimate ] for every recurrent point @xmath929 , for every neighborhood @xmath205 of @xmath125 in @xmath682 and for every @xmath475 there is an open neighborhood @xmath697 of @xmath125 in @xmath682 and a number @xmath378 such that @xmath971 let @xmath929 be recurrent and let @xmath205 be an open neighborhood of @xmath125 which satisfies the conclusion of lemma [ quasigeodesic ] for some @xmath941 .
let @xmath475 . with the notations from section 4 ,
let @xmath972 be such that @xmath973 .
choose a number @xmath728 which is sufficiently small that for every @xmath974 the jacobian at @xmath105 of the homeomorphism @xmath975\ ] ] with respect to the measures @xmath3 and @xmath976 is contained in the interval @xmath977 $ ] .
we may assume that any two points in a component @xmath978 of the preimage of @xmath734 can be connected in @xmath978 by a smooth curve whose projection to @xmath29 is of length at most @xmath979 .
let @xmath765 be as in lemma [ expansion ] .
let @xmath319 be a lift of @xmath125 to a component @xmath289 of the preimage of @xmath0 in @xmath30 .
recall from section 2 the definition of the map @xmath980 .
since @xmath125 is recurrent , the horizontal and the vertical measured geodesic laminations of @xmath319 are uniquely ergodic @xcite .
let @xmath981 be neighborhoods of @xmath125 as in the proof of lemma [ upperestimate ] and let @xmath982 be components of lifts of @xmath983 to @xmath289 which contain @xmath319 .
these sets have the following property . 1 .
there are closed sets @xmath984 with dense interior @xmath354 and there are numbers @xmath845 such @xmath985 .
2 . for every @xmath986
the @xmath769-diameter of @xmath987 and of @xmath988 is not bigger than @xmath444 .
3 . @xmath989 .
4 . there is a number @xmath990 with the following property . if @xmath991 and if @xmath992 ( or @xmath993 ) is an open neighborhood of @xmath105 such that the @xmath138-diameter of @xmath994 ( or of @xmath995 ) is not bigger than @xmath152 then @xmath996 and the @xmath2-orbit of every point of @xmath162 intersects @xmath997 in an arc of length @xmath384 .
let @xmath998 be the canonical projection .
by lemma [ asymptotic ] and lemma [ expansion ] , there is a number @xmath962 and there is a borel subset @xmath999 with @xmath1000 such that for every @xmath1001 and every @xmath1002 we have @xmath1003 where @xmath98 is as in the estimate ( [ deltacomparison ] ) .
we may assume that @xmath1004 for some borel set @xmath1005 .
in particular , we conclude as in the proof of lemma [ upperestimate ] ( see the estimate ( [ lambdaestimate ] ) that ( with some a - priori adjustment of the constant @xmath476 ) the following holds true .
let @xmath1006 and let @xmath1002 be such that @xmath1007 .
let @xmath706 be the connected component containing @xmath1008 of the intersection @xmath1009 .
then the lebesgue measure of the intersection @xmath1010 is not bigger than @xmath1011 on the other hand , since the lebesgue measure is mixing , for sufficiently large @xmath1012 we have @xmath1013 together this implies that the number of such intersection components is at least @xmath1014 next we claim that for sufficiently large @xmath1015 and for a point @xmath1006 with @xmath1016 there is a periodic orbit for the flow @xmath2 which intersects @xmath1017 in an arc of length at least @xmath384 and whose period is contained in the interval @xmath1018 $ ] . to this end
let @xmath1019 ; then the conclusion of lemma [ quasigeodesic ] is satisfied for every periodic @xmath1020-pseudo - orbit beginning at a point @xmath1021 .
let @xmath1022 be such that @xmath1023 for some @xmath1024 .
let @xmath8 be a characteristic curve of the periodic @xmath1020-pseudo - orbit @xmath1025 which we obtain by connecting @xmath1026 with @xmath1022 by a smooth arc contained in @xmath809 .
up to replacing @xmath242 by @xmath1027 for some @xmath1028 \subset [ \epsilon/2,\epsilon/2]$ ] we may assume that @xmath1029 .
let @xmath573 be a lift of @xmath8 to @xmath289 with starting point @xmath1030 .
then @xmath573 is invariant under a mapping class @xmath1031 whose conjugacy class defines the homotopy class of @xmath8 in @xmath682 . a fundamental domain for the action of @xmath1032 on @xmath573 projects to a smooth arc in @xmath29 of length at most @xmath1033 . by lemma [ quasigeodesic ] and the choice of @xmath1034 the curve @xmath1035 is an unparametrized @xmath77-quasi - geodesic in @xmath59 of infinite diameter .
up to perhaps a uniformly bounded modification , this quasi - geodesic is invariant under the mapping class @xmath1031 , and @xmath1032 acts on the quasi - geodesic @xmath1036 as a translation . as a consequence , @xmath1032 acts on @xmath59 with unbounded orbits and
hence it is pseudo - anosov . by invariance of @xmath573 under @xmath1032 , the attracting fixed point of @xmath1032 is just the endpoint of @xmath1036 in @xmath63 .
since @xmath1032 is pseudo - anosov , there is a closed orbit @xmath1037 for @xmath2 on @xmath19 which is the projection of a @xmath1032-invariant flow line @xmath1038 for @xmath2 in @xmath30 .
the length of the orbit is at most @xmath1039 .
the image under the map @xmath1040 of the orbit @xmath1038 in @xmath30 is an unparametrized @xmath85-quasi - geodesic in @xmath59 which connects the two fixed points for the action of @xmath1032 on @xmath63 .
assume that the characteristic curve @xmath8 is parametrized on @xmath1041 $ ] with @xmath1042 .
as in the proof of theorem 4.3 of @xcite , we claim that for every @xmath1043 we have @xmath1044 ( note that this makes sense since the points @xmath1045 are lifts of recurrent points in @xmath19 by assumption ) . to see this we proceed by induction on @xmath174 .
the case @xmath1046 follows from the definition and from ( [ deltacomparison ] ) above , so assume that the claim is known for all @xmath1047 and some @xmath1048 .
by equivariance under the action of the mapping class group we have @xmath1049 moreover the distances @xmath1050 are @xmath761-bilipschitz equivalent .
now @xmath1051 for all @xmath1052 by the induction hypothesis and therefore @xmath1053 on the other hand , by the choice of @xmath152 and the choice of @xmath15 and the fact that @xmath1054 we obtain that @xmath1055 together this implies the above claim . as a consequence ,
the attracting fixed point @xmath46 for the action of the pseudo - anosov element @xmath1032 on @xmath60 is contained in the ball @xmath1056 , moreover it is contained in the closure of the set @xmath1057 .
the same argument also shows that the repelling fixed point of @xmath1032 is contained in the intersection of @xmath1058 with the closure of @xmath1059 .
since the map @xmath165 is closed we conclude that the axis of @xmath1032 is contained in the closure of @xmath285 .
since @xmath1060 , by property 4 ) above , this axis passes through the lift @xmath997 of @xmath1017 containing @xmath319 .
in other words , the projection of this axis to @xmath516 passes through @xmath1017 , and , in particular , it is contained in @xmath0 .
moreover , it intersects the component of @xmath1061 which contains @xmath1062 . as a consequence ,
the length of the axis is contained in @xmath1063\subset [ n-\epsilon , n+\epsilon]$ ] .
to summarize , there is an injective assignment which associates to every @xmath1064 and to every connected component of the intersection @xmath1065 for @xmath1066 which contains points in @xmath1067 a subarc of length @xmath384 of the intersection with @xmath1017 of a periodic orbit for @xmath2 whose period is contained in @xmath1018 $ ] .
together with the above discussion , this completes the proof of the proposition . by proposition [ absolute ] and proposition [ lowerestimate ]
, it suffices to show the following .
let @xmath316 be birecurrent and let @xmath475 . for @xmath618 let @xmath619 be the set of all periodic orbits of @xmath2 in @xmath0 of period at most @xmath15 .
then there is a compact neighborhood @xmath1068 of @xmath125 in @xmath0 and there is a number @xmath931 such that for every @xmath1069 the measure @xmath1070 assigns the mass @xmath1071\ ] ] to @xmath1068 .
however , this holds true by proposition [ absolute ] and proposition [ lowerestimate ] .
this completes the proof of the theorem .
* acknowledgement : * this work was carried out in fall 2007 while i participated in the program on teichmller theory and kleinian groups at the msri in berkeley .
i thank the organizers for inviting me to participate , and i thank the msri for its hospitality .
i also thank juan souto for raising the question which is answered in this note .
u. hamenstdt , _ train tracks and the gromov boundary of the complex of curves _ , in `` spaces of kleinian groups '' ( y. minsky , m. sakuma , c. series , eds . ) , london math .
notes 329 ( 2006 ) , 187207 . | let @xmath0 be a connected component of a stratum in the moduli space of abelian or quadratic differentials for a non - exceptional riemann surface @xmath1 of finite type .
we show that the probability measure on @xmath0 in the lebesgue measure class which is invariant under the teichmller flow is obtained by bowen s construction . |
the excited states of the nucleon have been studied experimentally since the 1950 s @xcite .
they contributed to the discovery of the quark model in 1964 by gell - mann and zweig @xcite , and were critical for the discovery of `` color '' degrees of freedom as introduced by greenberg @xcite .
the quark structure of baryons resulted in the prediction of a wealth of excited states with underlying spin - flavor and orbital symmetry of @xmath2 , and led to a broad experimental effort to search for these states .
most of the initially observed states were found with hadronic probes .
however , of the many excited states predicted in the quark model , only a fraction have been observed to date .
search for the `` missing '' states and detailed studies of the resonance structure are now mostly carried out using electromagnetic probes and have been a major focus of hadron physics for the past decade @xcite .
a broad experimental effort is currently underway with measurements of exclusive meson photoproduction and electroproduction reactions , including many polarization observables .
precision data and the development of multi - channel partial wave analysis procedures have resulted in the discovery of several new excited states of the nucleon , which have been entered in the review of particle physics @xcite .
a quantitative description of baryon spectroscopy and the structure of excited nucleons must eventually involve solving qcd for a complex strongly interacting multi - particle system .
recent advances in lattice qcd led to predictions of the nucleon spectrum in qcd with dynamical quarks @xcite , albeit with still large pion masses of 396 mev .
lattice prediction can therefore only be taken as indicative of the quantum numbers of excited states and not of the masses of specific states . in parallel ,
the development of dynamical coupled channel models is being pursued with new vigor .
the ebac group at jlab has shown @xcite that dynamical effects can result in significant mass shifts of the excited states . as a particularly striking result
, a very large shift was found for the roper resonance pole mass to 1365 mev downward from its bare core mass of 1736 mev .
this result has clarified the longstanding puzzle of the incorrect mass ordering of @xmath3 and @xmath4 resonances in the constituent quark model .
developments on the phenomenological side go hand in hand with a world - wide experimental effort to produce high precision data in many different channel as a basis for a determination of the light - quark baryon resonance spectrum . on the example of experimental results from clas ,
the strong impact of precise meson photoproduction data is discussed .
several reviews have recently been published on this and related subjects @xcite . it is interesting to point out recent findings that relate the observed baryon spectrum of different quark flavors with the baryon densities in the freeze out temperature in heavy ion collisions , which show evidence for missing baryons in the strangeness and the charm baryon sector @xcite .
these data hint that an improved baryon model including further unobserved light quark baryons may resolve the current discrepancy between lattice qcd results and the results obtained using a baryon model that includes only states listed by the pdg .
a complete accounting of excited baryon states of all flavors seems essential for a quantitative description of the occurrence of baryons in the evolution of the microsecond old universe .
accounting for the complete excitation spectrum of the nucleon ( protons and neutrons ) and understanding the effective degrees of freedom is perhaps the most important and certainly the most challenging task of hadron physics .
the experimental n * program currently focusses on the search for new excited states in the mass range from 2 gev to 2.5 gev using energy - tagged photon beams in the few gev range , and the study of the internal structure of prominent resonances in meson electroproduction .
the complex structure of the light - quark ( u & d quarks ) baryon excitation spectrum complicates the experimental search for individual states . as a result of the strong interaction ,
resonances are wide , often 200 mev to 400 mev , and are difficult to uniquely identify when only differential cross sections are measured .
most of the excited nucleon states listed in the review of particle properties prior to 2012 have been observed in elastic pion scattering @xmath5 .
however there are important limitations in the sensitivity to the higher mass nucleon states that may have small @xmath6 decay widths .
the extraction of resonance contributions then becomes exceedingly difficult in @xmath7 scattering .
estimates for alternative decay channels have been made in quark model calculations@xcite for various channels .
this has led to a major experimental effort at jefferson lab , elsa , graal , and mami to chart differential cross sections and polarization observables for a variety of meson photoproduction channels . at jlab with clas , several final states have been measured with high precision@xcite that are now employed in multi - channel analyses . here
one focus has recently been on measurements of @xmath8 , using a polarized photon beam several polarization observables can be measured by analyzing the parity violating decay of the recoil @xmath9 . it is well known that the energy - dependence of a partial - wave amplitude for one particular channel is influenced by other reaction channels due to unitarity constraints . to fully describe the energy - dependence of an amplitude one has to include other reaction channels in a coupled - channel approach .
such analyses have been developed by the bonn - gatchina group@xcite , at jlab@xcite , at jlich@xcite and other groups .
the data sets with the highest impact on resonance amplitudes in the mass range above 1.7 gev have been kaon - hyperon production using a spin - polarized photon beam and where the polarization of the @xmath10 or @xmath11 is also measured .
the high precision cross section and polarization data @xcite provide nearly full polar angle coverage and span the @xmath12 invariant mass range from threshold to 2.9 gev , hence covering the full nucleon resonance domain where new states might be discovered .
the backward angle @xmath12 data in fig.[klambda - crs ] show clear resonance - like structures at 1.7 gev and 1.9 gev that are particularly prominent and well - separated from other structures at backward angles , while at more forward angles ( not shown ) t - channel processes become prominent and dominate the cross section . the broad enhancement at 2.2 gev may also indicate resonant behavior although it is less visible at more central angles with larger background contributions .
the @xmath13 channel also indicates significant resonant behavior as seen in fig .
[ ksigma - crs ] .
the peak structure at 1.9 gev is present at all angles with a maximum strength near 90 degrees , consistent with the behavior of a @xmath14 p - wave .
other structures near 2.2 to 2.3 gev are also visible .
still , only a full partial wave analysis can determine the underlying resonances , their masses and spin - parity .
the task is somewhat easier for the @xmath15 channel , as the iso - scalar nature of the @xmath10 selects isospin-@xmath16 states to contribute to the @xmath15 final state , while both isospin-@xmath16 and isospin-@xmath17 states can contribute to the @xmath18 final state . these cross section data together with the @xmath10 and @xmath19
recoil polarization and polarization transfer data to the @xmath10 and @xmath19 had strong impact on the discovery of several new nucleon states .
they also provided new evidence for several candidate states that had been observed previously but lacked confirmation as shown in fig .
[ pdg2014 ] .
it is interesting to observe that five of the observed nucleon states have nearly degenerate masses near 1.9 gev .
similarly , the new @xmath0 state appears to complete a mass degenerate multiplet near 1.9 gev as well .
there is no obvious mechanism for this apparent degeneracy .
nonetheless , all new states may be accommodated within the symmetric constituent quark model based on @xmath20 symmetry group as far as quantum numbers are concerned . as discussed in section [ intro ] for the case of the roper resonance @xmath3
, the masses of all pure quark model states need to be corrected for dynamical coupled channel effects to compare them with observed resonances .
the same applies to the recent lattice qcd predictions @xcite for the nucleon and delta spectrum . in the mass range above 2.0 gev resonances
tend to decouple from simple final states like @xmath21 , @xmath22 , and @xmath15 .
we have to consider more complex final states with multi - mesons or vector mesons , such as @xmath23 , @xmath24 , and @xmath25 .
the study of such final states adds significant complexity as more amplitudes can contribute to spin-1 mesons photoproduction compared to pseudo - scalar meson production . as is the case for @xmath22 production ,
the @xmath23 channel is selective to isospin @xmath26 nucleon states only .
clas has collected a tremendous amount of data in the @xmath27 @xcite and @xmath28 @xcite final states on differential cross sections and spin - density matrix elements that are now entering into the more complex multi - channel analyses such as bonn - gatchina @xcite . the clas collaboration performed a single channel event - based analysis , whose results are shown in fig .
[ omega ] , and provide further evidence for the @xmath29 .
photoproduction of @xmath30 mesons is also considered a potentially rich source of new excited nucleon states in the mass range above 2 gev .
some lower mass states such as @xmath4 may have significant @xmath31 components @xcite .
such components may result in states coupling to @xmath28 with significant strength above threshold .
differential cross sections and spin - density matrix elements have been measured for @xmath32 in a mass range up to nearly 3 gev . in fig .
[ phi ] structures are seen near 2.2 gev in the forward most angle bins and at very backward angles for both decay channels @xmath33 and @xmath34 , and with the exception of the smallest forward angle bin the structures are more prominent at backward angles .
only a multi - channel partial wave analysis will be able to pull out any significant resonance strength .
[ phi_forward ] shows the differential cross section @xmath35 of the most forward angle bin . a broad structure at 2.2 gev is present , but does not show the typical breit - wigner behavior of a single resonance .
it also does not fit the data in a larger angle range , which indicates that contributions other than genuine resonances may be significant .
the forward and backward angle structures may also hint at the presence of dynamical effects possibly due to molecular contributions such as diquark - anti - triquark contributions @xcite , the strangeness equivalent to the recently observed hidden charm @xmath36 states .
another process that has promise in the search for new excited baryon states , including those with isospin-@xmath37 is @xmath38 @xcite . in distinction to the vector mesons discussed above , diffractive processes do not play a role in this channel , which then should allow better direct access to s - channel resonance production .
meson photoproduction has become an essential tool in the search for new excited baryons .
the exploration of the internal structure of excited states and the effective degrees of freedom contributing to s - channel resonance excitation requires use of electron beams where the virtuality ( @xmath39 ) of the exchanged photon can be varied to probe the spatial structure ( fig .
[ su6 ] ) .
electroproduction of final states with pseudoscalar mesons ( e.g. @xmath21 , @xmath40 , @xmath15 ) have been employed with clas , leading to new insights into the scale dependence of effective degrees of freedom , e.g. meson - baryon , constituent quark , and dressed quark contributions .
several excited states , shown in fig . [ su6 ] assigned to their primary @xmath2 supermultiplets have been studied .
the @xmath41 transition is now well measured in a large range of @xmath39 @xcite .
two of the prominent higher mass states , the roper resonance @xmath3 and @xmath4 are shown in fig .
[ p11_s11 ] as representative examples @xcite from a wide program at jlab @xcite .
for these two states advanced relativistic quark model calculations @xcite and qcd calculations from dyson - schwinger equation @xcite and light cone sum rule @xcite have recently become available , for the first time employing qcd - based modeling of the excitation of the quark core .
there is agreement with the data at @xmath42 gev@xmath43 .
the calculations deviate significantly from the data at lower @xmath39 , which indicates significant non quark core effects . for the roper resonance
such contributions have been described successfully in dynamical meson - baryon models @xcite and in effective field theory @xcite .
knowledge of the helicity amplitudes in a large @xmath39 allows for the determination of the transition charge densities on the light cone in transverse impact parameter space ( @xmath44 ) @xcite .
figure [ charge_densities ] shows the comparison of @xmath3 and @xmath4 .
there are clear differences in the charge transition densities between the two states .
the roper state has a softer positive core and a wider negative outer cloud than @xmath45 and develops a larger shift in @xmath46 when the proton is polarized along the @xmath47 axis .
over the past five years eight baryon states in the mass range from 1.85 to 2.15 gev have been either discovered or evidence for the existence of states has been significantly strengthened . to a large degree this is the result of adding very precise photoproduction data in open strangeness channels to the data base that is included in multi - channel partial wave analyses , especially the bonn - gatchina pwa .
the possibility to measure polarization observables in these processes has been critical @xcite . in the mass range above 2 gev more complex processes such as vector mesons or @xmath48
may have sensitivity to states with higher masses but require more complex analyses techniques to be brought to bear .
precision data in such channels have been available for a few years but remain to be fully incorporated in multi - channel partial wave analyses processes .
the light - quark baryon spectrum is likely also populated with hybrid excitations @xcite where the gluonic admixtures to the wave function are dominating the excitation .
these states appear with the same quantum numbers as ordinary quark excitations , and can only be isolated from ordinary states due to the @xmath39 dependence of their helicity amplitudes @xcite , which is expected to be quite different from ordinary quark excitations .
this requires new electroproduction data especially at low @xmath39 @xcite with different final states and at masses above 2 gev .
despite the very significant progress made in recent years to further establish the light - quark baryon spectrum and explore the internal structure of excited states , much remains to be done .
a vast amount of precision data already collected needs to be included in the multi - channel analysis frameworks , and polarization data are still to be analyzed @xcite .
there are approved proposals to study resonance excitations at much higher @xmath39 and with higher precision at jefferson lab with clas12 @xcite that may reveal the transition to the bare quark core contributions at short distances .
i like to thank inna aznauryan and viktor mokeev for numerous discussions on the subjects discussed in this presentation .
this work was supported by the us department of energy under contract no .
de - ac05 - 06or23177 . | recent results of meson photo - production at the existing electron machines with polarized real photon beams and the measurement of polarization observables of the final state baryons have provided high precision data that led to the discovery of new excited nucleon and @xmath0 states using multi - channel partial wave analyses procedures .
the internal structure of several prominent excited states has been revealed employing meson electroproduction processes . on the theoretical front
, lattice qcd is now predicting the baryon spectrum with very similar characteristics as the constituent quark model , and continuum qcd , such as is represented in the dyson - schwinger equations approach and in light front relativistic quark models , describes the non - perturbative behavior of resonance excitations at photon virtuality of @xmath1 . in this
talk i discuss the need to continue a vigorous program of nucleon spectroscopy and the study of the internal structure of excited states as a way to reveal the effective degrees of freedom underlying the excited states and their dependence on the distance scale probed .
example.eps gsave newpath 20 20 moveto 20 220 lineto 220 220 lineto 220 20 lineto closepath 2 setlinewidth gsave .4 setgray fill grestore stroke grestore |
heii recombination emission indicates the presence of very hard ionizing radiation with photon energies @xmath2 54ev .
star - forming galaxies with lower metallicities tend to have larger nebular heii@xmath04686 line intensities compared to those with higher metallicities ( e.g. guseva et al .
2000 ; schaerer 2003 ) .
while nebular heii emission has been observed in some local low metallicity ( z ) starbursts ( e.g. schaerer et al .
1999 ; guseva et al .
2000 ; kehrig et al . 2004 ; thuan & izotov 2005 ) , heii - emitters are apparently more frequent among high - redshift ( z ) galaxies than for local objects .
recent work has found that @xmath2 3@xmath3 of the global galaxy population at z@xmath43 show narrow heii lines ( cassata et al . 2013 ) , while this number is much lower at z@xmath40 ( kehrig et al .
the heii lines have been suggested as a good tracer of population iii stars ( popiii - stars ; the first very hot metal - free stars ) in high - z galaxies ( e.g. schaerer 2003 , 2008 ) .
these stars , which should produce a large amount of hard ionizing radiation , are believed to have contributed significantly to the universe s reionization , a challenging subject in contemporary cosmology ( e.g. bromm 2013 ) . before interpreting the emission - line spectra of distant star - forming galaxies ,
it is crucial to understand the formation of high - ionization lines in the nearby universe . the ideal place to perform this study is in extremely metal - poor nearby galaxies with nebular heii emission , which are the natural local counterparts of distant heii - emitters .
in this regard , we have been carrying out a programme to investigate nearby low - z starburst systems using the integral field spectroscopy technique ( e.g. kehrig et al . 2008,2013 ; perez - montero et al .
2011,2013 ) . as a part of this programme
, we have recently obtained new deep integral field spectroscopic ( ifs ) data of izw18 .
this is a nearby ( d = 18.2 mpc ; aloisi et al .
2007 ) hii galaxy , well known for its extremely low z @xmath4 1/32 solar ( e.g. vlchez & iglesias - pramo 1998 ) , which keeps izw18 among the three most metal - poor galaxies known in the local universe ( e.g. thuan et al . 2004 ) .
its observational characteristics make izw18 an excellent local analogue of primeval systems ( see e.g. lebouteiller et al . 2013 and references therein ) .
the presence of the nebular heii@xmath04686 line in the spectrum of izw18 has been reported before , though the precise location and extension of this heii@xmath04686 emission is not known ( e.g. garnett et al .
1991 ; izotov et al .
1997 ; legrand et al . 1997 ; vlchez & iglesias - pramo 1998 ) .
our unique ifs data unveil for the first time the entire heii@xmath04686-emitting region and its structure in izw18 ( see section 3.2 ) . despite various attempts to explain it
, the origin of the nebular heii emission in hii galaxies / regions still remains difficult to understand in many cases ; several potential mechanisms ( e.g. hot wolf - rayet ( wr ) stars , shocks from supernovae remnants , x - ray sources ) are proposed to account ( in part or fully ) for the heii ionization in these objects ( e.g. garnett et al .
1991 ; schaerer 1996 ; dopita & sutherland 1996 ; cervio , mas - hesse & kunth 2002 ; thuan & izotov 2005 ; kehrig et al . 2011 ; shirazi & brinchmann 2012 ) .
though hot wrs have been suggested before as the source of heii - ionizing photons in izw18 ( e.g. izotov et al .
1997 ; de mello et al .
1998 ) , the main mechanism powering the nebular heii emission in this galaxy is still an open issue . in this letter , using new ifs data ,
we derive for the first time the total heii@xmath04686-ionizing flux in izw18 and provide new clues to constrain the sources of high - ionization .
we carried out new ifs observations of izw18 using the potsdam multi - aperture spectrophotometer ( pmas ; roth et al .
2005 ) on the 3.5 m telescope at the calar alto observatory ( almeria , spain ) .
the data were taken in 2012 december with a typical seeing of 1 `` .
each spaxel has a spatial sampling of 1@xmath51 '' on the sky resulting in a field - of - view ( fov ) of 16@xmath516 ( @xmath4 1.4 kpc @xmath5 1.4 kpc on izw18 ; see figure [ pmas_ifu ] ) .
one pointing of izw18 , encompassing its main body which hosts the two brightest stellar clusters ( referred to as the nw and se knots ) , was taken during 2.5 hours integration split into six exposures of 1500 seconds each .
we used the v500 grating which covers from @xmath4 3640 to 7200 and provides a linear dispersion of @xmath4 2 / pixel , and full width at half - maximum effective spectral resolution of @xmath4 3.6 .
calibration images ( exposures of standard star , arc and continuum lamps ) were also obtained .
the data reduction was performed following the procedure described in kehrig et al .
( 2013 ) . from palomar ,
green = far - uv / galex , red = sdss r ) .
the box represents the fov ( 16@xmath516 ) of the pmas spectrograph over the galaxy main body and the extended h@xmath6 halo .
the pmas fov is centered on the coordinates ra(j2000.0)= 09@xmath7:34@xmath8:[email protected] and dec.(j2000.0)= + 55@xmath10:14@xmath11:25@xmath12,width=302 ]
the emission - line fluxes were measured using the iraf task splot .
the flux of each emission line was derived by integrating between two points given by the position of a local continuum placed by eye . for each line
, this procedure was repeated several times by varying the local continuum position .
the final flux of each line and its associated uncertainty were assumed to be the average and standard deviation of the independent , repeated measurements ( e.g. kehrig et al .
2006 ) .
we used our own idl scripts to create the emission - line maps presented in figure [ maps ] .
the spaxels where we measure heii@xmath04686 are indicated with pluses
. 4425 - 5100 ) of the integrated spectrum of the heii@xmath04686-emitting region of izw18.,width=321 ] an extended heii@xmath04686-emitting region with a diameter of @xmath13 5 ( @xmath13 440 pc ) is revealed from our ifs data ( see figure [ maps ] ) .
the narrow line profile for the heii@xmath04686 emission and its spatial extent are evidence for its nebular nature .
the spectra of certain of stars may exhibit somewhat narrow heii@xmath04686 lines ( e.g. massey et al . 2004 ) , so heii emission observed in starburst galaxies could be thought to arise in the atmospheres of such stars ( e.g. bergeron 1977 ) .
if these of stars were present in appreciable numbers in izw18 , broad h@xmath6 emission would however be expected ( massey et al . 2004 ) .
thus , the non - detection of this feature in our spectra supports the nebular origin for the heii@xmath04686 line in izw18 . by adding the emission from the spaxels showing heii@xmath04686 ( see figure [ maps ] )
, we created the one - dimensional spectrum for the heii@xmath04686 region ( see figure [ spectrum ] ) . using this spectrum
, we obtained a very small logarithmic extinction coefficient c(h@xmath1 ) = 0.08 @xmath14 0.02 from the observed ratio h@xmath6/h@xmath1 = 2.92 @xmath14 0.04 , assuming an intrinsic case b recombination h@xmath6/h@xmath1 = 2.75 ( osterbrock & ferland 2006 , of06 ) . using pyneb ( luridiana , morisset & shaw 2015 ) ,
we derived the electron temperature , @xmath15 = ( 2.18 @xmath14 0.05)@xmath510@xmath16 k , and density , @xmath17 @xmath18 100 @xmath19 , from the [ oiii]@xmath04363/[oiii]@xmath0@xmath04959,5007 and [ sii]@xmath06717/[sii]@xmath06731 ratios , respectively .
the integrated flux of the heii@xmath04686 line , corrected for reddening , is ( 2.84 @xmath14 0.18)@xmath510@xmath20 erg s@xmath21 @xmath22 which translates to a heii@xmath04686 luminosity of l@xmath23 = ( 1.12 @xmath14 0.07)@xmath510@xmath24 erg s@xmath21 .
the corresponding heii ionizing photon flux , q(heii)@xmath25 = ( 1.33 @xmath14 0.08)@xmath510@xmath26 photon s@xmath21 , was derived from the measured l@xmath23 using the relation q(heii ) = l@xmath23/[j(@xmath04686)/@xmath27(heii ) ] ( assuming case b recombination , and @xmath15 = 2@xmath510@xmath16 k ; of06 ) .
we checked that different cloudy models ( ferland et al . 2013 ) computed for the typical @xmath17 and @xmath15 in the heii@xmath04686 region , and considering several effective temperatures and geometries with no dust , give a ratio l(heii@xmath04686)/q(heii ) which agrees with the assumed ratio ( of06 ) within 10@xmath3 .
the total q(heii)@xmath25 , a quantity not reported before for izw18 , will allow us to constrain possible ionizing sources of heii in izw18 .
one widely favored mechanism for heii ionization in hii galaxies involves hot wrs ( e.g. schaerer 1996 ) . nevertheless , it has been demonstrated that nebular heii@xmath04686 does not appear to be always associated with wrs , as it is the case of the heii nebulae lmc - n44c , lmc - n159f and m33-bclmp651 , among others ( kehrig et al . 2011 and references therein ) .
this indicates that wrs can not explain heii ionization in all cases , particularly at low - z ( e.g. guseva et al . 2000 ; kehrig et al .
2008,2013 ; schaerer 2003 ; shirazi & brinchmann 2012 ) .
besides , it has been shown that , for a sample of heii@xmath04686-emitting star - forming galaxies , current models of massive stars predict heii@xmath04686 emission , and heii@xmath04686/h@xmath1 ratios , only for z @xmath28 0.20 z@xmath29 instantaneous bursts ( shirazi & brinchmann 2012 ) . in the particular case of izw18 , previous work has claimed that the ratio heii@xmath04686/h@xmath1 could be reproduced using highly density - bounded photoionization models while underpredicting the electron temperature measurement ( stasiska & schaerer 1999 ) ; these models have been challenged by pquignot ( 2008 ) .
faint broad emission signatures , attributable to wrs , are observed in the spectrum of izw18 despite its low - z ( e.g. izotov et al .
1997 , legrand et al .
1997 , brown et al . 2002 ) .
a comprehensive study of wrs in izw18 , using uv stis spectroscopy , revealed signatures of carbon - type wrs ( wc ) in two clusters : one in the nw star - forming region and a second one on the outskirts of this region ( brown et al .
2002 ) . here
we deal with the nw one , since it is in the heii@xmath04686-emitting region ( see figure [ maps ] ) .
the civ@xmath01550 flux measured from the nw wr cluster ( brown et al . 2002 ) gives a civ@xmath01550 luminosity of l@xmath30 = 4.67@xmath510@xmath31 erg s@xmath21 .
taking the l@xmath30 luminosity of the metal - poor early - type wc ( wce ) model by crowther & hadfield ( 2006 , ch06 ) , which mimics a single wce star in izw18 , this implies @xmath4 9 izw18-like wce stars present in the nw cluster . from these 9 wce stars a total flux q(heii )
= 2.8 @xmath510@xmath32 photon s@xmath21 is expected [ assuming q(heii ) = 10@xmath33 photon s@xmath21 for one izw18-like wce ; ch06 ] , i.e. about 48 times lower than the q(heii)@xmath25=([email protected])@xmath510@xmath26 photon s@xmath21 , derived from our data ( see section 3.2 ) . based on the heii - ionizing flux expected from these izw18-like wrs
, a very large wr population is required to explain the heii - ionization budget measured ; for instance , taking the q(heii)= 10@xmath33 photon s@xmath21 for one izw18-like wce ( ch06 ) , the number of these wces needed to explain our derived q(heii)@xmath25 = ( [email protected])@xmath510@xmath26 photon s@xmath21 would be @xmath28 400 . in principle , the presence of hundreds of wrs in izw18 should not be discarded on the basis of empirical arguments for reduced wr line luminosity at low - z ( ch06 ) . however , assuming a salpeter initial mass function ( imf ; salpeter 1955 ; m@xmath34=150m@xmath35 ) and the initial mass needed for a star to certainly become a wc ( maeder & maynet 2005 ) , a cluster with @xmath28 8 times the total stellar mass of the nw region ( m@xmath36 = 2.9@xmath510@xmath37 m@xmath29 from stasiska & schaerer ( 1999 ) scaled to 18.2 mpc distance ) is required to provide @xmath28 400 wc in izw18 . also , such a high number of wrs is not supported by state - of - the - art stellar evolutionary models for single ( rotating and non - rotating ) massive stars in metal - poor environments ( leitherer et al .
furthermore , given the decrease in the ratio of wr / o stars with decreasing metallicity , shown by observations and theoretical models ( maeder & meynet 2012 and references therein ) , such a large number of wrs appears clearly unreasonable considering the extremely low - z and the o star content of izw18 ( ch06 ) .
all this suggests that wrs are not the sole responsible of the heii@xmath04686 emission in izw18 .
the binary channel in massive star evolution is suggested to increase the wr population ( e.g. eldridge et al . 2008 ) , but the wr population in local group galaxies does not show an increased binary rate at lower - z ( e.g. foellmi et al . 2003 ; neugent & massey 2014 ) ; thus the binary channel does not seem to favour the formation of wrs at lower - z , in contrast to what we need .
nevertheless , we should bear in mind the possible uncertainties yet unsolved in the models ( maeder & meynet 2012 ) .
further investigation awaits the calculation of evolutionary models for binary stars at very low - z . as mentioned before
, nebular heii@xmath04686 emission observed in the spectra of hii galaxies and extragalactic hii regions has often been attributed also to shocks and x - ray sources ( e.g. pakull & mirioni 2002 ; garnett et al .
1991 ; thuan & izotov 2005 ) . in
the following we discuss on these two candidate sources for heii ionization in izw18 .
the x - ray emission from izw18 is dominated by a single x - ray binary apparently located in the field of the nw knot ( thuan et al .
we have computed a cloudy photoionization model ( ferland et al .
2013 ) using as input a power - law spectral energy distribution ( sed ) with the same x - ray luminosity , column density and slope that have been reported for izw18 ( thuan et al . 2004 ) .
this cloudy model gives as an output a heii@xmath04686 luminosity l@xmath23 = 10@xmath38 erg s@xmath21 which is @xmath4 100 times lower than the l@xmath23 measured .
this result rules out the x - ray binary as the main source of heii ionizing photons in izw18 .
we note here that the emission from x - ray ionized nebulae has been successfully reproduced by cloudy models before ( e.g. pakull & mirioni 2002 ) .
guided by the existence of heii@xmath04686 emission associated with supernova remnants ( e.g. kehrig et al .
2011 ) we explored the conjecture that the heii@xmath04686 region represents such a shock - ionized nebula .
the [ oi]@xmath06300 line , often strong in remnants , has been frequently used as a sensitive shock - emission test ( e.g. skillman 1985 ) .
we checked that , in fact , most of the [ oi]@xmath06300 emission in izw18 is concentrated on the se knot ( see the [ oi]@xmath06300 map in figure [ maps ] ) , with only 12@xmath3 of the heii@xmath04686-emitting spectra showing [ oi]@xmath06300 flux above the 3@xmath39 detection limit .
additionally , we find no evidence for [ sii ] enhancement ( a usual sign of shock excitation ; e.g. dopita & sutherland 1996 ) associated to the location of the heii@xmath04686 region ( see the [ sii ] map in figure [ maps ] ) . therefore , the heii@xmath04686-emitting zone in izw18 is unlikely to be produced by shocks .
our new observations have allowed us to empirically demonstrate why conventional heii - ionizing sources ( e.g. wrs , shocks , x - ray binaries ) can not account for the total heii - ionization budget in izw18 .
what could the nebular heii@xmath04686 emission in izw18 originate from ?
we have also explored the possibilty of very massive , metal - poor o stars to account for our observations of izw18 . using current wind models of very massive o stars at low - z we can derive their heii - ionizing fluxes ( kudritzki 2002 , k02 ) . according to the hottest models ( @xmath15=60,000 k ) , between 10 - 20 super - massive stars with 300m@xmath29 [ with q(heii ) @xmath13 ( 0.70 - 1.4)@xmath510@xmath40 photon
s@xmath21 each ] would be enough to explain the derived q(heii)@xmath25 budget .
very massive stars of up to 300m@xmath29 were claimed to exist in the lmc - r136 cluster ( crowther et al .
2010 ) ; however the existence of such super - massive 300m@xmath29 stars remains heavily debated ( vink 2014 ) . additionally , assuming a salpeter imf , 10 - 20 stars with 290 @xmath18 m@xmath41/m@xmath29 @xmath18 310 would imply a cluster mass of @xmath4 10 - 20 @xmath5 m@xmath36 .
we should bear in mind that an extrapolation of the imf predicting 300m@xmath29 stars remains unchecked up to now . if we consider instead the 150m@xmath29 star hottest models [ with q(heii )
@xmath18 1.9@xmath510@xmath42 photons@xmath21 each , for z @xmath18 1/32 z@xmath29 ; k02 ] , the number of these stars required to explain the q(heii)@xmath25 would be @xmath28 650 . for a salpeter imf ,
650 stars with 145 @xmath18 m@xmath41/m@xmath29 @xmath18 155 would require a cluster mass @xmath4 200 @xmath5 m@xmath36 .
besides the q(heii)@xmath25 budget , in the heii@xmath04686 region we have measured heii@xmath04686/h@xmath1 ratios as high as 0.08 .
these values appear too big to be explained even by the models for the hottest , most metal - poor super - massive 300m@xmath29 stars ( k02 ) under ionization - bounded conditions .
further constraints to the observations should await the calculation of new evolutionary tracks and seds for single o stars at the metallicity of izw18 including rotation .
searches for very metal - poor starbursts and popiii - hosting galaxies have been carried out in the distant universe using the heii lines ( e.g. schaerer 2008 ; cassata et al .
this search is based on the high effective temperature for popiii - stars which will emit a large number of photons with energy above 54ev , and also on the expected increase of the heii recombination lines with decreasing z ( e.g. guseva et al .
2000 ; schaerer 2003,2008 ) .
predictions for burst models of different metallicities show how their corresponding q(heii ) can increase by up to @xmath4 10@xmath43 when going from z=10@xmath44 to 0 ( schaerer 2003 ) .
these models can not explain the q(heii)@xmath25 when z @xmath2 10@xmath44 [ for a salpeter imf , m@xmath34 = 100m@xmath29 ; see table 3 in schaerer 2003 ] even assuming that the total m@xmath36 would come from heii - ionizing stars .
so another more speculative possibility , to explain the derived q(heii)@xmath25 could be based on nearly metal - free ionizing stars .
these stars should ionize heii via their strong uv radiation expected at nearly zero metallicity ( e.g. tumlinson & shull 2000 ; schaerer 2003 ) . as an approximation of nearly metal - free single stars ,
we have compared our observations with the heii - ionizing radiation expected from state - of - the - art models for rotating z=0 stars ( yoon , dierks & langer 2012 ) . according to these models
, we found that a handful of such stars could explain our derived q(heii)@xmath25 [ e.g. @xmath4 8 - 10 stars with mass m@xmath45=150m@xmath29 or @xmath4 13 - 15 stars with m@xmath45=100m@xmath29 ; with q(heii ) @xmath13 1.4@xmath510@xmath40 ( 0.9@xmath510@xmath40 ) photon s@xmath21 for each star with m@xmath45 = 150m@xmath29(100m@xmath29 ) ] .
additionally , we note that the ionizing spectra produced by these star models are harder than the ones expected from the hottest models of super - massive 300m@xmath29 stars ( k02 ) , so they would also explain the highest heii@xmath04686/h@xmath1 values observed , providing that ionization - bounded conditions are met . while gas in izw18 is very metal - deficient but not primordial , lebouteiller et al .
( 2013 ) have pointed out that the hi envelope of izw18 near the nw knot contains essentially metal - free gas pockets .
these gas pockets could provide the raw material for making such nearly metal - free stars .
clearly , in this hypothetical scenario , these extremely metal - poor stars can not belong to the nw cluster , which hosts more chemically evolved stars .
this letter reports on new optical ifs observations of the nearby dwarf galaxy izw18 .
this is an extremely metal - poor system , which is our best local laboratory for probing the conditions dominating in distant low - z starbursts .
our ifs data reveal for the first time the total spatial extent ( @xmath13 440 pc diameter ) of the heii@xmath04686-emitting region and corresponding total heii - ionizing photon flux in izw18 .
the metal - poor sensitivity of the heii line is a primary motivation to develop diagnostics for unevolved starbursts , and strong nebular heii emission is expected to be one of the best signatures of massive popiii - stars ( e.g. schaerer 2003,2008 ) .
heii emission has been observed to be more frequent at higher - z than locally ( kehrig et al .
2011 ; cassata et al .
thus the analysis of the origin of the heii@xmath04686 nebular line in relatively close ionized regions , which can be studied in more detail , can yield insight into the ionizing sources in the distant universe .
our observations combined with stellar models predictions point out that conventional excitation sources ( e.g. wrs , shocks , x - ray binaries ) can not convincingly explain the heii - ionizing energy budget derived for izw18 .
other mechanisms are probably also at work
. if the heii - ionization in izw18 is due to stellar sources , these might be peculiar very hot stars ( perhaps uncommon in local starbursts but somewhat more frequent in distant galaxies ) : according to theoretical stellar models , either super - massive o stars or nearly metal - free ionizing stars could in principle account for the total q(heii)@xmath25 of izw18 .
however the super - massive o stars scenario would imply a cluster mass much higher than the mass of the nw knot derived from observations . on the other hand , though metal - free gas pockets were previously reported in izw18 ( lebouteiller et al .
2013 ) , we highlight that the existence of nearly metal - free ionizing stars is not yet confirmed observationally
. the work presented here can help in the preparation of prospective searches for primeval objects , one of the main science drivers for next - generation telescopes ( e.g. bromm 2013 ) .
this work has been partially funded by research projects aya2010 - 21887-c04 - 01 from the spanish pnaya , and pex2011-fqm7058 from junta de andalucia .
fd and dk gratefully acknowledge support from the centre national detudes spatiales .
we thank the anonymous referee for a constructive report which improved this letter .
we express our appreciation to leslie sage for his help and suggestions .
we also would like to thank manfred pakull for his useful comments on this letter . | new integral field spectroscopy has been obtained for izw18 , the nearby lowest - metallicity galaxy considered our best local analog of systems forming at high - z . here
we report the spatially resolved spectral map of the nebular heii@xmath04686 emission in izw18 , from which we derived for the first time its total heii - ionizing flux .
nebular heii emission implies the existence of a hard radiation field .
heii - emitters are observed to be more frequent among high - z galaxies than for local objects .
so investigating the heii - ionizing source(s ) in izw18 may reveal the ionization processes at high - z .
heii emission in star - forming galaxies , has been suggested to be mainly associated with wolf - rayet stars ( wrs ) , but wrs can not satisfactorily explain the heii - ionization at all times , in particular at lowest metallicities .
shocks from supernova remnants , or x - ray binaries , have been proposed as additional potential sources of heii - ionizing photons .
our data indicate that conventional heii - ionizing sources ( wrs , shocks , x - ray binaries ) are not sufficient to explain the observed nebular heii@xmath04686 emission in izw18 .
we find that the heii - ionizing radiation expected from models for either low - metallicity super - massive o stars or rotating metal - free stars could account for the heii - ionization budget measured , while only the latter models could explain the highest values of heii@xmath04686/h@xmath1 observed .
the presence of such peculiar stars in izw18 is suggestive and further investigation in this regard is needed .
this letter highlights that some of the clues of the early universe can be found here in our cosmic backyard . |
in paper @xcite , we surveyed period variations of superhumps in su uma - type dwarf novae ( for general information of su uma - type dwarf novae and superhumps , see @xcite ) .
@xcite indicated that evolution of superhump period ( @xmath3 ) is generally composed of three distinct stages : early evolutionary stage with a longer superhump period ( stage a ) , middle stage with systematically varying periods ( stage b ) , final stage with a shorter , stable superhump period ( c ) .
it was also shown that the period derivatives ( @xmath4 ) during stage b is correlated with @xmath3 , or binary mass - ratios ( @xmath5 ) .
although this relation commonly applies to classical su uma - type dwarf novae , wz sge - type dwarf novae , a subtype of su uma - type dwarf novae with very infrequent superoutbursts , tend to deviate from this picture : they rarely show a distinct stage b c transition , and
some objects show relatively small period derivatives , and they frequently exhibit unusual multiple post - superoutburst rebrightenings . the origin of these relations is not yet well understood .
we have extended the survey in order to test whether this picture is applicable to new su uma - type systems and to the degree of diversity between systems in a larger sample .
we include in this paper newly recorded objects and superoutbursts since the publication of @xcite .
some of new observations have led to revisions of analysis in @xcite .
we also include a few past superoutburst not analyzed in the previous studies .
the structure of the paper follows the scheme in @xcite , in which we mostly restricted to superhump timing analysis .
we also include some more details if the paper provides the first solid presentation of individual objects .
the data were obtained under campaigns led by the vsnet collaboration @xcite . in some objects , we used archival data for published papers , and the public data from the aavso international databasehttp://www.aavso.org / data / download/@xmath6 . ] .
the majority of the data were acquired by time - resolved ccd photometry by using 30 cm - class telescopes , whose observational details on individual objects will be presented in future papers dealing with analysis and discussion on individual objects .
the list of outbursts and observers is summarized in table [ tab : outobs ] .
the data analysis was performed just in the same way described in @xcite . in this paper
, we introduce bayesian extension to traditional method ( appendix ) , and used them if they give significantly improved results . the derived @xmath3 , @xmath7 and other parameters are listed in table [ tab : perlist ] as in the same format in @xcite .
the definitions of parameters @xmath8 and @xmath7 are the same as in @xcite . as in @xcite , we present comparisons of @xmath0 diagrams between different superoutbursts since this has been one of the motivations of these surveys ( cf .
@xcite ) .
ccccc subsection & object & year & observers or references & i d + [ obj : kxaql ] & kx aql & 2010 & mhh , aavso , kis , got , dpv , oku & + [ obj : nncam ] & nn cam & 2009 & aavso , dem , mhh , jsh , imi & + [ obj : v591cen ] & v591 cen & 2010 & mlf , gbo & + [ obj : zcha ] & z cha & 2010 & aavso , kis & + [ obj : pucma ] & pu cma & 2009 & ous , mhh , kis & + [ obj : aqcmi ] & aq cmi & 2010 & gbo & + [ obj : gzcnc ] & gz cnc & 2010 & oku , oao , mhh , dem , ous & + [ obj : gocom ] & go com & 2010 & dem , mhh , oku , dpv & + [ obj : tvcrv ] & tv crv & 2009 & kis & + [ obj : v337cyg ] & v337 cyg & 2010 & imi , ku , oku , dpv & + [ obj : v1113cyg ] & v1113 cyg & 2003 & @xcite & + [ obj : v1454cyg ] & v1454 cyg & 2009 & hmb , mhh , imi , aavso & + [ obj : aqeri ] & aq eri & 2010 & nel , mhh , kis , ioh & + [ obj : vxfor ] & vx for & 2009 & mlf , nyr , nel , aavso , sto , mhh , kis & + [ obj : awgem ] & aw gem & 2010 & hyn & + [ obj : irgem ] & ir gem & 2010 & dem & + [ obj : v592her ] & v592 her & 2010 & aavso , ous , mhh , oku , imi , dpv , rit , hmb , ku , boy , oao , mev & + [ obj : v660her ] & v660 her & 2009 & aavso & + [ obj : v844her ] & v844 her & 2009 & dpv , aavso & + & v844 her & 2010 & mhh & + [ obj : cthya ] & ct hya & 2010 & mhh & + [ obj : v699oph ] & v699 oph & 2010 & dem , ous & + [ obj : v1032oph ] & v1032 oph & 2010 & mhh , dem , aavso , kis , rui & + [ obj : v2051oph ] & v2051 oph & 2010 & sto , ous , kis & + [ obj : efpeg ] & ef peg & 2009 & mhh , imi , ous , aavso & + [ obj : v368peg ] & v368 peg & 2009 & ctx , aavso , imi , nyr , hmb & + [ obj : uvper ] & uv per & 2010 & dpv , pxr , dem , aavso & + [ obj : eipsc ] & ei psc & 2009 & got , kis & + [ obj : ektra ] & ek tra & 2009 & nel & + [ obj : suuma ] & su uma & 2010 & dpv , sac & + [ obj : bcuma ] & bc uma & 2009 & nyr , mhh & + [ obj : eluma ] & el uma & 2010 & mhh , ioh , dpv , hmb & + [ obj : iyuma ] & iy uma & 2009 & @xcite , dpv , aavso & + [ obj : ksuma ] & ks uma & 2010 & ous & + [ obj : mruma ] & mr uma & 2010 & ous , oku & + [ obj : tyvul ] & ty vul & 2010 & mev , aavso , imi , bst & + [ obj : j0423 ] & 1rxs j0423 & 2010 & ous , mhh , dem & + [ obj : j0532 ] & 1rxs j0532 & 2009 & dpv , mhh & + [ obj : asas2243 ] & asas j2243 & 2009 & hmb , kra , mhh , jsh , gbo , imi , dem , ioh , aavso & + [ obj : lanning420 ] & lanning 420 & 2010 & ku , aavso , dem , bxs , mhh , hmb , rui & + [ obj : pg0149 ] & pg 0149 & 2009 & imi , hmb , mhh & + [ obj : j1715 ] & rx j1715 & 2009 & @xcite , mhh & + [ obj : j0129 ] & sdss j0129 & 2009 & jsh , mhh , aavso , boy & + [ obj : j0310 ] & sdss j0310 & 2009b & imi & + [ obj : j0732 ] & sdss j0732 & 2010 & aavso , jsh & + [ obj : j0838 ] & sdss j0838 & 2010 & mhh , dem & + [ obj : j0839 ] & sdss j0839 & 2010 & dem , jsh & + [ obj : j0903 ] & sdss j0903 & 2010 & dem , vol , aavso , mas , mhh & + + + + + + + + + + + [ cols="^,^,^,^,^",options="header " , ]
kx aql has long been known as a dwarf nova with a low outburst frequency . @xcite and @xcite reported two outbursts in 1967 and 1972 from aavso observations .
@xcite examined photographic plate archives and estimated an interval between outbursts to be @xmath9 1000 d. the object underwent a very bright ( @xmath10 ) outburst in 1980 november , which was first recorded by s. fujino on november 9 .
the long duration of the outburst inferred from the vsolj and aavso observations is sufficient to imply the su uma - type nature .
additional short outbursts were recorded in 1994 decemberhttp://www.kusastro.kyoto - u.ac.jp / vsnet / mail/1994/vsnet - obs / msg00155.html@xmath6 . ] and 2007 may ( baavss - alert 582 ) , which later turned out to be a normal outburst .
@xcite first presented a spectrum of this object clearly showing a dwarf nova at a low mass - transfer rate .
@xcite listed kx aql as a candidate for a wz sge - type dwarf nova .
@xcite obtained a @xmath2 of 0.06035(3 ) d from a radial - velocity study .
the long - waited superoutburst was finally recorded in 2010 march by t. gomez ( cvnet - outburst 3626 ) .
subsequent observations detected superhumps ( vsnet - alert 11859 , 11862 , 11884 )
. the outburst was followed by a rebrightening ( vsnet - alert 11895 ) .
the times of superhump maxima are listed in table [ tab : kxaqloc2010 ] .
there was an apparent break in the period evolution around @xmath11 .
we identified this break as a stage b
c transition since the observed duration of the outburst after the break was only 7 d ( in contrast to @xmath914 d for typical durations of superoutbursts of short-@xmath2 su uma - type dwarf novae ) and the superhump period was almost constant after the break .
the relatively low recorded maximum brightness ( @xmath12 ) for a superoutburst also suggests that the true maximum was missed . the period for stage b
was not determined due to the shortness of the observed segment .
we adopted a mean superhump period of 0.06128(2 ) d with the pdm method ( figure [ fig : kxaqlshpdm ] ) .
the recorded @xmath13 of 1.5 % is slightly low for this @xmath2 ( cf .
figure 15 in @xcite ) , but not as low as those of wz sge - type dwarf novae with similar@xmath2 .
this interpretation is consistent with the presence of a single rebrightening , rather than multiple ones in wz sge - type systems with similar @xmath2 .
the scaled @xmath14 is 0.08 based on the refined relation in @xcite .
( 88mm,110mm)fig1.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55265.6366 & 0.0007 & @xmath170.0110 & 145 + 1 & 55265.6941 & 0.0018 & @xmath170.0149 & 52 + 7 & 55266.0852 & 0.0025 & 0.0077 & 56 + 17 & 55266.6967 & 0.0028 & 0.0050 & 53 + 26 & 55267.2439 & 0.0010 & @xmath170.0005 & 183 + 27 & 55267.3082 & 0.0005 & 0.0023 & 148 + 38 & 55267.9827 & 0.0004 & 0.0012 & 59 + 43 & 55268.2993 & 0.0007 & 0.0108 & 74 + 50 & 55268.7192 & 0.0014 & 0.0008 & 43 + 54 & 55268.9648 & 0.0003 & 0.0008 & 126 + 55 & 55269.0288 & 0.0007 & 0.0033 & 46 + 59 & 55269.2728 & 0.0006 & 0.0016 & 238 + 70 & 55269.9473 & 0.0004 & 0.0006 & 53 + 71 & 55270.0091 & 0.0008 & 0.0010 & 60 + 75 & 55270.2611 & 0.0025 & 0.0073 & 127 + 76 & 55270.3120 & 0.0017 & @xmath170.0032 & 113 + 82 & 55270.6850 & 0.0019 & 0.0013 & 52 + 87 & 55270.9918 & 0.0026 & 0.0010 & 7 + 99 & 55271.7243 & 0.0026 & @xmath170.0035 & 44 + 103 & 55271.9691 & 0.0008 & @xmath170.0044 & 62 + 119 & 55272.9490 & 0.0017 & @xmath170.0072 & 16 + + + + the superoutburst in 2009
november is the first superoutburst of this object whose entire evolution was first observed in detail [ for general information of this object , see @xcite ] .
the times of superhump maxima are listed in table [ tab : nncamoc2009 ] .
the @xmath0 diagram clearly shows all a c stages .
there was a rise in the light curve in accordance with the stage a b transition ( figure [ fig : nncam2009oc ] ) . the mean @xmath3 and @xmath7 for stage b ( @xmath18 ) were 0.07426(1 ) d and @xmath19 , respectively .
there was a distinct stage b
c transition around @xmath15 = 87 .
the mean period for stage a was calculated for @xmath20 although there was significant period shortening even during stage a. @xcite also recently reported on two superoutbursts of this object in 2007 and 2009 .
@xcite did not distinguish stages a
c during the 2009 superoutburst treated in this paper probably due to fragmentary observational coverage .
( 88mm,90mm)fig2.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55137.7356 & 0.0010 & @xmath170.0396 & 71 + 1 & 55137.8129 & 0.0019 & @xmath170.0366 & 69 + 2 & 55137.8989 & 0.0009 & @xmath170.0249 & 71 + 12 & 55138.6733 & 0.0006 & 0.0064 & 72 + 13 & 55138.7472 & 0.0004 & 0.0060 & 70 + 14 & 55138.8239 & 0.0003 & 0.0084 & 67 + 15 & 55138.8974 & 0.0003 & 0.0076 & 63 + 19 & 55139.1977 & 0.0002 & 0.0107 & 109 + 20 & 55139.2719 & 0.0002 & 0.0106 & 147 + 21 & 55139.3475 & 0.0003 & 0.0119 & 90 + 39 & 55140.6834 & 0.0004 & 0.0102 & 46 + 40 & 55140.7574 & 0.0003 & 0.0099 & 69 + 41 & 55140.8316 & 0.0003 & 0.0098 & 68 + 42 & 55140.9059 & 0.0003 & 0.0098 & 63 + 61 & 55142.3164 & 0.0003 & 0.0085 & 152 + 76 & 55143.4293 & 0.0008 & 0.0068 & 68 + 77 & 55143.5065 & 0.0011 & 0.0097 & 33 + 78 & 55143.5796 & 0.0016 & 0.0084 & 25 + 79 & 55143.6544 & 0.0009 & 0.0089 & 36 + 87 & 55144.2482 & 0.0005 & 0.0083 & 140 + 103 & 55145.4305 & 0.0009 & 0.0017 & 41 + 104 & 55145.5047 & 0.0007 & 0.0015 & 39 + 105 & 55145.5767 & 0.0012 & @xmath170.0008 & 30 + 106 & 55145.6509 & 0.0007 & @xmath170.0008 & 40 + 107 & 55145.7255 & 0.0013 & @xmath170.0006 & 32 + 116 & 55146.3882 & 0.0010 & @xmath170.0066 & 65 + 117 & 55146.4641 & 0.0006 & @xmath170.0050 & 63 + 118 & 55146.5391 & 0.0005 & @xmath170.0043 & 104 + 119 & 55146.6146 & 0.0007 & @xmath170.0031 & 39 + 129 & 55147.3495 & 0.0012 & @xmath170.0113 & 33 + 130 & 55147.4291 & 0.0028 & @xmath170.0061 & 37 + 134 & 55147.7171 & 0.0013 & @xmath170.0153 & 39 + + + + v591 cen was discovered by @xcite ( see also @xcite ) .
although the discovery observation was already suggestive of a superoutburst , this object had long been overlooked due to the apparent small outburst amplitude .
the lack of cv - type signature in spectroscopic observation by @xcite even led to a non - cv classification . monitoring for outbursts ,
however , continued because of its outbursting nature and the likely presence of superoutbursts . in 2006 , b. monard succeeded in obtaining astrometry of the outbursting object , which was 3 arcsec distant from the supposed quiescent counterpart ( vsnet - alert 9158 ) .
the true v591 cen was invisible down to 20 mag on dss images .
the object again underwent a bright outburst in 2010 april ( vsnet - alert 11915 ) , and the existence of superhumps was finally confirmed ( vsnet - alert 11919 ; figure [ fig : v591censhpdm ] ) .
astrometry of the outbursting object also confirmed the identification in 2006 ( vsnet - alert 11922 ) .
the times of superhump maxima are listed in table [ tab : v591cenoc2010 ] .
the data clearly shows a positive period derivative @xmath7 = @xmath21 , typical for this short @xmath3 .
( 88mm,110mm)fig3.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55296.1418 & 0.0008 & 0.0022 & 111 + 1 & 55296.2051 & 0.0007 & 0.0052 & 109 + 2 & 55296.2634 & 0.0011 & 0.0033 & 111 + 17 & 55297.1647 & 0.0009 & 0.0001 & 110 + 18 & 55297.2252 & 0.0008 & 0.0003 & 111 + 19 & 55297.2849 & 0.0011 & @xmath170.0003 & 112 + 20 & 55297.3425 & 0.0014 & @xmath170.0030 & 70 + 86 & 55301.3271 & 0.0032 & 0.0019 & 267 + 87 & 55301.3806 & 0.0020 & @xmath170.0050 & 268 + 88 & 55301.4362 & 0.0018 & @xmath170.0097 & 267 + 89 & 55301.4977 & 0.0017 & @xmath170.0084 & 268 + 102 & 55302.2868 & 0.0039 & @xmath170.0033 & 268 + 133 & 55304.1630 & 0.0018 & 0.0036 & 61 + 134 & 55304.2236 & 0.0030 & 0.0040 & 103 + 135 & 55304.2836 & 0.0015 & 0.0036 & 329 + 136 & 55304.3447 & 0.0014 & 0.0044 & 281 + 137 & 55304.4012 & 0.0014 & 0.0006 & 264 + 138 & 55304.4612 & 0.0027 & 0.0004 & 73
+ + + + we analyzed aavso and our observations of the 2010 january superoutburst of z cha .
the observation was performed during the late stage of the superoutburst . after removing observations within 0.10 @xmath2 of eclipses for the superoutburst plateau and 0.08 @xmath2 for later observations ,
we determined times of superhump maxima ( table [ tab : zchaoc2010 ] ) . the times for @xmath22 are secondary humps , which are likely traditional late superhumps .
the peaks of persistent ordinary superhumps were not determined because they fell on orbital humps and around eclipses .
these superhumps ( @xmath23 ) most likely correspond to stage c superhumps .
the period [ 0.07698(5 ) d ] determined from the superhump maxima is close to the period [ 0.07681(6 ) d ] of stage c superhumps in 1982 @xcite .
although post - superoutburst observation suggests the appearance of traditional late superhumps , this signal was not well traced due to the strong orbital modulation and limited coverage .
a better continuous observation is necessary to test whether dominant superhumps in the post - superoutburst stage are stage c superhumps or traditional late superhumps ( see also a discussion on the nature of late superhumps in @xcite ) .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55200.0935 & 0.0008 & 0.0032 & 152 + 13 & 55201.0794 & 0.0003 & @xmath170.0086 & 209 + 14 & 55201.1575 & 0.0003 & @xmath170.0072 & 165 + 15 & 55201.2328 & 0.0001 & @xmath170.0087 & 124 + 27 & 55202.1711 & 0.0021 & 0.0086 & 24 + 28 & 55202.2385 & 0.0015 & @xmath170.0007 & 22 + 65 & 55205.0878 & 0.0006 & 0.0088 & 158 + 66 & 55205.1638 & 0.0005 & 0.0081 & 159 + 67 & 55205.2385 & 0.0007 & 0.0060 & 94 + 78 & 55206.0940 & 0.0007 & 0.0173 & 130 + 79 & 55206.1667 & 0.0011 & 0.0132 & 62 + 80 & 55206.2399 & 0.0007 & 0.0097 & 159 + 91 & 55207.0442 & 0.0015 & @xmath170.0303 & 168 + 92 & 55207.1319 & 0.0009 & @xmath170.0194 & 169 + + + + we observed the 2009 superoutburst of this object ( table [ tab : pucmaoc2009 ] ) .
the nightly superhump profiles ( figure [ fig : pucma2009prof ] ) suggests that the superhump period first increased until bjd 2455164 , and then decreased .
the stages listed in table [ tab : perlist ] reflect this interpretation .
a comparison of @xmath0 diagrams between different superoutbursts is shown in figure [ fig : pucmacomp ] ) .
it would be worth noting that the @xmath0 diagram for the 2008 superoutburst ( superoutburst preceded by a precursor ) matched others only if @xmath15 was counted from the start of the main superoutburst , rather than from the precursor .
this suggests that superhumps were excited around the ignition of the main superoutburst , rather than during the precursor .
( 88mm,110mm)fig4.eps ( 88mm,70mm)fig5.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55159.1299 & 0.0009 & @xmath170.0110 & 66 + 1 & 55159.1892 & 0.0008 & @xmath170.0096 & 65 + 35 & 55161.1572 & 0.0007 & @xmath170.0083 & 53 + 36 & 55161.2158 & 0.0003 & @xmath170.0076 & 84 + 37 & 55161.2724 & 0.0005 & @xmath170.0088 & 82 + 53 & 55162.2057 & 0.0010 & @xmath170.0011 & 193 + 54 & 55162.2609 & 0.0004 & @xmath170.0038 & 186 + 55 & 55162.3181 & 0.0005 & @xmath170.0043 & 179 + 87 & 55164.1812 & 0.0008 & 0.0076 & 85 + 88 & 55164.2404 & 0.0008 & 0.0090 & 248 + 89 & 55164.2954 & 0.0009 & 0.0061 & 249 + 90 & 55164.3611 & 0.0019 & 0.0139 & 102 + 121 & 55166.1436 & 0.0011 & 0.0032 & 150 + 122 & 55166.2110 & 0.0007 & 0.0128 & 334 + 123 & 55166.2661 & 0.0005 & 0.0100 & 263 + 124 & 55166.3215 & 0.0006 & 0.0075 & 165 + 137 & 55167.0993 & 0.0035 & 0.0334 & 128 + 138 & 55167.1327 & 0.0019 & 0.0089 & 208 + 139 & 55167.1960 & 0.0006 & 0.0144 & 316 + 140 & 55167.2478 & 0.0005 & 0.0083 & 273 + 141 & 55167.3057 & 0.0013 & 0.0083 & 168 + 174 & 55169.1975 & 0.0022 & @xmath170.0087 & 84 + 175 & 55169.2371 & 0.0010 & @xmath170.0270 & 85 + 176 & 55169.2990 & 0.0007 & @xmath170.0229 & 82 + 192 & 55170.2172 & 0.0017 & @xmath170.0304 & 78 + + + + although the su uma - type nature of this object has been well established , http://www.kusastro.kyoto - u.ac.jp / vsnet / dne / aqcmi.html@xmath6 .
] there has unfortunately no solid publication on superhumps in this system .
the only known superoutburst since the 1997 observation was in 2008 september , which was too badly placed for time - series photometry .
the superoutburst in 2010 april brought the first chance to record superhumps since its recognition as an su uma - type dwarf nova .
the times of superhump maxima are listed in table [ tab : aqcmioc2010 ] .
although the course of the superhump evolution was not fully recorded , there appears to have been a discontinuous period change around @xmath24 .
we attributed this to stage a b transition because the observation recorded the early stage of the superoutburst .
the period derivative was not determined because of a gap in observation .
the mean @xmath3 during the entire observation was 0.06622(1 ) d ( figure [ fig : aqcmishpdm ] ) .
this period needs to be treated with caution because it was likely derived from different stages of superhump evolution .
( 88mm,110mm)fig6.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55295.9806 & 0.0006 & @xmath170.0022 & 116 + 1 & 55296.0489 & 0.0007 & @xmath170.0002 & 112 + 15 & 55296.9775 & 0.0005 & 0.0017 & 104 + 16 & 55297.0431 & 0.0005 & 0.0011 & 121 + 106 & 55302.9994 & 0.0009 & @xmath170.0004 & 68 + + + + gz cnc is a variable star discovered by takamizawa ( tmzv34 , vsnet - obs 10504 ) , which later turned out to be a dwarf nova @xcite .
@xcite observed an outburst in 2000 february .
based on the relatively slow rise to an outburst maximum and the lack of periodic modulations , they concluded that the object is likely an ss cyg - type dwarf nova with a long @xmath2 . in 2002 , @xcite noticed an unusually increase in the number of outburst detections , and suggested the similarity to a proposed intermediate polar v426 oph .
most surprisingly , radial - velocity studies by @xcite yielded an @xmath2 of 0.08825(28 ) d , placing the object at the lower edge of the period gap .
although this @xmath2 was seemingly incompatible with behavior of outbursts in this object , later observations have detected abundant short outbursts which are compatible with a short @xmath2-system .
the mystery , however , remained why the object did not develop superhumps during its long , 2000 february outburst .
later on , the object again exhibited a long outburst in 2007 december ( vsnet - alert 9783 ) , which was unfortunately not observed for searching superhumps . in 2009 january , this object underwent a long , bright outburst ( vsnet - alert 10984 ) .
follow - up observations , however , did not detect superhumps ( vsnet - alert 11002 ) .
gz cnc was then considered as a rare object below the period gap without a signature of an su uma - type dwarf nova . in 2010 march
, the object again underwent a long , bright outburst ( vsnet - alert 11855 , 11863 ) .
subsequent observations eventually detected superhumps ( vsnet - alert 11881 , 11888 , 11890 , 11894 ) .
the object is now confirmed to be a rare object exhibiting three classes of outbursts : normal ( narrow ) outbursts , long ss cyg - type ( wide ) outbursts and su uma - type superoutbursts ( figure [ fig : gzcnclc ] ) .
the mystery of the 2000 outburst could be understood if this outburst was a `` long '' outburst failing to trigger the tidal instability .
the only other known object having the same property is tu men ( cf .
@xcite ; @xcite ) , unusual su uma - type dwarf nova above the period gap .
if gz cnc is indeed a `` borderline '' su uma - type dwarf nova , the object may be analogous to bz uma , which also showed an usually slow rise to a full outburst ( cf .
@xcite ) .
the times of superhump maxima during the 2010 superoutburst are listed in table [ tab : gzcncoc2010 ] .
the @xmath0 diagram can be reasonably interpreted without a phase jump only if we assume a very large decrease in the period as in ax cap ( figure [ fig : gzcncoc ] ) .
the mean @xmath3 for @xmath25 was 0.09277(2 ) d ( pdm method ) , which is equivalent to @xmath13 = 5.1 % .
this fractional superhump excess is one of the largest among su uma - type dwarf novae below the period gap ( cf .
figure 15 in @xcite ) .
the mean period 0.08972(12 ) d after the period decrease corresponds to @xmath13 = 1.7 % .
since this period was fairly close to @xmath2 , this period decrease may have reflected the dominance of orbital humps during the late course of the superoutburst .
( 88mm,100mm)fig7.eps ( 88mm,100mm)fig8.eps ( 88mm,110mm)fig9.eps ( 88mm,110mm)fig10.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55269.0989 & 0.0007 & @xmath170.0405 & 64 + 1 & 55269.1935 & 0.0007 & @xmath170.0371 & 52 + 11 & 55270.1286 & 0.0008 & @xmath170.0146 & 175 + 26 & 55271.5221 & 0.0007 & 0.0099 & 86 + 27 & 55271.6139 & 0.0015 & 0.0104 & 55 + 32 & 55272.0723 & 0.0006 & 0.0124 & 226 + 33 & 55272.1535 & 0.0072 & 0.0024 & 48 + 36 & 55272.4347 & 0.0036 & 0.0099 & 57 + 37 & 55272.5312 & 0.0012 & 0.0150 & 93 + 42 & 55273.0023 & 0.0016 & 0.0298 & 53 + 43 & 55273.0952 & 0.0029 & 0.0315 & 74 + 44 & 55273.1874 & 0.0016 & 0.0325 & 88 + 64 & 55274.9832 & 0.0010 & 0.0029 & 97 + 65 & 55275.0761 & 0.0008 & 0.0046 & 276 + 66 & 55275.1721 & 0.0016 & 0.0093 & 160 + 87 & 55277.0421 & 0.0006 & @xmath170.0373 & 346 + 88 & 55277.1295 & 0.0007 & @xmath170.0412 & 173 + + + + the 2010 superoutburst was particularly well observed and stages b and c were very clearly defined ( table [ tab : gocomoc2010 ] ) .
these observations have confirmed the finding in @xcite and @xcite .
figure [ fig : gocomcomp2 ] illustrates a comparison of @xmath0 diagrams between different superoutbursts .
the 2010 observation better recorded the stage c superhumps with higher accuracy , while the 2003 observation well recorded the earlier part .
a combination of these superoutbursts very well demonstrates the `` canonical '' period variation in a short-@xmath3 system .
it is not clear whether the 2010 superoutburst was preceded by a precursor outburst as in the 2003 one .
( 88mm,70mm)fig11.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55286.1498 & 0.0006 & 0.0028 & 93 + 23 & 55287.5958 & 0.0004 & 0.0005 & 61 + 24 & 55287.6574 & 0.0004 & @xmath170.0009 & 37 + 38 & 55288.5362 & 0.0006 & @xmath170.0036 & 41 + 39 & 55288.5989 & 0.0003 & @xmath170.0038 & 61 + 40 & 55288.6630 & 0.0004 & @xmath170.0027 & 55 + 47 & 55289.1040 & 0.0002 & @xmath170.0026 & 127 + 48 & 55289.1663 & 0.0004 & @xmath170.0032 & 127 + 54 & 55289.5443 & 0.0004 & @xmath170.0030 & 50 + 55 & 55289.6075 & 0.0003 & @xmath170.0028 & 66 + 56 & 55289.6694 & 0.0005 & @xmath170.0039 & 66 + 63 & 55290.1132 & 0.0014 & @xmath170.0008 & 40 + 64 & 55290.1758 & 0.0004 & @xmath170.0012 & 40 + 65 & 55290.2384 & 0.0009 & @xmath170.0015 & 40 + 70 & 55290.5554 & 0.0005 & 0.0006 & 64 + 71 & 55290.6181 & 0.0004 & 0.0004 & 61 + 79 & 55291.1213 & 0.0019 & @xmath170.0002 & 36 + 94 & 55292.0701 & 0.0008 & 0.0041 & 135 + 95 & 55292.1369 & 0.0006 & 0.0079 & 159 + 96 & 55292.1997 & 0.0006 & 0.0077 & 96 + 101 & 55292.5144 & 0.0005 & 0.0076 & 60 + 102 & 55292.5763 & 0.0004 & 0.0065 & 64 + 103 & 55292.6419 & 0.0005 & 0.0091 & 66 + 117 & 55293.5184 & 0.0004 & 0.0041 & 60 + 118 & 55293.5817 & 0.0003 & 0.0045 & 39 + 130 & 55294.3331 & 0.0006 & 0.0002 & 66 + 131 & 55294.3949 & 0.0008 & @xmath170.0009 & 62 + 132 & 55294.4595 & 0.0004 & 0.0007 & 77 + 133 & 55294.5220 & 0.0005 & 0.0003 & 114 + 134 & 55294.5837 & 0.0004 & @xmath170.0011 & 132 + 143 & 55295.1460 & 0.0009 & @xmath170.0055 & 178 + 144 & 55295.2120 & 0.0014 & @xmath170.0024 & 38 + 150 & 55295.5881 & 0.0005 & @xmath170.0041 & 63 + 158 & 55296.0832 & 0.0028 & @xmath170.0128 & 28 + + + + we observed an superoutburst in 2009 december .
the times of superhump maxima are listed in table [ tab : tvcrvoc2009 ] . a combined @xmath0 diagram ( figure [ fig : tvcrvcomp2 ] )
strengthens the assertion in @xcite that the @xmath0 behavior is common between different superoutbursts .
( 88mm,70mm)fig12.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55183.2495 & 0.0004 & 0.0028 & 125 + 1 & 55183.3107 & 0.0003 & @xmath170.0010 & 144 + 16 & 55184.2870 & 0.0005 & @xmath170.0005 & 108 + 47 & 55186.3019 & 0.0006 & @xmath170.0024 & 143 + 62 & 55187.2796 & 0.0008 & @xmath170.0005 & 142 + 63 & 55187.3426 & 0.0007 & @xmath170.0025 & 121 + 77 & 55188.2587 & 0.0008 & 0.0028 & 143 + 78 & 55188.3227 & 0.0006 & 0.0018 & 144 + 92 & 55189.2347 & 0.0116 & 0.0030 & 114 + 93 & 55189.2941 & 0.0009 & @xmath170.0027 & 144 + 94 & 55189.3610 & 0.0015 & @xmath170.0008 & 82 + + + + we observed the 2010 superoutburst of this object .
the times of superhump maxima are listed in table [ tab : v337cygoc2010 ] .
although the period derivative was not well determined , we detected a stage b
c transition .
the obtained parameters are listed in table [ tab : perlist ] .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55421.4715 & 0.0012 & @xmath170.0046 & 118 + 1 & 55421.5390 & 0.0007 & @xmath170.0072 & 153 + 29 & 55423.5087 & 0.0007 & @xmath170.0009 & 75 + 42 & 55424.4236 & 0.0007 & 0.0024 & 73 + 43 & 55424.4944 & 0.0006 & 0.0032 & 72 + 44 & 55424.5626 & 0.0007 & 0.0012 & 77 + 45 & 55424.6340 & 0.0011 & 0.0025 & 54 + 51 & 55425.0473 & 0.0078 & @xmath170.0050 & 36 + 52 & 55425.1288 & 0.0019 & 0.0064 & 63 + 95 & 55428.1525 & 0.0013 & 0.0149 & 30 + 126 & 55430.3163 & 0.0065 & 0.0050 & 119 + 138 & 55431.1449 & 0.0010 & @xmath170.0079 & 128 + 139 & 55431.2128 & 0.0020 & @xmath170.0101 & 147 + + + + @xcite recently reported new observations of 2003 and 2005 superoutbursts , and claimed the presence of a large negative period derivative .
upon examination of their observations , it has become evident that they observed the apparently late stage of a superoutburst in 2003 .
using the typical duration ( @xmath26 d ) of superoutbursts in this system , their observation probably started @xmath9 94 cycles after the start of the superoutburst , and they most likely caught a stage b c transition .
a combined @xmath0 diagram based on this interpretation , supplemented by early observations reported in @xcite , is shown in figure [ fig : v1113cygcomp ] . as judged from this figure , the evolution of the superhump period is not particularly unusual in this system .
( 88mm,70mm)fig13.eps this object was only partly observed during the 2006 superoutburst @xcite .
the 2009 superoutburst was detected during its early stage ( i. miller , baavss - alert 2020 , vsnet - alert 11376 ) . a delay of @xmath276.5 d in the full growth of ordinary superhumps
was recorded ( vsnet - alert 11393 , 11395 ) .
the new observation now clarified the alias selection [ 0.05769(2 ) d with the pdm method ] and safely excluded the 0.0610 d earlier reported ( figure [ fig : v1454cygshpdm ] ) .
the @xmath7 for stage b was @xmath28 , fairly common for this short @xmath3 .
although there was likely a stage b
c transition after @xmath29 , we could not determine the period of stage c superhumps due to the lack of observations .
an analysis of the earlier observation ( bjd before 2455058 ) has yielded a weak signal of 0.05777(2 ) d. although the extraction of times of superhump maxima was difficult due to the low amplitudes , this period suggests that this interval involved stage a evolution rather than a manifestation of early superhumps .
( 88mm,110mm)fig14.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55058.1434 & 0.0004 & 0.0055 & 122 + 1 & 55058.2036 & 0.0013 & 0.0081 & 65 + 39 & 55060.3868 & 0.0010 & 0.0001 & 57 + 40 & 55060.4431 & 0.0024 & @xmath170.0013 & 54 + 41 & 55060.4996 & 0.0010 & @xmath170.0025 & 56 + 42 & 55060.5540 & 0.0035 & @xmath170.0057 & 54 + 50 & 55061.0219 & 0.0009 & 0.0009 & 109 + 51 & 55061.0756 & 0.0019 & @xmath170.0031 & 107 + 52 & 55061.1361 & 0.0024 & @xmath170.0003 & 121 + 53 & 55061.1919 & 0.0020 & @xmath170.0021 & 121 + 54 & 55061.2405 & 0.0033 & @xmath170.0112 & 115 + 56 & 55061.3703 & 0.0016 & 0.0033 & 54 + 57 & 55061.4258 & 0.0009 & 0.0012 & 52 + 58 & 55061.4806 & 0.0009 & @xmath170.0017 & 56 + 59 & 55061.5380 & 0.0011 & @xmath170.0020 & 57 + 60 & 55061.5956 & 0.0042 & @xmath170.0021 & 52 + 75 & 55062.4672 & 0.0019 & 0.0046 & 57 + 76 & 55062.5230 & 0.0027 & 0.0027 & 57 + 86 & 55063.0867 & 0.0032 & @xmath170.0101 & 114 + 87 & 55063.1513 & 0.0064 & @xmath170.0032 & 121 + 88 & 55063.2097 & 0.0155 & @xmath170.0024 & 86 + 92 & 55063.4463 & 0.0024 & 0.0035 & 56 + 93 & 55063.5010 & 0.0033 & 0.0005 & 57 + 94 & 55063.5635 & 0.0023 & 0.0054 & 57 + 95 & 55063.6254 & 0.0029 & 0.0095 & 30 + 160 & 55067.3705 & 0.0019 & 0.0066 & 57 + 161 & 55067.4330 & 0.0020 & 0.0115 & 54 + 163 & 55067.5267 & 0.0037 & @xmath170.0102 & 57 + 164 & 55067.5890 & 0.0019 & @xmath170.0055 & 54 + + + + we observed another superoutburst in 2010 ( table [ tab : aqerioc2010 ] ) .
there was a relatively large scatter in the @xmath0 diagram after the rapid fading due to its faintness .
we therefore did not attempt to determine the period during stage c. the mean @xmath3 appears to confirm the previous period determination ( cf .
figure [ fig : aqericomp2 ] ) .
( 88mm,70mm)fig15.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55201.0094 & 0.0004 & 0.0067 & 54 + 1 & 55201.0715 & 0.0011 & 0.0064 & 37 + 16 & 55202.0038 & 0.0003 & 0.0025 & 102 + 31 & 55202.9335 & 0.0013 & @xmath170.0039 & 65 + 32 & 55202.9972 & 0.0003 & @xmath170.0027 & 223 + 33 & 55203.0595 & 0.0004 & @xmath170.0027 & 73 + 48 & 55203.9967 & 0.0006 & @xmath170.0016 & 90 + 49 & 55204.0482 & 0.0048 & @xmath170.0126 & 51 + 64 & 55204.9952 & 0.0008 & @xmath170.0017 & 100 + 81 & 55206.0660 & 0.0022 & 0.0081 & 45 + 82 & 55206.1207 & 0.0028 & 0.0004 & 37 + 145 & 55210.0545 & 0.0018 & 0.0023 & 142 + 146 & 55210.1137 & 0.0021 & @xmath170.0008 & 146 + 161 & 55211.0574 & 0.0012 & 0.0067 & 68 + 162 & 55211.1045 & 0.0059 & @xmath170.0086 & 65 + 163 & 55211.1695 & 0.0035 & @xmath170.0060 & 101 + 193 & 55213.0543 & 0.0016 & 0.0065 & 61 + 194 & 55213.1111 & 0.0015 & 0.0009 & 91 + + + + vx for was discovered in 1990 as a probable dwarf nova showing balmer , hei and possibly heii emission lines @xcite . the large outburst amplitude and the outburst lasting for more than 10 d @xcite were already suggestive of an su uma - type superoutburst .
the object has been listed as a good candidate for a wz sge - type dwarf nova @xcite . the 2009 outburst , first - ever since the initial discovery , was detected by r. stubbings on 2009 sep .
14 at a visual magnitude of 13.0 ( vsnet - alert 11471 ) .
ordinary superhump were soon observed following this outburst detection ( vsnet - alert 11474 ) .
although the early appearance of ordinary superhumps was initially considered as unfavorable for the wz sge - type interpretation , a later retrospective detection of earlier positive observation in the asas-3 data ( @xmath30 = 12.61 on september 10 ) indicated that the early stage of the outburst was missed when early superhumps were expected ( vsnet - alert 11492 ) . using the empirical classification of wz sge - type dwarf novae introduced in @xcite , t. kato suggested that the object is expected to undergo multiple rebrightenings as in eg cnc based on relatively small @xmath7 obtained from early observations and a relatively long @xmath3 ( vsnet - alert 11492 ) .
the object indeed underwent five rebrightenings ( vsnet - alert 11521 , 11526 , 11536 , 11577 , 11602 ; figure [ fig : vxforlc ] ) , making it the first object predicted for its multiple rebrightening in real - time .
the overall behavior is very similar to that of asas [email protected] @xcite .
the mean @xmath3 during the plateau phase was 0.061355(7 ) d ( pdm method , figure [ fig : vxforshpdm ] ) .
the times of superhump maxima during the plateau phase and fading stage are listed in table [ tab : vxforoc2009 ] .
the superhumps persisted during the post - superoutburst and rebrightening phase .
we obtained @xmath31 d , slightly short than that of the superoutburst plateau phase ( figure [ fig : vxforlatepdm ] ) .
there was some indication of a signal around @xmath32 d ( small excess signal in figure [ fig : vxforlatepdm ] ) and transient appearance of this period ( cf .
vsnet - alert 11513 ) .
although this signal needs to be confirmed with better data , this might be analogous to long - period late - stage superhumps in wz sge - type dwarf novae ( @xcite ; @xcite ) .
( 88mm,70mm)fig16.eps ( 88mm,110mm)fig17.eps ( 88mm,110mm)fig18.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55090.0917 & 0.0002 & @xmath170.0100 & 991 + 6 & 55090.4665 & 0.0004 & @xmath170.0026 & 258 + 7 & 55090.5274 & 0.0004 & @xmath170.0029 & 258 + 8 & 55090.5873 & 0.0003 & @xmath170.0042 & 230 + 19 & 55091.2630 & 0.0007 & @xmath170.0019 & 104 + 20 & 55091.3216 & 0.0006 & @xmath170.0046 & 104 + 21 & 55091.3893 & 0.0004 & 0.0019 & 252 + 22 & 55091.4483 & 0.0003 & @xmath170.0003 & 229 + 23 & 55091.5073 & 0.0015 & @xmath170.0025 & 51 + 24 & 55091.5683 & 0.0015 & @xmath170.0028 & 116 + 32 & 55092.0586 & 0.0010 & @xmath170.0022 & 42 + 33 & 55092.1223 & 0.0005 & 0.0003 & 68 + 34 & 55092.1822 & 0.0009 & @xmath170.0011 & 66 + 35 & 55092.2450 & 0.0008 & 0.0005 & 84 + 36 & 55092.3078 & 0.0007 & 0.0021 & 30 + 48 & 55093.0420 & 0.0003 & 0.0016 & 349 + 49 & 55093.0999 & 0.0003 & @xmath170.0018 & 127 + 50 & 55093.1617 & 0.0003 & @xmath170.0012 & 159 + 51 & 55093.2244 & 0.0006 & 0.0003 & 95 + 55 & 55093.4687 & 0.0003 & @xmath170.0003 & 223 + 56 & 55093.5293 & 0.0003 & @xmath170.0009 & 261 + 57 & 55093.5909 & 0.0003 & @xmath170.0006 & 261 + 66 & 55094.1453 & 0.0015 & 0.0028 & 51 + 67 & 55094.2044 & 0.0011 & 0.0007 & 41 + 68 & 55094.2659 & 0.0008 & 0.0010 & 162 + 69 & 55094.3264 & 0.0018 & 0.0002 & 68 + 85 & 55095.3094 & 0.0012 & 0.0037 & 126 + 87 & 55095.4308 & 0.0008 & 0.0026 & 258 + 88 & 55095.4932 & 0.0006 & 0.0038 & 214 + 89 & 55095.5545 & 0.0005 & 0.0039 & 213 + 99 & 55096.1696 & 0.0013 & 0.0068 & 52 + 100 & 55096.2289 & 0.0011 & 0.0049 & 46 + 101 & 55096.2960 & 0.0020 & 0.0108 & 33 + 115 & 55097.1400 & 0.0049 & @xmath170.0024 & 24 + 132 & 55098.1979 & 0.0017 & 0.0147 & 22 + 133 & 55098.2561 & 0.0012 & 0.0117 & 32 + 137 & 55098.4994 & 0.0014 & 0.0100 & 258 + 149 & 55099.2215 & 0.0059 & @xmath170.0025 & 23 + 185 & 55101.4220 & 0.0094 & @xmath170.0061 & 134 + 186 & 55101.4843 & 0.0031 & @xmath170.0050 & 134 + 187 & 55101.5307 & 0.0027 & @xmath170.0198 & 134 + 201 & 55102.3908 & 0.0103 & @xmath170.0169 & 125 + 202 & 55102.4749 & 0.0167 & 0.0061 & 135 + 203 & 55102.5323 & 0.0018 & 0.0022 & 135 + 204 & 55102.5907 & 0.0018 & @xmath170.0006 & 133 + + + + the times of superhump maxima during the 2010 superoutburst is listed in table [ tab : awgemoc2010 ] .
a stage b
c transition was well recorded .
although only parts of superoutbursts were observed in individual years , the combined @xmath0 diagram ( figure [ fig : awgemcomp ] ) clearly demonstrates the universal pattern of period evolution .
( 88mm,70mm)fig19.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55257.3860 & 0.0002 & @xmath170.0033 & 72 + 1 & 55257.4640 & 0.0003 & @xmath170.0040 & 40 + 2 & 55257.5425 & 0.0004 & @xmath170.0042 & 69 + 24 & 55259.2830 & 0.0004 & 0.0040 & 81 + 25 & 55259.3619 & 0.0004 & 0.0041 & 80 + 26 & 55259.4398 & 0.0005 & 0.0033 & 80 + 51 & 55261.4091 & 0.0004 & 0.0041 & 69 + 52 & 55261.4873 & 0.0005 & 0.0035 & 61 + 53 & 55261.5602 & 0.0012 & @xmath170.0024 & 37 + 101 & 55265.3439 & 0.0029 & 0.0018 & 43 + 102 & 55265.4189 & 0.0008 & @xmath170.0020 & 78 + 103 & 55265.4946 & 0.0006 & @xmath170.0050 & 79 + + + + the times of superhump maxima during the 2010 superoutburst is listed in table [ tab : irgemoc2010 ] . although the object was observed on only two nights , the period is in good agreement with earlier observations .
the observations was likely performed during stage b. ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55265.4201 & 0.0004 & 0.0003 & 65 + 1 & 55265.4903 & 0.0004 & @xmath170.0003 & 67 + 56 & 55269.3867 & 0.0003 & 0.0002 & 71 + 57 & 55269.4573 & 0.0004 & @xmath170.0000 & 73 + 58 & 55269.5280 & 0.0007 & @xmath170.0002 & 50 + + + + v592 her was discovered as a possible fast nova in 1968 on sonneberg plates @xcite . @xcite further discovered a second outburst in 1986 on historical plates . in 1998 ,
another outburst was recorded which led to the identification of this object as being a wz sge - type dwarf nova ( @xcite ; @xcite ) .
@xcite first identified the true superhump period of 0.05648(2 ) d , although their observation was not very sufficient to determine its period variation .
the 2010 outburst of this object was detected by m. reszelski at a relatively faint magnitude of 14.16 ( unfiltered ccd magnitude , vsnet - obs 67929 ) .
subsequent observations detected growing superhumps ( vsnet - alert 12092 , 12094 , 12095 ) .
this new outburst has confirmed the selection of the superhump period in ( @xcite ; figure [ fig : v592hershpdm ] ) .
the times of superhump maxima are listed in table [ tab : v592heroc2010 ] .
all a c stages for superhump evolution were clearly detected ( figure [ fig : v592heroc2010 ] ) .
we obtained a clearly positive @xmath7 of @xmath33 during stage b , which is likely a more reliable value than the 1998 estimate thanks to the greatly improved statistics .
as judged from this relatively large period derivative and the presence of stage c evolution , this object is less likely an extreme wz sge - type dwarf nova with a very small period variation what was supposed from the 1998 data @xcite .
the relatively short ( @xmath9 12 yr ) recurrence time would qualify the system as a wz sge - type dwarf nova similar to hv vir @xcite .
the lack of repetitive rebrightenings would also support this classification .
there was a weaker signal slightly shorter than the superhump period ( figure [ fig : v592hershpdm ] ) , which might be attributed to the orbital period .
although the detection was not statistically significant , a similar periodicity may have been present before the appearance of ordinary superhumps ( figure [ fig : v592hereshpdm ] ) . since this period agrees with the candidate period from radial - velocity study ( @xcite
; see @xcite for the alias selection ) , we adopted it as a candidate for the orbital period , which corresponds to a fractional superhump excess of 0.9 % .
( 88mm,110mm)fig20.eps ( 88mm,90mm)fig21.eps ( 88mm,110mm)fig22.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55413.7374 & 0.0014 & @xmath170.0009 & 53 + 1 & 55413.7957 & 0.0016 & 0.0008 & 53 + 7 & 55414.1260 & 0.0019 & @xmath170.0086 & 82 + 12 & 55414.4196 & 0.0008 & 0.0018 & 53 + 13 & 55414.4805 & 0.0012 & 0.0061 & 54 + 17 & 55414.7105 & 0.0005 & 0.0096 & 48 + 18 & 55414.7679 & 0.0003 & 0.0105 & 53 + 35 & 55415.7294 & 0.0004 & 0.0094 & 52 + 36 & 55415.7856 & 0.0006 & 0.0089 & 38 + 40 & 55416.0066 & 0.0015 & 0.0035 & 58 + 41 & 55416.0661 & 0.0023 & 0.0063 & 79 + 42 & 55416.1268 & 0.0061 & 0.0104 & 127 + 46 & 55416.3490 & 0.0003 & 0.0062 & 33 + 47 & 55416.4049 & 0.0002 & 0.0054 & 104 + 48 & 55416.4612 & 0.0003 & 0.0051 & 71 + 49 & 55416.5176 & 0.0006 & 0.0049 & 42 + 51 & 55416.6290 & 0.0003 & 0.0031 & 62 + 52 & 55416.6836 & 0.0005 & 0.0010 & 69 + 53 & 55416.7427 & 0.0004 & 0.0035 & 52 + 57 & 55416.9637 & 0.0049 & @xmath170.0020 & 71 + 58 & 55417.0236 & 0.0006 & 0.0013 & 175 + 59 & 55417.0835 & 0.0017 & 0.0046 & 89 + 60 & 55417.1378 & 0.0013 & 0.0022 & 82 + 61 & 55417.1935 & 0.0064 & 0.0013 & 50 + 64 & 55417.3606 & 0.0005 & @xmath170.0015 & 89 + 65 & 55417.4180 & 0.0005 & @xmath170.0007 & 143 + 66 & 55417.4725 & 0.0009 & @xmath170.0027 & 55 + 70 & 55417.7001 & 0.0008 & @xmath170.0017 & 35 + 71 & 55417.7584 & 0.0005 & 0.0000 & 51 + 88 & 55418.7157 & 0.0010 & @xmath170.0053 & 52 + 89 & 55418.7719 & 0.0005 & @xmath170.0057 & 51 + 94 & 55419.0581 & 0.0008 & @xmath170.0026 & 27 + 95 & 55419.1071 & 0.0019 & @xmath170.0102 & 68 + 96 & 55419.1677 & 0.0021 & @xmath170.0062 & 66 + 99 & 55419.3370 & 0.0008 & @xmath170.0068 & 61 + 100 & 55419.3968 & 0.0009 & @xmath170.0036 & 71 + 101 & 55419.4498 & 0.0009 & @xmath170.0072 & 75 + 102 & 55419.5057 & 0.0004 & @xmath170.0080 & 62 + 103 & 55419.5641 & 0.0006 & @xmath170.0061 & 38 + 104 & 55419.6192 & 0.0007 & @xmath170.0077 & 62 + 105 & 55419.6739 & 0.0006 & @xmath170.0096 & 72 + 106 & 55419.7355 & 0.0009 & @xmath170.0046 & 53 + 117 & 55420.3571 & 0.0011 & @xmath170.0058 & 62 + 118 & 55420.4101 & 0.0011 & @xmath170.0095 & 69 + 119 & 55420.4693 & 0.0011 & @xmath170.0069 & 20 + 120 & 55420.5254 & 0.0006 & @xmath170.0074 & 52 + 130 & 55421.0909 & 0.0033 & @xmath170.0081 & 62 + 135 & 55421.3703 & 0.0018 & @xmath170.0119 & 60 + 136 & 55421.4374 & 0.0021 & @xmath170.0013 & 68 + 141 & 55421.7181 & 0.0012 & @xmath170.0037 & 52 + 142 & 55421.7706 & 0.0016 & @xmath170.0078 & 41 + 159 & 55422.7367 & 0.0016 & @xmath170.0043 & 52 + 176 & 55423.6891 & 0.0089 & @xmath170.0144 & 42 + 181 & 55423.9909 & 0.0046 & 0.0042 & 149 + 182 & 55424.0429 & 0.0032 & @xmath170.0004 & 113 + 183 & 55424.1055 & 0.0044 & 0.0056 & 60 + 194 & 55424.7301 & 0.0029 & 0.0073 & 52 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 195 & 55424.7736 & 0.0030 & @xmath170.0058 & 42 + 209 & 55425.5776 & 0.0017 & 0.0056 & 47 + 210 & 55425.6348 & 0.0026 & 0.0061 & 59 + 212 & 55425.7491 & 0.0019 & 0.0072 & 53 + 216 & 55425.9826 & 0.0042 & 0.0142 & 162 + 229 & 55426.7172 & 0.0027 & 0.0127 & 53 + 230 & 55426.7657 & 0.0012 & 0.0046 & 45 + 234 & 55426.9957 & 0.0026 & 0.0081 & 82 + 235 & 55427.0454 & 0.0012 & 0.0012 & 113 + 247 & 55427.7308 & 0.0014 & 0.0071 & 53 + 252 & 55428.0234 & 0.0046 & 0.0166 & 62 + 264 & 55428.6724 & 0.0063 & @xmath170.0138 & 36 + 265 & 55428.7468 & 0.0027 & 0.0039 & 52 + 282 & 55429.7099 & 0.0048 & 0.0045 & 52 + 283 & 55429.7621 & 0.0097 & 0.0000 & 52 + 300 & 55430.7126 & 0.0048 & @xmath170.0120 & 50 + + + + the times of superhump maxima during the 2009 superoutburst are listed in table [ tab : v660heroc2009 ] . since the late stage of the superoutburst
was only observed , the recorded superhumps are likely stage c superhumps .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55056.4789 & 0.0008 & @xmath170.0032 & 30 + 12 & 55057.4501 & 0.0007 & @xmath170.0004 & 36 + 25 & 55058.5061 & 0.0019 & 0.0067 & 9 + 49 & 55060.4337 & 0.0049 & @xmath170.0023 & 17 + 61 & 55061.4035 & 0.0098 & @xmath170.0008 & 23 + + + + two further superoutbursts in 2009 february march ( table [ tab : v844heroc2009 ] ) and in 2010 april may ( table [ tab : v844heroc2010 ] ; see also vsnet - alert 11959 for the outburst detection ) were observed .
both superoutbursts were relatively faint and short ones . during the former superoutburst we obtained a clearly positive @xmath7 of @xmath34 ( likely for stage b ) .
the object underwent yet another superoutburst between them : 2009 october
november ( vsnet - alert 11622 ) .
there was also a normal outburst in 2009 june ( vsnet - alert 11289 , 11290 ) .
the object was thus unusually active for this star ( cf .
@xcite ; @xcite ) .
as suggested in @xcite and @xcite , these superoutburst would provide an excellent opportunity to study the dependence of period derivatives on the extent of the superoutburst .
as already seen , the @xmath7 of the 2009 superoutburst is not significantly different from those of longer superoutbursts .
although the superhumps had already developed at the epoch of initial observation ( @xmath9 1.5 d after the initial outburst detection ) , we could nt severely constrain the delay time in evolution of superhumps since there was a three - day gap of observation before this detection . the 2010 observation
started @xmath9 1 d after the rising phase of the outburst , and superhumps were already present .
although these superhumps were possibly stage a superhumps ( see figure [ fig : v844hercomp2 ] ) , this observations seems to support the hypothesis in @xcite that the delay time in development of superhumps is shorter in smaller superoutbursts .
( 88mm,70mm)fig23.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 54887.4721 & 0.0003 & 0.0024 & 87 + 1 & 54887.5268 & 0.0002 & 0.0012 & 104 + 2 & 54887.5827 & 0.0002 & 0.0012 & 109 + 3 & 54887.6379 & 0.0002 & 0.0005 & 125 + 72 & 54891.4923 & 0.0003 & @xmath170.0038 & 101 + 73 & 54891.5472 & 0.0003 & @xmath170.0049 & 108 + 74 & 54891.6031 & 0.0004 & @xmath170.0049 & 108 + 90 & 54892.4986 & 0.0008 & @xmath170.0042 & 105 + 91 & 54892.5626 & 0.0008 & 0.0039 & 108 + 92 & 54892.6132 & 0.0006 & @xmath170.0014 & 107 + 93 & 54892.6681 & 0.0009 & @xmath170.0024 & 93 + 108 & 54893.5116 & 0.0006 & 0.0023 & 106 + 109 & 54893.5667 & 0.0005 & 0.0014 & 103 + 110 & 54893.6264 & 0.0023 & 0.0052 & 60 + 111 & 54893.6807 & 0.0015 & 0.0035 & 51 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55316.1758 & 0.0002 & @xmath170.0000 & 101 + 53 & 55319.1315 & 0.0002 & 0.0001 & 134 + 54 & 55319.1870 & 0.0002 & @xmath170.0001 & 177 + + + + the times of superhump maxima during the 2010 superoutburst is listed in table [ tab : cthyaoc2010 ] . this observation first time
recorded a stage a b transition and growth of superhumps .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55269.0068 & 0.0049 & @xmath170.0021 & 91 + 1 & 55269.0698 & 0.0024 & @xmath170.0057 & 143 + 14 & 55269.9458 & 0.0013 & 0.0047 & 86 + 15 & 55270.0122 & 0.0005 & 0.0045 & 129 + 90 & 55275.0002 & 0.0009 & @xmath170.0014 & 142 + + + + our new observation of the 2010 superoutburst ( table [ tab : v699ophoc2010 ] ) clearly caught a stage b
c transition .
a comparison of @xmath0 diagrams is shown in figure [ fig : v699ophcomp ] .
( 88mm,70mm)fig24.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55362.5280 & 0.0023 & @xmath170.0044 & 12 + 14 & 55363.5125 & 0.0007 & @xmath170.0008 & 77 + 22 & 55364.0674 & 0.0041 & @xmath170.0064 & 47 + 27 & 55364.4254 & 0.0016 & 0.0014 & 41 + 28 & 55364.4981 & 0.0007 & 0.0039 & 68 + 42 & 55365.4813 & 0.0007 & 0.0063 & 65 + 56 & 55366.4592 & 0.0008 & 0.0033 & 69 + 57 & 55366.5288 & 0.0008 & 0.0029 & 60 + 70 & 55367.4392 & 0.0013 & 0.0025 & 62 + 71 & 55367.5086 & 0.0008 & 0.0018 & 62 + 84 & 55368.4083 & 0.0132 & @xmath170.0092 & 39 + 85 & 55368.4882 & 0.0020 & 0.0005 & 67 + 99 & 55369.4666 & 0.0047 & @xmath170.0019 & 65 + + + + v1032 oph was originally discovered as a candidate rr lyr - type variable star @xcite . based on its bright uv emission
, @xcite conducted a systematic observation which led to a conclusion that the object is a likely su uma - type dwarf nova .
the 2010 outburst was detected by e. muyllaert ( vsnet - alert 11898 ) .
e. de miguel and h. maehara confirmed that this object is an eclipsing su uma - type dwarf nova ( vsnet - alert 11904 , 11905 ) .
the times of recorded eclipses , determined with the kwee and van woerden ( kw ) method @xcite , are summarized in table [ tab : v1032ophecl ] .
we obtained an ephemeris of @xmath35 in the following analysis , we removed observations within 0.07 @xmath2 of eclipses .
the times of superhump maxima are listed in table [ tab : v1032ophoc2010 ] . due to the faintness of the object and overlapping eclipsing feature ,
the scatter is relatively large .
the last maximum ( @xmath36 ) may have been a traditional late superhump with an @xmath90.5 phase offset , when the observation was performed after a brightness drop from the superoutburst .
a linear fit to the observed epochs for @xmath37 yielded a mean period of 0.08529(12 ) d. this period is in agreement with 0.08534(5 ) d determined with the pdm method ( figure [ fig : v1032ophshpdm ] ) .
we adopted the latter because of a smaller error . although the period derivative could not be unambiguously determined , restricting to well - defined maxima ( @xmath38 ) and allowing a large period variation ( as in mn dra ) , we can derive a global @xmath7 of @xmath39 .
this value needs to be confirmed by future observations .
the fractional superhump excess amounts to 5.3 % , which is one of the largest among su uma - type dwarf novae below the period gap ( cf .
subsection [ obj : gzcnc ] ) .
( 88mm,110mm)fig25.eps cccc @xmath15 & minimum & error & @xmath0 + 0 & 55286.6829 & 0.0008 & 0.0002 + 32 & 55289.2765 & 0.0006 & 0.0001 + 49 & 55290.6544 & 0.0007 & -0.0000 + 69 & 55292.2763 & 0.0010 & 0.0008 + 74 & 55292.6809 & 0.0006 & 0.0001 + 89 & 55293.8967 & 0.0005 & 0.0000 + 90 & 55293.9775 & 0.0004 & -0.0002 + 99 & 55294.7075 & 0.0006 & 0.0002 + 102 & 55294.9505 & 0.0004 & 0.0001 + 103 & 55295.0310 & 0.0006 & -0.0004 + 111 & 55295.6801 & 0.0006 & 0.0001 + 114 & 55295.9231 & 0.0004 & 0.0000 + 115 & 55296.0041 & 0.0005 & -0.0001 + 122 & 55296.5710 & 0.0005 & -0.0005 + 123 & 55296.6517 & 0.0004 & -0.0009 + 212 & 55303.8664 & 0.0011 & -0.0002 + 213 & 55303.9478 & 0.0009 & 0.0001 + 238 & 55305.9747 & 0.0006 & 0.0006 + 296 & 55310.6754 & 0.0005 & 0.0001 + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55286.6842 & 0.0023 & 0.0062 & 74 + 24 & 55288.7046 & 0.0018 & @xmath170.0174 & 24 + 30 & 55289.2294 & 0.0011 & @xmath170.0036 & 167 + 41 & 55290.1909 & 0.0040 & 0.0210 & 126 + 43 & 55290.3086 & 0.0017 & @xmath170.0316 & 73 + 47 & 55290.6862 & 0.0006 & 0.0053 & 73 + 65 & 55292.2223 & 0.0023 & 0.0083 & 147 + 66 & 55292.3013 & 0.0072 & 0.0022 & 141 + 85 & 55293.9201 & 0.0012 & 0.0028 & 34 + 86 & 55294.0063 & 0.0015 & 0.0039 & 32 + 94 & 55294.6949 & 0.0015 & 0.0110 & 65 + 97 & 55294.9549 & 0.0018 & 0.0156 & 37 + 98 & 55295.0222 & 0.0083 & @xmath170.0023 & 26 + 100 & 55295.2186 & 0.0035 & 0.0238 & 37 + 106 & 55295.6961 & 0.0060 & @xmath170.0097 & 60 + 117 & 55296.6072 & 0.0006 & @xmath170.0355 & 69 + + + + the times of superhump maxima during the 2010 superoutburst are listed in table [ tab : v2051ophoc2010 ] .
although the outburst was detected during its relatively early stage ( vsnet - alert 12049 ) , the subsequent observational coverage was rather insufficient .
the observed superhumps were likely stage b superhumps
. we could not meaningfully determine @xcite .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55382.9240 & 0.0003 & 0.0014 & 107 + 1 & 55382.9876 & 0.0003 & 0.0008 & 104 + 2 & 55383.0492 & 0.0002 & @xmath170.0019 & 78 + 3 & 55383.1165 & 0.0003 & 0.0012 & 94 + 14 & 55383.8220 & 0.0014 & 0.0000 & 34 + 17 & 55384.0142 & 0.0006 & @xmath170.0004 & 92 + 18 & 55384.0865 & 0.0005 & 0.0076 & 89 + 34 & 55385.1019 & 0.0009 & @xmath170.0048 & 71 + 35 & 55385.1613 & 0.0027 & @xmath170.0096 & 40 + 65 &
55387.1037 & 0.0013 & 0.0057 & 37 + + + + ef peg is one of representatives of long-@xmath3 su uma - type dwarf novae with infrequent outbursts .
@xcite reported mildly negative @xmath7 for the 1991 and 1997 superoutbursts .
the object underwent an outburst in 2009 december ( vsnet - alert 11738 , baavss - alert 2177 , vsnet - alert 11740 ) , first time since its 2001 superoutburst .
the outburst was caught during its early stage and the development of superhumps was recorded .
the times of superhump maxima are listed in table [ tab : efpegoc2009 ] . due to the short visibility in the evening sky ,
the number of maxima was relatively small .
the @xmath0 diagram , however , clearly shows the presence of stage a during the evolutionary stage of superhumps ( @xmath40 ) , and subsequent phase of a slow period decrease .
we attributed the latter phase to a transition from stage b to c and identified the periods in table [ tab : perlist ] .
a comparison of @xmath0 variations between different superoutbursts is presented in figure [ fig : efpegcomp2 ] . among long-@xmath3 systems ,
the @xmath0 variation looks similar to those of ax cap and sdss j1627 ( cf .
@xcite ) with a discontinuous period variation between stages b and c. it would be noteworthy that all systems are known to show only rare superoutbursts .
( 88mm,70mm)fig26.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55187.3273 & 0.0007 & @xmath170.0296 & 229 + 6 & 55187.8804 & 0.0008 & 0.0006 & 353 + 7 & 55187.9609 & 0.0007 & @xmath170.0061 & 521 + 10 & 55188.2402 & 0.0005 & 0.0117 & 287 + 11 & 55188.3234 & 0.0004 & 0.0078 & 277 + 30 & 55189.9852 & 0.0014 & 0.0135 & 296 + 68 & 55193.2913 & 0.0005 & 0.0073 & 268 + 75 & 55193.8931 & 0.0008 & @xmath170.0010 & 464 + 79 & 55194.2507 & 0.0005 & 0.0080 & 80 + 87 & 55194.9431 & 0.0008 & 0.0031 & 440 + 125 & 55198.2372 & 0.0006 & @xmath170.0151 & 77 + + + + we also observed the 2009 superoutburst .
this outburst was one of the brightest in recent years ( cf .
vsnet - alert 11507 ) .
the times of superhump maxima are listed in table [ tab : v368pegoc2009 ] , which clearly shows a stage b
c transition .
the superhumps were not very apparent on the first night ( bjd 2455102 ) , and it was likely that the development of superhumps took more than 1 d. the relatively large @xmath7 for stage b strongly depends on @xmath41 , and may not be real .
a comparison of @xmath0 diagrams between different superoutbursts is shown in figure [ fig : v368pegcomp2 ] . despite its brightness , the behavior of the 2009 superoutburst was not strikingly different from that of other superoutbursts .
( 88mm,70mm)fig27.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55102.8556 & 0.0009 & @xmath170.0046 & 109 + 35 & 55105.3047 & 0.0030 & @xmath170.0110 & 31 + 36 & 55105.3797 & 0.0003 & @xmath170.0062 & 73 + 37 & 55105.4512 & 0.0004 & @xmath170.0048 & 69 + 38 & 55105.5209 & 0.0005 & @xmath170.0053 & 43 + 40 & 55105.6602 & 0.0008 & @xmath170.0063 & 24 + 55 & 55106.7133 & 0.0003 & @xmath170.0056 & 110 + 97 & 55109.6732 & 0.0006 & 0.0078 & 153 + 98 & 55109.7456 & 0.0005 & 0.0100 & 156 + 99 & 55109.8159 & 0.0004 & 0.0102 & 148 + 113 & 55110.7965 & 0.0006 & 0.0087 & 36 + 114 & 55110.8657 & 0.0006 & 0.0076 & 33 + 115 & 55110.9358 & 0.0008 & 0.0076 & 26 + 125 & 55111.6361 & 0.0004 & 0.0064 & 74 + 126 & 55111.7060 & 0.0004 & 0.0061 & 102 + 127 & 55111.7756 & 0.0006 & 0.0056 & 56 + 128 & 55111.8544 & 0.0065 & 0.0142 & 13 + 137 & 55112.4750 & 0.0006 & 0.0033 & 57 + 138 & 55112.5466 & 0.0007 & 0.0049 & 73 + 177 & 55115.2731 & 0.0007 & @xmath170.0048 & 49 + 178 & 55115.3412 & 0.0009 & @xmath170.0068 & 46 + 179 & 55115.4090 & 0.0016 & @xmath170.0092 & 50 + 180 & 55115.4822 & 0.0018 & @xmath170.0061 & 50 + 181 & 55115.5469 & 0.0018 & @xmath170.0115 & 49 + 220 & 55118.2843 & 0.0009 & @xmath170.0103 & 30 + + + + we observed the 2010 superoutburst during this middle and final stages ( table [ tab : uvperoc2010 ] ) . although a stage b c transition was recorded , the long gap in observation hindered precise determination of periods .
we only list representative values for stage b. late - stage superhumps superimposed on the rapid fading from the superoutburst plateau were clearly recorded as in the 1992 superoutburst .
a comparison of @xmath0 diagrams between different superoutbursts is shown in figure [ fig : uvpercomp2 ] .
this figure is an improvement of the corresponding one presented in @xcite .
( 88mm,70mm)fig28.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55203.4188 & 0.0003 & @xmath170.0041 & 74 + 1 & 55203.4864 & 0.0004 & @xmath170.0029 & 74 + 2 & 55203.5521 & 0.0003 & @xmath170.0037 & 73 + 3 & 55203.6189 & 0.0004 & @xmath170.0033 & 62 + 13 & 55204.2845 & 0.0004 & @xmath170.0016 & 71 + 15 & 55204.4177 & 0.0005 & @xmath170.0012 & 40 + 16 & 55204.4847 & 0.0004 & @xmath170.0006 & 46 + 17 & 55204.5518 & 0.0006 & 0.0001 & 34 + 29 & 55205.3527 & 0.0005 & 0.0042 & 41 + 30 & 55205.4212 & 0.0005 & 0.0063 & 35 + 31 & 55205.4878 & 0.0006 & 0.0065 & 37 + 32 & 55205.5524 & 0.0009 & 0.0047 & 28 + 133 & 55212.2538 & 0.0004 & @xmath170.0001 & 63 + 134 & 55212.3179 & 0.0008 & @xmath170.0024 & 67 + 135 & 55212.3847 & 0.0009 & @xmath170.0019 & 62 + 136 & 55212.4535 & 0.0018 & 0.0004 & 48 + 137 & 55212.5190 & 0.0011 & @xmath170.0004 & 62 + + + + we observed the 2009 superoutburst first detected by asas-3 on june 8 at @xmath42 . due to the poor seasonal condition , we could only determine the mean superhump period of 0.04635(5 ) d ( with the pdm method ) , in agreement with previous measurements .
the times of superhump maxima are listed in table [ tab : eipscoc2009 ] .
the overall light curve of the outburst suggests that the main superoutburst plateau lasted less than 10 d and experienced a rebrightening on june 18 .
the course of the outburst could have been similar to the 2005 one ( @xcite ; @xcite ) .
the shortness of the superoutburst plateau in such short-@xmath2 systems with evolved secondaries would require a special explanation .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 54994.1841 & 0.0011 & 0.0006 & 44 + 1 & 54994.2292 & 0.0005 & @xmath170.0006 & 67 + 10 & 54994.6467 & 0.0004 & 0.0001 & 29 + + + + we observed the 2009 superoutburst .
a clear stae a b transition was recorded ( table [ tab : ektraoc2009 ] ) .
although the period variation was alost absent during the supposed stage b ( @xmath43 ) , this may have been a result of fragmentary observations of stages b and c. ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55027.0310 & 0.0029 & @xmath170.0009 & 74 + 1 & 55027.0787 & 0.0026 & @xmath170.0181 & 118 + 29 & 55028.9255 & 0.0002 & 0.0101 & 159 + 30 & 55028.9901 & 0.0003 & 0.0098 & 100 + 76 & 55031.9722 & 0.0003 & 0.0043 & 180 + 77 & 55032.0358 & 0.0003 & 0.0030 & 184 + 137 & 55035.9273 & 0.0011 & @xmath170.0023 & 108 + 138 & 55035.9924 & 0.0006 & @xmath170.0022 & 137 + 139 & 55036.0559 & 0.0005 & @xmath170.0036 & 139 + + + + the 2010 january superoutburst was observed for its early and late stages ( table [ tab : suumaoc2010 ] ) . although we could only measure the mean period of stage b , the value is in agreement with those obtained during previous superoutbursts .
the humps observed for @xmath44 may not be genuine superhumps .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55219.5098 & 0.0002 & @xmath170.0003 & 132 + 1 & 55219.5883 & 0.0002 & @xmath170.0009 & 149 + 2 & 55219.6689 & 0.0002 & 0.0005 & 150 + 7 & 55220.0702 & 0.0014 & 0.0061 & 89 + 11 & 55220.3809 & 0.0002 & 0.0002 & 145 + 12 & 55220.4613 & 0.0002 & 0.0014 & 156 + 13 & 55220.5392 & 0.0002 & 0.0002 & 153 + 14 & 55220.6182 & 0.0004 & @xmath170.0001 & 157 + 15 & 55220.6987 & 0.0003 & 0.0013 & 122 + 25 & 55221.4871 & 0.0004 & @xmath170.0018 & 155 + 26 & 55221.5668 & 0.0004 & @xmath170.0012 & 151 + 27 & 55221.6454 & 0.0004 & @xmath170.0018 & 154 + 50 & 55223.4634 & 0.0004 & @xmath170.0044 & 96 + 51 & 55223.5457 & 0.0004 & @xmath170.0012 & 118 + 163 & 55232.4126 & 0.0012 & 0.0006 & 75 + 164 & 55232.4926 & 0.0012 & 0.0014 & 83 + + + + bc uma underwent a superoutburst in 2009 september october after a period of 6.7 yr ( cf .
vsnet - alert 11514 ) .
the start of the outburst was not well constrained dur to the poor visibility in the morning sky .
superhumps were observed despite unfavorable seasonal conditions ( cf .
vsnet - alert 11540 , 11550 , 11559 ) .
the times of superhump maxima are listed in table [ tab : bcumaoc2009 ] .
there was a clear stage b
c transition and the @xmath7 for stage b ( @xmath45 ) was @xmath46 .
we attributed the interval @xmath47 to likely stage a superhumps based on comparison with other superoutbursts ( figure [ fig : bcumacomp2 ] ) .
if this identification is correct , the superhumps during this superoutburst appears to have taken a longer time to fully develop .
( 88mm,70mm)fig29.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55105.2681 & 0.0022 & @xmath170.0189 & 70 + 2 & 55105.3853 & 0.0013 & @xmath170.0310 & 43 + 56 & 55108.9240 & 0.0016 & 0.0181 & 182 + 92 & 55111.2424 & 0.0004 & 0.0101 & 49 + 125 & 55113.3725 & 0.0007 & 0.0076 & 45 + 126 & 55113.4388 & 0.0004 & 0.0093 & 84 + 138 & 55114.2105 & 0.0021 & 0.0056 & 40 + 139 & 55114.2800 & 0.0004 & 0.0104 & 86 + 140 & 55114.3451 & 0.0005 & 0.0109 & 72 + 141 & 55114.4106 & 0.0005 & 0.0117 & 85 + 142 & 55114.4740 & 0.0005 & 0.0105 & 86 + 143 & 55114.5395 & 0.0005 & 0.0114 & 87 + 144 & 55114.6037 & 0.0004 & 0.0109 & 87 + 170 & 55116.2768 & 0.0005 & 0.0039 & 113 + 203 & 55118.3985 & 0.0013 & @xmath170.0070 & 67 + 204 & 55118.4593 & 0.0007 & @xmath170.0108 & 72 + 205 & 55118.5248 & 0.0006 & @xmath170.0099 & 73 + 206 & 55118.5888 & 0.0005 & @xmath170.0105 & 73 + 207 & 55118.6512 & 0.0014 & @xmath170.0128 & 45 + 219 & 55119.4198 & 0.0007 & @xmath170.0196 & 46 + + + + although el uma was discovered as an eruptive object relatively early in the history @xcite , only little had been known until recent years . @xcite
listed this object among candidate wz sge - type dwarf novae .
based on the similarity of its quiescent sdss color to those of known wz sge - type dwarf novae , we started monitoring since 2008 .
another outburst at @xmath48 = 13.7 in 2003 april was found on an archival image @xcite . on 2009 january 13 , h. maehara finally detected this object in outburst at an unfiltered ccd magnitude of 17.4 ( vsnet - alert 11771 ) the object further brightened to a magnitude of 14.9 on january 16 ( vsnet - alert 11772 ) .
the object has been confirmed to be a hydrogen - rich dwarf niva in outburst by spectroscopy ( takahashi and kinugasa , private communication ) . although the nature of this brightening was unclear at the time , the detection of additional sequence of outbursts ( vsnet - alert 11789 , 11808 ) , led to an interpretation that they are post - superoutburst rebrightenings of a wz sge - type superoutburst , whose main superoutburst was missed ( vsnet - alert 11795 ) .
the detection of modulations attributable to superhumps seems to strengthen this interpretation ( vsnet - alert 11799 ) .
we include this object based on this interpretation . a period analysis of the rebrightening phase , after subtracting the trends of outbursts ( cf .
@xcite ) , has yielded a period of 0.06045(6 ) d ( figure [ fig : elumashpdm ] ) .
although this period needs to be confirmed by future observations , its relatively long @xmath3 appears to be similar to that of vx for ( subsection [ obj : vxfor ] ) that underwent multiple rebrightenings ( figure [ fig : elumalc ] ) .
there is a common tendency that the quiescent interval preceding the last rebrightening is longer than the other intervals between rebrightenings .
this object is a good candidate for a cv passed the period minimum in evolution ( see a discussion in @xcite ) .
further radial - velocity study in quiescence in encouraged in order to determine the orbital period and the nature of the period during the rebrightening phase .
( 88mm,110mm)fig30.eps ( 88mm,110mm)fig31.eps although the 2009 superoutburst of this object was reported in @xcite , we provide greatly improved results by combining newly available observations .
the times of superhump maxima ( table [ tab : iyumaoc2009 ] ) now clearly illustrate the presence of all a
c stages , and a definitely positive @xmath7 = @xmath49 for stage b superhumps , confirming the suggestion in @xcite based on the combined @xmath0 diagram ( cf .
figure [ fig : iyumacomp2 ] ) .
although there was a signature of distinct stages during stage a , we listed a mean period in table [ tab : perlist ] .
there was a very clear signature of a stage b
c transition thanks to the high quality of data .
although the phases of superhump at late epochs ( @xmath50 ) deviate from extrapolations of stage c superhumps , they can still be interpreted as a continuation of stage c superhump , rather than traditional late superhumps .
this object showed a strong `` textbook '' beat phenomenon between superhumps and orbital modulations ( figure [ fig : iyumabeatoc ] ) .
the period of the beat phenomenon was shorter ( @xmath9 2.5 d ) during stage b , while it became longer ( @xmath9 3.0 d ) during stage c. these periods are in very good agreement with the expected beat periods for stage b and c superhumps , 2.45 d and 3.08 d , respectively .
this can be understood if the variation of the beat period reflects the variation of the angular velocity of the apsidal motion of the elliptical accretion disk .
the close correlation between the beat period and the superhump period suggests that the change in the angular velocity of the global apsidal motion is more responsible for the stage b
c transition rather than the appearance of a more localized new component .
( 88mm,70mm)fig32.eps ( 88mm,110mm)fig33.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 54934.6865 & 0.0078 & @xmath170.0294 & 52 + 9 & 54935.3637 & 0.0011 & @xmath170.0372 & 77 + 10 & 54935.4455 & 0.0021 & @xmath170.0315 & 74 + 11 & 54935.5222 & 0.0017 & @xmath170.0309 & 64 + 12 & 54935.5987 & 0.0025 & @xmath170.0305 & 33 + 22 & 54936.3877 & 0.0013 & @xmath170.0026 & 105 + 23 & 54936.4637 & 0.0010 & @xmath170.0027 & 121 + 24 & 54936.5418 & 0.0010 & @xmath170.0007 & 82 + 30 & 54937.0005 & 0.0003 & 0.0014 & 144 + 31 & 54937.0753 & 0.0002 & 0.0001 & 201 + 34 & 54937.3054 & 0.0002 & 0.0019 & 59 + 35 & 54937.3817 & 0.0003 & 0.0021 & 68 + 36 & 54937.4578 & 0.0003 & 0.0020 & 63 + 37 & 54937.5331 & 0.0003 & 0.0012 & 66 + 43 & 54937.9895 & 0.0007 & 0.0010 & 119 + 47 & 54938.3025 & 0.0010 & 0.0096 & 66 + 48 & 54938.3787 & 0.0009 & 0.0097 & 68 + 49 & 54938.4553 & 0.0008 & 0.0102 & 67 + 50 & 54938.5309 & 0.0007 & 0.0097 & 58 + 56 & 54938.9851 & 0.0003 & 0.0073 & 123 + 57 & 54939.0607 & 0.0002 & 0.0067 & 125 + 58 & 54939.1363 & 0.0003 & 0.0062 & 82 + 64 & 54939.5901 & 0.0007 & 0.0034 & 86 + 65 & 54939.6675 & 0.0002 & 0.0047 & 122 + 69 & 54939.9721 & 0.0003 & 0.0049 & 83 + 70 & 54940.0471 & 0.0004 & 0.0038 & 123 + 71 & 54940.1227 & 0.0007 & 0.0033 & 108 + 74 & 54940.3503 & 0.0009 & 0.0026 & 61 + 83 & 54941.0404 & 0.0017 & 0.0077 & 104 + 84 & 54941.1189 & 0.0009 & 0.0101 & 97 + 87 & 54941.3452 & 0.0003 & 0.0080 & 136 + 88 & 54941.4247 & 0.0004 & 0.0115 & 133 + 89 & 54941.4983 & 0.0004 & 0.0090 & 105 + 90 & 54941.5746 & 0.0002 & 0.0091 & 130 + 100 & 54942.3361 & 0.0002 & 0.0096 & 151 + 101 & 54942.4111 & 0.0002 & 0.0085 & 254 + 102 & 54942.4877 & 0.0003 & 0.0090 & 203 + 103 & 54942.5638 & 0.0004 & 0.0090 & 134 + 108 & 54942.9498 & 0.0008 & 0.0144 & 53 + 109 & 54943.0254 & 0.0006 & 0.0139 & 88 + 110 & 54943.1010 & 0.0004 & 0.0134 & 62 + 113 & 54943.3315 & 0.0012 & 0.0156 & 68 + 114 & 54943.4078 & 0.0008 & 0.0158 & 67 + 115 & 54943.4832 & 0.0007 & 0.0151 & 62 + 116 & 54943.5539 & 0.0008 & 0.0097 & 27 + 122 & 54944.0106 & 0.0005 & 0.0098 & 86 + 123 & 54944.0867 & 0.0005 & 0.0097 & 88 + 128 & 54944.4631 & 0.0004 & 0.0056 & 51 + 130 & 54944.6139 & 0.0008 & 0.0042 & 65 + 135 & 54944.9922 & 0.0004 & 0.0020 & 177 + 136 & 54945.0690 & 0.0004 & 0.0027 & 213 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 137 & 54945.1445 & 0.0006 & 0.0021 & 79 + 138 & 54945.2196 & 0.0008 & 0.0011 & 79 + 141 & 54945.4462 & 0.0010 & @xmath170.0006 & 37 + 142 & 54945.5220 & 0.0010 & @xmath170.0010 & 55 + 153 & 54946.3601 & 0.0009 & @xmath170.0000 & 67 + 154 & 54946.4323 & 0.0008 & @xmath170.0040 & 66 + 155 & 54946.5090 & 0.0006 & @xmath170.0033 & 65 + 156 & 54946.5854 & 0.0010 & @xmath170.0030 & 30 + 161 & 54946.9622 & 0.0016 & @xmath170.0067 & 44 + 162 & 54947.0420 & 0.0016 & @xmath170.0030 & 61 + 166 & 54947.3439 & 0.0007 & @xmath170.0055 & 72 + 167 & 54947.4212 & 0.0007 & @xmath170.0044 & 156 + 168 & 54947.4962 & 0.0005 & @xmath170.0055 & 134 + 169 & 54947.5722 & 0.0006 & @xmath170.0056 & 103 + 175 & 54948.0264 & 0.0008 & @xmath170.0081 & 139 + 179 & 54948.3261 & 0.0006 & @xmath170.0128 & 67 + 180 & 54948.4012 & 0.0005 & @xmath170.0138 & 67 + 181 & 54948.4776 & 0.0008 & @xmath170.0135 & 64 + 182 & 54948.5506 & 0.0015 & @xmath170.0166 & 64 + 189 & 54949.0981 & 0.0026 & @xmath170.0018 & 191 + 194 & 54949.4680 & 0.0016 & @xmath170.0124 & 66 + 195 & 54949.5410 & 0.0007 & @xmath170.0156 & 67 + 201 & 54950.0120 & 0.0041 & @xmath170.0012 & 169 + 202 & 54950.0831 & 0.0091 & @xmath170.0062 & 149 + 215 & 54951.0703 & 0.0056 & @xmath170.0084 & 108 + + + + the times of superhump maxima during the 2010 superoutburst are listed in table [ tab : ksumaoc2010 ] .
since there was a large gap in observations , we did not attempt to determine @xmath7 from these data . the observation was probably recorded during stages b and c. ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55304.0436 & 0.0004 & @xmath170.0009 & 92 + 1 & 55304.1148 & 0.0003 & 0.0002 & 100 + 2 & 55304.1837 & 0.0003 & @xmath170.0011 & 99 + 3 & 55304.2563 & 0.0006 & 0.0013 & 99 + 99 & 55310.9915 & 0.0013 & 0.0034 & 99 + 100 & 55311.0610 & 0.0011 & 0.0028 & 100 + 101 & 55311.1262 & 0.0011 & @xmath170.0021 & 96 + 102 & 55311.1953 & 0.0011 & @xmath170.0031 & 100 + 103 & 55311.2691 & 0.0012 & 0.0005 & 93 + 113 & 55311.9785 & 0.0015 & 0.0086 & 87 + 114 & 55312.0382 & 0.0023 & @xmath170.0019 & 95 + 115 & 55312.1023 & 0.0018 & @xmath170.0079 & 75 + + + + the times of superhump maxima during the 2010 superoutburst are listed in table [ tab : mrumaoc2010 ] .
since there were relatively large gaps in observations , we did not attempt to determine @xmath7 from these data . the period of the presumable stage c superhumps is listed in table [ tab : perlist ] .
an updated comparison of @xmath0 diagrams between different superoutbursts is given in figure [ fig : mrumacomp2 ] .
( 88mm,70mm)fig34.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55303.9991 & 0.0002 & 0.0004 & 155 + 1 & 55304.0645 & 0.0005 & 0.0007 & 107 + 2 & 55304.1286 & 0.0005 & @xmath170.0001 & 97 + 61 & 55307.9730 & 0.0012 & 0.0074 & 42 + 108 & 55311.0210 & 0.0009 & @xmath170.0011 & 150 + 109 & 55311.0877 & 0.0009 & 0.0007 & 157 + 110 & 55311.1519 & 0.0006 & @xmath170.0002 & 158 + 111 & 55311.2214 & 0.0011 & 0.0043 & 113 + 112 & 55311.2787 & 0.0012 & @xmath170.0035 & 89 + 123 & 55311.9932 & 0.0012 & @xmath170.0043 & 138 + 124 & 55312.0565 & 0.0016 & @xmath170.0061 & 155 + 125 & 55312.1196 & 0.0047 & @xmath170.0079 & 71 + 184 & 55315.9675 & 0.0006 & 0.0031 & 37 + 185 & 55316.0361 & 0.0019 & 0.0067 & 44 + + + + the 2010 superoutburst of this su uma - type dwarf nova was relatively well - observed during its later stage .
the times of superhump maxima are listed in table [ tab : tyvuloc2010 ] .
the period evolution now clearly demonstrates the presence of stage b
c transition , whose existence was suggested from the 2003 observation @xcite .
the sudden change in the superhump period at this transition favors the suggestion in @xcite that this object is analogous to ax cap and sdss j1627 .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55368.7467 & 0.0008 & @xmath170.0048 & 41 + 1 & 55368.8262 & 0.0004 & @xmath170.0055 & 39 + 9 & 55369.4721 & 0.0007 & @xmath170.0017 & 202 + 10 & 55369.5530 & 0.0003 & @xmath170.0010 & 188 + 13 & 55369.7944 & 0.0009 & @xmath170.0003 & 40 + 14 & 55369.8748 & 0.0008 & @xmath170.0002 & 39 + 21 & 55370.4365 & 0.0007 & @xmath170.0003 & 135 + 22 & 55370.5183 & 0.0008 & 0.0012 & 160 + 23 & 55370.5991 & 0.0008 & 0.0018 & 89 + 25 & 55370.7518 & 0.0030 & @xmath170.0060 & 20 + 26 & 55370.8390 & 0.0007 & 0.0009 & 40 + 27 & 55370.9180 & 0.0008 & @xmath170.0004 & 32 + 33 & 55371.3982 & 0.0090 & @xmath170.0017 & 64 + 34 & 55371.4837 & 0.0011 & 0.0036 & 158 + 35 & 55371.5635 & 0.0008 & 0.0031 & 154 + 37 & 55371.7267 & 0.0017 & 0.0059 & 37 + 38 & 55371.8037 & 0.0033 & 0.0026 & 39 + 39 & 55371.8884 & 0.0033 & 0.0070 & 22 + 46 & 55372.4457 & 0.0004 & 0.0026 & 105 + 47 & 55372.5236 & 0.0004 & 0.0002 & 163 + 48 & 55372.6074 & 0.0005 & 0.0037 & 74 + 58 & 55373.4079 & 0.0011 & 0.0017 & 51 + 59 & 55373.4872 & 0.0005 & 0.0007 & 156 + 60 & 55373.5673 & 0.0006 & 0.0005 & 146 + 75 & 55374.7689 & 0.0014 & @xmath170.0016 & 27 + 76 & 55374.8473 & 0.0019 & @xmath170.0035 & 27 + 88 & 55375.8077 & 0.0021 & @xmath170.0061 & 58 + 89 & 55375.8919 & 0.0033 & @xmath170.0022 & 109 + + + + in @xcite , we reported observations of the 2008 superoutburst of this object ( = hs 0417@xmath17445 , hereafter 1rxs j0423 ) .
the 2010 superoutburst was again fortunately detected during its rising stage and early stage evolution of superhumps was recorded .
the times of superhump maxima are listed in table [ tab : j0423oc2010 ] . a stage
a b and transition was clearly recorded . the shorter superhump period after
@xmath51 probably corresponds to stage c superhumps . although the later half of the stage b and the stage b c transition itself
were not observed , there was a possible indication of a positive @xmath7 during the stage b. the parameters are listed in table [ tab : perlist ] .
figure [ fig : j0423occomp ] illustrates the comparison of @xmath0 diagrams between the 2008 and 2010 superoutbursts . although the later parts of the @xmath0 diagrams were similar , there was a distinction during the early stage
this may have been a result of the presence of a precursor outburst and early appearance of superhumps during the 2008 superoutburst ( cf .
the @xmath0 evolution in the 2010 superoutburst resembled those of ordinary su uma - type dwarf novae than in the 2008 superoutburst .
( 88mm,70mm)fig35.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55442.1696 & 0.0006 & @xmath170.0155 & 124 + 1 & 55442.2491 & 0.0005 & @xmath170.0141 & 94 + 12 & 55443.1233 & 0.0003 & @xmath170.0007 & 114 + 13 & 55443.1995 & 0.0004 & @xmath170.0028 & 309 + 14 & 55443.2795 & 0.0008 & @xmath170.0010 & 108 + 25 & 55444.1410 & 0.0004 & @xmath170.0003 & 268 + 26 & 55444.2188 & 0.0002 & @xmath170.0007 & 470 + 27 & 55444.2979 & 0.0003 & 0.0001 & 406 + 38 & 55445.1623 & 0.0007 & 0.0037 & 67 + 39 & 55445.2402 & 0.0018 & 0.0034 & 42 + 40 & 55445.3196 & 0.0007 & 0.0045 & 73 + 79 & 55448.3718 & 0.0009 & 0.0049 & 47 + 80 & 55448.4501 & 0.0004 & 0.0050 & 79 + 81 & 55448.5259 & 0.0004 & 0.0025 & 76 + 82 & 55448.6063 & 0.0005 & 0.0047 & 79 + 83 & 55448.6825 & 0.0010 & 0.0026 & 62 + 88 & 55449.0717 & 0.0020 & 0.0005 & 157 + 89 & 55449.1554 & 0.0008 & 0.0061 & 240 + 90 & 55449.2320 & 0.0009 & 0.0043 & 475 + 91 & 55449.3099 & 0.0005 & 0.0041 & 205 + 92 & 55449.3840 & 0.0024 & @xmath170.0001 & 49 + 93 & 55449.4716 & 0.0016 & 0.0093 & 80 + 94 & 55449.5447 & 0.0015 & 0.0041 & 78 + 95 & 55449.6240 & 0.0009 & 0.0052 & 79 + 96 & 55449.7020 & 0.0016 & 0.0048 & 42 + 102 & 55450.1716 & 0.0009 & 0.0050 & 218 + 103 & 55450.2496 & 0.0007 & 0.0047 & 197 + 114 & 55451.1109 & 0.0021 & 0.0053 & 90 + 115 & 55451.1783 & 0.0080 & @xmath170.0056 & 114 + 116 & 55451.2728 & 0.0026 & 0.0106 & 108 + 140 & 55453.1084 & 0.0032 & @xmath170.0318 & 77 + 141 & 55453.2028 & 0.0042 & @xmath170.0157 & 87 + 142 & 55453.2896 & 0.0033 & @xmath170.0071 & 85 + + + + the period evolution of this su uma - type dwarf nova during the 2005 and 2008 superoutbursts has been described in @xcite and @xcite .
we also observed the 2009 superoutburst which was accompanied by a precursor outburst , as in the 2005 one ( figure [ fig : j05322009oc ] ) .
the times of superhump maxima are listed in table [ tab : j0532oc2009 ] .
superhumps were already present during the fading stage of the rebrightening , and the period was smoothly decreasing as in the 2005 one @xcite .
this indicates that the stage b , rather than stage a , already started during the precursor outburst .
this behavior very well reproduced the features observed during the 2005 superoutburst .
the observed @xmath7 for stage b ( @xmath52 ) was @xmath53 , similar to the one observed in 2008 , but is larger than in 2005 .
the 2009 outburst was also well - observed during the post - superoutburst stage .
although times of individual superhumps were not sufficiently measured due to strong flickering , the period analysis has yielded a periodicity of 0.05690(2 ) d ( bjd 2455081.22455090.6 , figure [ fig : j0532lateshpdm ] ) , which is in agreement with the period of stage c superhumps .
the signal from the orbital period , if present , was still much smaller than the superhump signal .
this analysis suggests the long endurance of superhumps even after the termination of the superoutburst .
( 88mm,90mm)fig36.eps ( 88mm,110mm)fig37.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55068.3087 & 0.0016 & 0.0047 & 106 + 4 & 55068.5335 & 0.0008 & 0.0011 & 114 + 5 & 55068.5967 & 0.0016 & 0.0072 & 72 + 19 & 55069.3875 & 0.0004 & @xmath170.0013 & 106 + 20 & 55069.4457 & 0.0005 & @xmath170.0003 & 114 + 21 & 55069.5014 & 0.0004 & @xmath170.0016 & 111 + 22 & 55069.5574 & 0.0003 & @xmath170.0027 & 113 + 71 & 55072.3504 & 0.0003 & @xmath170.0075 & 112 + 72 & 55072.4075 & 0.0004 & @xmath170.0075 & 114 + 73 & 55072.4669 & 0.0009 & @xmath170.0052 & 36 + 74 & 55072.5226 & 0.0004 & @xmath170.0066 & 72 + 75 & 55072.5791 & 0.0003 & @xmath170.0071 & 107 + 105 & 55074.2935 & 0.0008 & @xmath170.0056 & 108 + 123 & 55075.3249 & 0.0019 & @xmath170.0019 & 115 + 124 & 55075.3828 & 0.0014 & @xmath170.0012 & 102 + 125 & 55075.4411 & 0.0009 & 0.0000 & 97 + 126 & 55075.4958 & 0.0011 & @xmath170.0023 & 108 + 127 & 55075.5598 & 0.0011 & 0.0045 & 114 + 128 & 55075.6117 & 0.0015 & @xmath170.0006 & 77 + 140 & 55076.3047 & 0.0014 & 0.0072 & 112 + 141 & 55076.3688 & 0.0013 & 0.0142 & 114 + 142 & 55076.4220 & 0.0010 & 0.0103 & 113 + 143 & 55076.4764 & 0.0009 & 0.0076 & 112 + 144 & 55076.5335 & 0.0006 & 0.0076 & 107 + 145 & 55076.5907 & 0.0010 & 0.0077 & 110 + 209 & 55080.2372 & 0.0005 & 0.0001 & 145 + 210 & 55080.2932 & 0.0004 & @xmath170.0011 & 172 + 213 & 55080.4643 & 0.0011 & @xmath170.0013 & 63 + 214 & 55080.5134 & 0.0017 & @xmath170.0093 & 54 + 215 & 55080.5705 & 0.0006 & @xmath170.0092 & 44 + + + + asas [email protected] ( hereafter asas j2243 ) was selected as a dwarf nova by p. wils ( cvnet - discussion 1320 ; @xcite ) .
the 2009 outburst and the superhumps were detected by i. miller ( cvnet - outburst 3358 ) .
the object showed well - developed superhumps and their time - evolution was intensively studied .
the times of superhump maxima are listed in table [ tab : asas2243oc2009 ] . as also reported in @xcite , a textbook stage b
c transition was observed .
the @xmath7 during stage b was @xmath54 ( @xmath55 ) .
of @xmath56 .
our analysis of their timing data yielded @xmath7 of @xmath57 ( @xmath58 ) by our definition . ]
the object also showed a post - superoutburst rebrightening ( figure [ fig : asas22432009oc ] ) .
( 88mm,90mm)fig38.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55113.0454 & 0.0002 & 0.0005 & 132 + 1 & 55113.1144 & 0.0003 & @xmath170.0003 & 131 + 2 & 55113.1842 & 0.0003 & @xmath170.0001 & 132 + 5 & 55113.3918 & 0.0003 & @xmath170.0017 & 64 + 6 & 55113.4596 & 0.0003 & @xmath170.0036 & 75 + 7 & 55113.5328 & 0.0003 & @xmath170.0001 & 67 + 10 & 55113.7392 & 0.0002 & @xmath170.0028 & 182 + 11 & 55113.8079 & 0.0002 & @xmath170.0039 & 182 + 15 & 55114.0866 & 0.0003 & @xmath170.0040 & 134 + 23 & 55114.6442 & 0.0002 & @xmath170.0041 & 184 + 24 & 55114.7131 & 0.0003 & @xmath170.0049 & 180 + 25 & 55114.7841 & 0.0003 & @xmath170.0036 & 179 + 26 & 55114.8539 & 0.0005 & @xmath170.0035 & 104 + 29 & 55115.0633 & 0.0004 & @xmath170.0033 & 109 + 30 & 55115.1365 & 0.0007 & 0.0003 & 64 + 36 & 55115.5479 & 0.0010 & @xmath170.0067 & 16 + 38 & 55115.6896 & 0.0004 & @xmath170.0044 & 55 + 47 & 55116.3174 & 0.0004 & @xmath170.0039 & 57 + 48 & 55116.3881 & 0.0003 & @xmath170.0030 & 83 + 49 & 55116.4581 & 0.0005 & @xmath170.0027 & 69 + 50 & 55116.5269 & 0.0007 & @xmath170.0036 & 31 + 51 & 55116.6073 & 0.0017 & 0.0071 & 39 + 52 & 55116.6714 & 0.0013 & 0.0014 & 49 + 53 & 55116.7374 & 0.0005 & @xmath170.0023 & 104 + 54 & 55116.8085 & 0.0022 & @xmath170.0009 & 69 + 56 & 55116.9463 & 0.0019 & @xmath170.0025 & 30 + 62 & 55117.3650 & 0.0009 & @xmath170.0020 & 80 + 63 & 55117.4344 & 0.0005 & @xmath170.0023 & 87 + 64 & 55117.5049 & 0.0005 & @xmath170.0015 & 76 + 65 & 55117.5769 & 0.0018 & 0.0008 & 22 + 71 & 55117.9965 & 0.0006 & 0.0020 & 196 + 72 & 55118.0654 & 0.0005 & 0.0012 & 176 + 75 & 55118.2775 & 0.0011 & 0.0042 & 26 + 76 & 55118.3461 & 0.0009 & 0.0030 & 35 + 77 & 55118.4148 & 0.0010 & 0.0021 & 58 + 78 & 55118.4790 & 0.0013 & @xmath170.0034 & 56 + 79 & 55118.5559 & 0.0013 & 0.0038 & 57 + 89 & 55119.2535 & 0.0025 & 0.0043 & 24 + 91 & 55119.3946 & 0.0012 & 0.0059 & 56 + 92 & 55119.4650 & 0.0007 & 0.0066 & 65 + 93 & 55119.5342 & 0.0009 & 0.0060 & 32 + 94 & 55119.6046 & 0.0007 & 0.0068 & 90 + 95 & 55119.6740 & 0.0005 & 0.0065 & 81 + 96 & 55119.7461 & 0.0005 & 0.0088 & 83 + 97 & 55119.8138 & 0.0005 & 0.0068 & 79 + 100 & 55120.0233 & 0.0004 & 0.0072 & 143 + 101 & 55120.0937 & 0.0004 & 0.0079 & 149 + 106 & 55120.4399 & 0.0005 & 0.0055 & 63 + 108 & 55120.5797 & 0.0007 & 0.0060 & 73 + 109 & 55120.6490 & 0.0004 & 0.0055 & 91 + 110 & 55120.7186 & 0.0004 & 0.0054 & 88 + 111 & 55120.7870 & 0.0006 & 0.0041 & 88 + 118 & 55121.2762 & 0.0015 & 0.0053 & 113 + 119 & 55121.3420 & 0.0004 & 0.0014 & 156 + 120 & 55121.4116 & 0.0007 & 0.0013 & 182 + 121 & 55121.4811 & 0.0008 & 0.0011 & 149 + 122 & 55121.5515 & 0.0009 & 0.0017 & 92 + 123 & 55121.6190 & 0.0003 & @xmath170.0005 & 182 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 124 & 55121.6885 & 0.0003 & @xmath170.0006 & 182 + 125 & 55121.7587 & 0.0005 & @xmath170.0002 & 184 + 126 & 55121.8274 & 0.0004 & @xmath170.0012 & 130 + 131 & 55122.1763 & 0.0009 & @xmath170.0008 & 131 + 133 & 55122.3165 & 0.0015 & @xmath170.0001 & 24 + 134 & 55122.3975 & 0.0012 & 0.0112 & 140 + 135 & 55122.4514 & 0.0007 & @xmath170.0046 & 220 + 136 & 55122.5235 & 0.0057 & @xmath170.0022 & 85 + 137 & 55122.6033 & 0.0005 & 0.0079 & 178 + 138 & 55122.6636 & 0.0007 & @xmath170.0016 & 181 + 139 & 55122.7466 & 0.0010 & 0.0117 & 178 + 140 & 55122.7979 & 0.0007 & @xmath170.0067 & 180 + 151 & 55123.5594 & 0.0026 & @xmath170.0120 & 94 + 152 & 55123.6415 & 0.0013 & 0.0004 & 178 + 157 & 55123.9831 & 0.0009 & @xmath170.0066 & 147 + 167 & 55124.6724 & 0.0008 & @xmath170.0144 & 92 + 173 & 55125.0856 & 0.0013 & @xmath170.0195 & 149 + 174 & 55125.1588 & 0.0046 & @xmath170.0160 & 73
+ + + + lanning 420 was selected as a uv - bright transient object @xcite , which was considered to be a possible nova because of the absence on the digitized sky survey image .
@xcite listed lanning cvs and further investigated these objects .
s. brady indeed detected an outburst in 2007 and another one in 2010 , which turned out to be a superoutburst ( baavss alert 2374 ) .
follow - up observation confirmed the presence of superhumps ( vsnet - alert 12131 , 12132 , 12138 ; figure [ fig : lanning420shpdm ] ) .
the times of superhump maxima are listed in table [ tab : lanning4202010 ] .
there was a distinct shortening of the superhump period around @xmath59 , which we interpreted as a stage b
c transition .
the last two epochs were measured during the early post - superoutburst stage .
we included these epochs because short-@xmath3 tend to show persistent superhumps during this stage and the times of maxima were in good agreement with extrapolated stage c superhumps .
the measured period derivative during the stage b had a relatively large error because only the late stage of the stage b was observed .
observations of the earlier stage are crucial to better characterize the period variation in this system .
( 88mm,110mm)fig39.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55438.5841 & 0.0005 & @xmath170.0084 & 20 + 1 & 55438.6469 & 0.0005 & @xmath170.0070 & 26 + 2 & 55438.7074 & 0.0005 & @xmath170.0078 & 24 + 3 & 55438.7691 & 0.0005 & @xmath170.0075 & 25 + 4 & 55438.8277 & 0.0018 & @xmath170.0102 & 24 + 14 & 55439.4478 & 0.0011 & @xmath170.0033 & 51 + 15 & 55439.5051 & 0.0018 & @xmath170.0074 & 46 + 16 & 55439.5672 & 0.0011 & @xmath170.0066 & 89 + 17 & 55439.6314 & 0.0012 & @xmath170.0037 & 86 + 18 & 55439.6895 & 0.0010 & @xmath170.0069 & 88 + 19 & 55439.7535 & 0.0008 & @xmath170.0042 & 76 + 20 & 55439.8132 & 0.0012 & @xmath170.0059 & 46 + 21 & 55439.8767 & 0.0009 & @xmath170.0037 & 60 + 22 & 55439.9369 & 0.0008 & @xmath170.0048 & 56 + 23 & 55439.9982 & 0.0012 & @xmath170.0049 & 45 + 24 & 55440.0589 & 0.0015 & @xmath170.0055 & 125 + 25 & 55440.1225 & 0.0030 & @xmath170.0032 & 115 + 31 & 55440.4990 & 0.0049 & 0.0054 & 20 + 32 & 55440.5511 & 0.0022 & @xmath170.0039 & 21 + 33 & 55440.6082 & 0.0073 & @xmath170.0081 & 19 + 34 & 55440.6785 & 0.0020 & 0.0008 & 49 + 35 & 55440.7379 & 0.0006 & @xmath170.0010 & 56 + 36 & 55440.7997 & 0.0009 & @xmath170.0006 & 31 + 37 & 55440.8611 & 0.0006 & @xmath170.0005 & 45 + 38 & 55440.9243 & 0.0007 & 0.0014 & 56 + 39 & 55440.9852 & 0.0023 & 0.0009 & 163 + 40 & 55441.0392 & 0.0054 & @xmath170.0064 & 128 + 41 & 55441.1128 & 0.0018 & 0.0059 & 253 + 42 & 55441.1713 & 0.0019 & 0.0031 & 115 + 46 & 55441.4149 & 0.0030 & 0.0014 & 15 + 47 & 55441.4792 & 0.0028 & 0.0044 & 21 + 48 & 55441.5345 & 0.0020 & @xmath170.0017 & 25 + 49 & 55441.5992 & 0.0038 & 0.0017 & 39 + 50 & 55441.6608 & 0.0016 & 0.0019 & 48 + 51 & 55441.7225 & 0.0008 & 0.0023 & 73 + 52 & 55441.7874 & 0.0017 & 0.0059 & 60 + 53 & 55441.8443 & 0.0020 & 0.0015 & 53 + 54 & 55441.9112 & 0.0015 & 0.0071 & 56 + 55 & 55441.9710 & 0.0019 & 0.0056 & 56 + 62 & 55442.4036 & 0.0011 & 0.0088 & 21 + 63 & 55442.4636 & 0.0010 & 0.0075 & 71 + 64 & 55442.5285 & 0.0026 & 0.0111 & 21 + 65 & 55442.5846 & 0.0016 & 0.0059 & 21 + 66 & 55442.6469 & 0.0029 & 0.0068 & 21 + 76 & 55443.2586 & 0.0018 & 0.0053 & 110 + 78 & 55443.3918 & 0.0051 & 0.0159 & 35 + 79 & 55443.4458 & 0.0014 & 0.0085 & 48 + 80 & 55443.5039 & 0.0015 & 0.0053 & 46 + 81 & 55443.5651 & 0.0017 & 0.0051 & 48 + 82 & 55443.6274 & 0.0015 & 0.0061 & 45 + 83 & 55443.6835 & 0.0083 & 0.0009 & 17 + 92 & 55444.2430 & 0.0024 & 0.0085 & 122 + 93 & 55444.3012 & 0.0021 & 0.0053 & 63 + 94 & 55444.3635 & 0.0024 & 0.0064 & 15 + 95 & 55444.4301 & 0.0130 & 0.0116 & 21 + 96 & 55444.4863 & 0.0023 & 0.0065 & 20 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 97 & 55444.5442 & 0.0020 & 0.0031 & 15 + 98 & 55444.6082 & 0.0026 & 0.0058 & 19 + 99 & 55444.6637 & 0.0014 & @xmath170.0001 & 19 + 109 & 55445.2851 & 0.0155 & 0.0081 & 65 + 116 & 55445.7083 & 0.0024 & 0.0020 & 17 + 117 & 55445.7670 & 0.0032 & @xmath170.0006 & 16 + 118 & 55445.8309 & 0.0017 & 0.0019 & 16 + 124 & 55446.1980 & 0.0028 & 0.0011 & 111 + 125 & 55446.2583 & 0.0046 & 0.0001 & 80 + 126 & 55446.3132 & 0.0059 & @xmath170.0063 & 60 + 174 & 55449.2331 & 0.0071 & @xmath170.0301 & 117 + 175 & 55449.2880 & 0.0028 & @xmath170.0365 & 109 + + + + this object ( hereafter pg 0149 ) was originally discovered as an eruptive object with strong uv excess , and was suspected to be a supernova @xcite .
@xcite selected this object during the course of the sdss survey and classified it as a cv ( dwarf nova ) .
the object as been monitored as a dwarf nova since then .
@xcite photometrically identified a @xmath2 of 0.08242(3 ) d. the 2009 september outburst of this object was detected by the the catalina real - time transient survey ( crts , @xcite)http://nesssi.cacr.caltech.edu / catalina/@xmath6 . for the information of the individual catalina cvs , see @xmath60http://nesssi.cacr.caltech.edu / catalina / allcv.html@xmath6 . ]
(= css090911:015152@xmath1140047 ) .
superhumps were detected immediately following the announcement ( vsnet - alert 11465 ) .
the mean @xmath3 with the pdm method was 0.08495(2 ) d ( figure [ fig : pg0149shpdm ] ) .
the times of superhump maxima are summarized in table [ tab : pg0149oc2009 ] .
there was a clear stage b
c transition with a positive @xmath7 = @xmath61 ( @xmath62 ) .
such a positive @xmath7 is rare for su uma - type dwarf novae with this @xmath3 .
the behavior in the superhump period resemble that of long-@xmath3 systems like qw ser ( cf .
@xcite ; @xcite ) .
( 88mm,110mm)fig40.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55086.5129 & 0.0008 & @xmath170.0038 & 44 + 1 & 55086.5987 & 0.0003 & @xmath170.0028 & 89 + 11 & 55087.4448 & 0.0007 & @xmath170.0055 & 57 + 12 & 55087.5312 & 0.0005 & @xmath170.0040 & 114 + 13 & 55087.6173 & 0.0005 & @xmath170.0029 & 89 + 23 & 55088.4654 & 0.0006 & @xmath170.0036 & 40 + 24 & 55088.5484 & 0.0006 & @xmath170.0054 & 83 + 55 & 55091.1917 & 0.0006 & 0.0065 & 180 + 59 & 55091.5320 & 0.0007 & 0.0073 & 77 + 60 & 55091.6176 & 0.0008 & 0.0080 & 89 + 70 & 55092.4661 & 0.0006 & 0.0076 & 91 + 71 & 55092.5517 & 0.0007 & 0.0083 & 130 + 72 & 55092.6371 & 0.0006 & 0.0089 & 111 + 93 & 55094.4113 & 0.0014 & 0.0005 & 40 + 94 & 55094.4977 & 0.0009 & 0.0020 & 45 + 95 & 55094.5776 & 0.0009 & @xmath170.0029 & 44 + 103 & 55095.2556 & 0.0008 & @xmath170.0039 & 146 + 154 & 55099.5741 & 0.0013 & @xmath170.0145 & 78
+ + + + this object ( hereafter rx j1715 ) is a cv identified from the rosat north ecliptic pole survey @xcite .
@xcite reported the detection of a 1.64(2 ) hr period from radial - velocity study .
@xcite identified the dwarf - nova type behavior and detected superhumps during a superoutburst in 2009 august .
we analyzed the combined data from @xcite ( from the aavso database ) and our own observations .
as reported in @xcite , these observations covered the final stage of the superoutburst .
the times of maxima are listed in table [ tab : j1715oc2009 ] .
since the @xmath0 s for maxima for @xmath63 largely deviate from the earlier trend , these humps are less likely a continuation of superhumps observed during the plateau phase ( @xmath64 ) . restricting to @xmath64
, we obtained a mean @xmath3 of 0.07074(4 ) d with the pdm method ( figure [ fig : j1715shpdm ] ) . and
is in agreement with the corresponding period of 0.07086(78 ) d by @xcite within respective errors .
the @xmath13 against @xmath2 was 3.5 % . since the observation covered the final part of the superoutburst , these superhumps are likely stage c superhumps .
although @xcite interpreted observed humps for @xmath63 ( rapid fading stage and post - superoutburst stage ) as orbital humps , the reported period [ 0.06944(90 ) d ] is somewhat different from the one reported from radial - velocity study .
our analysis also confirmed a significant phase offset during the rapid fading stage ( figure [ fig : j1715prof ] ) .
we could not , however , determine whether the newly arising signals were from orbital humps or were from traditional late superhumps because the amplitude on jd 2455069 was very small and it was difficult to derive a meaningful period based on the last two nights of observations .
future observations are needed to confirm the nature of these large - amplitude humps . since stage c superhumps frequently endure during the post - superoutburst stage in many well - observed systems ( cf .
@xcite ) , the early disappearance of superhumps in rx j1715 appears to be rather unique .
the case may be similar to dt oct ( cf .
@xcite ) , which showed an early switch to traditional late superhumps .
although both relatively low outburst amplitudes and the relatively strong x - ray emission are common to these objects , the frequency of outbursts in rx j1715 appears to be smaller @xcite .
a further detailed comparison between these objects might be fruitful .
( 88mm,110mm)fig41.eps ( 88mm,110mm)fig42.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55065.3776 & 0.0007 & 0.0037 & 61 + 1 & 55065.4528 & 0.0010 & 0.0078 & 87 + 2 & 55065.5181 & 0.0008 & 0.0021 & 30 + 23 & 55067.0079 & 0.0080 & @xmath170.0000 & 86 + 39 & 55068.1405 & 0.0021 & @xmath170.0041 & 134 + 42 & 55068.3635 & 0.0014 & 0.0058 & 62 + 43 & 55068.4183 & 0.0014 & @xmath170.0105 & 90 + 44 & 55068.4953 & 0.0015 & @xmath170.0045 & 101 + 47 & 55068.7020 & 0.0021 & @xmath170.0110 & 37 + 48 & 55068.7722 & 0.0024 & @xmath170.0118 & 27 + 63 & 55069.8143 & 0.0045 & @xmath170.0353 & 37 + 63 & 55069.8756 & 0.0019 & 0.0259 & 38 + 64 & 55069.9378 & 0.0010 & 0.0171 & 32 + 75 & 55070.7114 & 0.0019 & 0.0092 & 27 + 76 & 55070.7752 & 0.0016 & 0.0020 & 25 + 77 & 55070.8431 & 0.0024 & @xmath170.0012 & 26 + 78 & 55070.9205 & 0.0012 & 0.0051 & 24
+ + + + this object ( hereafter sdss j0129 ) is an am cvn - type cv selected during the course of the sdss @xcite , who reported broad hei emission lines . the orbital period has not been reported yet .
the object was reported in outburst on 2009 november 29 at a unfiltered ccd magnitude of 14.5 ( cvnet - outburst 3479 , vsnet - alert 11702 ) .
two days after this detection , the object started to fade rapidly . immediately following this transient fading ,
the object experienced rebrightenings on december 4 and 18 ( vsnet - alert 11707 , 11737 ) .
the overall behavior of the outburst ( figure [ fig : j0129lc ] ) was extremely similar to the 2003 august superoutburst of v803 cen ( @xcite ; vsnet - alert 11709 ) .
the object developed hump signals during its terminal stage of the plateau phase ( figures [ fig : j0129shpdm ] , [ fig : j0129shprof ] ) .
although these humps may reflect orbital modulations , we identified them as superhumps based on extreme analogy with v803 cen @xcite .
the mean period determined with the pdm method was 0.01805(10 ) d. the period is also very similar to that of v803 cen ( @xmath65 = 0.018686(4 ) d = 1614.5(4 ) s , @xcite ) .
the times of maxima during the fading branch from the superoutburst are listed in table [ tab : j0129oc2009 ] .
( 88mm,110mm)fig43.eps ( 88mm,110mm)fig44.eps ( 88mm,110mm)fig45.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55166.3297 & 0.0030 & @xmath170.0020 & 40 + 1 & 55166.3491 & 0.0037 & @xmath170.0007 & 40 + 2 & 55166.3686 & 0.0012 & 0.0007 & 41 + 4 & 55166.4080 & 0.0013 & 0.0039 & 41 + 5 & 55166.4231 & 0.0006 & 0.0009 & 52 + 6 & 55166.4433 & 0.0013 & 0.0031 & 62 + 7 & 55166.4567 & 0.0012 & @xmath170.0016 & 61 + 8 & 55166.4739 & 0.0015 & @xmath170.0026 & 22 + 18 & 55166.6542 & 0.0011 & @xmath170.0032 & 17 + 19 & 55166.6730 & 0.0011 & @xmath170.0024 & 17 + 20 & 55166.6947 & 0.0017 & 0.0012 & 17 + 22 & 55166.7329 & 0.0019 & 0.0031 & 17 + 23 & 55166.7458 & 0.0017 & @xmath170.0020 & 17 + 25 & 55166.7840 & 0.0010 & 0.0000 & 17 + 26 & 55166.8034 & 0.0012 & 0.0013 & 17 + 28 & 55166.8384 & 0.0010 & 0.0001 & 17 + 29 & 55166.8547 & 0.0017 & @xmath170.0016 & 16 + 30 & 55166.8764 &
0.0008 & 0.0020 & 16 + + + + in addition to the 2004 superoutburst of this object ( hereafter sdss j0310 ) , two superoutbursts were recorded in 2009 .
table [ tab : j0310oc2009b ] gives the times of superhump maxima during its second superoutburst in 2009 ( 2009 november , designated as 2009b in table in order to avoid confusion with the 2009 superoutburst described in @xcite ) .
the mean @xmath3 during this superoutburst was 0.06786(3 ) d ( pdm method , alias selected based on the 2004 observation ) , which is shorter than the period obtained during the 2004 superoutburst .
although this difference may have resulted from different stages observed in different outbursts , the difference appears to be larger than those associated with typical stage b c transitions @xcite .
this needs to be clarified by future observations .
the shortest interval between superoutbursts was 284 d , which is likely the supercycle of this object .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55144.5155 & 0.0003 & 0.0028 & 67 + 1 & 55144.5777 & 0.0003 & @xmath170.0029 & 73 + 31 & 55146.6161 & 0.0009 & 0.0001 & 61
+ + + + this object ( hereafter sdss j0732 ) was selected using sdss and crts data as a candidate dwarf nova by @xcite .
the object was detected in outburst by j. shears on 2009 december 31 at an unfiltered ccd magnitude of 16.2 ( cvnet - outburst 3528 ; @xcite ) .
subsequent observations confirmed the presence of superhumps ( cvnet - outburst 3535 , figure [ fig : j0732shpdm ] ) .
the times of superhump maxima determined from aavso observations are listed in table [ tab : j0732oc2010 ] .
the object showed a clear stage b
c transition around @xmath66 .
the @xmath7 for stage b was @xmath67 .
of @xmath68 .
they used a data set including our present data set .
we do not attempt to choose a better @xmath7 between ours and theirs , since @xmath7 is dependent on the segment used , and the observed baseline for the stage b was too short .
the @xmath7 determined from the timing data in @xcite was @xmath69 ( @xmath70 ) while it was @xmath71 for @xmath72 ( both values are by our definition of @xmath7 ) .
] relatively few objects with similar @xmath3 , including rz leo and qy per , are known to show positive @xmath7 @xcite .
although the relatively large ( 5.2 mag ) outburst amplitude also might suggest a low mass - transfer rate as in rz leo and qy per , the outburst frequency estimated in @xcite is much higher than in these objects .
an exact determination of @xmath7 during the next superoutburst is desired .
( 88mm,110mm)fig46.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55199.6270 & 0.0020 & @xmath170.0080 & 42 + 1 & 55199.7096 & 0.0007 & @xmath170.0050 & 83 + 2 & 55199.7887 & 0.0006 & @xmath170.0054 & 76 + 3 & 55199.8666 & 0.0006 & @xmath170.0071 & 81 + 4 & 55199.9462 & 0.0008 & @xmath170.0071 & 82 + 10 & 55200.4238 & 0.0018 & @xmath170.0071 & 60 + 14 & 55200.7446 & 0.0006 & @xmath170.0047 & 82 + 15 & 55200.8244 & 0.0007 & @xmath170.0044 & 68 + 16 & 55200.9040 & 0.0006 & @xmath170.0044 & 82 + 17 & 55200.9833 & 0.0011 & @xmath170.0047 & 73 + 26 & 55201.7046 & 0.0006 & 0.0003 & 82 + 27 & 55201.7859 & 0.0007 & 0.0020 & 82 + 28 & 55201.8656 & 0.0008 & 0.0020 & 74 + 29 & 55201.9427 & 0.0010 & @xmath170.0004 & 82 + 30 & 55202.0233 & 0.0012 & 0.0005 & 83 + 38 & 55202.6663 & 0.0010 & 0.0068 & 71 + 39 & 55202.7456 & 0.0007 & 0.0065 & 82 + 40 & 55202.8253 & 0.0006 & 0.0066 & 76 + 41 & 55202.9044 & 0.0009 & 0.0061 & 82 + 42 & 55202.9841 & 0.0011 & 0.0062 & 75 + 59 & 55204.3438 & 0.0017 & 0.0129 & 88 + 60 & 55204.4230 & 0.0011 & 0.0125 & 90 + 75 & 55205.6122 & 0.0006 & 0.0077 & 73 + 76 & 55205.6924 & 0.0007 & 0.0083 & 83 + 77 & 55205.7727 & 0.0008 & 0.0090 & 82 + 78 & 55205.8501 & 0.0007 & 0.0069 & 79 + 79 & 55205.9295 & 0.0007 & 0.0067 & 82 + 90 & 55206.8014 & 0.0006 & 0.0031 & 76 + 91 & 55206.8798 & 0.0008 & 0.0019 & 82 + 92 & 55206.9602 & 0.0010 & 0.0027 & 67 + 114 & 55208.7037 & 0.0013 & @xmath170.0049 & 82 + 115 & 55208.7840 & 0.0011 & @xmath170.0042 & 75 + 116 & 55208.8664 & 0.0010 & @xmath170.0014 & 83 + 117 & 55208.9413 & 0.0011 & @xmath170.0061 & 75 + 118 & 55209.0165 & 0.0023 & @xmath170.0104 & 82 + 126 & 55209.6628 & 0.0019 & @xmath170.0009 & 83 + 127 & 55209.7402 & 0.0013 & @xmath170.0031 & 76 + 128 & 55209.8098 & 0.0014 & @xmath170.0130 & 76 + 129 & 55209.8958 & 0.0014 & @xmath170.0067 & 82 + + + + this object ( hereafter sdss j0838 ) underwent another superoutburst in 2010 in addition to the 2007 and 2009 ones discussed in @xcite .
the 2010 superoutburst was notable in that it was preceded by a prominent precursor outburst ( vsnet - alert 11910 ) .
although only two superhump maxima were measured , we listed them in table [ tab : j0838oc2010 ] .
the interval between maxima is in agreement with the period of stage c superhumps in 2009 .
since the 2010 observation was undertaken during the early stage of the superoutburst , the period is expected to be close to the @xmath3 at the start of stage b. as discussed in @xcite , this period is expected to be close to @xmath3 for stage c superhumps .
the present finding is consistent with this interpretation .
there was a slight indication of modulation at a period around 0.0709 d and an amplitude of 0.07 mag .
although the presence of this signal is not conclusive , the periodicity might suggest a rare evolution of @xmath3 starting from @xmath2 , as recorded in qz vir @xcite .
this needs to be confirmed by future observations .
cccc @xmath15 & max & error & @xmath16 + 0 & 55294.9551 & 0.0008 & 99 + 1 & 55295.0289 & 0.0004 & 153
+ + + this object ( hereafter sdss j0839 ) was selected as a cv during the course of sdss @xcite .
although the spectrum was suggestive of that of a dwarf nova , no outburst had been recorded . on 2010
april 8 , k. itagaki detected an outbursting object which can be identified with this cv ( vsnet - alert 11911 ) .
subsequent observations have confirmed the presence of superhumps ( vsnet - alert 11916 , 11921 , 11926 , 11930 ) .
the times of superhump maxima are listed in table [ tab : j0839oc2010 ] .
there was an apparent break in the @xmath0 diagram around @xmath24 , which is likely a stage b
c transition .
the periods given in table [ tab : perlist ] are based on this interpretation .
the mean @xmath3 determined with the pdm method is 0.078423(7 ) d ( figure [ fig : j0839shpdm ] ) .
( 88mm,110mm)fig47.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55295.3531 & 0.0002 & @xmath170.0003 & 190 + 1 & 55295.4309 & 0.0002 & @xmath170.0009 & 220 + 2 & 55295.5099 & 0.0004 & @xmath170.0003 & 83 + 13 & 55296.3732 & 0.0004 & 0.0006 & 79 + 14 & 55296.4518 & 0.0003 & 0.0007 & 78 + 15 & 55296.5311 & 0.0007 & 0.0016 & 41 + 26 & 55297.3919 & 0.0008 & @xmath170.0000 & 78 + 27 & 55297.4694 & 0.0005 & @xmath170.0009 & 78 + 38 & 55298.3319 & 0.0006 & @xmath170.0008 & 128 + 39 & 55298.4115 & 0.0003 & 0.0003 & 246 + + + + this object ( hereafter sdss j0903 ) is a cv selected during the course of the sdss @xcite , who suspected its eclipsing nature .
@xcite confirmed that this object is a deeply eclipsing cv with a short orbital period of 0.059073543(9 ) d. the 2010 outburst , the first - ever outburst reported in real - time , was detected by crts (= css100522:090351@xmath1330036 ; cf .
vsnet - alert 11994 ) .
subsequent observations have confirmed the eclipsing su uma - type nature of this object ( vsnet - alert 12006 ; figure [ fig : j0903shpdm ] ) .
the times of mid - eclipses were determined with the kw method ( table [ tab : j0903ecl ] ) .
we obtained an updated orbital ephemeris ( equation [ equ : j0903ecl ] ) using our data and times of eclipses in @xcite .
ccccc @xmath15 & minimum & error & @xmath0 & source + 0 & 53800.394700 & 0.000006 & -0.00001 & 1 + 2 & 53800.512854 & 0.000006 & -0.00000 & 1 + 34 & 53802.403198 & 0.000006 & -0.00001 & 1 + 35 & 53802.462279 & 0.000006 & -0.00000 & 1 + 36 & 53802.521361 & 0.000006 & 0.00001 & 1 + 37 & 53802.580442 & 0.000006 & 0.00001 & 1 + 38 & 53802.639501 & 0.000006 & -0.00000 & 1 + 50 & 53803.348386 & 0.000006 & 0.00000 & 1 + 51 & 53803.407462 & 0.000006 & 0.00000 & 1 + 52 & 53803.466529 & 0.000006 & -0.00000 & 1 + 53 & 53803.525601 & 0.000006 & -0.00000 & 1 + 26086 & 55341.38690 & 0.00030 & 0.00022 & 2 + 26087 & 55341.44648 & 0.00036 & 0.00073 & 2 + 26102 & 55342.33163 & 0.00038 & -0.00022 & 2 + 26104 & 55342.44915 & 0.00038 & -0.00085 & 2 + 26119 & 55343.33624 & 0.00039 & 0.00013 & 2 + 26136 & 55344.34126 & 0.00034 & 0.00091 & 2 + 26137 & 55344.39964 & 0.00039 & 0.00020 & 2 + 26152 & 55345.28500 & 0.00030 & -0.00053 & 2 + 26154 & 55345.40339 & 0.00046 & -0.00029 & 2 + 26169 & 55346.29034 & 0.00029 & 0.00056 & 2 + 26170 & 55346.34966 & 0.00032 & 0.00080 & 2 + 26171 & 55346.40821 & 0.00045 & 0.00028 & 2 + 26188 & 55347.41127 & 0.00056 & -0.00091 & 2 + 26193 & 55347.70712 & 0.00027 & -0.00043 & 2 + 26204 & 55348.35745 & 0.00037 & 0.00009 & 2 + 26205 & 55348.41652 & 0.00035 & 0.00009 & 2 + 26209 & 55348.65271 & 0.00030 & -0.00001 & 2 + 26210 & 55348.71183 & 0.00036 & 0.00003 & 2 + 26221 & 55349.36176 & 0.00037 & 0.00015 & 2 + 26222 & 55349.42067 & 0.00040 & -0.00001 & 2 + 26243 & 55350.66096 & 0.00034 & -0.00026 & 2 + 26244 & 55350.71996 & 0.00044 & -0.00034 & 2 + 26255 & 55351.36974 & 0.00028 & -0.00037 & 2 + 26256 & 55351.42824 & 0.00041 & -0.00095 & 2 + 26272 & 55352.37433 & 0.00072 &
-0.00003 & 2 + + + + @xmath73 in the following analysis , we removed observations within 0.08 @xmath2 of eclipses .
the times of superhump maxima are listed in table [ tab : j0903oc2010 ] .
there was a clear stage b
c transition around @xmath74 .
the @xmath7 for stage b was @xmath75 .
this clearly positive @xmath7 in deeply eclipsing short - period su uma - type dwarf novae confirmed the finding in xz eri ( @xcite ; @xcite ) .
the fractional superhump excesses for stages b and c were 2.1 % and 1.7 % , respectively .
other parameters are listed in table [ tab : perlist ] .
( 88mm,110mm)fig48.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55340.4222 & 0.0009 & 0.0072 & 51 + 31 & 55342.2882 & 0.0009 & 0.0033 & 26 + 32 & 55342.3424 & 0.0007 & @xmath170.0028 & 57 + 33 & 55342.4003 & 0.0019 & @xmath170.0052 & 31 + 48 & 55343.3142 & 0.0013 & 0.0039 & 19 + 49 & 55343.3624 & 0.0016 & @xmath170.0083 & 55 + 50 & 55343.4267 & 0.0017 & @xmath170.0043 & 54 + 65 & 55344.3326 & 0.0008 & @xmath170.0032 & 57 + 81 & 55345.2974 & 0.0008 & @xmath170.0035 & 57 + 82 & 55345.3554 & 0.0047 & @xmath170.0058 & 48 + 83 & 55345.4122 & 0.0040 & @xmath170.0093 & 51 + 98 & 55346.3301 & 0.0017 & 0.0038 & 83 + 99 & 55346.3923 & 0.0017 & 0.0056 & 47 + 114 & 55347.3027 & 0.0090 & 0.0112 & 74 + 116 & 55347.4195 & 0.0024 & 0.0075 & 51 + 121 & 55347.7186 & 0.0012 & 0.0050 & 52 + 132 & 55348.3800 & 0.0009 & 0.0028 & 54 + 133 & 55348.4419 & 0.0017 & 0.0044 & 29 + 137 & 55348.6809 & 0.0007 & 0.0021 & 62 + 149 & 55349.4016 & 0.0010 & @xmath170.0010 & 50 + 170 & 55350.6692 & 0.0019 & @xmath170.0002 & 49 + 182 & 55351.3798 & 0.0022 & @xmath170.0133 & 29
+ + + + this object ( hereafter sdss j1152 ) is a cv selected during the course of the sdss @xcite , who suspected its eclipsing nature .
@xcite established that this cv is deeply eclipsing and determined its period to be 0.06770(28 ) d. the 2009 june outburst was detected by e. muyllaert at a ccd magnitude of 16.4 on june 9 ( cvnet - outburst 3158 ) .
superhumps and eclipses were soon detected ( vsnet - alert 11288 ) , establishing the su uma - type nature of this object .
the times of superhump maxima determined observations outside the eclipses are listed in table [ tab : j1152oc2009 ] .
the intervening clouds made the epoch of @xmath15 = 39 rather uncertain .
a period analysis with the pdm method has yielded a @xmath3 of 0.0689(1 ) d ( figure [ fig : j1152shpdm ] ) .
since the signal is broad due to the poor phase coverage , we used a bayesian modeling of the light curve using the template superhump light curve ( see appendix ) .
the best period by this method is 0.06887(4 ) d , corresponding to @xmath13 of 1.7 % .
we adopted this period in the figure and table .
although the basic nature of the object is well - established , both orbital and superhump periods need to be refined by further observations .
( 88mm,110mm)fig49.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 54994.0621 & 0.0005 & 0.0056 & 125 + 39 & 54996.7376 & 0.0068 & @xmath170.0135 & 9 + 67 & 54998.6919 & 0.0021 & 0.0062 & 18 + 68 & 54998.7564 & 0.0021 & 0.0016 & 28
+ + + + this object ( hereafter j1250 ) is a cv selected during the course of the sdss @xcite , who suggested a high inclination cv .
@xcite confirmed that this is indeed a deeply eclipsing cv with a very short orbital period of 0.058735687(4 ) d. shortly before @xcite becomes available , the object was found to be in outburst in 2008 january ( s. brady , cvnet - discussion 1104 ) , who successfully detected eclipses and reported a period of 0.059 d. based on these and 2009 observations , we have determined eclipse times and updated the ephemeris .
both the 2008 and 2009 outbursts were superoutbursts , and the times of superhump maxima are listed in tables [ tab : j1250oc2008 ] and [ tab : j1250oc2009 ] determined from observations outside the eclipses . a pdm analysis of the 2008 observations has yielded a mean @xmath3 of 0.06032(5 ) d ( figure [ fig : j1250shpdm ] ) .
the @xmath7 for @xmath76 , apparently stage b , was @xmath77 .
the period apparently decreased after this , suggesting a transition to stage c. the @xmath13 of 2.7 % for the mean @xmath3 is relatively large for this @xmath2 . using the @xmath13@xmath14 relation in @xcite
, this @xmath13 corresponds to @xmath14 = 0.14 . assuming a moderate white - dwarf mass , this @xmath14 would place the object around the upper ( massive secondary ) boundary defined by sdss j0903 and sdss j1507 in figure 2 of @xcite rather than around the period bounce .
this identification appears to be favored by the relatively large @xmath7 and rather frequent outbursts ( the interval of two superoutburst was @xmath9 650 d ) .
further determination of system parameters of this object will improve our knowledge of cv evolution near the period minimum .
( 88mm,110mm)fig50.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 54491.9263 & 0.0010 & 0.0039 & 6 + 44 & 54494.5762 & 0.0018 & @xmath170.0002 & 18 + 45 & 54494.6331 & 0.0012 & @xmath170.0037 & 41 + 46 & 54494.6926 & 0.0014 & @xmath170.0044 & 39 + 47 & 54494.7553 & 0.0008 & @xmath170.0020 & 42 + 48 & 54494.8165 & 0.0007 & @xmath170.0011 & 41 + 77 & 54496.5690 & 0.0022 & 0.0021 & 17 + 78 & 54496.6287 & 0.0011 & 0.0015 & 24 + 79 & 54496.6906 & 0.0015 & 0.0031 & 24 + 80 & 54496.7482 & 0.0011 & 0.0003 & 22 + 81 & 54496.8082 & 0.0009 & @xmath170.0000 & 24 + 82 & 54496.8696 & 0.0013 & 0.0011 & 24 + 83 & 54496.9293 & 0.0010 & 0.0005 & 20 + 146 & 54500.7289 & 0.0033 & 0.0000 & 22 + 147 & 54500.7882 & 0.0029 & @xmath170.0010 & 19 + + + + ccccc @xmath15 & max & error & @xmath16 + 0 & 55144.2715 & 0.0015 & 77 + 1 & 55144.3338 & 0.0004 & 54
+ + + this object ( hereafter sdss j1502 ) is an eclipsing cv selected during the course of the sdss @xcite . @xcite
performed high - speed photometry in quiescence and obtained orbital parameters [ @xmath7 = 0.05890961(5 ) d and estimated @xmath14 = 0.109(3 ) ] .
@xcite attributed this cv to a cv before the `` period bounce '' .
the 2009 july superoutburst of this object was detected by j. shears ( baavss - alert 1997 ) .
subsequent observations detected superhumps ( cf .
vsnet - alert 11332 ; figure [ fig : j1502shpdm ] ) .
@xcite also reported an analysis of the same superoutburst using the slightly different data set from ours .
the times of mid - eclipses were determined with the kw method ( table [ tab : j1502ecl ] ) .
we obtained an updated orbital ephemeris ( equation [ equ : j1502ecl ] ) using our data and times of eclipses in @xcite and @xcite .
we only used times of eclipses in @xcite which were not covered by our observations .
we also disregarded @xmath78 eclipse due to its large @xmath0 .
ccccc @xmath15 & minimum & error & @xmath0 & source + 0 & 53799.640618 & 0.000004 & 0.00013 & 1 + 2 & 53799.758414 & 0.000007 & 0.00011 & 1 + 17 & 53800.642070 & 0.000006 & 0.00012 & 1 + 18 & 53800.700966 & 0.000006 & 0.00011 & 1 + 19 & 53800.759901 & 0.000006 & 0.00014 & 1 + 52 & 53802.703911 & 0.000002 & 0.00013 & 1 + 68 & 53803.646461 & 0.000003 & 0.00013 & 1 + 69 & 53803.705371 & 0.000006 & 0.00013 & 1 + 70 & 53803.764277 & 0.000003 & 0.00013 & 1 + 854 & 53849.94908 & 0.00035 & -0.00009 & 2 + 866 & 53850.65615 & 0.00012 & 0.00006 & 2 + 867 & 53850.71498 & 0.00011 & -0.00002 & 2 + 868 & 53850.77384 & 0.00021 & -0.00007 & 2 + 869 & 53850.83284 & 0.00012 & 0.00002 & 2 + 870 & 53850.89165 & 0.00014 & -0.00008 & 2 + 871 & 53850.95066 & 0.00013 & 0.00003 & 2 + 883 & 53851.65760 & 0.00014 & 0.00005 & 2 + 884 & 53851.71630 & 0.00014 & -0.00016 & 2 + 886 & 53851.83418 & 0.00016 & -0.00010 & 2 + 887 & 53851.89334 & 0.00014 & 0.00015 & 2 + 888 & 53851.95250 & 0.00014 & 0.00040 & 2 + 900 & 53852.65899 & 0.00018 & -0.00002 & 2 + 901 & 53852.71800 & 0.00015 & 0.00008 & 2 + 902 & 53852.77661 & 0.00011 & -0.00022 & 2 + 903 & 53852.83568 & 0.00014 & -0.00006 & 2 + 1308 & 53876.69411 & 0.00004 & 0.00004 & 2 + 6454 & 54179.84201 & 0.00009 & -0.00021 & 2 + 6455 & 54179.90009 & 0.00009 & -0.00104 & 2 + 6457 & 54180.01899 & 0.00011 & 0.00004 & 2 + 13787 & 54611.82514 & 0.00012 & -0.00025 & 2 + 13788 & 54611.88411 & 0.00012 & -0.00019 & 2 + 13789 & 54611.94325 & 0.00017 & 0.00004 & 2 + 13802 & 54612.70876 & 0.00019 & -0.00027 & 2 + 13803 & 54612.76795 & 0.00012 & 0.00001 & 2 + 13804 & 54612.82684 & 0.00018 & -0.00001 & 2 + 13805 & 54612.88550 & 0.00012 & -0.00026 & 2 + 13806 & 54612.94456 & 0.00027 & -0.00011 & 2 + 20758 & 55022.48413 & 0.00005 & 0.00081 & 2 + 20760 & 55022.60121 & 0.00032 & 0.00007 & 3 + 20761 & 55022.66052 & 0.00036 & 0.00047 & 3 + 20762 & 55022.71932 & 0.00026 & 0.00036 & 3 + 20763 & 55022.77857 & 0.00031 & 0.00070 & 2 + 20777 & 55023.60263 & 0.00029 & 0.00003 & 3 + 20778 & 55023.66133 & 0.00028 & -0.00018 & 2 + 20779 & 55023.72066 & 0.00021 & 0.00024 & 3 + 20780 & 55023.77951 & 0.00018 & 0.00018 & 3 + 20781 & 55023.83814 & 0.00024 & -0.00010 & 3 + 20782 & 55023.89720 & 0.00040 & 0.00005 & 3 + 20794 & 55024.60306 & 0.00050 & -0.00101 & 2 + 20796 & 55024.72187 & 0.00026 & -0.00001 & 3 + 20797 & 55024.78077 & 0.00022 & -0.00002 & 3 + 20798 & 55024.83956 & 0.00028 & -0.00014 & 3 + 20807 & 55025.37023 & 0.00027 & 0.00034 & 3 + + + + + ccccc @xmath15 & minimum & error & @xmath0 & source + 20808 & 55025.42910 & 0.00030 & 0.00030 & 3 + 20811 & 55025.60573 & 0.00042 & 0.00020 & 3 + 20812 & 55025.66472 & 0.00014 & 0.00028 & 3 + 20813 & 55025.72362 & 0.00016 & 0.00027 & 3 + 20814 & 55025.78256 & 0.00017 & 0.00030 & 3 + 20815 & 55025.84150 & 0.00023 & 0.00034 & 3 + 20824 & 55026.37130 & 0.00026 & -0.00005 & 3 + 20825 & 55026.43013 & 0.00033 & -0.00013 & 3 + 20828 & 55026.60726 & 0.00011 & 0.00027 & 2 + 20829 & 55026.66588 & 0.00029 & -0.00002 & 3 + 20830 & 55026.72440 & 0.00022 & -0.00041 & 3 + 20831 & 55026.78290 & 0.00033 & -0.00082 & 3 + 20832 & 55026.84210 & 0.00100 & -0.00053 & 3 + 20835 & 55027.01943 & 0.00032 & 0.00008 & 3 + 20836 & 55027.07836 & 0.00028 & 0.00010 & 3 + 20837 & 55027.13703 & 0.00053 & -0.00014 & 3 + 20841 & 55027.37299 & 0.00026 & 0.00018 & 3 + 20845 & 55027.60852 & 0.00038 & 0.00007 & 3 + 20846 & 55027.66772 & 0.00040 & 0.00036 & 3 + 20847 & 55027.72659 & 0.00021 & 0.00032 & 3 + 20848 & 55027.78561 & 0.00022 & 0.00043 & 2 + 20849 & 55027.84470 & 0.00051 & 0.00061 & 3 + 20858 & 55028.37421 & 0.00027 & -0.00006 & 3 + 20859 & 55028.43318 & 0.00019 & -0.00000 & 3 + 20860 & 55028.49206 & 0.00023 & -0.00003 & 3 + 20861 & 55028.55083 & 0.00026 & -0.00017 & 3 + 20864 & 55028.72779 & 0.00037 & 0.00006 & 3 + 20881 & 55029.72935 & 0.00035 & 0.00016 & 3 + 20882 & 55029.78814 & 0.00036 & 0.00004 & 3 + 20883 & 55029.84720 & 0.00033 & 0.00019 & 3 + 20894 & 55030.49541 & 0.00105 & 0.00040 & 2 + 20895 & 55030.55404 & 0.00024 & 0.00012 & 2 + 20898 & 55030.73046 & 0.00014 & -0.00019 & 2 + 20898 & 55030.73050 & 0.00033 & -0.00015 & 3 + 20899 & 55030.78951 & 0.00034 & -0.00005 & 3 + 20900 & 55030.84853 & 0.00035 & 0.00006 & 3 + 20915 & 55031.73159 & 0.00076 & -0.00052 & 3 + 20916 & 55031.79037 & 0.00043 & -0.00065 & 2 + 20917 & 55031.84916 & 0.00138 & -0.00077 & 3 + 20928 & 55032.49776 & 0.00068 & -0.00017 & 3 + 20929 & 55032.55681 & 0.00079 & -0.00003 & 3 + 20932 & 55032.73351 & 0.00064 & -0.00006 & 3 + 20934 & 55032.85137 & 0.00065 & -0.00002 & 3 + 20949 & 55033.73470 & 0.00075 & -0.00033 & 3 + 20950 & 55033.79383 & 0.00077 & -0.00011 & 3 + 20966 & 55034.73608 & 0.00118 & -0.00041 & 2 + 20968 & 55034.85449 & 0.00133 & 0.00018 & 3 + 21031 & 55038.56538 & 0.00065 & -0.00023 & 3 + 21758 & 55081.39289 & 0.00168 & 0.00009 & 2 + + + + + @xmath79 in the following analysis , we removed observations within 0.08 @xmath2 of eclipses .
the times of superhump maxima are listed in table [ tab : j1502oc2009 ] .
there was a likely stage b
c transition around @xmath80 .
the values given in table [ tab : perlist ] were estimated following this interpretation .
the @xmath7 for stage b was @xmath81 , which may have been underestimated because the earliest part of the outburst was not sufficiently observed .
although @xcite detected a pattern of @xmath0 variation basically to ours , we used our values based on larger set of data .
was reported to be @xmath82 , our analysis of their timing data for @xmath83 yielded a @xmath7 of @xmath84 by our definition . ] the maxima for @xmath85 may not be superhumps , probably strongly affected by orbital modulations , and are excluded from this period analysis .
( 88mm,110mm)fig51.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55022.6300 & 0.0004 & 0.0039 & 92 + 1 & 55022.6895 & 0.0003 & 0.0029 & 78 + 16 & 55023.5921 & 0.0004 & @xmath170.0009 & 69 + 17 & 55023.6541 & 0.0008 & 0.0007 & 9 + 18 & 55023.7122 & 0.0004 & @xmath170.0016 & 145 + 19 & 55023.7736 & 0.0003 & @xmath170.0007 & 179 + 20 & 55023.8343 & 0.0004 & @xmath170.0004 & 147 + 21 & 55023.8946 & 0.0012 & @xmath170.0005 & 74 + 34 & 55024.6790 & 0.0005 & @xmath170.0017 & 73 + 35 & 55024.7401 & 0.0002 & @xmath170.0011 & 95 + 36 & 55024.8015 & 0.0003 & @xmath170.0002 & 95 + 46 & 55025.4060 & 0.0003 & 0.0000 & 91 + 47 & 55025.4652 & 0.0003 & @xmath170.0012 & 91 + 49 & 55025.5904 & 0.0004 & 0.0031 & 29 + 50 & 55025.6449 & 0.0002 & @xmath170.0028 & 154 + 51 & 55025.7061 & 0.0001 & @xmath170.0020 & 279 + 52 & 55025.7648 & 0.0001 & @xmath170.0038 & 230 + 53 & 55025.8260 & 0.0003 & @xmath170.0029 & 187 + 61 & 55026.3178 & 0.0011 & 0.0054 & 70 + 62 & 55026.3694 & 0.0005 & @xmath170.0035 & 110 + 63 & 55026.4387 & 0.0005 & 0.0054 & 108 + 66 & 55026.6156 & 0.0014 & 0.0010 & 24 + 67 & 55026.6726 & 0.0004 & @xmath170.0024 & 127 + 68 & 55026.7367 & 0.0005 & 0.0012 & 115 + 69 & 55026.7966 & 0.0006 & 0.0007 & 111 + 73 & 55027.0410 & 0.0006 & 0.0033 & 91 + 74 & 55027.1001 & 0.0003 & 0.0020 & 109 + 79 & 55027.4019 & 0.0003 & 0.0017 & 94 + 82 & 55027.5840 & 0.0005 & 0.0025 & 52 + 83 & 55027.6446 & 0.0005 & 0.0027 & 71 + 84 & 55027.7070 & 0.0005 & 0.0046 & 75 + 86 & 55027.8266 & 0.0003 & 0.0033 & 75 + 89 & 55028.0090 & 0.0006 & 0.0045 & 54 + 90 & 55028.0678 & 0.0003 & 0.0028 & 131 + 91 & 55028.1284 & 0.0007 & 0.0030 & 85 + 95 & 55028.3691 & 0.0003 & 0.0020 & 105 + 96 & 55028.4293 & 0.0003 & 0.0017 & 135 + 97 & 55028.4904 & 0.0004 & 0.0024 & 157 + 98 & 55028.5513 & 0.0003 & 0.0029 & 101 + 101 & 55028.7293 & 0.0006 & @xmath170.0005 & 59 + 118 & 55029.7516 & 0.0007 & @xmath170.0055 & 88 + 119 & 55029.8127 & 0.0006 & @xmath170.0048 & 92 + 120 & 55029.8733 & 0.0006 & @xmath170.0046 & 81 + 134 & 55030.7147 & 0.0005 & @xmath170.0093 & 76 + 135 & 55030.7763 & 0.0005 & @xmath170.0081 & 87 + 136 & 55030.8392 & 0.0007 & @xmath170.0057 & 88 + 152 & 55031.7872 & 0.0023 & @xmath170.0246 & 34 + 153 & 55031.8347 & 0.0023 & @xmath170.0375 & 34 + 163 & 55032.4860 & 0.0016 & 0.0095 & 47 + 164 & 55032.5491 & 0.0017 & 0.0121 & 46 + 167 & 55032.7254 & 0.0010 & 0.0072 & 48 + 168 & 55032.7880 & 0.0008 & 0.0093 & 46 + 169 & 55032.8425 & 0.0011 & 0.0034 & 49 + 183 & 55033.6874 & 0.0037 & 0.0023 & 31 + 184 & 55033.7398 & 0.0070 & @xmath170.0058 & 48 + 185 & 55033.7996 & 0.0020 & @xmath170.0065 & 49 + 186 & 55033.8721 & 0.0038 & 0.0056 & 27 + 200 & 55034.7195 & 0.0013 & 0.0070 & 47 + 201 & 55034.7818 & 0.0016 & 0.0088 & 49 + 202 & 55034.8427 & 0.0026 & 0.0093 & 49
+ + + + this object ( hereafter sdss j1610 ) was initially selected as a dwarf nova by @xcite .
the 2009 july outburst was detected by the crts (= css090727:161028 + 090739 ) .
double - peaked strong emission lines , characteristic to a relatively high inclination dwarf nova , in sdss spectrum was announced ( vsnet - alert 11350 ) . a remarkable growth of superhumps was observed four days after the outburst detection ( vsnet - alert 11366 ) , suggesting a substantial delay in the growth of superhumps .
further observations clarified the potential wz sge - type characters ( vsnet - alert 11367 , 11368 , 11381 ; figure [ fig : j1610shpdm ] .
the times of ordinary superhumps are listed in table [ tab : j1610oc2009 ] .
there was a clear stage a b transition around @xmath86 .
the @xmath7 for the well - observed segment of stage b was @xmath87 ( @xmath88 ) .
although an extrapolation to @xmath89 has yielded a @xmath7 of @xmath90 , the identification of superhumps was slightly ambiguous due to the faintness .
we thus adopted the former value for the @xmath7 of this object .
an analysis of the light curve before the growth of ordinary superhumps detected early superhumps ( figure [ fig : j1610eshpdm ] ) with a period of 0.05687(1 ) d ( pdm and bayesian methods ) , confirming both the wz sge - type nature and a relatively high inclination .
the resultant @xmath13 of 1.6 % is relatively large among wz sge - type dwarf novae ( cf .
@xcite ) , consistent with a relatively large @xmath7 .
( 88mm,110mm)fig52.eps ( 88mm,110mm)fig53.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55042.4471 & 0.0006 & @xmath170.0220 & 120 + 10 & 55043.0435 & 0.0016 & @xmath170.0053 & 99 + 11 & 55043.1014 & 0.0011 & @xmath170.0053 & 184 + 16 & 55043.3959 & 0.0012 & @xmath170.0007 & 29 + 17 & 55043.4515 & 0.0016 & @xmath170.0031 & 28 + 26 & 55043.9811 & 0.0003 & 0.0048 & 61 + 27 & 55044.0393 & 0.0002 & 0.0049 & 60 + 28 & 55044.0986 & 0.0004 & 0.0063 & 60 + 33 & 55044.3907 & 0.0003 & 0.0085 & 23 + 34 & 55044.4468 & 0.0006 & 0.0066 & 26 + 35 & 55044.5061 & 0.0013 & 0.0080 & 17 + 43 & 55044.9678 & 0.0004 & 0.0059 & 58 + 44 & 55045.0245 & 0.0006 & 0.0045 & 56 + 45 & 55045.0839 & 0.0004 & 0.0060 & 46 + 51 & 55045.4319 & 0.0005 & 0.0061 & 114 + 52 & 55045.4864 & 0.0005 & 0.0026 & 76 + 61 & 55046.0067 & 0.0006 & 0.0012 & 49 + 62 & 55046.0641 & 0.0011 & 0.0006 & 23 + 68 & 55046.4086 & 0.0009 & @xmath170.0028 & 29 + 68 & 55046.4086 & 0.0010 & @xmath170.0027 & 30 + 78 & 55046.9894 & 0.0012 & @xmath170.0017 & 181 + 79 & 55047.0467 & 0.0012 & @xmath170.0023 & 167 + 80 & 55047.1042 & 0.0008 & @xmath170.0029 & 60 + 85 & 55047.3919 & 0.0008 & @xmath170.0050 & 31 + 86 & 55047.4519 & 0.0009 & @xmath170.0029 & 27 + 102 & 55048.3792 & 0.0009 & @xmath170.0033 & 22 + 103 & 55048.4370 & 0.0010 & @xmath170.0034 & 28 + 137 & 55050.4060 & 0.0028 & @xmath170.0055 & 28 + 268 & 55058.0072 & 0.0025 & 0.0009 & 121 + 269 & 55058.0659 & 0.0035 & 0.0017 & 123 + + + + this object was selected as a cv candidate by @xcite .
although the object was not recognized as a cv during the course of the sdss , we employed the sdss designation in this paper ( hereafter sdss j1625 ) .
the 2010 outburst of this object was detected by the crts (= css100705:162520@xmath1120309 ) , a discussion on the sdss spectrum and the earlier discovery of the object by @xcite can be found in vsnet - alert 12052 , 12053 .
short - term variations were detected soon after this outburst detection ( vsnet - alert 12054 , 12059 ) , which later turned out to be developing superhumps ( vsnet - alert 12061 , 12062 ) .
fully developed superhumps and the course of period evolution were subsequently observed ( vsnet - alert 12064 , 12065 , 12066 , 12068 , 12071 ; figure [ fig : j1625shpdm ] ) .
the object entered the rapid decline phase 45 d after the development of superhumps , which was unexpectedly early ( vsnet - alert 12079 ) .
the object was also unusual in its rebrightening phenomenon soon after the rapid decline ( vsnet - alert 12087 ) .
the times of superhump maxima are listed in table [ tab : j1625oc2010 ] .
the epochs for @xmath91 correspond to the growing stage of superhumps ( stage a ) .
the epochs for @xmath92 were maxima after the rebrightening and these humps may not be true superhumps .
we also excluded @xmath93 ( just prior to the rebrightening ) in determining period variation due to its low signal .
the @xmath7 for @xmath94 was @xmath95 , which is unusually large for this long @xmath3 = 0.09605(5 ) d ( mean period based on timing analysis ) .
there was also little indication of a stage b
c transition during the superoutburst plateau .
these unusual evolution may be related to the unexpectedly early fading during the superoutburst plateau ( see figure [ fig : j1625oc2010 ] ) .
this object appears to add another variety of superhump evolution in long-@xmath3 systems ( cf .
@xcite , subsection 4.10 ) .
( 88mm,110mm)fig54.eps ( 88mm,90mm)fig55.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55384.4489 & 0.0003 & @xmath170.0363 & 363 + 2 & 55384.6553 & 0.0002 & @xmath170.0224 & 176 + 6 & 55385.0497 & 0.0003 & @xmath170.0128 & 161 + 10 & 55385.4459 & 0.0002 & @xmath170.0015 & 587 + 11 & 55385.5446 & 0.0002 & 0.0010 & 365 + 12 & 55385.6431 & 0.0001 & 0.0032 & 179 + 20 & 55386.4255 & 0.0001 & 0.0159 & 732 + 21 & 55386.5198 & 0.0001 & 0.0140 & 662 + 30 & 55387.3886 & 0.0003 & 0.0169 & 265 + 31 & 55387.4833 & 0.0003 & 0.0153 & 289 + 41 & 55388.4373 & 0.0002 & 0.0071 & 328 + 42 & 55388.5327 & 0.0002 & 0.0064 & 291 + 43 & 55388.6296 & 0.0002 & 0.0070 & 194 + 44 & 55388.7250 & 0.0002 & 0.0062 & 158 + 51 & 55389.3941 & 0.0003 & 0.0018 & 382 + 52 & 55389.4895 & 0.0003 & 0.0010 & 430 + 53 & 55389.5855 & 0.0002 & 0.0007 & 189 + 54 & 55389.6808 & 0.0003 & @xmath170.0001 & 200 + 61 & 55390.3514 & 0.0027 & @xmath170.0031 & 135 + 62 & 55390.4523 & 0.0004 & 0.0016 & 338 + 63 & 55390.5440 & 0.0006 & @xmath170.0029 & 144 + 72 & 55391.4104 & 0.0005 & @xmath170.0025 & 193 + 73 & 55391.5071 & 0.0007 & @xmath170.0021 & 170 + 82 & 55392.3723 & 0.0010 & @xmath170.0027 & 102 + 83 & 55392.4739 & 0.0008 & 0.0026 & 130 + 84 & 55392.5750 & 0.0009 & 0.0075 & 144 + 85 & 55392.6648 & 0.0008 & 0.0010 & 190 + 93 & 55393.4316 & 0.0034 & @xmath170.0019 & 153 + 94 & 55393.5346 & 0.0025 & 0.0049 & 52 + 104 & 55394.4882 & 0.0013 & @xmath170.0037 & 43 + 114 & 55395.4413 & 0.0026 & @xmath170.0127 & 49 + 130 & 55397.0091 & 0.0028 & 0.0156 & 200 + 134 & 55397.3893 & 0.0035 & 0.0108 & 94 + 135 & 55397.4389 & 0.0042 & @xmath170.0358 & 149
+ + + + this object ( hereafter sdss j1637 ) is a cv selected during the course of the sdss @xcite .
@xcite reported the presence of high and low states and suggested the dwarf nova - type classification .
the 2004 outburst of this object was detected by r. stubbings at a visual magnitude of 15.0 on 2004 march 27 ( vsnet - outburst 6212 ) .
the presence of superhumps was soon confirmed ( vsnet - alert 8084 , 8086 , 8088 ) . the mean superhump period determined from the observation was 0.06910(4 ) d ( pdm method , figure [ fig : j1637shpdm ] ) , and the times of superhump maxima are listed in table [ tab : j1637oc2004 ] . since the outburst entered its rapid decline stage @xmath9 4 d after the initial detection , these superhumps were very likely stage c superhumps .
using the orbital period of 0.06739(1 ) d determined by @xcite , we obtained an @xmath13 of 2.5 % .
( 88mm,110mm)fig56.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 53093.2412 & 0.0016 & @xmath170.0023 & 145 + 1 & 53093.3132 & 0.0012 & 0.0004 & 76 + 2 & 53093.3832 & 0.0012 & 0.0014 & 37 + 4 & 53093.5206 & 0.0015 & 0.0004 & 18 + 5 & 53093.5896 & 0.0008 & 0.0002 & 155 + 29 & 53095.2464 & 0.0027 & @xmath170.0030 & 36 + 30 & 53095.3206 & 0.0025 & 0.0020 & 38 + 31 & 53095.3890 & 0.0022 & 0.0013 & 30 + 43 & 53096.2189 & 0.0034 & 0.0012 & 87 + 44 & 53096.2853 & 0.0028 & @xmath170.0016 & 147
+ + + + this object ( hereafter sdss j1653 ) is a cv selected during the course of the sdss @xcite , who detected superhumps with a period of 1.58 hr during one of its superoutburst .
the 2010 superoutburst was detected by the crts ( cf .
vsnet - alert 11936 ) .
the times of superhump maxima are listed in table [ tab : j1653oc2010 ] .
it is evident from these data that these observations recorded a stage b c transition .
the derived periods for each stage are listed in table [ tab : perlist ] .
although the observation only covered the late part of the superoutburst , the evolution of superhump period appears to be typical .
as judged from these periods , @xcite appears to have recorded stage b superhumps .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55304.1037 & 0.0009 & @xmath170.0007 & 67 + 1 & 55304.1673 & 0.0007 & @xmath170.0022 & 59 + 2 & 55304.2313 & 0.0004 & @xmath170.0032 & 66 + 3 & 55304.2970 & 0.0005 & @xmath170.0025 & 66 + 5 & 55304.4272 & 0.0003 & @xmath170.0024 & 73 + 6 & 55304.4913 & 0.0004 & @xmath170.0033 & 67 + 7 & 55304.5579 & 0.0004 & @xmath170.0018 & 72 + 50 & 55307.3670 & 0.0017 & 0.0110 & 55 + 51 & 55307.4263 & 0.0014 & 0.0052 & 65 + 52 & 55307.4906 & 0.0010 & 0.0046 & 69 + 93 & 55310.1499 & 0.0009 & @xmath170.0025 & 45 + 94 & 55310.2197 & 0.0015 & 0.0023 & 40 + 215 & 55318.0817 & 0.0041 & @xmath170.0045 & 22
+ + + + this object ( hereafter sdss j2048 ) is a cv selected during the course of the sdss @xcite , who reported the detection of a white dwarf in the spectrum , suggesting a dwarf nova with a low mass - transfer rate . the outburst in 2009 october was detected by e. muyllaert ( cvnet - outburst 3367 ) .
subsequent observations by i. miller detected superhumps ( cvnet - outburst 3383 ) .
@xcite observed the object in quiescence and obtained an orbital period of 0.060597(2 ) d. we identified the superhump period of 0.06166(2 ) d based om this @xmath2 ( figure [ fig : j2048shpdm ] .
the times of superhump maxima are listed in table [ tab : j2048oc2009 ] .
the @xmath13 was 1.8 % .
( 88mm,110mm)fig57.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55119.4241 & 0.0008 & @xmath170.0013 & 45 + 16 & 55120.4152 & 0.0012 & 0.0029 & 32 + 31 & 55121.3347 & 0.0007 & @xmath170.0029 & 63 + 32 & 55121.4006 & 0.0011 & 0.0014 & 56
+ + + + this object ( hereafter ot j0406 ) was discovered by k. itagaki in 2008 @xcite .
the 2010 outburst , detected by the crts , is the second known superoutburst of this object .
the outburst was apparently detected in its late stage .
the times of superhump maxima are listed in table [ tab : j0406oc2010 ] .
we attribute these superhumps to stage c superhumps .
the mean period with the pdm method was 0.07996(3 ) d , close to that ( 0.07992 d ) recorded during the 2008 superoutburst @xcite .
we adopted this period in table [ tab : perlist ] .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55246.9899 & 0.0009 & @xmath170.0011 & 126 + 12 & 55247.9497 & 0.0010 & @xmath170.0000 & 132 + 25 & 55248.9899 & 0.0011 & 0.0015 & 205 + 49 & 55250.9069 & 0.0030 & 0.0010 & 156 + 62 & 55251.9432 & 0.0027 & @xmath170.0014 & 207 + + + + this object ( = [email protected] 1257@xmath170089884 , hereafter ot j0506 ) is a dwarf nova discovered by @xcite ( see also vsnet - alert 11686 for the initial announcement ) .
the object was rising at the time of the discovery and superhumps were subsequently detected ( vsnet - alert 11688 ; @xcite ) . since the discovery announcement was made sufficiently early ,
the early stages of the outburst was well observed .
the times of superhump maxima are listed in table [ tab : j0506oc2009 ] .
the mean superhump period excluding the initial night was 0.06928(2 ) d ( pdm method , figure [ fig : j0506shpdm ] ) .
( 88mm,110mm)fig58.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55160.5113 & 0.0034 & @xmath170.0010 & 24 + 1 & 55160.5777 & 0.0008 & @xmath170.0039 & 48 + 2 & 55160.6485 & 0.0007 & @xmath170.0024 & 48 + 23 & 55162.1113 & 0.0006 & 0.0045 & 130 + 24 & 55162.1810 & 0.0007 & 0.0049 & 149 + 52 & 55164.1160 & 0.0010 & @xmath170.0012 & 102 + 53 & 55164.1854 & 0.0011 & @xmath170.0011 & 140 + 56 & 55164.3920 & 0.0009 & @xmath170.0025 & 12 + 81 & 55166.1293 & 0.0023 & 0.0016 & 146 + 82 & 55166.2028 & 0.0099 & 0.0058 & 92 + 85 & 55166.4085 & 0.0029 & 0.0035 & 7 + 88 & 55166.6129 & 0.0018 & @xmath170.0001 & 13 + 96 & 55167.1688 & 0.0020 & 0.0013 & 157 + 97 & 55167.2387 & 0.0020 & 0.0018 & 14 + 98 & 55167.3056 & 0.0017 & @xmath170.0006 & 12 + 99 & 55167.3755 & 0.0011 & @xmath170.0000 & 14 + 100 & 55167.4428 & 0.0015 & @xmath170.0021 & 12 + 101 & 55167.5116 & 0.0016 & @xmath170.0026 & 13 + 102 & 55167.5774 & 0.0022 & @xmath170.0061 & 12
+ + + + this object ( hereafter ot j1026 ) is a dwarf nova discovered by k. itagaki @xcite .
although @xcite reported on late - stage observations of the 2009 superoutburst , the observational coverage was short .
we observed the 2010 superoutburst , detected by i. miller ( baavss - alert 2245 ) .
we first time succeeded in recording both stages b and c ( table [ tab : j1026oc2010 ] ) .
the resultant period is in disagreement with the 2009 result .
since the present coverage is much better than the 2009 observation , the present values are more reliable .
a reanalysis of the 2009 data could not yield a continuous @xmath0 diagram , suggesting a discontinuous phase jump ( or appearance of stronger secondary superhump peaks ) in the final stage of the 2009 superoutburst .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55270.5082 & 0.0002 & @xmath170.0012 & 67 + 1 & 55270.5783 & 0.0003 & 0.0002 & 75 + 2 & 55270.6456 & 0.0003 & @xmath170.0011 & 75 + 26 & 55272.2894 & 0.0009 & @xmath170.0048 & 30 + 27 & 55272.3616 & 0.0005 & @xmath170.0013 & 76 + 28 & 55272.4296 & 0.0005 & @xmath170.0020 & 58 + 29 & 55272.5006 & 0.0006 & 0.0004 & 68 + 43 & 55273.4636 & 0.0004 & 0.0024 & 60 + 44 & 55273.5311 & 0.0008 & 0.0012 & 36 + 57 & 55274.4268 & 0.0039 & 0.0045 & 47 + 58 & 55274.4972 & 0.0021 & 0.0063 & 67 + 59 & 55274.5603 & 0.0012 & 0.0007 & 72 + 113 & 55278.2660 & 0.0013 & @xmath170.0004 & 53 + 114 & 55278.3322 & 0.0012 & @xmath170.0029 & 76 + 115 & 55278.4048 & 0.0011 & 0.0010 & 69 + 116 & 55278.4718 & 0.0008 & @xmath170.0006 & 76 + 117 & 55278.5393 & 0.0008 & @xmath170.0018 & 61 + 118 & 55278.6091 & 0.0007 & @xmath170.0006 & 72 + + + + this object (= css100217:104411@xmath1211307 , hereafter ot j1044 ) was discovered by the crts on 2010 february 17 .
the large outburst amplitude was already suggestive of a wz sge - type like outburst ( vsnet - alert 11817 ) .
subsequent observation soon established the presence of early superhumps ( vsnet - alert 11820 , 11828 , 11829 ; figure [ fig : j1044shpdm ] ) .
the object later developed ordinary superhumps ( vsnet - alert 11836 , 11837 ) .
there was a single post - superoutburst rebrightening ( vsnet - alert 11885 ) .
the times of ordinary superhumps are listed in table [ tab : j1044oc2010 ] .
it was most likely that we only sufficiently observed stage a ( @xmath70 ) and stage c ( @xmath96 ) .
although the @xmath0 analysis suggests a mean period of @xmath90.0605 d during stage b , we did not adopt this value in table [ tab : perlist ] . the superhump period during stage c
was , however , reliably determined ( figure [ fig : j1044shpdm ] ) .
the period [ 0.06024(4 ) d ] suggests an @xmath13 of 1.9 % against the period [ 0.05909 ( 1 ) d ] of early superhumps .
the waveform of superhumps became double - humped during the rapid fading stage ( @xmath97 ) , we only listed maxima having the same phases as in @xmath98 .
the last two points ( @xmath99 , @xmath100 ) were obtained during the fading stage of a rebrightening .
although these humps may have been different from earlier superhumps , we included them because wz sge - type dwarf novae are known to exhibit long - enduring superhumps even during the rebrightening phase .
all the observed features , including the outburst amplitude of @xmath96 mag , the existence of early superhumps , relatively large @xmath13 , the likely existence of stage c and the single post - superoutburst rebrightening resemble those of large - amplitude borderline wz sge - type dwarf novae such as bc uma and rz leo ( objects showing `` type - c '' wz sge - type outbursts in @xcite ) .
( 88mm,110mm)fig59.eps ( 88mm,110mm)fig60.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55250.5223 & 0.0010 & @xmath170.0081 & 65 + 1 & 55250.5847 & 0.0008 & @xmath170.0062 & 49 + 2 & 55250.6440 & 0.0013 & @xmath170.0074 & 25 + 8 & 55251.0082 & 0.0030 & @xmath170.0065 & 60 + 9 & 55251.0696 & 0.0004 & @xmath170.0056 & 95 + 10 & 55251.1309 & 0.0003 & @xmath170.0049 & 228 + 11 & 55251.1920 & 0.0002 & @xmath170.0044 & 229 + 24 & 55251.9822 & 0.0004 & @xmath170.0011 & 101 + 25 & 55252.0459 & 0.0009 & 0.0019 & 100 + 26 & 55252.1050 & 0.0003 & 0.0005 & 312 + 27 & 55252.1643 & 0.0002 & @xmath170.0008 & 302 + 28 & 55252.2251 & 0.0004 & @xmath170.0004 & 218 + 41 & 55253.0074 & 0.0019 & @xmath170.0052 & 61 + 92 & 55256.1063 & 0.0017 & 0.0060 & 105 + 113 & 55257.3823 & 0.0041 & 0.0106 & 32 + 114 & 55257.4322 & 0.0019 & @xmath170.0000 & 61 + 115 & 55257.4891 & 0.0011 & @xmath170.0037 & 67 + 116 & 55257.5505 & 0.0013 & @xmath170.0028 & 65 + 195 & 55262.3421 & 0.0014 & 0.0059 & 21 + 196 & 55262.4117 & 0.0011 & 0.0149 & 87 + 197 & 55262.4689 & 0.0009 & 0.0116 & 85 + 198 & 55262.5285 & 0.0008 & 0.0106 & 89 + 199 & 55262.5863 & 0.0010 & 0.0079 & 96 + 200 & 55262.6474 & 0.0011 & 0.0084 & 76 + 212 & 55263.3763 & 0.0028 & 0.0109 & 30 + 213 & 55263.4288 & 0.0010 & 0.0028 & 92 + 214 & 55263.4911 & 0.0010 & 0.0046 & 93 + 215 & 55263.5524 & 0.0017 & 0.0053 & 94 + 216 & 55263.6158 & 0.0012 & 0.0082 & 97 + 217 & 55263.6732 & 0.0017 & 0.0051 & 49 + 227 & 55264.2828 & 0.0014 & 0.0093 & 62 + 228 & 55264.3384 & 0.0012 & 0.0043 & 76 + 229 & 55264.3995 & 0.0016 & 0.0048 & 86 + 230 & 55264.4557 & 0.0015 & 0.0005 & 96 + 231 & 55264.5205 & 0.0016 & 0.0047 & 82 + 232 & 55264.5774 & 0.0011 & 0.0011 & 93 + 233 & 55264.6271 & 0.0026 & @xmath170.0097 & 86 + 244 & 55265.2995 & 0.0022 & @xmath170.0033 & 78 + 245 & 55265.3571 & 0.0012 & @xmath170.0062 & 91 + 246 & 55265.4192 & 0.0027 & @xmath170.0047 & 97 + 247 & 55265.4772 & 0.0013 & @xmath170.0072 & 92 + 248 & 55265.5405 & 0.0026 & @xmath170.0045 & 100 + 249 & 55265.6008 & 0.0043 & @xmath170.0047 & 85 + 376 & 55273.2741 & 0.0018 & @xmath170.0204 & 63 + 377 & 55273.3328 & 0.0023 & @xmath170.0222 & 63 + + + + this object (= css100603:112253@xmath17111037 , hereafter ot j1122 ) is a transient discovered by the crts .
the object was soon suspected to be a large - amplitude dwarf nova ( vsnet - alert 12020 ) .
h. maehara detected short - period superhumps ( vsnet - alert 12025 ; figure [ fig : j1122shpdm ] ) indicating that the object is an unusual short-@xmath2 cv with an evolved secondary . j.
greaves pointed out that this object is included in the public sdss dr7 archive .
the spectrum indicates that the object is hydrogen - rich , rather than an am cvn - type helium cv ( vsnet - alert 12025 ) , although the strength of helium emission lines suggests an helium enrichment ( vsnet - alert 12026 ) .
the times of superhump maxima are listed in table [ tab : j1122oc2010 ] .
there is an indication of a systematic period decrease after @xmath101 .
we attributed this to a stage b
c transition and give the derived parameters based in this interpretation in table [ tab : perlist ] .
more detailed analysis and discussion will be presented in maehara et al .
, in preparation .
( 88mm,110mm)fig61.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55350.9981 & 0.0034 & @xmath170.0001 & 95 + 1 & 55351.0405 & 0.0033 & @xmath170.0032 & 89 + 36 & 55352.6341 & 0.0007 & 0.0019 & 54 + 37 & 55352.6784 & 0.0007 & 0.0007 & 83 + 44 & 55352.9942 & 0.0015 & @xmath170.0012 & 96 + 45 & 55353.0346 & 0.0033 & @xmath170.0061 & 95 + 49 & 55353.2263 & 0.0004 & 0.0040 & 148 + 52 & 55353.3564 & 0.0022 & @xmath170.0021 & 45 + 72 & 55354.2685 & 0.0006 & 0.0022 & 189 + 73 & 55354.3118 & 0.0012 & 0.0001 & 189 + 93 & 55355.2223 & 0.0007 & 0.0029 & 188 + 94 & 55355.2656 & 0.0007 & 0.0007 & 219 + 95 & 55355.3096 & 0.0012 &
@xmath170.0006 & 211 + 96 & 55355.3646 & 0.0048 & 0.0089 & 48 + 132 & 55356.9831 & 0.0030 & @xmath170.0066 & 95 + 137 & 55357.2187 & 0.0012 & 0.0020 & 83 + 138 & 55357.2635 & 0.0012 & 0.0014 & 94 + 139 & 55357.3070 & 0.0011 & @xmath170.0005 & 101 + 154 & 55357.9839 & 0.0046 & @xmath170.0044 & 42 + + + + we reported on the detection of superhumps and period variation in this object (= css090530:144011@xmath1494734 , hereafter ot j1440 ) in @xcite .
we present here a re - analysis after incorporating the data in @xcite .
the times of superhump maxima are listed in table [ tab : j1440oc2009 ] .
it has now become evident that @xmath102 corresponds to stage a , when the superhumps were indeed still in development and the object once started to fade temporarily ( figure [ fig : j14402009oc ] ) .
the period break reported in @xcite and @xcite after @xmath103 this is most likely a stage b c transition .
it is unusual for such a short-@xmath3 system to show a stage b
c transition at such an early stage .
the apparent lack of distinct positive @xmath7 during the stage b is also unusual .
the development of the superhump period the presence of a likely precursor ( or a stagnation ) in the light curve are similar to those of bz uma , which has been proposed to critically exhibit the su uma - type phenomenon @xcite .
since bz uma is known to show many normal outbursts in addition to a very rare superoutburst ( @xcite ; @xcite ) , a further systematic observation of ot j1440 is encouraged to explore the similarity between these objects and potential cause of the unusual period behavior .
( 88mm,90mm)fig62.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 54983.0245 & 0.0032 & @xmath170.0095 & 69 + 1 & 54983.0946 & 0.0101 & @xmath170.0039 & 64 + 7 & 54983.4733 & 0.0017 & @xmath170.0123 & 68 + 8 & 54983.5397 & 0.0015 & @xmath170.0104 & 68 + 15 & 54984.0046 & 0.0014 & 0.0029 & 191 + 16 & 54984.0697 & 0.0011 & 0.0034 & 223 + 22 & 54984.4581 & 0.0004 & 0.0048 & 164 + 23 & 54984.5231 & 0.0005 & 0.0052 & 168 + 24 & 54984.5869 & 0.0006 & 0.0046 & 74 + 38 & 54985.4914 & 0.0007 & 0.0059 & 106 + 39 & 54985.5564 & 0.0010 & 0.0063 & 105 + 53 & 54986.4604 & 0.0006 & 0.0072 & 71 + 54 & 54986.5254 & 0.0005 & 0.0077 & 64 + 69 & 54987.4837 & 0.0010 & @xmath170.0017 & 110 + 70 & 54987.5528 & 0.0014 & 0.0029 & 125 + 119 & 54990.7082 & 0.0008 & @xmath170.0028 & 47 + 120 & 54990.7752 & 0.0014 & @xmath170.0003 & 102 + 121 & 54990.8374 & 0.0014 & @xmath170.0026 & 169 + 122 & 54990.8997 & 0.0015 & @xmath170.0049 & 143 + 135 & 54991.7401 & 0.0078 & @xmath170.0031 & 98 + 136 & 54991.8083 & 0.0044 & 0.0006 & 119 + + + + this object (= css080505:163121@xmath1103134 , hereafter ot j1631 ) underwent another superoutburst in 2010 ( i. miller , baavss - alert 2268 ) .
the times of superhump maxima are listed in table [ tab : j1631oc2010 ] . although there seems to have a change in the period between @xmath104 and @xmath105 , the change was apparently too large to be attributed to a stage a b or b c transition .
it is likely that observation at @xmath104 was insufficient to derive a reliable maximum . disregarding observations before bjd 2455307
, we obtained a period of 0.06395(2 ) d with the pdm method .
ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55306.6220 & 0.0008 & @xmath170.0103 & 42 + 14 & 55307.5341 & 0.0009 & 0.0047 & 68 + 15 & 55307.6010 & 0.0010 & 0.0074 & 69 + 55 & 55310.1595 & 0.0018 & 0.0026 & 133 + 71 & 55311.1836 & 0.0020 & 0.0014 & 138 + 72 & 55311.2404 & 0.0012 & @xmath170.0059 & 108 + + + + this object (= css090622:170344@xmath1090835 , hereafter ot j1703 ) was discovered by the crts on 2009 june 22 .
the detection of superhumps ( vsnet - alert 11297 , 11298 ) led to a classification as an su uma - type dwarf nova .
although there still remain possibilities of aliases , we have adopted @xmath3 = 0.06085(2 ) d ( figure [ fig : j1703shpdm ] ) based on the best period determined from the single - night observation ( vsnet - alert 11298 ) .
the times of superhump maxima are listed in table [ tab : j1703oc2009 ] .
the outburst appears to have been caught when the amplitude of superhumps decayed ( the initial observation was performed @xmath9 5 d after the crts detection ) . according to crts observations ,
this object were further detected in outburst on 2009 july 22 and 29 .
since there were no previous crts detections before 2009 june , there may have been an enhanced outburst activity ( either normal outbursts or post - superoutburst rebrightenings ) in 2009 july .
( 88mm,110mm)fig63.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55009.4145 & 0.0021 & 0.0011 & 94 + 1 & 55009.4716 & 0.0012 & @xmath170.0026 & 100 + 2 & 55009.5366 & 0.0014 & 0.0016 & 103 + 51 & 55012.5157 & 0.0004 & @xmath170.0000 & 58
+ + + + this object ( hereafter ot j1821 ) was discovered by k. itagaki ( vsnet - alert 11952 ) .
the sdss color suggested that the object is a dwarf nova in outburst .
superhumps were subsequently detected ( vsnet - alert 11956 ; figure [ fig : j1821shpdm ] ) .
the times of superhump maxima are listed in table [ tab : j1821oc2010 ] .
these observations covered the final part of the superoutburst , and the recorded superhumps were likely stage c superhumps .
the lack of variation in the period is compatible with this interpretation .
( 88mm,110mm)fig64.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55312.1638 & 0.0008 & @xmath170.0034 & 80 + 1 & 55312.2544 & 0.0039 & 0.0052 & 50 + 18 & 55313.6433 & 0.0018 & @xmath170.0016 & 53 + 48 & 55316.1072 & 0.0011 & @xmath170.0005 & 46 + 49 & 55316.1883 & 0.0021 & @xmath170.0014 & 59 + 74 & 55318.2444 & 0.0055 & 0.0023 & 118 + 85 & 55319.1458 & 0.0119 & 0.0007 & 19 + 86 & 55319.2259 & 0.0284 & @xmath170.0013 & 31
+ + + + this transient ( hereafter ot j2138 ) was independently discovered by d .- a .
yi @xcite and s. kaneko @xcite .
@xcite spectroscopically confirmed that this is an outbursting dwarf nova ( see also vsnet - alert 11971 ) .
the spectrum by @xcite indicated the presence of highly excited emission lines , suggesting that this is a wz sge - type outburst [ cf .
vsnet - alert 11974 , see also vsnet - alert 11987 ( k. kinugasa , gunma astronomical observatory ) and follow - up spectroscopy by @xcite ] .
the earliest indication of short - period modulations ( likely early superhumps ) was detected by g. masi ( vsnet - alert 11977 , see also 11978 , 11985 ; figure [ fig : j2138eshpdm ] ) .
ordinary superhumps subsequently appeared ( vsnet - alert 11990 ; figure [ fig : j2138shpdm ] ) , strengthening the identification as a wz sge - type dwarf nova .
@xcite reported a detection of another outburst in 1942 , which is the only known outburst other than the present one .
the times of maxima for ordinary superhumps are listed in table [ tab : j2138oc2010 ] .
although there was an indication of double maxima during the late plateau phase , we did not attempt to distinguish the different peaks , and listed most prominent ones .
the overall @xmath0 diagram ( figure [ fig : j2138humpall ] ) bears a high degree of similarity to that of asas j0025 in 2004 @xcite . and the low amplitudes of post - superoutburst superhumps . ]
the main difference is that ot j2138 did not fade from the plateau before entering stage c. combined with the apparent presence of multiple hump peaks , this stage in ot j2138 appears to phenomenologically correspond to the phase between the rapid fading from the plateau and rebrightening in asas j0025 .
the apparent absence of a rebrightening in ot j2138 may be explained if this object somehow succeeded in maintaining the plateau phase ( or avoiding the quenching of the outbursting state ) when asas j0025 once returned to quiescence before the rebrightening .
the predominance of type - d superoutbursts ( superoutbursts without a rebrightening , see @xcite ) in very short-@xmath2 systems may be understood as this kind of variety . in table
[ tab : perlist ] , we listed intervals exhibiting best - defined maxima and omitted a phase with multiple peaks .
the post - superoutburst stage with @xmath106 showed a slightly shorter superhump period and the signal was persistent even during the very late post - superoutburst stage ( @xmath65 = 0.05487(1 ) d , figure [ fig : j2138latepdm ] ) .
there was no indication of a longer late - stage superhumps as observed in gw lib , v455 and , eg cnc , wz sge and several newly discovered wz sge - type dwarf novae ( @xcite ; @xcite ) .
more detailed analysis and discussion will be presented in maehara et al .
, in preparation .
( 88mm,110mm)fig65.eps ( 88mm,110mm)fig66.eps ( 160mm,160mm)fig67.eps ( 88mm,110mm)fig68.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55330.8319 & 0.0006 & 0.0019 & 89 + 6 & 55331.1881 & 0.0013 & 0.0274 & 309 + 36 & 55332.8555 & 0.0002 & 0.0412 & 76 + 37 & 55332.9109 & 0.0002 & 0.0414 & 83 + 38 & 55332.9650 & 0.0002 & 0.0404 & 116 + 41 & 55333.1300 & 0.0005 & 0.0401 & 125 + 42 & 55333.1838 & 0.0002 & 0.0388 & 662 + 43 & 55333.2391 & 0.0002 & 0.0390 & 832 + 44 & 55333.2940 & 0.0002 & 0.0387 & 503 + 54 & 55333.8428 & 0.0002 & 0.0363 & 128 + 55 & 55333.8963 & 0.0002 & 0.0347 & 149 + 56 & 55333.9519 & 0.0002 & 0.0351 & 106 + 60 & 55334.1724 & 0.0004 & 0.0352 & 145 + 61 & 55334.2265 & 0.0003 & 0.0342 & 477 + 62 & 55334.2810 & 0.0004 & 0.0335 & 427 + 66 & 55334.5003 & 0.0009 & 0.0324 & 35 + 67 & 55334.5554 & 0.0004 & 0.0324 & 53 + 68 & 55334.6094 & 0.0003 & 0.0313 & 53 + 73 & 55334.8845 & 0.0002 & 0.0307 & 152 + 80 & 55335.2689 & 0.0002 & 0.0293 & 110 + 86 & 55335.5978 & 0.0005 & 0.0275 & 151 + 87 & 55335.6535 & 0.0005 & 0.0281 & 117 + 90 & 55335.8173 & 0.0003 & 0.0265 & 102 + 91 & 55335.8749 & 0.0004 & 0.0290 & 130 + 103 & 55336.5340 & 0.0014 & 0.0267 & 52 + 104 & 55336.5871 & 0.0008 & 0.0246 & 134 + 105 & 55336.6417 & 0.0009 & 0.0241 & 171 + 108 & 55336.8067 & 0.0003 & 0.0238 & 153 + 109 & 55336.8625 & 0.0002 & 0.0244 & 244 + 110 & 55336.9171 & 0.0002 & 0.0239 & 145 + 114 & 55337.1408 & 0.0024 & 0.0271 & 56 + 115 & 55337.1938 & 0.0005 & 0.0250 & 97 + 116 & 55337.2511 & 0.0002 & 0.0272 & 387 + 122 & 55337.5806 & 0.0007 & 0.0260 & 179 + 123 & 55337.6346 & 0.0007 & 0.0249 & 182 + 127 & 55337.8545 & 0.0002 & 0.0242 & 104 + 128 & 55337.9101 & 0.0002 & 0.0247 & 106 + 129 & 55337.9647 & 0.0004 & 0.0242 & 146 + 132 & 55338.1396 & 0.0018 & 0.0338 & 131 + 133 & 55338.1893 & 0.0005 & 0.0284 & 500 + 134 & 55338.2408 & 0.0004 & 0.0248 & 578 + 137 & 55338.4063 & 0.0006 & 0.0248 & 92 + 138 & 55338.4633 & 0.0003 & 0.0267 & 113 + 139 & 55338.5180 & 0.0005 & 0.0263 & 163 + 140 & 55338.5746 & 0.0008 & 0.0278 & 141 + 141 & 55338.6259 & 0.0008 & 0.0239 & 159 + 145 & 55338.8472 & 0.0004 & 0.0248 & 89 + 146 & 55338.9023 & 0.0003 & 0.0247 & 74 + 156 & 55339.4574 & 0.0005 & 0.0287 & 175 + 157 & 55339.5065 & 0.0005 & 0.0227 & 208 + 158 & 55339.5705 & 0.0017 & 0.0315 & 85 + 159 & 55339.6238 & 0.0017 & 0.0297 & 76 + 162 & 55339.7897 & 0.0029 & 0.0302 & 51 + 163 & 55339.8426 & 0.0006 & 0.0280 & 82 + 164 & 55339.8972 & 0.0007 & 0.0275 & 82 + 187 & 55341.1909 & 0.0016 & 0.0535 & 82 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 188 & 55341.2308 & 0.0011 & 0.0383 & 264 + 189 & 55341.2788 & 0.0010 & 0.0311 & 292 + 192 & 55341.4479 & 0.0015 & 0.0348 & 78 + 193 & 55341.5139 & 0.0028 & 0.0457 & 75 + 199 & 55341.8388 & 0.0019 & 0.0400 & 118 + 200 & 55341.8928 & 0.0009 & 0.0388 & 119 + 201 & 55341.9568 & 0.0011 & 0.0477 & 158 + 204 & 55342.1406 & 0.0011 & 0.0662 & 44 + 209 & 55342.3890 & 0.0021 & 0.0390 & 46 + 210 & 55342.4457 & 0.0012 & 0.0405 & 113 + 212 & 55342.5794 & 0.0020 & 0.0640 & 101 + 213 & 55342.6214 & 0.0016 & 0.0508 & 84 + 217 & 55342.8330 & 0.0023 & 0.0419 & 169 + 218 & 55342.8921 & 0.0011 & 0.0460 & 174 + 228 & 55343.4363 & 0.0005 & 0.0389 & 116 + 229 & 55343.4904 & 0.0008 & 0.0379 & 115 + 230 & 55343.5673 & 0.0012 & 0.0597 & 89 + 231 & 55343.6096 & 0.0092 & 0.0469 & 99 + 233 & 55343.7196 & 0.0036 & 0.0466 & 51 + 234 & 55343.7661 & 0.0036 & 0.0381 & 62 + 235 & 55343.8313 & 0.0023 & 0.0481 & 95 + 236 & 55343.8757 & 0.0011 & 0.0374 & 116 + 241 & 55344.1558 & 0.0013 & 0.0419 & 90 + 242 & 55344.2173 & 0.0009 & 0.0483 & 103 + 246 & 55344.4345 & 0.0012 & 0.0449 & 107 + 247 & 55344.4832 & 0.0011 & 0.0386 & 107 + 248 & 55344.5569 & 0.0009 & 0.0571 & 88 + 249 & 55344.5973 & 0.0021 & 0.0424 & 57 + 267 & 55345.5896 & 0.0014 & 0.0425 & 81 + 268 & 55345.6420 & 0.0013 & 0.0399 & 83 + 270 & 55345.7540 & 0.0050 & 0.0416 & 60 + 271 & 55345.8132 & 0.0038 & 0.0457 & 59 + 272 & 55345.8721 & 0.0171 & 0.0495 & 44 + 274 & 55345.9678 & 0.0007 & 0.0350 & 83 + 278 & 55346.1947 & 0.0004 & 0.0413 & 398 + 279 & 55346.2512 & 0.0004 & 0.0427 & 508 + 284 & 55346.5280 & 0.0011 & 0.0439 & 46 + 285 & 55346.5840 & 0.0022 & 0.0448 & 130 + 286 & 55346.6330 & 0.0022 & 0.0387 & 116 + 290 & 55346.8566 & 0.0019 & 0.0418 & 104 + 291 & 55346.9111 & 0.0011 & 0.0411 & 104 + 292 & 55346.9569 & 0.0019 & 0.0319 & 165 + 295 & 55347.1368 & 0.0010 & 0.0464 & 177 + 296 & 55347.1814 & 0.0025 & 0.0359 & 198 + 297 & 55347.2432 & 0.0016 & 0.0425 & 80 + 302 & 55347.5203 & 0.0019 & 0.0440 & 47 + 303 & 55347.5682 & 0.0049 & 0.0368 & 84 + 304 & 55347.6304 & 0.0009 & 0.0440 & 113 + 308 & 55347.8539 & 0.0005 & 0.0469 & 103 + 309 & 55347.9080 & 0.0006 & 0.0459 & 104 + 310 & 55347.9612 & 0.0008 & 0.0440 & 67 + 313 & 55348.1229 & 0.0014 & 0.0403 & 98 + 314 & 55348.1771 & 0.0002 & 0.0394 & 2499 + 315 & 55348.2376 & 0.0018 & 0.0448 & 76 + 321 & 55348.5653 & 0.0005 & 0.0418 & 76 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 322 & 55348.6266 & 0.0011 & 0.0480 & 101 + 323 & 55348.6800 & 0.0008 & 0.0462 & 45 + 331 & 55349.1111 & 0.0008 & 0.0364 & 119 + 332 & 55349.1650 & 0.0007 & 0.0352 & 264 + 333 & 55349.2184 & 0.0008 & 0.0334 & 358 + 334 & 55349.2800 & 0.0004 & 0.0399 & 396 + 339 & 55349.5518 & 0.0004 & 0.0362 & 72 + 340 & 55349.6091 & 0.0017 & 0.0383 & 82 + 341 & 55349.6629 & 0.0007 & 0.0370 & 50 + 346 & 55349.9285 & 0.0013 & 0.0270 & 49 + 347 & 55349.9937 & 0.0006 & 0.0371 & 53 + 349 & 55350.1040 & 0.0011 & 0.0371 & 142 + 350 & 55350.1519 & 0.0008 & 0.0299 & 262 + 351 & 55350.2145 & 0.0012 & 0.0374 & 363 + 352 & 55350.2652 & 0.0005 & 0.0329 & 325 + 362 & 55350.8202 & 0.0011 & 0.0367 & 53 + 363 & 55350.8695 & 0.0005 & 0.0309 & 73 + 364 & 55350.9235 & 0.0007 & 0.0299 & 73 + 368 & 55351.1500 & 0.0012 & 0.0358 & 291 + 369 & 55351.2004 & 0.0004 & 0.0311 & 374 + 370 & 55351.2553 & 0.0004 & 0.0309 & 383 + 376 & 55351.5870 & 0.0013 & 0.0319 & 57 + 377 & 55351.6347 & 0.0004 & 0.0244 & 57 + 387 & 55352.1884 & 0.0006 & 0.0270 & 47 + 388 & 55352.2459 & 0.0008 & 0.0293 & 236 + 394 & 55352.5725 & 0.0009 & 0.0252 & 58 + 395 & 55352.6220 & 0.0009 & 0.0196 & 59 + 399 & 55352.8451 & 0.0006 & 0.0223 & 61 + 400 & 55352.9027 & 0.0005 & 0.0247 & 73 + 401 & 55352.9523 & 0.0006 & 0.0191 & 151 + 406 & 55353.2247 & 0.0008 & 0.0160 & 240 + 407 & 55353.2892 & 0.0011 & 0.0254 & 175 + 410 & 55353.4490 & 0.0002 & 0.0198 & 184 + 411 & 55353.5024 & 0.0006 & 0.0181 & 172 + 412 & 55353.5589 & 0.0006 & 0.0195 & 104 + 413 & 55353.6161 & 0.0015 & 0.0216 & 58 + 414 & 55353.6714 & 0.0007 & 0.0218 & 41 + 417 & 55353.8271 & 0.0016 & 0.0120 & 59 + 418 & 55353.8895 & 0.0013 & 0.0194 & 73 + 419 & 55353.9565 & 0.0010 & 0.0312 & 121 + 420 & 55353.9933 & 0.0008 & 0.0129 & 56 + 428 & 55354.4397 & 0.0005 & 0.0183 & 120 + 429 & 55354.4925 & 0.0004 & 0.0160 & 123 + 430 & 55354.5502 & 0.0008 & 0.0186 & 80 + 431 & 55354.5990 & 0.0003 & 0.0123 & 58 + 432 & 55354.6537 & 0.0018 & 0.0119 & 58 + 446 & 55355.4283 & 0.0004 & 0.0148 & 123 + 447 & 55355.4806 & 0.0006 & 0.0119 & 122 + 448 & 55355.5368 & 0.0010 & 0.0130 & 72 + 449 & 55355.5925 & 0.0008 & 0.0136 & 56 + 450 & 55355.6435 & 0.0004 & 0.0095 & 56 + 456 & 55355.9745 & 0.0008 & 0.0098 & 52 + 465 & 55356.4668 & 0.0003 & 0.0060 & 155 + 466 & 55356.5203 & 0.0004 & 0.0044 & 169 + 467 & 55356.5748 & 0.0010 & 0.0037 & 25 + 477 & 55357.1070 & 0.0034 & @xmath170.0152 & 54 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 478 & 55357.1882 & 0.0026 & 0.0108 & 256 + 479 & 55357.2414 & 0.0009 & 0.0089 & 272 + 480 & 55357.2894 & 0.0009 & 0.0018 & 119 + 482 & 55357.4003 & 0.0016 & 0.0024 & 61 + 483 & 55357.4555 & 0.0010 & 0.0025 & 62 + 484 & 55357.5187 & 0.0014 & 0.0107 & 63 + 485 & 55357.5628 & 0.0011 & @xmath170.0004 & 52 + 486 & 55357.6172 & 0.0004 & @xmath170.0011 & 54 + 487 & 55357.6766 & 0.0027 & 0.0032 & 29 + 496 & 55358.1624 & 0.0008 & @xmath170.0071 & 259 + 497 & 55358.2249 & 0.0011 & 0.0003 & 301 + 498 & 55358.2725 & 0.0052 & @xmath170.0072 & 269 + 501 & 55358.4491 & 0.0007 & 0.0040 & 59 + 502 & 55358.4989 & 0.0006 & @xmath170.0013 & 61 + 503 & 55358.5534 & 0.0006 & @xmath170.0020 & 46 + 504 & 55358.6081 & 0.0017 & @xmath170.0024 & 54 + 505 & 55358.6608 & 0.0010 & @xmath170.0048 & 45 + 508 & 55358.8288 & 0.0004 & @xmath170.0022 & 59 + 509 & 55358.8805 & 0.0006 & @xmath170.0055 & 73 + 510 & 55358.9432 & 0.0006 & 0.0020 & 73 + 514 & 55359.1547 & 0.0013 & @xmath170.0070 & 177 + 515 & 55359.2161 & 0.0009 & @xmath170.0007 & 222 + 516 & 55359.2561 & 0.0006 & @xmath170.0159 & 224 + 519 & 55359.4297 & 0.0011 & @xmath170.0075 & 60 + 520 & 55359.4892 & 0.0004 & @xmath170.0033 & 61 + 521 & 55359.5382 & 0.0007 & @xmath170.0093 & 35 + 527 & 55359.8668 & 0.0004 & @xmath170.0115 & 73 + 528 & 55359.9224 & 0.0019 & @xmath170.0110 & 73 + 537 & 55360.4107 & 0.0006 & @xmath170.0187 & 61 + 538 & 55360.4622 & 0.0019 & @xmath170.0223 & 61 + 545 & 55360.8531 & 0.0004 & @xmath170.0173 & 73 + 546 & 55360.9034 & 0.0005 & @xmath170.0221 & 73 + 547 & 55360.9613 & 0.0011 & @xmath170.0193 & 49 + 555 & 55361.4070 & 0.0009 & @xmath170.0146 & 53 + 556 & 55361.4620 & 0.0008 & @xmath170.0147 & 62 + 557 & 55361.5089 & 0.0022 & @xmath170.0229 & 54 + 558 & 55361.5579 & 0.0083 & @xmath170.0291 & 60 + 559 & 55361.6198 & 0.0009 & @xmath170.0223 & 56 + 577 & 55362.5991 & 0.0005 & @xmath170.0352 & 56 + 578 & 55362.6583 & 0.0006 & @xmath170.0311 & 42 + 594 & 55363.5412 & 0.0005 & @xmath170.0301 & 39 + 595 & 55363.5908 & 0.0020 & @xmath170.0356 & 56 + 596 & 55363.6413 & 0.0010 & @xmath170.0402 & 54 + 613 & 55364.5834 & 0.0011 & @xmath170.0351 & 55 + 614 & 55364.6486 & 0.0036 & @xmath170.0251 & 52 + 631 & 55365.5662 & 0.0010 & @xmath170.0446 & 56 + 632 & 55365.6194 & 0.0026 & @xmath170.0464 & 54 + 638 & 55365.9491 & 0.0005 & @xmath170.0475 & 85 + 649 & 55366.5640 & 0.0007 & @xmath170.0389 & 42 + 650 & 55366.6105 & 0.0009 & @xmath170.0475 & 55 + 651 & 55366.6678 & 0.0011 & @xmath170.0453 & 29 + 668 & 55367.6003 & 0.0006 & @xmath170.0498 & 52 + 669 & 55367.6548 & 0.0011 & @xmath170.0505 & 46 + 685 & 55368.5408 & 0.0009 & @xmath170.0464 & 36 + 686 & 55368.6132 & 0.0028 & @xmath170.0291 & 53 + 687 & 55368.6551 & 0.0008 & @xmath170.0423 & 47 + + + + ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 704 & 55369.5719 & 0.0015 & @xmath170.0626 & 55 + 705 & 55369.6239 & 0.0012 & @xmath170.0657 & 52 + 722 & 55370.5701 & 0.0011 & @xmath170.0566 & 50 + 777 & 55373.5876 & 0.0012 & @xmath170.0706 & 47 + 778 & 55373.6503 & 0.0012 & @xmath170.0630 & 50 + 813 & 55375.5696 & 0.0014 & @xmath170.0729 & 55 + 814 & 55375.6237 & 0.0045 & @xmath170.0740 & 56 + 815 & 55375.6692 & 0.0025 & @xmath170.0836 & 35 + 831 & 55376.5545 & 0.0034 & @xmath170.0802 & 49 + 833 & 55376.6535 & 0.0149 & @xmath170.0914 & 53 + 849 & 55377.5468 & 0.0006 & @xmath170.0800 & 37 + 850 & 55377.5936 & 0.0013 & @xmath170.0884 & 56 + 851 & 55377.6485 & 0.0013 & @xmath170.0886 & 54 + 868 & 55378.5821 & 0.0014 & @xmath170.0921 & 57 + 869 & 55378.6184 & 0.0171 & @xmath170.1109 & 56 + 886 & 55379.5563 & 0.0020 & @xmath170.1100 & 45 + 887 & 55379.6170 & 0.0012 & @xmath170.1044 & 46 + 888 & 55379.6683 & 0.0015 & @xmath170.1082 & 39 + 923 & 55381.5977 & 0.0025 & @xmath170.1081 & 55 + 924 & 55381.6300 & 0.0050 & @xmath170.1309 & 54 + 1104 & 55391.5100 & 0.0004 & @xmath170.1725 & 49 + 1104 & 55391.5492 & 0.0018 & @xmath170.1333 & 58 + 1140 & 55393.5096 & 0.0004 & @xmath170.1572 & 92 + 1155 & 55394.3357 & 0.0019 & @xmath170.1579 & 53 + 1156 & 55394.3778 & 0.0007 & @xmath170.1710 & 71 + 1174 & 55395.3714 & 0.0011 & @xmath170.1695 & 82 + 1176 & 55395.4992 & 0.0019 & @xmath170.1519 & 54 + 1233 & 55398.5953 & 0.0007 & @xmath170.1977 & 54 + 1234 & 55398.6551 & 0.0006 & @xmath170.1930 & 43 + 1251 & 55399.6012 & 0.0011 & @xmath170.1839 & 53 + 1342 & 55404.5667 & 0.0038 &
@xmath170.2344 & 42 + 1343 & 55404.6583 & 0.0030 & @xmath170.1978 & 42 + + + + this object (= css100615:215815@xmath1094709 , hereafter ot j2158 ) was discovered by the crts in 2010 june .
astrokolkhoz team immediately detected superhumps ( vsnet - outburst 11306 ) .
the times of superhump maxima are listed in table [ tab : j2158oc2010 ] .
the mean period determined with the pdm method was 0.07755(9 ) d. the large amplitudes ( 0.35 mag , figure [ fig : j2158shpdm ] ) and waveform suggest that these superhumps are stage b superhumps at their early evolution .
( 88mm,110mm)fig69.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55366.5092 & 0.0008 & 0.0001 & 25 + 1 & 55366.5868 & 0.0003 & 0.0000 & 50 + 2 & 55366.6644 & 0.0005 & @xmath170.0000 & 56 + 4 & 55366.8190 & 0.0003 & @xmath170.0007 & 108 + 5 & 55366.8979 & 0.0005 & 0.0005 & 74 + + + + this object (= css090727:223003@xmath17145835 , hereafter ot j2230 ) was discovered by the crts in 2009 july . the large outburst amplitude exceeding 6 mag
was immediately noted ( vsnet - alert 11351 ) .
superhump - like modulations were soon reported ( vsnet - alert 11353 ) .
it became evident by later observations that these modulations were early superhumps of a wz sge - type dwarf nova rather than ordinary superhumps ( vsnet - alert 11365 , 11388 ) .
we obtained a mean period of 0.05841(1 ) d with the pdm method ( figure [ fig : j2230eshpdm ] ) .
doubly humped modulations characteristic to early superhumps ( cf .
@xcite ) are clearly seen . due to the faintness of the object ,
the appearance of ordinary superhumps was not recorded .
the outburst lasted for at least 18 d. ( 88mm,110mm)fig70.eps this object (= mls100904:234441@xmath17001206 , hereafter ot j2344 ) was discovered by the crts in 2010 september during the course of the mount lemmon survey .
p. wils notified this event ( cvnet - outburst 3828 ) . immediately following this announcement ,
superhumps were clearly detected ( vsnet - alert 12143 , 12144 ; figure [ fig : j2344shpdm ] ) .
the times of superhump maxima are listed in table [ tab : j2344oc2010 ] .
these superhumps were likely stage c superhumps since there was little evidence of period variation and the object started a rapid fading 4 d after the start of the observation .
( 88mm,110mm)fig71.eps ccccc @xmath15 & max & error & @xmath0 & @xmath16 + 0 & 55444.4661 & 0.0009 & 0.0024 & 38 + 1 & 55444.5425 & 0.0005 & 0.0020 & 77 + 2 & 55444.6163 & 0.0005 & @xmath170.0008 & 104 + 9 & 55445.1507 & 0.0066 & @xmath170.0035 & 76 + 12 & 55445.3861 & 0.0010 & 0.0019 & 38 + 13 & 55445.4615 & 0.0012 & 0.0005 & 39 + 14 & 55445.5387 & 0.0005 & 0.0010 & 39 + 15 & 55445.6111 & 0.0011 & @xmath170.0033 & 39 + 26 & 55446.4576 & 0.0005 & @xmath170.0006 & 170 + 27 & 55446.5316 & 0.0007 & @xmath170.0034 & 195 + 28 & 55446.6107 & 0.0011 & @xmath170.0009 & 149 + 38 & 55447.3778 & 0.0009 & @xmath170.0009 & 172 + 39 & 55447.4547 & 0.0009 & @xmath170.0008 & 173 + 40 & 55447.5373 & 0.0020 & 0.0052 & 112 + 48 & 55448.1513 & 0.0081 & 0.0054 & 111 + 51 & 55448.3756 & 0.0011 & @xmath170.0005 & 172 + 52 & 55448.4519 & 0.0014 & @xmath170.0009 & 172 + 53 & 55448.5283 & 0.0018 & @xmath170.0011 & 173 + 60 & 55449.0629 & 0.0026 & @xmath170.0035 & 40 + 61 & 55449.1448 & 0.0042 & 0.0017 & 21 + + + +
the new data for su uma - type dwarf novae have improved the statistics presented in @xcite .
although the presentations in @xcite covered a wide range of relations , we restrict only to key figures after combining the samples with @xcite in this paper .
figure [ fig : pdotpsh2 ] represents the relation between @xmath3 and @xmath7 during stage b. the enlarged figure ( corresponding to figure 10 in @xcite ) is only shown here .
the estimated density map is overlaid for a better visualization .
the new data generally confirmed the predominance of positive period derivatives during stage b in systems with superhump periods shorter than 0.07 d , in agreement with the tendency reported in @xcite .
( 120mm,80mm)fig72.eps in figure [ fig : pdotpsh2 ] , there appears to be a systematic difference in @xmath7 for systems with @xmath3 longer than 0.075 d. we have selected samples for @xmath107 from @xcite and this paper . after using the same criterion for samples described in @xcite and rejecting poorly determined ( nominal error of @xmath7 exceeding @xmath108 ) superoutbursts
, we applied student s @xmath109-test .
the test indicated that these samples are different at a significance level of 0.09 .
this test suggests that the properties of samples are different between these two data sets .
it is already apparent all well - determined @xmath7 s in this paper for this @xmath3 region has positive @xmath7 s , while many systems had negative @xmath7 s in @xcite .
this might have been caused by a small number of long-@xmath3 samples in this paper , since most of newly discovered su uma - type dwarf novae have short @xmath3 .
there would also be a possibility , however , that most of long-@xmath3 with frequent superoutbursts were already discovered at the time of @xcite and that long-@xmath3 systems in this papers were heavily biased toward systems with infrequent superoutbursts .
this possibility would be supported considering that all systems with supercycles longer than 1000 d ( v1251 cyg , rz leo , qy per ) have large positive @xmath7 ( see also a discussion in @xcite subsection 4.10 ) while systems with @xmath110 are restricted to systems with supercycles shorter than 400 d ( dh aql , v1316 cyg , cu vel ) .
although a direct correlation coefficient ( 0.20 ) between @xmath111 supercycle and @xmath7 is not significant , these results seem to strengthen the idea in @xcite that the degree of period variation is more diverse in long-@xmath3 systems .
there is , however , a case of iy uma , with frequent superoutbursts , which showed a positive @xmath7 in this new sample .
it is not yet clear whether some of long-@xmath3 have different @xmath7 s between different superoutbursts or whether this was a mere consequence of poor sampling in the past .
future studies in these viewpoints would be fruitful .
the apparent lack of systematic difference between different superoutbursts of the same system is one of the main conclusions in @xcite .
we further examined this issue using new materials .
the new data strengthen this conclusions , including well - observed superoutbursts preceded by precursor outbursts ( e.g. pu cma and 1rxs j0532 .
it would be noteworthy 1rxs j0532 almost completely reproduced the behavior in 2005 , which has been reported to be rather unusual @xcite .
this suggests that peculiarities found in certain systems are specific to these systems , rather than a chance occurrence .
this might suggest the existence of parameters specific to systems in addition to general parameters such as @xmath14 or @xmath2 . in v844 her ,
our observations give additional support to the hypothesis that the delay time in development of superhumps is shorter in significantly smaller superoutbursts . in the 2009 superoutburst of iy uma
, we recorded a strong beat phenomenon ( in the total luminosity ) between superhumps and orbital variation .
the beat period largely varied between stage b and c , and were in very good agreement with the beat periods expected from superhump periods during the corresponding stages .
the close correlation between the beat period and the superhump period suggests that the change in the angular velocity of the global apsidal motion is more responsible for the stage b
c transition rather than the appearance of a more localized new component with a different period ( or angular velocity ) , such as a bright spot on the disk .
future research focusing on this relation between the beat period and superhump period will be an important key in clarifying the nature of period variations of superhumps .
the new sample includes three new wz sge - type objects with established early superhumps , sdss j1610 , ot j1044 and ot j2230 , and one wz sge - type object with likely early superhumps ( ot j2138 ) .
all objects have @xmath3 shorter than 0.061 d. we also suggest that two systems , vx for and el uma , are wz sge - type dwarf novae with multiple rebrightenings . although the stage with early superhumps was not observed in vx for , the observed @xmath7 for the first time predicted for its multiple rebrightening in real - time .
this success strengthens the relationship between @xmath7 and type of rebrightenings in wz sge - type dwarf novae proposed in @xcite .
although el uma was only observed for its supposed rebrightening phase , all the known properties are strongly suggestive of a wz sge - type system with multiple rebrightenings .
the @xmath0 analysis of ot j2138 and its comparison to asas j0025 suggest an interpretation that the frequent absence of rebrightenings in very short-@xmath2 objects can be a result of sustained superoutburst plateau when usual su uma - type dwarf novae return to quiescence preceding a rebrightening . although this phenomenon may be somehow related to a small binary separation resulting stronger tidal torque and thereby stronger tidal dissipation to maintain the outbursting state , the exact mechanism should await further clarification .
the authors are grateful to observers of vsnet collaboration and vsolj observers who supplied vital data .
we acknowledge with thanks the variable star observations from the aavso international database contributed by observers worldwide and used in this research .
this work is deeply indebted to outburst detections and announcement by a number of variable star observers worldwide , including participants of cvnet , baa vss alert and avson networks .
we are grateful to d. boyd and his collaborators for making the data for ot j1440 available .
the ccd operation of the bronberg observatory is partly sponsored by the center for backyard astrophysics .
the ccd operation by peter nelson is on loan from the aavso , funded by the curry foundation .
we are grateful to the catalina real - time transient survey team for making their real - time detection of transient objects available to the public .
we are also grateful to h. takahashi and k. kinugasa for making unpublished spectra of el uma available to us .
we are grateful to the anonymous referee for pointing out the apparent difference of distribution of @xmath7 between this paper and @xcite .
we explored three applications of bayesian statistics in period analysis employed in this paper .
bayesian statistics ( see e.g. @xcite ) provide a framework in estimating a posterior probability density function ( pdf ) of model parameters from a combination of the observed data , a likelihood function defined by the model , and a prior pdf of the model parameters . according to the bayes theorem , the posterior pdf of the model parameters @xmath112 is we usually employ markov - chain monte carlo ( mcmc ) method in order to obtain the pdf . the metropolis - hastings algorithm ( @xcite ; @xcite ) , one of the best known of mcmc implementation , is as follows : starting with @xmath119 , and at the @xmath120-th step of mcmc , we obtain @xmath121 .
we then move to a new point @xmath122 following a proposal density distribution @xmath123 .
the new point is accepted with a probability multivariate gaussian movements are frequently used to cancel out @xmath125 and @xmath123 . if the proposal is rejected , @xmath126 . after discarding initial ` burn - in ' steps , we sample the remaining steps of the markov chain to obtain the pdf . typical numbers of steps are @xmath127@xmath128 for the entire chain and @xmath129@xmath130 for the ` burn - in ' steps . as reviewed in @xcite ,
the @xmath0 diagrams of superhump maxima in su uma - type dwarf novae generally follow discrete stages : most frequently a segment with a positive @xmath7 and a later segment with a discontinuously shorter constant period ( stages b and c in @xcite ) . determining the parameters for periods and @xmath7 and
the time of transition is not a trivial task for the classical statistics . @xcite
dealt with this problem using the bayesian statistics and the mcmc method .
we formulate this problem for a wider usage . using this likelihood , or its combination with priors ,
we can obtain the pdf with the mcmc method .
figure [ fig : ocsamp1 ] is a sample application to superhump timings in qz vir in 2004 @xcite .
the initial @xmath136 was chosen as 85 and other initial parameters were determined by linear least - squares fitting .
the @xmath140 s of the multivariate gaussian distribution for the proposal density were 0.04 times 1-@xmath140 errors of the initial fitting .
the total length of the chain was @xmath130 and the first 5000 samples were discarded for obtaining the pdf .
the mean parameters and standard errors were @xmath141 , @xmath142 , @xmath143 , @xmath144 and @xmath145 .
the posterior means of two parameters @xmath146 and @xmath136 are shown in figure [ fig : ocsamp2 ] . when expected values of some parameters are empirically known .
this application of priors would be useful when @xmath7 of a certain object is known from observation of other superoutbursts while a particular observation has a significant gap to determine parameters by itself .
figure [ fig : ocsampprior ] demonstrates the effect of priors for observations with a gaps .
the data are times of superhumps of qz vir during the 2007 superoutburst .
an unconstrained model gives a negative @xmath7 of @xmath148 and a break at @xmath149 ( dashed curve ) .
incorporating prior knowledge that this object has a positive @xmath7 , i.e. setting a prior with @xmath150 and @xmath151 , we get a reasonable fit ( solid curve ) with a break at @xmath152 .
since this example is for a demonstration purpose of bayesian approach , we did not use these values in the main text .
although a proper way of using priors in such a problem need to be further investigated , this formulation would provide a way of analyzing a badly sampled observations when we have firm knowledge in choosing the prior .
we consider a bayesian extension of parameter fitting of observed light curves with a cyclic function . in this case , @xmath153 is the observations at @xmath154 , and the parameter space is @xmath155 defined by the model although a directed application of the original mcmc to solve the aliasing problem is limited due to the slow mixing between widely separate peaks , it is well suited for sampling the detailed structure of individual peaks . figure [ fig : v1454cygmcmc ] illustrates the dependence on the model functions .
although a spline - interpolated model light curve usually gives a smaller standard error for parameter estimates , the difference is not usually very big . in the present application to v1454 cyg , spline and sine
fits give @xmath164 and @xmath165 , respectively .
these analysis has confirmed the results , both period and error estimate , of the pdm analysis .
this method would be advantageous when the model description becomes more complex than what the original pdm method describes .
the pdm method evaluates dispersions of the observed data against phase - binned averages .
this prescription sometimes produces spurious scatters in the resultant theta diagram when the number of data ( and data in each bin ) is small . instead of getting phase - averaged light curves ( at given trial periods ) using a small number of discrete bins ,
we introduce a bayesian approach for obtaining continuous phase - averaged light curves .
we model light curves by a set of parameters @xmath166 ( at a trial period ) for phases @xmath167 ) , where @xmath16 is the number of phase bins , which are large ( @xmath168 ) enough to describe continuous functions . following a standard technique in bayesian analysis
, we then express the condition of smoothness of @xmath112 by introducing a prior function assuming that second order differences of @xmath175 follow a normal distribution @xmath176 @xmath177 } , \ ] ] where @xmath178 and @xmath179 reflecting the cyclic condition .
although the smoothing parameter @xmath183 can be estimated from individual fits by different trial periods , we adopt a constant @xmath183 for all periods .
the value is estimated from the best fit , i.e. from the period the giving smallest dispersion . using this smooth phase - average light curve
, we can then estimate overall variance as prescribed as in @xcite . | as an extension of the project in @xcite , we collected times of superhump maxima for 61 su uma - type dwarf novae mainly observed during the 20092010 season .
the newly obtained data confirmed the basic findings reported in @xcite : the presence of stages a c , as well as the predominance of positive period derivatives during stage b in systems with superhump periods shorter than 0.07 d. there was a systematic difference in period derivatives for systems with superhump periods longer than 0.075 d between this study and @xcite .
we suggest that this difference is possibly caused by the relative lack of frequently outbursting su uma - type dwarf novae in this period regime in the present study .
we recorded a strong beat phenomenon during the 2009 superoutburst of iy uma .
the close correlation between the beat period and superhump period suggests that the changing angular velocity of the apsidal motion of the elliptical disk is responsible for the variation of superhump periods .
we also described three new wz sge - type objects with established early superhumps and one with likely early superhumps .
we also suggest that two systems , vx for and el uma , are wz sge - type dwarf novae with multiple rebrightenings . the @xmath0 variation in ot j213806.6@xmath1261957 suggests that the frequent absence of rebrightenings in very short-@xmath2 objects can be a result of sustained superoutburst plateau at the epoch when usual su uma - type dwarf novae return to quiescence preceding a rebrightening .
we also present a formulation for a variety of bayesian extension to traditional period analyses . |
in relativistic heavy - ion collisions at top rhic and the lhc energies , a hot and dense matter the quark - gluon plasma ( qgp ) has been created . during its expansion ,
the qgp fireball quickly cools down , which then undergos the phase transition and produces a large amount of hadrons .
these hadrons subsequently experience frequent inelastic and elastic collisions during the hadronic evolution . with the termination of inelastic and elastic collisions ,
the evolving system reach the chemical and thermal freeze - out , respectively . traditionally , the chemical freeze - out temperature @xmath2 and the baryon chemical potential @xmath3 are extracted from the particle yields of various hadrons using the statistical model @xcite , which gives a uniform @xmath2 and @xmath3 for all hadrons species .
similarly , pure hydrodynamic simulations implement a uniform thermal freeze - out temperature to construct the freeze - out hyper - surface , which decouples various hadrons from the bulk matter @xcite .
microscopically , the related inelastic and elastic scattering channels are different for each hadron species . as a result ,
the chemical and thermal freeze - out procedures are very possibly hadron - species dependent . in this proceeding
, we will briefly summarize our recent investigation on the chemical and thermal freeze - out of various hadrons in 2.76 a tev pb+pb collisions @xcite , based on dynamical simulations from vishnu hybrid model . , @xmath4 , @xmath5 , @xmath6 , @xmath0 , and @xmath1 in the most central pb+pb collisions at @xmath7=2.76 tev .
the experimental data are from the alice collaboration @xcite .
theoretical curves are from vishnu hybrid model calculations , using mc - kln initial conditions , @xmath8 and @xmath9.,scaledwidth=80.0% ]
our calculations implement vishnu hybrid model @xcite that combines ( 2 + 1)-d relativistic viscous hydrodynamics ( vish2 + 1 ) @xcite for the qgp fluid expansion with a microscopic hadronic transport model ( urqmd ) @xcite for the hadron resonance gas evolution .
the hydrodynamic calculations input the equation of state ( eos ) s95p - pce @xcite with the transition to urqmd at a constant temperature of 165 mev . following ref .
@xcite , we input smooth initial conditions generated from the mc - kln model @xcite , where the normalization factor of the initial entropy density profiles is tuned to reproduce the multiplicity of all charged hadrons in the most central pb+pb collisions @xcite .
the hydrodynamic simulations start at @xmath10 with the specific shear viscosity @xmath11 set to 0.16 , and the bulk viscosity set to zero @xcite .
the chemical and thermal freeze - out of the evolving system are respectively related to the terminations of inelastic and elastic collisions .
after the chemical freeze - out , the hadron yields no longer change .
after the thermal freeze - out , the momentum distributions of various hadrons no longer change . in this section
, we will briefly summarize our recent investigations on the chemical and thermal freeze - out of various hadron species at the lhc , based on dynamical simulations from vishnu hybrid model .
figure 1 presents the vishnu calculations of the particle yields for @xmath12 , @xmath4 , @xmath5 , @xmath6 , @xmath0 , and @xmath1 in the most central pb+pb collisions , which nicely fit the corresponding experimental data measured by alice .
it was found that the baryon and anti - baryon ( @xmath13-@xmath14 ) annihilations during the hadronic evolution reduce the proton and anti - proton ( @xmath15 ) yields by @xmath16 , largely improving the description of the @xmath15 data @xcite .
@xcite and @xcite also showed that vishnu hybrid model also nicely fit the multiplicities , @xmath17 spectra and differential elliptic flow of @xmath12 , @xmath4 , @xmath5 , @xmath6 , @xmath0 , and @xmath1 at various centrality bins . with the good descriptions of these related soft hadron data , it is meaningful to further investigate the chemical and thermal freeze - out procedures of various hadrons at the lhc . for @xmath12 , @xmath4 , @xmath5 , @xmath6 , @xmath0 , and @xmath1 during the urqmd expansion of vishnu . here ,
@xmath18 and @xmath19 denote the particle yield density at mid - rapidity at later evolution time and at the starting time , respectively .
right panels ( g ) and ( h ) : time evolution of the changing rates for the corresponding particle yield densities .
solid / dashed lines denote the vishnu simulations with / without @xmath20-@xmath14 annihilations.,scaledwidth=98.0%,height=340 ] figure [ fig : timedndy ] shows the time evolution for the particle yields of @xmath12 , @xmath4 , @xmath5 , @xmath6 , @xmath0 and @xmath1 during the hadronic evolution , simulated from vishnu hybrid model with and without @xmath20-@xmath14 annihilations . for the simulations without @xmath20-@xmath14 annihilations , the yields of these two multi - strange baryons @xmath0 and @xmath1 almost do not change . we also notice that their yields quickly reach saturation for the case with @xmath20-@xmath14 annihilations .
this reveals that @xmath0 and @xmath1 experience early chemical freeze - out after the hadronazition .
in contrast , the yields of @xmath5 slowly decreases during the hadronic evolution especially for the case with @xmath20-@xmath14 annihilations , indicating later chemical freeze - out of protons . on the production of @xmath6 and @xmath4 , the @xmath20-@xmath14 annihilations almost balance with other inelastic collision channels . as a result ,
the yields of @xmath6 and @xmath4 only slightly change during the hadronic evolution with @xmath20-@xmath14 annihilations @xcite .
the right panels of fig . 1 presents the changing rates of the particle yields of these identified hadrons . for the simulations without @xmath20-@xmath14 annihilations , the changing rates for @xmath4 , @xmath5 , and @xmath6 show wide peaks along the time axis , illustrating that the associated inelastic collisions are still frequent after 10 - 20 @xmath21 . due to small hadronic cross sections , multi - strange hadrons @xmath0 and @xmath1 rarely interact with other hadrons . while , the annihilations with their own anti - particles lead to wide peaks of the changing rates for @xmath0 and @xmath1 in panel ( h ) , which delay their chemical freeze - out , when compared with the case without @xmath20-@xmath14 annihilations .
however , these multi - strange hadrons still experience earlier chemical freeze - out in the full vishnu simulations compared with other identified hadrons @xcite .
figure [ fig : wbb05 ] presents the thermal freeze - out time distributions of @xmath12 , @xmath4 , @xmath5 , @xmath6 , @xmath0 and @xmath1 in the most central pb+pb collisions . here
, the thermal freeze - out time distributions are extracted through analyzing the space - time information of the last collisions for various hadron species . for @xmath0 and @xmath1 ,
the peaks of their thermal freeze - out time distributions are located around 10 @xmath21 . while the peaks of @xmath5 and @xmath6 distributions are around 20 - 30 @xmath21 .
due to a large amount of hadronic scatterings during the late urqmd evolution , the thermal freeze - out time distributions of @xmath12 and @xmath4 widely spread along the time axis .
3 thus illustrates that thermal freeze - out procedures are hadron species dependent .
compared with other hadrons , the two multi - strange hadrons @xmath0 and @xmath1 experience early thermal freeze - out due to their much smaller hadronic cross sections @xcite .
in this proceeding , we briefly review our recent investigations on the chemical and thermal freeze - out of various hadron species in 2.76 a tev pb+pb collisions within the framework of vishnu hybrid model .
we found that the chemical and thermal freeze - out procedures are hadron - species dependent . due to their much smaller hadronic cross sections , the two multi - strange hadrons @xmath0 and @xmath1 experience earlier chemical and thermal freeze - out than other hadrons . to extract the possible effective chemical freeze - out temperatures of various hadron species
requires detailed analysis of the space - time distributions of the last inelastic collisions , which we would like to leave it to future study .
this work was supported by the nsfc and the most under grants no .
11435001 and no . 2015cb856900 and the china postdoctoral science foundation under grant no .
we gratefully acknowledge extensive computing resources provided to us on tianhe-1a by the national supercomputing center in tianjin , china .
, k , p , @xmath6 , @xmath0 and @xmath1 in the most central pb+pb collisions , calculated from vishnu with @xmath20-@xmath14 annihilations.,scaledwidth=80.0%,height=264 ]
9 p. braun - munzinger , d. magestro , k. redlich , and j. stachel 2001 _ phys .
_ b * 518 * 41 a. andronic , p. braun - munzinger and j. stachel 2006 _ nucl .
phys . _ a * 772 * 167 - 199 f. becattini , j. manninen and m. gazdzicki 2006 _ phys .
_ c * 73 * 044905 p. f. kolb , j. sollfrank and u. w. heinz , phys .
c * 62 * ( 2000 ) 054909 h. song and u. heinz 2008 _ phys .
_ b * 658 * 279 ; 2008 _ phys .
_ c * 77 * 064901 ; 2008 _ phys .
_ c * 78 * 024902 ; h. song 2009 _ thesis proceeding _
( _ preprint _
arxiv:0908.3656 [ nucl - th ] ) s. a. bass _ et al .
_ 1998 _ prog . part .
phys . _ * 41 * 255 ; m. bleicher _ et al .
_ 1999 _ j. phys .
_ g * 25 * 1859 p. huovinen and p. petreczky 2010 _ nucl .
phys . _ a * 837 * 26 - 53 h. song , s. a. bass and u. heinz 2011 _ phys .
_ c * 83 * 054912 h. song , s. bass and u. w. heinz 2014 _ phys .
_ c * 89 * 034919 h. j. drescher and y. nara 2007 _ phys .
_ c * 75 * 034905 b. abelev _ et al . _
( alice collaboration ) 2013 _ phys .
_ c * 88 * 044910 b. b. abelev _ et al . _
( alice collaboration ) 2013 _ phys .
lett . _ * 111 * 222301 b. b. abelev _ et al . _
( alice collaboration ) 2014 _ phys .
_ b * 728 * 216 | this proceeding briefly summarizes our recent vishnu hybrid model investigations on the chemical and thermal freeze - out of various hadrons species in 2.76 a tev pb+pb collisions .
detailed analysis on the evolution of particle yields and the last elastic collisions distributions during the hadronic evolution reveals that the two multi - strange hadrons , @xmath0 and @xmath1 , experience early chemical and thermal freeze - out when compared with other hadron species . |
recent x - ray studies of seyfert 2 galaxies with and have shown that the 0.110 kev spectra of these sources are rich in emission lines at both soft and hard energies ( e.g. , guainazzi et al . 1999 ; turner et al .
the interpretation of the emission lines is problematic because of ambiguities about line blending , line profiles , and line flux distribution that are all poorly constrained by and .
the data are consistent with emission from gas in photoionization equilibrium ( e.g. , netzer , turner , & george 1998 and references therein ) , but there are also attempts to fit the spectra by a two - temperature gas in collisional equilibrium ( ueno et al . 1994 ) , presumably due to starburst emission .
x - ray observations at high resolution both spatially and spectrally are crucial to determining the origin of the x - ray lines in seyfert 2s , a task for which is uniquely suited . here
we present a 60 ks hetgs spectrum of the seyfert 2 galaxy .
previous and observations of this source revealed several emission lines at both soft and hard energies ( sako et al . 2000a ; guainazzi et al .
1999 ; matt et al .
1996 ) . in a companion paper , focusing on the zeroth - order acis image , we established that several components contribute to the x - ray emission from . in particular
, we found that @xmath2 60% of the x - ray flux at 2 kev is due to an extended component on scales @xmath2 2.3 , while at harder energies the contribution from a compact ( 0.8 ) region prevails .
importantly , the acis spectrum of the latter component exhibits several emission lines including a prominent ( ew @xmath2 2.5 kev ) line . in this paper , we concentrate on the analysis of the hetgs spectrum and on the implications for the physical conditions of the emitting gas .
we also use a simultaneous observation to derive useful constraints on the intrinsic nuclear x - ray continuum . at the distance of the galaxy ( @xmath2 4 mpc ) , 1=19 pc .
was observed with the high energy transmission grating spectrometer ( hetgs ; canizares et al .
2000 , in prep . ) on 2000 june 6 , with acis - s ( garmire et al . 2000 ) in the focal plane .
the total net exposure was 60,223 s. details on the observation are given in our companion paper .
the hetgs carries two mirror assemblies , the high energy grating ( heg ) and medium energy grating ( meg ) .
the nuclear heg and meg spectra were extracted in a narrow ( 15 pixel ) rectangular region centered on the zeroth - order position , avoiding contamination from the dispersed spectra of the nearby serendipitous sources .
although a few of the serendipitous sources exhibit emission lines in their acis spectra at both soft and hard x - rays , the 0.58 kev flux of the brightest one is a factor 5 weaker than the nucleus .
we thus believe that contamination to the hetgs nuclear spectrum ( in the regions of overlap of the dispersed spectra ) is negligible . the heg and meg spectra were gain corrected and flux calibrated using ancillary response files generated with the ` ciao ` software .
only the first order heg and meg spectra were used for the analysis , as higher orders contain only a few ( 3 per bin ) counts and are not useful .
the spectra were also corrected for cosmological redshift and galactic absorption , @xmath3 ( freeman et al . 1977 ) .
we used a simultaneous 30 ks exposure with to constrain the higher - energy x - ray continuum emission from .
the data were reduced following standard criteria ; here we use only data from the pca detector and report only on the results most relevant for the modeling of the hetgs data , leaving more details to a future paper .
the source = 10.5 cm -0.8 cm was detected with the pca up to @xmath2 30 kev , with a 230 kev count rate of 7.26 @xmath4 0.04 counts s@xmath5 .
we fitted the pca data in the energy range 630 kev , where relatively few lines are present ( mainly at 6.4 kev , fek@xmath6 at 7.1 kev , and /nik@xmath1 at 7.9 kev ; guainazzi et al .
we find that an excellent description of the pca data ( @xmath7=56 for 62 degrees of freedom ) is obtained with a model including a `` pure '' reflection continuum from neutral gas , plus a heavily absorbed ( @xmath8 ) power law dominant at 10 kev , plus the three gaussians lines .
the best - fit model and fitted parameters are in complete agreement with a previous observation of ( matt et al .
1999 , guainazzi et al .
we mainly stress here the results for the intrinsic nuclear continuum : power - law photon index @xmath9 ( 90% confidence errors ) and intrinsic ( absorption - corrected ) 210 kev luminosity in the range @xmath10 , in agreement with the data ( matt et al .
the first order meg spectra agree well with each other within the resolution of the grating ( @xmath2 0.023 , twice as for the heg ) .
the two heg spectra also agree with each other and with the meg spectra .
therefore , in order to increase the signal - to - noise ratio , the four spectra were averaged after rebinning the heg data to the meg resolution .
while this procedure sacrifices the higher resolution of the heg , in most cases the lines are unresolved and the loss of resolution is thus well compensated by a cleaner detection of the lines .
figure [ model - data ] shows a wide wavelength coverage view of the flux - calibrated spectrum , binned at 0.02 .
a plethora of emission lines are apparent in the spectrum at all energies .
the most prominent one is the line at 6.4 kev , with an ew @xmath11 kev , an unresolved core , and a hint of a broader - base component which will be discussed elsewhere .
a detailed view of the = 10.5 cm -0.8 cm fe line complex ( figure [ feline ] ) shows also the fe k@xmath6line at @xmath2 7.1 kev , with an intensity of @xmath12 that of the line .
we have identified another feature at around 6.6 kev , consistent with the k@xmath1 line . as shown below
, our model can not explain the intensity of this feature .
the high absorption column in the direction to , and the relatively faint x - ray flux of the source , do not allow any clear line detections below @xmath13 kev .
table 1 gives quantitative information about the detected emission lines , including the line identification ( note some question marks due to uncertain identifications ) , measured fluxes , and ews .
many h - like and he - like lines of mg , si , s and possibly ar are detected in the spectrum , together with weaker `` neutral '' lines of si , s , ar , and perhaps ca .
the high resolution allows deblending of the forbidden , intercombination , and resonance lines of several of the he - like ions ( figure [ model - data ] ) .
the significance of the lines detections can be judged from the corresponding uncertainties on the line fluxes .
the poor signal - to - noise ratio did not allow a meaningful line - width determination , and all fwhms are consistent with the instrumental resolution .
the ews are calculated with respect to the total observed continuum , which was evaluated using two independent methods .
first , we selected line - free regions of the spectrum ( mostly above 23 kev ) , rebinned the data heavily , and fitted them with a smooth curve .
second , we used the zeroth - order spectrum from an extraction radius consistent with the extraction width of the hetgs spectrum ( @xmath2 3 ) .
the latter was fitted with a power law plus free @xmath14 , plus narrow ( width=0.05 kev ) gaussians representing all the emission lines detected in the hetgs spectrum .
both methods gave consistent results .
the 3 - 7 kev continuum can be described by an inverted power law with photon index @xmath15 and flux f@xmath16 .
this continuum was used in the ew measurements .
the uncertainty on the continuum flux is 22% or better at all wavelengths .
we modeled the observed spectrum using our new line measurements , as well as the information obtained from the radial flux distribution ( from our companion paper ) and the high en- [ cols="<,^,^,^,^,^ " , ] ergy observations from ( matt et al .
1999 ) and .
the modeling is based on the following observations : ( a ) the 210 continuum is flat , in @xmath17 , at long wavelengths ( as expected from an `` ionized mirror '' ) and hardens below about 5 ( as expected from a `` neutral mirror '' ) .
this , plus the detection of low ionization species , suggests two distinct components ; ( b ) the ew of the line is very large , indicating iron over abundance ( e.g. netzer et al .
1998 ) ; ( c ) the extended spectrum ( outside of the central 0.8 ) is flatter than the central spectrum , suggesting that the more neutral component contributes less at larger radii .
modeling is done using ion00 , the 2000 version of the photoionization code ion ( netzer 1996 ) .
this includes the computation of the steady state ionization and thermal structure of the gas , and the emergent spectrum .
the main ingredients of the model are : ( a ) central power - law continuum with @xmath18 , extending from 0.1100 kev and normalized to produce @xmath19 , in agreement with the and observations ; ( b ) two component gas with twice - solar metallicity .
we have experimented with the density and column density of the two and found the following satisfactory combination : 1 . an _
ionized component _ with a radial density distribution of @xmath20 where @xmath21 pc and @xmath22 @xmath23 .
this distribution is consistent with our measured radial flux distribution assuming most of the 0.83 flux is due to scattered continuum .
2 . a _ neutral component _ with a similar ( yet unconstrained ) radial distribution with the same @xmath24 , but with @xmath25 @xmath23 .
the first component is allowed to extend all the way to 300 pc while the second is limited by its column density , arbitrarily chosen at @xmath26 @xmath27 , and is thus terminated well inside the inner 8 pc .
this component can perhaps be viewed as the wall of the inner torus .
the covering factors of the two components are free parameters of the model ; ( c ) the gas turbulent velocity can range from no turbulence ( pure thermal motion ) up to @xmath28 km s@xmath5 .
the increased line width results in an increased intensity of all resonance lines due to continuum fluorescence ( krolik and kriss 1995 ; netzer 1996 ) .
our best model requires no turbulent motion .
several models have been computed , with various covering factors .
a satisfactory solution is found for @xmath29 = 0.40.5 , and @xmath30 ( ionized ) = 0.10.2 , consistent with the expected opening angle of the ( hypothetical ) torus , where @xmath31 is the solid angle subtended by the gas to the illuminating source .
figure [ model - data ] shows the two components alongside with the observations and figure [ ratio ] shows observed over computed intensities for the strongest lines .
note that , given the gas density and location , the only free parameters are the covering factors of the two components .
the overall agreement between the model and the observations is good , given the uncertainties .
in particular : ( a ) the observed fluxes of most emission lines are reproduced , to within a factor of two .
the 1.511 continuum shape is reproduced too .
these results support the idea of the two component model , = 10.5 cm with the assumed levels of ionization and metallicity ; ( b ) the covering fractions are consistent with the suggested source geometry .
the radial density distribution reproduces well the highly peaked emission of this source and the relative weak flux outside the central 16 pc . the calculated scattered continuum and extended emission lines in the inner 1657 pc are in good agreement with the ( highly uncertain ,
see our companion paper ) observations and there is no need to assume an additional starburst source ; ( c ) the `` neutral '' iron k@xmath1 and k@xmath6 lines are consistent with the compact , neutral component gas and the assumed metallicity .
notable difficulties are the over - prediction of the fe l - shell lines around 912 and the under - prediction of the 6.6 kev feature if due to .
all argon lines are also under - predicted by the model
. the measured intensity of these lines are highly uncertain and any suggestion for their origin ( e.g. unusual composition ) must await better observations .
we also note that the predicted k@xmath1 flux , outside the central 20 pc , is below the observed value
. this may be due to the already noted uncertainty in flux measurement .
we also note that a `` typical '' nlr , can produce a sizeable fraction of this k@xmath1 emission . regarding earlier x - ray observations of this source , the discovery paper by matt et al .
( 1996 ) reports an spectrum including both neutral and ionized species .
netzer et al .
( 1998 ) re - analyzed the data and reported the measurements of six ionized lines plus the iron k - lines , all in reasonable agreement with the present observations .
the paper includes a detailed photoionization model of the source and addresses also the k@xmath1/h@xmath6 line ratio .
guainazzi et al .
( 1999 ) reported on data that include the measurement of , , and lines , as well as neutral fe - k lines .
the observed fluxes are in good agreement except for whose intensity exceeds our estimate by a factor 4 . the present observations , being far superior in terms of the spatial and spectral resolution , yet limited in signal - to - noise ratio , confirm several of the suspected features of this source , such as the gas location , metallicity , and the various components .
this , and the recently published observations of mkn 3 ( sako et al .
2000b ) show , for the first time , that centrally illuminated ionized gas in seyfert 2 galaxies can have very different distributions in different sources .
while in mkn 3 most of the line emission is spread over several hundred parsecs , this in not the case in , where the ionized gas is highly concentrated near the center and most of its flux originates within the central @xmath32 pc .
moreover , the new observations show that the more neutral gas is even more concentrated , and its dimension may be as small as a few parsecs . while we do not want to speculate about the origin of this difference , we note the large difference in luminosity between these two sources .
it is therefore possible that the highly ionized gas in seyfert galaxies has dimensions that are regulated by the central source s x - ray luminosity .
the new observations of the seyfert 2 galaxy enable the very first detailed analysis of the physical conditions and the gas distribution in the inner 200 pc of this source .
our observations show the emission to be highly concentrated within the inner 60 pc and suggest that emission within this volume is entirely due to the reprocessing of the obscured central source s radiation .
an even smaller , more neutral component is seen through emission of low ionization iron lines and hard reflected continuum .
this is the first determination of the ionized gas distribution in the inner 100 pc region of a seyfert 2 galaxy . | results from a 60 ks hetgs observation of the nearby seyfert 2 are presented .
the spectrum shows a wealth of emission lines at both soft and hard x - rays , including lines of ne , mg , si , s , ar , ca , and fe , and a prominent line at 6.4 kev .
we identify several of the he - like components and measure several of the lyman lines of the h - like ions .
the lines profiles are unresolved at the limited signal - to - noise ratio of the data . our analysis of the zeroth - order image in a companion paper constrains the size of the emission region to be 2060 pc , suggesting that emission within this volume is almost entirely due to the reprocessing of the obscured central source . here we show that a model containing two distinct components can reproduce almost all the observed properties of this gas .
the ionized component can explain the observed intensities of the ionized species , assuming twice - solar composition and an @xmath0 density distribution .
the neutral component is highly concentrated , well within the 0.8 point source , and is responsible for almost all of the observed k@xmath1 ( 6.4 kev ) emission
. seems to be different than mkn 3 in terms of its gas distribution . |
the central idea of algebraic quantum field theory is the representation of observables by a net of local @xmath0-algebras associated to bounded regions of a spacetime ( see ( haag , 1992 ) ) .
this correspondence is established _ via _ the standard axioms of the theory such as isotony , locality ( also called einstein causality or microcausality ) and covariance .
a state @xmath1 in such a local quantum theory is defined as a normalized positive linear functional on the quasilocal observable algebra @xmath2 which is an inductive limit of local observable algebras .
the representation @xmath3 corresponding to the state @xmath1 converts the net of @xmath0-algebras into a net of von neumann observable algebras by closures in the weak topology .
since von neumann algebras are rich in projections , they offer a nice representation of _ quantum events _ : projections can be interpreted as 0 - 1valued observables and their expectation value defines the probability of the event that the observable takes on the value 1 in the appropriate quantum state .
although due to the axiom of locality two projections @xmath4 and @xmath5 commute if they are contained in local algebras supported in spacelike separated regions , they can still _ be correlating _ in a state @xmath1 , that is @xmath6 in general .
due to their spacelike separated supports the correlation between these events is said to be _
superluminal_. since spacelike separation excludes direct causal influence , one may look for a causal explanation of these superluminal correlations in terms of _ common causes_. the first probabilistic definition of the common cause is due to hans reichenbach ( 1956 ) .
reichenbach s definition has been first generalized to algebraic quantum field theory by rdei ( 1997 , 1998 ) and the question was posed whether superluminal correlations can be given a common causal explanation . as a positive answer , rdei and summers ( 2002 , 2007 )
have shown that in the von neumann algebraic setting with type iii local algebras there always exists a common cause commuting with the correlating projections in the union of their causal past called the _ weak past_. hofer - szab and vecsernys ( 2012a ) have shown , however , that this is not a general fact in _ all _ local quantum theories : there exist correlations in algebraic quantum field theory with locally _
finite _ degrees of freedom for which no commutative common causal explanation can be given . finally , hofer - szab and vecsernys ( 2012b ) have shown that by allowing common causes that do _ not _ commute with the correlating events the noncommutative version of the result of rdei and summers can be regained . in the same paper
it also was argued why demanding commutativity between the common causes and the correlating events is unjustified and should be given up .
this paper is a further step into the investigation of noncommuting common causes : we will apply them for the common causal explanation of not _
one _ but _ several _ superluminal correlations together violating the clauser
horne inequality . to be more specific ,
consider four events @xmath7 , @xmath8 , @xmath9 and @xmath10 such that @xmath11 and @xmath12 ( @xmath13 ) where @xmath14 and @xmath15 are von neumann algebras supported in the spacelike separated regions @xmath16 and @xmath17 , respectively .
suppose that the four pairs @xmath18 are correlating in a state @xmath1 that is @xmath19 suppose furthermore that they violate the clauser
horne ( ch ) inequality ( clauser and horne , 1974 ) defined as follows : @xmath20 sometimes in the epr - bell literature another bell - type inequality is used instead of ( [ ch_int ] ) : the clauser horne shimony
holte ( chsh ) inequality ( clauser , horne , shimony and holt , 1969 ) defined in the following way : @xmath21 where @xmath22 and @xmath23 are self - adjoint _ contractions _ ( that is @xmath24 for @xmath13 ) located in @xmath14 and @xmath15 , respectively .
it is easy to show , however , that in a given state @xmath1 the set @xmath25 violates the ch inequality ( [ ch_int ] ) _ if and only if _ the set @xmath26 of self - adjoint contractions given by @xmath27 violates the chsh inequality ( [ chsh_int ] ) .
therefore we will concentrate only on the ch - type bell inequalities .
a number of important results in ( summers , werner 1988 ) , ( halvorson , clifton 2000 ) prove that poincar covariant algebraic field theories abound in normal states establishing superluminal correlations between spacelike separated projections / contractions violating the ch / chsh inequality .
here we produce such states in algebraic field theories with locally finite degrees of freedom , namely in local uhf - type quantum theories ( hofer - szab and vecsernys , 2012b ) . in this case the events @xmath28 and @xmath29 will be elements of such commuting local algebras @xmath14 and @xmath15 with spacelike separated supports @xmath16 and @xmath17 that are isomorphic to simple matrix algebras @xmath30 and @xmath31 .
moreover , they generate a tensor product algebra : @xmath32 .
thus the meaning of the mentioned inequalities can be enlightened in terms of quantum information theory ( see for example ( horodeczki et al . , 2009 ) ) : the existence of a correlation ( [ correlation ] ) simply means that @xmath1 is not a product state on @xmath33 .
correlations can be obtained from separable states , i.e. from convex combination of product states . since a simple matrix algebra admits a unique normalized trace , states can be uniquely given by density matrices .
hence , product states are characterized by tensor product density matrices and separable states are characterized by separable density matrices , i.e. by convex combinations of tensor product density matrices . since in the real linear space of selfadjoint elements in @xmath33 positive elements @xmath34 form a self - dual cone with respect to the scalar product @xmath35 , the ( convex ) set of states , that is the ( convex ) set of density matrices is even more rich : there also exist non - separable that is entangled states on @xmath33 .
this is due to the existence of positive elements in @xmath33 that can not be written as positive linear combinations of tensor product of positive elements in @xmath30 and @xmath31 .
entangled states lead to inherent and numerically stronger correlations among events contained in different tensor factors of @xmath33 .
separable and entangled states can be distinguished by witness operators @xmath36 which are the non - positive elements in the dual cone @xmath37 of the cone @xmath38 spanned by separable density matrices .
clearly , a witness operator @xmath39 gives rise to inequalities since @xmath40 for any separable state @xmath1 on @xmath33 .
basically , these properties are behind the inequalities for superluminal correlations discussed above . coming back to the superluminal correlations ( [ corrs ] ) violating the ch inequality ( [ ch_int])/chsh inequality ( [ chsh_int ] )
, it is well known that the violation of these inequalities leads to various no - go theorems excuding a local common causal explanation of the set ( see for example ( shimony 2004 ) and the literature therein ) .
however , all these common causal explanations assume that the common cause is _ commuting _ with the correlating events . hence , a natural question arises : does the noncommutative generalization of the common causes in algebraic quantum field theory give enough freedom to find _ noncommuting _ common causes for a set of correlations violating the ch / chsh inequality ? in this paper
we will answer this question in the affirmative by showing that the violation of the ch / chsh inequality does _ not _ exclude the existence of a common causal explanation of the set of correlations in question _ if _ we abandon commutativity between the common cause and the outcome events .
moreover , in the given example the noncommuting common cause turns out to be local and supported in the common past i.e. in the intersection of the causal pasts of the correlating events . in section 2
we reproduce the epr - bohm scenario in the local quantum ising model , which is a prototype of local uhf - type quantum theories . giving a faithful state on the corresponding observable algebra
we show that there is a set @xmath25 of projections having such correlations that maximally violates the ch inequality . in section 3
we introduce the main concepts of a common causal explanation of superluminal correlations and we explicitly construct a common cause of the above four correlations .
we conclude in section 4 .
consider the net of ` intervals ' @xmath41 of half - integers .
the set of half - integers can be interpreted as the space coordinates of the center @xmath42 and @xmath43 of minimal double cones @xmath44 of unit diameter on a thickened cauchy surface in two dimensional minkowski space @xmath45 .
( see fig .
[ strong_bell1 ] . )
an interval @xmath46 can be interpreted as the smallest double cone @xmath47 containing both @xmath48 and @xmath49 .
they determine a directed subset @xmath50 of double cones in @xmath51 , which is left invariant by the group of space - translations with integer values .
the net of local algebras is defined as follows .
the ` one - point ' observable algebras @xmath52 associated to the minimal double cone @xmath53 are defined to be isomorphic to the direct sum matrix algebra @xmath54 .
let @xmath55 denote the selfadjoint unitary generator of the algebra @xmath56
. between the generators one demands the following commutation relations : @xmath57 now , the local algebras @xmath58 are defined to be the linear span of the monoms @xmath59 where @xmath60 .
specifically , the two - point algebra @xmath61 is spanned by the selfadjoint unitary monoms @xmath62 ( where @xmath63 is the imaginary unit ) .
they satisfy the same commutation relations like the pauli matrices @xmath64 and @xmath65 in @xmath66 .
the minimal projections in @xmath67 can be parametrized by a unit vector @xmath68 in @xmath69 as follows : @xmath70 and can be interpreted as the event localized in @xmath71 pertaining to the generalized spin measurement in direction @xmath72 .
similarly , the three - point algebra @xmath73 is linearly spanned by the selfadjount unitary monoms @xmath74 and any minimal projection in @xmath75 can be written in the form @xmath76 where @xmath72 is again a unit vector in @xmath69 .
since the local algebras @xmath77 for @xmath78 are isomorphic to the full matrix algebra @xmath79 the quasilocal observable algebra @xmath2 is a uniformly hyperfinite ( uhf ) @xmath0-algebra , which possesses a unique ( non - degenerate ) normalized trace @xmath80 .
now , consider the subset @xmath81 of double cones in @xmath51 that are spanned by minimal double cones being integer time translates of those in @xmath50 .
the directed set @xmath82 is left invariant by integer space and time translations . in order to extend the
cauchy surface net
@xmath83 to the net @xmath84 in a causal and time translation covariant manner one has to classify causal ( integer valued ) time evolutions in the local quantum ising model .
this classification was given in ( mller , vecsernys ) and it also was shown that the extended net satisfies isotony , einstein causality , algebraic haag duality @xmath85 and primitive causality : @xmath86 here @xmath87 is a finite connected piece of a thickened cauchy surface ( composed of minimal double cones ) and @xmath88 denotes the double spacelike complement of @xmath87 , which is the smallest double cone in @xmath81 containing @xmath87 .
we note that @xmath89 is defined to be the @xmath0-algebra in @xmath2 generated by the algebras @xmath90 .
] moreover , the commuting ( unit ) time and ( unit ) space translation automorphisms @xmath91 and @xmath92 of the quasilocal algebra @xmath2 act covariantly on the local algebras , i.e. @xmath84 is a @xmath93-covariant local quantum theory .
the causal time translation automorphisms @xmath91 of @xmath2 can be parametrized by @xmath94 with @xmath95 and @xmath96 and they are given on the algebraic generator set @xmath97 of @xmath98 .
the automorphisms @xmath99 of @xmath98 corresponding to causal time translations by a unit read as @xmath100 where @xmath101 .
it was also shown ( see mller , vecsernys ) that the following algebra isomorphisms hold : if @xmath102 is a double cone containing @xmath103 and @xmath104 minimal double cones in the right forward and left forward lightlike directions , respectively , then @xmath105 , the linear dimension of the corresponding local algebra is @xmath106 and @xmath107 after this general introduction let us specify our model .
let @xmath108 denote the minimal double cone with time @xmath109 and space @xmath63 coordinates of its center .
consider the double cones @xmath110 and @xmath111 and the two - point algebras @xmath112 and @xmath113 pertaining to them . since they do
_ not _ lie in the original cauchy surface ( see fig .
[ strong_bell2 ] ) , where the generators @xmath114 are supported , we have to use a causal time evolution @xmath91 to give elements of them .
let @xmath115 and @xmath116 be minimal projections in @xmath112 and in @xmath113 , respectively , given by @xmath117 where @xmath118 and @xmath119 are four ( @xmath120 ) unit vectors in @xmath69 .
the state @xmath121 on @xmath2 will be given by a density matrix @xmath122 using the normalized trace on @xmath2 : @xmath123 .
the density matrix @xmath122 is chosen in @xmath124 to describe the usual ` singlet state ' by restricting @xmath121 to this subalgebra of @xmath2 .
that is in terms of matrix units @xmath125 : @xmath126 in terms of local generators the matrix units are given as @xmath127 hence , substituting ( [ p1])-([p4 ] ) in ( [ rho1 ] ) one gets @xmath128 since @xmath122 is a scalar multiple of a minimal projection in @xmath129 , the state @xmath121 is not locally faithful . to get
a locally faithful state one can use the convex combination @xmath130 which gives back the singlet state for @xmath131 .
the correlation between @xmath132 and @xmath133 in the state @xmath134 is defined as the deviation from the product state value : @xmath135 using ( [ a_m])-([b_n ] ) and ( [ rho3 ] ) this correlation reads as @xmath136 where @xmath137 is the scalar product in @xmath69 .
therefore @xmath132 and @xmath133 will correlate whenever @xmath138 and @xmath139 are not orthogonal .
now , let @xmath138 and @xmath139 be chosen as follows : @xmath140 for this setting the clauser
horne inequality ( [ ch_int ] ) will be violated at the lower bound since @xmath141 which is smaller than @xmath142 if @xmath143 . or , equivalently , the clauser
horne shimony
holte inequality ( [ chsh_int ] ) ( with @xmath144 and @xmath145 defined in ( [ u_a])-([u_b ] ) ) will be violated for @xmath143 since then @xmath146 both the ch and the chsh inequality are maximally violated if @xmath131 that is for the ` singlet state ' , which is a maximally entangled state for @xmath147 .
then the question arises : does the above set @xmath25 violating the ch inequality have a common causal explanation ? to answer this question we have to characterize what a common causal explanation consists in .
as mentioned in the introduction , the first characterization of the common cause in terms of a classical probability measure space is due to reichenbach ( 1956 ) .
although reichenbach s original defintion is inappropriate to explain correlations in algebraic quantum field theory for several reasons ( it is classical , it does not take into account the spacetime localization of the events , etc .
; for details see ( hofer - szab and vecsernys , 2012a ) ) , one can easily generalize the notion of the common cause for the non - classical case . here
we repeat only the definition given in ( hofer - szab and vecsernys , 2012a ) because it also embraces the commuting common causes defined in ( rdei 1997 , 1998 ) .
let @xmath148 be the ( non - distributive ) lattice of projections ( events ) in a von neumann algebra @xmath149 and let @xmath150 be a state on it .
a set of mutually orthogonal projections @xmath151 is called a _ partition of the unit _
@xmath152 if @xmath153 . such a partition defines a _ conditional expectation _
@xmath154 that is @xmath155 is a unit preserving positive surjection onto the unital @xmath0-subalgebra @xmath156 obeying the bimodule property @xmath157 .
we note that @xmath158 contains exactly those elements of @xmath149 that commute with @xmath159 .
since @xmath160 is also a state on @xmath161 we can give the following [ ncqccs ] a partition of the unit @xmath162 is said to be a _ common cause system _ of the commuting events @xmath163 , which correlate in the state @xmath150 , if @xmath164 for @xmath165 with @xmath166 . if @xmath167 commutes with both @xmath4 and @xmath5 for all @xmath165 we call @xmath168 a _ commuting _ common cause system , otherwise a _ noncommuting _ one . a common cause system of size @xmath169
is called a _ common cause_.
some remarks are in place here .
first , in case of a commuting common cause system @xmath160 can be replaced by @xmath1 in ( [ ncqccs1 ] ) since @xmath170 .
second , using the decompositions of the unit , @xmath171 , ( [ ncqccs1 ] ) can be rewritten in an equivalent form : @xmath172 one can even allow here the case @xmath173 since then both sides of ( [ ncqccsrew ] ) are zero .
we also have to specify the spacetime localization of the possible common causes of the correlations .
one can define different pasts of the bounded contractible regions @xmath174 and @xmath175 in a spacetime @xmath176 as : @xmath177 where @xmath178 denotes the union of the backward light cones @xmath179 of every point @xmath180 in @xmath87 ( rdei , summers 2007 ) . clearly , @xmath181 . with these different localizations of the common cause in hand
now we can define various common cause systems in local quantum theories according to ( i ) whether commutativity is required and ( ii ) where the common cause system is localized .
thus we can speak about _ commuting / noncommuting ( weak / strong ) common cause systems_. since in our model of the previous section we have not one pair of correlating events but four , we need also the notion of the joint common cause system : [ lcccs ] let @xmath182 be the observable net in a local quantum theory and let @xmath183 and @xmath184 be finite sets of projections in the algebras @xmath185 and @xmath186 , respectively , supported in spacelike separated regions @xmath187 .
suppose that all pair of spacelike separated projections @xmath188 correlate in a state @xmath1 of @xmath2 .
then the set of correlations is said to possess a commuting / noncommuting ( weak / strong ) _ joint _ common cause system if there exists _ a single _ commuting / noncommuting ( weak / strong ) common cause system for _ all _ correlations @xmath189 .
having defined what a common causal explanation means we answer the question raised at the end of the previous section in the affirmative : we explicitly contruct a noncommuting joint common cause for the set @xmath25 of correlations violating the ch inequalities ( [ ch_int ] ) .
first , observe that without specifying the dynamics we only can talk about _ weak _ ( commuting or noncommuting ) common cause or joint common cause .
the reason for this is the following . to be able to talk about the correlation between @xmath132 and @xmath133 we need to consider a local algebra containing both @xmath132 and @xmath133 . to be able to talk about a common cause system @xmath190 of this correlation
we need a local algebra that contains @xmath190 too .
let us consider the piece @xmath191 of a cauchy surface contained in the _ weak _ past of @xmath192 and @xmath193 , which is the shadowed region in fig .
[ strong_bell3 ] , that is the support of a hypothetical weak common cause system @xmath190 of the correlation . due to primitive causality
the algebra corresponding to @xmath191 is equal to the algebra @xmath194 , where @xmath195 is the smallest double cone containing @xmath191 .
since @xmath196 contains both @xmath192 and @xmath193 isotony implies that @xmath197 for any causal dynamics . hence , the hypothetical weak common cause system @xmath190 and the events @xmath198 are in the common algebra @xmath194 . in ( hofer - szab , vecsernys 2012a )
it was shown that a weak noncommuting common cause can be found for a _ single _ correlation @xmath199 in local uhf - type quantum theories without the explicit knowledge of the dynamics .
on the other hand , if @xmath200 is the piece of a cauchy surface contained in the _ common _ past of @xmath192 and @xmath193 then neither @xmath192 nor @xmath193 are contained in @xmath201 .
therefore @xmath129 and @xmath202 are not related by isotony , and one should find a proper cauchy surface piece @xmath200 in the common past and an integer time shift @xmath203 such that @xmath204 holds .
( see fig .
[ strong_bell3b ] . )
then the corresponding algebras are related by the dynamics @xmath205 where the last equality is due to time translation covariance .
thus for finding a joint common cause in the local quantum ising model one has to specify the dynamics .
let us therefore choose a special set of dynamics by fixing two parameters of the causal dynamics : @xmath206 . by this choice the time translation automorphism ( [ causal_automorph2 ] ) reads as follows : @xmath207 for every @xmath101 .
this specification of the dynamics converts events @xmath132 and @xmath133 in ( [ a_m])-([b_n ] ) into @xmath208 and the density matrix ( [ rho3 ] ) into @xmath209 since @xmath210 is in the common past of @xmath192 and @xmath193 , it would be a possible choice for the support of a hypothetical joint common cause system .
however , as it will turn out soon , using this dynamics we will be able to localize a ( noncommuting ) common cause @xmath211 of the correlation in a smaller spacetime region within @xmath212 , namely in @xmath213 .
( see fig .
[ strong_bell4 ] . )
the common cause @xmath214 localized in @xmath215 will be the common cause system of all the four correlations @xmath25 , i.e. @xmath214 serves as a joint common cause . to see the concrete form of @xmath216 recall that any non - central projection of rank 2 in @xmath217 can be written as @xmath218 where @xmath219 and @xmath220 are unit vectors in @xmath69 . with these notations in hand
we have the following [ nccccs ] let @xmath221 be four projections defined in ( [ a_m])-([b_n ] ) where @xmath138 and @xmath139 are non - orthogonal unit vectors in @xmath69 establishing four correlations @xmath25 in the state defined by the density matrix in ( [ rho4 ] ) .
then @xmath216 given in ( [ c ] ) is a noncommuting joint common cause of the correlations @xmath222 localized in @xmath215 if @xmath223 for any @xmath13 and @xmath224 .
_ @xmath225 and @xmath226 are localized in the common past of the events @xmath132 and @xmath133 .
so if they satisfy the criterium @xmath227 for any @xmath228 then they qualify as a noncommuting joint common cause of the correlations @xmath222 .
( [ ncqccsrew ] ) can be rewritten in the equivalent form : @xmath229 where @xmath230 .
using ( [ rho4 ] ) and ( [ c ] ) one can easily calculate @xmath231 : @xmath232 @xmath233 can be readily obtained by substituting @xmath234 for @xmath235 . for the choice @xmath236 we obtain the density matrices @xmath237 derived from the singlet state . substituting ( [ a_m ] ) , ( [ b_n ] ) and
( [ rho_c ] ) into ( [ ncqccsrew ] ) we arrive at the following equation : @xmath238 = \nonumber \\
\lambda c_1 c_2 \ , ( -a^m_1 b^n_2 + a^m_2 b^n_1 ) -
\lambda c^2_2 \ , ( a^m_1 b^n_1 + a^m_2 b^n_2 ) - \left(\lambda -\frac{(1+\lambda)^2}{4}c^2_3
+ \frac{(1-\lambda)^2}{4}\tilde{c}^2_3 \right ) \ , a^m_3 b^n_3\end{aligned}\ ] ] for @xmath239 ; which is zero ( among others ) if @xmath223 for any @xmath13 and @xmath224 .
hence , @xmath240 in ( [ c ] ) with @xmath224 is a noncommuting joint common cause localized in @xmath215 of the correlations @xmath25 with @xmath223 . observe that for the directions @xmath138 and @xmath139 defined in ( [ a1])-([b2 ] ) the requirement @xmath223 holds for any @xmath13 , hence the correlations ( maximally ) violating the ch inequality _ do _ have a joint common cause
any @xmath241 of form ( [ c ] ) with @xmath224 .
finally , we show that there exists no commuting joint common cause for these correlations even without any restriction to their localization .
[ nocommjcc ] let @xmath242 be projections defined in ( [ a_m])-([b_n ] ) with @xmath138 and @xmath139 given in ( [ a1])-([b2 ] ) .
the correlations @xmath25 in the state ( [ rho4 ] ) do _ not _ have a _ commuting _ joint common cause @xmath243 in @xmath2 .
_ since two noncommuting projections @xmath244 in @xmath66 already generate @xmath66 , @xmath245 , we have @xmath246 hence , a commuting joint common cause @xmath247 should be in @xmath248 .
since not only @xmath249 but also @xmath250 is in @xmath129 only those @xmath251-monoms in @xmath167 give non - zero contribution to @xmath252 ( and to the similar expressions containing @xmath253 or @xmath254 ) that are also in @xmath255 .
however , it is a simple matrix algebra therefore @xmath256 . denoting the coefficient of @xmath257 in @xmath167 by @xmath258 , which is the dimension @xmath259 of the projection @xmath167
, one obtains that the correlations containing @xmath167 are the original correlations ( [ corr ] ) multiplied by @xmath260 : @xmath261 which is non - zero for all @xmath138 and @xmath139 defined in ( [ a1])-([b2 ] ) ( if @xmath262 ) .
therefore @xmath247 is not a common cause of any of the correlations @xmath25 .
in this paper it was shown that a set of correlations between spacelike separated projections maximally violating the clauser
horne inequality can be given a joint common causal explanation if common causes _ not _ commuting with the correlating events are allowed . moreover , this noncommuting common cause could be located in the common past of the correlating events .
* acknowledgements . *
this work has been supported by the hungarian scientific research fund , otka k-68195 and by the fulbright research grant while g. h - sz .
was a visiting fellow at the center for philosophy of science in the university of pittsburgh .
j. f. clauser , m.a .
horne , a. shimony and r. a. holt , proposed experiment to test local hidden - variable theories , _ phys
_ , * 23 * , 880 - 884 . | in the paper the epr - bohm scenario will be reproduced in an algebraic quantum field theoretical setting with locally finite degrees of freedom .
it will be shown that for a set of spatially separated correlating events ( projections ) maximally violating the clauser
horne inequality there can be given a common causal explanation _ if _ commutativity is abandoned between the common cause and the correlating events .
moreover , the noncommuting common cause will be local and supported in the common past of the correlating events .
* key words : * clauser horne inequality , common cause , noncommutativity , algebraic quantum field theory . |
the channel capacity of multiple - input - multiple - output ( mimo ) channels linearly increases with the number of antennas @xcite .
maximum - likelihood ( ml ) detection provides the minimum error rate .
however , the computational complexity of the simple ml detection algorithm grows exponentially with the number of transmit antennas .
thus , we need an efficient algorithm that achieves similar error rate to the ml detection .
the qr decomposition m - algorithm ( qrd - mld ) @xcite and sphere decoding ( sd ) @xcite are possibly the most promising algorithms . in @xcite , to reduce the computational complexity , dijkstra s algorithm is applied to sd which achieves same error rate as ml detection . both the qrd - mld and dijkstra s algorithm are tree search based algorithms .
dijkstra s algorithm uses the list of unlimited size to keep detection candidates .
however , the computational complexities of the qrd - mld and dijkstra s algorithm are still high . to reduce the computational complexity , we propose dijkstra s algorithm with bounded list size .
when proposed algorithm s list size is set to achieve the almost same symbol error rate ( ser ) as the qrd - mld , the computational complexity of proposed algorithm is lower than the qrd - mld .
this paper is organized as follows . in section 2
, we introduce the system model of mimo channels . in section 3
, we review the qrd - mld and its improvement , then propose dijkstra s algorithm with bounded list size . in section 4 , we show the comparison between the computational complexity of the qrd - mlds and proposed algorithm by computer simulations . finally , we give the conclusion in section 5 .
we consider the uncoded system with @xmath0 transmit antennas and @xmath1 receive antennas , and we assume @xmath2 .
we assume that the noise at each receive antenna is the additive white gaussian noise ( awgn ) .
let @xmath3 be a @xmath4 vector consisting of complex envelopes of transmitted signals with the signal constellation @xmath5 , @xmath6 an @xmath7 fading matrix whose @xmath8 entry is a complex fading coefficient between @xmath9-th transmit antenna and @xmath10-th receive antenna , @xmath11 an @xmath12 complex vector whose component is noise at each receive antenna , and @xmath13 an @xmath12 complex vector whose component is the received signal component at each receive antenna .
the model of this channel is written as @xmath14 we assume that the receiver knows the channel state information ` h ` perfectly . in this case , the ml detection of the transmitted signal over the channel ( 1 ) can be formulated as finding @xmath15
in this section , we propose the new near ml detection algorithm .
first , to calculate ( 2 ) efficiently , we explain how to find the ml signal by tree search algorithm in section 3.1 .
then , we review near ml detection algorithms called qrd - mld @xcite and its improvement @xcite in section 3.2 . finally , we propose dijkstra s algorithm with bounded list size in section 3.3 .
to calculate ( 2 ) efficiently , we compute a qr decomposition of @xmath6 and obtain an upper triangular matrix @xmath16 and a unitary matrix @xmath17 with @xmath18 .
since @xmath17 is unitary , @xmath19 let @xmath20 . the ml detection problem ( 2 )
can be reformulated as finding @xmath21 @xmath22 the second equality above follows as the second term in the second equation is irrelevant to @xmath3 . to calculate ( 4 ) efficiently , we consider a weighted directed graph as follows .
the decisions on @xmath23 construct a tree where nodes at @xmath10-th depth are correspond to the candidate of @xmath24 @xcite , and the root node is placed at depth 0 .
then , the metric value , which is the weight of branch , between a node @xmath25 that has @xmath26 as ancestor nodes from the root node to its parent node is defined by @xmath27 the distance of each node from the root node , which is called the accumulated metric value in this paper , is equal to the sum of the metric values of branches from the root node to the node itself .
the accumulated metric value from the root node to the bottom node whose depth is @xmath0 is @xmath28 because @xmath29 that makes ( 5 ) minimum is equal to @xmath30 of ( 4 ) , the shortest path from the root node to the bottom node corresponds to the ml signal @xcite .
the qrd - mld @xcite , which is a breadth - first tree search based algorithm , finds a near ml signal .
the qrd - mld keeps only @xmath31 nodes at each depth with the smallest accumulated metric values @xcite , instead of testing all the candidate in @xmath32 according to ( 4 ) . at each depth , only @xmath31 nodes make their child nodes .
we call a node that makes its child node detection node in this paper .
an improvement to qrd - mld proposed in @xcite reduces the number of detection nodes from the original qrd - mld .
this improved qrd - mld has threshold value at each depth .
the depth @xmath33 s threshold value @xmath34 is defined by @xmath35 where @xmath36 is the smallest accumulated metric value of the node at @xmath33-th depth in the nodes whose parent node is a detection node .
@xmath37 is a fixed constant number , and @xmath38 is the noise variance . at each depth , select the nodes that have smaller accumulated metric value than threshold value @xmath34 .
if the number of selected nodes is more than @xmath31 , only @xmath31 nodes with smallest accumulated metric values are selected .
note that both algorithms do not always find the ml signal . for small to
medium @xmath31 values , the complexity is substantially lower than the simple ml detection algorithm .
however , the final result is no longer guaranteed to be the ml signal .
dijkstra s algorithm is an efficient algorithm to find the shortest path from a point to a destination in a weighted graph @xcite .
dijkstra s algorithm uses the list of unlimited size to keep candidate nodes .
if we use dijkstra s algorithm to find the shortest path from the root to one of nodes at the bottom depth , we can get the node with minimum @xmath39 among all nodes at the bottom depth and it corresponds to the ml estimate @xcite
. however , this algorithm still has high computational complexity . to reduce the computational complxity
, we propose a modified version of dijkstra s algorithm whose list keeps only @xmath40 nodes with the smallest accumulated metric values in the list .
we show dijkstra s algorithm with bounded list size . 1 .
create an empty list for nodes .
2 . insert all nodes at the first level into the list .
3 . select the node a having smallest accumulated metric value in the list and remove it from the list .
if the depth of a is @xmath0 , then output the node a and its ancestor nodes as the ml signal and finish this algorithm .
4 . insert all a s child nodes into the list . 5 .
arrange the nodes in the list according to the accumulated metric value by the quick sort .
if the list has more than @xmath40 nodes , select the @xmath40 nodes with the smallest accumulated metric values in the list , and discard other nodes from the list .
go back to step 3 .
the node whose child nodes are inserted into the list is called detection node in this paper . because the discarded nodes , which are decided at step 5 , and their descendant nodes are not examined , the proposed algorithm dose not examine all the candidate in @xmath32 according to ( 4 ) .
thus , the proposed algorithm dose not always find the ml signal .
when we use ldpc codes @xcite or turbo codes @xcite after detection , we have to compute @xmath41 most likely signals @xcite .
such signals can be computed by this algorithm s modification that is finished after output @xmath41 signals with the smallest accumulated metric value .
) symbol error rate ] ) maximum computational complexity ] ) maximum computational complexity ] ) maximum number of detection nodes ] ) maximum number of detection nodes ] ) maximum number of comparisons of real numbers ] ) maximum number of comparisons of real numbers ] ) symbol error rate ] ) maximum computational complexity ] ) maximum computational complexity ] ) maximum number of detection nodes ] ) maximum number of detection nodes ] ) maximum number of comparisons of real numbers ] ) maximum number of comparisons of real numbers ] in this section , we compare the computational complexity , the number of detection nodes and the number of comparisons of real numbers among the proposed algorithm and the qrd - mlds . throughout the simulations ,
we consider the following system model . *
we do two simulations . in the first simulation , the number of transmit antennas @xmath42 , and the number of receive antennas @xmath43 . in the second simulation ,
the number of transmit antennas @xmath44 , and the number of receive antennas @xmath45 . * the signal constellation at each transmit antenna is 16-qam and all signals are drawn according to the uniform i.i.d .
distribution . *
the fading coefficients obey the @xmath46 distribution , and the receiver knows it perfectly . *
the noise at each recieve antenna obeys the @xmath47 distribution .
@xmath48 is caluculated by @xmath49 , where @xmath50 is the average symbol energy .
* we transmit 100000 signals , which is 400000 symbols if the number of transmit antennas is @xmath51 and 600000 symbols if the number of transmit antennas is @xmath52 , and every 100 signals , change the fading matrix .
if @xmath53 is used and the signal constellation is 16-qam , qrd - mld has symbol error rate ( ser ) near to the ml detection @xcite .
so , we use @xmath53 . in qrd - mld s improvement
, we use @xmath54 in ( 6 ) as used in @xcite . in order for the proposed algorithm to have the similar ser to qrd - mld and its improvement , we use two versions of proposed algorithm whose list sizes are @xmath55 and @xmath56 .
figures 1 and 8 show that the proposed algorithm with @xmath55 , the original qrd - mld and the ml algorithm have almost the same ser throughout this simulations .
the proposed algorithm with @xmath56 and qrd - mld s improvement also have similar ser throughout this simulations .
we count the number of multiplications and divisions of complex numbers as the computational complexity .
since the part of qr decomposition is the common part of all compared algorithms , we do not include that part in comparison of complexity . in qrd - mlds
, we use the quick sort to arrange the nodes and decide @xmath31 nodes with the smallest accumulated metric value at each depth . because the qrd - mld keeps @xmath31 nodes at each depth , the number of detection nodes and the computational complexity are completely determined by @xmath31 .
however , in the proposed algorithm and qrd - mld s improvement , the number of detection nodes and the computational complexity are not fixed .
figures 17 are the results of first simulation whose number of transmit antennas and receive antennas are 4 .
figures 814 are the results of second simulation whose number of transmit antennas and receive antennas are 6 . at first , we discuss the result of first simulation . according to figures 2 , 4 and 6 ,
the propose algorithm with @xmath55 reduece the average computational complexity , average number of detection nodes and average number of comparisons of real numbers from the original qrd - mld .
moreover , in the case of high snr , although the proposed algorithm with @xmath55 has much smaller ser than qrd - mld s improvement according to figure 1 , the average computational complexity of proposed algorithm with @xmath55 is almost the same as qrd - mld s improvement . in the case of low snr ,
the average computational complexity , average number of detection nodes and average number of comparisons of real numbers of the proposed algorithm with @xmath56 are lower than qrd - mld s improvement . in the case of high snr , the average computational complexity , average number of detection nodes and average number of comparisons of real numbers of proposed algorithm with @xmath56 are almost same as qrd - mld s improvement while the proposed algorithm has smaller ser according to figure 1 . according to figures 3 , 5 and 7 , in the case of low snr , maximum computational complexity , maximum number of detection nodes and the maximum number of comparisons of real numbers of the proposed algorithm with @xmath55 are higher than qrd - mlds . however , because the average computational complexity , the average number of detection nodes and the average number of comparisons of real number of the proposed algorithm with @xmath55 are lower than qrd - mlds , we find that the proposed algorithm rarely gets high computational complexity , large number of detection nodes or large number of comparisons of real numbers .
according to figure 814 , which is the result of second simulation , the characteristic of proposed algorithm dose not change with the number of antennas .
in this paper , we propose a near ml detection algorithm .
when the list size is adjusted so that the proposed algorithm has the almost same symbol error rate ( ser ) as the original qrd - mld , the average of the computational complexity and the number of detection nodes are reduced . when the list size is adjusted so that the proposed algorithm has the almost same symbol error rate ( ser ) as the qrd - mld s improvement , in the case of low snr , both the average computational complexity and average number of detection nodes are reduced and in the case of high snr , the computational complexity and average number of detection nodes of proposed algorithm is almost same as qrd - mld s improvement while ser of the proposed algorithm becomes smaller than qrd - mld s improvement .
we would like to thank prof .
kiyomichi araki for drawing our attention to the reference [ 6 ] .
this research is partly supported by the international communications foundation .
99 e. telatar , `` capacity of multi - antenna gaussian channels , '' europ .
telecommun . , vol.10 , pp.585595 , nov .
g. j. foschini , `` layered space - time architecture for wireless communication in a fading environment when using multi - element antennas , '' bell labs tech .
j. , vol.1 , pp.4159 , 1996 .
m. o. damen , a. chkeif and j. c. belfiore , `` lattice code decoder for space - time codes , '' ieee commun .
lett . , vol.36 , no.5 , pp.166168 , jan .
b. hassibi and h. vikalo , `` on the sphere - decoding algorithm i. expected complexity , '' ieee trans .
signal proc . , vol.53 , no.8 ,
pp.28062818 , aug . 2005 . k. j. kim , and r. a. iltis , `` joint detection and channel estimation algorithms for qs - cdma signals over time - varying channels , '' ieee trans .
845855 , may 2002 .
j. yue , k. j. kim , g. d. gibson , and r. a. iltis , `` channel estimation and data detection for mimo - ofdm systems , '' proc .
ieee globecom , vol.2 , pp.581585 , dec .
y. dai , s. sun , and z. lei , `` a comparative study of qrd - m detection and sphere decoding for mimo - ofdm systems , '' proc .
ieee pimrc , vol.1 , pp.186190 , sept .
h. kawai , k. higuchi , n. maeda , and m. sawahashi , `` adaptive control of surviving symbol replica candidates in qrm - mld for ofdm mimo multiplexing , '' ieee jsac , vol .
24 , no . 6 , pp.11301140 june 2006 . a. v. aho , j. e. hopcoft , and j. d. ullman , data structures and algorithms , chapter 6.3 , addison - wesley , reading , ma , 1983 .
t. fukatani , r. matsumoto , and t. uyematsu , `` two methods for decreasing the computational complexity of the mimo ml decoder , '' ieice trans .
fundamentals , vol .
e87-a , no.10 , pp.25712576 , oct . 2004 . b. m. hochwald and s. ten brink , `` achieving near - capacity on a multiple - antenna channel , '' ieee trans .
51 , vol.51 , no.3 , pp.389399 , mar .
r. g. gallager , `` low density parity check codes , '' mit press , cambridge , ma , 1963 . c. berrou and a. glavieux , `` near optimum error correcting coding and decoding : turbo - codes , '' ieee trans .
commun . , vol.44 , no.10 , pp.12611271 , oct 1996 . | we propose dijkstra s algorithm with bounded list size after qr decomposition for decreasing the computational complexity of near maximum - likelihood ( ml ) detection of signals over multiple - input - multiple - output ( mimo ) channels .
after that , we compare the performances of proposed algorithm , qr decomposition m - algorithm ( qrd - mld ) , and its improvement . when the list size is set to achieve the almost same symbol error rate ( ser ) as the qrd - mld , the proposed algorithm has smaller average computational complexity . |
51 ophiuchi is a be ( b9.5iiie ) star , located at 131 pc with a rotational velocity @xmath1 = 267 @xmath2 5 km / s @xcite .
51 oph shows a large infrared excess , that was first noticed by waters et al .
1988 @xcite .
this infrared excess was explained as being originated by the circumstellar dust around the star .
the 51 oph circumstellar gas and dust origin is still unknown .
since its infrared excess was noticed , this object has been observed in many different wavelengths , from the infrared to the ultraviolet . in 1993 ,
grady and silvis @xcite found the first evidence for the presence of dust by detecting a 10 @xmath0 m silicate feature in emission .
51 oph has been compared to @xmath3 pic due to the presence of a gas envelope with a variable column density , high density gas accreting to the star and the fact that this gas is also collisionally ionized .
there was no detection of an extended disk at 18 @xmath0 m , which suggested that the hot dust must be located in the close proximity of the star @xcite .
van den ancker et al .
@xcite analyzed 51 oph iso archive data and suggested a few different scenarios for the nature of this system .
the authors have studied the composition of the circumstellar gas and detected spectral features due to hot gas - phase molecules not typical of ae / be stars . they have suggested that 51 oph might not be a young system but a highly evolved one instead .
nevertheless , the presence of all the circumstellar material can not be explained if the system is really an evolved one .
the other scenario includes the presence of a companion in order to explain the hot gas observed .
a more exotic explanation consists of assuming that the detected material is a result of a recent collision of two gas - rich planets or the accretion of a solid body as the star increases its size at the end of its main - sequence life @xcite .
roberge et al .
@xcite observed 51 oph circumstellar disk with the far ultraviolet spectroscopic explorer and found that the composition of the infalling gas is highly nonsolar .
however , all studies suggest that the circumstellar matter is most likely in the form of a keplerian disk rather than in a spherical shell . in 2004
the first midi observations of herbig ae / be stars were published by leinert et al . @xcite .
the authors have studied a sample of 7 herbig ae / be stars but could not fit a dullemond et al .
( 2001 ) type disk model to the 51 oph data like they did for the other stars . in this paper ,
we have used the data obtained by leinert et al .
@xcite on 51 oph and constrained the size and geometry of the 51 oph circumstellar disk by using a standard disk model . in section 2
the observations and data reduction are presented .
the results are introduced and discussed in sections 3 and 4 , and summarized in section 5 .
51 oph was observed during midi science demonstration time ( sdt ) at the vlti on june 15th and 16th 2003 .
midi is a 2 beam combiner that operates in the mid - infrared , 8 to 13 @xmath0 m .
the baseline used was the ut1-ut3 ( 102 m ) , with a maximum full spatial resolution of 20 mas at 10 @xmath0 m .
the observations log is presented in table [ gil : log ] .
we have obtained 5 visibility points , that were dispersed at a resolution of 30 in the n band .
we have reduced the 51 oph data using the idl software written for midi .
the data reduction is described in detail in leinert et al .
( 2004 ) @xcite .
the data analysis and modeling was done with the interpreted programming language yorick .
[ gil : log ] .log of the observations @xcite [ cols="^,^,^,^,^ " , ] for a radius of 7 @xmath4 , we have found a good fit for the spectral energy distribution ( sed ) , with an inner disk radius of 0.55 au and an outer disk radius of 7 au ( see fig . [
gil : sed ] ) . in this figure
one can see the sed for 51 oph well reproduced by a standard disk model .
the photometry points used in these figure were taken from waters et al .
( 1988 ) @xcite .
[ gil : sed ] [ gil : uv ] in figure [ gil : uv ] on the left , we can see the visibility model for the 51 oph circumstellar disk and the @xmath5-track for our measurements .
the 5 visibility points are marked with crosses over the @xmath5-track . in this figure north
is up and west is left , so we note that the position angle is obtained after a fourier transform of the visibility model ( -12@xmath6 + 90@xmath6=78@xmath6 ) .
we have also studied the visibilities behavior against projected baselines and hour angle . in figure
[ gil : uv ] on the right , we have plotted the visibilities as a function of hour angle . for both cases we have found a good fit for all the visibility points by using the disk model parameters in table [ gil : model ] .
the midi observations are compatible with a flat circumstellar disk , inclined of 88@xmath6 with a position angle of 78@xmath6 .
these results are compatible with the recent results obtained by thi et al .
@xcite , who have found a high column density of co rotating in a disk in the inner astronomical unit of 51 oph , seen nearly edge - on .
we have determined a stellar radius of 7 @xmath4 , a disk inner radius of 0.55 au , an outer radius of 7 au , a disk inclination angle of about 88@xmath6 and a disk position angle of about 78@xmath6 . due to the nature of the object
, our interpretation relies critically on one visibility point and thus drew some criticism .
four of the visibility points are nearly identical , while the fifth one is significantly higher .
this higher visibility point was obtained with the lowest projected baseline at 85.6 m .
there is consequently a potential degeneracy of information between a base length effect and an angle effect . in order to secure our data
set , we plan to obtain a few extra visibility points with midi .
* 51 oph was observed for the first time in the mid - infrared at high - angular resolution . *
we have modeled 51 oph visibilities and were able to constrain the size and geometry of the 51 oph circumstellar disk . *
the best fit to the data corresponds to a flat circumstellar disk , inclined of 88@xmath6 with a position angle of 78@xmath6 .
* we will need more visibility measurements in order to secure our data set and confirm the results obtained . | we present interferometric observations of the be star 51 ophiuchi .
these observations were obtained during the science demonstration phase of the midi instrument at the very large telescope interferometer ( vlti ) .
using midi , a michelson 2 beam combiner that operates at the n band ( 8 to 13 @xmath0 m ) , we obtained for the first time observations of 51 oph in the mid - infrared at high - angular resolution .
it is currently known that this object presents a circumstellar dust and gas disk that shows a very different composition from other herbig ae disks .
the nature of the 51 oph system is still a mystery to be solved .
does it have a companion ?
is it a protoplanetary system ? we still do nt know .
observations with midi at the vlti allowed us to reach high - angular resolution ( 20 mas).we have several uv points that allowed us to constrain the disk model .
we have modeled 51 oph visibilities and were able to constrain the size and geometry of the 51 oph circumstellar disk . |
the _ constraint satisfaction problem _ ( csp ) provides a common framework for many theoretical problems in computer science as well as for many real - life applications .
a csp instance consists of a number of variables , a domain , and constraints imposed on the variables with the goal to determine whether the instance is satisfiable , that is , whether there is an assignment of domain values to all the variables in such a way that all the constraints are satisfied .
the general csp is np - complete and thus a major research direction is to identify restrictions on the csp that render the problem _ tractable _ , that is , solvable in polynomial time .
a substantial body of work exists from the past two decades on applications of universal algebra in the computational complexity of and the applicability of algorithmic paradigms to csps .
moreover , a number of celebrated results have been obtained through this method ; see @xcite for a recent survey . however , the algebraic approach to csps is only applicable to _ language - based _ csps , that is , classes of csps defined by the set of allowed constraint relations but with arbitrary interactions of the constraint scopes .
for instance , the well - known 2-sat problem is a class of language - based csps on the boolean domain @xmath1 with all constraint relations being _ binary _ , that is , of arity at most two .
on the other side of the spectrum are _ structure - based _ csps , that is , classes of csps defined by the allowed interactions of the constraint scopes but with arbitrary constraint relations . here
the methods that have been successfully used to establish complete complexity classifications come from graph theory @xcite .
the complexity of csps that are neither language - based nor structure - based , and thus are often called _ hybrid _ csps , is much less understood ; see @xcite for a recent survey .
one approach to hybrid csps that has been rather successful studies the classes of csps defined by _ forbidden patterns _ ; that is , by forbidding certain generic subinstances .
the focus of this paper is on such csps .
we remark that we deal with _ binary _ csps but , unlike in most papers on ( the algebraic approach to ) language - based csps , the domain is _ not _ fixed and is part of the input .
an example of a pattern is given in figure [ fig : btpmc](a ) .
this is the so - called _ broken triangle _ pattern ( btp ) @xcite ( a formal definition is given in section [ sec : prelim ] ) .
btp is an example of a _
tractable _ pattern , which means that any binary csp instance in which btp does not occur is solvable in polynomial time .
the class of csp instances defined by forbidding btp includes , for instance , all tree - structured binary csps @xcite .
there are several generalisations of btp , for instance , to quantified csps @xcite , to existential patterns @xcite , to patterns on more variables @xcite , and other classes @xcite .
the framework of forbidden patterns is general enough to capture language - based csps in terms of their polymorphisms .
for instance , the pattern in figure [ fig : btpmc](b ) captures the notion of binary relations that are max - closed @xcite . surprisingly , there are essentially only two classes of algorithms ( and their combinations ) known for establishing tractability of csps .
these are , firstly , a generalisation of gaussian elimination @xcite , whose applicability for language - based csps is known @xcite , and , secondly , problems solvable by _ local consistency methods _ , which originated in artificial intelligence ; see references in @xcite .
the latter can be defined in many equivalent ways including pebble games , datalog , treewidth , and proof complexity @xcite .
intuitively , a class of csp instances is solvable by @xmath2-consistency if unsatisfiable instances can always be refuted while only keeping partial solutions of size @xmath2 `` in memory '' .
for instance , the 2-sat problem is solvable by local consistency methods . for structure - based csps ,
the power of consistency methods is well understood : a class of structures can be solved by @xmath2-consistency if and only if the treewidth ( modulo homomorphic equivalence ) is at most @xmath2 @xcite .
consequently , consistency methods solve all tractable cases of structurally - restricted bounded - arity csps @xcite . for language - restricted csps
, the power of consistency methods has only recently been characterised @xcite .
our ultimate goal is to understand the power of local consistency methods for hybrid csps . on this quest , we focus in this article on the power of the first level of local consistency , known as _ arc consistency _ ( ac ) , for classes of binary hybrid csps defined by forbidden ( partially - ordered ) patterns .
the class of csps defined by forbidding btp from figure [ fig : btpmc](a ) is in fact solvable by ac .
but as it turns out , btp is not the only pattern with this property . as our main contribution , we give , in theorem [ thm : order ] , a _
complete characterisation _ of so - called simple partially - ordered forbidden patterns which have this ac - solvability property . here
the partial orders are on variables and domain values .
it turns out that btp is just one of five such ac - solvable patterns .
the four other patterns allow us to exhibit new tractable classes , one of which in particular we expect to lead to new applications since it defines a strict generalisation of binary max - closed constraints which have already found applications in computer vision @xcite and temporal reasoning @xcite .
we also provide results on the associated meta problem of deciding whether a csp instance falls into one of these new tractable classes .
given that ac is the first level of local consistency methods and is implemented in _ all _ constraint solvers , an understanding of the power of ac is paramount .
we note that focusing on classes of csps defined by forbidden patterns is very natural as ac _ can not _ introduce forbidden patterns .
while simple patterns do not cover all partially - ordered patterns it is a natural , interesting , and broad enough concept that covers btp and four other novel and non - trivial tractable classes .
we expect our results and techniques to be used in future work on the power of ac . computational complexity classifications have been obtained for binary csps defined by forbidden negative patterns ( i.e. , only pairwise incompatible assignments are specified ) @xcite and for binary csps defined by patterns on 2 constraints @xcite
. moreover , ( generalisations of ) forbidden patterns have been studied in the context of variable and value elimination rules @xcite .
finally , the idea of forbidding patterns as topological minors has recently been investigated @xcite .
@xcite recently considered the possible extensions of the algebraic approach from the language to the hybrid setting .
the power of the valued version of ac @xcite has recently been characterised @xcite .
moreover , the valued version of ac is known to solve all tractable finite - valued language - based csps @xcite .
a pattern can be seen as a generalisation of the concept of a binary csp instance that leaves the consistency of some assignments to pairs of variables undefined .
a _ pattern _ is a four - tuple @xmath3 where : * @xmath4 is a finite set of _ variables _ ; * @xmath5 is a finite set of _ values _ ; * @xmath6 is the set of possible variable - value assignments called _ points _ ; the _ domain _ of @xmath7 is its non - empty set @xmath8 of possible values : @xmath9 ; * @xmath10 is a partial _ compatibility function _ from the set of unordered pairs of points @xmath11 to @xmath12 . if @xmath13 @xmath14 true ( resp . ,
false ) we say that @xmath15 and @xmath16 are _ compatible _ ( resp . , _
incompatible _ ) . for simplicity , we write @xmath17 for @xmath18
. we will use a simple figurative drawing for patterns .
each variable will be drawn as an oval containing dots for each of its possible points .
pairs in the domain of the function @xmath10 will be represented by lines between points : solid lines ( called _ positive _ ) for compatibility and dashed lines ( called _ negative _ ) for incompatibility . the pattern in figure [ fig : lx ] is called lx .
it consists of three variables , five points , six positive edges , and two negative edges .
we refine patterns to give a definition of a csp instance . a _ binary csp instance _
@xmath19 is a pattern @xmath3 where @xmath20 is a total function , _
i.e. _ the domain of @xmath20 is precisely @xmath21 , @xmath22 , @xmath23 . *
the _ relation _ @xmath24 on @xmath25 is @xmath26 .
* a _ partial solution _ to @xmath19 on @xmath27 is a mapping @xmath28 where , for all @xmath29 we have @xmath30 . * a _ solution _ to @xmath19 is a partial solution on @xmath4 . for notational simplicity
we have assumed that there is _ exactly one _ binary constraint between each pair of variables . in particular
, this means that the absence of a constraint between variables @xmath31 is modelled by a complete relation @xmath32 allowing every possible pair of assignments to @xmath33 and @xmath34 .
we say that there is a _ non - trivial _
constraint on variables @xmath31 if @xmath35 .
we also use the simpler notation @xmath36 for @xmath37 .
the main focus of this paper is on ordered patterns , which additionally allow for variable and value orders .
an _ ordered pattern _ is a six - tuple @xmath38 where : * @xmath39 is a pattern ; * @xmath40 is a ( possibly partial ) strict order on @xmath4 ; and * @xmath41 is a ( possibly partial ) strict order on @xmath5 .
a pattern @xmath42 can be seen as an ordered pattern with empty variable and value orders , i.e. @xmath43 . throughout the paper when we say `` pattern '' we implicitly mean `` ordered pattern '' and use the word `` unordered '' to emphasize , if needed , that the pattern in question is not ordered .
we do not consider patterns with structure ( such as equality or order ) between elements in the domains of _ distinct _ variables .
a pattern @xmath44 is called _ basic _ if ( 1 ) @xmath8 and @xmath45 do not intersect for distinct @xmath46 , and ( 2 ) @xmath41 only contains pairs of elements @xmath47 from the domain of the same variable , i.e. , @xmath48 for some @xmath49 .
the pattern shown in figure [ fig : btpmc](a ) is known as the _ broken triangle _ pattern ( btp ) @xcite .
btp consists of three variables , four points , three positive edges , two negative edges , @xmath50 , and @xmath51 .
given a basic pattern , we can refer to a point @xmath52 in the pattern as simply @xmath53 when the variable is clear from the context or a figure .
for instance , the point @xmath54 in figure [ fig : btpmc](a ) can be referred to as @xmath55 . the pattern in figure [ fig : btpmc](b )
is the ( binary ) _ max - closed _ pattern ( mc ) .
the pattern mc consists of two variables , four points , two positive edges , one negative edge , @xmath56 , and @xmath57 .
mc ( figure [ fig : btpmc](b ) ) together with the extra structure @xmath58 is an example of a pattern that is not basic .
( 400,100)(0,0 ) ( 0,0 ) ( 170,100)(0,0 ) ( 10,60 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath33 ( 20,58)(0,0)@xmath34 ( 100,42)(0,0)@xmath59 ( 140,60 ) ( 115,80)(0,0)@xmath55 ( 115,60)(0,0)@xmath60 ( 15,10)(0,0)(a ) ( 220,0 ) ( 180,100)(0,0 ) ( 10,50 ) ( 90,50 ) ( 20,80)(100,80 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,42)(0,0)@xmath33 ( 100,42)(0,0)@xmath34 ( 140,60 ) ( 5,80)(0,0)@xmath61 ( 5,60)(0,0)@xmath62 ( 115,80)(0,0)@xmath55 ( 115,60)(0,0)@xmath60 ( 15,10)(0,0)(b ) for some of the proofs we will require patterns with additional structure , namely , the ability to enforce certain points to be distinct .
a pattern with a disequality structure is a seven - tuple @xmath63 where : * @xmath64 is a pattern ; and * @xmath65 is a set of pairs of domain values that are distinct .
an example of such a pattern is given in figure [ fig : badpatterns](b ) .
some points in a pattern are indistinguishable with respect to the rest of the pattern .
two points @xmath66 are _ mergeable _ in a pattern @xmath64 if there is no point @xmath67 for which @xmath68 , @xmath69 are both defined and @xmath70 .
a pattern is called _ unmergeable _ if it does not contain any mergeable points . the points @xmath55 and @xmath60 in btp ( figure [ fig : btpmc](a ) ) are not mergeable since they have different compatibility with , for instance , the point in variable @xmath33 . the pattern lx ( figure [ fig : lx ] ) is unmergeable
. some points in a pattern ( known as dangling points ) are redundant in arc - consistent csp instances and hence can be removed .
let @xmath44 be a pattern .
a point @xmath71 is called _ dangling _ if it is not ordered by @xmath41 and if there is at most one point @xmath72 for which @xmath17 is defined , and furthermore ( if defined ) @xmath73 .
the point @xmath62 in the pattern mc ( figure [ fig : btpmc](b ) ) is not dangling since it is ordered . in order to use ( the absence of ) patterns for ac - solvability we need to define what we mean when we say that a pattern _ occurs _ in a csp instance .
we define the slightly more general notion of occurrence of a pattern in another pattern , thus extending the definitions for unordered patterns @xcite .
recall that a csp instance corresponds to the special case of a pattern whose compatibility function is total .
essentially pattern @xmath19 occurs in pattern @xmath74 if @xmath19 is homomorphic to a subpattern of @xmath74 via an injective renaming of variables and a ( possibly non - injective ) renaming of points @xcite .
we first make the observation that dangling points in a pattern provide no useful information since we assume that all csp instances are arc consistent , which explains why dangling points can be eliminated from patterns .
[ def : simple ] a pattern is _ simple _ if it is ( i ) basic , ( ii ) has no mergeable points , and ( iii ) has no dangling points . from a given pattern it is possible to create an infinite number of equivalent patterns by adding dangling points or by duplicating points .
by restricting our attention to simple patterns we avoid having to consider such patterns .
let @xmath75 and @xmath76 be two patterns .
a _ homomorphism _ from @xmath77 to @xmath19 is a mapping @xmath78 which satisfies : * if @xmath79 is defined , then @xmath80 . *
the mapping @xmath81 , given by @xmath82 if @xmath83 such that @xmath84 , is well - defined and injective . * if @xmath85 then @xmath86 . *
if @xmath87 , @xmath88 , @xmath89 and @xmath90 then @xmath91 .
a _ consistent linear extension _ of a pattern @xmath92 is a pattern @xmath93 obtained from @xmath19 by first identifying any number of pairs of points @xmath94 which are both mergeable and incomparable ( according to @xmath41 ) and then extending the orders on the variables and the domain values to total orders .
let @xmath75 and @xmath76 be two patterns .
@xmath77 _ occurs _ in @xmath19 if for all consistent linear extensions @xmath93 of @xmath19 , there is a homomorphism from @xmath77 to @xmath93 .
we use the notation csp@xmath95 to represent the set of binary csp instances in which the pattern @xmath19 does _ not _ occur .
this definition extends in a natural way to patterns with a disequality structure .
we can add @xmath96 to a pattern , without changing its semantics , when @xmath97 or @xmath52 and @xmath98 are joined by negative and positive edges to some point @xmath99 .
furthermore , all domain values @xmath100 in an _ instance _ are distinct so there is an implicit @xmath96 .
the pattern mc ( figure [ fig : btpmc](b ) ) occurs in pattern emc ( figure [ fig : emc ] ) but not in patterns btp ( figure [ fig : btpmc](a ) ) or btx ( figure [ fig : btx ] ) . for a pattern @xmath19 ,
we denote by unordered(@xmath19 ) the underlying unordered pattern , that is , @xmath101 for instance , the pattern unordered(btp ) is the pattern from figure [ fig : btpmc](a ) _ without _ the structure @xmath102 .
( 400,110)(0,0 ) ( 0,0 ) ( 170,100)(0,0 ) ( 10,60 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath33 ( 20,58)(0,0)@xmath34 ( 100,42)(0,0)@xmath59 ( 140,60 ) ( 115,80)(0,0)@xmath55 ( 115,60)(0,0)@xmath60 ( 15,10)(0,0)(a ) ( 220,0 ) ( 170,100)(0,0 ) ( 10,60 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath33 ( 20,58)(0,0)@xmath34 ( 100,42)(0,0)@xmath59 ( 140,60 ) ( 115,80)(0,0)@xmath55 ( 115,60)(0,0)@xmath60 ( 15,10)(0,0)(b ) the following three simple lemmas follow from the definitions .
[ lem : occ - transitive ] if @xmath19 occurs in @xmath74 and @xmath74 occurs in @xmath103 , then @xmath19 occurs in @xmath103 .
[ lem : occ - sup ] if @xmath19 occurs in @xmath74 and @xmath19 does not occur in @xmath104 , then @xmath74 does not occur in @xmath104 , i.e. csp@xmath105 @xmath106 csp@xmath107 .
[ lem : occ - unordered ] for any pattern @xmath19 , unordered(@xmath19 ) occurs in @xmath19 .
arc consistency ( ac ) is a fundamental concept for csps .
let @xmath108 be a csp instance .
a point @xmath109 is called _ arc consistent _ if , for all variables @xmath110 in @xmath4 there is some point @xmath111 compatible with @xmath15 . the csp instance @xmath39 is called _ arc consistent _ if @xmath112 and every point in @xmath113 is arc consistent .
points that are not arc - consistent can not be part of a solution so can safely be removed .
there are optimal @xmath114 algorithms for establishing arc consistency which repeatedly remove such points @xcite , where @xmath115 is the number of non - trivial constraints and @xmath116 the maximum domain size .
algorithms establishing arc consistency are implemented in all constraint solvers .
ac is a _ decision procedure _ for a csp instance if , after establishing arc consistency , non - empty domains for all variables guarantee the existence of a solution .
( note that a solution can then be found without backtrack by maintaining ac during search ) .
ac is a decision procedure for a class of csp instances if ac is a decision procedure for every instance from the class .
a pattern @xmath19 is called _ ac - solvable _ if ac is a decision procedure for csp@xmath105 .
the following lemma is a straightforward consequence of the definitions .
[ lem : not - ac ] a pattern @xmath19 is not ac - solvable if and only if there is an instance @xmath117 csp@xmath105 that is arc consistent and has no solution . the following lemma follows directly from lemmas [ lem : occ - sup ] and [ lem : not - ac ] .
[ lem : hered ] if @xmath19 occurs in @xmath74 and @xmath19 is not ac - solvable , then @xmath74 is not ac - solvable . as our main result we will , in theorem [ thm : order ] , characterise all simple patterns that are ac - solvable . for an ordered pattern @xmath19 ,
we denote by invdom(@xmath19 ) , invvar(@xmath19 ) the patterns obtained from @xmath19 by inversing the domain order or the variable order , respectively .
[ lem : inv ] if @xmath19 is not ac - solvable , then neither is any of invdom(@xmath19 ) , invvar(@xmath19 ) or invdom(invvar(@xmath19 ) ) .
the claims follow from inversing the respective orders in the instance @xmath104 of lemma [ lem : not - ac ] proving that @xmath19 is not ac - solvable .
some patterns define the same classes of csp instances .
patterns @xmath19 and @xmath77 are _ equivalent _ if @xmath118 if @xmath19 occurs in @xmath77 and @xmath77 occurs in @xmath19 , then @xmath119 are equivalent .
[ ex : lx < ] let lx@xmath120 be the pattern obtained from lx ( figure [ fig : lx ] ) by adding the partial variable order @xmath121 . due to the symmetry of lx , observe that lx and lx@xmath120 are equivalent .
the two patterns shown in figure [ fig : btp ] are also equivalent : ( a ) btp@xmath122 with structure @xmath123 and @xmath124 , and ( b ) btp@xmath125 with variable order @xmath126 .
we will call these the _ variable - ordered _ and _ domain - ordered _ versions of btp , respectively , when it is necessary to make the distinction between the two .
btp ( figure [ fig : btpmc](a ) ) will refer to the same pattern with the only structure @xmath102 which again , by symmetry , is equivalent to both btp@xmath122 and btp@xmath125 .
our search for a characterisation of all simple patterns decided by arc consistency surprisingly uncovered four new tractable patterns , which we describe in this section .
the first pattern we study is shown in figure [ fig : emc ] .
it is a proper generalisation of the mc pattern ( figure [ fig : btpmc](b ) ) since it has an extra variable and three extra edges .
( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath33 ( 20,42)(0,0)@xmath34 ( 100,42)(0,0)@xmath59 ( 160,60 ) ( 5,80)(0,0)@xmath61 ( 5,60)(0,0)@xmath62 ( 115,80)(0,0)@xmath55 ( 115,60)(0,0)@xmath60 ( 75,30)(0,0)@xmath127 [ thm : emc ] ac is a decision procedure for csp@xmath128 where emc is the pattern shown in figure [ fig : emc ] .
since establishing arc consistency only eliminates domain elements , and hence can not introduce the pattern , it suffices to show that every arc - consistent instance @xmath129 @xmath130 csp@xmath128 has a solution .
we give a constructive proof .
let @xmath131 be an ordering of @xmath4 such that emc does not occur in @xmath104 .
define an assignment @xmath132 to the variables @xmath133 recursively as follows : @xmath134 and , for @xmath135 , @xmath136 for @xmath135 , we denote by @xmath137 a value of @xmath138 such that @xmath139 .
arc consistency guarantees that @xmath140 exists and hence that @xmath141 and @xmath137 are well defined .
we claim that @xmath132 is a solution .
suppose , for a contradiction , that @xmath142 for some @xmath143 .
if there is more than one such pair @xmath144 , then choose @xmath2 to be minimal and then for this value of @xmath2 choose @xmath145 to be minimal .
we prove our claim that @xmath132 is a solution to @xmath104 by induction on @xmath146 .
the claim trivially holds for @xmath147 since @xmath148 .
it remains to show that if the claim holds for instances of size less than @xmath146 then it holds for instances of size @xmath146 .
let @xmath149 and @xmath150 for @xmath151 if @xmath152 .
let @xmath153 be such that @xmath154 . by definition of @xmath155
, we have @xmath156 which implies that this series is finite and hence that @xmath153 is well - defined .
we distinguish two cases : ( 1 ) @xmath157 , and ( 2 ) @xmath158 . since @xmath142 and @xmath159 we know that @xmath160 .
[ [ case-1-j - m_1 ] ] case ( 1 ) @xmath157 : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + define @xmath161 . by definition of @xmath162 , we know that @xmath163 . since @xmath142 and @xmath164
, we have @xmath165 . by our choice of @xmath145 to be minimal , and since @xmath157 we know that @xmath166 for @xmath167 . indeed , by minimality of @xmath2 , we already had @xmath168 for @xmath169 .
thus , since @xmath170 , we have @xmath171 ( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,8)(0,0)@xmath172 ( 20,42)(0,0)@xmath173 ( 100,42)(0,0)@xmath174 ( 160,60 ) ( 3,80)(0,0)@xmath175 ( 5,60)(0,0)@xmath176 ( 116,80)(0,0)@xmath177 ( 116,60)(0,0)@xmath162 ( 75,30)(0,0)@xmath178 by arc consistency , @xmath179 such that @xmath180 .
we have @xmath181 by minimality of @xmath2 and since @xmath182 . since @xmath183 and hence @xmath184 , we have @xmath185 and @xmath186 by the maximality of @xmath187 in equation ( [ eq : defai ] ) .
we thus have the situation illustrated in figure [ fig : almostemc ] for @xmath188 .
since the pattern emc does not occur in @xmath104 , we must have @xmath189 .
( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,60)(100,80 ) ( 60,30)(100,60 ) ( 20,60)(1,0)80 ( 20,80)(1,0)80 ( 20,60)(4,-3)40 ( 60,30)(4,5)40 ( 60,8)(0,0)@xmath190 ( 20,42)(0,0)@xmath191 ( 100,42)(0,0)@xmath192 ( 157,60 ) ( 5,80)(0,0)@xmath193 ( 2,60)(0,0)@xmath194 ( 120,80)(0,0)@xmath195 ( 123,60)(0,0)@xmath196 ( 80,30)(0,0)@xmath197 for @xmath198 , let @xmath199 be the following hypothesis .
@xmath199 : : @xmath200 , @xmath201 , @xmath202 , with @xmath203 , such that we have the situation shown in figure [ fig : hremc ] .
we have just shown that @xmath204 holds ( with @xmath205 and @xmath206 ) .
we now show , for @xmath207 , that @xmath208 . we know
that @xmath209 and @xmath210 , since @xmath211 and by maximality of @xmath212 in equation ( [ eq : defai ] ) .
let @xmath213 be minimal such that @xmath214 .
we distinguish two cases : ( a ) @xmath215 , and ( b ) @xmath216 .
if @xmath215 , then we have @xmath217 ( from equation ( [ eq : posedges ] ) , since @xmath170 ) , @xmath218 ( since @xmath215 ) , @xmath219 ( by minimality of @xmath2 , since @xmath220 ) . by arc consistency , @xmath221 such that @xmath222 .
we then have the situation illustrated in figure [ fig : almostemc ] for @xmath223 .
as above , from the absence of pattern emc , we can deduce that @xmath224 .
we thus have @xmath225 ( with @xmath226 and @xmath227 ) .
( 190,100)(-10,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,8)(0,0)@xmath228 ( 20,42)(0,0)@xmath229 ( 100,42)(0,0)@xmath230 ( 150,60 ) ( -2,80)(0,0)@xmath231 ( 1,60)(0,0)@xmath232 ( 116,80)(0,0)@xmath233 ( 119,60)(0,0)@xmath234 ( 81,30)(0,0)@xmath235 if @xmath216 , then @xmath236 implies that @xmath237 holds . by minimality of @xmath238 , we know that @xmath239 since @xmath240 . we know that @xmath241 from equation ( [ eq : posedges ] ) , and that @xmath242 by definition of @xmath238 . we know
that @xmath243 and @xmath244 from @xmath237 . by arc consistency , @xmath245 such that @xmath246 .
we then have the situation illustrated in figure [ fig : almostemc2 ] . as
above , from the absence of pattern emc , we can deduce that @xmath247 .
we thus have @xmath225 ( with @xmath248 and @xmath249 ) .
[ [ case-2-j - m_1 ] ] case ( 2 ) @xmath158 : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + consider the subproblem @xmath250 of @xmath104 on variables @xmath251 . since @xmath252 does not belong to the set of variables of @xmath250 , this instance has size strictly less than @xmath146 , and hence by our inductive hypothesis has a solution .
the values of @xmath141 may differ between @xmath104 and @xmath250 .
however , we can see from its definition given in equation ( [ eq : defai ] ) , that the value of @xmath141 depends uniquely on the subproblem on previous variables @xmath253
. showing the dependence on the instance by a superscript , we thus have @xmath254 ( @xmath255 ) although @xmath256 may ( and , in fact , does ) differ from @xmath257 . by our inductive hypothesis ,
@xmath258 is a solution to @xmath250 . setting @xmath259
, it follows that @xmath260 for @xmath261 .
in particular , since @xmath158 , we have @xmath262
. now @xmath263 , since @xmath250 is a subinstance of @xmath104 ( and so , from equation ( [ eq : defai ] ) , @xmath257 is the minimum of a superset over which @xmath256 is a minimum ) .
thus @xmath264 , since @xmath265 and @xmath142 .
by arc consistency , @xmath179 such that @xmath266 .
as in case ( 1 ) , we have the situation illustrated in figure [ fig : almostemc ] for @xmath188 . since the pattern emc does not occur in @xmath104 , we must have @xmath189 .
consider the hypothesis @xmath199 stated in case ( 1 ) and illustrated in figure [ fig : hremc ] .
we have just shown that @xmath204 holds ( with @xmath205 and @xmath206 ) .
we now show , for @xmath207 , that @xmath267 . as in case
( 1 ) , we know that @xmath268 and @xmath210 .
let @xmath269 be minimal such that @xmath242 .
we have seen above that @xmath270 ( since @xmath229 , @xmath271 are assigned , respectively , the values @xmath231 , @xmath177 in a solution to @xmath250 ) .
therefore , we can deduce that @xmath216 .
therefore @xmath272 implies that @xmath237 holds . by minimality of @xmath2 , and since @xmath273 , we know that @xmath274 . as in case ( 1 ) , by minimality of @xmath238 , we know that @xmath275 . by arc consistency , @xmath221 such that @xmath246 .
we thus have the situation illustrated in figure [ fig : almostemc2 ] .
again , from the absence of pattern emc , we can deduce that @xmath247 .
we thus again have @xmath225 with @xmath248 and @xmath249 .
thus , by induction on @xmath276 , we have shown in both cases that @xmath277 holds . but recall that @xmath154 and that @xmath278 was chosen to the maximal element of @xmath279 and hence @xmath280 such that @xmath281 .
this contradiction shows that @xmath132 is a solution , as claimed .
the next two patterns we study in this section , shown in figure [ fig : btx ] and figure [ fig : bti ] , are similar to emc but the three patterns are incomparable ( in the sense that none occurs in another ) due to the different orders on the three variables .
( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath33 ( 20,42)(0,0)@xmath34 ( 100,42)(0,0)@xmath59 ( 160,60 ) ( 5,80)(0,0)@xmath61 ( 5,60)(0,0)@xmath62 ( 115,80)(0,0)@xmath55 ( 115,60)(0,0)@xmath60 ( 75,30)(0,0)@xmath127 ( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath33 ( 20,42)(0,0)@xmath34 ( 100,42)(0,0)@xmath59 ( 160,60 ) ( 5,80)(0,0)@xmath61 ( 5,60)(0,0)@xmath62 ( 115,80)(0,0)@xmath55 ( 115,60)(0,0)@xmath60 ( 75,30)(0,0)@xmath127 [ thm : btx ] ac is a decision procedure for csp@xmath282 where btx is the pattern shown in figure [ fig : btx ] .
since establishing arc consistency only eliminates domain elements , and hence can not introduce the pattern , we only need to show that every arc - consistent instance @xmath283 csp@xmath282 has a solution .
let @xmath284 be an ordering of @xmath4 such that btx does not occur in @xmath104 .
in fact we will show a stronger result by proving that the hypothesis @xmath285 , below , holds for all @xmath286 .
@xmath285 : : for all arc - consistent instances @xmath287 from csp@xmath282 with @xmath288 , if @xmath289 , then @xmath104 has a solution @xmath290 with @xmath291 .
trivially , @xmath204 holds .
suppose that @xmath292 holds where @xmath293 .
we will show that this implies @xmath285 , which will complete the proof by induction .
let @xmath294 ( @xmath295 ) .
denote by @xmath250 the subproblem of @xmath104 on variables @xmath296 and domains @xmath297 ( @xmath295 ) . to complete the inductive proof ,
it is sufficient to show that @xmath250 is arc consistent : since @xmath298 csp@xmath282 and has @xmath299 variables , by @xmath292 , if @xmath250 is arc consistent it has a solution which can clearly be extended to a solution to @xmath104 by adding the assignment @xmath300 .
consider any two variables @xmath301 such that @xmath302 .
we know by arc consistency of @xmath104 that @xmath297 is non - empty .
let @xmath303 ( i.e. @xmath304 ) . to complete the proof , it suffices to show that @xmath305 has a support in @xmath306 . by arc consistency of @xmath104
, we can deduce the existence of @xmath307 such that @xmath308 , and then @xmath309 such that @xmath310 , as well as @xmath311 such that @xmath312 ( i.e. @xmath313 ) .
if @xmath305 has no support in @xmath306 , then we must have @xmath314 ( i.e. @xmath315 ) and @xmath316 . since @xmath53 is the maximum element of @xmath279 , we know that @xmath317 .
indeed , since @xmath315 and @xmath310 , we have that @xmath318 .
but then the pattern btx occurs in @xmath104 , as shown in figure [ fig : occurrencebtx ] .
this contradiction shows that @xmath305 does have a support in @xmath306 and hence that @xmath250 is arc consistent , as required .
( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath319 ( 20,42)(0,0)@xmath320 ( 100,42)(0,0)@xmath172 ( 160,60 ) ( 5,80)(0,0)@xmath53 ( 5,60)(0,0)@xmath321 ( 115,80)(0,0)@xmath115 ( 115,60)(0,0)@xmath116 ( 75,30)(0,0)@xmath305 ( 360,220)(0,0 ) ( 0,120 ) ( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,7)(0,0)@xmath172 ( 20,42)(0,0)@xmath322 ( 100,42)(0,0)@xmath319 ( 110,10 ) ( -3,80)(0,0)@xmath323 ( 5,60)(0,0)@xmath115 ( 115,80)(0,0)@xmath305 ( 115,60)(0,0)@xmath53 ( 83,30)(0,0)@xmath324 ( 10,10)(0,0)(a ) ( 220,120 ) ( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,7)(0,0)@xmath322 ( 20,42)(0,0)@xmath172 ( 100,42)(0,0)@xmath319 ( 110,10 ) ( -3,80)(0,0)@xmath324 ( 5,60)(0,0)@xmath116 ( 115,80)(0,0)@xmath53 ( 115,60)(0,0)@xmath305 ( 83,30)(0,0)@xmath323 ( 10,10)(0,0)(b ) ( 0,0 ) ( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,7)(0,0)@xmath174 ( 20,42)(0,0)@xmath319 ( 100,42)(0,0)@xmath172 ( 110,10 ) ( 5,80)(0,0)@xmath53 ( 5,60)(0,0)@xmath305 ( 123,80)(0,0)@xmath324 ( 115,60)(0,0)@xmath116 ( 83,30)(0,0)@xmath325 ( 10,10)(0,0)(c ) ( 220,0 ) ( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,7)(0,0)@xmath326 ( 20,42)(0,0)@xmath319 ( 100,42)(0,0)@xmath322 ( 110,10 ) ( 5,80)(0,0)@xmath305 ( 5,60)(0,0)@xmath53 ( 123,80)(0,0)@xmath323 ( 115,60)(0,0)@xmath115 ( 83,30)(0,0)@xmath327 ( 10,10)(0,0)(d ) [ thm : bti ] ac is a decision procedure for csp@xmath328 where bti is the pattern shown in figure [ fig : bti ] . since establishing arc consistency only eliminates domain elements , and hence can not introduce the pattern , we only need to show that every arc - consistent instance @xmath283 csp@xmath328 has a solution
let @xmath131 be an ordering of @xmath4 such that bti does not occur in @xmath104 .
in fact , we will show a stronger result by proving that , for all arc - consistent instances @xmath283 csp@xmath328 , @xmath104 has a solution @xmath290 defined recursively by : @xmath329 is the maximum value in @xmath330 compatible with all previous assignments @xmath331 .
let @xmath332 denote the set of values in @xmath330 compatible with the assignments @xmath331 ( defined recursively as above ) .
if each @xmath332 ( @xmath333 ) is non - empty , then @xmath334 is a solution to @xmath104 .
let @xmath335 denote the subinstance of @xmath104 on the first @xmath336 variables @xmath337 .
@xmath338 and is non - empty since @xmath104 is arc consistent , and hence @xmath339 is a solution to @xmath340 .
suppose that @xmath331 , as defined above , is a solution to @xmath341 .
we will show that @xmath329 exists and hence that @xmath342 is a solution to @xmath343 , which by a simple induction will complete the proof .
suppose for a contradiction that @xmath329 does not exist , i.e. that @xmath344 .
let @xmath345 denote the subset of @xmath330 which is compatible with the assignment of @xmath346 to @xmath172 , i.e. @xmath347 . for @xmath348 , let @xmath349 be the intersection of the sets @xmath350 ( @xmath351 ) . by our hypothesis that @xmath344 , we know that @xmath352 . by arc consistency , @xmath353 .
let @xmath354 be minimal such that @xmath355 .
thus @xmath356 by arc consistency , @xmath357 such that @xmath358 ( i.e. @xmath359 ) . by equation ( [ eq : rtj ] ) and definition of @xmath360
, there is some @xmath361 such that @xmath362 .
choose @xmath276 to be minimal .
consider @xmath363 ( which is non - empty since @xmath145 was chosen to be minimal ) .
then @xmath364 , by definition of @xmath360 , since @xmath365 . but
@xmath366 by ( [ eq : rtj ] ) . by arc consistency , @xmath367 such that @xmath368 and @xmath369 such that @xmath370 . by our inductive hypothesis that @xmath331 is a solution to @xmath341 , we know that @xmath371 since @xmath372 .
because of their different compatibilities with , respectively , @xmath305 and @xmath53 , we know that @xmath373 and @xmath374 . if @xmath375 , then , since @xmath376 , the pattern bti occurs in @xmath104 , as shown in figure [ fig : occurrencesbti](a ) ; so we can deduce that @xmath377 .
similarly , if @xmath378 , then , since @xmath138 , the pattern bti occurs in @xmath104 , as shown in figure [ fig : occurrencesbti](b ) ; so we can deduce that @xmath379 .
but , by definition of @xmath290 , @xmath323 is the maximal element of @xmath380 .
so , since @xmath377 , there must be some @xmath381 such that @xmath382 . by minimality of @xmath276
, we know that @xmath383 . similarly , there must be some @xmath384 such that @xmath385 since @xmath379 and @xmath324 is the maximal element of @xmath386 .
since @xmath387 and @xmath384 , we know that @xmath388 . by our inductive hypothesis that @xmath331 is a solution to @xmath341 , we know that @xmath389 and @xmath390 since @xmath391 . we know that @xmath392 because they have different compatibilities with @xmath324 .
if @xmath393 , then , since @xmath384 , the pattern bti occurs in @xmath104 , as shown in figure [ fig : occurrencesbti](c ) . and ,
if @xmath394 , then the pattern bti occurs in @xmath104 , as shown in figure [ fig : occurrencesbti](d ) .
this contradiction shows that @xmath395 ( for each @xmath333 ) and hence that @xmath104 has a solution @xmath290 .
we conclude this section with a pattern which is essentially different from the patterns emc , btx , and bti , since it includes two negative edges that meet but has no domain or variable order .
the tractability of this pattern was previously unknown @xcite .
( 120,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,60)(60,30 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,8)(0,0)@xmath33 ( 20,42)(0,0)@xmath34 ( 100,42)(0,0)@xmath59 ( 5,80)(0,0)@xmath61 ( 5,60)(0,0)@xmath62 ( 115,80)(0,0)@xmath55 ( 115,60)(0,0)@xmath60 ( 75,30)(0,0)@xmath127 [ thm : lx ] ac is a decision procedure for csp@xmath396 where lx is the pattern shown in figure [ fig : lx ] . since establishing arc consistency only eliminates domain elements , and hence can not introduce the pattern , we only need to show that every arc - consistent instance @xmath117 csp@xmath396 has a solution .
in fact we will show a stronger result by proving that the hypothesis @xmath285 , below , holds for all @xmath397 .
@xmath285 : : for all arc - consistent instances @xmath398 csp@xmath396 with @xmath399 , @xmath400 , @xmath401 , @xmath104 has a solution @xmath290 such that @xmath402 .
trivially , @xmath204 holds .
suppose that @xmath292 holds where @xmath293 .
we will show that this implies @xmath285 , which will complete the proof by induction .
consider an arc - consistent instance @xmath129 from csp@xmath396 with @xmath403 and let @xmath404 where @xmath405 .
let @xmath406 denote the subproblem of @xmath104 on variables @xmath407 .
for any solution @xmath290 of @xmath406 , we denote by @xmath408 the set of variables in @xmath407 on which @xmath290 is compatible with the unary assignment @xmath409 , i.e. @xmath410 consider two distinct solutions @xmath411 to @xmath406 . if we have @xmath412 and @xmath413 , then the pattern lx occurs in @xmath104 under the mapping @xmath414 , @xmath415 , @xmath416 , @xmath417 , @xmath418 , @xmath419 , @xmath420 , @xmath421 ( see figure [ fig : lx ] ) . since lx does not occur in @xmath104 , we can deduce that the sets @xmath408 , as @xmath290 varies over all solutions to @xmath406 , form a nested family of sets .
let @xmath422 be a solution to @xmath406 such that @xmath423 is maximal for inclusion .
consider any @xmath424 . by arc consistency , @xmath425 such that @xmath426 . by our inductive hypothesis @xmath292
, there is a solution @xmath290 to @xmath406 such that @xmath427 . since @xmath428
, we have @xmath429 . by maximality of @xmath422 , this implies @xmath430 , i.e. @xmath431 . since this is true for any @xmath424 , we can deduce that @xmath422 can be extended to a solution to @xmath104 ( which assigns @xmath53 to @xmath319 ) by simply adding the assignment @xmath409 to @xmath422 .
for an unordered pattern @xmath19 of size @xmath2 , checking for ( the non - occurrence of ) @xmath19 in a csp instance @xmath104 is solvable in time @xmath432 by simple exhaustive search .
consequently , checking for ( the non - occurrence of ) unordered patterns of constant size is solvable in polynomial time .
however , the situation is less obvious for ordered patterns since we have to test all possible orderings of @xmath104 .
the following result was shown in @xcite .
[ thm : recog - btp ] given a binary csp instance @xmath104 with a fixed total order on the domain , there is a polynomial - time algorithm to find a total variable ordering such that btp does not occur in @xmath104 ( or to determine that no such ordering exists ) .
we show that the same result holds for the other three ordered patterns studied in this paper , namely bti , btx , and emc .
[ thm : recog - bti ] given a binary csp instance @xmath104 with a fixed total order on the domain and a pattern @xmath433bti , btx , emc@xmath434 , there is a polynomial - time algorithm to find a total variable ordering such that @xmath19 does not occur in @xmath104 ( or to determine that no such ordering exists ) .
we give a proof only for btx as the same idea works for the other two patterns as well . given a binary csp instance @xmath104 with @xmath146 variables @xmath435 , we define an associated csp instance @xmath436 that has a solution precisely when there exists a suitable variable ordering for @xmath104 .
to construct @xmath436 , let @xmath437 be variables taking values in @xmath438 representing positions in the ordering .
we impose the ternary constraint @xmath439 for all triples of variables @xmath440 in @xmath104 such that the btx pattern occurs for some @xmath441 with @xmath442 , @xmath443 , and @xmath444 when the variables are ordered @xmath445 .
the instance @xmath436 has a solution precisely if there is an ordering of the variables @xmath435 of @xmath104 for which btx does not occur .
note that if the solution obtained represents a partial order ( i.e. if @xmath446 and @xmath447 are assigned the same value for some @xmath448 ) , then it can be extended to a total order which still satisfies all the constraints by arbitrarily choosing the order of those @xmath446 s that are assigned the same value .
this reduction is polynomial in the size of @xmath104 .
we now show that all constraints in @xmath436 are ternary max - closed and thus @xmath436 can be solved in polynomial time @xcite .
let @xmath449 and @xmath450 satisfy any constraint in @xmath436
. then @xmath451 and @xmath452 , and thus @xmath453 @xmath14 @xmath454 .
consequently , @xmath455 , @xmath456 , @xmath457 also satisfies the constraint .
we can deduce that all constraints in @xmath436 are max - closed . using the same technique
, we can also show the following .
[ thm : recog - dom - bti ] given a binary csp instance @xmath104 with a fixed total variable order and a pattern @xmath433bti , btx@xmath434 , there is a polynomial - time algorithm to find a total domain ordering such that @xmath19 does not occur in @xmath104 ( or determine that no such ordering exists ) .
it is known that determining a domain order for which mc does not occur is np - hard @xcite .
not surprisingly , for emc when the domain order is not known , detection becomes np - hard .
for the case of btx and bti , if neither the domain nor variable order is known , finding orders for which the pattern does not occur is again np - hard .
[ thm : hardness ] for the pattern emc , even for a fixed total variable order of an arc - consistent binary csp instance @xmath104 , it is np - hard to find a total domain ordering of @xmath104 such that the pattern does not occur in @xmath104 . for patterns btx and bti ,
it is np - hard to find total variable and domain orderings of an arc - consistent binary csp instance @xmath104 such that the pattern does not occur in @xmath104 . to show this
, we exhibit a polynomial reduction from 3sat . given an @xmath146-variable instance @xmath458 of 3sat
, we create a domain of size @xmath459 with a value @xmath141 for each variable @xmath460 in @xmath458 and another @xmath461 for its negation @xmath462 .
this is the domain of each variable @xmath319 in a binary csp instance @xmath463 . for each total ordering @xmath464 of the domain
there is a corresponding assignment to the variables of @xmath458 given by @xmath465 true if and only if @xmath466 . to complete the reduction we have to show how to impose a clause , e.g. @xmath467 .
the basic construction of @xmath463 is composed of the following elements : @xmath468 variables linked by equality constraints @xmath469 ( @xmath470 ) . consider first the case of emc . to impose a clause
we can construct a gadget on any three variables @xmath471 which are not linked by any other constraints in our construction ( in particular , not consecutive variables linked by equality constraints ) .
we add four extra values @xmath472 ( of which @xmath473 depend on the clause ) in the domain of each of these three variables @xmath471 .
the value @xmath474 is compatible with all values in the domains of the other variables .
this ensures arc consistency for all domain values , and if @xmath475 for all domain values @xmath476 , then the pattern can not occur on @xmath474 .
we then place negative edges between all pairs of values other than @xmath474 except for three positive edges in each constraint . in the constraint between @xmath477 and @xmath478
we add the three positive edges : @xmath479 , @xmath480 , @xmath481 .
if @xmath482 in @xmath483 and @xmath484 in @xmath485 , then the pattern emc occurs in @xmath463 .
the third variable of the pattern is @xmath486 or @xmath487 which is linked by an equality constraint to @xmath478 .
thus , to avoid the pattern occurring in @xmath463 the domain order must respect @xmath488 by adding similar constraints between @xmath489 and @xmath490 , we can also impose @xmath491 by imposing these three inequalities , we impose @xmath492 ( since we can not simultaneously have @xmath493 ) which corresponds to the clause @xmath494 in @xmath458 . by inversing the roles of @xmath141 and @xmath495 we can clearly impose clauses involving negative literals .
this completes the reduction from @xmath458 to the problem of finding a domain ordering of a binary csp instance so that emc does not occur .
since this reduction is clearly polynomial , we can conclude that the problem of testing the existence of a domain order so that emc does not occur is np - hard .
now consider the case of the pattern btx .
we use a similar construction to the case of emc , above . again , to simulate a clause @xmath496 in @xmath458 we need to impose @xmath497 .
this can achieved by imposing : @xmath498 for example , to impose @xmath499 we place the same gadget as above ( i.e. positive edges @xmath479 , @xmath480 , @xmath481 ) on each of the three pairs of variables @xmath500 , @xmath501 and @xmath502 . to avoid btx on variables @xmath503 , on variables @xmath504 and on variables @xmath505 , we must have @xmath506 since there is a total strict ordering on the variables , this is logically equivalent to imposing @xmath499 , as required ( provided none of the variables @xmath507 are used in other gadgets ) . finally , we consider the pattern bti .
but this is an easier case than btx .
we just need to place the gadget on @xmath500 to impose @xmath508 to avoid the pattern bti . thus , emc is np - hard to detect when the domain order of the instance is not fixed , and btx , bti are np - hard to detect when neither the domain order nor the variable order of the instance is fixed .
the results from this section are summarised in table [ tab : summary ] .
we use the star to denote uninteresting cases .
note that since lx is an unordered pattern the questions of determining variable and/or domain orders are not interesting .
similarly , since pattern btp only orders variables the question of determining a domain order is not interesting .
.summary of recognition problems . [ cols="^,^,^,^,^,^",options="header " , ]
we first give a set of instances , each of which is arc consistent and has no solution .
if for any of these instances @xmath104 , we have @xmath117 csp@xmath105 , then this constitutes a proof , by lemma [ lem : not - ac ] , that pattern @xmath19 is not solved by arc consistency . for simplicity of presentation , in each of the following instances , we suppose the variable order given by @xmath509 if @xmath510 .
( 160,160)(10,10 ) ( 30,90)(18,58 ) ( 30,70)(0,0)@xmath0 ( 30,90)(0,0)@xmath0 ( 30,110)(0,0)@xmath0 ( 150,90)(18,58 ) ( 150,70)(0,0)@xmath0 ( 150,90)(0,0)@xmath0 ( 150,110)(0,0)@xmath0 ( 90,30)(58,18 ) ( 70,30)(0,0)@xmath0 ( 90,30)(0,0)@xmath0 ( 110,30)(0,0)@xmath0 ( 90,150)(58,18 ) ( 70,150)(0,0)@xmath0 ( 90,150)(0,0)@xmath0 ( 110,150)(0,0)@xmath0 ( 53,150)(0,0)@xmath320 ( 150,127)(0,0)@xmath511 ( 53,30)(0,0)@xmath512 ( 30,127)(0,0)@xmath513 ( 30,110)(70,150 ) ( 30,110)(90,150 ) ( 30,90)(70,150 ) ( 30,90)(90,150 ) ( 30,70)(150,70 ) ( 30,70)(150,110 ) ( 30,110)(150,70 ) ( 30,110)(150,110 ) ( 70,30)(70,150 ) ( 70,30)(110,150 ) ( 110,30)(70,150 ) ( 110,30)(110,150 ) ( 90,150)(150,90 ) ( 90,150)(150,110 ) ( 110,150)(150,90 ) ( 110,150)(150,110 ) ( 30,70)(70,30 ) ( 30,70)(90,30 ) ( 30,90)(70,30 ) ( 30,90)(90,30 ) ( 90,30)(150,90 ) ( 90,30)(150,70 ) ( 110,30)(150,90 ) ( 110,30)(150,70 ) ( 30,90)(1,0)120 ( 30,90)(6,1)120 ( 30,90)(6,-1)120 ( 150,90)(-6,1)120 ( 150,90)(-6,-1)120 ( 90,30)(0,1)120 ( 90,30)(1,6)20 ( 90,30)(-1,6)20 ( 90,150)(1,-6)20 ( 90,150)(-1,-6)20 ( 30,70)(1,2)40 ( 30,70)(3,4)60 ( 30,70)(1,1)80 ( 110,150)(-2,-1)80 ( 110,150)(-4,-3)80 ( 30,110)(1,-2)40 ( 30,110)(3,-4)60 ( 30,110)(1,-1)80 ( 110,30)(-2,1)80 ( 110,30)(-4,3)80 ( 70,150)(2,-1)80 ( 70,150)(4,-3)80 ( 70,150)(1,-1)80 ( 150,70)(-1,2)40 ( 150,70)(-3,4)60 ( 70,30)(2,1)80 ( 70,30)(4,3)80 ( 70,30)(1,1)80 ( 150,110)(-1,-2)40 ( 150,110)(-3,-4)60 ( 70,164)(0,0)1 ( 90,164)(0,0)2 ( 110,164)(0,0)3 ( 70,16)(0,0)3 ( 90,16)(0,0)2 ( 110,16)(0,0)1 ( 16,70)(0,0)1 ( 16,90)(0,0)2 ( 16,110)(0,0)3 ( 164,70)(0,0)3 ( 164,90)(0,0)2 ( 164,110)(0,0)1 ( 170,160)(10,10 ) ( 30,90)(18,58 ) ( 30,70)(0,0)@xmath0 ( 30,90)(0,0)@xmath0 ( 30,110)(0,0)@xmath0 ( 150,100)(18,38 ) ( 150,90)(0,0)@xmath0 ( 150,110)(0,0)@xmath0 ( 80,30)(38,18 ) ( 70,30)(0,0)@xmath0 ( 90,30)(0,0)@xmath0 ( 80,150)(38,18 ) ( 70,150)(0,0)@xmath0 ( 90,150)(0,0)@xmath0 ( 55,150)(0,0)@xmath320 ( 150,125)(0,0)@xmath512 ( 55,30)(0,0)@xmath511 ( 30,125)(0,0)@xmath513 ( 30,110)(70,150 ) ( 30,90)(70,150 ) ( 30,70)(70,30 ) ( 30,110)(70,30 ) ( 30,70)(150,110 ) ( 30,90)(150,110 ) ( 90,150)(150,90 ) ( 90,30)(150,90 ) ( 90,30)(90,150 ) ( 30,90)(1,0)120 ( 30,110)(1,0)120 ( 150,90)(-6,1)120 ( 150,90)(-6,-1)120 ( 30,90)(1,1)60 ( 30,70)(1,2)40 ( 30,110)(3,2)60 ( 30,70)(3,4)60 ( 30,90)(1,-1)60 ( 30,90)(2,-3)40 ( 30,110)(3,-4)60 ( 30,70)(3,-2)60 ( 70,30)(0,1)120 ( 70,30)(1,6)20 ( 90,30)(-1,6)20 ( 70,150)(2,-1)80 ( 70,150)(4,-3)80 ( 90,150)(3,-2)60 ( 70,30)(1,1)80 ( 70,30)(4,3)80 ( 90,30)(3,4)60 ( 70,164)(0,0)1 ( 90,164)(0,0)0 ( 70,16)(0,0)1 ( 90,16)(0,0)0 ( 16,70)(0,0)1 ( 16,90)(0,0)2 ( 16,110)(0,0)3 ( 164,90)(0,0)0 ( 164,110)(0,0)1 ( 20,40 ) ( 20,40)(0,0 ) ( 10,20)(18,38 ) ( 10,10)(0,0)@xmath0 ( 10,30)(0,0)@xmath0 ( 280,220)(0,0 ) ( 0,0 ) ( 120,220)(0,-20 ) ( 10,30 ) ( 50,-10 ) ( 90,70 ) ( 10,110 ) ( 50,150 ) ( 20,120)(60,160 ) ( 20,140)(100,80 ) ( 60,180)(100,80 ) ( 20,60)(100,100 ) ( 20,40)(60,0 ) ( 60,20)(100,100 ) ( 20,140)(1,1)40 ( 20,140)(2,1)40 ( 20,120)(2,3)40 ( 20,140)(2,-1)80 ( 20,120)(2,-1)80 ( 20,120)(4,-1)80 ( 60,180)(1,-2)40 ( 60,160)(1,-2)40 ( 60,160)(2,-3)40 ( 20,60)(4,1)80 ( 20,40)(4,3)80 ( 20,40)(2,1)80 ( 20,60)(2,-3)40 ( 20,60)(1,-1)40 ( 20,40)(2,-1)40 ( 60,20)(2,3)40 ( 60,0)(1,2)40 ( 60,0)(2,5)40 ( 6,140)(0,0)1 ( 6,120)(0,0)0 ( 6,60)(0,0)1 ( 6,40)(0,0)0 ( 75,180)(0,0)1 ( 75,160)(0,0)0 ( 75,20)(0,0)1 ( 75,0)(0,0)0 ( 114,100)(0,0)1 ( 114,80)(0,0)0 ( 20,103)(0,0)@xmath320 ( 60,143)(0,0)@xmath511 ( 20,23)(0,0)@xmath512 ( 60,-17)(0,0)@xmath513 ( 102,63)(0,0)@xmath514 ( 15,-15)(0,0)(a ) ( 160,60 ) ( 280,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,60)(100,60 ) ( 20,80)(60,40 ) ( 60,20)(100,60 ) ( 20,80)(100,80 ) ( 20,60)(60,20 ) ( 60,40)(100,80 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(2,-3)40 ( 20,60)(2,-1)40 ( 60,40)(2,1)40 ( 60,20)(2,3)40 ( 60,2)(0,0)@xmath511 ( 20,42)(0,0)@xmath320 ( 100,42)(0,0)@xmath512 ( 15,80)(0,0)1 ( 15,60)(0,0)0 ( 60,35)(0,0)1 ( 60,15)(0,0)0 ( 105,80)(0,0)1 ( 105,60)(0,0)0 ( 15,4)(0,0)(b ) ( 200,170)(10,10 ) ( 30,110)(18,58 ) ( 30,90)(0,0)@xmath0 ( 30,110)(0,0)@xmath0 ( 30,130)(0,0)@xmath0 ( 110,40)(18,38 ) ( 110,30)(0,0)@xmath0 ( 110,50)(0,0)@xmath0 ( 110,100)(18,38 ) ( 110,90)(0,0)@xmath0 ( 110,110)(0,0)@xmath0 ( 110,160)(18,38 ) ( 110,150)(0,0)@xmath0 ( 110,170)(0,0)@xmath0 ( 190,110)(18,58 ) ( 190,90)(0,0)@xmath0 ( 190,110)(0,0)@xmath0 ( 190,130)(0,0)@xmath0 ( 16,90)(0,0)3 ( 16,110)(0,0)2 ( 16,130)(0,0)1 ( 204,90)(0,0)3 ( 204,110)(0,0)2 ( 204,130)(0,0)1 ( 110,165)(0,0)1 ( 110,145)(0,0)0 ( 110,105)(0,0)1 ( 110,85)(0,0)0 ( 110,45)(0,0)1 ( 110,25)(0,0)0 ( 30,145)(0,0)@xmath320 ( 190,145)(0,0)@xmath511 ( 110,134)(0,0)@xmath515 ( 110,74)(0,0)@xmath516 ( 110,14)(0,0)@xmath517 ( 30,90)(110,170 ) ( 30,110)(110,170 ) ( 30,90)(110,110 ) ( 30,130)(110,110 ) ( 30,110)(110,50 ) ( 30,130)(110,50 ) ( 190,110)(110,150 ) ( 190,90)(110,150 ) ( 190,130)(110,90 ) ( 190,90)(110,90 ) ( 190,110)(110,30 ) ( 190,130)(110,30 ) ( 30,130)(2,1)80 ( 30,130)(4,1)80 ( 30,130)(2,-1)80 ( 30,130)(4,-5)80 ( 30,110)(2,1)80 ( 30,110)(1,0)80 ( 30,110)(4,-1)80 ( 30,110)(1,-1)80 ( 30,90)(4,3)80 ( 30,90)(1,0)80 ( 30,90)(2,-1)80 ( 30,90)(4,-3)80 ( 190,130)(-2,1)80 ( 190,130)(-4,1)80 ( 190,130)(-4,-1)80 ( 190,130)(-1,-1)80 ( 190,110)(-4,3)80 ( 190,110)(-1,0)80 ( 190,110)(-4,-1)80 ( 190,110)(-4,-3)80 ( 190,90)(-1,1)80 ( 190,90)(-4,1)80 ( 190,90)(-2,-1)80 ( 190,90)(-4,-3)80 ( 20,40 ) ( 20,40)(0,0 ) ( 10,20)(18,38 ) ( 10,10)(0,0)@xmath0 ( 10,30)(0,0)@xmath0 ( 280,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 170,50 ) ( 210,10 ) ( 250,50 ) ( 20,60)(100,60 ) ( 20,80)(60,40 ) ( 60,20)(100,60 ) ( 100,80)(180,80 ) ( 180,60)(260,60 ) ( 180,60)(220,20 ) ( 220,40)(260,80 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(1,0)80 ( 20,80)(2,-3)40 ( 20,60)(2,-1)40 ( 20,60)(1,-1)40 ( 60,40)(1,1)40 ( 60,40)(2,1)40 ( 60,20)(2,3)40 ( 180,80)(2,-3)40 ( 180,80)(1,-1)40 ( 180,60)(2,-1)40 ( 220,40)(2,1)40 ( 220,20)(1,1)40 ( 220,20)(2,3)40 ( 100,80)(4,-1)80 ( 100,60)(4,1)80 ( 100,60)(1,0)80 ( 180,80)(4,-1)80 ( 180,60)(4,1)80 ( 180,80)(1,0)80 ( 60,2)(0,0)@xmath511 ( 20,42)(0,0)@xmath320 ( 100,42)(0,0)@xmath512 ( 220,2)(0,0)@xmath514 ( 180,42)(0,0)@xmath513 ( 260,42)(0,0)@xmath518 ( 15,80)(0,0)1 ( 15,60)(0,0)0 ( 60,35)(0,0)1 ( 60,15)(0,0)0 ( 102,75)(0,0)1 ( 102,55)(0,0)0 ( 178,75)(0,0)1 ( 178,55)(0,0)0 ( 220,35)(0,0)1 ( 220,15)(0,0)0 ( 265,80)(0,0)1 ( 265,60)(0,0)0 ( 160,120)(0,0 ) ( 20,90)(18,38 ) ( 20,80)(0,0)@xmath0 ( 20,100)(0,0)@xmath0 ( 40,30)(18,38 ) ( 40,20)(0,0)@xmath0 ( 40,40)(0,0)@xmath0 ( 120,30)(18,38 ) ( 120,20)(0,0)@xmath0 ( 120,40)(0,0)@xmath0 ( 140,90)(18,38 ) ( 140,80)(0,0)@xmath0 ( 140,100)(0,0)@xmath0 ( 16,100)(0,0)1 ( 16,80)(0,0)0 ( 40,35)(0,0)1 ( 40,15)(0,0)0 ( 120,35)(0,0)1 ( 120,16)(0,0)0 ( 144,100)(0,0)1 ( 144,80)(0,0)0 ( 10,65)(0,0)@xmath320 ( 40,5)(0,0)@xmath511 ( 120,5)(0,0)@xmath512 ( 150,65)(0,0)@xmath513 ( 20,80)(140,100 ) ( 20,80)(120,40 ) ( 20,100)(40,40 ) ( 40,20)(140,100 ) ( 40,20)(120,40 ) ( 120,20)(140,80 ) ( 20,100)(6,-1)120 ( 20,100)(1,0)120 ( 20,100)(5,-3)100 ( 20,100)(5,-4)100 ( 20,100)(1,-4)20 ( 20,80)(1,0)120 ( 20,80)(5,-3)100 ( 20,80)(1,-2)20 ( 20,80)(1,-3)20 ( 40,40)(5,3)100 ( 40,40)(5,2)100 ( 40,40)(1,0)80 ( 40,40)(4,-1)80 ( 40,20)(5,3)100 ( 40,20)(1,0)80 ( 120,40)(1,3)20 ( 120,40)(1,2)20 ( 120,20)(1,4)20 * @xmath519 ( shown in figure [ fig : ik4 ] ) is composed of four variables with domains @xmath520 ( @xmath521 ) , and the following constraints : @xmath522 ( @xmath523 ) and @xmath524 ( @xmath525 ) . *
@xmath526 ( shown in figure [ fig : i4 ] ) is composed of four variables with domains @xmath527 , @xmath528 ( @xmath529 ) , and the following constraints : @xmath530 ( @xmath531 ) and @xmath532 ( @xmath533 ) .
* @xmath534 ( shown in figure [ fig : isat2delta](a ) ) is composed of five boolean variables and the following constraints : @xmath535 , @xmath536 , @xmath537 , @xmath538 , @xmath539 , @xmath540 .
* @xmath541 ( shown in figure [ fig : i5 ] ) is composed of five variables with domains @xmath542 ( @xmath529 ) , @xmath520 , and the constraints : @xmath543 ( @xmath529 ) and @xmath544 ( @xmath529 ) . in this instance
the variable order is @xmath545 .
* @xmath546 ( shown in figure [ fig : isat6 ] ) is composed of six boolean variables and the following constraints : @xmath547 , @xmath548 , @xmath549 , @xmath550 , @xmath551 , @xmath552 , @xmath553 . *
@xmath554 ( shown in figure [ fig : isatk4 ] ) is composed of four boolean variables and the following constraints : @xmath547 , @xmath536 and @xmath555 ( for @xmath556 , @xmath557 , @xmath558 , @xmath559 ) .
* @xmath560 ( shown in figure [ fig : isat2delta](b ) ) is composed of three boolean variables and the three inequality constraints : @xmath561 ( @xmath531 ) .
( 500,620)(0,-20 ) ( 0,500 ) ( 160,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,60)(4,-3)40 ( 60,30)(4,3)40 ( 5,10)(0,0)(a ) ( 165,500 ) ( 160,90)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 20,60)(4,1)80 ( 60,20)(1,1)40 ( 60,20)(-1,1)40 ( 60,40)(100,60 ) ( 4,80)(0,0)@xmath53 ( 4,60)(0,0)@xmath305 ( 116,80)(0,0)@xmath115 ( 116,60)(0,0)@xmath116 ( 74,40)(0,0)@xmath562 ( 74,20)(0,0)@xmath563 ( 90,10 ) ( 5,10)(0,0)(b ) ( 330,500 ) ( 150,90)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(60,30 ) ( 4,80)(0,0)@xmath53 ( 4,60)(0,0)@xmath305 ( 20,43)(0,0)@xmath336 ( 100,47)(0,0)@xmath2 ( 60,7)(0,0)@xmath145 ( 90,15 ) ( 5,10)(0,0)(c ) ( 0,400 ) ( 160,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,70)(60,40 ) ( 60,40)(100,70 ) ( 20,70)(4,-5)40 ( 60,20)(4,5)40 ( 5,10)(0,0)(d ) ( 165,400 ) ( 160,80)(0,0 ) ( 10,50 ) ( 50,20 ) ( 90,50 ) ( 20,70)(60,40 ) ( 60,15 ) ( 100,45 ) ( 60,40)(100,70 ) ( 90,20 ) ( 5,10)(0,0)(e ) ( 330,400 ) ( 160,80)(0,0 ) ( 10,50 ) ( 50,20 ) ( 90,50 ) ( 20,70)(60,50 ) ( 20,45 ) ( 60,13 ) ( 100,45 ) ( 60,30)(100,70 ) ( 44,50)(0,0)@xmath562 ( 44,30)(0,0)@xmath563 ( 90,15 ) ( 5,10)(0,0)(f ) ( 0,300 ) ( 160,70)(0,0 ) ( 10,30 ) ( 90,30 ) ( 20,40)(100,40 ) ( 20,60)(100,60 ) ( 4,60)(0,0)@xmath53 ( 4,40)(0,0)@xmath305 ( 80,15 ) ( 5,10)(0,0)(g ) ( 165,300 ) ( 160,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,70 ) ( 60,30)(100,70 ) ( 4,80)(0,0)@xmath53 ( 4,60)(0,0)@xmath305 ( 90,20 ) ( 5,10)(0,0)(h ) ( 330,300 ) ( 160,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,70)(100,80 ) ( 60,30)(100,80 ) ( 116,80)(0,0)@xmath115 ( 116,60)(0,0)@xmath116 ( 90,20 ) ( 5,10)(0,0)(i ) ( 0,200 ) ( 150,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,70 ) ( 4,80)(0,0)@xmath53 ( 4,60)(0,0)@xmath305 ( 20,42)(0,0)@xmath336 ( 60,7)(0,0)@xmath145 ( 100,47)(0,0)@xmath2 ( 90,15 ) ( 5,10)(0,0)(j ) ( 165,200 ) ( 150,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 4,80)(0,0)@xmath53 ( 4,60)(0,0)@xmath305 ( 116,80)(0,0)@xmath115 ( 116,60)(0,0)@xmath116 ( 100,44)(0,0)@xmath2 ( 60,7)(0,0)@xmath145 ( 90,8 ) ( 5,10)(0,0)(k ) ( 330,200 ) ( 150,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,70)(100,60 ) ( 60,30)(100,80 ) ( 116,80)(0,0)@xmath115 ( 116,60)(0,0)@xmath116 ( 20,47)(0,0)@xmath336 ( 60,7)(0,0)@xmath145 ( 90,15 ) ( 5,10)(0,0)(l ) ( 0,90 ) ( 150,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,20)(100,60 ) ( 4,80)(0,0)@xmath53 ( 4,60)(0,0)@xmath305 ( 44,40)(0,0)@xmath562 ( 44,20)(0,0)@xmath563 ( 90,15 ) ( 5,10)(0,0)(m ) ( 165,90 ) ( 150,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,70)(100,80 ) ( 60,30)(-1,1)40 ( 60,30)(4,3)40 ( 116,80)(0,0)@xmath115 ( 116,60)(0,0)@xmath116 ( 90,20 ) ( 5,10)(0,0)(n ) ( 330,120 ) ( 150,60)(0,-10 ) ( 10,10 ) ( 90,10 ) ( 4,40)(0,0)@xmath53 ( 4,20)(0,0)@xmath305 ( 116,40)(0,0)@xmath115 ( 116,20)(0,0)@xmath116 ( 20,20)(100,40 ) ( 125,22 ) ( 5,0)(0,0)(o ) ( 0,-10 ) ( 150,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 60,30)(-1,1)40 ( 20,70)(1,0)80 ( 60,30)(1,1)40 ( 5,10)(0,0)(p ) ( 165,-10 ) ( 150,80)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 60,30)(-1,1)40 ( 20,70)(100,70 ) ( 60,30)(100,70 ) ( 5,10)(0,0)(q ) ( 330,50 ) ( 150,60)(0,-10 ) ( 10,10 ) ( 90,10 ) ( 20,20)(1,0)80 ( 20,40)(100,40 ) ( 4,40)(0,0)@xmath53 ( 4,20)(0,0)@xmath305 ( 116,40)(0,0)@xmath115 ( 116,20)(0,0)@xmath116 ( 125,22 ) ( 5,0)(0,0)(r ) ( 330,-20 ) ( 150,60)(0,-10 ) ( 10,10 ) ( 90,20 ) ( 20,20)(4,1)80 ( 20,40)(1,0)80 ( 4,40)(0,0)@xmath53 ( 4,20)(0,0)@xmath305 ( 120,25 ) ( 5,0)(0,0)(s ) we illustrate four of these instances in figure [ fig : ik4 ] , figure [ fig : i4 ] , figure [ fig : i5 ] and figure [ fig : isatk4 ] . in figures representing csp instances , similarly to patterns , ovals represent variables , the set of points inside an oval the elements of the variable s domain , a dashed ( respectively , solid ) line joining two points represents the incompatibility ( respectively , compatibility ) of the two points . in order not to clutter up
figures representing instances , trivial constraints containing only positive edges ( i.e. corresponding to complete relations ) are not shown .
figure [ fig : badpatterns](a ) is a pattern which does not occur in the instance @xmath519 ( figure [ fig : ik4 ] ) .
similarly , figure [ fig : badpatterns](b ) is a pattern which does not occur in the instance @xmath564 ( figure [ fig : i4 ] ) , and the pattern in figure [ fig : badpatterns](c ) does not occur in instance @xmath534 .
figure [ fig : badpatterns](d ) , ( e ) and ( f ) are three patterns which do not occur in the instance @xmath541 ( figure [ fig : i5 ] ) .
the pattern ( known as @xmath565 ) shown in figure [ fig : badpatterns](d ) is , in fact , a tractable pattern @xcite , but the fact that it does not occur in @xmath541 ( an arc - consistent instance which has no solution ) shows that arc consistency is not a decision procedure for csp@xmath566 .
this instance was constructed using certain known properties of the pattern @xmath565 @xcite .
it can easily be verified that the three patterns figure [ fig : badpatterns](g ) , ( h ) , ( i ) do not occur in @xmath567 .
similarly , the four patterns in figure [ fig : badpatterns](j),(k),(l),(m ) do not occur in the instance @xmath554 ( figure [ fig : isatk4 ] ) . the instance @xmath568 is the problem of colouring a complete graph on three vertices with only two colours
it is arc consistent but clearly has no solution .
it is easy to verify that none of the six patterns in figure [ fig : badpatterns](n),(o),(p),(q),(r),(s ) occur in @xmath568 .
furthermore , trivially , no pattern on four or more variables occurs in @xmath568 and no pattern with three or more distinct values in the same domain occurs in @xmath568 . by lemma [ lem : not - ac ] , we know that if a pattern @xmath19 does not occur in any of the instances @xmath519 , @xmath526 , @xmath534 , @xmath569 , @xmath546 , @xmath554 , @xmath560 , then it is not ac - solvable
. let @xmath19 be any of the patterns shown in figure [ fig : badpatterns ] .
by lemma [ lem : hered ] , any pattern @xmath74 in which @xmath19 occurs is not ac - solvable . by the pattern in figure [ fig : badpatterns](g ) ,
a simple ac - solvable pattern can not contain two negative edges between the same pair of variables .
since instance @xmath568 contains only three variables and instance @xmath541 contains no triple of variables which have a negative edge between each pair of variables , an ac - solvable pattern can contain _ at most three variables and at most two negative edges_. thus to identify simple ac - solvable patterns we only need to consider patterns on at most three variables , at most two points per variable and with none , one or two negative edges .
furthermore , in the case of two negative edges these negative edges can not be between the same pair of variables . in this subsection
, we consider only patterns @xmath19 that have no associated structure ( i.e. with @xmath570 ) .
we prove the following characterisation of unstructured ac - solvable patterns .
[ thm : noorder ] if @xmath19 is a simple unordered pattern , then @xmath19 is ac - solvable if and only if @xmath19 occurs in the pattern lx ( figure [ fig : lx ] ) or in the pattern unordered(btp ) . by the discussion in section [ sec:5.1 ] , we only need to consider patterns @xmath19 with at most three variable , at most two points per variable and at most two negative edges ( with these edges not being between the same pair of variables ) .
we consider separately the cases of a pattern with 0 , 1 or 2 negative edges .
the only simple unordered pattern with no negative edges is the triangle of positive edges shown in figure [ fig : badpatterns](p ) and this pattern is not ac - solvable since it does not occur in @xmath568 .
let @xmath19 be a simple pattern with one negative edge @xmath571 ( between variables @xmath34 and @xmath59 ) and at most two points per variable .
if the domain of @xmath34 ( respectively , @xmath59 ) contains another point @xmath305 ( respectively , @xmath116 ) , then for @xmath100 ( respectively , @xmath572 ) to be non - mergeable , there must be a positive edge @xmath573 ( respectively , @xmath574 ) .
furthermore , for @xmath305 ( respectively , @xmath116 ) not to be a dangling point , it must belong to another positive edge .
any two distinct points in the domain of a third variable @xmath33 would be mergeable , so we can assume that @xmath19 has at most one point in the domain of @xmath33 . since this point is not a dangling point , it must be connected by positive edges to at least two points . by a simple exhaustive search
we can easily deduce that _ either _ @xmath19 is ( a subpattern of ) a triangle on three variables composed of one negative edge and two positive edges ( in which case @xmath19 occurs in the pattern lx shown in figure [ fig : lx ] ) , _ or _
one of the patterns shown in figure [ fig : badpatterns](a ) , figure [ fig : badpatterns](p ) or figure [ fig : badpatterns](s ) occurs in @xmath19 , in which case , by lemma [ lem : hered ] , @xmath19 is not ac - solvable .
let @xmath19 be a simple pattern containing exactly two negative edges @xmath575 ( between variables @xmath31 ) and @xmath571 ( between variables @xmath576 ) that meet at the point @xmath53 of variable @xmath33 .
suppose first that @xmath33 has no other point .
if @xmath19 does not occur in lx , then @xmath19 must have a positive edge between variables @xmath34 and @xmath59 which is either @xmath573 or @xmath577 where @xmath578 and @xmath579 . in the latter case , to avoid points @xmath580 ( respectively , @xmath581 ) being mergeable , @xmath19 must have the positive edge @xmath574 ( respectively , @xmath582 ) .
we can deduce that , if @xmath19 does not occur in lx , then one of the patterns figure [ fig : badpatterns](p ) or figure [ fig : badpatterns](q ) occurs in @xmath19 .
suppose now that @xmath19 has two points @xmath583 in the domain of variable @xmath33 . for @xmath583 not to be mergeable
, @xmath19 must have either the positive edge @xmath584 or the positive edge @xmath585 .
if it has both , then the pattern figure [ fig : badpatterns](d ) occurs in @xmath19 . if @xmath19 has just one , which without loss of generality we can suppose is the positive edge @xmath584 , then for @xmath563 not to be a dangling point , @xmath563 must belong to another positive edge @xmath586 ( where @xmath578 ) or @xmath587 ( where @xmath588 ) . in the latter case , for @xmath589 not to be mergeable , @xmath19 must also have the positive edge @xmath582 . in both cases ,
the pattern figure [ fig : badpatterns](s ) occurs in @xmath19 .
thus , if @xmath19 does not occur in lx , then , by lemma [ lem : hered ] , @xmath19 is not ac - solvable .
finally , let @xmath19 be a simple pattern containing exactly two negative edges @xmath575 ( between variables @xmath31 ) and @xmath590 ( between variables @xmath576 ) with two distinct points @xmath591 in the domain of variable @xmath33 . to avoid @xmath592 being mergeable
, @xmath19 must have a positive edge @xmath573 or @xmath574 .
without loss of generality , suppose @xmath19 has the positive edge @xmath573 .
if @xmath19 does not occur in unordered(btp ) , then at least one of the variables @xmath593 must have two distinct points .
if @xmath34 has a point @xmath594 , then for @xmath595 not to be mergeable , @xmath19 must have the positive edge @xmath582 .
similarly , if @xmath59 has a point @xmath596 , then @xmath19 must have the positive edge @xmath597 .
but then , to avoid dangling points , we have to add other positive edges to @xmath19 and we find that one of the patterns figure [ fig : badpatterns](a ) , figure [ fig : badpatterns](n ) , figure [ fig : badpatterns](p ) or figure [ fig : badpatterns](s ) occurs in @xmath19 , and so , by lemma [ lem : hered ] , @xmath19 is not ac - solvable . by lemma [ lem : occ - unordered ] , if @xmath19 occurs in unordered(btp
) , then @xmath19 occurs in btp , and thus is ac - solvable @xcite . in this subsection
we consider simple patterns @xmath19 which have no domain order , ( i.e. @xmath598 ) , but do have a partial order on the variables . we first require the following lemma . [ lem : p < ] if @xmath599 is a pattern whose only structure is a partial order on its variables and @xmath600 , then 1 .
@xmath599 is simple if and only if @xmath601 is simple .
@xmath599 is ac - solvable only if @xmath601 is ac - solvable .
the property of being simple is ( definition [ def : simple ] ) independent of any variable order , hence @xmath599 is simple if and only if @xmath601 is simple . by lemma [ lem : occ - unordered ]
, @xmath601 occurs in @xmath602 .
the fact that @xmath599 is ac - solvable only if @xmath601 is ac - solvable then follows from lemma [ lem : hered ] .
recall pattern lx@xmath603 from example [ ex : lx < ] that is obtained from the pattern lx ( figure [ fig : lx ] ) by adding the partial variable order @xmath121 .
lemma [ lem : p < ] allows us to give the following characterisation of variable - ordered ac - solvable patterns .
[ thm : varorder ] if @xmath19 is a simple pattern whose only structure is a partial order on its variables , then @xmath19 is ac - solvable if and only if @xmath19 occurs in the pattern lx@xmath603 ( example [ ex : lx < ] ) , the pattern btp@xmath125 ( figure [ fig : btp ] ) or the pattern invvar(btp@xmath125 ) . by lemma [ lem : p < ] and theorem [ thm : noorder ] , we only need to consider patterns @xmath601 occurring in lx or unordered(btp ) to which we add a partial order on the variables to produce a pattern @xmath19 .
we first consider the case of a pattern @xmath19 in which two negative edges that meet . by lemma [ lem : p < ] and theorem [ thm : noorder ] , for @xmath19 to be ac - solvable , @xmath601 must occur in lx . by theorem [ thm : lx ] and
example [ ex : lx < ] , we have that lx@xmath120 is ac - solvable .
let @xmath604 be the pattern composed of three variables @xmath605 and two negative edges which meet at a point @xmath606 ( i.e. @xmath604 is the pattern lx without its positive edges )
. then figure [ fig : badpatterns](e ) and lemma [ lem : inv ] tell us that placing any order between @xmath33 and @xmath34 or between @xmath33 and @xmath59 turns @xmath604 into a pattern which is not ac - solvable .
it follows from lemma [ lem : hered ] that an ordered pattern @xmath19 containing two negative edges that meet is ac - solvable if and only if @xmath19 occurs in @xmath607 .
now consider simple patterns @xmath19 which contain two negative edges that do not meet . by lemma [ lem : p < ] and
theorem [ thm : noorder ] , for @xmath19 to be ac - solvable , @xmath601 must occur in unordered(btp ) .
adding almost any partial variable order to unordered(btp ) produces a pattern which occurs in btp@xmath125 ( figure [ fig : btp](b ) ) or invvar(btp@xmath125 ) .
the only order for which this is not the case , is the total order @xmath608 ( or its inverse ) , where the variables @xmath605 are as shown in figure [ fig : btp](b ) .
let @xmath19 be any simple pattern on three variables @xmath605 , containing two negative edges ( between @xmath576 and @xmath593 ) that do not meet and with the variable order @xmath608 .
then the pattern shown in figure [ fig : badpatterns](f ) occurs in @xmath19 and hence , by lemma [ lem : hered ] , @xmath19 is not ac - solvable .
if @xmath19 is any simple pattern such that @xmath601 occurs in lx or unordered(btp ) and contains at most one negative edge , then @xmath19 occurs in the three - variable triangle pattern composed of one negative and two positive edges ; it is then easy to check that , whatever the ordering of is variables , @xmath19 occurs in btp@xmath125 or invvar(btp@xmath125 ) . in this subsection
we consider simple patterns @xmath19 with a partial order on domains but no ordering on the variables .
let emc@xmath609 be the no - variable - order version of the pattern emc depicted in figure [ fig : emc ] .
we prove the following characterisation of domain - ordered ac - solvable patterns .
[ thm : domorder ] if @xmath19 is a simple pattern whose only structure is a partial order on its domains , then @xmath19 is ac - solvable if and only if @xmath19 occurs in the pattern lx ( figure [ fig : lx ] ) , or the pattern emc@xmath609 , or the pattern invdom(emc@xmath609 ) . as in the proofs of theorem [ thm : noorder ] and [ thm : varorder ] , we only need to consider patterns on at most three variables , with at most two points per variable and with either no negative edges , one negative edge , two negative edges that meet or two negative edges that do not meet . however , we have more cases to consider than in theorem [ thm : noorder ] since patterns may now contain points , such as @xmath62 in the pattern emc shown in figure [ fig : emc ] , which would be a dangling point without the domain order @xmath610 . the only pattern with no negative edges and no mergeable points is the triangle of positive edges shown in figure [ fig : badpatterns](p ) which is not ac - solvable .
let @xmath19 be a simple pattern with one negative edge , at most two points per variable and just two variables .
if neither of the patterns shown in figure [ fig : badpatterns](o ) and figure [ fig : badpatterns](r ) occur in @xmath19 , then @xmath19 occurs in the pattern emc@xmath609 or in invdom(emc@xmath609 ) .
let @xmath19 be a simple pattern on three variables @xmath605 , with one negative edge @xmath571 ( between variables @xmath34 and @xmath59 ) and at most two points per variable .
if the domain of @xmath34 ( respectively , @xmath59 ) contains another point @xmath305 ( respectively , @xmath116 ) , then for @xmath100 ( respectively , @xmath572 ) to be non - mergeable , there must be a positive edge @xmath573 ( respectively , @xmath574 ) .
any two distinct points in the domain of variable @xmath33 would be mergeable , so we can assume that @xmath19 has exactly one point in the domain of @xmath33 . since this point @xmath562 is not a dangling point , it must be connected by positive edges to at least two points .
if none of the patterns in figure [ fig : badpatterns](a ) , figure [ fig : badpatterns](p ) and figure [ fig : badpatterns](s ) occurs in @xmath19 , then @xmath562 must belong to the two positive edges @xmath582 and @xmath611 , and no others . if neither of the patterns in figure [ fig : badpatterns](o ) or figure [ fig : badpatterns](r ) occurs in @xmath19 , then we can deduce that @xmath19 occurs in the pattern emc@xmath609 or in invdom(emc@xmath609 ) .
let @xmath19 be a simple pattern on three variables with at most two points per variable and with two negative edges that meet .
if @xmath19 has two points @xmath100 in the domain of the same variable together with an ordering @xmath612 , then one of the patterns in figure [ fig : badpatterns](h ) and figure [ fig : badpatterns](i ) ( or their domain - inversed version ) occurs in @xmath19 , and hence @xmath19 can not be ac - solvable .
this leaves only the case of unordered patterns @xmath19 . by the proof ( and in particular the part that deals with patterns containing exactly two negative edges that meet ) of theorem [ thm : noorder ]
, we can deduce that if @xmath19 is ac - solvable then it occurs in the pattern lx .
let @xmath19 be a simple pattern on three variables @xmath605 with at most two points per variable and with two negative edges @xmath571 ( between variables @xmath33 and @xmath34 ) and @xmath586 ( between variables @xmath34 and @xmath59 ) that do not meet ( i.e. @xmath613 ) .
if @xmath19 contains only these four points @xmath614 , then it necessarily occurs in emc@xmath609 .
if @xmath19 contains exactly five points , then without loss of generality , we can assume that there is a point @xmath615 in the domain of @xmath33 .
since @xmath100 are not mergeable , there must be a positive edge @xmath573 in @xmath19 .
if @xmath19 has a positive edge @xmath584 , then the pattern in figure [ fig : badpatterns](n ) occurs in @xmath19 ; if @xmath19 has a positive edge @xmath616 then the pattern in figure [ fig : badpatterns](s ) occurs in @xmath19 .
now , if the pattern in figure [ fig : badpatterns](o ) does not occur in @xmath19 , then whatever ordering is placed on @xmath100 and @xmath572 , @xmath19 occurs in the pattern emc@xmath609 or in invdom(emc@xmath609 ) .
if @xmath19 contains exactly six points , then there must be points @xmath617 in the domain of @xmath33 and @xmath618 in the domain of @xmath59 .
since both @xmath100 and @xmath619 , are not mergeable , there must be positive edges @xmath573 and @xmath620 .
if @xmath19 has a positive edge @xmath621 , then the pattern in figure [ fig : badpatterns](b ) occurs in @xmath19 ; if @xmath19 has a positive edge @xmath584 or @xmath582 , then the pattern in figure [ fig : badpatterns](n ) occurs in @xmath19 ; if @xmath19 has a positive edge @xmath616 or @xmath622 , then the pattern in figure [ fig : badpatterns](s ) occurs in @xmath19 .
but then in all other cases one of @xmath305 and @xmath562 is a dangling point unless @xmath19 has an order on both @xmath100 and @xmath619 .
but this then implies that at least one of the patterns in figure [ fig : badpatterns](m ) and figure [ fig : badpatterns](o ) occur in @xmath19 . in all these cases , by lemma [ lem : hered ] , @xmath19 is not ac - solvable . in this subsection
we consider the most general case of simple patterns @xmath19 which have a partial domain order and a partial variable order .
we prove the following characterisation of ac - solvable patterns with partial orders on domains and variables .
[ thm : order ] if @xmath19 is a simple pattern with a partial order on its domains and/or variables , then @xmath19 is ac - solvable if and only if @xmath19 occurs in one of the patterns lx@xmath603 , emc ( figure [ fig : emc ] ) , btp@xmath125 , btp@xmath122 ( figure [ fig : btp ] ) , btx ( figure [ fig : btx ] ) or bti ( figure [ fig : bti ] ) ( or versions of these patterns with inversed domain - order and/or variable - order ) .
let @xmath601 be the same pattern as @xmath19 but without the partial order on its variables .
if @xmath19 is ac - solvable , then , by lemma [ lem : occ - unordered ] , @xmath601 occurs in @xmath19 , and hence , by lemmas [ lem : occ - sup ] and [ lem : not - ac ] , @xmath623 is also ac - solvable .
thus , by theorem [ thm : domorder ] , @xmath19 must occur in either lx , emc@xmath609 or invdom(emc@xmath609 ) .
we consider the four cases : no negative edges , one negative edge , two negative edges that meet , two negative edges that do not meet in @xmath19 .
we have already seen in the proofs of theorems [ thm : noorder ] and [ thm : domorder ] that there are no simple ac - solvable patterns with only positive edges , so there remain three cases to consider .
( 380,100)(0,0 ) ( 0,0 ) ( 180,100)(0,0 ) ( 10,60 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 20,80)(4,-1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath145 ( 20,57)(0,0)@xmath336 ( 100,42)(0,0)@xmath2 ( 130,65 ) ( 5,80)(0,0)@xmath53 ( 115,80)(0,0)@xmath115 ( 115,60)(0,0)@xmath116 ( 15,5)(0,0)(a ) ( 200,0 ) ( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath145 ( 20,42)(0,0)@xmath336 ( 100,42)(0,0)@xmath2 ( 130,60 ) ( 5,80)(0,0)@xmath53 ( 5,60)(0,0)@xmath305 ( 115,80)(0,0)@xmath115 ( 115,60)(0,0)@xmath116 ( 15,5)(0,0)(b ) ( 380,100)(0,0 ) ( 0,0 ) ( 180,100)(0,0 ) ( 10,60 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath145 ( 20,57)(0,0)@xmath336 ( 100,42)(0,0)@xmath2 ( 130,65 ) ( 5,80)(0,0)@xmath53 ( 115,80)(0,0)@xmath115 ( 115,60)(0,0)@xmath116 ( 15,5)(0,0)(a ) ( 200,0 ) ( 180,100)(0,0 ) ( 10,50 ) ( 50,10 ) ( 90,50 ) ( 20,80)(100,80 ) ( 60,30)(100,60 ) ( 20,80)(4,-1)80 ( 20,60)(4,1)80 ( 20,80)(4,-5)40 ( 60,30)(4,5)40 ( 60,10)(0,0)@xmath145 ( 20,42)(0,0)@xmath336 ( 100,42)(0,0)@xmath2 ( 130,60 ) ( 5,80)(0,0)@xmath53 ( 5,60)(0,0)@xmath305 ( 115,80)(0,0)@xmath115 ( 115,60)(0,0)@xmath116 ( 15,5)(0,0)(b ) let @xmath19 be a simple pattern with exactly one negative edge .
if @xmath623 ( which also has one negative edge ) occurs in lx and is simple ( and hence has no dangling points ) then it must occur in a triangle @xmath624 consisting of one negative and two positive edges .
this triangle pattern @xmath624 occurs in btp or in invvar(btp ) whatever ordering we place on its three variables and hence the same is true of @xmath19 . if @xmath623 ( which has one negative edge and is unmergeable ) occurs in emc@xmath609 but not in the triangle @xmath624 , then @xmath623 occurs in one of the two patterns shown in figure [ fig : sub1emc ] ( or the domain - inversed versions of these patterns ) and includes the point @xmath116 ( in the case corresponding to figure [ fig : sub1emc](a ) ) or the points @xmath305 and @xmath116 ( in the case corresponding to figure [ fig : sub1emc](b ) ) . for @xmath116 ( respectively , @xmath305 ) not to be a dangling point in @xmath19 , we must have the order @xmath625 ( respectively @xmath626 ) in @xmath19 . if @xmath623 occurs in the pattern in figure [ fig : sub1emc](a ) and @xmath19 has the variable order @xmath627 ( or @xmath628 or @xmath629 )
, then @xmath19 occurs in btp@xmath122 .
if @xmath623 occurs in the pattern in figure [ fig : sub1emc](a ) and @xmath19 has the variable order @xmath630 , then @xmath19 occurs in invvar(bti ) . if @xmath623 occurs in the pattern in figure [ fig : sub1emc](a ) and @xmath19 includes the variable order @xmath631 and domain order @xmath625 , then the pattern in figure [ fig : badpatterns](j ) occurs in invvar(@xmath19 ) .
if @xmath623 occurs in the pattern in figure [ fig : sub1emc](a ) and @xmath19 includes the variable order @xmath632 and domain order @xmath625 , then the pattern in figure [ fig : badpatterns](c ) occurs in @xmath19 .
this covers all variable orderings of @xmath19 ( after taking into account the variable - inversed versions of each case ) when @xmath623 occurs in the pattern in figure [ fig : sub1emc](a ) .
now consider the case in which @xmath623 occurs in the pattern in figure [ fig : sub1emc](b ) .
if @xmath19 includes the variable order @xmath629 ( or @xmath138 ) together with the domain order @xmath626 and @xmath625 , then the pattern in figure [ fig : badpatterns](k ) occurs in @xmath19 .
if @xmath19 has the variable order @xmath628 then @xmath19 occurs in emc .
thus all one - negative - edge ac - solvable patterns occur in btp , bti or emc ( or their domain and/or variable - inversed versions ) .
let @xmath19 be a simple pattern with two negative edges that meet at a point .
@xmath623 necessarily occurs in the pattern lx .
let @xmath145 be the variable of @xmath19 where the two negative edges meet , and let @xmath633 be the other two variables . if @xmath19 includes the variable order @xmath629 ( or , by symmetry , the order @xmath138 ) , then the pattern in figure [ fig : badpatterns](e ) occurs in @xmath19 and hence @xmath19 is not ac - solvable .
if @xmath19 has the variable order @xmath628 , then @xmath19 occurs in the pattern @xmath607 ( the version of lx shown in figure [ fig : lx ] together with the variable order @xmath121 ) . by symmetry
, we have covered all possible cases .
finally , let @xmath19 be a simple pattern with two negative edges that do not meet at a point .
@xmath623 necessarily occurs in the pattern emc .
we distinguish two distinct cases : ( 1 ) @xmath19 occurs in the pattern in figure [ fig : sub2emc](a ) or ( 2 ) @xmath19 occurs in the pattern in figure [ fig : sub2emc](b ) and includes the point @xmath305 together with the order @xmath626 ( otherwise @xmath305 would be a dangling point ) .
we first consider case ( 1 ) .
if @xmath19 includes the order @xmath625 and @xmath630 , then the ( inversed domain - order version of the ) pattern in figure [ fig : badpatterns](l ) occurs in @xmath19 and hence @xmath19 is not ac - solvable .
if @xmath19 includes the order @xmath634 , then the pattern in figure [ fig : badpatterns](f ) occurs in @xmath19 .
if @xmath19 has the variable order @xmath635 and no domain order , then @xmath19 occurs in btp@xmath125 .
if @xmath19 has the variable order @xmath627 and the domain order @xmath625 , then @xmath19 occurs in btp@xmath122 .
all other patterns which fall in case ( 1 ) are covered by symmetry .
now we consider the case ( 2 ) .
first suppose that @xmath19 includes the domain order @xmath625 ( as well as @xmath626 ) .
if @xmath19 includes the variable order @xmath630 , then the ( domain - inversed version of the ) pattern in figure [ fig : badpatterns](l ) occurs in @xmath19 . if @xmath19 includes the variable order @xmath629 , then the pattern in figure [ fig : badpatterns](k ) occurs in @xmath19 .
if @xmath19 has the variable order @xmath628 , then @xmath19 occurs in emc . now
, suppose that @xmath19 does not include the domain order @xmath625 .
if @xmath19 includes the variable order @xmath627 , then the pattern in figure [ fig : badpatterns](j ) occurs in @xmath19 .
if @xmath19 includes the variable order @xmath632 , then the pattern in figure [ fig : badpatterns](c ) occurs in @xmath19 .
if @xmath19 has the variable order @xmath631 ( or @xmath630 or @xmath628 ) then @xmath19 occurs in btx .
if @xmath19 has the variable order @xmath629 then @xmath19 occurs in bti . by symmetry
we have covered all possible variable orderings of @xmath19 in case ( 2 ) .
we have identified 4 new tractable classes of binary csps . moreover , we have given a characterisation of all simple partially - ordered patterns decided by ac .
we finish with open problems . for future work
, we plan to study the wider class of unmergeable ordered patterns in which two points @xmath100 may be non - mergeable simply because there is an order @xmath612 on them . in the present paper ,
@xmath100 are mergeable unless they have different compatibilities with a third point @xmath115 .
is there a way of combining emc , btx and bti , since to find a solution after establishing arc consistency we use basically the same algorithm ?
any such generalisation will not be a simple forbidden pattern by theorem [ thm : order ] , but there is possibly some other way of combining these patterns .
are there interesting generalisations of these patterns to constraints of arbitrary arity , valued constraints , infinite domains or qcsp ?
btp has been generalised to constraints of arbitrary arity @xcite as well as to qcsps @xcite .
max - closed constraints have been generalised to vcsps @xcite .
infinite domains is an interesting avenue of future research because simple temporal constraints are binary max - closed @xcite . in this paper
, we only focused on classes of csp instances with totally ordered domains ( but defined by partially - ordered patterns ) .
however , the framework of forbidden patterns captures language - based csps with partially - ordered domains , such as csps with a semi - lattice polymorphism . in the future
, we plan to investigate classes of csp instances with partially - ordered domains .
d. a. cohen , m. c. cooper , g. escamocher , and s. ivn .
variable and value elimination in binary constraint satisfaction via forbidden patterns .
_ journal of computer and system sciences _ , 810 ( 7):0 11271143 , 2015 . t. feder and m. y. vardi . the computational structure of monotone monadic snp and constraint satisfaction : a study through datalog and group theory . _ siam journal on computing _ , 280 ( 1):0 57104 , 1998 . | characterising tractable fragments of the constraint satisfaction problem ( csp ) is an important challenge in theoretical computer science and artificial intelligence .
forbidding patterns ( generic sub - instances ) provides a means of defining csp fragments which are neither exclusively language - based nor exclusively structure - based .
it is known that the class of binary csp instances in which the broken - triangle pattern ( btp ) does not occur , a class which includes all tree - structured instances , are decided by arc consistency ( ac ) , a ubiquitous reduction operation in constraint solvers . we provide a characterisation of simple partially - ordered forbidden patterns which have this ac - solvability property .
it turns out that btp is just one of five such ac - solvable patterns .
the four other patterns allow us to exhibit new tractable classes .
[ defn]definition [ exmp]example supported by epsrc grant ep / l021226/1 ] supported by epsrc grant ep / l021226/1 and a royal society university research fellowship .
part of this work was done while the second author was visiting the simons institute for the theory of computing at uc berkeley ] ( 20,40 ) ( 20,40)(0,0 ) ( 10,20)(18,38 ) ( 10,10)(0,0)@xmath0 ( 10,30)(0,0)@xmath0 ( 20,40 ) ( 20,40)(0,0 ) ( 10,20)(18,28 ) ( 10,20)(0,0)@xmath0 |
a group @xmath1 is said to be _ residually finite _ if for any distinct @xmath7 there is a finite group @xmath8 and a homomorphism @xmath9 such that @xmath10 in @xmath8 . informally speaking this means that @xmath1 can be approximated by its finite quotients , in which case these quotients can be used to study the group @xmath1 .
for example , two classical theorems of malcev state that finitely presented residually finite groups have solvable word problem @xcite , and finitely generated residually finite groups are hopfian @xcite .
thus residual finiteness is a very basic property , so given any infinite group @xmath1 , one of the first questions we could ask about @xmath1 is whether it is residually finite .
the goal of this work is to prove residual finiteness of the outer automorphism group @xmath11 , where @xmath1 belongs to one of the following large classes of groups : * the class of virtually compact special groups ( in the sense of f. haglund and d. wise @xcite ) , * the class of fundamental groups of compact @xmath2-manifolds . before formulating the main results ,
let us recall some background of the problem .
a well - known theorem of g. baumslag @xcite asserts that for a finitely generated residually finite group @xmath1 its group of automorphisms , @xmath12 , is also residually finite . in @xcite
the second author and osin showed that in the case when @xmath1 has infinitely many ends , the same assumptions on @xmath1 ( finite generation and residual finiteness ) also imply that the group of outer automorphisms @xmath0 is residually finite .
however , this ` outer ' version of baumslag s result may not hold if @xmath1 is @xmath13-ended .
indeed , bumagin and wise @xcite proved that for any finitely presented group @xmath14 there is a finitely generated residually finite ( @xmath13-ended ) group @xmath1 such that @xmath15 .
a classical criterion for establishing residual finiteness of outer automorphism groups was discovered by grossman @xcite .
this criterion imposes stronger assumptions on the group : @xmath1 must be conjugacy separable and any pointwise inner automorphism of @xmath1 must be inner . recall that a group is _ conjugacy separable _
if for any pair of non - conjugate elements @xmath7 there exists a finite quotient @xmath8 , of @xmath1 , in which the images of @xmath16 are non - conjugate .
an automorphism @xmath17 is said to be _ pointwise inner _ if @xmath18 is conjugate to @xmath6 for each @xmath4 .
the set of all pointwise inner automorphism , @xmath19 , forms a normal subgroup of @xmath12
. in @xcite grossman proved the following theorem : if @xmath1 is a finitely generated conjugacy separable group such that @xmath20 then @xmath0 is residually finite . unfortunately , it is usually hard to show that a given group is conjugacy separable , as it is a much more delicate condition than residual finiteness ( for example conjugacy separability may not pass to subgroups or overgroups of finite index see @xcite ) .
one class of groups for which conjugacy separability is known is the class @xmath21 , virtual retracts of finitely generated right angled artin groups
see @xcite .
let @xmath22 denote the class of groups which contain a finite index subgroup from the class @xmath21 .
the significance of the class @xmath22 can be seen from the work of haglund and wise @xcite , who proved that every virtually compact special group belongs to this class ( recall that a group @xmath1 is said to be _ virtually compact special _ if @xmath1 contains a finite index subgroup which is the fundamental group of a compact special cube complex in the sense of @xcite ) .
the list of virtually compact special groups is quite large and includes most coxeter groups , @xmath13-relator groups with torsion and finitely generated fully residually free ( limit ) groups
see @xcite .
therefore the following theorem covers a wide range of groups : [ thm : avr ] for any group @xmath23 the group @xmath0 is residually finite .
as discussed above , theorem [ thm : avr ] together with results from @xcite yields if @xmath1 is virtually compact special then @xmath0 is residually finite .
it is worth mentioning that residual finiteness of @xmath0 , when @xmath1 is itself a finitely generated right angled artin group , was proved by the second author in @xcite and , independently , by charney and vogtmann in @xcite . on the other hand , there exist finitely generated groups @xmath24 such that @xmath24 is a subgroup of some finitely generated right angled artin group and @xmath25 is not residually finite .
such examples can be easily found using the modification of the rips s construction proposed by haglund and wise in @xcite .
it is easy to produce examples of groups from @xmath22 which possess pointwise inner automorphisms that are not inner ( one can simply take the direct product of the free group of rank @xmath26 with a finite group @xmath27 for which @xmath28 see @xcite for a construction of such finite groups ) . also , it is unknown whether all groups from @xmath29 are conjugacy separable , thus to prove theorem [ thm : avr ] one can not simply apply grossman s criterion , and a more elaborate approach is required .
the original application of grossman s criterion was the proof that @xmath30 is residually finite for any compact orientable surface @xmath31 ( see @xcite ) .
our second theorem extends this result to @xmath32 , where @xmath33 is a compact @xmath2-manifold : [ thm:3-man ] let @xmath1 be a group containing a finite index subgroup that is isomorphic to the fundamental group of some compact @xmath2-manifold @xmath33 .
then @xmath0 is residually finite . for fundamental groups of seifert
fibered @xmath2-manifolds ( with two exceptions ) , the residual finiteness of outer automorphism groups was proved by allenby , kim and tang @xcite . evidently , theorem [ thm:3-man ] can not be further generalized to @xmath34-dimensional manifolds , as it is well - known that for any finitely presented group @xmath1 there is a closed @xmath34-manifold @xmath33 such that @xmath35 .
it is well - known that , for a manifold @xmath33 , the group @xmath32 is closely related to the _ mapping class group _ ( i.e. , the group of isotopy classes of self - homeomorphisms ) @xmath36 of @xmath33 . for example , waldhausen @xcite proved that if @xmath33 is an irreducible orientable haken @xmath2-manifold with incompressible boundary such that @xmath33 is not a line bundle , then @xmath36 embeds into @xmath32 .
a similar statement when @xmath33 is non - orientable ( but still haken and @xmath37-irreducible ) was proved in @xcite .
if @xmath33 is not irreducible then the natural homomorphism @xmath38 will not , in general , be injective
see @xcite .
thus theorem [ thm:3-man ] yields suppose that @xmath33 is a compact irreducible orientable haken @xmath2-manifold with incompressible boundary that is not a line bundle .
then the mapping class group @xmath36 is residually finite .
thanks to a recent result of hamilton , wilton and zalesskii @xcite , stating that the fundamental group of any orientable compact @xmath2-manifold is conjugacy separable , we can prove theorem [ thm:3-man ] using an approach which is similar to the one employed in theorem [ thm : avr ] .
namely , in both theorems we use the techniques from geometric group theory , of groups acting on hyperbolic spaces , to prove a strong version of the fact that pointwise inner automorphisms are inner , which constitutes the second ingredient of grossman s criterion .
let us now discuss how the two theorems above are proved in more detail .
a group @xmath1 is called _ acylindrically hyperbolic _ if it admits a non - elementary acylindrical action on a hyperbolic metric space
see subsection [ subsec : acyl_hyp ] .
this definition was originally proposed by d. osin in @xcite , where he proved that the class of such groups coincides with other large classes , previously studied by bestvina and fujiwara @xcite , dahmani , guirardel and osin @xcite , hamenst " adt @xcite and the third author @xcite .
the class of acylindrically hyperbolic groups is rather extensive : it includes all non - elementary relatively hyperbolic groups , non-(virtually cyclic ) groups acting properly on proper cat(@xmath39)-spaces with at least one rank @xmath13-element ( see @xcite ) , mapping class groups of compact surfaces of genus at least @xmath13 , outer automorphism groups of free groups of rank at least @xmath26 ( see @xcite ) , many groups acting on simplicial trees ( see @xcite ) , etc . in @xcite
the second author and osin proved that for any compact irreducible @xmath2-manifold @xmath33 , @xmath40 is either acylindrically hyperbolic or virtually polycyclic , or @xmath33 is seifert fibered
. two elements @xmath41 of a group @xmath1 are said to be _
commensurable _ if there are @xmath42 and @xmath43 such that @xmath44 in @xmath1 . in this we case
we will write @xmath45 . otherwise , if @xmath6 and @xmath46 are non - commensurable , we will write @xmath47 . note that commensurability is an equivalence relation on the set of elements of @xmath1 .
given a subgroup @xmath24 of a group @xmath1 and a homomorphism @xmath48 , we will say that @xmath3 is _ commensurating _ if @xmath49 for all @xmath50 .
commensurating homomorphisms were introduced and studied by the second author and osin in the context of relatively hyperbolic groups in @xcite . to prove theorems [ thm : avr ] and [ thm:3-man ] we study such homomorphisms for an acylindrically hyperbolic group @xmath1 .
our main technical result ( theorem [ thm : comm ] in section [ sec : comm_hom ] ) , generalizing the work from @xcite , states that if @xmath24 is a sufficiently large subgroup of @xmath1 , then every commensurating homomorphism @xmath51 is induced by an inner automorphism of @xmath1 modulo a small perturbation ( which disappears when one restricts to some finite index subgroup of @xmath24 ) . in theorem [ thm : comm_end ]
we apply this result to the case when @xmath52 to obtain a characterization of commensurating endomorphisms of an acylindrically hyperbolic group .
in particular , we get the following [ cor : conj_aut ] for any acylindrically hyperbolic group @xmath1 , @xmath53 has finite index in @xmath19 . moreover , if @xmath1 has no non - trivial finite normal subgroups then @xmath20 .
sections [ sec : adding ] [ sec : techn ] of the paper develop the theory of acylindrically hyperbolic groups , which is necessary to prove the main technical theorem .
in particular , in section [ sec : adding ] we investigate the necessary and sufficient conditions for adding a subgroup to an existing family of hyperbolically embedded subgroups , generalizing osin s work from @xcite ( this has also been independently done by m. hull @xcite ) .
the proof of theorem [ thm : avr ] uses the following quite general statement : [ thm : comm_end - raag ] let @xmath24 be a non - abelian subgroup of a finitely generated right angled artin group . then every commensurating endomorphism @xmath54 is an inner automorphism of @xmath24 . while theorem [ thm : comm_end - raag ] employs the characterization of commensurating endomorphisms of acylindrically hyperbolic groups from theorem [ thm : comm_end ] , it is not a straightforward consequence of it , as not all subgroups of right angled artin groups are acylindrically hyperbolic .
the proof of theorem [ thm : comm_end - raag ] occupies section [ sec : raag - end ] . the final ingredient in the proofs of theorems [ thm : avr ] and [ thm:3-man ] is a new criterion for residual finiteness of outer automorphism groups
see proposition [ prop : new_crit ] in section [ sec : crit ] .
this criterion could be of independent interest : for example , it gives a short proof of the fact that @xmath55 is residually finite for any seifert fibered space @xmath56 , which was conjectured by allenby , kim and tang in @xcite see lemma [ lem : seifert ] . the last application of theorem [ thm : comm_end ] that we discuss here concerns normal endomorphisms .
we will say that an endomorphism @xmath57 is _ normal _ if @xmath58 for every normal subgroup @xmath59 . normal automorphisms
( with a slightly more restrictive definition requiring that @xmath60 for all @xmath59 ) have been studied by several authors before .
for instance , lubotzky @xcite showed that all normal automorphisms of free groups are inner .
a similar statement was proved for non - trivial free products @xcite and non - elementary relatively hyperbolic groups with trivial finite radical @xcite ; see @xcite for more results and references .
it is known that every acylindrically hyperbolic group @xmath1 contains a unique maximal finite normal subgroup ( see ( * ? ? ?
2.24 ) or lemma [ lem : e_g(h ) ] below ) .
this subgroup , sometimes called the _ finite radical _ of @xmath1 , will be denoted by @xmath61 ( @xmath62 is the notation used in @xcite ) , in line with lemma [ lem : e_g(h ) ] below . combining theorem [ thm : comm_end ] with the theory of dehn fillings for hyperbolically embedded subgroups , developed by dahmani , guirardel and osin in @xcite we show that almost all normal endomorphisms of acylindrically hyperbolic groups are commensurating , and so their structure is described by theorem [ thm : comm_end ] .
[ thm : norm_end ] let @xmath1 be an acylindrically hyperbolic group and let @xmath57 be a normal endomorphism .
then either @xmath63 or @xmath3 is commensurating .
in particular , if @xmath64 and @xmath65 then @xmath3 is an inner automorphism of @xmath1 . * acknowledgements . *
the authors would like to thank denis osin for helpful discussions .
we are also grateful to henry wilton for suggesting and discussing the applications of the main result to @xmath2-manifold groups .
in this subsection we fix the notation and recall some basic concepts that will be used throughout the paper .
let @xmath66 be a metric space . given a subset @xmath67 and @xmath68 , we denote by @xmath69 the closed @xmath70-neighborhood of @xmath71 , i.e. , @xmath72 similarly
, we denote by @xmath73 , the closed ball of center @xmath74 and radius @xmath70 .
recall that for @xmath75 , the hausdorff distance is given by @xmath76 an isometric action of a group @xmath1 on @xmath66 is _ metrically proper _ if for any bounded subset @xmath77 , the set @xmath78 is finite .
recall that a _ path _ in @xmath79 is a continuous function @xmath80\to { \mathcal{s}}$ ] , and the _ length _ of @xmath81 is @xmath82 the path @xmath81 is rectifiable if @xmath83 is finite .
we denote by @xmath84 and @xmath85 the initial and the final points of @xmath81 .
the metric @xmath86 is a _ length metric _ if for every @xmath87 ,
@xmath88 if the metric @xmath86 is a length metric , @xmath66 is called a _ length space_.
if the infinum above is always realized ( i.e. , for any @xmath89 there is a rectifiable path @xmath81 with @xmath90 ) , then @xmath66 is said to be a _
geodesic metric space_. let @xmath1 be a group .
suppose that @xmath91 is a set equipped with a map @xmath92 .
we will say that @xmath1 generated by @xmath91 if @xmath93 . the set @xmath91 will be called _ symmetric _ if @xmath94 in @xmath1 . in this case
one can define the _ cayley graph _ @xmath95 , of @xmath1 with respect to @xmath91 and @xmath96 , as the graph with vertex set @xmath1 and edge set @xmath97 , where the initial vertex of @xmath98 is @xmath6 and the final vertex is @xmath99 . note that this definition allows the cayley graph to have multiple edges joining two vertices .
when the map @xmath96 is clear we will abuse the notation and simply write @xmath100 instead of @xmath95 . given a word @xmath101 over @xmath91 , @xmath102 will denote the length of @xmath101 .
for any other word @xmath103 over @xmath91 , we will write @xmath104 to denote the graphical ( letter - by - letter ) equality between words @xmath101 and @xmath103 . if @xmath91 generates @xmath1 and @xmath4 then @xmath105 will denote the length of a shortest word over @xmath91 representing @xmath6 in @xmath1 .
we will denote by @xmath106 the _ graph metric _ on @xmath100 , that is @xmath106 is the metric of the geometric realization of the graph where all the edges are isometric to the unit interval .
thus if @xmath107 then @xmath108 . in the context of graphs
, we will consider combinatorial paths .
a _ combinatorial path _ in @xmath100 is a formal sequence @xmath109 where @xmath110 are edges and the initial vertex of @xmath111 is the terminal vertex of @xmath112 , @xmath113 . in this case
, the _ length _ @xmath83 of @xmath81 is @xmath114 ; @xmath115 will be the path , inverse to @xmath81 ( i.e. , @xmath116 , where @xmath117 is the the edge inverse to @xmath118 ) .
furthermore , @xmath119 and @xmath120 will denote the initial and the terminal vertices of @xmath81 respectively .
if @xmath81 is a combinatorial path in a labelled directed graph ( e.g. , a cayley graph ) , we will use @xmath121 to denote its label .
given a subgroup @xmath24 of a group @xmath1 and a subset @xmath122 , @xmath123 will denote the centralizer of @xmath124 in @xmath24 , and @xmath125 will denote the normalizer of @xmath24 in @xmath1 . we will also use @xmath126 to denote the normal closure of @xmath124 in @xmath1 . a geodesic metric space @xmath127 is called _ @xmath128-hyperbolic _
if for any geodesic triangle , every side of the triangle is contained in the @xmath128-neighborhood of the union of the other two sides .
a metric space is said to be _ hyperbolic _ if it is geodesic and @xmath128-hyperbolic for some @xmath129 .
a subset @xmath71 of @xmath79 is _ @xmath130-quasi - convex _ , for some @xmath131 , if for every geodesic path @xmath81 in @xmath79 with @xmath132 , one has @xmath133 .
a set is _ quasi - convex _ if it is @xmath130-quasi - convex for some @xmath131 .
the following observation is an easy exercise on the definitions : [ rem : qc_orbits ] suppose that @xmath8 is a subgroup of a group @xmath1 acting by isometries on some @xmath128-hyperbolic space @xmath66 . if the orbit @xmath134 is @xmath130-quasi - convex for some @xmath135 and @xmath136 then for any @xmath137 the orbit @xmath138 is @xmath139-quasi - convex , where @xmath140 .
if @xmath141 is another metric space , then a map @xmath142 is a _ quasi - isometric embedding _ if there exist @xmath143 and @xmath144 such that @xmath145 if the quasi - isometric embedding @xmath146 is _ quasi - surjective _ , i.e. , @xmath147 for some @xmath148 , then @xmath146 is said to be a quasi - isometry .
the spaces @xmath141 and @xmath66 are quasi - isometric if there exists a quasi - isometry @xmath149 .
we will say that a path @xmath81 in @xmath127 is a @xmath150-_quasi - geodesic _ for some @xmath151 if for any subpath @xmath152 of @xmath81 we have @xmath153 where @xmath154 is the length of @xmath152 and @xmath155 , @xmath156 are the initial and terminal points of @xmath152 respectively .
we now collect a series of well known facts about quasi - geodesic paths in hyperbolic spaces . [ qg ] for any @xmath157 , @xmath158 , @xmath159
, there exists a constant @xmath160 such that in a @xmath128-hyperbolic space any two @xmath150-quasi - geodesics with the same endpoints belong to the closed @xmath161-neighborhoods of each other .
two paths @xmath162 in a metric space @xmath163 are called @xmath164-_connected _ , if @xmath165 the paths @xmath81 and @xmath152 are _
@xmath164-close _ for some @xmath166 if @xmath81 is @xmath164-connected with either @xmath152 or @xmath167 .
the next lemma is a simplification of lemma 25 from @xcite .
basically it says that if some sides of a geodesic polygon are much longer than the rest , then there is a pair of the long sides having sufficiently long subsegments which travel close to each other .
[ ols ] let @xmath168 be a geodesic @xmath114-gon in a @xmath128-hyperbolic space whose sides @xmath169 are divided into two subsets @xmath14 , @xmath170 .
denote the total length of all sides from @xmath14 by @xmath130 and the total length of all sides from @xmath170 by @xmath171 , and assume that @xmath172 for some @xmath173 .
then there are two distinct sides @xmath174 , and @xmath175-close subsegments @xmath176 and @xmath177 of @xmath178 and @xmath179 , respectively , such that @xmath180 . for our purposes
we need the following version of the varc - milnor lemma .
[ the varc - milnor lemma ] [ lem : svarc - milnor ] let @xmath127 be a length space .
suppose that a group @xmath1 acts by isometries on @xmath181 and the action is cobounded .
then there exists a symmetric generating set @xmath91 of @xmath1 such that for any @xmath182 , the map @xmath183 is a quasi - isometry from @xmath184 to @xmath127 .
moreover if the action is metrically proper then @xmath91 can be chosen to be finite .
this is proved in ( * ? ? ?
* i.8.19 ) with the assumption that the action is metrically proper , which is only used to conclude that @xmath91 is finite . [
lem : proper ] if @xmath1 acts by isometries on a hyperbolic space @xmath66 , @xmath185 and @xmath186 then the following are equivalent : 1 .
the orbit @xmath134 is quasi - convex and the induced action of @xmath8 on @xmath79 is metrically proper ; 2 .
@xmath8 is generated by a finite set of elements @xmath187 and there exist @xmath188 s.t .
@xmath189 for all @xmath190 assume ( 1 ) .
let @xmath182 be such that @xmath191 is @xmath130-quasi - convex for some @xmath192 .
let @xmath193 be the induced length metric on @xmath194 , i.e. , @xmath195 is the infimum of the lengths of all the paths from @xmath196 to @xmath197 contained in @xmath194 .
since @xmath191 is @xmath130-quasi - convex in @xmath181 , the inclusion map @xmath198 is a quasi - isometric embedding .
note that the action of @xmath8 on @xmath194 is metrically proper , by isometries and cobounded .
hence varc - milnor lemma ( lemma [ lem : svarc - milnor ] ) implies the existence of some finite generating set @xmath187 of @xmath8 such that @xmath199 and @xmath200 are quasi - isometric .
since the natural inclusion of @xmath201 into @xmath127 is a quasi - isometric embedding , there exist @xmath202 and @xmath159 such that @xmath203 for all @xmath204 , implying that ( 2 ) holds .
now , assume ( 2 ) . for every @xmath205 we have @xmath206 , so that the induced action of @xmath8 on @xmath79 is metrically proper . to prove that the orbit @xmath134 is @xmath130-quasi - convex , for some @xmath136 , take any geodesic path @xmath81 in @xmath181 with endpoints in @xmath191
we are going to show that @xmath207 , where @xmath130 will be determined later . since @xmath8 is a group acting by isometries on @xmath79 , without loss of generality we can assume that @xmath208 .
define @xmath209 and choose @xmath204 with @xmath210 .
suppose that @xmath211 is a shortest word in @xmath212 representing @xmath6 .
let @xmath152 be the path obtained by concatenating the geodesic segments @xmath213 $ ] of length at most @xmath214 , for @xmath215
. then @xmath216 and @xmath217 .
we are now going to show that @xmath152 is a quasi - geodesic .
consider an arbitrary subpath @xmath218 of @xmath152 . by the construction of @xmath152 ,
there is a subpath @xmath219 of @xmath152 such that @xmath220 , @xmath221 , for some @xmath222 , @xmath223 , @xmath224 and @xmath225 .
then we have @xmath226 on the other hand , recalling ( 2 ) we get @xmath227 combining the two inequalities above with the fact that @xmath228 , we obtain @xmath229 thus the path @xmath152 is @xmath230-quasi - geodesic in @xmath79 .
let @xmath231 be the constant provided by lemma [ qg ] , so that @xmath81 is lies in the @xmath161-neighborhood of @xmath152 . since @xmath232 , we see that @xmath207 , where @xmath233 .
let @xmath234 be a hyperbolic metric space and let @xmath1 be a group acting on @xmath181 by isometries .
an element @xmath235 will be called _ loxodromic _ ( with respect to the action on @xmath79 ) , if for some @xmath236 the map @xmath237 , @xmath238 is a quasi - isometric embedding . by lemma [ lem : proper ] , this is equivalent to the requirements that the orbit @xmath239 is quasi - convex and the induced action of @xmath240 on @xmath79 is metrically proper .
an element @xmath241 enjoys the _ weak proper discontinuity _ condition ( or @xmath46 is a _ wpd element _ ) if for every @xmath242 and any @xmath243 , there exists @xmath244 such that @xmath245 wpd elements originally were introduced by bestvina and fujiwara in @xcite .
further in the text we will use @xmath246 to denote the set of all elements @xmath4 that are loxodromic wpd with respect to the action of @xmath1 on @xmath79 .
[ rem : powers ] an element @xmath247
if and only if @xmath248 for any @xmath249 . to see this , fix some @xmath250 . since @xmath251 is finite for all @xmath252 , @xmath46 is loxodromic if and only if @xmath253 is loxodromic . on the other hand , assuming that the element @xmath46 is loxodromic , ( * ? ? ?
* lemma 6.4 ) shows that @xmath46 is wpd if and only if @xmath253 is wpd .
it is an easy exercise to prove the following [ rem : tame - conj ] suppose that @xmath41 are conjugate elements of @xmath1 . if @xmath6 is loxodromic wpd then so is @xmath46 .
remark [ rem : powers ] and [ rem : tame - conj ] together imply that if @xmath254 and @xmath255 then @xmath247 .
recall that a group is said to be _ elementary _ if it contains a cyclic subgroup of finite index .
* lemma 6.5 , corollary 6.6)[lem : elemrem ] suppose that @xmath79 is a hyperbolic space , @xmath1 is a group acting on @xmath181 by isometries and @xmath256 a loxodromic wpd element .
then there is a unique maximal elementary subgroup @xmath257 that contains @xmath46 .
moreover , for every @xmath258 the following are equivalent : * @xmath259 ; * @xmath260 for some @xmath261 ; * @xmath262 for some @xmath263 .
furthermore , set @xmath264 .
then @xmath265 is a subgroup of index at most @xmath26 in @xmath266 , and there exists @xmath267 such that @xmath268 .
[ rem : elem_inter ] suppose that @xmath107 are loxodromic wpd elements for an action of @xmath1 on some hyperbolic space @xmath79 . * if @xmath269 then @xmath270 for some @xmath271 .
* if @xmath270 for some @xmath271 , then @xmath272 .
indeed , the first claim immediately follows from lemma [ lem : elemrem ] , stating that @xmath273 and @xmath274 .
the second claim can be quickly derived from part ( b ) of this lemma . in this subsection
we recall some basic concepts which were originally introduced by dahmani , guirardel and osin in @xcite .
let @xmath1 be a group and let @xmath275 be a family of subgroups of @xmath1 .
suppose that @xmath91 is a _ relative generating set _ of @xmath1 with respect to @xmath275 ( i.e. , @xmath276 ) .
note that @xmath91 could be infinite ; we also assume that it is symmetric , i.e. , @xmath277 in @xmath1 .
denote @xmath278 as discussed in subsection [ subsec : not ] , the disjoint union @xmath279 can be considered as a ` generating alphabet ' for @xmath1 , even though some letters from @xmath280 may represent the same element in @xmath1 .
let @xmath281 be the corresponding cayley graph of @xmath1 .
we also let @xmath282 denote the cayley graphs @xmath283 , which we think of as complete subgraphs of @xmath281 . by @xmath284
we denote the set of edges of @xmath285 .
[ def : maindef ] for every @xmath286 , the _ relative metric _
@xmath287 is defined as follows . for any @xmath288
, @xmath289 is the length of a shortest path in @xmath290 that joins @xmath6 to @xmath46 . if there is no such path , one sets @xmath291 .
it is easy to see that @xmath292 is an extended metric on @xmath293 .
[ def : he ] the family @xmath294 is _ hyperbolically embedded _ in @xmath1 with respect to @xmath91 ( notation : @xmath295 ) , if the cayley graph @xmath281 is hyperbolic and the metric space @xmath296 is locally finite for every @xmath297 .
we will say that @xmath294 is _ hyperbolically embedded _ in @xmath1 ( notation : @xmath298 ) if there exists a ( possibly infinite ) relative generating set @xmath91 , of @xmath1 with respect to @xmath294 , such that @xmath295 .
the concept of hyperbolically embedded subgroups has been introduced by dahmani , guirardel and osin in @xcite , where they also give an equivalent definition in terms of relative isoperimetric functions ( * ? ? ?
* theorem 4.24 ) ( see also ( * ? ? ?
6.4 ) or theorem [ thm : addq ] below for more equivalent conditions ) .
definition [ def : he ] immediately implies the following observation ( cf .
* rem . 4.26 ) ) : [ rem : subset - hyp_emb ] consider any subset @xmath299 and set @xmath300 . if @xmath301 then @xmath302 , where @xmath303
. the following lemma will be useful : [ lem : changegenset ] suppose that @xmath1 is a group , @xmath294 is a family of subgroups of @xmath1 and @xmath304 are relative generating sets of @xmath1 , with respect to @xmath294 , such that @xmath305 .
then @xmath306 if and only if @xmath307 let @xmath152 be a path in the cayley graph @xmath281 and let @xmath297 .
a non - trivial subpath @xmath81 of @xmath152 is called an @xmath293-_component _ , if all the edges of @xmath81 are labelled by letters from @xmath308 and @xmath81 is a maximal subpath of @xmath8 with this property .
component _ of @xmath152 is an @xmath293-component for some @xmath297 .
two @xmath293-components @xmath309 , @xmath310 of paths @xmath311 , @xmath312 , respectively , in @xmath281 are said to be _ connected _ if all vertices of @xmath309 and @xmath310 lie in the same left coset of @xmath293 in @xmath1 ( this is equivalent to the existence of an edge @xmath313 between any two distinct vertices of @xmath309 and @xmath310 with @xmath314 ) .
a component @xmath81 of a path @xmath152 is _ isolated _ if it is not connected to any other component of @xmath152 .
a path @xmath152 in @xmath281 is said to be _ without backtracking _ if all of its components are isolated .
below we formulate one of the main technical tools for working with hyperbolically embedded subgroups .
this statement is essentially proved in @xcite and is analogous to the relatively hyperbolic case ( cf .
* lemma 2.7 ) ) .
[ lem : omega ] suppose that @xmath275 is hyperbolically embedded in @xmath315 .
then there exists a finite subset @xmath316 and a constant @xmath317 such that the following holds .
let @xmath152 be a cycle in @xmath281 , @xmath318 be a collection of isolated components of @xmath152 and @xmath319 be the elements of @xmath1 represented by @xmath320 respectively .
then @xmath319 belong to the subgroup @xmath321 and the word lengths of @xmath322 s with respect to @xmath323 satisfy @xmath324 by the assumptions , for every @xmath297 , @xmath325 is locally finite . by ( * ? ? ?
* theorem 4.24 ) there @xmath1 has a strongly bounded relative presentation @xmath326 with respect to @xmath91 and @xmath275 with a linear relative isoperimetric function .
( recall that a group presentation @xmath326 is _ strongly bounded _ if @xmath327 and the set of letters from @xmath328 that appear in relators @xmath329 is finite ) .
let @xmath323 be the set of letters from @xmath330 that appear in words of @xmath331 .
now the lemma follows from ( * ? ? ? * lemmas 4.10 , 4.11 ) .
suppose that a group @xmath1 acts by isometries on a metric space @xmath66 .
following bowditch @xcite we will say that this action is _ acylindrical _
if for every @xmath332 there exist @xmath333 such that for any pair of points @xmath89 with @xmath334 one has @xmath335 comparing this with the definition of a loxodromic wpd element above , we immediately obtain [ rem : acyl->tame ] if a group @xmath1 acts acylindrically on a hyperbolic space @xmath79 then every loxodromic element of @xmath1 satisfies the wpd condition .
the action of @xmath1 on @xmath79 is _ non - elementary _ if for some ( equivalently , for any ) @xmath185 , the set of limit points @xmath336 of the orbit @xmath337 in the gromov boundary @xmath338 has at least @xmath2 points . in @xcite osin
proved the following theorem : [ thm : acyl - equiv_def ] for any group @xmath1 the following are equivalent : * @xmath1 admits a non - elementary acylindrical action on some hyperbolic space ; * there is a symmetric generating subset @xmath91 of @xmath1 such that the cayley graph @xmath100 is hyperbolic , the natural action of @xmath1 on @xmath100 is acylindrical and the gromov boundary @xmath339 has more than two points ; * @xmath1 is non - elementary and there exists a hyperbolic space @xmath79 such that @xmath1 acts on @xmath79 coboundedly and by isometries and @xmath340 ; * @xmath1 contains a proper infinite hyperbolically embedded subgroup . remark that in ( * ? ? ? * thm .
1.2 ) the statement ( iii ) of theorem [ thm : acyl - equiv_def ] is formulated in a weaker form , without the requirement for the action to be cobounded . however , ( ii ) clearly implies ( iii ) with this additional condition : assuming ( ii ) , one can simply take @xmath79 to be the cayley graph @xmath100 on which @xmath1 acts acylindrically ( the hypothesis that @xmath341 implies that the unique @xmath1-orbit of vertices in @xmath342 is unbounded , hence @xmath343 by remark [ rem : acyl->tame ] and the classification of acylindrical actions of groups on hyperbolic spaces obtained by osin in ( * ? ? ?
1.1 ) ) .
theorem [ thm : acyl - equiv_def ] allows one to say that a group @xmath1 is _ acylindrically hyperbolic _ if it satisfies one of the equivalent conditions ( i)(iv ) from its claim .
in this section we give necessary and sufficient conditions that allow to add a finite family of subgroups to the existing family of hyperbolically embedded subgroups .
this is analogous to osin s theorem @xcite , where a similar criterion was developed for relatively hyperbolic groups . in this subsection
we suppose that @xmath1 is a group , @xmath344 is a generating set of @xmath1 and @xmath345 is a collection of subgroups of @xmath1 such that @xmath346 .
[ lem : he->fgqi]for each @xmath347 there exists a finite generating set @xmath348 of @xmath349 and constants @xmath350 , @xmath351 such that @xmath352 for all @xmath353 . obviously , it is enough to prove the statement for @xmath354 .
. then @xmath356 generates @xmath1 and @xmath357 by remark [ rem : subset - hyp_emb ] .
let @xmath358 and @xmath359 be the finite subset and the constant provided by lemma [ lem : omega ] . consider any element @xmath360 .
since @xmath356 generates @xmath1 , we can let @xmath361 to be a shortest word over @xmath356 such that @xmath362 in @xmath1 .
therefore , in the cayley graph @xmath363 , there is a cycle @xmath152 with @xmath364 .
evidently , @xmath152 has exactly one @xmath365-component labelled by @xmath366 , hence it must be isolated in it .
consequently , by lemma [ lem : omega ] , @xmath367 and @xmath368 thus @xmath365 is generated by the finite set @xmath323 and the required inequality for the word lengths is satisfied .
[ lem : he->geomsep ] let @xmath369 and @xmath370 . for every @xmath68
, there exists @xmath371 such that diameter @xmath372 whenever @xmath373 or @xmath374 and @xmath375 ( here @xmath376 ) .
this is a straightforward consequence of definition [ def : he ] .
indeed , this definition implies that for any @xmath68 the set @xmath377 is finite , where @xmath378 is the relative metric on @xmath379 induced from the cayley graph @xmath380 with @xmath381 .
hence we can let @xmath382 .
now , for any distinct @xmath383 there are @xmath384 such that @xmath385 and @xmath386 for @xmath387 .
therefore @xmath388 for some @xmath389 . for @xmath387
, let @xmath390 be a word over @xmath391 of length at most @xmath70 representing @xmath392 in @xmath1 , and let @xmath393 be the letter representing @xmath394 . consider the path @xmath81 in @xmath380 starting at @xmath13 and labelled by the word @xmath395 . if @xmath396 or @xmath397 then the path @xmath81 has no edges from @xmath398 ( indeed , if @xmath374 but @xmath397 then @xmath399 ) and @xmath400 . since @xmath401 we see that @xmath402 , which implies that @xmath403 as required .
[ hyp : add ] throughout this section @xmath1 is a group , @xmath344 is a generating set of @xmath1 such that @xmath404 is @xmath128-hyperbolic , for some @xmath405 , and @xmath406 is a finite collection of subgroups of @xmath1 .
we use @xmath86 to denote the graph metric on @xmath407 .
we will consider the following properties for the family @xmath406 . 1 .
( _ @xmath408 is geometrically separated _ ) for every @xmath68 there exists @xmath409 such that for @xmath370 if @xmath410 then @xmath374 and @xmath411 ( here the distances are measured with respect to the graph metric @xmath86 on @xmath407 ) .
( _ finite generation _ ) for each @xmath412 , there exists a finite subset @xmath413 generating @xmath349 .
( _ quasi - isometrically embedded _ ) there exist @xmath202 and @xmath159 such that for any @xmath347 and all @xmath353 one has @xmath414 . [ rem : he->q1 - 3 ] under the previous notation , suppose that @xmath415 is hyperbolically embedded in @xmath416 , then by lemma [ lem : he->geomsep ] the family @xmath415 satisfies ( q1 ) , and by lemma [ lem : he->fgqi ] @xmath415 satisfies ( q2 ) and ( q3 ) with @xmath417 and @xmath418 .
the goal of this section is prove the converse result .
namely , if @xmath406 satisfy ( q1)(q3 ) then @xmath419 .
the next lemma says that if a pair of geodesics , labelled by elements of some @xmath349 s , have sufficiently long @xmath164-connected subpaths , then the endpoints of these geodesics belong to the same coset of @xmath349 .
[ lem : q2 ] in the notation [ hyp : add ] , suppose that @xmath406 satisfy ( q1)(q3 ) . for every @xmath420 there exists @xmath421 such that the following holds .
suppose that @xmath81 , @xmath152 are geodesic paths in @xmath407 such that @xmath121 ( resp .
@xmath422 ) represents an element of @xmath349 ( resp .
@xmath423 ) , and that there exist two @xmath164-close subpaths @xmath176 and @xmath177 of @xmath81 and @xmath152 . if @xmath424 , then @xmath374 and the label of an arbitrary path connecting any endpoint of @xmath81 with any endpoint of @xmath152 represents an element of @xmath349 . by lemma [ lem :
proper ] and remark [ rem : qc_orbits ] , there exists @xmath136 such that @xmath349 and @xmath423 ( considered as subsets of @xmath407 ) are @xmath130-quasi - convex .
let @xmath425 and @xmath409 be given by ( q1 ) ; set @xmath426 .
without loss of generality we can assume that @xmath176 and @xmath177 are @xmath164-connected , @xmath427 and @xmath428 .
. then there are vertices @xmath430 and @xmath431 such that @xmath432 , @xmath433 , @xmath434 and @xmath435 .
consequently , @xmath436 , thus @xmath437 .
one can also note that @xmath438 hence @xmath374 and @xmath439 by the assumption ( q1 ) , finishing the proof of the lemma .
[ not : phi ] in the notation [ hyp : add ] , suppose that @xmath406 satisfy ( q1)(q3 ) and @xmath440 .
let @xmath441 and @xmath442 .
we will denote by @xmath443 the graph metric on @xmath444 . for every @xmath445 and every @xmath353 , fix a shortest word @xmath446 over @xmath447 representing @xmath46 .
since @xmath407 and @xmath444 have the same vertex set @xmath1 , we can define a map @xmath448 just by replacing each edge @xmath313 , labelled by some @xmath353 in @xmath444 , with the ( unique ) path @xmath449 , labelled by @xmath446 and having the same initial and terminal vertices as @xmath313 in @xmath407 .
in particular , @xmath450 and @xmath451 for any path @xmath81 in @xmath444 .
our goal now is to show that if @xmath81 is a geodesic in @xmath452 then @xmath453 is a quasi - geodesic in @xmath454 . in order to do
so we will use the following lemma that deals with the situation when for some path @xmath81 the path @xmath453 is `` far '' from being a geodesic .
the conclusion is that in this case @xmath81 backtracks , i.e. , it goes through a coset of some @xmath349 twice .
[ lem : paths ] in the notation [ not : phi ] , there exists @xmath455 such that for all @xmath456 , @xmath457 and every path @xmath81 in @xmath444 satisfying @xmath458 and @xmath459 , if @xmath460 then there exist @xmath461 and two distinct edges @xmath462 and @xmath463 of @xmath81 that are labelled by letters from @xmath464 , so that all the endpoints of @xmath462 and @xmath463 belong to the same left coset of @xmath379 in @xmath1 .
let @xmath465 be the constant provided by lemma [ lem : q2 ] , where @xmath128 is the hyperbolicity constant of @xmath407 , and set @xmath466 .
we now fix @xmath467 , @xmath457 and a path @xmath81 in @xmath444 such that @xmath458 and @xmath459 .
suppose that @xmath468 where each @xmath469 and each @xmath470 is a ( possibly empty ) word in @xmath344 , in particular @xmath471 .
we have that @xmath472 let @xmath473 be a shortest word over @xmath344 representing the same element of @xmath1 as @xmath470 , @xmath474 , and let @xmath475 be the shortest word over @xmath344 representing the element @xmath476 , @xmath477 . consider the path @xmath152 in @xmath407 with the same endpoints as @xmath453 and with @xmath478 .
recall that @xmath479 , hence , in view of ( q3 ) , we have @xmath480 observe that @xmath481 can be written as the concatenation of geodesic paths @xmath482 in @xmath407 , where @xmath483 and @xmath484 .
let @xmath485 be a geodesic path in @xmath454 from @xmath156 to @xmath486 ; then @xmath487 . the polygon @xmath488 is a geodesic @xmath489-gon in @xmath407 and we partition its sides into two subsets @xmath490 and @xmath491 . by the assumptions we have that @xmath492 , and @xmath493 by .
choose a constant @xmath455 ( independent of @xmath218 and @xmath164 ) so that @xmath494 and suppose that @xmath460 .
since @xmath495 , all the conditions of lemma [ ols ] will then be satisfied , hence there will be @xmath496 , @xmath396 , and two @xmath497-close subsegments @xmath176 of @xmath498 and @xmath177 of @xmath499 such that @xmath500 .
it remains to apply lemma [ lem : q2 ] , claiming that there is @xmath461 such that @xmath501 and all the endpoints of the corresponding edges of @xmath452 belong to the same left coset of @xmath379 .
we are now ready to show that @xmath453 is a geodesic when @xmath81 is a geodesic .
the key observation , which allows us to use the previous lemma , is that a geodesic does not backtrack .
( we also apply this to subpaths of @xmath81 . )
[ lem : geod - qgeod ] in the notation [ not : phi ] , let @xmath455 be the constant provided by lemma [ lem : paths ] .
then for any geodesic path @xmath81 in @xmath444 , the path @xmath453 is @xmath502-quasi - geodesic in @xmath407 . as before ,
suppose that @xmath468 where each @xmath469 and each @xmath470 is a ( possibly empty ) word in @xmath344 .
consider any ( combinatorial ) subpath @xmath503 of @xmath453 in @xmath407 .
let us assume that @xmath504 starts with a suffix @xmath505 of @xmath506 and ends with a prefix @xmath507 of @xmath508 for some @xmath509 , @xmath510 , as the other cases can be treated similarly .
thus @xmath511 .
since @xmath505 is a geodesic word over @xmath447 for some @xmath347 , it represents an element @xmath512 and @xmath513 .
let @xmath152 be the path in @xmath444 with @xmath514 and @xmath515 .
then @xmath516 , which implies that @xmath517 and @xmath518 .
let @xmath519 be the subpath of @xmath152 with @xmath520 .
then @xmath519 is geodesic in @xmath444 , as it is also a subpath of @xmath81 , @xmath521 , and the endpoints of @xmath519 lie at distance at most @xmath13 from the corresponding endpoints of @xmath152 in @xmath444 .
set @xmath522 and @xmath523 .
then @xmath524 because @xmath525 since @xmath81 is geodesic in @xmath444 , all @xmath526-components of @xmath81 consist of single edges and no two components of @xmath81 are connected .
the latter also holds for @xmath152 since any component of @xmath152 is connected to a component of @xmath81 .
therefore lemma [ lem : paths ] implies that @xmath527 .
consequently , @xmath528 which shows that @xmath453 is @xmath502-quasi - geodesic in @xmath407 .
the following is the main result of this section .
it generalizes ( * ? ? ?
* theorem 1.5 ) .
[ thm : addq ] suppose that @xmath1 is a group , @xmath275 is a collection of subgroups of @xmath1 and @xmath91 is a relative generating set of @xmath1 with respect to @xmath275 , such that @xmath301 .
set @xmath529 . a family @xmath530 of subgroups of @xmath1 satisfies ( q1)(q3 ) if and only if the family @xmath531 is hyperbolically embedded in @xmath315 . the necessity is given by remark [ rem : he->q1 - 3 ] , so we only have to show that if @xmath530 satisfies ( q1)(q3 ) then @xmath532 .
since the set @xmath533 is finite , without loss of generality we can suppose that @xmath534 ( see lemma [ lem : changegenset ] ) .
using notation [ not : phi ] , let @xmath455 be the constant provided by lemma [ lem : paths ] .
take any @xmath535 and @xmath536 .
we will denote by @xmath537 the cayley graph @xmath538 and by @xmath285 the cayley graph @xmath539 .
the set of edges of @xmath537 and @xmath285 will be denoted @xmath540 and @xmath284 respectively . by @xmath541 and @xmath542
we denote the metrics on @xmath293 induced by graph metric on @xmath543 and @xmath544 , respectively .
the metric @xmath545 on @xmath349 is defined similarly .
we now break the proof in three claims .
* claim 1 : * for every @xmath445 the metric space @xmath546 is locally finite .
let @xmath547 and let @xmath309 be a shortest path from @xmath13 to @xmath548 in @xmath549 .
let @xmath313 be the edge of @xmath537 from @xmath550 to @xmath551 .
define @xmath81 to be the cycle in @xmath444 obtained by concatenating @xmath309 with @xmath313 .
suppose that @xmath552 .
then , by lemma [ lem : paths ] , there are @xmath553 and two distinct edges @xmath462 and @xmath463 of @xmath81 , labelled by some letters from @xmath464 , such that all endpoints of these edges belong to the same left coset @xmath554 . note that if @xmath555 then @xmath397 , as otherwise both @xmath462 and @xmath463 would have belonged to @xmath540 , but the only edge of @xmath81 from @xmath540 is @xmath313 .
in particular , @xmath556 and @xmath557 .
it follows that the subsegment of @xmath309 starting with @xmath462 and ending with @xmath463 can be substituted by a single edge @xmath558 , labelled by a letter from @xmath379 , so that the resulting path @xmath559 still lies in @xmath560 , connects @xmath13 with @xmath548 and @xmath561 , which contradicts the choice of @xmath309 .
therefore @xmath562 by ( q2 ) and ( q3 ) , for each @xmath563 there are only finitely many elements in @xmath349 of @xmath344-length at most @xmath564 .
this completes the proof of claim 1 .
* claim 2 : * for each @xmath297 the metric space @xmath565 is locally finite .
recall that by hypothesis @xmath566 is locally finite .
arguing by contradiction , suppose that for some @xmath456 , there exist infinitely many @xmath567 such that @xmath568 .
since @xmath566 is locally finite , there exists @xmath569 such that @xmath570 and @xmath571 .
let @xmath81 be a shortest path in @xmath572 from @xmath13 to @xmath573 , with @xmath574 .
notice that , by construction , @xmath575 is a path in @xmath543 . since @xmath576 , the inequality @xmath571 implies that @xmath577 .
hence , we can use lemma [ lem : paths ] to argue as above that the path @xmath81 can be shortened , yielding the required contradiction .
* claim 3 : * the graph @xmath452 is @xmath578-hyperbolic .
consider any geodesic triangle @xmath579 in @xmath452 and vertex @xmath580 . by lemma [ lem : geod - qgeod ] ,
the triangle @xmath581 is @xmath502-quasi - geodesic in @xmath407 .
let @xmath582 be the constant from lemma [ qg ] .
note that @xmath177 is also a vertex of @xmath583 , and any vertex @xmath584 , regarded as an element of @xmath1 ( and thus as a vertex of @xmath444 ) , lies within @xmath443-distance @xmath13 of a vertex of @xmath178 in @xmath444 , @xmath585 .
now , since the graph @xmath407 is @xmath128-hyperbolic , there is a vertex @xmath586 such that @xmath587 .
observe that @xmath588 by definition , hence there is a vertex @xmath589 such that @xmath590 thus the graph @xmath444 is @xmath578-hyperbolic , for @xmath591 .
claims 13 imply that the family of subgroups @xmath531 is hyperbolically embedded in @xmath1 , and so the theorem is proved .
the following corollary gives and alternative proof of ( * ? ? ?
* theorem 4.42 ) when the action of @xmath1 on @xmath79 is cobounded . during the work on this paper
the authors learned that this corollary was independently proved by hull in ( * ? ? ? * thm .
see also ( * ? ? ?
6.4 ) for other equivalent conditions .
[ cor : add_q ] let @xmath1 be a group acting by isometries on a hyperbolic space @xmath127 .
suppose that this action is cobounded and @xmath530 is a finite family of subgroups of @xmath1 .
fix any @xmath185 .
then the following are equivalent . 1 .
the family @xmath530 satisfies the conditions : 1 .
@xmath592 is quasi - convex and the induced action of @xmath349 on @xmath79 is metrically proper , @xmath593 ; 2 . for every @xmath594 there exists @xmath563 such that for @xmath4 if @xmath595 then @xmath374 and @xmath596 .
the family @xmath530 is hyperbolically embedded in @xmath597 , where @xmath344 is a generating set of @xmath1 provided by lemma [ lem : svarc - milnor ] . by the varc - milnor lemma ( lemma [ lem : svarc - milnor ] ) ,
the map @xmath598 is a @xmath1-equivariant quasi - isometry between @xmath1 , endowed with the metric from @xmath599 , and @xmath127 . in particular
, @xmath599 is hyperbolic and @xmath600 .
if we show that ( i)-(ii ) are equivalent to ( q1)(q3 ) , the result will follow from theorem [ thm : addq ]
. indeed , by lemma [ lem : proper ] and as @xmath599 is quasi - isometric to @xmath127 , the family @xmath530 satisfies ( i ) if and only if it satisfies ( q2 ) and ( q3 ) . on the other hand , ( ii )
is a restatement of ( q1 ) . as a corollary
we obtain the following statement ( cf .
* cor . 4.14 ) ) : [ cor : wpd - he ]
let @xmath1 be a group acting coboundedly on a hyperbolic space @xmath601 and let @xmath344 be a generating set of @xmath1 given by lemma [ lem : svarc - milnor ] . if @xmath602 is a collection of pairwise non - commensurable loxodromic wpd elements with respect to the action of @xmath1 on @xmath79 then @xmath603 .
fix @xmath604 .
since @xmath322 is loxodromic , there is @xmath182 such that the orbit @xmath605 is quasi - convex and the action of @xmath606 on @xmath79 is metrically proper .
thus the condition ( a).(i ) from corollary [ cor : add_q ] is satisfied .
the geometric separability condition ( a).(ii ) from corollary [ cor : add_q ] for the family @xmath607 is proved in ( * ? ? ?
hence , @xmath608 by corollary [ cor : add_q ] .
one can note that corollary [ cor : wpd - he ] resembles ( * ? ? ?
* theorem 6.8 ) .
the main difference is that we require the action to be cobounded , but because of this we are able to specify that the relative generating set @xmath344 comes naturally from the action of @xmath1 on @xmath79 ( this will be important for the rest of the paper ) .
similarly , theorem [ thm : addq ] can also be used to obtain the following strengthening of corollary [ cor : wpd - he ] : let @xmath1 be a group with a family of subgroups @xmath275 and a relative generating set @xmath91 ( with respect to @xmath275 ) , such that @xmath301 . set @xmath609 .
if @xmath602 is a collection of pairwise non - commensurable loxodromic wpd elements with respect to the action of @xmath1 on @xmath610 then the family @xmath611 is hyperbolically embedded in @xmath315 .
this section provides some technical geometric tools which will later be used to develop the theory of acylindrically hyperbolic groups similarly to the theory of relatively hyperbolic groups .
let @xmath1 be a group , let @xmath275 be a family of subgroups of @xmath1 and let @xmath91 be a symmetric relative generating set of @xmath1 with respect to @xmath275 . as usual , we set @xmath612 .
[ defn : w ] suppose that @xmath613 and @xmath323 is a finite subset of @xmath1 .
define @xmath614 to be the set of all words @xmath361 over the alphabet @xmath615 that have the following form : @xmath616 where @xmath617 , @xmath618 ( if @xmath619 then @xmath361 is the empty word ; if @xmath620 then @xmath621 ) , @xmath322 and @xmath622 are considered as single letters and * for every @xmath623 either @xmath624 or @xmath622 is the empty word , and for each @xmath625 , there exists @xmath626 such that @xmath627 ; * if @xmath628 then @xmath629 for each @xmath630 ; * @xmath631 , @xmath632 . finally , let @xmath633 be defined as the subset of all words @xmath634 such that @xmath635 and if @xmath636 then @xmath637 .
thus @xmath633 can be thought of as the set of cyclically reduced words from @xmath614 . for the remainder of this section assume that @xmath275 is hyperbolically embedded in @xmath315 .
choose the finite subset @xmath638 and the constant @xmath359 according to the claim of lemma [ lem : omega ] .
the following lemmas are taken from ( * ? ? ?
* section 6 ) , where they were established for the case when @xmath1 is hyperbolic relative to the family @xmath275 .
their proofs only use the combinatorial properties of the paths with labels from @xmath639 , together with the claim of ( * ? ? ?
* lemma 6.1 ) . using lemma [ lem : omega ] instead of the latter , the proofs transfer verbatim to the more general settings of the present paper .
[ lem : no_back ] let @xmath152 be a path in the cayley graph @xmath281 with @xmath640 and @xmath641
. then @xmath152 is without backtracking .
see the proof of ( * ? ? ?
* lemma 6.2 ) .
[ lem : regul ] let @xmath642 be a cycle in the cayley graph @xmath281 , such that @xmath643 .
suppose that @xmath644 and denote @xmath645 . then * if @xmath646 then no component of @xmath152 or @xmath647 is isolated in @xmath648 ; * if @xmath649 then each of @xmath152 and @xmath647 can have at most @xmath650 isolated components ; * if @xmath651 is the number of components of @xmath152 , then at least @xmath652 of components of @xmath152 are connected to components of @xmath647 ; and two distinct components of @xmath152 can not be connected to the same component of @xmath647 .
similarly for @xmath647 . see the proof of ( * ? ? ?
* lemma 6.3 ) .
[ lem : conseq - reg ] in the notations of lemma [ lem : regul ] , let @xmath653 and @xmath654 . for any positive integer @xmath655
there exists a constant @xmath656 such that if @xmath657 then there are @xmath655 consecutive components @xmath658 of @xmath152 and @xmath659 of @xmath660 , so that @xmath661 is connected to @xmath662 for each @xmath663 .
see the proof of ( * ? ? ?
* lemma 6.5 ) .
[ cor : wgeod ] if @xmath664 then every path @xmath81 in @xmath281 , with @xmath665 , is @xmath666-quasi - geodesic .
_ let @xmath81 be a path in @xmath281 such that @xmath667 . then @xmath668 where @xmath669 are the edges labelled by elements of @xmath328 , and @xmath670 are either trivial paths or edges labelled by elements of @xmath91 .
let @xmath671 be the element represented by @xmath672 in @xmath1 , @xmath632 . since any combinatorial subpath @xmath503 of @xmath81 still satisfies @xmath673 , to prove the lemma it is enough to show that @xmath674 , where @xmath152 is a geodesic path from @xmath120 to @xmath119 in @xmath281 .
note that @xmath675 .
if @xmath676 the claim is obvious , so we assume that @xmath677 , hence @xmath678 .
note that by the definition of @xmath81 , each @xmath178 is a component of @xmath81 .
let @xmath679 be the set of all indices @xmath412 such that @xmath178 is not connected to a component of @xmath152 in @xmath281 .
lemma [ lem : no_back ] implies that for each @xmath680 such @xmath178 is an isolated component of the cycle @xmath681 .
therefore , by lemma [ lem : omega ] , we have @xmath682 and @xmath683 however , since for @xmath684 , @xmath685 , we achieve @xmath686 .
let @xmath687 .
then @xmath688 , and for every @xmath689 the component @xmath178 of @xmath81 is connected to a component of @xmath152 , and no two such components of @xmath81 can be connected to the same component of @xmath152 ( as @xmath81 is without backtracking by lemma [ lem : no_back ] ) .
therefore , @xmath152 has at least @xmath690 distinct components and hence @xmath691 the main result of this section is the following .
[ thm : combtame ] suppose that @xmath692 .
take @xmath693 and @xmath359 according to the claim of lemma [ lem : omega ] .
let @xmath361 be any word from @xmath694 and let @xmath370 be the element represented by the word @xmath361
. then @xmath6 is loxodromic wpd with respect to the action of @xmath1 on @xmath281 .
suppose that @xmath695 .
observe that according to the definition of @xmath694 , for any @xmath696 , @xmath697 , hence any path labelled by @xmath698 in @xmath281 is @xmath666-quasi - geodesic by corollary [ cor : wgeod ] .
it follows that the map @xmath699 is a quasi - isometric embedding from @xmath700 to @xmath281 .
hence @xmath6 is loxodromic with respect to the action of @xmath1 on @xmath281 .
let us prove the wpd property . fix any @xmath68 and @xmath258 , and
choose @xmath701 so that @xmath702 , where @xmath703 .
suppose that @xmath704 satisfies @xmath705 then @xmath706 .
similarly , @xmath707 . choose words @xmath563 and @xmath708 over @xmath709 representing the elements @xmath146 and @xmath710 in @xmath1 with @xmath711 .
let @xmath712 be the cycle in @xmath281 starting at @xmath13 such that @xmath713 , @xmath714 , @xmath715 and @xmath716 .
let @xmath717 and @xmath718 be the lists of components of @xmath152 and @xmath660 in the order of their occurrence . by lemmas [ lem : no_back ] and [ lem : regul ] ( c ) there exists @xmath719 , @xmath720 , such that @xmath721 is connected to a component of @xmath722 of @xmath660 and @xmath721 is not connected to any component of @xmath218 .
let @xmath519 be a geodesic path in @xmath281 from @xmath723 to @xmath724 ( see figure [ fig:3 ] ) .
then @xmath725 and one can consider the cycle @xmath726 in @xmath281 , where @xmath311 is the initial segment of @xmath152 from @xmath727 to @xmath723 and @xmath728 is the terminal segment of @xmath647 from @xmath724 to @xmath729 .
then @xmath730 is the list of components of @xmath731 and if @xmath732 , one can apply lemmas [ lem : no_back ] and [ lem : regul ] ( c ) again to the cycle @xmath733 to find @xmath734 such that @xmath735 is connected to a component @xmath736 of @xmath311 and is not connected to a component of @xmath218 . in this case
we replace @xmath164 with @xmath737 and @xmath738 with @xmath739 .
thus , without loss of generality , we can further assume that @xmath740 .
it follows that @xmath741 ; similarly , @xmath742 .
let @xmath743 and @xmath46 be the elements of @xmath1 represented by the words @xmath744 , @xmath745 and @xmath746 respectively . then @xmath747 in @xmath1
. by construction , @xmath743 belong to the subgroup of @xmath1 generated by the finite set of elements @xmath748 and @xmath749 . on the other hand , note that if @xmath750 in @xmath1 then @xmath751 for some @xmath297 and @xmath519 is an isolated @xmath293-component of the cycle @xmath733 . hence we can use lemma [ lem : omega ] to conclude that @xmath752 and @xmath753
. therefore any element @xmath146 satisfying belongs to the subset @xmath754 which is finite as it is a product of finite subsets .
thus we have shown that the element @xmath6 is wpd .
in this section we fix a group @xmath1 and a hyperbolic space @xmath66 where @xmath1 acts by isometries and coboundedly . by lemma [ lem : svarc - milnor ]
, there is a generating set @xmath91 of @xmath1 such that @xmath184 is equivariantly quasi - isometric to @xmath79 .
it follows that @xmath370 is a loxodromic wpd element with respect to the action of @xmath1 on @xmath79 if and only if @xmath6 is a loxodromic wpd element with respect to the action of @xmath1 on @xmath755 .
thus , without loss of generality , we can work with either @xmath79 or @xmath755 .
the following observation will be useful .
[ lem : tameins ] suppose that @xmath344 is a subset of @xmath1 containing @xmath91 . if @xmath6 is a loxodromic wpd element with respect to the action of @xmath1 on @xmath599 then @xmath756 .
it is enough to show that @xmath6 is loxodromic wpd with respect to the @xmath1-action on @xmath755 .
since the action of @xmath6 is loxodromic on @xmath599 there exist @xmath202 and @xmath159 such that @xmath757 for all @xmath758 .
since @xmath759 for all @xmath256 , we get @xmath760 for all @xmath758 , which shows that @xmath6 acts as a loxodromic element on @xmath755 . similarly , since @xmath761 for any @xmath762
, it easily follows that any wpd element with respect to the action of @xmath1 on @xmath599 is also a wpd element with respect to the @xmath1-action on @xmath755 .
the purpose of this section is to develop basic tools for working with loxodromic wpd elements and producing new loxodromic wpd elements from a number of old ones .
[ lem : tame1 ] let @xmath294 be a family of subgroups of @xmath1 that is hyperbolically embedded in @xmath315
. set @xmath763 and take an arbitrary finite subset @xmath764 , @xmath635 .
consider any subset @xmath765 of @xmath1 such that @xmath766 and if @xmath767 then @xmath768 .
then there exists a finite subset @xmath769 such that for any @xmath770 and @xmath771 @xmath632 , the element @xmath772 has the following properties : 1 .
@xmath6 is a loxodromic wpd element with respect to the action on @xmath281 ; in particular , @xmath756 ; 2 .
@xmath6 is not commensurable with any element @xmath773 in @xmath1 . by lemmas [ lem : changegenset ] and [ lem : tameins ]
, we can replace @xmath91 with @xmath774 to assume that @xmath775 .
let @xmath776 and @xmath777 be the finite subset and the constant from the claim of lemma [ lem : omega ] .
we can then define the finite subset @xmath778 by setting @xmath779 .
now part ( a ) follows from the assumptions together with the claims of theorem [ thm : combtame ] and lemma [ lem : tameins ] . to prove part ( b ) notice that for every @xmath773 , the cyclic subgroup @xmath780 acts with bounded orbits on the cayley graph @xmath281 . on the other hand ,
all the orbits of @xmath240 are unbounded because @xmath6 is loxodromic by part ( a ) .
thus a non - zero power of @xmath6 can not be conjugate to a power of @xmath46 in @xmath1 , i.e. , ( b ) holds . applying lemma [ lem : tame1 ] in the special case when @xmath767 we obtain the following statement , generalizing ( * ? ? ?
* corollary 6.12 ) : [ cor : elemhe ] suppose that @xmath781
. then for any @xmath297 and @xmath782 , there exists a finite subset @xmath783 such that for all @xmath784 the element @xmath785 is loxodromic wpd with respect to the action of @xmath1 on @xmath786 ; in particular , @xmath787 .
recall that by lemma [ lem : elemrem ] , every @xmath756 belongs to the virtually cyclic subgroup @xmath788 and @xmath789 .
this lemma also implies that @xmath790 if and only if @xmath791 has infinite center .
[ lem : newtame1 ] let @xmath792 be a non - empty family of pairwise non - commensurable loxodromic wpd elements with respect to the action of @xmath1 on @xmath79 .
consider any subset @xmath793 such that @xmath794 and if @xmath767 then @xmath795 .
then there exists @xmath796 such that for arbitrary @xmath770 and @xmath797 with @xmath798 , @xmath632 , the element @xmath799 belongs to @xmath246 and is not commensurable with any @xmath800 , @xmath801 .
moreover , 1 . if @xmath767 then for every @xmath802 there exist @xmath803 such that @xmath804 ; 2 .
if @xmath805 and @xmath806 then @xmath807 and for every @xmath802 there exist @xmath803 satisfying @xmath808 .
recall that by corollary [ cor : wpd - he ] the family @xmath809 is hyperbolically embedded in @xmath315 . as before , in view of lemmas [ lem : changegenset ] and [ lem : tameins ] , we can assume that @xmath775 . set @xmath810 and let @xmath638 and @xmath777 be chosen according to lemma [ lem : omega ] .
take @xmath811 so that @xmath812 for any @xmath813 , whenever @xmath814 .
consider any @xmath815 with @xmath816 and @xmath817 , @xmath632 . by lemma [ lem : tame1 ] , @xmath254
and it is not commensurable with with any @xmath800 , @xmath801 .
so , it remains to prove claims ( i ) and ( ii ) .
consider any @xmath802 . by lemma [ lem : elemrem ]
, there exist @xmath818 and @xmath819 such that @xmath820 let @xmath821 be the constant provided by lemma [ lem : conseq - reg ] , where @xmath822 .
evidently we can take @xmath214 in to be large enough so that @xmath823 .
let @xmath101 be a word over @xmath824 representing @xmath197 , with @xmath825 , and let @xmath826 be the word from @xmath827 representing @xmath6 , where @xmath828 .
consider a cycle @xmath642 in @xmath281 , where @xmath829 , @xmath830 , @xmath831 and @xmath832 .
then @xmath833 , hence by lemma [ lem : conseq - reg ] there are @xmath834 consecutive components of @xmath152 connected to @xmath834 consecutive components of @xmath660 .
suppose , first , that @xmath767 .
then there is an @xmath835-component @xmath81 of @xmath152 connected to @xmath503 , an @xmath835-component of @xmath660 .
that is , there is a path @xmath519 in @xmath281 with @xmath836 , @xmath837 such that @xmath838 represents an element @xmath839 .
note that @xmath840 and @xmath841 .
let @xmath311 be the subpath of @xmath152 starting at @xmath842 and ending at @xmath843 ; let @xmath728 be the subpath of @xmath647 starting at @xmath844 and ending at @xmath845 . consider the cycle @xmath846 in @xmath281 . if @xmath847 we see that @xmath848 for some integer @xmath849 and @xmath850 for an integer @xmath851 .
therefore @xmath852 in @xmath1 .
recall that @xmath853 and the left hand side of the latter equality belongs to @xmath791 , hence @xmath854 .
similarly , in the case when @xmath855 we see that @xmath856 for some @xmath857 .
thus part ( i ) is proved . to prove part ( ii ) ,
assume that @xmath805 .
then three consecutive components @xmath309 , @xmath310 , @xmath858 of @xmath152 , with @xmath859 , @xmath585 , are connected to three consecutive components @xmath559 , @xmath860 , @xmath861 , of @xmath660 . since an @xmath862-component
can not be connected to an @xmath863-component if @xmath396 , we see that @xmath864 must be labelled by @xmath865 , @xmath585 . however , if @xmath855 , any triple of consecutive components of @xmath660 would be labelled by a cyclic permutation of the sequence @xmath866 , @xmath867 , @xmath868 , which can not give the sequence @xmath868 , @xmath867 , @xmath866
. thus @xmath847 , implying that @xmath869 .
since the latter is true for any @xmath870 we can conclude that @xmath790 . for the last claim of part ( ii ) ,
suppose that @xmath806 and choose consecutive components @xmath871 and @xmath309 of @xmath152 that are connected to consecutive components @xmath872 and @xmath559 of @xmath660 , so that @xmath178 and @xmath864 are @xmath862-components of the corresponding paths for @xmath873 .
it follows that for any path @xmath519 in @xmath281 joining @xmath874 with @xmath875 , @xmath838 represents an element @xmath876 .
since @xmath806 and @xmath847 the label of the subpath of @xmath647 from @xmath877 to @xmath845 represents a negative power of @xmath6 , and the label of the subpath of @xmath152 from @xmath842 to @xmath878 represents a positive power of @xmath6 .
thus there are integers @xmath879 and @xmath880 such that @xmath881 .
this completes the proof of the lemma .
[ lem : newtamenc ] let @xmath756 and @xmath882 . for any finite subset @xmath187 of @xmath1 ,
there exists @xmath883 such that @xmath884 and is not commensurable with any @xmath885 whenever @xmath886 . by corollary [ cor : wpd - he ] @xmath887 .
let @xmath888 be a maximal subset of pairwise non - commensurable elements such that each @xmath889 is loxodromic wpd with respect to the action of @xmath1 on @xmath79 and is not commensurable with @xmath6 . by corollary [ cor : wpd - he ] , @xmath890 , hence we can apply lemma [ lem : tame1 ] to find @xmath891 such that the element @xmath892 belongs to @xmath246 and is not commensurable with any element from the subset @xmath893 whenever @xmath886 .
suppose that there is an integer @xmath114 such that @xmath894 and @xmath895 is commensurable with some @xmath896 .
then @xmath897 , @xmath898 is not commensurable with any element of @xmath893 and @xmath899 by remarks [ rem : powers ] and [ rem : tame - conj ] .
this contradicts the maximality of @xmath900 .
thus the lemma is proved .
let @xmath24 be a subgroup of @xmath1 . in this subsection
we develop the theory of @xmath24-special elements .
many ideas and statements in this subsection are similar to those of ( * ? ? ?
* section 3 ) ( see also ( * ? ? ? * subsection 6.2 ) for the case @xmath52 ) .
[ lem : e_g(h ) ] let @xmath24 be a non - elementary subgroup of @xmath1 such that @xmath901 . then the subgroup @xmath902 is the unique maximal finite subgroup of @xmath1 normalized by @xmath24 .
if a finite subgroup @xmath903 is normalized by @xmath24 , then @xmath904 , where @xmath905 denotes the centralizer of @xmath765 in @xmath24 .
therefore for every @xmath50 and @xmath906 , there is @xmath249 such that @xmath907 .
hence , by lemma [ lem : elemrem](b ) , @xmath908 for all @xmath909 , thus @xmath910 .
let @xmath911 .
since @xmath24 is non - elementary , there exists @xmath912 .
then @xmath913 by remark [ rem : tame - conj ] .
if the intersection @xmath914 is infinite then , according to remark [ rem : elem_inter ] , there exist @xmath915 such that @xmath916 , which implies that @xmath917 ( by lemma [ lem : elemrem].(c ) ) .
this contradiction shows that @xmath918 is finite . the fact that @xmath919 is normalized by @xmath24 follows from its definition together with remark [ rem : tame - conj ] and lemma [ lem : elemrem ] : the latter two statements imply that for any @xmath920 and any @xmath921 , @xmath922 and @xmath923 . in the case
when @xmath52 , the statement of lemma [ lem : e_g(h ) ] is proved in ( * ? ? ?
* lemma 6.15 ) , where @xmath62 is used to denote the largest finite normal subgroup of @xmath1 , which is @xmath61 in our notation . set @xmath924 .
[ lem : lwpd+ ] let @xmath925 be a non - elementary subgroup such that @xmath901 .
for every finite subset @xmath926 there exists @xmath927 that is not commensurable in @xmath1 with any element of @xmath187 .
in particular , @xmath928 contains infinitely many pairwise non - commensurable ( in @xmath1 ) elements .
let @xmath929 be a maximal subset consisting of pairwise non - commensurable loxodromic wpd elements ( thus any element from @xmath930 is commensurable to some element from @xmath900 ) . if @xmath620 we understand that @xmath900 is empty .
take any element @xmath931 .
since @xmath24 is non - elementary , there exists @xmath932 and we can apply lemma [ lem : newtamenc ] , to find @xmath267 such that @xmath933 and @xmath934 is not commensurable with any element of @xmath900 .
applying this lemma two more times , we get elements @xmath935 such that @xmath800 is not commensurable to @xmath936 whenever @xmath937 . now , by lemma [ lem : newtame1 ] , there is @xmath938 such that the element @xmath939 belongs to @xmath940 and is not commensurable with any element from @xmath941 .
finally , if @xmath46 was commensurable to some @xmath942 then @xmath899 ( by remarks [ rem : powers ] and [ rem : tame - conj ] ) and @xmath898 would be non - commensurable with any @xmath943 , contradicting the choice of @xmath900 .
thus the lemma is proved .
[ lem : intersec_eg ] given two non - commensurable elements @xmath944 , there exists @xmath945 with the properties that @xmath46 is not commensurable with @xmath800 , @xmath946 , @xmath947 and @xmath948 , that is @xmath949 . by lemma [ lem : elemrem ] , remarks [ rem : powers ] and [ rem : elem_inter ] , we can replace @xmath800 with its power to assume that @xmath800 is central in @xmath862 , @xmath946 .
the subgroup @xmath950 is non - elementary because @xmath951 and @xmath952 are non - commensurable , hence , according to lemma [ lem : lwpd+ ] , there is @xmath953 that is not commensurable with @xmath951 and @xmath952 .
now , by lemma [ lem : newtame1 ] , we can choose @xmath954 so that the element @xmath955 belongs to @xmath956 , is not commensurable with @xmath951 and @xmath952 , and satisfies @xmath957 .
thus @xmath958 . but each of @xmath951 and @xmath952 commutes with @xmath959 , hence so does @xmath46 , and so lemma [ lem : elemrem ] yields that @xmath960 .
thus @xmath961 .
finally , note that @xmath46 has infinite order and @xmath962 by remark [ rem : elem_inter ] , which implies that @xmath963 .
therefore @xmath964 , as claimed .
[ lem : e_g(h)+ ] let @xmath925 be a non - elementary subgroup such that @xmath965
. then @xmath966 by lemma [ lem : lwpd+ ] , there exist two non - commensurable elements @xmath967 .
then @xmath968 is finite ( remark [ rem : elem_inter ] ) , and therefore @xmath969 is finite .
notice that the set @xmath928 is closed under @xmath24-conjugation and @xmath970 for any @xmath254 and any @xmath971 . hence @xmath24 normalizes the finite subgroup @xmath972 .
clearly @xmath973 . to obtain the desired equality , it remains to recall that @xmath919 is the unique maximal finite subgroup of @xmath1 normalized by @xmath24 by lemma [ lem : e_g(h ) ] .
let @xmath24 be a non - elementary subgroup of @xmath1 .
an element @xmath974 will be called _ @xmath24-special _
if @xmath756 , @xmath975 and @xmath976 ( i.e. , @xmath977 ) .
the set of all @xmath24-special elements will be denoted by @xmath978 .
the next statement is an analogue of ( * ? ? ?
* lemma 3.8.(ii ) ) .
[ lem : special ] let @xmath925 be a non - elementary subgroup such that @xmath901 .
then @xmath978 is non - empty .
let @xmath979 be the set of all elements @xmath980 such that @xmath266 is the direct product of @xmath780 with some finite subgroup @xmath981 of @xmath1 .
by lemma [ lem : lwpd+ ] there exists two non - commensurable elements in @xmath928 , and so , by lemma [ lem : intersec_eg ] and remark [ rem : elem_inter ] , the set @xmath979 is non - empty .
let @xmath982 be such that @xmath983 is minimal .
we will show that @xmath984 and thus @xmath985 .
notice that @xmath986 , as @xmath987 and @xmath981 is the unique maximal finite subgroup of @xmath266 by definition .
arguing by contradiction , assume that there exists a finite order element @xmath988 .
then , according to lemma [ lem : e_g(h)+ ] , there is @xmath989 such that @xmath990 .
if @xmath6 and @xmath46 are non - commensurable , using lemma [ lem : intersec_eg ] we can find @xmath991 such that @xmath992 and @xmath993 .
moreover , remark [ rem : elem_inter ] shows that @xmath994 is finite , and so it is contained in @xmath981 .
thus @xmath995 and , as @xmath996 , we have that @xmath997 , contradicting the minimality of @xmath998 .
it remains to consider the case when @xmath6 is commensurable with @xmath46 . by lemma [ lem : lwpd+ ]
, there exists @xmath999 non - commensurable with @xmath6 .
then , by lemma [ lem : intersec_eg ] , we can find @xmath991 such that @xmath1000 , @xmath1001 and @xmath146 is not commensurable with @xmath6 , and hence @xmath1002 .
moreover , since @xmath990 , we have that @xmath1003 as the torsion of @xmath1004 is exactly @xmath1005 .
then @xmath146 has the same properties as @xmath6 in the previous paragraph , which leads to a contradiction with the minimality of @xmath998 .
therefore @xmath984 and so @xmath1006 .
the following lemma is similar to ( * ? ? ?
* lemma 3.6 ) : [ lem : sg ] suppose that @xmath1007 , @xmath1008 and @xmath1009 .
then there exists @xmath1010 such that @xmath1011 for any @xmath1012 with @xmath1013 . by lemma [ lem : newtame1 ]
there exists @xmath1014 such that for all @xmath696 with @xmath1013 , @xmath1015 and this element is not commensurable with @xmath6 . part ( i ) of this lemma
also shows that @xmath1016 .
since @xmath6 is @xmath24-special and the subgroup @xmath1017 is finite ( by remark [ rem : elem_inter ] ) , we see that @xmath1018 .
recalling lemma [ lem : e_g(h ) ] , we obtain @xmath1019 thus @xmath1020 .
it remains to observe that @xmath1021 because both @xmath6 and @xmath196 belong to this centralizer by the assumptions .
hence @xmath1022 , as claimed .
[ prop : special]let @xmath24 be a non - elementary subgroup of @xmath1 with @xmath901 .
then @xmath1023 is generated by the set @xmath978 .
in particular @xmath1024 has finite index in @xmath24 . the proof is omitted , as it is identical to the proof of ( * ? ? ?
* proposition 3.3 ) , modulo lemmas [ lem : special ] and [ lem : sg ] .
the goal of this section is to prove several auxiliary statements that will help in establishing the claim of the main theorem [ thm : comm ] .
all of these statements are analogous to the ones from ( * ? ? ?
* section 4 ) . throughout this section @xmath1 will denote a group acting coboundedly by isometries on a hyperbolic space @xmath66 .
let @xmath91 be the generating set of @xmath1 given by lemma [ lem : svarc - milnor ] , so that @xmath755 is equivariantly quasi - isometric to @xmath79 .
the main technical tool is the following lemma , which generalizes ( * ? ? ?
* lemma 4.4 ) . roughly speaking
, it says that the products of large powers of wpd loxodromic elements are commensurable only in the `` obvious '' cases .
[ lem : cyclicperm ] let @xmath1025 , @xmath1026 , be a set of pairwise non - commensurable loxodromic wpd elements .
let @xmath765 be a subset of @xmath1 such that @xmath1027 ( e.g. , @xmath765 could be finite ) .
there exists @xmath1028 such that for any permutation @xmath130 of @xmath1029 and arbitrary elements @xmath1030 , @xmath632 , of infinite order , the following holds .
suppose that @xmath1031 is conjugate to @xmath1032 in @xmath1 , for some @xmath770 , @xmath1033 , and @xmath1034 @xmath1035 , @xmath1036 for all @xmath632 .
then @xmath1037 and there is @xmath1038 such that @xmath130 is a cyclic shift by @xmath164 , that is @xmath1039 for all @xmath1040 , and @xmath1041 when @xmath1042 , @xmath1043 .
this proof is very similar to the proof of ( * ? ? ?
* lemma 4.4(2 ) ) , using the appropriate references . by corollary [ cor : wpd - he ] the family @xmath809 is hyperbolically embedded in @xmath315 , and , by lemma [ lem : changegenset ] , we can enlarge @xmath91 to ensure that @xmath775 .
set @xmath810 and let @xmath638 and @xmath777 be chosen according to lemma [ lem : omega ] .
let @xmath14 be the finite subset of @xmath1 given by @xmath1044 .
first , let us show that for each @xmath412 there is @xmath1045 such that @xmath1046 whenever @xmath1047 is an element of infinite order and @xmath1048 . indeed , since @xmath1049 we see that every infinite order element @xmath1050 in fact belongs to the subgroup @xmath1051 .
note that the center of @xmath1051 has finite index in it ( e.g. , by the last assertion of lemma [ lem : elemrem ] ) .
hence all the elements of finite order form a finite normal subgroup @xmath1052 , and the quotient @xmath1053 is an infinite cyclic group , generated by the coset @xmath1054 , for some @xmath1055 .
since @xmath197 has infinite order and the set @xmath1056 is finite , there exists @xmath1045 such that @xmath1057 provided @xmath1058 .
then for any infinite order element @xmath1047 there is @xmath1059 with @xmath1060 .
thus for any @xmath1061 , @xmath1062 .
but if @xmath1048 then @xmath1063 and hence @xmath1064 , implying that @xmath1046 , as required .
now , set @xmath1065 .
choose arbitrary elements @xmath1066 and assume that @xmath1067 in @xmath1 , for some infinite order elements @xmath1068 , where @xmath130 is a permutation of @xmath1069 , and some @xmath1070 , @xmath1071 , @xmath1072 with @xmath1073 , @xmath625 .
then , for every @xmath1074 we have @xmath1075 let @xmath473 , @xmath1076 and @xmath470 be the letters from @xmath328 and from @xmath91 representing the elements @xmath1077 , @xmath1078 and @xmath1079 , @xmath632 , respectively . by our choice of @xmath1080 and @xmath1081 , the words @xmath1082 , and @xmath1083 are in @xmath1084 for all @xmath1012 . choose a shortest word @xmath979 over @xmath1085 representing @xmath1086 in @xmath1 .
set @xmath1087 and let @xmath1088 be the constant given by lemma [ lem : conseq - reg ] . take @xmath267 to be sufficiently large so that @xmath1089 and @xmath1090 . in the cayley graph @xmath281 equation
gives rise to a cycle @xmath642 , in which @xmath1091 , @xmath1092 , @xmath1093 , @xmath1094 , @xmath1095 , @xmath1096 , @xmath1097 . by construction , the paths @xmath152 and @xmath647 have exactly @xmath1098 and @xmath1099 components respectively .
suppose that @xmath1100 .
by lemma [ lem : regul ] ( c ) , at least @xmath1101 components of @xmath152 must be connected to components of @xmath647 , hence two distinct components of @xmath152 will have to be connected to the same component of @xmath647 , contradicting lemma [ lem : regul ] ( c ) . hence @xmath1102 .
a symmetric argument shows that @xmath1103 .
consequently @xmath1104 .
since @xmath1105 , we can apply lemma [ lem : conseq - reg ] to find @xmath834 consecutive components of @xmath152 that are connected to @xmath834 consecutive components of @xmath660 .
therefore there are consecutive components @xmath1106 of @xmath152 and @xmath1107 of @xmath660 such that @xmath179 is connected to @xmath1108 for each @xmath1109 , and @xmath1110 for @xmath632 , @xmath1111 ( see figure [ pic:2 ] ) .
therefore @xmath1112 , @xmath632 , @xmath1113 . from the form of @xmath1114
it follows that there is @xmath1115 such that @xmath1116 for @xmath1117 ( indices are added modulo @xmath651 ) . thus @xmath1118 . on the other hand , @xmath1119 and it has infinite order by the assumptions , hence @xmath1120 is commensurable with @xmath936 in @xmath1 by remark [ rem : elem_inter ] .
the latter yields that @xmath1121 for all @xmath1109 .
therefore @xmath130 is a cyclic shift ( by @xmath1122 ) of @xmath1069 . to prove the last claim of the lemma , note that the subpath @xmath1123 of @xmath660 between @xmath1124 and @xmath1125 is labelled by @xmath1126 . as we showed , the vertex @xmath1127 is connected to @xmath1128 by a path @xmath498 with @xmath1129 , and to @xmath1130 by a path @xmath1131 with @xmath1132 , @xmath813 ( here we use the convention that @xmath1133 ) . considering the cycle @xmath1134 we achieve the desired inclusion : @xmath1135 , @xmath632 .
let @xmath1007 be a non - elementary subgroup such that @xmath1136 .
the following three lemmas are analogues of ( * ? ? ?
* lemmas 4.5 , 4.6 , 4.7 ) respectively . the proofs
are exactly the same as in @xcite once one uses lemma [ lem : newtame1 ] instead of ( * ? ? ?
* lemma 4.4.(i ) ) , lemma [ lem : cyclicperm ] instead of ( * ? ? ?
* lemma 4.4.(ii ) ) and lemma [ lem : elemrem].(c ) instead of ( * ? ? ?
* lemma 2.4.(b ) ) .
[ lem : prod_of_three ] suppose that @xmath1137 is a homomorphism such that @xmath1138 for all @xmath909 .
then for any @xmath1139 , satisfying @xmath1140 for @xmath373 , there exists @xmath1141 such that for arbitrary @xmath1142 , with @xmath1143 , @xmath585 , and for @xmath1144 , one has @xmath254 and @xmath1145 , for some @xmath1146 and @xmath1147 .
[ lem : first_step ] let @xmath1148 be non - commensurable elements and let @xmath1149 .
there exists @xmath1150 such that the following holds .
suppose that @xmath1151 for some integers @xmath1152 with @xmath1153
. then @xmath1154 and @xmath1155 .
[ lem : spec - image ] assume that @xmath1156 and @xmath1157 is a homomorphism satisfying @xmath1158 for some @xmath267 and @xmath1159 .
then there is @xmath1160 such that @xmath1161 .
this section is dedicated to proving our main technical theorem : [ thm : comm ] let @xmath1 be a group acting coboundedly by isometries on a hyperbolic space @xmath79 .
let @xmath925 be a non - elementary subgroup of @xmath1 and let @xmath1162 be a homomorphism .
suppose that @xmath901 and @xmath1163 for all @xmath909 .
then there exists a set map @xmath1164 , whose restriction to @xmath1023 is a homomorphism , and an element @xmath1165 such that for every @xmath50 , @xmath1166 .
moreover , if @xmath1167 then @xmath1168 .
we need two auxiliary lemmas in order to prove the theorem . as usual , @xmath1 is a group acting isometrically and coboundedly on a hyperbolic space @xmath79 and @xmath1007 is a non - elementary subgroup with @xmath901 .
[ lem : towards_inner ] let @xmath1169 be a homomorphism such that @xmath1170 for all @xmath909 .
suppose that @xmath1171 is a triple of pairwise non - commensurable ( in @xmath1 ) elements with @xmath1172 and @xmath1173 .
then for any @xmath1174 there are @xmath1175 such that @xmath1176 and @xmath1177 satisfy the following properties : * @xmath1178 ; * the elements @xmath1179 are pairwise non - commensurable in @xmath1 ; * there exist @xmath1180 , @xmath1181 such that @xmath1182 and @xmath1183 .
let @xmath1184 be given by lemma [ lem : newtame1 ] applied to the set @xmath1185 and @xmath1186 .
choose @xmath1187 according to an application of lemma [ lem : prod_of_three ] to @xmath1188 and let @xmath1189 . by lemma [ lem : newtamenc ] , there is @xmath1190 such that @xmath1191 and this element is not commensurable with @xmath951 in @xmath1 .
it follows that the element @xmath1192 has infinite order , and thus it can not belong to the virtually cyclic subgroup @xmath835 . since @xmath951 is @xmath24-special , we can use lemma [ lem : sg ] to find @xmath1193 such that @xmath1194 whenever @xmath1195 .
take @xmath1196 so that @xmath1197 , and apply lemma [ lem : newtamenc ] to find @xmath1198 such that the elements @xmath1199 and @xmath1200 are non - commensurable in @xmath1 . by lemma [ lem : sg ]
we have @xmath1178 , and by lemma [ lem : newtame1 ] neither of these two elements is commensurable to any @xmath800 , @xmath585 . finally , using lemma [ lem : prod_of_three ] , one can conclude that there exist @xmath1181 , @xmath1201 such that @xmath1202 and @xmath1203 .
[ lem : from2toall ] let @xmath1169 be a homomorphism such that @xmath1170 for all @xmath909 .
suppose that there are two non - commensurable elements @xmath1204 such that @xmath1205 and @xmath1206 for some @xmath1180 .
then for every @xmath1008 there is @xmath1207 such that @xmath1161 .
consider any @xmath1008 .
if @xmath1208 then there is @xmath261 such that @xmath1209 because @xmath1210 . hence @xmath1211 and
then by lemma [ lem : spec - image ] , @xmath1161 for some @xmath1212 .
suppose now that @xmath1213 .
recall that @xmath1214 because this element is @xmath24-special .
now , combining lemmas [ lem : sg ] and [ lem : newtamenc ] , we can find some @xmath1215 such that @xmath1216 and @xmath655 is not commensurable with @xmath548 and @xmath1086 in @xmath1 . by lemma [ lem : towards_inner ]
, we can find @xmath1217 such that @xmath1218 , @xmath1219 , @xmath1220 and @xmath1221 for some @xmath1222 and @xmath1223 . by lemma [ lem : newtame1 ] ,
@xmath1224 for every sufficiently large @xmath719 .
hence @xmath1225 whenever @xmath164 is sufficiently large .
so , lemma [ lem : first_step ] shows that @xmath1226 .
thus @xmath1227 for some @xmath1228 and @xmath1212 .
this implies that @xmath1229 since @xmath1230 .
similarly one proves that @xmath1231 , and thus there is @xmath1232 such that @xmath1233 .
hence @xmath1234 and therefore @xmath1235 by lemma [ lem : elemrem].(b ) .
assume , first , that @xmath1236 .
if @xmath1237 has finite order , then @xmath1238 because @xmath1239 . hence , @xmath1240 contradicting the assumption that @xmath1241 . thus @xmath1242 must have infinite order , and so there are @xmath1243 such that @xmath1244 . since @xmath1205 and @xmath1206 , by lemma [ lem : spec - image ] there exist @xmath1245 such that @xmath1246 and @xmath1247 .
since @xmath1248 we obtain that @xmath1249 then for @xmath1250 we get that @xmath1251 , implying that @xmath1252 , which contradicts @xmath1253 .
therefore , @xmath1254 and , thus , @xmath1255 . by lemma [ lem : spec - image ]
, there exists @xmath1256 such that @xmath1257 . since @xmath1258 and @xmath1259 , @xmath1260
are fixed by @xmath1261 , we see that @xmath1262 . applying lemma [ lem : spec - image ] again , we find @xmath1263 such that @xmath1264 . finally , since @xmath1265 , we achieve that @xmath1266 as needed .
we are now ready to prove the main result of this section . since @xmath901 , by lemma [ lem : special ] there is at least one element @xmath1267 .
since @xmath24 is non - elementary and @xmath1023 has finite index in it , @xmath1268 is non - elementary itself . on the other hand
, @xmath835 is elementary by lemma [ lem : elemrem ] , hence there exists @xmath1269 . by lemma [ lem : newtamenc ]
, there is @xmath1270 such that @xmath1271 and @xmath1272 . using the same lemma again
, we can find @xmath1273 such that @xmath1274 and @xmath1275 for @xmath946 .
note that , by construction , @xmath1173 , so one can use lemma [ lem : towards_inner ] to find non - commensurable elements @xmath1204 such that @xmath1276 and @xmath1277 for some @xmath1181 and @xmath1201 .
let @xmath1278 be the homomorphism defined by @xmath1279 for all @xmath50 .
then @xmath1280 , @xmath1281 note that @xmath1282 for every @xmath909 .
by lemma [ lem : newtame1 ] , @xmath1283 if @xmath719 is large enough .
therefore @xmath1284 consequently , by lemma [ lem : first_step ] , @xmath1285 , thus @xmath1286 for some @xmath1287 and @xmath1212 . hence @xmath1288 ,
and since @xmath1289 , @xmath1290 .
let @xmath1291 and let the homomorphism @xmath1169 be defined by @xmath1292 for all @xmath50 . by construction
@xmath1293 now we are under the hypothesis of lemma [ lem : from2toall ] , claiming that for every @xmath1008 there exists @xmath1207 such that @xmath1161 . by proposition [ prop : special ]
, @xmath1023 is generated by @xmath978 , therefore for each @xmath1294 there is @xmath1295 such that @xmath1296 since the map @xmath1261 is a homomorphism , the map @xmath1297 will also be a homomorphism . by construction
, we have @xmath1298 for all @xmath1299 .
now we need to extend the homomorphism @xmath1297 to a set map @xmath1300 .
define @xmath1301 , @xmath1302 and @xmath1303 . since @xmath919 is normalized by @xmath24 ,
the centralizer @xmath1023 is a normal subgroup of @xmath24 .
consequently , for any @xmath1304 we have that @xmath1305 and @xmath1306 take an arbitrary @xmath50 . for any @xmath911 we have @xmath1307 , implying that @xmath1308 . since @xmath6 was an arbitrary element of @xmath1309
, we conclude that @xmath1310 ( see lemma [ lem : e_g(h ) ] ) . after defining the @xmath1311 for each @xmath50
, one immediately sees that @xmath1312 is a map with the required properties .
evidently the restriction of @xmath70 to @xmath1023 is the homomorphism @xmath1313 .
it remains to prove the last claim of the theorem .
assume that @xmath1167 .
consider any element @xmath1212 . by the above assumption , for any @xmath911
there is @xmath50 such that @xmath1314 . recalling and the definition of @xmath1261 we achieve @xmath1315 .
but @xmath1316 , therefore @xmath1317 hence , @xmath1318 for every @xmath911 ; consequently @xmath1319 .
the latter implies that @xmath1320 and since @xmath919 is finite , we conclude that @xmath1321 normalizes @xmath919 .
observe that @xmath1322 is a subgroup of @xmath1 because @xmath919 is normalized by @xmath24 ( see lemma [ lem : e_g(h ) ] ) .
for any @xmath50 we have that @xmath1323 ; thus @xmath1324 . since @xmath1325 and @xmath1167 , one gets @xmath1326
therefore @xmath1327 @xmath1328 , i.e. , @xmath1329 .
this finishes the proof of the theorem .
theorem [ thm : comm ] allows us to generalize corollaries 5.3 and 5.4 from @xcite .
[ cor : comm_equiv_cond ] let @xmath1 be group acting coboundedly and by isometries on a hyperbolic space @xmath79 .
suppose that @xmath925 is a non - elementary subgroup , with @xmath901 , and @xmath1162 is a homomorphism .
the following are equivalent : 1 .
@xmath3 is commensurating ; 2 .
@xmath1330 for every @xmath911 ; 3 .
there is a set map @xmath1164 , whose restriction to @xmath1023 is a homomorphism , and an element @xmath1165 such that for every @xmath370 , @xmath1331 . in particular , if @xmath1332 then every commensurating homomorphism from @xmath24 to @xmath1 is the restriction to @xmath24 of an inner automorphism of @xmath1 .
\(a ) implies ( b ) by definition , and ( b ) implies ( c ) by theorem [ thm : comm ] .
it remains to show that ( c ) implies ( a ) .
indeed , let the homomorphism @xmath3 satisfy ( c ) , and let @xmath6 be an arbitrary element of @xmath24 .
thus @xmath1331 for some @xmath1165 and @xmath1333 . since @xmath919 is a finite subgroup of @xmath1 normalized by @xmath24 , the subgroup @xmath1023 is normal and of finite index in @xmath24 .
set @xmath1334 and @xmath1335 .
it follows that @xmath1336 and @xmath1337 in @xmath1 by the assumptions of ( c ) .
therefore @xmath1338 hence @xmath1339 for all @xmath1340 , as required .
an application of the above corollary to the case @xmath52 gives rise to the following characterization of commensurating endomorphisms , which is very similar to the result for relatively hyperbolic groups from ( * ? ? ?
1.4 ) : [ thm : comm_end ] let @xmath1 be an acylindrically hyperbolic group .
an endomorphism @xmath1341 is commensurating if and only if there is a set map @xmath1342 , whose restriction to @xmath1343 is a homomorphism , and an element @xmath1165 such that @xmath1331 for every @xmath370 . in particular , if @xmath64 then every commensurating endomorphism is an inner automorphism of @xmath1 . by theorem [ thm : acyl - equiv_def ] , since @xmath1 is acylindrically hyperbolic , it is non - elementary and admits a cobounded action on a hyperbolic space @xmath79 such that @xmath340 .
now the claim follows from corollary [ cor : comm_equiv_cond ] applied to the case when @xmath52 .
[ rem : f_many ] if @xmath1 is a finitely generated acylindrically hyperbolic group then theorem [ thm : comm_end ] easily implies that @xmath53 has finite index in the group @xmath1344 of all commensurating automorphisms of @xmath1 . on the other hand , it is not difficult to show that this is not true for @xmath1345 , the direct product of the free group of countably infinite rank and the cyclic group of order @xmath26 ( in fact this group has uncountably many commensurating automorphisms ) .
the above remark shows that to establish corollary [ cor : conj_aut ] we need to work a bit more since the group @xmath1 may not be finitely generated ( however , the proof is very similar to that of ( * ? ? ?
5.4 ) ) . again , by theorem
[ thm : acyl - equiv_def ] , @xmath1 is non - elementary and admits a cobounded action on a hyperbolic space @xmath79 such that @xmath340 . applying corollary [ cor : comm_equiv_cond ] to the case when @xmath52 , we see that for any automorphism @xmath1346 , there exist @xmath1347 and a map
@xmath1348 such that @xmath1349 for each @xmath4 .
take any element @xmath1350
. then @xmath46 commutes with @xmath1351 , and , consequently , @xmath1352 where @xmath1353 .
now , since @xmath1354 is a pointwise inner automorphism of @xmath1 , there is @xmath1355 such that @xmath1356 .
hence @xmath1357 , i.e. , @xmath1358 , hence @xmath1359 .
consequently , we have @xmath1360 , which implies that @xmath1361 .
since the latter holds for any @xmath1350 , it follows from proposition [ prop : special ] that @xmath1362 , where @xmath1363 .
note that @xmath1364 , hence there are @xmath1365 such that @xmath1366 . for any @xmath4
there are @xmath1367 and @xmath1368 such that @xmath1369 , and one has @xmath1370 therefore @xmath1371 , i.e. , the map @xmath70 is uniquely determined by the images of @xmath1372 . since @xmath1373 , the automorphism @xmath1346 , up to composition with an inner automorphism of @xmath1 ,
is completely determined by the finite collection of elements @xmath1374 , and since @xmath61 is finite , we can conclude that @xmath1375 .
finally , if @xmath64 we have @xmath1376 for all @xmath4 , that is @xmath1377 .
combining grossman s criterion with corollary [ cor : conj_aut ] , we obtain the following let @xmath1 be a finitely generated acylindrically hyperbolic group . if @xmath1 is conjugacy separable and contains no non - trivial finite normal subgroups then @xmath0 is residually finite . in ( * ? ? ?
1.6 ) caprace and the second author showed that any pointwise inner automorphism of a finitely generated coxeter group @xmath361 is inner .
theorem [ thm : comm_end ] can be used to say much more in the case when @xmath361 is acylindrically hyperbolic .
[ lem : cox - ahyp ] suppose that @xmath361 is a finitely generated infinite irreducible non - affine coxeter group .
then @xmath361 is acylindrically hyperbolic and @xmath1378 .
the assumptions imply that @xmath361 is not virtually cyclic and @xmath361 contains a rank @xmath13 isometry for the natural action on the associated davis cat(@xmath39 ) complex
see @xcite .
therefore @xmath361 is acylindrically hyperbolic by @xcite .
it remains to note that @xmath1378 because any finite normal subgroup of a coxeter group is contained in a finite normal parabolic subgroup , but an infinite irreducible coxeter group can not have any proper normal parabolic subgroups ( the normalizer of a parabolic subgroup @xmath1379 is itself a parabolic subgroup , which is isomorphic to the direct product @xmath1380 , where @xmath563 is the orthogonal complement of @xmath168 in @xmath361
see @xcite ) .
a combination of lemma [ lem : cox - ahyp ] with theorem [ thm : comm_end ] immediately yields the following : [ cor : cox ] if @xmath361 is a finitely generated infinite irreducible non - affine coxeter group then every commensurating endomorphism of @xmath361 is an inner automorphism .
this section is dedicated to proving theorem [ thm : norm_end ] .
our argument uses the powerful machinery of algebraic dehn fillings , developed for hyperbolically embedded subgroups by dahmani , guirardel and osin @xcite : [ thm : dehn_fil ] let @xmath1 be a group , @xmath91 a subset of @xmath1 , @xmath275 a collection of subgroups of @xmath1 .
suppose that @xmath301 .
then there exists a family of finite subsets @xmath1381 , @xmath1382 , such that for every collection of normal subgroups @xmath1383 , satisfying @xmath1384 for all @xmath297 , the following hold : * @xmath1385 for all @xmath297 , where @xmath1386 ; * every element of @xmath1387 is either conjugate to an element of @xmath1388 or is loxodromic with respect to the action of @xmath1 on @xmath610 , where @xmath1389 ; * @xmath1387 is isomorphic to the free product of copies of groups from @xmath1390 . combining the above result with corollary [ cor : wpd - he ] one obtains the following statement : [ lem : norm_sbgp_in_acyl_hyp ]
assume that @xmath1 is a group acting isometrically and coboundedly on a hyperbolic space @xmath79 .
for any element @xmath254 there exists @xmath1391 such that if @xmath1392 and @xmath1393 then the normal closure @xmath1394 is free and every non - trivial element in it is loxodromic ( with respect to the action of @xmath1 on @xmath79 ) .
let @xmath91 be a symmetric generating set of @xmath1 given by lemma [ lem : svarc - milnor ] .
by corollary [ cor : wpd - he ] , @xmath1395 , therefore we can apply theorem [ thm : dehn_fil ] , which claims that there exists a finite subset @xmath1396 such that for every normal subgroup @xmath1397 , with @xmath1398 , the normal closure @xmath1399 is isomorphic to the free product of some copies of @xmath1400 , and every element of @xmath1387 is either conjugate to an element of @xmath1400 in @xmath1 or is loxodromic with respect to the action of @xmath1 on the cayley graph @xmath1401 .
since the order of @xmath6 is infinite , there is @xmath1391 such that @xmath1402 whenever @xmath1392 .
so , if @xmath214 satisfies this inequality and @xmath1393 , by the previous paragraph we see that @xmath1394 is isomorphic to the free product of infinite cyclic groups ( hence , it is free ) and every element @xmath1403 is either conjugate to some non - zero power of @xmath6 in @xmath1 or is loxodromic with respect to the action of @xmath1 on @xmath1401 . therefore such @xmath46 is loxodromic with respect to the action of @xmath1 on @xmath100 : in the former case this is true because @xmath254 and in the latter case this is demonstrated in the first paragraph of the proof of lemma [ lem : tameins ] ( one can take @xmath1404 ) .
it follows that @xmath46 is loxodromic with respect to the action of @xmath1 on @xmath79 .
the argument will be split in two cases .
_ case 1 : _ @xmath64 .
we need to show that either @xmath1405 or @xmath3 is an inner automorphism of @xmath1 .
arguing by contradiction suppose that @xmath65 and @xmath1406 .
let us first prove the following claim : @xmath1407 since @xmath1 is acylindrically hyperbolic , it has a symmetric generating set @xmath91 such that @xmath1408 is hyperbolic , @xmath1409 , and @xmath1 acts on @xmath79 acylindrically .
then @xmath1 is non - elementary and @xmath340 ( as explained in theorem [ thm : acyl - equiv_def ] and in the paragraph after it ) , hence @xmath1 is generated by the @xmath1-special elements ( by proposition [ prop : special ] ) .
therefore there must exist @xmath1410 such that @xmath1411 .
choose @xmath1391 according to lemma [ lem : norm_sbgp_in_acyl_hyp ] .
then for any @xmath1412 , @xmath1413 and every non - trivial element of @xmath1414 is loxodromic .
it follows that @xmath1415 by remark [ rem : acyl->tame ] and the fact that the action of @xmath1 on @xmath79 is acylindrical . since @xmath1411
, there exists @xmath1412 such that @xmath1416 . on the other hand , @xmath58 as @xmath3 is a normal endomorphism , hence we can conclude that @xmath1417 .
consequently , @xmath1418 by remark [ rem : powers ] . if @xmath1419 , then claim is true for @xmath1420 .
so , suppose that @xmath1339 .
since @xmath64 and @xmath1406 , @xmath3 is not commensurating by theorem [ thm : comm_end ] . hence , according to corollary [ cor : comm_equiv_cond ] , there exists @xmath1421 such that @xmath1422 .
recall that @xmath266 is virtually cyclic , hence there is @xmath1423 such that @xmath1424 whenever @xmath651 is divisible by @xmath1425 .
therefore , we can apply lemma [ lem : norm_sbgp_in_acyl_hyp ] as before to find @xmath1426 such that @xmath1427 . again , since @xmath3 is normal , it must map this normal closure into itself .
so , if @xmath1428 then @xmath1429 , consequently @xmath1430 and @xmath1431 satisfies claim .
thus it remains to consider the case when @xmath1432 .
then @xmath1433 , and by lemma [ lem : newtamenc ] , there exists @xmath267 such that the element @xmath1434 belongs to @xmath246 and is not commensurable with @xmath6 in @xmath1 . but
@xmath1435 by the assumption above , therefore @xmath1436 ( by remarks [ rem : powers ] and [ rem : tame - conj ] ) and @xmath1437 ( as @xmath1438 ) .
thus we have shown the validity of claim .
so , let @xmath1439 be as in claim . then , according to corollary [ cor : wpd - he ] , the family @xmath1440 is hyperbolically embedded in @xmath1 .
now we can use the theory of algebraic dehn fillings : let @xmath1441 and @xmath1442 be the finite subsets given by theorem [ thm : dehn_fil ] . evidently , there is @xmath267 such that @xmath1443 and @xmath1444 .
then we can take @xmath1445 and @xmath1446 . since @xmath1447 , theorem [ thm : dehn_fil ] claims that for @xmath1448 one has @xmath1449 thus the image of @xmath951 in @xmath1450 has finite order @xmath261 and the image of @xmath1451 has infinite order in @xmath1450 .
on the other hand , since @xmath57 is a normal endomorphism , @xmath58 , hence it naturally induces an endomorphism @xmath1452 , defined by the formula @xmath1453 for all @xmath704 .
this yields a contradiction , as the order of @xmath1454 does not divide the order of @xmath1455 in @xmath1450 .
therefore , the proof under the assumption of case 1 is complete .
_ case 2 : _
@xmath1456 . in this case
@xmath1457 is also acylindrically hyperbolic and @xmath1458 ( see ( * ? ? ?
* lemma 5.10 ) ) .
it is also clear that @xmath1459 is a normal endomorphism of @xmath1460 .
therefore we are can apply case 1 to @xmath1460 and @xmath1459 , concluding that either @xmath1461 or there exists an element @xmath1462 such that @xmath1463 for all @xmath1464 . if @xmath1461 then @xmath63 , as required . in the remaining case ,
pick some preimage @xmath1347 of @xmath1465 .
then for every @xmath704 there exists @xmath1466 such that @xmath1467 .
clearly , since @xmath3 is an endomorphism , the restriction of @xmath70 to @xmath1343 is a homomorphism from @xmath1 to @xmath61 , hence , by corollary [ cor : comm_equiv_cond ] , @xmath3 is commensurating .
now that we have proved theorem [ thm : norm_end ] , one can show that if @xmath1 is acylindrically hyperbolic then @xmath53 has finite index in the group of all normal automorphisms @xmath1468 .
if @xmath1 is finitely generated , then this is a consequence of remark [ rem : f_many ] .
if @xmath1 is not finitely generated , then one can use a more involved argument similar to the one from ( * ? ? ?
6.4 and cor .
if the finite radical of an acylindrically hyperbolic group @xmath1 is non - trivial , then it may possess non - commensurating normal automorphisms with non - trivial finite images . indeed , let @xmath765 be the free group of rank @xmath26 and let @xmath8 be a non - abelian finite simple group .
let @xmath1469 be the direct product of @xmath765 and @xmath8 , so that @xmath1 is hyperbolic and @xmath1470 .
then @xmath1 has a natural endomorphism @xmath57 , which is the projection onto @xmath8 .
it is not difficult to check that every normal subgroup @xmath59 either contains @xmath8 or is contained in @xmath1471 .
it follows that @xmath3 is a normal endomorphism of @xmath1 with @xmath1472 .
the purpose of this section is to prove theorem [ thm : comm_end - raag ] from the introduction .
let @xmath1473 be a simplicial graph with the vertex set @xmath1474 and the edge set @xmath1475 .
the associated _ right angled artin group _
@xmath1476 is the group given by the presentation @xmath1477=1 , \
, \forall \{u , v\}\in e \right\rangle}}.\ ] ] the cardinality @xmath1478 is said to be the _ rank _ of @xmath71 .
algebraically , the rank of @xmath71 is exactly the smallest cardinality of a generating set of @xmath71 ( this can be justified by looking at the abelianization of @xmath71 , which is isomorphic to @xmath1479 ) .
right angled artin groups are special cases of _ graph products of groups _
, when all the vertex groups are infinite cyclic ( see ( * ? ? ?
* subsection 2.2 ) for some background on graph products ) .
[ lem : norm - centr ] suppose that @xmath71 is a right angled artin group and @xmath1480 is any subgroup .
* if @xmath1481 is a normal subgroup which does not contain non - abelian free subgroups , then @xmath1387 is central in @xmath24 . *
the quotient of @xmath24 by its center @xmath1482 is centerless . to prove ( i ) , suppose that @xmath1387 is not central in @xmath24 . then there exist @xmath1483 and @xmath1484 such that @xmath1485 . by a theorem of baudisch @xcite
( see also ( * ? ? ?
* cor . 1.6 ) ) , the latter implies that @xmath46 and @xmath6 generate a free subgroup @xmath765 , of rank @xmath26 , in @xmath71 .
since @xmath1486 , this intersection is a non - trivial normal subgroup of @xmath765 , hence it is a non - abelian free group .
this contradicts the assumption that @xmath1387 has no non - abelian free subgroups
. therefore @xmath1387 must be central in @xmath24 .
to verify ( ii ) , let @xmath1487 be the full preimage of the center of @xmath1488 under the homomorphism @xmath1489 .
then @xmath1387 is nilpotent of class at most @xmath26 , hence it satisfies the assumptions of ( i ) , and therefore it must be central in @xmath24 .
thus @xmath1490 ; on the other hand @xmath1491 by the definition of @xmath1387 .
it follows that @xmath1492 , and so the image of @xmath1387 in @xmath1488 ( i.e. , the center of @xmath1488 ) is trivial . for any subset @xmath101 of @xmath103 the subgroup @xmath1493 is said to be a _ full subgroup _ of @xmath71 .
it is not difficult to show that @xmath1494 is naturally isomorphic to the right angled artin group @xmath1495 , where @xmath1496 the full subgraph of @xmath407 spanned on the vertices from @xmath101 ( see , for example , ( * ? ? ?
* section 6 ) ) . for every @xmath1497
there is a _
canonical retraction _ @xmath1498 defined on the generators of @xmath71 by @xmath1499 , if @xmath1500 and @xmath1501 if @xmath1502 .
a subgroup @xmath1503 is called _ parabolic _ if it is conjugate to a full subgroup , i.e. , there exist @xmath1504 and @xmath1505 such that @xmath1506 ; we will say that @xmath24 is a _ proper parabolic subgroup _ of @xmath1507 if @xmath1508 .
if the graph @xmath407 is finite then any subgroup @xmath1509 is contained in a unique minimal parabolic subgroup @xmath1510 , called the _
parabolic closure _ of @xmath24 in @xmath1507 ( see ( * ? ? ?
* prop . 3.10 ) ) . using the terminology from @xcite
, we will say that a graph @xmath407 is _ reducible _ if there exists a partition @xmath1511 into non - empty disjoint subsets @xmath71 and @xmath979 such that every vertex from @xmath71 is adjacent to every vertex from @xmath979 in @xmath407 . otherwise , @xmath407 is said to be _
irreducible_. alternatively , @xmath407 is irreducible if and only if the _
complement graph _
@xmath1512 is connected ( recall that @xmath1512 is defined by @xmath1513 and @xmath1514 ) .
every finite graph @xmath407 can be decomposed into irreducible subgraphs ; this means that there is a partition @xmath1515 , where @xmath1516 , @xmath1517 is irreducible for @xmath1518 , and for any pair of indices @xmath396 , every vertex of @xmath473 is adjacent with every vertex of @xmath1519 in @xmath407 ( this corresponds to the decomposition of @xmath1512 into the union of its connected components ) . using this
we obtain the _ standard factorization _ of the right angled artin group @xmath1476 : @xmath1520 where @xmath1521 is a free abelian group ( i.e. , the right angled artin group corresponding to a complete subgraph of @xmath407 ) and each @xmath1522 , @xmath632 , is a right angled artin group corresponding to a full irreducible subgraph @xmath537 , of @xmath407 , with @xmath1523 .
we will say that @xmath1521 is the _ abelian factor _ of @xmath71 and @xmath1524 are the _
irreducible factors _ of @xmath71 .
note that @xmath1521 is central in @xmath71 by definition ( in fact @xmath1521 coincides with the center of @xmath71 , which , for example , follows from lemma [ lem : no_center ] below ) .
the following fact was proved in ( * ? ? ?
3.15 ) : [ prop_parab ] let @xmath407 be a finite irreducible graph and let @xmath1476 be the associated right angled artin group .
suppose that @xmath1480 and @xmath1481 is a non - trivial normal subgroup of @xmath24 .
if @xmath1525 then @xmath1526 . we will also need the following statement , which is a special case of ( * ? ? ?
* cor . 6.20 ) .
[ lem : heraags ] let @xmath1476 be a right angled artin group corresponding to some finite irreducible graph @xmath407 with @xmath1527
. then @xmath71 acts simplicially and coboundedly by isometries on a simplicial tree @xmath1528 so that the following holds . for any subgroup @xmath1529 with @xmath1530 one has @xmath1531 .
note that the geometric realization of a simplicial tree is @xmath39-hyperbolic .
therefore , lemma [ lem : heraags ] shows that the theory which we developed in section [ sec : comm_hom ] can be applied to any such @xmath24 . given a group @xmath1 and an element
@xmath241 we can define @xmath1532 one easily checks that @xmath266 is a subgroup of @xmath1 , containing the centralizer @xmath1533 .
[ lem : no_center ] let @xmath407 be a finite irreducible graph and let @xmath1476 be the corresponding right angled artin group .
suppose that @xmath1480 is a non - cyclic subgroup such that @xmath1525 .
then @xmath24 has trivial center and there is @xmath1534 such that @xmath1535 .
since @xmath24 is not cyclic , @xmath1527 , and so we can apply lemma [ lem : heraags ] to find a simplicial tree @xmath1528 such that @xmath71 acts on @xmath1528 isometrically and coboundedly , and @xmath1531 .
recall that right angled artin groups are torsion - free , hence @xmath1536 ( see lemma [ lem : e_g(h ) ] ) and @xmath24 is non - elementary ( because it is not cyclic , and a torsion - free elementary group is cyclic ) .
therefore we can apply lemma [ lem : special ] to find an infinite order element @xmath971 such that @xmath1537 .
moreover , by lemma [ lem : lwpd+ ] , there is an element @xmath1538 such that @xmath6 is not commensurable with @xmath46 in @xmath71 . in view of remark [ rem : elem_inter ]
, the latter implies that @xmath1539 . since this intersection contains the center of @xmath24 , @xmath24 must be centerless
. the following simple observation will be useful : [ rem : comm_end_of_ab ] if @xmath24 is a free abelian group then the only commensurating endomorphisms of @xmath24 are endomorphisms of the form @xmath1540 for some @xmath1541 and for all @xmath971 .
we can now prove the main result of this section .
choose a finite graph @xmath407 , with the smallest possible @xmath1542 , so that the corresponding right angled artin group @xmath1476 contains ( an isomorphic copy of ) @xmath24 .
let @xmath1543 be the standard factorization of @xmath71 , where @xmath1521 is the abelian factor of @xmath71 and @xmath1544 are the irreducible factors of @xmath71 .
observe that @xmath1521 is a finitely generated free abelian group and @xmath1545 as @xmath24 is non - abelian .
let @xmath1546 denote the canonical retraction ( in other words , @xmath1547 is the @xmath412-th coordinate projection ) , @xmath623 .
note that for every @xmath1548 , the image @xmath1549 can not be isomorphic to a subgroup of a right angled artin group @xmath1 whose rank is strictly smaller than the rank of @xmath1522 . indeed ,
otherwise @xmath24 would embed into the direct product @xmath1550 which would be a right angled artin group of smaller rank than @xmath71 , contradicting the choice of @xmath407 .
it follows that for each @xmath1548 , @xmath1549 can not be cyclic ( as the rank of @xmath1522 is at least @xmath26 by the definition of irreducible factors ) and the parabolic closure of @xmath1549 in @xmath1522 is @xmath1522 .
one can also deduce that @xmath1551 is non - trivial whenever @xmath813 , because otherwise @xmath24 would embed into the direct product of @xmath1552 , which is a right angled artin group of smaller rank than @xmath71 .
observe that @xmath1553 , hence @xmath1554 by lemma [ prop_parab ] , where @xmath537 is the full irreducible subgraph of @xmath407 corresponding to @xmath1522 , @xmath632 .
moreover , @xmath1555 can not be cyclic in view of lemma [ lem : norm - centr].(i ) as the center of @xmath1549 is trivial by lemma [ lem : no_center ] .
hence we can apply lemma [ lem : no_center ] to @xmath1522 and @xmath1555 in to find an element @xmath1556 such that @xmath1557 , @xmath632 .
now consider any commensurating endomorphism @xmath54 .
for each @xmath1558 let @xmath1559 denote the product of all @xmath1560 , @xmath1561 ; thus @xmath1562 and @xmath1563 . by the hypothesis , for any @xmath1564 , @xmath1565 and @xmath1566 for some @xmath1567 and @xmath1568 . and
since @xmath1569 is torsion - free , we can conclude that @xmath1570 .
the latter shows that @xmath3 preserves the kernel of the restriction of @xmath1547 to @xmath24 , @xmath623 .
therefore @xmath3 naturally induces an endomorphism @xmath1571 for @xmath623 , defined by the formula @xmath1572 for all @xmath1340 .
evidently , @xmath1573 will be a commensurating endomorphism of @xmath1549 for each @xmath623 .
therefore , according to remark [ rem : comm_end_of_ab ] , there must exist @xmath1574 such that @xmath1575 for all @xmath1576 .
on the other hand , if @xmath1548 , we can recall that @xmath1522 is an irreducible factor of @xmath71 and @xmath1549 is a non - elementary subgroup of @xmath1522 such that the parabolic closure of @xmath1549 in @xmath1522 is @xmath1522 .
therefore , in view of lemma [ lem : heraags ] , all the assumptions of theorem [ thm : comm ] are satisfied , hence there exists @xmath1577 such that @xmath1578 for all @xmath1579 ( here we used the fact that @xmath1580 as @xmath1522 is torsion - free ) , @xmath632 .
let @xmath1581 be the inner automorphism defined by @xmath1582 for all @xmath1583 , where @xmath1584 .
let us show that the endomorphism @xmath3 is actually the restriction of @xmath1261 to @xmath24 .
the preceding paragraph implies that this is true if the abelian factor @xmath1521 is trivial , because in this case for every @xmath1340 one would have @xmath1585 , and so @xmath1586 on the other hand , if @xmath1521 is non - trivial , then @xmath1587 is also non - trivial ( by the minimality of the rank of @xmath71 ) .
so , pick any @xmath1588 .
let @xmath1589 be the element constructed above
. since @xmath3 is commensurating and @xmath1590 , there must exist @xmath271 and @xmath1591 such that @xmath1592 but we also have @xmath1593 . therefore @xmath1594 applying @xmath1595 and @xmath1596 to the above equation we obtain @xmath1597 and @xmath1598 . the former yields that @xmath1599 ; and the latter shows that @xmath1600 , in particular this element commutes with @xmath1601 .
thus @xmath1602 , and so @xmath1603 .
consequently , @xmath1604 , which implies that @xmath1605 for all @xmath1340 .
if @xmath1606 then the proof would have been finished . however , this may not be the case , so one more step is needed .
let @xmath1607 , @xmath632 , be the elements constructed above so that @xmath1608 , and set @xmath1609 . by the assumption , there exist @xmath271 and @xmath1591 such that @xmath1610 .
on the other hand , we know that @xmath1611 .
combining these two equalities one gets @xmath1612 in @xmath71 . applying @xmath1547 yields that @xmath1613 , where @xmath1614 , for @xmath632 .
it follows that for every @xmath632 , there exists @xmath1615 such that @xmath1616 in @xmath1522 .
thus , denoting @xmath1617 , we achieve @xmath1618 where the element @xmath1619 belongs to @xmath24 by construction .
since @xmath1620 is central in @xmath71 , we see that @xmath1621 for all @xmath1340 , thus @xmath3 is indeed an inner automorphism of @xmath24
. the claim of theorem [ thm : comm_end - raag ] would be no longer true if one dropped the assumption that the ambient right angled artin group is finitely generated .
indeed , let @xmath1 be the direct product of infinitely ( countably ) many copies of the free group of rank @xmath26 .
then @xmath1 is a normal subgroup in the cartesian ( i.e. , unrestricted ) product @xmath168 of these free groups and any inner automorphism of @xmath168 induces a pointwise inner automorphism of @xmath1 .
it follows that @xmath1 has uncountably many pointwise inner ( hence , commensurating ) but non - inner automorphisms .
recall that , given a group @xmath1 , the _ profinite topology _ on @xmath1 is the topology whose basic open sets are cosets to normal subgroups of finite index in @xmath1 .
it is easy to see that group operations and group homomorphisms are continuous with respect to this topology .
in particular , @xmath1 , equipped with this topology , is a topological group .
one can also observe that the profinite topology is hausdorff if and only if @xmath1622 is a closed subset of @xmath1 if and only if @xmath1 is residually finite .
it follows that any finite subset of a residually finite group is closed ( in the profinite topology ) . if @xmath59 , then @xmath1450 is residually finite if and only if @xmath1387 is closed in @xmath1 .
thus if @xmath1 is residually finite and @xmath1623 then @xmath1450 is also residually finite .
finally , residual finiteness is preserved under taking subgroups or overgroups of finite index .
[ rem : rf ] suppose that @xmath1 is a group and for every @xmath1624 there is a homomorphism @xmath1261 from @xmath1 to a residually finite group @xmath1625 such that @xmath1626 .
then @xmath1 is residually finite . in this section
we discuss various conditions one can impose on @xmath1 to ensure residual finiteness of @xmath0 .
one set of conditions is given by grossman s criterion @xcite , mentioned in the introduction .
in particular , since any pointwise inner automorphism is commensurating , we can combine this criterion with theorem [ thm : comm_end - raag ] to obtain [ cor : cs+raag->out - rf ] let @xmath1 be a finitely generated conjugacy separable subgroup of a right angled artin group .
then @xmath0 is residually finite . in ( * ? ?
2.1 ) the second author proved that groups from the class @xmath1627 ( i.e. , virtual retracts of finitely generated right angled artin groups ) are conjugacy separable .
since these groups are finitely generated ( and even finitely presented ) , as virtual retracts of finitely presented groups , we can apply corollary [ cor : cs+raag->out - rf ] to achieve [ cor : vr ] if @xmath1628 then @xmath0 is residually finite .
another useful tool for establishing residual finiteness of @xmath0 is given by the following observation : [ lem : outnormsub ] suppose that @xmath1 is a finitely generated group , and @xmath1387 is a centerless normal subgroup of finite index in @xmath1
. then some finite index subgroup @xmath1629 is isomorphic to a quotient of a subgroup of @xmath1630 by a finite normal subgroup .
in particular , if @xmath1631 is residually finite then so is @xmath0 . for our purposes
we will also need a criterion ( see proposition [ prop : new_crit ] below ) which applies when the center of @xmath1387 is non - trivial .
given a subgroup @xmath1007 , define @xmath1632 , and let @xmath1633 be its image in @xmath0 .
since a finitely generated group contains only finitely many subgroups of any given finite index , the following observation can be made : [ rem : out(g;h ) ] if @xmath1 is a finitely generated group and @xmath1007 has finite index then @xmath1634 and @xmath1635 . if @xmath8 is an _ abelian _
group and @xmath267 then @xmath1636 is called a _ congruence subgroup _ of @xmath8 .
clearly , every finite index subgroup of @xmath8 contains @xmath1637 for some @xmath267 .
if , additionally , @xmath8 is finitely generated , then @xmath1638 for all @xmath267 , hence the profinite topology of @xmath8 is generated by the congruence subgroups .
it follows that for any fixed @xmath954 , the profinite topology on @xmath8 is also generated by the collection @xmath1639 .
let us also specify some notation . if @xmath16 are elements of a group @xmath1 , we will write @xmath1640 for the conjugate @xmath1641 and @xmath1642 $ ] for the commutator @xmath1643 . if @xmath122 then @xmath1644 and @xmath1645 $ ] will denote the subsets @xmath1646 and @xmath1647 \mid e \in e\ } \subseteq g$ ] respectively .
[ prop : new_crit ] let @xmath1 be a finitely generated group , let @xmath1648 be a normal subgroup of finite index such that the center @xmath1649 , of @xmath1387 , is finitely generated .
suppose that @xmath1650 is residually finite and there is @xmath954 such that @xmath1651 is residually finite for all @xmath267 .
then @xmath0 is also residually finite . in view of remark [ rem : out(g;h ) ] and since @xmath1652 ( because @xmath1653 is a characteristic subgroup of @xmath1387 ) , it is enough to prove that @xmath1654 is residually finite .
so , consider any @xmath1655 ( note that @xmath1656 as @xmath1657 ) and let @xmath1658 denote its image in @xmath0 .
note that @xmath1659 for every @xmath267 , hence @xmath1660 naturally induces an automorphism of @xmath1661 ( as it permutes the cosets of @xmath1662 in @xmath1 ) .
this gives rise to the following commutative diagram between automorphism groups : @xmath1663\ar[d ] & aut(g / z^n ; z / z^n)\ar[r]\ar[d ] & aut(g / z ) \ar[d ] \\
out(g;z ) \ar[r ] & out(g / z^n ; z / z^n)\ar[r ] & out(g / z ) } \ ] ] in view of the assumptions and remark [ rem : rf ] , to prove the proposition it is enough to show that there exists @xmath1664 such that the image of @xmath1665 in @xmath1666 ( coming from the commutative diagram above ) is non - trivial .
if @xmath1660 induces a non - inner automorphism of @xmath1667 , then the image of @xmath1665 will be non - trivial in @xmath1650 .
thus , we can now suppose that @xmath1660 induces an inner automorphism of @xmath1667 .
this means that we can replace @xmath1660 by its composition with an inner automorphism of @xmath1 ( this does not affect @xmath1665 ) to further assume that @xmath1660 induces the identity on @xmath1667 . in other words , @xmath1668 for all @xmath4 .
choose a finite generating set @xmath1669 of @xmath1 .
then for every @xmath1518 , there is @xmath1670 such that @xmath1671 .
let @xmath1672 be the full preimage in @xmath1 of the center @xmath1673 , and set @xmath1674 .
let @xmath1675 be the @xmath164-th direct power of @xmath1 , let @xmath1676 be the @xmath164-th direct power of @xmath1653 , and let @xmath1677 be the corresponding diagonal subgroup of @xmath168 .
observe that for any given @xmath267 , @xmath1660 induces an inner automorphism of @xmath1661 if and only if there exists @xmath1678 such that @xmath1679 for every @xmath1518 . the latter equality can be re - written as @xmath1680 \text { ( mod $ z^n$)}$ ] in @xmath1 .
thus @xmath1660 induces an inner automorphism of @xmath1661 if and only if @xmath1681 $ ] @xmath1682 , where @xmath1683 .
note that @xmath1684\subseteq q$ ] as @xmath1685 \in z$ ] for all @xmath1678 , @xmath4 , by the definition of @xmath1672 .
observe that the subgroup @xmath1686 has finite index in @xmath124 ( because @xmath1687 ) .
moreover , if @xmath1688 then @xmath1689=[c , g][c',g]$ ] for any @xmath4 .
this can be derived from the commutator identities , because @xmath1690 , [ c',g ] \in z$ ] , and @xmath1653 is an abelian subgroup centralized by @xmath1691 .
it follows that @xmath1692 $ ] is actually a subgroup of the finitely generated abelian group @xmath8 .
therefore , @xmath1692 $ ] is closed in the profinite topology of @xmath8 ( in fact any subgroup @xmath1693 is closed because the quotient @xmath1694 is again a finitely generated abelian group , and so it is residually finite as a direct sum of cyclic groups ) . by construction
, there exist @xmath1695 such that @xmath1696 . utilizing commutator identities once again
, we get @xmath1697=\bigcup_{j=1}^l [ e_1,(x_1,\dots , x_k)]^{e_j } [ e_j,(x_1,\dots , x_k)].\ ] ] this shows that @xmath1684 $ ] is also a closed subset of @xmath8 , as finite union of closed subsets .
recall , that @xmath1698 , therefore @xmath1699 $ ] in @xmath8 .
it follows that we can find @xmath267 such that @xmath1699 $ ] @xmath1700 .
the latter demonstrates that @xmath1660 induces a non - inner automorphism of @xmath1701 , which finishes the proof of the proposition .
the proof of proposition [ prop : new_crit ] actually shows that if @xmath1 is a finitely generated group and @xmath1648 is a finite index normal subgroup such that the center @xmath1653 , of @xmath1387 , is finitely generated then for any @xmath954 , @xmath1654 embeds into the cartesian product @xmath1702 .
in this section we will prove theorem [ thm : avr ] . in view of corollary
[ cor : vr ] and lemma [ lem : outnormsub ] , essentially it remains to deal with the case when a finite index normal subgroup @xmath1703 of a group @xmath1704 has non - trivial center .
[ lem : descr_of_center ] let @xmath71 be the right angled artin group corresponding to a finite graph @xmath407 and let @xmath1705 be its standard factorization , where @xmath1521 is the abelian factor and @xmath1544 are the irreducible factors of @xmath71 .
suppose that @xmath1480 is a subgroup such that @xmath1525 and @xmath1549 is not cyclic , for each @xmath632 , where @xmath1706 denotes the canonical retraction .
then the center of @xmath24 is equal to the intersection of @xmath24 with @xmath1521 .
consider any @xmath1548 and observe that if @xmath1549 is contained in a proper parabolic subgroup @xmath1707 of @xmath1522 ( where @xmath1708 and @xmath1709 is a full subgroup of @xmath1522 , and , hence , of @xmath71 ) , then @xmath24 is contained in the subgroup @xmath1710 , which is a proper parabolic subgroup of @xmath71 , contradicting the assumption that @xmath1525 .
therefore the parabolic closure of @xmath1549 in @xmath1522 is the whole of @xmath1522 , @xmath632 .
let @xmath1653 denote the center of @xmath24 .
then @xmath1711 is contained in the center of @xmath1549 , which is trivial for @xmath632 , by lemma [ lem : no_center ] .
thus @xmath1712 for each @xmath1548 , which implies that @xmath1713 .
evidently , @xmath1714 because @xmath1521 is central in @xmath71 , hence @xmath1715 , as claimed .
it is not difficult to see that the class @xmath1627 is closed under taking subgroups of finite index ( see ( * ? ? ?
* remark 9.4 ) ) . to prove the main result of this section we will also need the fact that this class is closed under taking quotients by the center : [ prop : quot_by_center ] let @xmath1672 be a finitely generated right angled artin group , let @xmath1716 be an arbitrary subgroup and let @xmath1653 be the center of @xmath24 .
* for any subgroup @xmath1717 , @xmath1718 is finitely generated and the quotient @xmath1719 is residually finite . * if @xmath24 is a virtual retract of @xmath1672 then @xmath1720 .
since @xmath1672 has finite rank , there exists a right angled artin subgroup @xmath1721 which contains @xmath24 and has minimal rank ( among all such subgroups of @xmath1672 ) .
let @xmath407 be the finite simplicial graph corresponding to @xmath71 and let @xmath1705 be the standard factorization of @xmath71 , where @xmath1521 is the abelian factor and @xmath1544 are the irreducible factors of @xmath71 . if @xmath620 then the groups @xmath24 and @xmath1722 are free abelian of finite rank , hence both statements are evidently true .
therefore we can assume that @xmath635 .
let @xmath1546 denote the canonical projection of @xmath71 onto @xmath1522 , @xmath623 .
we remark that @xmath1525 , by the choice of @xmath71 .
if @xmath1549 is a cyclic subgroup @xmath979 of @xmath1522 for some @xmath1548 , then @xmath24 embeds into the subgroup @xmath1723 where @xmath1724 which is a right angled artin group of strictly smaller rank than @xmath71 , contradicting the choice of @xmath71 .
therefore we can conclude that @xmath1549 is non - cyclic for every @xmath1725 .
thus we are able to apply lemma [ lem : descr_of_center ] , claiming that @xmath1726 .
consider any subgroup @xmath1727 . since @xmath1521 is a finitely generated abelian group
we see that @xmath1718 is also finitely generated .
moreover , the quotient @xmath1719 naturally embeds into the quotient @xmath1728 .
therefore @xmath1729 ( and hence @xmath1719 ) is residually finite , as a direct product of residually finite groups : @xmath1730 is a finitely generated abelian group and @xmath1522 , @xmath632 , are right angled artin groups , whose residual finiteness is well - known ( see ( * ? ? ?
3 , thm 1.1 ) or ( * ? ? ?
thus ( a ) is proved . to prove ( b ) assume that @xmath24 is a virtual retract of @xmath1672 .
this implies that for any subgroup @xmath1731 such that @xmath1732 , @xmath24 is a virtual retract of @xmath1733 . in particular
, @xmath24 will also be a virtual retract of @xmath71 .
thus @xmath71 contains a finite index subgroup @xmath1625 such that @xmath1734 and there is a retraction @xmath1735 . since @xmath1521 is central in @xmath71 , @xmath1736 is central in @xmath1625 , and so @xmath1737 .
consider the canonical projection @xmath1738 , and observe that @xmath1739 it follows ( see ( * ? ? ?
* lemma 4.1 ) ) that @xmath1740 naturally induces a retraction @xmath1741 of @xmath1742 onto its subgroup @xmath1743 .
thus @xmath1743 is a retract of @xmath1742 , and the latter has finite index in the finitely generated right angled artin group @xmath1744 , because @xmath1745 .
it remains to recall that @xmath1746 , hence @xmath1747 .
thus @xmath1720 , and the proposition is proved . 1 . 2 . part ( a ) of proposition [ prop : quot_by_center ] can actually be derived from more general results .
indeed , it is known that the finitely generated right angled artin group @xmath1672 can be embedded into @xmath1748 for some @xmath1749 .
therefore every solvable subgroup @xmath1750 is polycyclic ( hence , finitely generated ) and is closed in the profinite topology of @xmath1748 by a result of segal ( * ? ? ?
* , thm . 5 ) .
since the profinite topology of @xmath24 is finer than the topology induced by the profinite topology of @xmath1748 , we can conclude that @xmath979 is closed in @xmath24 .
3 . since right angled artin groups are cat(@xmath39 )
, it is easy to prove a weaker version of proposition [ prop : quot_by_center].(b ) , that @xmath1751 , using the flat torus theorem ( * ? ? ? *
ii.7.1.(5 ) ) . combining proposition [ prop : quot_by_center ] with lemma [ lem : norm - centr].(ii )
one immediately obtains suppose that @xmath1752 and @xmath1653 is the center of @xmath24
. then the group @xmath1488 is centerless and belongs to @xmath1627 .
let @xmath1 be any group from the class @xmath29 .
this means that @xmath1 contains a finite index subgroup @xmath1752 ; in particular , @xmath1 is finitely generated .
note that @xmath1753 is a finite index normal subgroup of @xmath1 , and @xmath1703 because the class @xmath1627 is closed under taking finite index subgroups .
let @xmath1653 denote the center of @xmath1387 .
we are going to check that all the assumptions of the criterion from proposition [ prop : new_crit ] are satisfied .
first , the fact that @xmath1653 is finitely generated follows from proposition [ prop : quot_by_center].(a ) .
second , take any @xmath267 and note that @xmath1754 is residually finite , also by proposition [ prop : quot_by_center].(a ) .
hence there is a finite index normal subgroup @xmath1755 such that @xmath27 has trivial intersection with the finite subgroup @xmath1756 in @xmath1754 .
again , we can replace @xmath27 with the intersection of all its conjugates in @xmath1661 to further assume that @xmath1757 . by construction
, @xmath27 injects into the quotient @xmath1758 under the natural epimorphism @xmath1759 .
let @xmath1760 denote the image of @xmath27 in @xmath1761 .
since @xmath1762 by proposition [ prop : quot_by_center].(b ) and @xmath1763 has finite index in @xmath1758 , we see that @xmath1764 .
hence , according to corollary [ cor : vr ] , @xmath1765 is residually finite .
moreover , since the center of @xmath1762 is trivial ( lemma [ lem : norm - centr].(ii ) ) , the center of @xmath1766 must be trivial as well ( because @xmath1766 is an abelian normal subgroup of @xmath1758 and so it is central in @xmath1758 by lemma [ lem : norm - centr].(i ) ) .
therefore @xmath1767 is a centerless finite index normal subgroup in @xmath1661 with a residually finite outer automorphism group .
consequently , lemma [ lem : outnormsub ] yields that @xmath1768 is residually finite .
since this works for arbitrary @xmath267 , we see that all the assumptions of proposition [ prop : new_crit ] are satisfied .
it remains to apply this proposition to conclude that @xmath0 is residually finite , which finishes the proof of the theorem .
this last section of the paper is dedicated to proving theorem [ thm:3-man ] .
we start with the following lemma , which allows to deal with the seifert fibered case .
[ lem : seifert ] suppose that @xmath1 is a finitely generated group containing a finite index subgroup @xmath24 that fits into the short exact sequence @xmath1769 where @xmath1625 is a cyclic group and @xmath1425 has a finite index subgroup which embeds into the fundamental group of a compact surface .
then @xmath0 is residually finite .
since @xmath1625 is cyclic , its automorphism group is finite . moreover ,
@xmath24 acts on @xmath1625 by conjugation because @xmath1770 , and the kernel of this action is the centralizer of @xmath1625 in @xmath24 .
it follows that @xmath1771 . combining this with the other assumptions on @xmath1 and @xmath24 , we can find a finite index normal subgroup @xmath1648 such that @xmath1772
, @xmath1773 is central in @xmath1387 and @xmath1758 is a subgroup of some compact surface group .
note that @xmath1387 is finitely generated , as this is true for @xmath1 by the hypothesis , hence the quotient @xmath1758 is itself isomorphic to the fundamental group of some compact surface @xmath31 ( because any finitely generated subgroup of a surface group is itself a surface group ) .
evidently we can also assume that @xmath31 is orientable .
it follows that @xmath1774 is either abelian ( isomorphic to @xmath1622 , @xmath700 or @xmath1775 ) or is non - elementary torsion - free hyperbolic . in the former case
, @xmath1387 is polycyclic , hence @xmath1 is virtually polycyclic and so @xmath0 is residually finite ( according to a theorem of wehrfritz @xcite , @xmath0 is linear over @xmath700 ) .
thus we can assume that @xmath1758 is a torsion - free hyperbolic group , which , in particular , implies that it is centerless and so the cyclic subgroup @xmath1653 is equal to the center of @xmath1387 .
now , in order to apply proposition [ prop : new_crit ] , we check that @xmath1768 is residually finite for any @xmath267 .
indeed , observe that @xmath1756 is a finite central subgroup of @xmath1754 such that the quotient @xmath1776 is isomorphic to the surface group @xmath1777 .
it follows that @xmath1754 possesses a finite index normal subgroup @xmath1755 which intersects @xmath1756 trivially ( see ( * ? ? ?
* lemma 4.2 ) ) .
thus the image @xmath1763 , of @xmath27 in @xmath1758 , is naturally isomorphic to @xmath27 and has finite index in @xmath1777 .
consequently , @xmath1763 is itself isomorphic to the fundamental group of a compact orientable surface , which finitely covers @xmath31 . by grossman
s theorem @xcite , @xmath1778 is residually finite ; moreover , @xmath1779 is centerless because it is a non - elementary torsion - free hyperbolic group ( as it has finite index in @xmath1758 ) .
since @xmath27 has finite index in @xmath1661 , lemma [ lem : outnormsub ] implies that @xmath1768 is residually finite for any @xmath267 .
therefore we can use proposition [ prop : new_crit ] to conclude that @xmath0 is residually finite .
one of the main ingredients of the proof of theorem [ thm:3-man ] is the following beautiful result of hamilton , wilton and zalesskii , which is based on the deep work of wise @xcite and agol @xcite mentioned in the introduction . 1 .
@xmath24 is acylindrically hyperbolic with @xmath1780 ; 2 .
@xmath24 contains an infinite cyclic normal subgroup @xmath1625 such that @xmath1781 is virtually a subgroup of the fundamental group of a compact surface ; 3 .
@xmath24 is virtually polycyclic . by the assumptions @xmath1782 has finite index in @xmath1 ,
thus both @xmath24 and @xmath1 are finitely generated because @xmath33 is compact .
moreover , after replacing @xmath24 with some finite index subgroup , we can assume that @xmath1783 . if @xmath24 is virtually polycyclic then so is @xmath1 , hence @xmath0 is residually finite by wehrfritz s theorem @xcite .
if @xmath24 satisfies condition ( ii ) of theorem [ thm : trichotomy ] then @xmath0 is residually finite by lemma [ lem : seifert ] .
thus , in view of theorem [ thm : trichotomy ] , we can assume that @xmath24 is acylindrically hyperbolic and @xmath1780
. therefore we can apply corollary [ cor : conj_aut ] , stating that every pointwise inner automorphism of @xmath24 is inner . recall that @xmath24 is finitely generated and conjugacy separable by theorem [ thm : hwz ] ,
hence @xmath25 is residually finite by grossman s criterion ( * ? ? ?
1 ) . it remains to observe that the center @xmath1784 , of @xmath24 , is finite because @xmath24 is acylindrically hyperbolic ( see ( * ? ? ?
4.34 ) ) , hence @xmath1785 , i.e. , @xmath24 is centerless .
consequently , lemma [ lem : outnormsub ] implies that @xmath0 is residually finite .
allenby , g. kim and c.y .
tang , _ outer automorphism groups of certain orientable seifert
3-manifold groups_. combinatorial group theory , discrete groups , and number theory , 1522 , contemp .
, * 421 * , amer .
soc . , providence , ri , 2006 .
r. charney , k. vogtmann , _ subgroups and quotients of automorphism groups of raags .
_ low - dimensional and symplectic topology , 927 , proc .
pure math .
, * 82 * , amer . math .
soc . , providence , ri , 2011 .
malcev , _ on the faithful representation of infinite groups by matrices . _
( english .
russian original ) am .
, ii . ser .
* 45 * ( 1965 ) , 118 ; translation from mat .
sb . , n. ser .
* 8 * ( 50)(1940 ) , no .
3 , 405422 .
d. mccullough , _ topological and algebraic automorphisms of 3-manifolds .
_ groups of self - equivalences and related topics ( montreal , pq , 1988 ) , 102113 , lecture notes in math . , 1425 , springer , berlin , 1990 . | we prove that the outer automorphism group @xmath0 is residually finite when the group @xmath1 is virtually compact special ( in the sense of haglund and wise ) or when @xmath1 is isomorphic to the fundamental group of some compact @xmath2-manifold . to prove these results we characterize commensurating endomorphisms of acylindrically hyperbolic groups .
an endomorphism @xmath3 of a group @xmath1 is said to be commensurating , if for every @xmath4 some non - zero power of @xmath5 is conjugate to a non - zero power of @xmath6 .
given an acylindrically hyperbolic group @xmath1 , we show that any commensurating endomorphism of @xmath1 is inner modulo a small perturbation .
this generalizes a theorem of minasyan and osin , which provided a similar statement in the case when @xmath1 is relatively hyperbolic .
we then use this result to study pointwise inner and normal endomorphisms of acylindrically hyperbolic groups . |
the problem of motion , one of the cardinal problems of einstein s gravitation theory , has received continuous attention over the years .
the early , classic works of lorentz - droste , eddington - clark , einstein - infeld - hoffmann , fock , papapetrou and others led to a good understanding of the equations of motion of @xmath7 bodies at the first post - newtonian ( 1pn ) approximationpn order refers to the terms of order @xmath8 in the equations of motion . ]
e.g. _ , @xcite for a general review of the problem of motion ) . in the 1970 s , an important series of works by a japanese group @xcite led to a nearly complete control of the problem of motion at the second post - newtonian ( 2pn ) approximation .
then , in the early 80 s , motivated by the observation of secular orbital effects in the hulse - taylor binary pulsar psr1913 + 16 , several groups solved the two - body problem at the 2.5pn level ( while completing on the way the derivation of the 2pn dynamics ) ( for more recent work on the 2.5pn dynamics see @xcite ) . in the late 90 s , motivated by the aim of deriving high - accuracy templates for the data analysis of the upcoming international network of interferometric gravitational - wave detectors
, two groups embarked on the derivation of the equations of motion at the third post - newtonian ( 3pn ) level .
one group used the arnowitt - deser - misner ( adm ) hamiltonian approach and worked in a corresponding adm - type coordinate system .
another group used a direct post - newtonian iteration of the equations of motion in harmonic coordinates .
the end results of these two approaches have been proved to be physically equivalent @xcite .
however , both approaches , even after exploiting all symmetries and pushing their hadamard - regularization - based methods to the maximum of their possibilities , left undetermined _ one and only one _ dimensionless parameter : @xmath9 in the adm approach and @xmath4 in the harmonic - coordinates one .
the unknown parameters in both approaches are related by @xmath10 as was deduced from the comparison between the invariant energy functions for circular orbits in the two approaches @xcite , and from two independent proofs of the equivalence between the two formalisms for general orbits @xcite .
the appearance of one ( and only one ) unknown parameter in the equations of motion is quite striking ; it is related with the choice of the regularization method used to cure the self - field divergencies of point particles
. both lines of works and @xcite regularized the self - field divergencies by some version of the hadamard regularization method . the second line of work defined an extended version of the hadamard regularization @xcite , which permitted a self - consistent derivation of the 3pn equations of motion , but its use still allowed for the presence of arbitrary parameters in the final equations . on the other hand , the hadamard regularization also yielded some arbitrary parameters in the gravitational radiation field of point - particle binaries at the 3pn order , the most important of which being the parameter @xmath11 entering the binary s energy flux @xcite .
let us notice that the regularization ( when dealing with point particles ) and the renormalization ( needed when dealing either with point particles or with extended bodies ) of self - field effects has recurrently plagued the general relativistic problem of motion .
even at the 1pn level , early works often contained incorrect treatments of self - field effects ( see , _ e.g. _ , section 6.14 of @xcite for a review ) .
at the 2pn level , the self - field divergencies are more severe than at the 1pn level .
for instance , they caused ref .
@xcite to incorrectly evaluate the `` static '' ( _ i.e. _ , velocity - independent ) part of the 2pn two - body hamiltonian .
the first correct and complete evaluation of the 2pn dynamics has been obtained by using the riesz analytical continuation method @xcite .
( see @xcite for a detailed discussion of the evaluation of the static 2pn two - body hamiltonian . ) in brief , the riesz analytical continuation method consists of replacing the problematic delta - function stress - energy tensor of a set of point particles @xmath12 , @xmath13^{-1/2 } \ , \delta^{(4 ) } ( x^{\lambda } - y_a^{\lambda } ( s_a))\ , , \label{tdelta4}\ ] ] [ where @xmath14 , @xmath15 by an auxiliary , smoother source @xmath16^{-1/2 } \ , z_a^{(4 ) } ( x^{\lambda } - y_a^{\lambda } ( s_a))\ , .
\label{triesz}\ ] ] [ actually , in the implementation of @xcite , one works with @xmath17 .
] in eq .
( [ triesz ] ) the four - dimensional delta function entering eq .
( [ tdelta4 ] ) has been replaced by the lorentzian is the lorentzian version of the euclidean kernel @xmath18 discussed in appendix [ formulae ] . ]
four - dimensional riesz kernel @xmath19 , which depends on the complex number @xmath20 .
when the real part of @xmath20 is large enough the source @xmath21 is an ordinary function of @xmath22 , which is smooth enough to lead to a well - defined iteration of the harmonically relaxed einstein field equations , involving no divergent integrals linked to the behavior of the integrands when @xmath23 .
one then analytically continues @xmath20 down to @xmath24 , where the kernel @xmath19 tends to @xmath25 .
the important point is that it has been shown @xcite that all the integrals appearing in the 2pn equations of motion are meromorphic functions of @xmath20 which admit a smooth continuation at @xmath26 ( without poles ) .
it was also shown there that the formal construction based on ( [ triesz ] ) does generate , at the 2.5pn level , the metric and equations of motion of @xmath7 `` compact '' bodies ( _ i.e. _ , bodies with radii comparable to their schwarzschild radii ) . the riesz analytic continuation method just sketched works within a normal 4-dimensional space - time ( as recalled by the superscript @xmath27 in ( [ triesz ] ) ) .
however , it was mentioned in @xcite that the same final result ( at the 2.5pn level ) is obtained by replacing @xmath19 by @xmath28 , _
i.e. _ , by formally considering delta - function sources in a space - time of complex dimension @xmath29 .
in other words , at the 2.5pn level , the riesz analytic continuation method is equivalent to the _ dimensional regularization _ method .
however , it was also mentioned at the time @xcite that the generalization of riesz analytic continuation beyond the 2.5pn level did not look straightforward because of the appearance of poles , proportional to @xmath30 , at the 3pn level ( when using harmonic coordinates ) .
recently , damour , jaranowski and schfer @xcite showed how to use dimensional regularization within the adm canonical formalism .
they found that the reduced hamiltonian describing the dynamics of two point masses in space - time dimension @xmath31 was _ finite _ ( no pole part ) as @xmath32 .
they also found that the unique 3pn hamiltonian defined by the analytic continuation of @xmath1 towards 3 had two properties : ( i ) the velocity - dependent terms had the unique structure compatible with global poincar invariance , , originally introduced in the adm approach @xcite , takes the unique value @xmath33 .
this value was obtained in @xcite using the result for the binary energy function in the case of circular orbits , as calculated in the harmonic - coordinates formalism , and also directly from the requirement of poincar invariance in the adm formalism @xcite . ] and ( ii ) the velocity - independent ( `` static '' ) terms led to an unambiguous determination of the unknown adm parameter @xmath9 , namely @xmath34 the fact that the dimensionally regularized 3pn adm hamiltonian ends up being globally poincar invariant is a confirmation of the consistency of dimensional regularization , because this symmetry is not at all manifest within the adm approach which uses a space - plus - time split from the start .
by contrast , the global poincar symmetry is manifest in harmonic coordinates , and indeed the 3pn harmonic - coordinates equations of motion derived in @xcite were found to be manifestly poincar invariant . in the present paper
, we shall show how to implement dimensional regularization ( henceforth often abbreviated as `` dim .
reg . '' or even `` dr '' ) in the computation of the equations of motion in harmonic coordinates , _
i.e. _ , following the same iterative post - newtonian formalism as in refs . @xcite .
similarly to the adm calculation of ref .
@xcite , our strategy will essentially consist of computing the _ difference _ between the @xmath1-dimensional result and the 3-dimensional one @xcite corresponding to hadamard regularization .
this difference is computed in the form of a laurent expansion in @xmath35 , where @xmath1 denotes the spatial dimension .
the main reason for computing the @xmath36-expansion of the difference is that it depends only on the singular behavior of various metric coefficients in the vicinity of the point particles , so that the functions involved in the delicate divergent integrals can all be computed in @xmath1 dimensions in the form of local expansions in powers of @xmath37 or @xmath38 ( where @xmath39 ; @xmath40 , @xmath41 , denoting the locations of the two point masses ) .
dimensional regularization as we use it here can then be seen as a powerful argument for completing the 3-dimensional hadamard - regularization results of @xcite and fixing the value of the unknown parameter .
we leave to future work the task of an exact calculation of the @xmath1-dimensional equations of motion , instead of the calculation of the first few terms in a laurent expansion in @xmath36 around @xmath42 , as done here .
the first step towards such a calculation is taken in appendix [ littleg ] , where we give the explicit expression of the basic quadratically non - linear green function @xmath43 in @xmath1 dimensions . the detailed way of computing the difference between dim .
reg . and hadamard s reg .
will turn out to be significantly more intricate than in the adm case .
this added complexity has several sources .
a first source of complexity is that the harmonic - gauge @xmath1-dimensional calculation will be seen to contain ( as anticipated long ago @xcite ) poles , proportional to @xmath44 , by contrast to the adm calculation which is finite as @xmath32 . a second source of complexity
is that the end results @xcite for the 3-dimensional 3pn equations of motion have been derived using systematically an _ extended _ version of the hadamard regularization method , incorporating both a generalized theory of singular pseudo - functions and their associated ( generalized ) distributional derivatives @xcite , and an improved definition of the finite part as @xmath45 , say @xmath46_1 $ ] , of a singular function @xmath47 , designed so as to respect the global poincar symmetry of the problem @xcite .
we shall then find it technically convenient to subtract the various contributions to the end results of @xcite which arose because of the specific use of the extended hadamard regularization methods of @xcite before considering the difference with the @xmath1-dimensional result .
a third source of added complexity ( with respect to the adm case derived in @xcite and used ( in its @xmath1-dimensional generalization ) in @xcite was written , on purpose , in a way which does not involve any hidden distributional terms ( the only delta - function contributions it contains being explicit contact terms @xmath48 ) .
this allowed one to estimate the difference between the @xmath1-dimensional hamiltonian @xmath49 and the hadamard - regularized 3-dimensional one hr@xmath50 $ ] without having to worry about distributional derivatives .
however , as a check on the consistency of dim .
, the authors of @xcite did perform another calculation of @xmath51 based on a starting form of the hamiltonian which involved hidden distributional terms , with the same final result . ] ) comes from the presence in the harmonic - gauge integrals we shall evaluate of `` hidden - distributional '' terms in the integrands . by hidden distributional terms we mean terms proportional to the second spatial derivatives of the poisson kernel @xmath52 , or to the fourth spatial derivatives of the iterated poisson kernel @xmath53 . such terms , @xmath54 or @xmath55 , considered as schwartz distributional derivatives @xcite , contain pieces proportional to the delta function @xmath56 , which need to be treated with care .
the generalized distributional derivative defined in @xcite , and used to compute the end results of @xcite , led to an improved way , compared to the normal schwartz distributional derivative , of evaluating contributions coming from the product of a singular function and a derivative of the type @xmath57 or @xmath58 , and more generally of any derivatives of singular functions in a certain class .
we shall find it convenient to subtract these additional non - schwartzian contributions to the 3pn equations of motion before applying dimensional regularization .
however , we shall note at the end that dim .
reg . automatically incorporates all of these non - schwartzian contributions .
a fourth , but minor , source of complexity concerns the dependence of the end results of @xcite for the 3pn acceleration of the first particle ( label @xmath59 ) , say @xmath60 , on two arbitrary length scales @xmath61 and @xmath62 , and on the `` ambiguity '' parameter @xmath4 .
explicitly , we define @xmath63 \equiv \text{r.h.s . of eq.~(7.16 ) in ref .
\cite{blanchet:2000ub}}\ , .
\label{a1bf}\ ] ] here the acceleration is considered as a function of the two masses @xmath64 and @xmath65 , the relative distance @xmath66 ( where @xmath67 is the unit vector directed from particle 2 to particle 1 ) , the two coordinate velocities @xmath68 and @xmath69 , and also , as emphasized in ( [ a1bf ] ) , the parameter @xmath4 as well as two regularization length scales @xmath70 and @xmath71 .
the latter length scales enter the equations of motion at the 3pn level through the logarithms @xmath72 and @xmath73 .
they come from the regularization as the field point @xmath74 tends to @xmath75 or @xmath76 of poisson - type integrals ( see section [ hadamardpoisson ] below )
. the length scales @xmath70 , @xmath71 are `` pure gauge '' in the sense that they can be removed by the effect induced on the world - lines of a coordinate transformation of the bulk metric @xcite . on the other hand
, the dimensionless parameter @xmath4 entering the final result ( [ a1bf ] ) corresponds to genuine physical effects .
it was introduced by requiring that the 3pn equations of motion admit a conserved energy ( and more generally be derivable from a lagrangian ) .
this extra requirement imposed _ two relations _ between the two length scales @xmath61 , @xmath62 and two other length scales @xmath77 , @xmath78 entering originally into the formalism , namely the constants @xmath77 and @xmath78 parametrizing the hadamard partie finie of an integral as defined by eq .
( [ pf ] ) below .
these relations were found to be of the form @xmath79 where the so introduced _ single _ dimensionless parameter @xmath4 has been proved to be a purely numerical coefficient ( independent of the two masses ) . when estimating the difference between dim .
reg . and hadamard reg
. it will be convenient to insert eq .
( [ lnr2s2 ] ) into ( [ a1bf ] ) and to reexpress the acceleration of particle 1 in terms of the _ original _ regularization length scales entering the hadamard regularization of @xmath80 , which were in fact @xmath61 and @xmath78 .
thus we can consider alternatively @xmath81 \equiv \mathbf{a}_1^\text{bf } \bigl[\lambda ; r_1 ' , r'_2 ( s_2 , \lambda)\bigr]~~\text{and $ 1\leftrightarrow 2$}\ , , \label{foncta1bf}\ ] ] where the regularization constants are subject to the constraints ( [ lnr2s2 ] ) [ we will then check that the @xmath4-dependence on the r.h.s . of ( [ foncta1bf ] )
disappears when using eq .
( [ lnr2s2 ] ) to replace @xmath62 as a function of @xmath78 and @xmath4 ] .
our strategy will consist of _ two steps_. the _ first step _ consists of subtracting all the extra contributions to eq .
( [ a1bf ] ) , or equivalently eq . ( [ foncta1bf ] ) , which were specific consequences of the extended hadamard regularization defined in @xcite .
as we shall detail below , there are _ seven _ such extra contributions @xmath83 , @xmath84 .
essentially , subtracting these contributions boils down to estimating the value of @xmath80 that would be obtained by using a `` pure '' hadamard regularization , together with schwartz distributional derivatives .
such a `` pure hadamard - schwartz '' ( phs ) acceleration was in fact essentially the result of the first stage of the calculation of @xmath80 , as reported in the ( unpublished ) thesis @xcite .
it is given by @xmath85=\mathbf{a}_1^\text{bf } [ r'_1 , s_2]-\sum_{a=1}^7\delta^a\mathbf{a}_1\ , .
\label{accph}\ ] ] the _ second step _ of our method consists of evaluating the laurent expansion , in powers of @xmath86 , of the _ difference _ between the @xmath1-dimensional and the pure hadamard - schwartz ( 3-dimensional ) computations of the acceleration @xmath80 .
we shall see that this difference makes a contribution only when a term generates a _ pole _ @xmath87 , in which case dim .
reg . adds an extra contribution , made of the pole and the finite part associated with the pole [ we consistently neglect all terms @xmath88 .
one must then be especially wary of combinations of terms whose pole parts finally cancel ( `` cancelled poles '' ) but whose dimensionally regularized finite parts generally do not , and must be evaluated with care .
we denote the above defined difference @xmath89\equiv\mathcal{d}\mathbf{a}_1 \bigl[\varepsilon,\ell_0;\lambda ; r_1 ' , r'_2\bigr]\ , .
\label{deltaacc}\ ] ] it depends both on the hadamard regularization scales @xmath70 and @xmath78 ( or equivalently on @xmath4 and @xmath70 , @xmath71 ) and on the regularizing parameters of dimensional regularization , namely @xmath36 and the characteristic length @xmath90 associated with dim . reg . and
introduced in eq .
( [ l0 ] ) below .
we shall explain in detail below the techniques we have used to compute @xmath91 ( see section [ difference ] ) .
finally , our main result will be the explicit computation of the @xmath36-expansion of the dim .
acceleration as @xmath92 = \mathbf{a}_1^\text{phs } [ r'_1 , s_2 ] + \mathcal{d}\mathbf{a}_1 [ \varepsilon , \ell_0 ; r'_1 , s_2]\ , .
\label{a1dimreg}\ ] ] with this result in hands , we shall prove ( in section [ renormalise ] ) two theorems .
* * theorem 1**__the pole part @xmath93 of the dimensionally - regularized acceleration ( [ a1dimreg ] ) , as well as of the metric field @xmath94 outside the particles , can be re - absorbed ( _ i.e. _ , renormalized away ) into some shifts of the two `` bare '' world - lines : @xmath95 , with , say , @xmath96 ( `` minimal subtraction '' ; ms ) , so that the result , expressed in terms of the `` dressed '' quantities , is finite when @xmath97 . _
_ the situation in harmonic coordinates is to be contrasted with the calculation in adm - type coordinates within the hamiltonian formalism @xcite , where it was shown that all pole parts directly cancel out in the total 3pn hamiltonian ( no shifts of the world - lines were needed ) .
the central result of the paper is then as follows . * * theorem 2**__the `` renormalized '' ( finite ) dimensionally - regularized acceleration is physically equivalent to the extended - hadamard - regularized acceleration ( end result of ref .
@xcite ) , in the sense that there exist some shift vectors @xmath98 and @xmath99 , such that @xmath100 = \lim_{\varepsilon\rightarrow 0 } \ , \bigl[\mathbf{a}_1^\mathrm{dr } [ \varepsilon , \ell_0 ] + \delta_{\bm{\xi } ( \varepsilon , \ell_0 ; r'_1 , r'_2 ) } \ , \mathbf{a}_1 \bigr ] \label{eta}\ ] ] ( where @xmath101 denotes the effect of the shifts on the acceleration below ) . ] ) , if and only if the heretofore unknown parameter @xmath4 entering the harmonic - coordinates equations of motion takes the value @xmath102 _ _ the precise shifts @xmath103 needed in theorem 2 involve not only a pole contribution @xmath93 , which defines the `` minimal '' ( ms ) shifts considered in theorem 1 , but also a finite contribution when @xmath104 .
their explicit expressions read : @xmath105 \mathbf{a}_{n1}~~\text{and}~~1\leftrightarrow 2\,,\ ] ] where @xmath106 is the usual newton s constant [ see eq .
( [ l0 ] ) below ] , @xmath107 denotes the acceleration of the particle 1 ( in @xmath1 dimensions ) at the newtonian level , and @xmath108 depends on the euler constant @xmath109 .
[ the detailed proofs of theorems 1 and 2 will consist of our investigations expounded in the successive sections of the paper , and will be completed at the end of sections [ shift ] and [ lambdadetermined ] respectively , taking also into account the results of section [ kinetic ] . ]
notice that an alternative way of presenting our central result is to say that , in fact , each choice of specific renormalization prescription ( within dim .
reg . ) , such as `` minimal subtraction '' as assumed in theorem 1 for conceptual simplicity , below to use a modified minimal subtraction that we shall denote @xmath110 . ] leads to renormalized equations of motion which depend only on the dim .
characteristic length scale @xmath90 through the logarithm @xmath111 , and that any of these renormalized equations of motion are physically equivalent to the final results of @xcite .
in particular , this means , as we shall see below , that each choice of renormalization prescription within dim .
reg . determines the two regularization length scales @xmath61 , @xmath62 entering eq .
( [ a1bf ] ) .
of course , what is important is not the particular values these constants can take in a particular renormalization scheme [ indeed @xmath61 and @xmath62 are simply `` gauge '' constants which can anyway be removed by a coordinate transformation ] , but the fact that the different renormalization prescriptions yield equations of motion falling into the `` parametric '' class ( _ i.e. _ , parametrized by @xmath61 and @xmath62 ) of equations of motion obtained in @xcite . an alternative way to phrase the result ( [ eta])-([lambda ] ) , is to combine eqs .
( [ accph ] ) and ( [ a1dimreg ] ) in order to arrive at @xmath112 + \delta_{\bm{\xi } ( \varepsilon , \ell_0 ; r'_1 , r'_2 ) } \ , \mathbf{a}_1 \bigr ] = \sum_{a=1}^7\delta^a\mathbf{a}_1\ , .
\label{equiveta}\ ] ] under this form one sees that the sum of the additional terms @xmath113 differs by a mere shift , _ when and only when _
@xmath4 takes the value ( [ lambda ] ) , from the specific contribution @xmath91 we shall evaluate in this paper , which comes directly from dimensional regularization .
therefore one can say that , when @xmath114 , the extended - hadamard regularization @xcite is in fact ( physically ) equivalent to dimensional regularization .
however the extended - hadamard regularization is incomplete , both because it is unable to determine @xmath4 , and also because it necessitates some `` external '' requirements such as the imposition of the link ( [ lnr2s2 ] ) in order to ensure the existence of a conserved energy and in fact of the ten first integrals linked to the poincar group .
by contrast dim .
reg . succeeds automatically ( without extra inputs ) in guaranteeing the existence of the ten conserved integrals of the poincar group , as already found in ref .
@xcite . in view of the necessary link ( [ lambdaomegas ] ) provided by the equivalence between the adm - hamiltonian and the harmonic - coordinates equations of motion ,
our result ( [ lambda ] ) is in perfect agreement with the previous result ( [ omegas ] ) obtained in @xcite . is a complicated rational fraction while @xmath9 is so simple
this is because @xmath9 was introduced precisely to measure the amount of ambiguities of certain integrals , and that the adm hamiltonian reported in @xcite was put in a minimally ambiguous form , already in three dimensions , for which an _ a posteriori _ look at the `` ambiguities '' discussed in the appendix a of @xcite already showed that @xmath6 . by contrast , @xmath4 has been introduced as the only possible unknown constant in the link between the four arbitrary scales @xmath115 ( which has _ a priori _ nothing to do with ambiguities of integrals ) , in a framework where the use of the extended hadamard regularization makes in fact the calculation to be unambiguous . ]
our result is also in agreement with the recent finding of itoh and futamase @xcite ( see also @xcite ) , who derived the 3pn equations of motion in harmonic gauge using a `` surface - integral '' approach , aimed at describing _ extended _ relativistic compact binary systems in the strong - field point particle limit .
the surface - integral approach of refs .
@xcite is interesting because , like the matching method used at 2.5pn order in @xcite , it is based on the physical notion of extended compact bodies . in this respect , we recall that the matching method used in @xcite showed that the internal structure ( love numbers ) of the constituent bodies would start influencing the equations of motion of ( non - spinning ) compact bodies only at the 5pn level .
effacement property _ strongly suggests that it is possible to model , in a physically preferred manner , two compact bodies as being two point - like particles , described by two masses and two world - lines , up to the 4.5pn level included .
it remains , however , to prove that the dimensional regularization of delta - function sources does yield the physically unique equations of motion of two compact bodies up to the 4.5 pn order .
the work @xcite proved it at the 2.5 pn level , and the agreement of the present results with those of @xcite indicates that this is also true at the 3pn level .
besides the independent confirmation of the value of @xmath9 or @xmath4 , let us also mention that our work provides a confirmation of the _ consistency _ of dim . reg . , because our explicit calculations [ which involved combinations of hundreds of laurent expansions of the form @xmath116 are entirely different from the ones of @xcite : we use harmonic coordinates ( instead of adm - type ones )
, we work at the level of the equations of motion ( instead of the hamiltonian ) , we use a different form of einstein s field equations and we solve them by a different iteration scheme .
finally , from a practical point of view our confirmation of the value of @xmath9 or@xmath4 allows one to use the full 3pn accuracy in the analytical computation of the dynamics of the last orbits of binary orbits @xcite .
it remains , however , the task of computing , using dimensional regularization , the parameter @xmath11 entering the 3.5pn gravitational energy flux @xcite to be able to have full 3.5pn accuracy in the computation of the gravitational waveforms emitted by inspiralling compact binaries ( see , _
e.g. _ , @xcite and references therein ) .
the organization of the paper is as follows . in section [ fieldeq ]
we derive our basic 3pn solution of the field equations for general fluid sources in @xmath1 spatial dimensions , using @xmath1-dimensional generalizations of the elementary potentials introduced in ref .
section [ hadamard ] collects all the additional terms included in @xcite which are due specifically to the extended hadamard regularization , and derives the pure hadamard - schwartz ( phs ) contribution to the equations of motion .
the differences between the dimensional and phs regularizations for all the potentials and their gradients are computed in section [ difference ] .
then the dim .
equations of motion are obtained in section [ results ] , where we comment also on their interpretation in terms of space - time diagrams .
section [ renormalise ] is devoted to the renormalization of the dim .
equations by means of suitable shifts of the particles world - lines , and to the equivalence with the end results of @xcite when eq .
( [ lambda ] ) holds . at this stage ,
the proofs of theorems 1 and 2 stated above are finally completed .
we end the paper with some conclusions ( section [ conclusions ] ) and three appendices .
appendix [ expfieldeq ] provides further material on the @xmath1-dimensional metric and geodesic equation , appendix [ formulae ] gives a compendium of useful formulae for working in @xmath1 dimensions , and appendix [ littleg ] generalizes the well - known quadratic - order elementary kernel @xmath117 to @xmath1 dimensions .
the latter calculation of the @xmath1-dimensional kernel @xmath118 is not directly employed in the present paper , but represents a first step in obtaining the equations of motion in any dimension @xmath1 ( not necessarily of the form @xmath119 ) .
this section is devoted to the field equations of general relativity in @xmath120 space - time dimensions , and to the geodesic equation describing the motion of point particles .
we use the sign conventions of ref .
@xcite , and in particular our metric signature is mostly @xmath121 .
space - time indices are denoted by greek letters , and spatial indices by latin letters ( @xmath122 run from 1 to @xmath1 ) .
a summation is understood for any pair of repeated indices .
we work in the harmonic gauge , which is such that @xmath123 as usual , @xmath124 denotes the inverse metric and @xmath125 the christoffel symbols . using this gauge condition , one can easily prove that the ricci tensor reads _ in any dimension _
@xmath126 where a comma denotes partial derivation .
note that the spatial dimension @xmath1 does not appear explicitly in this expression , whereas some @xmath1-dependent coefficients do appear when expressing the ricci tensor in terms of the so - called `` gothic '' metric @xmath127 [ see eq .
( [ riccigoth ] ) in appendix [ expfieldeq ] below ] . in any dimension , the einstein field equations read @xmath128 where @xmath129 denotes the matter stress - energy tensor , given by the functional derivative @xmath130 of the matter action @xmath131 with respect to the metric tensor .
_ by definition _
, @xmath132 denotes the constant involved in eq .
( [ einstein ] ) , which shows that its dimension is such that @xmath133 where @xmath106 is the usual newton constant ( in 3 spatial dimensions ) and @xmath90 is an arbitrary length scale .
this scale will be involved in our dimensionally regularized results below , but we will finally show that the physical observables do not depend on it . as is well known , the combination of eq .
( [ einstein ] ) with its trace allows us to rewrite it as @xmath134 in which the spatial dimension @xmath1 now appears explicitly .
we wish to expand in powers of @xmath135 the field equations resulting from ( [ ricciharm ] ) and ( [ ricci ] ) .
the basic idea is to introduce a sequence of `` elementary potentials '' , @xmath136 , @xmath137 , @xmath138 , which allow one to parametrize conveniently the successive post - minkowskian contributions to the metric @xmath139 .
for instance , at the first post - minkowskian order it is convenient to parametrize the metric as @xmath140 where the so - introduced elementary potentials @xmath136 and @xmath137 satisfy equations of the form @xmath141 where @xmath142 denotes the flat dalembertian and where , _ by definition _ , the sources @xmath143 and @xmath144 are linear combinations of the contravariant components @xmath129 of the stress - energy tensor of the matter .
let us underline that the factor @xmath145 in these equations is a _
choice_. we could of course introduce here a functional dependence on the spatial dimension @xmath1 , for instance by replacing the factor @xmath146 by the surface of the unit @xmath147-dimensional sphere [ see eq .
( [ omegad1 ] ) in appendix [ formulae ] ] , but this would only complicate the intermediate expressions without changing our final result . the matter sources @xmath143 and @xmath144 _ defined _ by eqs .
( [ glowest ] ) , ( [ vlowest ] ) read ( in @xmath1 spatial dimensions ) : @xmath148 the definition for @xmath149 has been added for future use . note that @xmath144 and @xmath149 take the same forms as usual in 3 dimensions ( see eqs .
( 3.9 ) of ref .
@xcite ) , but that the definition of @xmath143 involves an explicit dependence on @xmath1 .
conversely , the first and third of these equations allow us to express @xmath150 in terms of the above matter sources : @xmath151 . a simple consequence of the expression of @xmath143 is that the @xmath1-dimensional _ newtonian potential _ generated by a mass @xmath152 located at @xmath40 reads explicitly @xmath153 where the factor @xmath154 comes from @xmath143 ( _
i.e. _ , from einstein s equations ) , while the factor @xmath155 comes from the expression of the green function of the laplacian in @xmath1 dimensions ( see eq .
( [ green ] ) below and appendix [ formulae ] ) .
we give below the simplest forms of the metric and of the potential equations that we could obtain .
we will explain afterwards which rules we followed to simplify them .
let us first define the useful combination @xmath156 then the metric components can be written in a rather compact form : [ metric ] @xmath157 + \frac{8\hat r_i}{c^5 } + \frac{16}{c^7}\left[\hat y_i+\frac{1}{2}\hat w_{ij}v_j\right]\right\ } + { \mathcal{o}}\left(\frac{1}{c^9}\right),\nonumber\\ \label{g0i}\\ g_{ij}&=&e^{\frac{2\mathcal{v}}{(d-2)c^2}}\left\{\delta_{ij } + \frac{4}{c^4 } \hat w_{ij } + \frac{16}{c^6 } \left[\hat z_{ij}-v_i v_j + \frac{1}{2(d-2)}\ , \delta_{ij } v_k v_k \right]\right\ } + { \mathcal{o}}\left(\frac{1}{c^8}\right ) .
\label{gij}\end{aligned}\ ] ] the various elementary potentials @xmath136 , @xmath137 , @xmath158 , @xmath159 , @xmath160 , @xmath161 , @xmath162 , @xmath163 and @xmath164 introduced in these definitions are @xmath1-dimensional analogues of those used in eqs .
( 3.24 ) of ref .
actually , an extra potential is needed for @xmath165 , and it has been denoted @xmath158 in eq .
( [ calv ] ) above .
we give in appendix [ expfieldeq ] the explicit expansion of this metric in powers of @xmath135 , as well as its inverse @xmath166 and its determinant @xmath167 , which can be useful for future works .
note that the first post - newtonian order of the spatial metric , @xmath168 + { \mathcal{o}}(1/ c^4)$ ] , explicitly depends on @xmath1 contrary to our choice ( [ glowest ] ) for @xmath169 .
this dissymmetry between @xmath169 and @xmath170 is imposed by the field equations ( [ ricci ] ) .
the successive post - newtonian truncations of the field equations ( [ ricciharm])-([ricci ] ) give us the sources for these various potentials .
the equations for @xmath171 and @xmath172 have already been written in eqs .
( [ vlowest ] ) above .
we get for the remaining potentials : [ potentialeq ] @xmath173\nonumber\\ & & + \hat w_{ij}\ , \partial_{ij}v + 2 v_i\,\partial_t\partial_i v + \frac{1}{2}\left(\frac{d-1}{d-2}\right ) v \partial^2_t v \nonumber\\ & & + \frac{d(d-1)}{4(d-2)^2}\left(\partial_t v\right)^2 -2 \partial_i v_j\ , \partial_j v_i\ , \label{dalx}\\ \box\hat z_{ij}&=&-\frac{4\pi g}{d-2}\ , v\left(\sigma_{ij } -\delta_{ij}\,\frac{\sigma_{kk}}{d-2}\right ) -\frac{d-1}{d-2}\ , \partial_t v_{(i}\ , \partial_{j)}v + \partial_i v_k\ , \partial_j v_k\nonumber\\ & & + \partial_k v_i\ , \partial_k v_j -2 \partial_k v_{(i}\ , \partial_{j)}v_k -\frac{\delta_{ij}}{d-2}\ , \partial_k v_m \left(\partial_k v_m -\partial_m v_k\right)\nonumber\\ & & -\frac{d(d-1)}{8(d-2)^3}\ , \delta_{ij}\left(\partial_t v\right)^2 + \frac{(d-1)(d-3)}{2(d-2)^2}\ , \partial_{(i } v\partial_{j ) } k\ , , \label{dalzij}\\ \box\hat y_i&=&-4\pi g \biggl[-\frac{1}{2}\left(\frac{d-1}{d-2}\right)\sigma\hat r_i -\frac{(5-d)(d-1)}{4(d-2)^2}\ , \sigma v v_i + \frac{1}{2}\ , \sigma_k\hat w_{ik } + \frac{1}{2}\sigma_{ik } v_k\nonumber\\ & & \hphantom{-4\pi g \biggl[}+\frac{1}{2(d-2)}\ , \sigma_{kk}v_i -\frac{d-3}{(d-2)^2}\ , \sigma_i \left(v^2 + \frac{5-d}{2}\ , k\right)\biggr ] \nonumber\\ & & + \hat w_{kl}\ , \partial_{kl } v_i -\frac{1}{2}\left(\frac{d-1}{d-2}\right ) \partial_t\hat w_{ik}\ , \partial_k v + \partial_i\hat w_{kl}\ , \partial_k v_l -\partial_k\hat w_{il}\ , \partial_l v_k \nonumber\\ & & -\frac{d-1}{d-2}\ , \partial_k v \partial_i \hat r_k -\frac{d(d-1)}{4 ( d-2)^2}\ , v_k\ , \partial_i v \partial_k v -\frac{d(d-1)^2}{8 ( d-2)^3}\ , v\partial_t v\partial_i v \nonumber\\ & & -\frac{1}{2}\left(\frac{d-1}{d-2}\right)^2 v \partial_k v \partial_k v_i + \frac{1}{2}\left(\frac{d-1}{d-2}\right ) v \partial^2_tv_i + 2 v_k\ , \partial_k\partial_t v_i \nonumber\\ & & + \frac{(d-1)(d-3)}{(d-2)^2}\ , \partial_k k \partial_i v_k + \frac{d(d-1)(d-3)}{4(d-2)^3 } \left(\partial_t v\partial_i k + \partial_i v\partial_t k\right ) , \label{dalyi}\\ \box\hat t&=&-4\pi g\biggl[\frac{1}{2(d-1)}\ , \sigma_{ij } \hat w_{ij } + \frac{5-d}{4(d-2)^2}\ , v^2\sigma_{ii } + \frac{1}{d-2}\ , \sigma v_i
v_i -\frac{1}{2}\left(\frac{d-3}{d-2}\right)\sigma\hat x\nonumber\\ & & \hphantom{-4\pi g\biggl[}-\frac{1}{12}\left(\frac{d-3}{d-2}\right)^3 \sigma v^3 -\frac{1}{2}\left(\frac{d-3}{d-2}\right)^3 \sigma v k + \frac{(5-d)(d-3)}{2(d-1)(d-2)}\ , \sigma_i v_i v\nonumber\\ & & \hphantom{-4\pi g\biggl[}+\frac{d-3}{d-1}\ , \sigma_i\hat r_i -\frac{d-3}{2(d-2)^2}\ , \sigma_{ii } k\biggr ] + \hat z_{ij}\ , \partial_{ij}v + \hat r_i\ , \partial_t\partial_i v\nonumber\\ & & -2 \partial_i v_j\ , \partial_j\hat r_i -\partial_i v_j\ , \partial_t\hat w_{ij } + \frac{1}{2}\left(\frac{d-1}{d-2}\right ) v v_i\ , \partial_t\partial_i v + \frac{d-1}{d-2}\ , v_i\ , \partial_j v_i\ , \partial_j v\nonumber\\ & & + \frac{d(d-1)}{4(d-2)^2}\ , v_i\ , \partial_t v\ , \partial_i v + \frac{1}{8}\left(\frac{d-1}{d-2}\right)^2 v^2\partial^2_t v + \frac{d(d-1)^2}{8(d-2)^3}\ , v\left(\partial_t v\right)^2\nonumber\\ & & -\frac{1}{2}\left(\partial_t v_i\right)^2 -\frac{(d-1)(d-3)}{4(d-2)^2}\ , v \partial^2_t k -\frac{d(d-1)(d-3)}{4(d-2)^3}\ , \partial_t v\ , \partial_t k\nonumber\\ & & -\frac{(d-1)(d-3)}{4(d-2)^2}\ , k \partial^2_t v -\frac{d-3}{d-2}\ , v_i\ , \partial_t\partial_i k -\frac{1}{2}\left(\frac{d-3}{d-2}\right)\hat w_{ij}\ , \partial_{ij } k\ , . \label{dalt}\end{aligned}\ ] ] in eq .
( [ dalzij ] ) , parentheses around indices mean their symmetrization , _
i.e. _ , @xmath174 . for @xmath42 ,
the above set of equations ( [ potentialeq ] ) reduces to eqs .
( 3.26 ) and ( 3.27 ) of ref .
the order of the terms and their writing has been chosen to be as close as possible to this reference . the harmonic gauge conditions ( [ harmgauge ] ) impose the following differential identities between the potentials : [ gaugeidentities ] @xmath175\nonumber\\ & & \hphantom{\partial_t\biggl\{}+\frac{2}{c^4 } \left(\frac{d-1}{d-2}\right)\biggl[\hat x + \frac{d-2}{d-1}\,\hat z -\frac{d-3}{d-1}\,v_k v_k + \frac{1}{2}\,v\hat w\nonumber\\ & & \hphantom{\partial_t\biggl\{+\frac{2}{c^4 } \left(\frac{d-1}{d-2}\right)\biggl[}+\frac{(d-1)^2}{6(d-2)}\ , v^3 -\frac{(d-1)(d-3)}{(d-2)^2}\ , v k \biggr ] \biggr\}\nonumber\\ & & + \partial_i\biggl\{v_i + \frac{2}{c^2}\left[\hat r_i + \frac{1}{2}\left(\frac{d-1}{d-2}\right ) v v_i\right]\nonumber\\ & & \hphantom{+\partial_i\biggl\{}+\frac{4}{c^4}\biggl[\hat y_i -\frac{1}{2}\ , \hat w_{ij } v_j + \frac{1}{2}\ , \hat w v_i + \frac{1}{2}\left(\frac{d-1}{d-2}\right ) v \hat r_i + \frac{1}{4}\left(\frac{d-1}{d-2}\right)^2 v^2 v_i\nonumber\\ & & \hphantom{+\partial_i\biggl\{+\frac{4}{c^4}\biggl[}- \frac{(d-1)(d-3)}{2 ( d-2)^2}\ , k v_i
\biggr ] \biggr\ } = { \mathcal{o}}\left(\frac{1}{c^6}\right ) , \label{divi}\\ g^{\mu\nu}\gamma^i_{\mu\nu } = 0&\rightarrow&\partial_t\left\ { v_i + \frac{2}{c^2}\left [ \hat r_i + \frac{1}{2}\left(\frac{d-1}{d-2}\right)v v_i \right]\right\}\nonumber\\ & & + \partial_j\left\ { \hat w_{ij } -\frac{1}{2}\ , \hat w \delta_{ij } + \frac{4}{c^2}\left[\hat z_{ij } -\frac{1}{2}\ , \hat z \delta_{ij}\right ] \right\ } = { \mathcal{o}}\left(\frac{1}{c^4}\right ) , \label{djwij}\end{aligned}\ ] ] @xmath176 where @xmath177 and @xmath178 denote the traces of potentials @xmath159 and @xmath162 . for @xmath42 ,
these identities reduce to eqs .
( 3.28 ) of ref .
@xcite . in this paper
we shall check ( see sections [ iteratedeinstein ] and [ poles ] ) that all the dimensionally - regularized potentials we use obey , at the indicated accuracy , the differential identities ( [ gaugeidentities ] ) equivalent to the harmonic gauge conditions . in order to simplify as much as possible the above equations ( [ potentialeq ] ) for the potentials
, we used the following rules : 1 .
we used the harmonic gauge condition ( [ divi ] ) to replace everywhere @xmath179 in terms of @xmath180 and higher post - newtonian order terms , and the gauge condition ( [ djwij ] ) to replace @xmath181 in terms of @xmath182 and @xmath183 [ our knowledge of the higher order terms @xmath184 in eq .
( [ djwij ] ) was actually not necessary for the simplification of eqs .
( [ potentialeq ] ) ] .
we also used the lowest order terms of eqs .
( [ gaugeidentities ] ) to simplify their own higher order contributions .
if the source of a potential @xmath185 contained a double ( contracted ) gradient of the form @xmath186 , where @xmath20 and @xmath187 were two lower - order potentials , we got rid of the double gradient by defining another potential @xmath188 .
we could then write its equation as @xmath189 , in which @xmath190 and @xmath191 were replaced by their own explicit sources .
the contribution proportional to @xmath192 was then transferred into the source of a higher order potential .
3 . at order @xmath193 , equation ( [ ricciharm])-([ricci ] ) for @xmath194 ( _
i.e. _ , for @xmath195 ) contains the term @xmath196 , that we introduced in the source of potential @xmath161 , eq .
( [ dalx ] ) . in all other equations involving the same source @xmath196
, we used @xmath197 to eliminate it , instead of reintroducing it in the sources of other potentials .
this is the reason why @xmath161 is involved in the spatial metric @xmath170 too at order @xmath193 [ _ via _ the exponential of @xmath198 in eq .
( [ gij ] ) ] , and why @xmath199 appears again in @xmath169 at order @xmath200 .
see the expanded form of the metric ( [ metricexp ] ) in appendix [ expfieldeq ] .
4 . in the equation for @xmath194 at order @xmath200
, we _ chose _ to eliminate a source proportional to @xmath201 , by including an all - integrated term @xmath202 in the definition of @xmath169 , eq .
( [ g00 ] ) . on the other hand
, we could not eliminate at the same time the source term proportional to @xmath203 in @xmath195 , although it involves a double ( contracted ) gradient too .
this is the reason why such a term appears in eq .
( [ dalt ] ) .
the above simplification rules have been applied systematically with a single exception .
indeed , eq . ( [ dalzij ] ) for @xmath204 involves double ( contracted ) gradients @xmath205 and @xmath206 .
therefore , the application of rule ( ii ) would have yielded another potential @xmath207 such that no double gradient appears in its source ( but extra compact sources @xmath208 and @xmath209 would have been involved ) .
although this modified potential @xmath210 actually simplifies slightly some equations ( but not all of them ) , we have chosen to use @xmath162 which is the direct @xmath1-dimensional analogue of the potential written in eq .
( 3.27c ) of ref .
indeed , as explained in the following sections , the 3-dimensional results of this reference will be necessary for our @xmath1-dimensional calculations , and it is more convenient to keep the same notation .
notice also that after the above simplifications , the resulting metric involves only potentials which are at most cubically non - linear ( like for the term @xmath196 in the potential @xmath161 using the terminology of section [ dimregstat ] below ) . there is no need to introduce any quartically non - linear elementary potential because it turns out that it is possible to `` integrate directly '' all of them ( at the 3pn level ) in terms of other potentials .
the only quartic contributions are the terms composed of @xmath211 and @xmath212 in the metric component @xmath169 [ see eq .
( [ g00exp ] ) in appendix [ expfieldeq ] ] .
the fact that there are no intrinsically quartic potentials at the 3pn order made the closed - form calculation in @xcite possible .
we shall comment more on this interesting fact in section [ iteratedeinstein ] .
let us now apply the general potential parametrization of the metric defined above to the specific case of ( monopolar ) point particles , _
i.e. _ , to the action @xmath213 the stress - energy tensor @xmath214 \ , \delta \ , s_\text{matter } / \delta \ , g_{\mu\nu } ( x)$ ] deduced from this action reads @xmath215^{-\frac{1}{2 } } \ , \delta^{(d+1 ) } ( x^{\lambda } - y_a^{\lambda } ( s_a))\ , , \label{tmunu}\ ] ] where @xmath216 is ( @xmath217 times ) the proper time along the world - line of the @xmath218 particle and where @xmath219 is the dirac density in @xmath120 dimensions ( @xmath220 ) . here
, we take advantage of the fact ( emphasized in @xcite ) that dim .
respects the basic properties of the algebraic and differential calculus : associativity , commutativity and distributivity of point - wise addition and multiplication , leibniz s rule , schwarz s rule @xmath221 , integration by parts , etc .
in addition , the post - newtonian expansion of @xmath222 yields `` @xmath1-dimensional functions '' which are formally as smooth as wished ( by taking the real part of @xmath1 small enough ) in the vicinity of the world - lines : see for instance eq .
( [ newton ] ) .
this allows us to work with self - gravitating point particles essentially as if they were _ test _ particles .
for instance , we can use @xmath223 \ , \delta^{(d ) } \ , ( \mathbf{x}-\mathbf{y}_a ) = f [ g_{\mu\nu } ( y_a ) ] \ , \delta^{(d ) } \ , ( \mathbf{x}-\mathbf{y}_a)$ ] .
in particular , the @xmath224-evaluated determinant factor @xmath225^{-\frac{1}{2}}$ ] in ( [ tmunu ] ) came from the field - point dependent factor @xmath226^{-\frac{1}{2}}$ ] in the definition of @xmath227 .
similarly , the usual derivation of the equations of motion of a test particle formally generalizes to the case of self - gravitating point particles in @xmath1 dimensions .
one then finds that the equations of motion of point particles can equivalently be written as @xmath228 or as the usual geodesic equations .
the latter can be written either in covariant form @xmath229 ( @xmath230 ) , _ i.e. _ , explicitly @xmath231 \ , \frac{dy_a^{\mu}}{ds_a } \ , \frac{dy_a^{\nu}}{ds_a } = 0 \ , , \label{geod2}\ ] ] where @xmath232 as usual , or in the explicit form corresponding to using the coordinate time @xmath233 as parameter along the world - lines , which is easily derived from the covariant expression with a lower index , @xmath234 . like in 3 dimensions , cf .
( 3.32)-(3.33 ) of ref .
@xcite , it can thus be put in the form @xmath235 where @xmath236 @xmath237 denoting the coordinate velocity .
let us emphasize again that in @xmath1 dimensions , all the non - linear functions of @xmath238 and @xmath239 that will enter our calculation of ( [ eqgeod])-([pifi ] ) can be treated as in the @xmath240 evaluation of smooth functions of @xmath241 .
for instance , denoting for simplicity @xmath242 , the newtonian approximation , say @xmath243 , of the basic scalar potential @xmath244 , reads , in the vicinity of @xmath245 , @xmath246 where @xmath247 is ( in any @xmath1 ) an indefinitely differentiable function of @xmath241 near @xmath75 .
analytically continuing @xmath1 to sufficiently `` low '' ( and even with negative real part , if needed ) values , we see not only that @xmath248_{\mathbf{x } = \mathbf{y}_1 } = u_2 ( \mathbf{y}_1)$ ] , but that @xmath249_{\mathbf{x } = \mathbf{y}_1 } = ( u_2 ( \mathbf{y}_1))^n$ ] , and , _
e.g. _ , @xmath250_{\mathbf{x } = \mathbf{y}_1 } = ( u_2 ( \mathbf{y}_1))^p \ , \partial_i \ , u_2 ( \mathbf{y}_1)$ ] , etc .
although the expressions ( [ pifi ] ) do not depend explicitly on the dimension @xmath1 , the metric ( [ metric ] ) does depend on it , and therefore the post - newtonian expansions of eqs .
( [ pifi ] ) involve many @xmath1-dependent coefficients .
we give their full expressions in appendix [ expfieldeq ] , eqs .
( [ exppi])-([expfi ] ) , but we quote below only their newtonian orders and the very few terms which will contribute to the poles @xmath251 in our dimensionally regularized calculations : @xmath252 the acceleration @xmath253 can thus be written as @xmath254 + \frac{4}{c^4}\biggl[\partial_i\hat x + 2\frac{d\hat r_i}{dt}+\cdots\biggr ] + \frac{16}{c^6}\biggl[\partial_i\hat t + \frac{d\hat y_i}{dt}+\cdots\biggr ] + { \mathcal{o}}\left(\frac{1}{c^{8}}\right).\nonumber\end{aligned}\ ] ] in section [ iteratedeinstein ] we shall give flesh to the formal expressions written above by explaining by which algorithm one can compute , with the required accuracy , the explicit @xmath1-dimensional expansions near @xmath255 [ analogous to the simple case ( [ u ] ) ] of the various elementary potentials entering eq . ( [ smallaccel ] ) , and notably of the crucial ones @xmath161 , @xmath160 ( to be computed with 1pn accuracy ) and @xmath164 , @xmath163 ( to be computed at newtonian order only ) .
the main aim of this section is to complete the _ first step _ of the strategy outlined in the introduction , _
i.e. _ to collect a complete list of the additional contributions to the equations of motion which are specific consequences of the use of the _ extended _ hadamard regularization methods defined in @xcite . however , to do that we need to start by recalling some material concerning the hadamard regularization in 3 dimensions , and by contrasting it with dimensional regularization . such material is needed for understanding our computation based on the `` difference '' in section [ difference ] .
we shall start by recalling the definition of the `` ordinary '' hadamard regularization and complete it by defining what we shall call the `` pure '' hadamard regularization .
then we shall recall the main new features of the _ extended _ hadamard regularization defined in @xcite , and collect the additional contributions to the equations of motion which are specific consequences of the use of the extended hadamard regularization ( there are seven such additional contributions ) .
the phrase `` hadamard regularization '' covers two distinct concepts : ( i ) the regularization of the `` limit '' @xmath256 where @xmath257 belongs to a class @xmath258 of singular functions ( generated by the iteration of einstein s equations ) , and ( ii ) the regularization of the 3-dimensional integral @xmath259 of some function @xmath260 .
the class of functions @xmath261 consists of all functions @xmath262 on @xmath263 that are smooth except at @xmath75 and @xmath76 , around which they admit laurent - type expansions in powers of @xmath37 or @xmath38 ( see section ii.a of @xcite for the precise definition of @xmath261 ) . when @xmath264 ( _ i.e. _ , around singularity 1 ) we have , @xmath265 , @xmath266 where the landau @xmath267-symbol takes its usual meaning , and the @xmath268 s denote the coefficients of the various powers of @xmath37 , which are functions of the positions and velocities of the particles , and of the unit direction @xmath269 of approach to singularity 1 .
the powers of @xmath37 are relative integers , @xmath270 , bounded from below by some typically negative @xmath271 depending on the @xmath272 in question .
the hadamard `` _ _ partie finie _ _ '' of the singular function @xmath272 at the location of the singular point 1 ( first meaning of hadamard regularization ) is defined as the angular average of the zeroth - order coefficient in the expansion ( [ fx ] ) .
it is denoted @xmath273 , so that @xmath274 where @xmath275 denotes the usual surface element on the 2-dimensional sphere centered on 1 .
we shall employ systematically a bracket notation @xmath276 for the angular average of a function of the angles ( either @xmath277 or @xmath278).-dimensional ( as we shall see later there can be no confusion about this ) . ]
a distinctive feature of the hadamard partie finie ( [ f1 ] ) is its `` non - distributivity '' in the sense that @xmath279 the non - distributivity represents a crucial departure away from the simple algebraic properties of the analog of @xmath273 in dim .
. which is merely @xmath280 .
it is an interesting fact that in a post - newtonian expansion the non - distributivity starts playing a role only at the 3pn order ( because the functions there become singular enough ) . up to the 2pn order
one can show that @xmath281 for all the functions involved in the equations of motion in harmonic coordinates @xcite .
several of the problems of the hadamard self - field regularization ( in the `` ordinary '' sense ) when applied at the 3pn level ( _ e.g. _ the occurrence of the unknown constant @xmath4 ) are related to the latter non - distributivity .
the second notion of hadamard _ partie finie _ ( denoted @xmath282 in the following ) is to give a meaning to the generally divergent integral @xmath283 . in this work
we shall have to consider only the ultra - violet ( uv ) divergencies of the integrals , _
i.e. _ , at the locations of the two local singularities @xmath75 and @xmath76 .
all functions involved at the 3pn order are such that there are no infra - red ( ir ) divergencies when @xmath284 ( this is true not only in 3 dimensions but also for any dimension @xmath285 in a neighborhood of @xmath42 ) .
the hadamard partie finie of the ( uv ) divergencies is then defined as @xmath286 the description of this formula in words is as follows .
one first excises two _ spherical _ balls @xmath287 and @xmath288 surrounding the two singularities ( each one having the same radius @xmath289 ) , and one computes the integral on the volume external to these balls , _
i.e. _ , @xmath290 first term in eq .
( [ pf ] ) .
that integral tends to infinity when @xmath291 , but we can subtract from it its purely divergent part , which is given by the additional terms in ( [ pf ] ) ( which obviously are to be duplicated when there are 2 singularities , _ cf . _ the symbol @xmath292 ) .
the limit @xmath293 then exists ( by definition ) and defines hadamard s partie finie .
notice the crucial dependence of the partie finie on two constants @xmath77 and @xmath78 entering the log - terms .
these constants have the dimension of length .
we shall say that @xmath77 is the regularization length scale associated to the hadamard regularization of the divergencies near @xmath255 ( similarly for @xmath78 ) .
note also that the hadamard partie finie does not depend ( modulo changing the values of @xmath77 and @xmath78 ) on the _ shape _ of the regularization volumes @xmath294 and @xmath295 , above chosen as simple spherical balls ( see the discussion in ref .
@xcite ) . an important consequence of the definition ( [ pf ] ) is that , in general , the integral of a gradient @xmath296 is not zero , because the surface integrals surrounding the singularities become infinite when the surface areas tend to zero , and may possess a finite part .
we find ( see eq .
( 3.4 ) in @xcite ) @xmath297 for a general @xmath298 the r.h.s . is typically non - zero .
this fact shows that the application of the ordinary hadamard regularization in the post - newtonian iteration has to be supplemented by a notion of distributional derivatives , in order to ensure that the integrals of gradients are zero like in the case of regular functions .
notice that the constants @xmath77 and @xmath78 disappear from the result ( [ pfdiv ] ) .
[ we shall also see the need , within dim .
reg . , to consider some derivatives in the sense of distribution theory . ]
let us apply the definition ( [ pf ] ) to the integral of a compact - support or `` contact '' term , _
i.e. _ , made of the product of some @xmath272 and a dirac delta - function at the point 1 .
let us formally assume that and which takes the property ( [ pfdelta ] ) .
this is exactly what we do in the context of the extended hadamard regularization . ]
@xmath299 which is the most natural way , within hadamard s regularization , to give a sense to such integral .
now the problem with that definition is that if we want to dispose of some _ local _ meaning ( at any field point @xmath241 ) for the product of @xmath272 with the delta - function , then as a consequence of the non - distributivity we can not simply equate @xmath300 with @xmath301 , _
i.e. _ : @xmath302 indeed , if it were true that @xmath303 [ for simplicity we denote @xmath304 , then multiplying by any @xmath132 we would have @xmath305 , and by integrating over @xmath263 following the rule ( [ pfdelta ] ) this would yield @xmath281 in contradiction with the violation of distributivity ( [ nondistr ] ) . therefore , both the violation of distributivity ( [ nondistr ] ) and the consequence ( [ deltanondistr ] ) are unescapable in the ordinary hadamard regularization .
the previous situation should be contrasted with the @xmath1-dimensional case for which the distributivity is always satisfied , as we have simply @xmath306 and @xmath307 finally , taking the @xmath32 limit , we see that the dim .
way of regularizing a three - dimensional `` contact term '' , _
i.e. _ a term like @xmath308 , is by considering it as the @xmath309 limit of its @xmath1-dimensional analogue ( [ contactd ] ) .
that is , @xmath310 \equiv \left ( \lim_{d\rightarrow 3 } f^{(d ) } ( \mathbf{y}_1 ) \right ) \delta^{(3 ) } ( \mathbf{x } - \mathbf{y}_1)\ , , \label{purecontact}\ ] ] where @xmath311 is the @xmath1-dimensional version of @xmath272 , as obtained by solving einstein s equations in @xmath1 dimensions ( using the method explained in section [ iteratedeinstein ] below ) .
there are no poles in the calculation of the `` contact '' terms in any of the potentials at the 3pn order so the limit @xmath309 in eq .
( [ purecontact ] ) always exists .
once again the dim .
prescription ( [ purecontact ] ) owns all the good features one wishes , notably the distributivity as we have emphasized in eqs .
( [ distrd])-([contactd ] ) . in the following
it will be convenient , in order to compare the present dim .- reg .
calculation with the hadamard - based work @xcite , to introduce the terminology _ pure hadamard _
regularization to refer to the following `` minimal '' version of the hadamard regularization : ( a ) an integral @xmath312 , where @xmath272 is made of some product of derivatives of the non - linear potentials @xmath136 , @xmath137 , @xmath313 , is regularized by the ordinary hadamard partie finie prescription ( [ pf ] ) , without bringing in any distributional contributions ( see below for the treatment of these ) ; ( b ) the regularization of a product of potentials @xmath136 ,
@xmath137 , @xmath314 , @xmath313 ( and their gradients ) is assumed to be distributive , which means that the value at the singular point @xmath75 of some polynomial in @xmath136 , @xmath137 , @xmath314 , @xmath313 and their gradients , say @xmath315 $ ] , is given by the replacement rule @xmath316\bigr)_1\longrightarrow { \cal f } [ ( v)_1 , ( v_i)_1 , ( \hat{w}_{ij})_1 , ( \partial_i v)_1 , \cdots]\,;\label{calf1}\ ] ] and ( c ) a contact term , _
i.e. _ of the form @xmath300 , appearing in the calculation of the _ sources _ of the non - linear potentials , is regularized by using the rule @xmath317 \delta^{(3)}(\mathbf{x}-\mathbf{y}_1)\longrightarrow { \cal f } [ ( v)_1 , ( v_i)_1 , ( \hat{w}_{ij})_1 , \cdots]\delta^{(3 ) } ( \mathbf{x}-\mathbf{y}_1)\,,\label{calfdelta}\ ] ] ( there are no gradients of potentials in the contact terms ) .
the rules ( [ calf1])-([calfdelta ] ) of the pure - hadamard regularization are formally equivalent to assuming the replacement rules @xmath318 together with @xmath319 , in the case where @xmath272 and @xmath132 are made of products of our elementary potentials and their gradients . or @xmath320
^ 3(\partial_iv)_1 $ ] , but not , for instance , @xmath321 ^ 2 $ ] . ] the rules of the pure hadamard regularization are , however , well defined , and are not submitted ( by their very definition ) to the consequences of the _ ordinary _ hadamard regularization ( [ nondistr ] ) and ( [ deltanondistr ] ) .
note also that , as done in previous computations of the 3pn adm - hamiltonian @xcite and the 3pn binary s energy flux @xcite , one can formally use ( [ calf1])-([calfdelta ] ) at the price of adding a limited number of arbitrary parameters ( considered as unknown ) .
let us give some reminders of the way we apply the considerations of section [ purehadamard ] to the computation of hadamard - regularized potentials having the form of poisson or poisson - like integrals .
let us first discuss the prescription one has taken in @xmath42 to define the `` value at @xmath335 '' of a ( singular ) poisson potential @xmath336 . in @xmath42 ,
the poisson integral @xmath336 , at some field point @xmath337 , of some singular source function @xmath262 in the class @xmath338 is defined in the sense of the partie - finie integral ( [ pf ] ) , namely @xmath339 where @xmath77 and @xmath78 are the two constants introduced in eq .
( [ pf ] ) . at first sight
we could think that a good choice for defining the pure hadamard value @xmath340_{\mathbf{x } ' = \mathbf{y}_1}$ ] is simply to replace @xmath341 in ( [ px ] ) , _ i.e. _ , @xmath342 however , the work on the 3pn equations of motion @xcite suggested that the definition ( [ py1 ] ) is not acceptable : it did not seem to be able to yield equations of motion compatible with basic physical properties such as energy conservation .
the choice adopted in @xcite is to define the regularized `` value at @xmath335 '' of the function @xmath336 by taking the hadamard partie finie in the singular limit @xmath343 .
notice first that @xmath336 does not belong ( in general ) to the class @xmath261 because the poisson integral will generate some _ logarithms _ of @xmath70 in its expansion when @xmath344 .
thus , we shall have , rather than an expansion of type ( [ fx ] ) , @xmath345+o({r_1'}^n)\ , , \label{expandpx}\ ] ] where the coefficients @xmath346 and @xmath347 depend on the angles @xmath348 , and also on the constants @xmath77 and @xmath78 , in such a way that when combining together the terms in ( [ expandpx ] ) the constant @xmath70 always appears in `` adimensionalized '' form like in @xmath349 .
then we define the hadamard partie finie at point 1 exactly in the same way as in eq .
( [ f1 ] ) , except that we now include a contribution linked to the ( divergent ) logarithm of @xmath70 , which is possibly present into the zeroth - order power of @xmath70 .
more precisely , we define @xmath350 where we introduced a _ new regularization length scale _ denoted @xmath61 , which can be seen as some `` small '' but finite cut - off length scale [ so that @xmath351 in eq .
( [ p1def ] ) is a finite , but `` large '' cut - off dependent contribution ] .
we shall see later that the dependence on @xmath61 disappears ( as it should ) when adding to @xmath352 the difference @xmath353 . to compute the partie finie
one must apply the definition ( [ p1def ] ) to the poisson integral ( [ px ] ) , which involves evaluating correctly the angular integration therein .
the result , proved in theorem 3 of @xcite , is @xmath354\bigl<\mathop{f}_1{}_{-2}\bigr>\ , .
\label{p1}\ ] ] we recover in the first term the value of the potential at the point 1 : @xmath355 , given by eq .
( [ py1 ] ) .
the supplementary term makes the partie finie to differ from the `` nave '' guess @xmath355 in a way which was found to play a significant role in the computations of @xcite .
the apparent dependence of the result ( [ p1 ] ) on the scale @xmath77 is illusory .
the @xmath77-dependence of the r.h.s . of eq .
( [ p1 ] ) cancels between the first and the second term , so the result depends only on the constants @xmath70 and @xmath78 , and we have in fact the following simpler rewriting of ( [ p1 ] ) , @xmath356 similarly the regularization performed at point 2 will depend on @xmath71 and @xmath77 , so that the binary s point - particle dynamics in hadamard s regularization depends on four ( _ a priori _ independent ) length scales @xmath70 , @xmath78 and @xmath71 , @xmath77 .
the explicit expression of the result ( [ p1 ] ) is readily obtained from the definition of the partie - finie integral ( [ pf ] ) .
we find ( see the details in ref .
@xcite ) @xmath357 \bigl<\mathop{f}_1{}_{-2}\bigr>\nonumber\\ & -&\sum_{\ell=0}^{+\infty}\frac{(-)^\ell}{\ell!}\partial_l \left(\frac{1}{r_{12}}\right)\left[\sum_{p+\ell+3<0 } \frac{s^{p+\ell+3}}{p+\ell+3}\bigl <
n_2^l\mathop{f}_2{}_p\bigr > + \ln\left(\frac{s}{s_2}\right)\bigl < n_2^l \mathop{f}_2{}_{-\ell-3}\bigr>\right]\biggr\}.\nonumber\\ \label{p1result}\end{aligned}\ ] ] note that the terms corresponding to singularity 2 involve the multipolar expansion around the point @xmath76 of the factor @xmath358 present into the integrand . ]
employing our usual notation where capital letters denote multi - indices : @xmath359 , and , for instance , @xmath360 .
the expansion is symmetric - trace - free ( stf ) because @xmath361 . here
@xmath362 is a short - hand for @xmath363 partial derivatives @xmath364 of @xmath365 . the multipole expansion in @xmath1 dimensions ( also stf ) is given by eq .
( [ r1ofr2 ] ) below . ] because we work at the level of the equations of motion , many of the terms we shall need in this paper are in the form of the _ gradient _ of a poisson - like potential . for the gradient
we have a formula analogous to ( [ p1 ] ) and given by eq .
( 5.17a ) of @xcite , namely @xmath366 where we have taken into account ( in the rewriting of the second line ) the always correct fact that the constant @xmath77 cancels out and gets `` replaced '' by @xmath70 .
notice that in ( [ dip1 ] ) there is no additional term to the partie finie integral similar to the last term in ( [ p1 ] ) .
the corresponding explicit expression is @xmath367\biggr\}. \nonumber\\ \label{dip1result}\end{aligned}\ ] ] finally we must also treat the more general case of potentials in the form of retarded integrals [ see eqs .
( [ potentialeq ] ) ] , but because we shall have to consider ( in section [ diff ] ) only the _ difference _ between the dimensional and hadamard regularizations , it will turn out that in fact the first - order retardation ( 1pn relative order ) is sufficient for this purpose .
actually , in this paper we are not interested in radiation - reaction effects , so we shall use the symmetric ( half - retarded plus half - advanced ) integral . at the 1pn order we thus have to evaluate @xmath368 where @xmath336 is given by ( [ px ] ) , and where @xmath369 denotes ( two times ) the double or `` twice - iterated '' poisson integral of the second - time derivative , still endowed with a prescription of taking the hadamard partie finie , namely @xmath370 in the case of @xmath369 the results concerning the partie finie at point 1 were given by eqs .
( 5.16 ) and ( 5.17b ) of @xcite , [ q1]@xmath371 where the @xmath372 s denote the analogues of the coefficients @xmath373 , parametrizing the expansion of @xmath272 when @xmath374 , but corresponding to the double time - derivative @xmath375 instead of @xmath272 . [ in the following we shall not need the explicit forms of the results ( [ q1 ] ) . ] let us clarify an important point concerning the treatment of the repeated time derivative @xmath376 in eqs .
( [ q1 ] ) .
as we are talking here about hadamard - regularized integrals ( which excise small balls around both @xmath75 and @xmath76 ) , the value of @xmath377 can be simply taken in the sense of ordinary functions , _
i.e. _ , without including eventual `` distributional '' contributions proportional to @xmath378 or @xmath379 and their derivatives . however , we know that such terms are necessary for the consistency of the calculation . this is why we must also include somewhere in our formalism the _ difference _ between the evaluation of these distributional terms in @xmath1 dimensions , and the specific distributional contributions issued from the generalized framework used in @xcite .
this difference will be included in section [ additional ] below , among the complete list of additional contributions specifically related to the use of the extended regularization approach we shall now describe .
the `` _ _ extended - hadamard _ _ '' regularization , proposed in refs .
@xcite , tackles the particular properties of the ordinary hadamard regularization , notably the non - distributivity of eqs .
( [ nondistr ] ) and ( [ deltanondistr ] ) , and the fact that the integral of a gradient is not zero [ eq .
( [ pfdiv ] ) ] .
these properties are implemented within a theory of pseudo - functions , _ viz _ linear forms defined on the set of singular functions @xmath261 .
the use of pseudo - functions in this context enables one to give a precise meaning to the object @xmath380 needed in the computation of the contact terms , and which is otherwise ill - defined in distribution theory . furthermore the use of some generalized versions of distributional derivatives permits a systematic treatment of integrals and a natural implementation of the property that the integral of a gradient is always zero . in this paper
we shall content ourselves with recalling the principle of the extended - hadamard regularization , and with presenting its `` ready - to - use '' consequences . to any @xmath298 we associate the `` partie finie '' pseudo - function @xmath381 , which is the linear form on @xmath261 defined by the duality bracket @xmath382 which means that the action of @xmath381 on any @xmath383 is the partie - finie integral , as given by ( [ pf ] ) , of the ordinary product . the pseudo - function @xmath381 reduces to a distribution in the ordinary sense of schwartz @xcite when restricted to the usual set @xmath384 of smooth functions with compact support on @xmath263 .
the product of pseudo - functions coincides , by definition , with the ordinary point - wise product , namely @xmath385 . in the class of pseudo - functions constructed in ref .
@xcite , the `` dirac - delta '' pseudo - function @xmath386 is defined by @xmath387 where @xmath273 denotes hadamard s partie finie ( [ f1 ] ) .
this definition , which obviously yields a natural extension of the dirac function @xmath388 in the context of hadamard s regularization , leads also to new objects which have no equivalent in distribution theory , the most important one being the pseudo - function @xmath389 which played a crucial role in @xcite for the calculation of the compact - support parts of potentials as well as the purely distributional parts of derivatives .
it is given by @xmath390 where one should be reminded that it is in general not allowed to replace the r.h.s . by the product of regularizations : @xmath391 .
in the actual computation @xcite the pseudo - function @xmath389 acts always on smooth functions with compact support ( @xmath392 ) , in which case it reduces to a distribution in the ordinary sense , which was shown to admit the `` intrinsic '' form @xmath393 here @xmath394 denotes a multi - index composed of @xmath363 multipolar indices @xmath395 , @xmath396 means a product of @xmath363 partial derivatives @xmath397 , and @xmath398 a product of @xmath363 unit vectors ( we do not write the @xmath363 summation symbols , from 1 to 3 , over the indices composing @xmath399 ) .
notice that the sum in eq .
( [ pffdelta ] ) is finite because @xmath272 admits some maximal order of divergency when @xmath400 .
now we discover that the `` monopole '' term in the latter multipolar sum , having @xmath401 , is nothing but @xmath402 which is exactly the result we would get following the pure - hadamard regularization rule ( [ calfdelta ] ) . [ indeed , as we are considering here only the contact terms entering the source terms for the 3pn - level nonlinear potentials , the `` ordinary '' hadamard regularization @xmath273 coincides with the `` pure '' hadamard regularization ( [ calfdelta ] ) . ]
the sum of the other terms then define what we can call some non - distributive contributions because their appearance is the direct consequence of the violation of distributivity ( [ nondistr ] ) .
thus , @xmath403 in the work @xcite care has been taken of all such non - distributivity terms .
consider for instance the poisson integral of a compact - support term @xmath389 ( say , proportional to the matter source densities @xmath143 , @xmath144 or @xmath149 ) . using ( [ pffdelta ] ) the poisson integral reads ) we assume that @xmath74 is distinct from the 2 singularities @xmath75 and @xmath76 ; see @xcite for more details . ] @xmath404 [ where @xmath405 should be better written @xmath406 .
evaluating now the partie finie ( [ f1 ] ) at both singular points [ _ i.e. _ , when @xmath374 and @xmath407 we obtain [ poissonfdeltaab]@xmath408 the result ( [ poissonfdelta1 ] ) is in agreement with the pure - hadamard regularization ; however eq .
( [ poissonfdelta2 ] ) does involve some extra terms with respect to the pure - hadamard calculation , since the latter is easily seen to simply yield @xmath409 , which is nothing but the `` monopolar '' term @xmath401 of the multipolar sum in the r.h.s . of ( [ poissonfdelta2 ] ) .
therefore we decompose ( [ poissonfdelta2 ] ) as @xmath410 the second ingredient of the extended - hadamard regularization concerns the treatment of partial derivatives in some ( extended ) distributional sense .
essentially , one requires @xcite that the derivative reduces to the ordinary derivative in the case of regular functions , and is such that one can integrate by parts any integrals .
the latter property ( valid for the spatial derivative ) translates into @xmath411 this rule contains the standard definition of the distributional derivative @xcite as a particular case .
it implies the important property that the integral of a divergence is zero .
let us pose @xmath412\ , , \label{derivedistr}\ ] ] where @xmath413 denotes the derivative of @xmath272 viewed as an `` ordinary '' pseudo - function , and @xmath414 $ ] represents the purely distributional part of the spatial derivative ( with support concentrated on @xmath75 or @xmath76 ) .
looking for explicit solutions of the basic relation ( [ intparts ] ) we have found @xcite , with the help of eq .
( [ pfdiv ] ) : @xmath415 = 4\pi \ , \text{pf } \biggl ( n_1^i \biggl [ \frac{1}{2 } \ , r_1 \ , \mathop{f}_1{}_{-1}+\sum_{k\geq 0 } \frac{1}{r_1^k } \ , \mathop{f}_1{}_{-2-k}\biggl ] \delta_1 \biggr ) + 1 \leftrightarrow 2 \ , .
\label{distrpart}\ ] ] notice that @xmath414 $ ] depends only on the _ singular _ coefficients of @xmath272 ( coefficients of negative powers of @xmath37 in the expansion of @xmath272 ) . the derivative operator defined by eqs .
( [ derivedistr])-([distrpart ] ) does not represent the unique solution of ( [ intparts ] ) , but it has been checked during the calculation @xcite that using another possible solution results in _ physically equivalent _ equations of motion at 3pn order ( _ i.e. _ , reducing to each other by a gauge transformation ) .
concerning multiple derivatives we dispose of the general formula @xmath416=\sum_{k=1}^\ell\partial_{i_1\dots i_{k-1 } } \text{d}_{i_{k}}[\partial_{i_{k+1}\dots i_\ell}f]\ , , \label{multdistr}\ ] ] giving the distributional term associated with the @xmath363-th spatial derivative , @xmath417\equiv\partial_l\text{pf}f -\text{pf}\partial_l f$ ] ( where @xmath418 ) , in terms of the single derivative @xmath414 $ ] .
as an example , to treat the second - derivative of the newtonian potential , @xmath419 where @xmath420 , one uses @xmath421=-\frac{4\pi}{3}\ , \text{pf}\left(\delta^{ij}+\frac{15}{2}\hat{n}_1^{ij}\right)\delta_1 \ , \label{ex}\ ] ] where @xmath422 . therefore the extended distributional derivative differs in general from the usual schwartz s derivative [ _ cf .
_ the second term in ( [ ex ] ) ] .
[ this is unavoidable if one wants to respect the basic rule of integration by parts ( [ intparts ] ) for general functions in the class @xmath261 . ]
notice also that we do find a distributional term in the case of the first derivative : @xmath423=2\pi\ , \text{pf}\left(r_1 n_1^i \delta_1\right)$ ] .
we recall also ( for future use ) the case of the partial time - derivative , @xmath424 $ ] , whose distributional term is given by ( following ref .
@xcite ) @xmath425=v_1^i\mathop{\text{d}}_1{}_i[f ] + v_2^i\mathop{\text{d}}_2{}_i[f]\ , , \label{dtf}\ ] ] in terms of the partial derivatives with respect to the _ source _ points @xmath75 and @xmath76 , namely @xmath426 $ ] and @xmath427 $ ] .
the explicit expression reads @xmath425
= -4\pi \ , \text{pf } \biggl ( ( n_1v_1)\biggl [ \frac{1}{2 } \ , r_1 \ , \mathop{f}_1{}_{-1}+\sum_{k\geq 0 } \frac{1}{r_1^k } \ , \mathop{f}_1{}_{-2-k}\biggl ] \delta_1 \biggr ) + 1 \leftrightarrow 2\ , , \label{distrtpart}\ ] ] where @xmath428 denotes the ordinary scalar product [ notice the over - all sign difference with respect to eq . ( [ distrpart ] ) ] .
multiple time - derivatives can be treated accordingly to eq .
( [ multdistr ] ) .
for instance , @xmath429 = \text{d}_t[\partial_tf]+\partial_t\text{d}_t[f]\ , .
\label{dtt}\ ] ] following the regularization @xcite all the distributional terms [ of type @xmath389 and @xmath430 issued from the latter distributional derivatives are to be treated when computing the potentials according to the extended contact term definitions of eqs .
( [ poissonfdelta1])-([poissonfdelta2 ] ) .
finally let us turn to the extension of the hadamard regularization ( introduced in @xcite ) concerning the definition of a new operation of regularization , denoted @xmath46_1 $ ] , consisting of performing the hadamard regularization @xmath273 within the spatial hypersurface that is geometrically orthogonal ( in a minkowskian sense ) to the four - velocity of the particle 1 .
the regularization @xmath46_1 $ ] differs from @xmath273 by a series of relativistic corrections calculated in @xcite .
together with the other improvements of the extended - hadamard regularization , it resulted in equations of motion in harmonic - coordinates which are manifestly lorentz invariant at the 3pn order @xcite . here
we give a formula , sufficient for the present purpose , for expressing @xmath46_1 $ ] in terms of the basic regularization @xmath273 , defined by ( [ f1 ] ) , at the 1pn order : @xmath431_1=\biggl(f+\frac{1}{c^2}(\mathbf{r}_1.\mathbf{v}_1 ) \bigl[\partial_tf+\frac{1}{2}v^i_1\partial_i f\bigr]\biggr)_1+\mathcal{o}\left(\frac{1}{c^4}\right ) .
\label{lorentzian}\ ] ] the first term is simply @xmath273 , while the other terms define a set of relativistic corrections required for ensuring the lorentz invariance of the final equations in hadamard s regularization .
hence , we decompose ( [ lorentzian ] ) into @xmath431_1=(f)_1~+~\text{``lorentz '' contributions}\ , . \label{lorentzian'}\ ] ] after the reminders of the last subsections , we are now in position to explain the origin of all the contributions [ included in the final result ( [ a1bf ] ) ] which were due to the specific use of the extended - hadamard regularization .
actually , we shall list here the contributions due to the use of the full prescriptions of @xcite with respect to those that would follow from using what we shall call a `` _ _ pure hadamard - schwartz _ _ '' ( phs ) regularization . by this
we mean : ( 1 ) treating the contact terms of all the non - linear potentials @xmath136 , @xmath137 , @xmath314 , @xmath313 as in ( [ calfdelta ] ) [ we have checked that for all the potentials involved this is equivalent to ( [ purecontact ] ) ] ; ( 2 ) treating the distributional part of an integrand such as @xmath432 in the normal schwartz s distributional way , for instance ) . ]
@xmath433 and evaluating the contact term generated by the delta function in the `` pure hadamard '' way ( [ calfdelta ] ) ; ( 3 ) regularize any three - dimensional integral by the ordinary hadamard prescription ( [ pf ] ) ; and , finally , ( 4 ) using systematically , in the last stage of the calculation where one replaces the metric into the geodesic equations , the pure hadamard replacement rule appearing in ( [ calf1 ] ) [ for instance , we write @xmath320 ^ 3(\partial_iv)_1 $ ] , creating therefore a net difference with respect to the ordinary and/or extended hadamard regularizations for which @xmath434 ^ 3(\partial_iv)_1 $ ] ] .
the usefulness of the definition of such a phs regularization is to `` localize '' the additional contributions brought by dim .
reg . to the occurrence of poles @xmath93 ( or `` cancelled '' poles ) in @xmath1 dimensions .
our complete list of additional contributions contains seven items .
first of all there are four `` non - distributivity '' contributions of the type given by eq .
( [ poissonfdelta2nondistr ] ) : 1 .
the so - called `` self '' terms , for which the delta - function in @xmath389 comes from the purely distributional part of the distributional derivative given by ( [ distrpart ] ) .
the self terms were derived in eq .
( 6.20 ) in @xcite ; they read explicitly @xmath435 , \label{deltaself0}\ ] ] where @xmath436 denotes the usual scalar product between @xmath67 and @xmath69 , and where @xmath437 . the expression ( [ deltaself0 ] ) can be rewritten in a simpler way as ( where we denote for simplicity @xmath438 ) @xmath439 2 .
the so - called `` leibniz '' terms , which are additional contributions due to the extended distributional derivative , taking into account the violation of the leibniz rule when performing some simplifications of the non - linear potentials at the 3pn order ( see the explanations in section iii b in @xcite ) .
the leibniz terms were written in eq .
( 6.19 ) in @xcite , and read @xmath440 we emphasize that the contributions ( [ deltaself ] ) and ( [ deltaleibniz ] ) represent some additive effects of the use of the distributional derivative introduced in ref .
@xcite , when compared to the effect of the schwartz derivative in the phs regularization .
note that both eqs .
( [ deltaself ] ) and ( [ deltaleibniz ] ) depend on the choice of distributional derivative , and we have given them here in the case of the `` particular '' derivative defined by ( [ distrpart ] ) .
3 . a special non - distributivity in the compact - support potential @xmath136 when it is computed at the 3pn order . in this case
the @xmath389 comes simply from the compact - support point - particle source @xmath143 of the potential .
( 4.17 ) of @xcite gives for that term @xmath441 4 .
a contribution coming from the compact - support part of the potential @xmath442 parametrizing the metric at the 3pn order , and derived at the end of section iv a in @xcite : @xmath443 besides the non - distributivity of the type ( [ poissonfdelta2nondistr ] ) , we have also the more `` direct '' non - distributivity due to the fact that the pure hadamard prescription for the regularization of the value of an expression `` at @xmath75 '' , eq .
( [ calf1 ] ) , differs from the ordinary and/or extended hadamard ones [ see for instance eq .
( [ u4had ] ) ] .
it plays a role only in the last stage of the computation of the 3pn equations of motion , once we substitute all the potentials computed at the right pn order into the geodesic equations .
we thus have 1 . a `` direct '' non - distributivity contribution , which can be called non - distributivity in the equations of motion ( eom ) , and given by eq . ( 6.34 ) in @xcite , @xmath444n_{12}^i-\frac{779}{420}\frac{g^3m_1 ^ 2m_2}{c^6}\ , v_{12}^{jk}\,\partial_{ijk}\left(\frac{1}{r_{12}}\right ) , \label{deltaeom}\ ] ] where @xmath445 and @xmath446 .
this term involves some combinations of masses different from those in eqs .
( [ deltaself])-([deltat ] ) .
note that because @xmath391 the non - distributivity in the eom depends on which prescription has been chosen for the stress - energy tensor of point - particles .
( [ deltaeom ] ) corresponds to the particular prescription advocated in section v of @xcite .
however it was checked in @xcite that different prescriptions yield physically equivalent equations of motion .
the next correction brought about by the extended - hadamard regularization is the one due to the regularization @xmath46_1 $ ] , performed in the lorentzian rest frame of the particle . in practice the effect of such `` lorentzian '' regularization boils down to applying eq .
( [ lorentzian ] ) .
it turned out that the only new contribution of this type came from the regularization of the potential @xmath447 at the 1pn order [ and also when deriving the result for @xmath448 , which is galilean - invariant , in eq .
( [ deltaeom ] ) ] , leading to 1 . the so - called `` lorentz '' contribution to the acceleration , given by eq .
( 5.35 ) in @xcite as @xmath449 \partial_{ijk}\left(\frac{1}{r_{12}}\right ) .
\label{deltalorentz}\ ] ] this term has been crucial for ensuring the lorentz invariance of the final 3pn equations of motion in @xcite . finally , one must also take care of one additional contribution ( with respect to the `` phs '' definitions ) due to the non - schwartzian way of treating distributional derivatives .
we have already mentioned two contributions coming from this origin : ( i ) and ( ii ) above .
actually , there is a third one with the same origin and which comes from our computation ( see section [ difference ] below ) of the `` difference '' between the dimensional and hadamard regularizations of retarded potentials , namely the crucial potentials @xmath161 and @xmath450 which must both be expanded to 1pn fractional accuracy .
more precisely , this contribution is due to the repeated time - derivative operator @xmath451 coming when expanding the time - symmetric green function of the dalembertian as @xmath452
. we shall explicitly exhibit in eqs .
( [ drdirresult ] ) below the way these derivatives enter our calculation of the difference . for technical reasons
the time - derivative @xmath451 must be kept _ inside _ the integrals , so it has to be considered in a distributional sense , and we have therefore to take into account the different ways of treating the distributional derivatives in both regularizations . in the extended - hadamard regularization
the distributional terms are given by @xmath453 $ ] which is shown in eqs .
( [ dtf])-([dtt ] ) , and when they enter the source of some poisson - type integral they are evaluated according to eqs .
( [ poissonfdeltaab ] ) . on the other hand , in
one uses the ordinary schwartz derivative ( in @xmath1 dimensions ) which is described in section [ distrderiv ] . in this case
the double time - derivative @xmath451 is computed with the help of the gelfand - shilov formulae ( [ gelfand])-([gelfandt ] ) below . when examining the difference between the contact terms in @xmath453 $ ] and those issued from @xmath454 , we find that only the source for the 1pn potential @xmath161 ( or rather for the combination @xmath455 which enters the equations of motion ) contributes . this gives : 1 .
the following `` time - derivative '' contribution to the acceleration , @xmath456 this term was part of the final result of @xcite .
however it is not mentioned in @xcite because this reference never tried to compare the results of the extended distributional derivative with those given by the ordinary schwartz derivative in 3 dimensions , except in those cases , _ items _ ( i ) and ( ii ) above , for which the schwartz derivative yielded in fact some ill - defined ( formally infinite ) expressions in 3 dimensions .
[ the latter expressions turn out to be rigorously zero when computed in dimensional regularization . ] in summary , there are in all _ seven _ different terms , ( i)(vii ) , which are specifically due to the extended version of the hadamard regularization .
the `` pure - hadamard - schwartz '' equations of motion are then obtained from the end result of @xcite , _
i.e. _ , @xmath60 given by eq .
( 7.16 ) of @xcite , by subtracting these terms .
therefore we define ( see also section [ purehadstat ] below ) @xmath457 and _ idem _ with @xmath292 for the other particle .
in this section we come to the core of our technique for evaluating the difference between the @xmath1-dimensional equations of motion and their pure - hadamard - schwartz expressions , defined above and given in practice by eq .
( [ defph ] ) .
let us start by indicating how we solved ( with sufficient accuracy ) einstein s field equations in @xmath1 dimensions .
one writes the post - minkowskian expansion of einstein s equations in the guise of explicit formulae for the elementary potentials @xmath458 , as given in section [ fieldeq ] .
note that it is crucial to take into account the explicit @xmath1-dependence of the coefficients entering these equations .
the first step of the formalism is to get sufficiently accurate explicit expressions for the basic linear potentials @xmath136 and @xmath137 . as we do not need to consider here radiation reaction effects ( which do not mix with the uv divergencies arising at the 3pn level ) it is enough to solve eqs .
( [ vlowest ] ) by means of the pn expansion of the time - symmetric green function .
for instance , we have @xmath459 from eq .
( [ tmunu ] ) we see that the source @xmath143 reads @xmath460 + 1 \leftrightarrow 2\,,\ ] ] where @xmath461 note the presence of many `` contact '' evaluations of field quantities in @xmath143 .
such terms are unambiguously defined in dimensional regularization .
they are computed by successive iterations ( _ e.g. _ to get @xmath462 to 1pn fractional accuracy we need to have already computed @xmath463 to order @xmath464 included ) .
those evaluations do not give rise to pole terms in @xmath143 , up to the 3pn accuracy .
hence , as we said above , we can consider that their @xmath32 limits define a certain ( three - dimensional ) way of estimating contact terms , that we have checked to be in full agreement with the `` pure - hadamard '' prescription defined in the previous section .
coming now to the spatial dependence of the scalar potential @xmath136 we get from eq .
( [ gsigma ] ) @xmath465 + \cdots + 1 \leftrightarrow 2\ , , \label{vexpand}\ ] ] where we introduced the following elementary solutions @xmath466 , @xmath467 , etc .
, whose explicit forms are [ u1v1]@xmath468 where @xmath469 is related to the usual eulerian @xmath470-function by adopted here is related by @xmath471 to the constant @xmath472 chosen in @xcite .
our present choice is motivated by the easy - to - remember fact that @xmath473 . ]
@xmath474 inserting the explicit expression ( [ vexpand ] ) of @xmath136 into , say , the non - linear terms in the r.h.s . of eq .
( [ dalwij ] ) , yields a dalembert equation for the non - linear potential @xmath159 with a `` source function '' which is the sum of some contact terms @xmath475 and of an extended non - linear source @xmath476 which belongs to the @xmath1-dimensional analogue of the class @xmath258 , say @xmath477 .
more precisely , at each stage of the iteration we find inhomogeneous wave equations of the type @xmath478 where the extended source function @xmath476 is regular everywhere except at the points @xmath75 and @xmath76 , in the vicinity of which it admits an expansion of the general form ( @xmath479 ) @xmath480 where @xmath481 and @xmath482 are relative integers ( @xmath483 ) , whose values are limited by some @xmath271 , @xmath484 and @xmath485 as indicated .
the expansion ( [ fd ] ) differs from the corresponding expansion in 3 dimensions , as given in ( [ fx ] ) , by the appearance of integer powers of @xmath486 where @xmath35 .
the coefficients @xmath487 depend on the unit vector @xmath277 in @xmath1 dimensions , on the positions and coordinate velocities of the particles , and also on the characteristic length scale @xmath90 of dimensional regularization . because @xmath488 when @xmath309 we necessarily have the constraint ( @xmath489 ) @xmath490 the iteration continues by inverting the wave operator by means of the time - symmetric expansion ( [ gsigma ] ) .
the basic terms of this expansion which will turn out to be crucial for our 3pn calculation based on the _ difference _ are in fact the first two terms .
focussing on the terms generated by the extended source @xmath476 ( rather than the simpler contact terms ) we can write the @xmath1-dimensional analogue of ( [ rpq ] ) as @xmath491 = p^{(d)}(\mathbf{x } ' ) + \frac{1}{2c^2 } q^{(d ) } ( \mathbf{x } ' ) + \mathcal{o}\left(\frac{1}{c^4}\right ) , \label{rd}\ ] ] where the @xmath1-dimensional poisson integral of @xmath311 reads @xmath492\equiv -\frac{\tilde{k}}{4\pi } \int\frac{d^d\mathbf{x}}{\vert\mathbf{x}-\mathbf{x}'\vert^{d-2 } } f^{(d)}(\mathbf{x})\ , .
\label{pdx}\ ] ] we have used the fact , already mentioned above , that the @xmath1-dimensional elementary solution of the laplacian reads @xmath493 ( see appendix [ formulae ] for a proof of eq .
( [ green ] ) and for other useful formulae valid in @xmath1 dim . ) , while the 1pn term is given by @xmath494= -\frac{\tilde{k}}{4\pi(4-d ) } \int d^d\mathbf{x}\,\vert\mathbf{x}-\mathbf{x}'\vert^{4-d } \partial_t^2f^{(d)}(\mathbf{x})\ , .
\label{qdx}\ ] ] note the important point that in @xmath1 dimensions , as in 3 dimensions , the time - derivative operator @xmath451 present in the integrand of ( [ qdx ] ) is to be considered in the sense of distributions ( see further discussion in section [ distrderiv ] below ) .
an important technical aspect of the @xmath1-dimensional pn iteration of the elementary potentials @xmath495 is the existence of the generalization ( [ green ] ) of the usual green s function for the laplace equation , as well as of its higher pn analogues @xmath496 , allowing one to explicitly compute the spatial dependence of the _ linear _ potentials @xmath136 , @xmath137 and @xmath158 for instance .
however , starting with @xmath314 we need to poisson - integrate _ non - linear _ sources , such as @xmath497 . in three dimensions ,
these non - linear contributions are reducible to the knowledge of the basic non - linear potential @xmath498 , such that @xmath499 .
we have succeeded in explicitly computing the @xmath1-dimensional analogue of the @xmath498 potential , namely @xmath500 our result is reported in appendix [ littleg ] . as indicated there , if we wished to explicitly compute some of the higher pn potentials needed to write the closed form of the non - linear sources relevant to the 3pn equations of motion , we should extend the calculation of the potential @xmath118 to the potentials @xmath501 and @xmath502 of appendix [ littleg ] .
luckily , it is not needed to use a closed - form expression for any of the non - linear potentials . indeed , similarly to what was used long ago @xcite when discussing the iteration generated by riesz - type sources , eq .
( [ triesz ] ) , one can control the uv _ singular _ part of @xmath503 from the knowledge of the uv singular part of its non - linear source @xmath262 .
for any small enough value of @xmath86 .
] more precisely , in the vicinity say of @xmath75 , at each iteration stage we can decompose the source in @xmath504 where the _ singular _
part @xmath505 ( with respect to @xmath75 ) is a sum of terms of the form eq .
( [ fd ] ) , which are not ( in the limit @xmath32 ) smooth functions of @xmath506 , and where the _ regular _ part @xmath507 is a smooth @xmath508 function of @xmath506 .
[ the simplest example of this decomposition is eq .
( [ u ] ) with , near point @xmath75 , @xmath509 and @xmath510 . ]
if , for concreteness , we then consider @xmath511 , the above decomposition entails a corresponding decomposition of @xmath512 , and it is easy to see that @xmath513 . from this result , we can uniquely determine @xmath514 from @xmath515 using , _
e.g. _ , the formula ( [ intform ] ) in appendix [ formulae ]
. this local procedure does not allow one to compute the regular part of the poisson potential in @xmath1 dimensions .
fortunately , thanks to particular simplifications that occur in the structure of einstein s field equations , the knowledge of @xmath516 in @xmath1 dimensions for the complicated non - linear sources is not needed .
indeed , one can see on our explicit solution of einstein s field equations at 3pn given in section [ fieldeq ] that there are no `` quartically non - linear '' source terms of the form , say , @xmath517 or @xmath518 for @xmath169 at the 3pn order ( see fig .
[ fig5 ] below ) . as explained in section [ dimregstat ] below , a nice way to understand the origin of the poles @xmath519 appearing in the 3pn equations of motion is to use a diagrammatic representation .
a pole can arise in @xmath80 only when three propagator lines ( including the extra one coming from @xmath520 when solving @xmath521 non - linear source ) can all shrink towards the first world - line . if terms of the type above ( _ e.g. _ @xmath522 ) were present in the source one could have a diagram where the three shrinking propagators come from @xmath520 , @xmath523 and @xmath524 . then @xmath525 $ ]
would remain as an external attachment to this diagram ( and would then fork into two `` feet '' on the second world - line ) . in view of the pole @xmath526 ( with @xmath35 ) arising from the triplet of shrinking propagators , one would need to know @xmath527 $ ] up to @xmath36 accuracy , _
i.e. _ , @xmath527 = \text{reg}^{(3 ) } [ \hat{w}_{ij } ( \mathbf{x } ) ] + \varepsilon \
, \hat{w}'_{ij } ( \mathbf{x } ) + { \cal o } ( \varepsilon^2)$ ] [ in which @xmath528 is defined by this expansion ] . if such a term had been present we would have needed to use the full @xmath1-dimensional , globally determined @xmath498-potential given in appendix [ littleg ] to determine @xmath529 , which would have entered the final , renormalized equations of motion .
however , because all such terms are absent at the 3pn order , the only external attachments to the dangerous shrinking diagrams are simple lines , such for instance as the lines ending on @xmath530 in figs .
[ fig2]d , [ fig3]b or [ fig4]b presented below .
such lines do need to be evaluated to accuracy @xmath36 , but this is easy because they represent linear potentials such as @xmath136 or @xmath137 which are known in dimension @xmath1 _ via _ eq .
( [ vexpand ] ) . in conclusion ,
the algorithm we use to solve , with sufficient accuracy , einstein s equations in @xmath1 dimensions consists of : ( 1 ) starting from the fully @xmath1-dimensional expressions for the linear potentials @xmath136 , @xmath137 ( and more generally for the parts of the non - linear potentials with delta - function sources ) ; ( 2 ) determining the local expansions , near @xmath40 , of the singular parts of the non - linear potentials by inverting @xmath531 _ via _ formulae ( [ intform ] ) of appendix [ formulae ] ; ( 3 ) completing @xmath532 by adding to @xmath533 the limit when @xmath534 of @xmath535 , namely @xmath536 which is known from the previous work on the 3pn equations of motion in 3-dimensions @xcite .
note that we denote by @xmath536 a formal @xmath1-dimensional function , @xmath537 , the explicit expression of which in terms of @xmath37 , @xmath277 , etc .
coincides with its 3-dimensional counterpart .
for instance @xmath538 denotes the usual regular part of @xmath539 , obtained by subtracting from @xmath540 the two three - dimensional locally singular expansions of @xmath541 around @xmath75 and @xmath76 as given by the @xmath32 limit of eq .
( [ eqd9 ] ) and its @xmath542 analog . after this double subtraction ,
@xmath538 is considered as a function in @xmath543 , and we can use as approximation to @xmath544 the explicit expression @xmath545 . more generally , in our calculations we use as approximation to @xmath546 [ which symbolizes here the non - linear potentials @xmath547 and @xmath548 at newtonian order , and @xmath314 at the 1pn order ] the expression @xmath549 .
evidently , the subtraction of the singular part needs to be performed only up to some finite order in @xmath550 and @xmath551 .
we have checked the choice we made of @xmath7 in each calculation by doing two separate calculations for the values @xmath7 and @xmath552 , and checking that the corresponding final results are the same .
we performed also direct checks of the independence of the final results on the precise @xmath1-dimensional extensions of the `` regular '' part of the non - linear potentials , such as @xmath553 [ in which @xmath554 is defined by this expansion ] .
we systematically added in all our non - linear potentials @xmath547 , @xmath313 , @xmath314 some smooth contributions to @xmath536 vanishing with @xmath36 , _
i.e. _ , some substitutes for the actual @xmath554 .
these `` substitutes '' were determined in such a way that ( i ) they are homogeneous solutions of the dalembertian equation at the required post - newtonian order , ( ii ) the differential identities obeyed by the potentials in @xmath1 dimensions , eqs .
( [ divi])-([djwij ] ) , are indeed satisfied up to the order @xmath36 , and with the required precision @xmath7 in powers ) has been done only in the vicinity of the two particles . ] of @xmath37 or @xmath38 . and we checked that our final results are totally insensitive to the introduction of such substitutes for the function @xmath554 . finally , when evaluating the equations of motion , as given by eq .
( [ smallaccel ] ) , we must evaluate the value at @xmath555 of many terms given either by poisson integrals of the form ( [ pdx ] ) or their 1pn generalizations ( [ qdx ] ) .
this is quite easy to do in dim .
reg . , because the nice properties of analytic continuation allow simply to get @xmath556_{\mathbf{x } ' = \mathbf{y}_1}$ ] ( say ) by replacing @xmath74 by @xmath75 in the explicit integral form ( [ pdx ] ) .
finally , we simply have for the values at @xmath335 of the potentials , [ pqd ] @xmath557 as well as for their spatial gradients , [ dpqd]@xmath558 as said above , the main technical step of our strategy will then consist of computing the _ difference _ between such @xmath1-dimensional poisson - type potentials ( [ pqd ] ) or ( [ dpqd ] ) , and their `` pure hadamard - schwartz '' 3-dimensional counterparts , which were already obtained in section [ hadamardpoisson ] .
we denote the difference between the prescriptions of dimensional and `` pure hadamard - schwartz '' regularizations by means of the script letter @xmath384 . given the results @xmath352 and @xmath559 of the two regularizations [ respectively obtained in eqs .
( [ p1 ] ) and ( [ pd ] ) ] we pose @xmath560 that is , @xmath561 is what we shall have to _ add _ to the pure hadamard - schwartz result ( [ defph ] ) in order to get the correct @xmath1-dimensional result .
note that , in this paper , we shall only compute the first two terms , @xmath562 , of the laurent expansion of @xmath561 when @xmath563 .
this is the information we shall need to fix the value of the parameter @xmath4 . we leave to future work an eventual computation of the @xmath1-dimensional equations of motion as an exact function of the complex number @xmath1 .
similarly to the evaluation of the difference @xmath564 $ ] in ref .
@xcite , the difference ( [ dp1 ] ) can be obtained by splitting the @xmath1-dimensional integral ( [ pd ] ) into three volumes , two spherical balls @xmath565 and @xmath566 of radius @xmath289 and centered on the two singularities , and the external volume @xmath567 .
when @xmath309 ( with fixed @xmath289 ) , @xmath565 and @xmath566 tend to the regularization volumes @xmath287 and @xmath288 we introduced in eq .
( [ p1result ] ) . consider first , for a given value @xmath568 , the external integral , over @xmath569 .
[ if wished , two balls with different radii could be used , with the same result . ]
since the integrand is regular on this domain , it is clear that the external integral reduces in the limit @xmath104 to the one in 3 dimensions that is part of the hadamard regularization ( [ p1result ] ) .
so we can write ( for any @xmath568 ) @xmath570 and we see that when computing the difference @xmath561 the exterior contributions will cancel out modulo @xmath571 .
thus we obtain , after this preliminary step [ following eq .
( [ p1result ] ) ] , @xmath572 \bigl<\mathop{f}_1{}_{-2}\bigr>\nonumber\\ & & + \sum_{\ell\geq 0}\frac{(-)^\ell}{\ell!}\partial_l \left(\frac{1}{r_{12}}\right)\left[\sum_{p\leq -\ell-4}\frac{s^{p+\ell+3}}{p+\ell+3}\bigl < n_2^l \mathop{f}_2{}_p\bigr>+\ln\left(\frac{s}{s_2 } \right)\bigl < n_2^l\mathop{f}_2{}_{-\ell-3}\bigr>\right]\biggr\ } \nonumber\\ & & + \mathcal{o}(\varepsilon)\ , .
\label{dp1result}\end{aligned}\ ] ] see section iv of ref .
@xcite for a careful justification of the formal interversions of limits @xmath573 and @xmath563 that we shall do here .
the point is that in order to obtain the difference @xmath561 we do not need the expression of @xmath311 for an arbitrary source point @xmath574 but only in the vicinity of the two singularities : indeed the two local integrals over @xmath565 and @xmath566 in eq .
( [ dp1result ] ) can be computed by replacing @xmath311 by its expansions when @xmath400 and @xmath575 respectively .
we substitute the @xmath37-expansion eq .
( [ fd ] ) into the local integral over @xmath565 , and integrate that expansion term by term .
this readily leads to @xmath576 where we still use the bracket notation to denote the angular average , but now performed in @xmath1 dimensions , _
i.e. _ , @xmath577 here @xmath578is the solid angle element around the direction @xmath277 , and @xmath579 is the volume of the unit sphere with @xmath580 dimensions ( see appendix [ formulae ] for more discussion ) .
to derive ( [ int1 ] ) we used the following relation linking @xmath469 and @xmath581 , @xmath582 concerning the other local integral , over @xmath566 , things are a little bit more involved because we need to perform a multipolar re - expansion of the factor @xmath583 present in that integral around the point @xmath76 . writing down this multipole expansion presents no problem , and in symmetric - trace - free ( stf ) form it reads in @xmath1 dimensions ( in the sense of functions )
. see appendix [ formulae ] for a compendium of @xmath1-dimensional formulae on stf expansions .
see also eq .
( [ eqd6 ] ) in appendix [ littleg ] . ] @xmath584 the multipole expansion being then correctly taken into account , we obtain @xmath585 as we can see , simple poles @xmath87 will occur into our two local integrals , as determined by ( [ int1 ] ) and ( [ int2 ] ) , only for the `` critical '' values @xmath586 and @xmath587 respectively . next we replace the explicit expressions ( [ int1 ] ) and ( [ int2 ] ) into the formula ( [ dp1result ] ) we had for the `` difference '' . as expected we find that the divergencies when @xmath293 , some value @xmath588 being given , cancel out between eqs .
( [ int1])-([int2 ] ) and the remaining terms in ( [ dp1result ] ) , so that the result is finite for any @xmath588 .
furthermore , we find that if we neglect terms of order @xmath571 , the only contributions which remain are the ones coming from the poles ( and their associated finite part ) , _ i.e. _ , for the latter critical values @xmath586 in the case of singularity 1 and @xmath587 in the case of singularity 2 .
the other contributions in ( [ int1 ] ) and ( [ int2 ] ) have a finite limit when @xmath104 which is therefore cancelled by the corresponding terms in hadamard s regularization . as a result
we obtain the following closed - form expression for the difference , which will constitute the basis of all the practical calculations of the present paper , @xmath589\right)\bigl<\mathop { f}_1{}_{-2,q}^{(\varepsilon)}\bigr>\nonumber\\ & -&\frac{1}{\varepsilon ( 1+\varepsilon)}\sum_{q_0\leq q\leq q_1}\left(\frac{1}{q+1}+\varepsilon\ln s_2\right ) \sum_{\ell=0}^{+\infty}\frac{(-)^\ell}{\ell!}\partial_l \left(\frac{1}{r_{12}^{1+\varepsilon}}\right)\bigl
< n_2^l\mathop{f}_2{}_{-\ell-3,q}^{(\varepsilon)}\bigr>\nonumber\\ & + & \mathcal{o}(\varepsilon)\ , . \label{dp1total}\end{aligned}\ ] ] notice that ( [ dp1total ] ) depends on the two `` constants '' @xmath590 and @xmath591 .
as we shall check these @xmath592 and @xmath591 will exactly cancel out the same constants present in the `` pure - hadamard '' calculation , so that the dimensionally regularized acceleration will be finally free of the constants @xmath70 and @xmath78 .
note also that the coefficients @xmath487 and @xmath593 in @xmath1 dimensions depend on the length scale @xmath90 associated with dimensional regularization [ see eq .
( [ l0 ] ) ] . taking this dependence into account one can verify that @xmath70 and @xmath78 in ( [ dp1total ] ) appear only in the combinations @xmath594 and @xmath595 .
let us give also ( without proof ) the formula for the difference between the _ gradients _ of potentials , _
i.e. _ , @xmath596 the formula is readily obtained by the same method as before , and we have @xmath597 formulae ( [ dp1total ] ) and ( [ ddip1total ] ) correspond to the difference of poisson integrals .
but we have already discussed that we shall need also the difference of inverse dalembertian integrals at the 1pn order .
to express as simply as possible the 1pn - accurate generalizations of eqs .
( [ dp1total ] ) and ( [ ddip1total ] ) , let us define two _ functionals _ @xmath598 and @xmath599 which are such that their actions on any @xmath1-dimensional function @xmath311 is given by the r.h.s.s of eqs .
( [ dp1total ] ) and ( [ ddip1total ] ) , _ i.e. _ , so that [ calffi ] @xmath600\ , , \label{calf}\\ \mathcal{d}\partial_ip(1)&=&\mathcal{h}_i\left[f^{(d)}\right]\ , .
\label{calfi}\end{aligned}\ ] ] the difference of 1pn - retarded potentials and gradients of potentials is denoted [ drdir ] @xmath601 where in 3 dimensions the potential @xmath602 is defined by eq .
( [ rpq ] ) and the regularized values @xmath603 and @xmath604 follow from ( [ p1 ] ) , ( [ dip1 ] ) , ( [ q1 ] ) , and where in @xmath1 dimensions @xmath605 and @xmath606 are given by ( [ rd ] ) , ( [ pqd ] ) , ( [ dpqd ] ) . with this notation
we now have our result , which will be stated without proof , that the difference in the case of such 1pn - expanded potentials reads in terms of the above defined functionals @xmath598 and @xmath599 as [ drdirresult ] @xmath607 -\frac{3}{4c^2}\bigl<\mathop{k}_1{}_{-4}\bigr > + \mathcal{o}\left(\frac{1}{c^4}\right ) , \label{dr1result}\\ \mathcal{d}\partial_ir(1)&=&\mathcal{h}_i\left[f^{(d ) } -\frac{r_1 ^ 2}{2c^2(d-2)}\partial_t^2f^{(d)}\right ] -\frac{1}{4c^2}\bigl <
n_1^i\mathop{k}_1{}_{-3}\bigr > + \mathcal{o}\left(\frac{1}{c^4}\right ) .
\label{ddir1result}\end{aligned}\ ] ] these formulae involve some `` effective '' functions which are to be inserted into the functional brackets of @xmath598 and @xmath599 .
beware of the fact that the effective functions are not the same in the cases of a potential and the gradient of that potential .
note the presence , besides the main terms @xmath608 $ ] and @xmath609 $ ] , of some extra terms , purely of order 1pn , in eqs .
( [ drdirresult ] ) .
these terms are made of the average of some coefficients @xmath372 of the powers @xmath610 in the expansion when @xmath400 of the _ second - time - derivative _ of @xmath272 , namely @xmath375 .
they do not seem to admit a simple interpretation .
they are important to get the final correct result .
let us end this section by explaining in more details how we dealt with distributional derivatives in @xmath1 dimensions .
first , it is clear that if we were dealing with @xmath1-dimensional integrals of the type @xmath611 where @xmath612 is some ( formally ) everywhere smooth function of @xmath537 , with fast enough decay at infinity , and where @xmath613 is the elementary newtonian potential in @xmath1 dimensions [ see eqs .
( [ u1 ] ) above ] , we should , in a straightforward @xmath1-continuation of schwartz distributional derivatives , consider that @xmath614 contains , besides an `` ordinary '' singular function @xmath615 ( treated as a pseudo - function in the sense of schwartz ) , a distributional part proportional to @xmath616 .
in other words , we would write [ du1v1]@xmath617 where the indication `` ord '' refers to the `` ordinary '' ( pseudo - function ) part of the repeated derivative .
we have also added the corresponding result for the fourth derivatives of the `` less singular '' kernel @xmath618 , eqs .
( [ v1 ] ) .
note that the decompositions above of @xmath619 or @xmath620 into `` ordinary '' and `` distributional '' pieces arise because of our working in ( @xmath1-dimensional ) @xmath241-space , and of explicitly computing some derivatives , say as @xmath621 \
, r_1^{n-2}$ ] . if we were working in the ( @xmath1-dimensional ) fourier - transform space @xmath622 ( which is where dimensional continuation is most clearly defined @xcite ) , the corresponding decomposition would be simply algebraic : _ e.g. _
@xmath623 , where @xmath624 denotes the stf part of @xmath625 .
the decompositions ( [ du1v1 ] ) are clearly needed when dealing with simple integrals of the type ( [ i ] ) ( with a smooth @xmath626 ) to ensure consistency with the requirement that one may integrate by parts ( which is one of the defining properties of dim .
@xcite ) , and we shall therefore employ it , when applicable . on the other hand ,
most of the singular integrals that we have to deal with look like ( [ i ] ) but contain a _ singular _
function @xmath627 , of the type of eq .
( [ fd ] ) .
it is , however , a very simplifying feature of dim . reg . that when considering integrals like ( [ i ] ) with some _ singular _ @xmath626 we can simply ignore any distributional contributions @xmath628 or its derivatives . indeed , as long as the integer @xmath482 in the powers @xmath629 present in ( [ fd ] ) is different from zero ( which is precisely the case of all delicate terms involving several propagators shrinking towards a particle world - line ) , the `` singular '' expansion ( [ fd ] ) can be considered , in dim .
, as defining a sufficiently smooth function [ by taking both @xmath630 and @xmath7 large enough in ( [ fd ] ) ] which _ vanishes _ , as well as its derivatives , at @xmath255 .
therefore , all the `` dangerous '' terms of the form @xmath631 unambiguously vanish in dim .
let us now consider the consequences of this fact for the time derivatives occurring in expansions such as eqs .
( [ drdirresult ] ) .
the distributional time - derivatives , acting in our present example on @xmath632 or @xmath633 , _
i.e. _ , on functions of @xmath634 , can be treated in a simple way from the rule @xmath635 applicable to the purely distributional part of the derivative .
for instance we can write [ dtu1v1]@xmath636 we have checked using these formulae that all the @xmath1-dimensional terms coming from second - order derivatives of potentials , taken in the distributional sense ( for instance the term @xmath637 in the source of the @xmath638-potential , but also fourth - order derivatives acting on @xmath633 . ] ) yield the _ same _ purely distributional contributions , in the limit @xmath563 , as the ones that would be computed using what we called above a `` pure schwartz '' , three - dimensional computation of such contributions [ to `` smooth '' integrals ( [ i ] ) ] .
on the other hand , the extended version of distributional derivatives introduced in @xcite does yield some specific additional contributions , two of which were already mentioned in @xcite and are reported in the _ items _ ( i ) and ( ii ) of section [ additional ] above , and a third one ( also included in @xcite ) which comes in connection with the second time - derivatives in our formulae for the difference , eqs .
( [ drdirresult ] ) .
let us indicate here that the distributional second - time - derivatives in @xmath1 dimensions have been obtained by using the following ( generalizations of ) gelfand - shilov formulae @xcite , valid for general functions @xmath639 admitting some expansions of the type ( [ fd ] ) : namely , for the spatial derivative , @xmath640 where @xmath641 is the @xmath363-th partial derivative of the @xmath1-dimensional dirac delta - function at the point 1 ( @xmath359 ) and where the angular average is performed over the @xmath147-dimensional sphere having total volume @xmath581 ; and , concerning the time derivative , @xmath642 from the latter formula one deduces the second time - derivative in a way similar to eqs .
( [ dtt ] ) .
we have indicated in the _ item _ ( vii ) of section [ additional ] the correction it leads to when comparing with the extended - hadamard prescription for the second time - derivative , and we have subtracted it from @xmath60 to define the pure hadamard - schwartz result ( [ defph ] ) .
therefore , we consistently do not need to include such an effect into the differences @xmath643 discussed here .
finally we are now in position to obtain the supplement of acceleration @xmath91 induced by dimensional regularization , which is composed of the sum of all the differences of potentials and their gradients computed by means of the generals formulae of ( [ dp1total ] ) , ( [ ddip1total ] ) and ( [ drdirresult ] ) . the term @xmath91 when added to the `` pure - hadamard - schwartz '' acceleration defined by ( [ defph ] ) , gives our result for the dimensionally regularized ( `` dr '' ) acceleration @xmath644 more details on the practical computation of @xmath645 ( which parts of the potentials contribute ; what is the diagrammatic picture ) will be given in section [ dimregstat ] .
the preceding section has explained the method we used to compute the dimensionally regularized equations of motion as the sum ( @xmath646 ; considered modulo 2 ) @xmath647 = \mathbf{a}_a^\text{phs } [ r'_a , s_{a+1 } ] + { \mathcal d } \mathbf{a}_a [ r'_a , s_{a+1 } ; \varepsilon , \ell_0]\,,\ ] ] where the label `` phs '' refers to the `` pure hadamard - schwartz '' definition of the acceleration ( _ i.e. _ , the `` raw '' result of @xcite , after subtraction of the additional contributions quoted in section [ additional ] above , eq .
( [ defph ] ) ) , and where @xmath648 is the difference induced when using dimensional continuation as regularization method , instead of hadamard s one .
a first check on our results will be that , as indicated in ( [ eq5.1 ] ) , the four regularization parameters ( with dimension of length ) , @xmath649 , that enter the hadamard method must cancel between @xmath650 and @xmath648 to leave a result for the dimensionally regularized accelerations @xmath651 which depends only on the two regularization parameters of dimensional continuation : @xmath35 and the basic length scale @xmath90 entering newton s constant in @xmath1 dimensions , @xmath652 , where we recall that @xmath106 denotes the usual three - dimensional newton constant .
the dimensionally regularized acceleration ( [ eq5.1 ] ) has the structure @xmath653 & = & \mathbf{a}_{\text{n}a } [ \mathbf{y}_{12 } ] + \mathbf{a}_{1\text{pn}a } [ \mathbf{y}_{12 } , \mathbf{v}_1 , \mathbf{v}_2 ] \nonumber \\ & + & \mathbf{a}_{2\text{pn}a } [ \mathbf{y}_{12 } , \mathbf{v}_1 , \mathbf{v}_2 ] + \mathbf{a}_{2.5\text{pn}a } [ \mathbf{y}_{12 } , \mathbf{v}_1 , \mathbf{v}_2 ] + \mathbf{a}_{3\text{pn}a } [ \mathbf{y}_{12 } , \mathbf{v}_1 , \mathbf{v}_2]\,,\end{aligned}\ ] ] where we denote @xmath654 .
the 3pn term ( which is the only one to have a pole at @xmath655 ) has a tensor structure of the form ( say for the first particle , @xmath59 ) @xmath656 where , as usual , @xmath657 denotes the unit vector directed from particle 2 to particle 1 .
the scalar coefficients @xmath20 , @xmath658 , @xmath659 entering the equation of motion of @xmath75 can be decomposed in powers of the masses , say [ eq5.4 ] @xmath660 where @xmath661 and @xmath662 are natural integers , with the restrictions indicated .
note that , in eqs .
( [ eq5.4 ] ) , we have conventionally factored out an integer power of the `` full '' ( @xmath1-dimensional ) gravitational constant @xmath132 , and a corresponding integer power of @xmath663 .
this creates a mismatch between the usual 3-dimensional dimension of , say , @xmath664 and the dimension of @xmath665 . using @xmath652 one sees that it is the combination @xmath666 which has the same dimension as @xmath664 .
alternatively said , the ensuing fact that @xmath667 has the same dimension as @xmath664 implies , as indicated in eqs .
( [ eq5.4 ] ) , a dependence of @xmath668 on @xmath669 when @xmath563 .
notice also that in eq .
( [ eq5.3 ] ) we have introduced separate notations for the coefficient of @xmath68 and that of @xmath69 .
actually , the poincar invariance of the equations of motion imposes the restriction @xmath670
so that the last two terms in eq . ( [ eq5.3 ] ) are proportional to the relative velocity @xmath671 .
[ note , however , that @xmath658 is not a function of @xmath672 only ; it depends both on @xmath68 and @xmath69 . ] because the calculation of the separate contributions @xmath650 and @xmath673 to the equations of notion breaks the over - all poincar invariance of the formalism , our computation of the separate pieces @xmath650 and @xmath673 will involve partial contributions to @xmath658 and @xmath659 that do not coincide .
it is only at the end of the calculation that the equality @xmath674 will be satisfied , so that finally @xmath675 most of the coefficients @xmath668 , @xmath676 , @xmath677 entering the 3pn acceleration are well behaved when @xmath678 , in the sense that their evaluation never involves any poles @xmath93 . by this
we mean that whatever be the ( reasonable ) way of decomposing the integral giving a coefficient in separate contributions , the latter contributions do not involve poles @xmath93 .
the subset of coefficients whose evaluation involves poles coincides with the set of `` delicate '' coefficients in the hadamard regularization , namely the nine coefficients contributing to terms of the following form in the acceleration of the first particle : @xmath679 \mathbf{n}_{12 } \nonumber\\ & & \qquad + \frac{g^3 \ , m_1 ^ 2 \ , m_2}{c^6 \
, r_{12}^4 } \bigl [ c_{21 } ( \mathbf{v}_1 , \mathbf{v}_2 ) \mathbf{n}_{12 } \ , + c'_{21 } ( \mathbf{v}_1 , \mathbf{v}_2 ) \mathbf{v}_1 - c''_{21 } ( \mathbf{v}_1 , \mathbf{v}_2)\ , \mathbf{v}_2\bigr ] \nonumber\\ & & \qquad + \frac{g^3 \ , m_2 ^ 3}{c^6 \ ,
r_{12}^4 } \bigl[c_{03 } ( \mathbf{v}_1 , \mathbf{v}_2 ) \
, \mathbf{n}_{12 } + c'_{03 } ( \mathbf{v}_1 , \mathbf{v}_2)\ , \mathbf{v}_1 - c''_{03 } ( \mathbf{v}_1 , \mathbf{v}_2)\ , \mathbf{v}_2 \bigr]\,.\end{aligned}\ ] ] the first three terms in eq . ( [ eq5.5 ] ) do not depend on velocities and will be referred to as the _
static _ delicate contributions , by contrast to the _ kinetic _ delicate contributions involving the velocity - dependent coefficients @xmath680 , @xmath681 , @xmath682 , @xmath683 , @xmath684 , and @xmath685 ( they depend on @xmath68 , @xmath69 and also on @xmath67 ) . correspondingly to the decomposition ( [ eq5.1 ] ) of the equations of motion , the dimensionally regularized static contributions , unambiguously obtained from the test - mass limit @xmath686 . ]
@xmath687 , @xmath688 , @xmath689 to the acceleration of the first particle can be written as the sum ( @xmath690 , @xmath691 , @xmath692 ) @xmath693 = c_{mn}^\text{phs } [ r'_1 , s_2 ] + { \mathcal d } c_{mn } [ r'_1 , s_2 , \varepsilon]\,.\ ] ] in this subsection , we discuss the explicit evaluation of the pure hadamard - schwartz static coefficients @xmath694 .
as explained in the previous section , the phs static contributions @xmath695 $ ] are obtained from the results reported in @xcite by undoing two things .
first , the `` bf '' results reported there for @xmath80 ( eq .
( 7.16 ) of @xcite ) were expressed in terms of the three parameters @xmath61 , @xmath62 and @xmath4 , instead of the two pure hadamard parameters @xmath61 and @xmath78 more relevant for the present purpose .
the introduction of the parameter @xmath4 was motivated by requiring that the full set of equations of motion ( which _ a priori _ depended on four independent regularizing parameters @xmath61 , @xmath62 , @xmath77 , @xmath78 ) admit a conserved energy .
this led to the link , eqs .
( 7.9 ) in @xcite,$ ] given by ( [ distrpart ] ) above .
another derivative , the `` correct '' one , was also considered in @xcite and shown to yield equivalent equations of motion . the pure hadamard result does not depend on this choice because we shall subtract below the specific additional contributions coming from the distributional derivative @xmath414 $ ] . ] @xmath696 when inserting ( [ eq5.8 ] ) in the expression of @xmath697 $ ] we find , as it should be , that the result simplifies to an expression depending only on the two pure hadamard parameters @xmath61 and @xmath78 .
this leads to the following net results from @xcite , [ eq5.9]@xmath698 & = & - \frac{3187}{1260 } + \frac{44}{3 } \ln \left ( \frac{r_{12}}{r'_1 } \right ) , \\
c_{22}^\text{bf } [ r'_1 , s_2 ] & = & \frac{34763}{210 } - \frac{41}{16 } \ , \pi^2\ , , \\
c_{13}^\text{bf } [ r'_1 , s_2 ] & = & \frac{1565}{9 } - \frac{41}{16 } \ , \pi^2 - \frac{44}{3 } \ln \left ( \frac{r_{12}}{s_2 } \right).\end{aligned}\ ] ] second , ref .
@xcite obtained their results for the equations of motion by adding to the pure hadamard - schwartz contributions 7 additional corrections , imposed by their extended - hadamard regularization and explained in section [ additional ] above : see the _ items _ ( i)-(vii ) there .
note that these various corrections affect the `` delicate '' contributions to @xmath80 , in general both the static and kinetic ones , but only five of them contribute to the static part .
these are the self term ( [ deltaself ] ) , the leibniz term ( [ deltaleibniz ] ) , the @xmath136-correction given by ( [ deltav ] ) , the eom non - distributivity ( [ deltaeom ] ) , and the distributional time - derivative one ( [ deltatime ] ) .
following eq .
( [ defph ] ) , and focussing on the static contributions , we now _ subtract _ these static terms from the result ( [ eq5.9 ] ) in order to get the looked - for pure hadamard contributions : [ eq5.14]@xmath699 & = & c_{31}^\text{bf } [ r'_1 , s_2 ] - \frac{779}{210}\ , , \\ c_{22}^\text{phs } [ r'_1 , s_2 ] & = & c_{22}^\text{bf } [ r'_1 , s_2 ] + \frac{97}{210}\ , , \\ c_{13}^\text{phs } [ r'_1 , s_2 ] & = & c_{13}^\text{bf } [ r'_1 , s_2 ] - 5 + \frac{88}{9 } - \frac{151}{9 } + \frac{2}{15}\,,\end{aligned}\ ] ] _ i.e. _ , explicitly , [ eq5.15]@xmath699 & = & - \frac{1123}{180 } + \frac{44}{3 } \ln \left ( \frac{r_{12}}{r'_1 } \right ) , \\ c_{22}^\text{phs } [ r'_1 , s_2 ] & = & 166 - \frac{41}{16 } \ , \pi^2\,,\\ c_{13}^\text{phs } [ r'_1 , s_2 ] & = & \frac{7291}{45 } - \frac{41}{16 } \ , \pi^2 - \frac{44}{3 } \ln \left ( \frac{r_{12}}{s_2 } \right).\end{aligned}\ ] ] note in passing that though the coefficient @xmath700 does not contain regularization logarithms , its evaluation involves many intermediate logarithmic divergencies that cancel in the final result
. such `` cancelled logs '' lead to as much ambiguity in the final result than uncancelled ones that explicitly depend on an arbitrary regularization scale such as @xmath61 or @xmath78 in @xmath701 or @xmath702 .
we now turn to the evaluation of the `` dim .- reg .
minus pure - hadamard '' differences @xmath703 in eq .
( [ eq5.7 ] ) , coming from the differences @xmath704 in eq .
( [ eq5.1 ] ) .
we start from the @xmath1-dimensional expression for the acceleration @xmath80 [ see eq .
( [ smallaccel ] ) for a short - hand form ] , which is itself expressed in terms of the @xmath1-dimensional elementary potentials @xmath136 , @xmath137 , @xmath158 , @xmath159 , @xmath450 , @xmath161 , @xmath705 , @xmath163 and @xmath164 defined in section [ fieldeq ] .
each elementary potential can be naturally decomposed in a `` compact '' ( or , equivalently , `` contact '' ) piece ( whose source is compact , _
i.e. _ , involves the basic delta - function sources @xmath143 , @xmath144 , @xmath149 ) and a `` non - compact '' one ( whose source is non - linearly generated and extends all over space ) . the potentials @xmath136 , @xmath137 and @xmath158 are purely `` compact '' , @xmath706 , @xmath707 , @xmath708 , while all the other potentials admit a decomposition of the form @xmath709 , etc . for instance , the `` compact '' part of @xmath159 is defined by @xmath710 while its `` non - compact '' part is defined by @xmath711 a more complicated example is the potential @xmath712 with @xmath713\,,\ ] ] and
@xmath714 the @xmath715 contribution can be further decomposed into the piece of @xmath716 whose source is quadratic in compact potentials , namely @xmath717 and its `` cubically non - compact '' piece given by @xmath718 to get a feeling of the actual evaluation of the difference @xmath719 let us consider a specific contribution to @xmath80 , say the term @xmath720 \equiv \frac{4}{c^4 } \ , ( \partial_i \ , \hat x)_{\mathbf{x } = \mathbf{y}_1}\,.\ ] ] it can be decomposed into : ( i ) its `` compact '' piece @xmath721 $ ] , ( ii ) its `` quadratically non - compact '' one @xmath722 $ ] , ( iii ) its `` cubically non - compact '' part @xmath723 $ ] .
it is sometimes convenient to think of the various contributions to @xmath80 in terms of space - time diagrams .
if we represent the basic delta - function sources [ proportional to @xmath724 and @xmath725 as two world - lines and each propagator @xmath520 as a dotted line , a `` compact '' contribution to @xmath80 will be represented by one of the diagrams in fig .
[ fig1 ] . .
the dotted line represents @xmath520 , the cross represents the field point @xmath241 ( here taken on the first worldline ) , and the bullet represents either a source point or ( in the figures below ) an intermediate nonlinear vertex . ] for instance , fig .
[ fig1]a can represent a term @xmath726 in @xmath727 in which the ( compact ) source @xmath143 of @xmath136 is proportional to @xmath728 and involves no further powers of the masses , while fig .
[ fig1]b represents a _ self - action _
term , @xmath78 ) the self - action diagrams , such as fig .
[ fig1]b or fig .
[ fig1]d , are the first divergencies that one encounters and must then renormalize away , dimensional regularization has the technically useful property of setting all of these diagrams to zero .
indeed , when using a time - symmetric propagator @xmath729 these diagrams are seen to involve the coinciding - point limits of @xmath730 , which vanish when @xmath731 by dimensional continuation in @xmath1 . ] in @xmath732 with source proportional to @xmath733 . by contrast ,
[ fig1]c might correspond to another term in @xmath732 where the compact source @xmath143 is concentrated at @xmath76 , @xmath734 , and where a part of the `` effective mass '' , @xmath735 contains , besides the overall factor @xmath65 , another factor @xmath64 . as all the sources of @xmath736 contain , besides some `` basic '' @xmath737 , a potential ( @xmath136 , @xmath137 , @xmath738 or @xmath158 ) , the diagrams contained in @xmath739 $ ] will be at least of the form of fig .
[ fig1]c , [ fig1]d , [ fig1]e , [ fig1]f , or will involve a more complicated mass - dependence .
the quadratically non - compact terms @xmath740 $ ] will then contain diagrams of the type of fig .
[ fig2 ] , while the cubically non - compact term @xmath741 $ ] contains many subdiagrams of the type sketched in fig .
[ fig3 ] . .
] . ] the particular term ( [ eq5.20 ] ) that we considered contains only diagrams of the general type of figs .
[ fig1 ] , [ fig2 ] or [ fig3 ] .
note , however , that there are also more non - linear contributions to @xmath727 , such as some terms in @xmath742 = \frac{16}{c^6 } \ , ( \partial_i \ , \hat t)_1\,,\ ] ] corresponding to diagrams of the type sketched in fig .
[ fig4 ] . .
] similarly to the diagrams fig .
[ fig1]c or fig .
[ fig2]d , all the diagrams above can be modified by the presence of additional lines propagating directly between the two world - lines and corresponding to `` potential '' modifications of compact - support sources . as underlined in section [ iteratedeinstein ] above , the 3pn equations of motion do _ not _ involve `` quartically non - linear '' contributions corresponding to diagrams such as those of fig .
terms like @xmath743 or @xmath744 are of this form , and they do occur in the 3pn acceleration @xmath80 , but since they involve double contracted gradients , it was possible to integrate them away thanks to rule ( ii ) of section [ fieldeq ] ; see eq .
( [ expfi ] ) in appendix [ expfieldeq ] below . on the other hand , terms of the form @xmath745 or @xmath746
do not occur at the 3pn order @xmath747 , although they are of the third post-_minkowskian _ order @xmath748 . at the 3pn order . ]
drawing diagrams often helps to highlight the nature of the uv singularities contained in the integrals they represent . as a rule of thumb , the `` delicate '' diagrams , that might involve poles , or cancelled poles , when @xmath563 ( corresponding to logarithms , or cancelled logarithms , in @xmath42 ) are characterized by the presence of a subdiagram containing three propagator lines that can simultaneously shrink to zero size , as a subset of vertices coalesce together on one of the two world - lines .
examples of such uv dangerous diagrams are fig .
[ fig2]d and fig . [ fig3]b [ for vertices coalescing towards @xmath749 or fig .
[ fig3]d and fig . [ fig4]d [ for vertices coalescing on the second world - line ] .
the former diagrams can give poles proportional to @xmath750 ( with some velocity dependence , or some extra mass dependence due to an extra line propagating between the two world - lines ) , while the latter can give poles proportional to @xmath751 ( possibly with some extra velocity or mass dependence ) .
the reason why three simultaneously shrinking propagators can yield poles as @xmath678 is easy to see in the approximation where the relativistic propagators @xmath520 are replaced by non - relativistic ones @xmath752 . indeed ,
when three such propagators shrink simultaneously , the overall integral contains a subintegral of the form @xmath753 . on the other hand , beyond our obtaining a heuristic feeling of what are the origins of the poles in @xmath80
, we did not use a diagrammatic technique for evaluating the equations of motion .
[ note , however , that a generalization of the ( 2pn level ) work @xcite would lead to a diagrammatic technique for evaluating the fokker lagrangian of two point masses . ] our actual computations used the techniques elaborated in the previous sections .
we evaluated the contributions to the difference @xmath754 coming from all the terms in the expression for @xmath80 deduced from ( [ eqgeod ] ) together with the complete expanded forms ( [ exppi])-([expfi ] ) .
however , as expected from various arguments diagrammatic analysis , existence of ( possibly cancelled ) logarithms in the corresponding @xmath42 evaluation most of the terms lead to a vanishing difference @xmath754 .
the only terms that give non - vanishing contributions to @xmath704 are the four terms given in eq .
( [ smallaccel ] ) , @xmath755 & = & \frac{4}{c^4 } \ , ( \partial_i \ , \hat x)_1\ , , \quad a_1^i [ \hat t ] = \frac{16}{c^6 } \ , ( \partial_i \ , \hat t)_1 \ , , \nonumber\\ a_1^i [ \hat r_i ] & = & \frac{8}{c^4 } \ , \frac{d}{dt } \ , ( \hat r_i)_1\ , , \quad a_1^i [ \hat y_i ] = \frac{16}{c^6 } \ , \frac{d}{dt } \ , ( \hat y_i)_1\ , . \label{eq5.23}\end{aligned}\ ] ] note that , for the contributions associated to @xmath161 and @xmath450 , one needs a 1pn - accurate treatment of both their respective sources and the propagator @xmath520 .
apart from the compact support terms in the sources for @xmath161 , @xmath164 , @xmath450 and @xmath163 which lead to zero difference , most of the non - compact terms do lead to some non - vanishing contributions to the difference of acceleration @xmath704 .
we give in tables [ table1]-[table4 ] the contributions to @xmath703 associated to the various individual source terms of the `` delicate '' potentials @xmath161 , @xmath164 , @xmath450 and @xmath163 , which were displayed in section [ fieldeq ] , eqs .
( [ potentialeq ] ) [ of course , we limit ourselves to non - compact source terms ] . in these tables , we use the simplifying notation @xmath756 where @xmath757 denotes the euler constant .
.static contributions of @xmath758 .
all the results are presented modulo some neglected terms @xmath571 .
the `` principal '' part of a term corresponds to the term @xmath759 in the 1pn symmetric propagator @xmath760 , while the `` retarded '' part corresponds to the purely 1pn piece @xmath761 .
the `` extra term '' refers to the last term in the r.h.s .
( [ ddir1result ] ) .
note that , in view of eqs .
( [ eq5.23 ] ) , one must multiply the results by a factor 4 in order to get the contributions to the coefficients @xmath762 in the equations of motion .
[ cols="^,^,^,^",options="header " , ] summing up the separate non - vanishing contributions displayed in tables [ table1]-[table4 ] , we get the following total differences [ eq5.28]@xmath763 finally , adding eqs .
( [ eq5.28 ] ) to the pure hadamard - schwartz result ( [ eq5.15 ] ) , we get the dimensionally regularized static contributions to @xmath80 : [ eq5.29 ] @xmath764 as expected the two hadamard regularization length scales @xmath61 and @xmath78 have cancelled between @xmath694 and @xmath765 to leave a result which depends only on the dim .
regularization parameter @xmath86 .
one might be surprised by the presence in @xmath766 of terms @xmath767 compared to corresponding terms @xmath768 in @xmath694 , and by the absence of any adimensionalizing length scale in these logarithms of @xmath769 .
these two properties can be understood when one remembers from the discussion above that the coefficients which have the same physical dimension as @xmath770 are the combinations @xmath771 . in the present case
, this means that @xmath772 are dimensionless .
it is easy to see , thanks to the pole terms @xmath773 and the expansion @xmath774 , that the combinations @xmath775 and @xmath776 do indeed depend only on the dimensionless quantities @xmath36 and @xmath777 .
the first computation of the dimensional continuation of the 3pn gravitational interaction of point masses was done in adm coordinates and resulted into a _ finite _ ( _ i.e. _ , without @xmath778 poles ) answer @xcite .
our task in analyzing the physical meaning of the harmonic - coordinates result ( [ eq5.29 ] ) is to interpret the presence of @xmath778 poles in it . for this
we have to remember that , as in quantum field theory ( qft ) , dimensional continuation is a _ regularization _ method which , like all regularization methods , transforms truly infinite results , say containing @xmath779 , into finite , but `` large '' ones , which depend on some cut - off parameter , _
e.g. _ @xmath780 or @xmath781 .
any regularization must be followed by a
_ renormalization _
process which allows one to absorb the cut - off dependent terms in some of the basic _ bare parameters _ of the theory . in order to have a clearer understanding of the poles in the ( static ) equations of motion ( [ eq5.29 ] )
[ we shall prove below that our discussion extends to the full , velocity - dependent equations of motion ] , we need to analyze the presence of poles in the `` bulk '' metric , _
i.e. _ , the metric @xmath782 evaluated at a generic field point @xmath241 , away from the two world - lines .
indeed , if we were considering the gravitational field generated by regular ( _ i.e. _ , non point - like ) sources , a complete physical description of their gravitational effects would necessitate the simultaneous consideration of the bulk metric and of the equations of motion of the ( extended ) sources .
similarly , in the present formal study of two - point - like sources , we need to consider both the equations of motion @xmath783 and the bulk metric @xmath784 .
it is here that the diagrammatic representation introduced above plays a useful role in highlighting the structure of divergencies in the equations of motion and in the bulk metric .
indeed , it is clear that the divergent diagrams of the equations of motion of the first particle , where the @xmath778 pole is due to the presence of a subdivergence induced by three propagators shrinking onto the second world - line ( such as in fig . [ fig3]d or fig .
[ fig4]d ) , will correspond to similar @xmath778 poles in the bulk metric , for the corresponding `` bulk diagrams '' where the special point marked by a cross in the diagrams above ( denoting the coincidence @xmath245 ) is detached from the first world - line to end at an arbitrary point in the bulk , as indicated in fig .
[ fig6 ] . , is detached from the first worldline . ] evidently , in addition to such diagrams as fig .
[ fig6]a and [ fig6]b which will contain ( at least ) a factor @xmath751 , there will exist `` mirror diagrams '' , containing a factor @xmath785 , and obtained by exchanging the labels 1 and 2 . on the other hand , note that the bulk poles @xmath786 of the type of fig .
[ fig6]c and [ fig6]d do not ( necessarily ) correspond to poles in the equations of motion of @xmath75 because their coincidence limits @xmath45 induce diagrams of the type of fig .
[ fig3]a or [ fig4]a containing four shrinking propagators instead of three .
[ though such diagrams would exhibit worse divergences in a dimensionful cut - off regularization schemes , they are generally non dangerous in dimensional regularization because the integral @xmath787 has no pole as @xmath563 . ] a careful analysis of the possible presence of poles in the various potentials @xmath788 we use to parametrize the bulk metric ( aided by the structure of potentially dangerous terms sketched in fig .
[ fig6 ] ) shows that , at the 3pn approximation , in @xmath169 , @xmath789 in @xmath790 , and @xmath791 in @xmath170 . ]
such poles can only be present in the 1pn - level expansion of @xmath161 and in the newtonian - level approximation of @xmath164 .
drawing on the results of @xcite and @xcite we can also see that all velocity - dependent terms in the poles present in @xmath792 and @xmath164 ( _ i.e. _ , the terms proportional to @xmath793 or @xmath794 ) exactly cancel in the combination @xmath795 that matters for the bulk metric .
[ this shows up , for instance , in eq .
( 7.1 ) of @xcite which implies that the divergencies linked to the second world - line , characterized by the presence of @xmath591 , do not depend on velocities .
this shows up also in the absence of poles in the velocity - dependent contributions to @xmath80 proportional to @xmath751 , see eqs .
( [ eq5.76c])-([eq5.76d ] ) below .
] we are therefore left with evaluating the poles present in the _ static limit _
@xmath796 of @xmath797 and @xmath164 .
clearly , from fig . [ fig6 ] , the poles in @xmath792 and @xmath164 will come only from cubically non - compact ( @xmath798 ) sources .
finally , as we are only interested in the pole part we can neglect the @xmath1-dependence of the coefficients in the sources of @xmath161 and @xmath164 ( which we indicate by using a symbol @xmath799 ) .
thus , these poles can only come from [ eq5.31ab ] @xmath800 \nonumber\\ & \simeq & \box^{-1 } \bigl[\partial_{ij } \
, v \ , \box^{-1 } \bigl(- \partial_i \ , v \partial_j \ , v\bigr)\bigr ] \ , , \label{eq5.30}\\ \hat t^{cnc}_\text{static } & \simeq & \box^{-1 } \biggl [ \frac{1}{2 } \ , v^2 \ , \partial_t^2 \ ,
v + \partial_{ij } \ ,
v\ , \hat{z}_{ij}^{nc } \biggr ] \nonumber \\ & \simeq & \box^{-1 } \biggl[\frac{1}{2 } \ , v^2 \ , \partial_t^2 \ , v + \partial_{ij } \ , v \ , \box^{-1 } \bigl(- 2\partial_i \ , v \partial_t \ , v_j \bigr ) \biggr]\ , .
\label{eq5.31}\end{aligned}\ ] ] the static poles ( involving factors @xmath785 or @xmath751 ) in eqs .
( [ eq5.31ab ] ) are then obtained by : ( 1 ) considering sources involving three times @xmath801 or three times @xmath802 ( where @xmath803 denotes the piece @xmath804 in @xmath136 ) , ( 2 ) evaluating the time derivatives in the static limit , using for instance @xmath805 and ( 3 ) expanding up to the required accuracy the ( time - symmetric ) propagators according to @xmath806 .
as an example among the simplest terms , let us consider the @xmath785 contribution coming from the first term on the r.h.s . of eq .
( [ eq5.31 ] ) , @xmath807_\text{static } = - \frac{1}{2 } \ , \delta^{-1 } [ v_1 ^ 2 \ , a_1^j \ , \partial_j \ , v_1 ] = - \frac{1}{6 } \ , a_1^j \ ,
, \delta^{-1 } [ v_1 ^ 3]\,.\ ] ] using @xmath808 and @xmath809 $ ] , one finds that the pole part of ( [ eq5.33 ] ) reads @xmath810 similarly an analysis of the second source term in eq .
( [ eq5.31 ] ) yields @xmath811 so that the full ( static ) contribution of @xmath164 is @xmath812 the analysis of the pole part in the static limit of @xmath161 , eq .
( [ eq5.30 ] ) , is more intricate because one must expand to 1pn accuracy both @xmath813 and the propagator @xmath520 .
this yields @xmath814 let us now consider the improved @xmath136-potential ( [ calv ] ) that makes up the essential part of @xmath169 , @xmath815 such that @xmath816 $ ] , see eq .
( [ g00 ] ) . combining the results above
, we find that the only @xmath778 poles in the bulk metric @xmath817 show up in @xmath169 at the 3pn level and are ( when expressed in terms of the improved potential @xmath198 , and after cancellation of @xmath818 terms between @xmath161 and @xmath164 of the following static form @xmath819\,,\ ] ] where @xmath820 , @xmath821,@xmath822 s are finite when @xmath104 .
to understand better the structure of result ( [ eq5.39 ] ) let us introduce the notation @xmath823 where @xmath106 is the 3-dimensional newton constant and @xmath824 the @xmath1-dimensional acceleration of @xmath825 .
[ this definition ensures that @xmath826 has the physical dimension of a length . ] in terms of the definition ( [ eq5.40 ] ) , the result ( [ eq5.39 ] ) can be equivalently written as @xmath827 + \frac{1}{c^2 } \mathcal{v}_2 + \frac{1}{c^4 } \mathcal{v}_4 + \frac{1}{c^6 } \mathcal{v}_6\,,\ ] ] where the pole part is entirely contained in the terms proportional to @xmath828 and @xmath829 [ @xmath830 here differs from @xmath822 in eq .
( [ eq5.39 ] ) by some finite corrections when @xmath104 ] .
the fact that poles appear only in @xmath198 , at order @xmath791 , implies that there are no divergencies in the harmonic gauge conditions ( [ gaugeidentities ] ) in the bulk .
indeed , ( [ divi ] ) needs @xmath198 at order @xmath832 only , and ( [ djwij ] ) at newtonian order only .
result ( [ eq5.41 ] ) indicates a simple way of renormalizing away the poles present in the bulk metric .
indeed , the logic up to now has been to describe in the simplest possible manner a gravitationally interacting two - particle system , parametrized by the following _ bare _ parameters : @xmath833 , @xmath834 , @xmath835 , @xmath836 , @xmath837 , considered in everywhere - harmonic coordinates , @xmath838 . in particular , the internal structure of each particle has been , up to now , entirely described by a monopolar stress - energy distribution , _
i.e. _ , @xmath839 . in other words
, we have set to zero any higher multipolar structure .
( [ eq5.41 ] ) is most simply interpreted by saying that the non - linear interactions ( see fig . [ fig6 ] ) dress each particle by a cloud of gravitational energy which generates , at the 3pn order , a divergent _ dipole _ in the newtonian - like potential .
therefore , to get a net , finite bulk gravitational field we must endow each initial particle by an infinite , bare dipole , corresponding to a counterterm @xmath840 , which will cancel the non - linearly generated one ( [ eq5.41 ] ) .
an equivalent , but technically simpler way of endowing each particle by a bare structure able to cancel the dipolar pole terms in ( [ eq5.41 ] ) is simply to say that the central _ bare _ world - lines used in our derivations up to now , henceforth denoted as @xmath841 , can be decomposed in a finite _ renormalized _
part @xmath842 and a formally infinite shift @xmath843 involving a pole @xmath93 , @xmath844 the gravitational potential of two point particles [ @xmath845 is then @xmath846 \\ & + & \frac{1}{c^2 } \ , \mathcal{v}_2 ( \mathbf{x } , \mathbf{y}_a^\text{ren } + \bm{\xi}_a ) + \frac{1}{c^4 } \ , \mathcal{v}_4 ( \mathbf{x } , \mathbf{y}_a^\text{ren } + \bm{\xi}_a ) + \frac{1}{c^6 } \ , \mathcal{v}_6 ( \mathbf{x } , \mathbf{y}_a^\text{ren } + \bm{\xi}_a)\ , .
\nonumber\end{aligned}\ ] ] assuming that the vector @xmath843 is of 3pn order [ _ i.e. _ , @xmath847 , we can rewrite eq .
( [ eq5.430 ] ) as @xmath848 \\ & + & \frac{1}{c^2 } \
, \mathcal{v}_2 ( \mathbf{x } , \mathbf{y}_a^\text{ren } ) + \frac{1}{c^4 } \ , \mathcal{v}_4 ( \mathbf{x } , \mathbf{y}_a^\text{ren } ) + \frac{1}{c^6 } \ , \mathcal{v}_6 ( \mathbf{x } , \mathbf{y}_a^\text{ren } ) + \mathcal{o}\left(\frac{1}{c^8}\right ) , \nonumber\end{aligned}\ ] ] which makes it clear that the potential will be finite ( at 3pn accuracy ) when @xmath563 if we choose @xmath849 where by @xmath850 we mean a term finite when @xmath563 and of the 3pn order .
we shall henceforth refer to @xmath843 in eq .
( [ eq5.44 ] ) as a _ shift _ of the @xmath218 world - line .
the reasoning above shows that the introduction of such shifts , at the 3pn order and having the pole structure ( [ eq5.40 ] ) , is _ necessary _ to renormalize away the poles present in the _
bulk metric_. it remains to show that these shifts are also _ sufficient _ to renormalize away the poles present in the _ equations of motion_. the effect of 3pn - level shifts @xmath843 on the equations of motion is easy to obtain .
indeed , the equations of motion we computed above concern the original , _ bare _ world - lines @xmath841 .
for the first particle , they had the structure ( in dimensional regularization ) @xmath851 where @xmath852 . here @xmath853 denotes the dimensionally continued newtonian - level acceleration , @xmath854 where by a slight abuse of notation we pose @xmath855 , where @xmath856 denotes the @xmath1-dimensional gravitational constant , and where the @xmath1-dependent correcting factors @xmath857 tend to 1 as @xmath563 , but will play a significant role below .
when inserting the redefinitions ( [ eq5.42 ] ) into ( [ eq5.45 ] ) one easily finds that the _ renormalized _ equations of motion , _
i.e. _ , the equations for @xmath842 , read [ using only @xmath858 at this stage ] @xmath859 where @xmath860 with @xmath861
let us note that the effect on the equations of motion of a ( 3pn ) shift of the world - lines , eq .
( [ eq5.50 ] ) , is technically identical to the effect on the equations of motion of the restriction to the world - lines of a 3pn - level coordinate transformation , say @xmath862 and @xmath863 .
indeed , a coordinate transformation has two effects : ( i ) it changes the bulk metric by @xmath864 , where @xmath865 denotes the lie derivative along @xmath866 , and ( ii ) it induces a shift of the world - lines @xmath867 ( plus non - linear terms in @xmath866 ) , where we denote the coordinate change at the location of the @xmath868-th particle by @xmath869_{\mathbf{x } = \mathbf{y}_a}$ ] . because of the diffeomorphism invariance of the total action , the effect ( i ) does not change the action , , ensures that it does not contribute to first order in @xmath866 . ]
so that the net effect of a coordinate transformation on the equations of motion reduces to the effect ( [ eq5.50 ] ) of the following shift induced on the world - lines : @xmath870 the coordinate transformations considered in @xcite , see eq .
( 6.11 ) there , were of the type @xmath871 , where the @xmath872 s are some coefficients , so that we see that the latter induced shift reduces at the 3pn order to the ( regularization of the ) purely spatial coordinate transformation evaluated on the world - line : @xmath873 we have checked that the formula given by eq . ( 6.15 ) in @xcite for the coordinate transformation of the acceleration of the particle 1 gives exactly the same result as the one computed from the effect of the shift ( [ eq5.50 ] ) .
[ the agreement extends to @xmath1 dimensions if we consider the straightforward extension of the latter coordinate transformation @xmath866 to @xmath1 dimensions , namely @xmath874 . ]
note that the coordinate transformations @xmath875 were considered in @xcite only in terms of their effects , eqs .
( [ xia])-([xia ] ) , on the equations of motion .
this was sufficient to prove , for instance , that the two constants @xmath70 and @xmath71 in the 3pn equations of motion are not physical , because they can be gauged away in 3 dimensions and therefore will never appear in any physical result .
however , we remark that the extension to @xmath1 dimensions of the coordinate transformation @xmath875 of the _ bulk _ metric , say @xmath874 ( with coefficients @xmath876 , as needed to remove the poles in the equations of motion ) , does not lead to a bulk metric free of poles .
indeed , assuming @xmath876 , we see that the pole in the spatial coordinate transformation @xmath877 would then induce a pole in the spatial components of the metric , @xmath878 , but this is inadmissible because we have proved above eqs .
( [ eq5.31ab ] ) that , at 3pn order , only the time - time component of the bulk metric contained a pole .
a bulk coordinate transformation of the type above can then remove the poles in the time - time component of the bulk metric only at the price of creating a pole in the initially pole - free spatial metric .
we shall leave to future work a complete clarification of the possibility of using , within our dim .
context , a coordinate transformation to induce the shifts ( [ eq5.44 ] ) . for the time being
, what is important is that our introduction above of shifts of the world - lines ( _ a priori _ unconnected to any coordinate transformation ) is a consistent way of renormalizing away the poles in the metric , and that its effect on the equations of motion , eq .
( [ eq5.50 ] ) , is identical to the transformations of the acceleration obtained in ref .
@xcite .
it remains now to show that the same world - line shifts ( [ eq5.44 ] ) that renormalize away the poles in the bulk metric , eq . ( [ eq5.43 ] ) , do renormalize away also the poles present in the original bare equations of motion [ see ( [ eq5.29 ] ) for the static contributions and ( [ eq5.76 ] ) below for the kinetic ones ] . for this purpose
let us consider a shift of the more general form @xmath879 where @xmath880 represents a certain numerical coefficient depending on @xmath1 , and where @xmath824 denotes the @xmath1-dimensional acceleration of @xmath825 given by its newtonian approximation ( [ eq5.46 ] ) [ but , for notational simplicity , we henceforth drop the label @xmath7 on such accelerations entering 3pn effects ] . inserting ( [ eq5.51 ] ) into ( [ eq5.50 ] ) yields ( for the index @xmath59 ) @xmath881\,,\ ] ] where @xmath882 and @xmath883 [ and also , as before , @xmath884 . before further evaluating ( [ eq5.52 ] ) by inserting the explicit expression ( [ eq5.46 ] ) for the acceleration , we shall consider some simple but important consequences of the structure ( [ eq5.52 ] ) . as recalled in the introduction , previous work on the 3pn equations of motion in _ harmonic coordinates
_ has shown that these equations necessarily belonged to a three - parameter class of equations of motion , say @xmath885 the dimensionless parameter @xmath4 could not be determined by the previous work in harmonic coordinates .
however , comparison with the work in adm coordinates , has shown @xcite that , _ if there were consistency _ between the two calculations one should have the following link between @xmath4 and the corresponding adm `` static ambiguity '' parameter @xmath9 , @xmath886 if dimensional regularization is a fully consistent regularization scheme for classical perturbative gravity , we then expect that the dim .
determination of @xmath9 in adm coordinates @xcite , namely @xmath887 , should lead to a dim .
reg . direct determination of @xmath4 ( in harmonic coordinates ) of @xmath888 .
we will turn to this verification in a moment .
the two other parameters , denoted above @xmath61 , @xmath62 , entering the general `` parametric '' harmonic equations of motion ( [ eq5.53 ] ) have the dimension of length and have the character of gauge parameters .
indeed , they can be chosen at will ( except that one can not set them to zero ) by the effect of shifts of the world - line , induced for instance [ but not necessarily , _ cf . _ a discussion in section [ shift ] above ] by some gauge transformations . in the way they were originally introduced @xcite ,
the two parameters @xmath70 and @xmath71 can be interpreted as some infinitesimal radial distances used as cut - offs when the field point tends towards the two singularities @xmath75 and @xmath76 .
therefore in principle @xmath592 and @xmath889 should initially be thought of as being ( formally ) infinite .
however , it is trivial to show that by a ( formally infinite ) gauge transformation , involving the logarithmic ratios @xmath890 and @xmath891 , where @xmath892 and @xmath893 denote any two _ finite _ length scales , one can replace everywhere @xmath70 , @xmath71 by the finite scales @xmath892 , @xmath893 . by this process
it is therefore possible to identify the two sets of scales and thereby to think of the scales @xmath70 , @xmath71 as being in fact finite , as implicitly done in ref .
@xcite . in the language of renormalization theory ,
the original ( infinitesimal ) scales @xmath70 and @xmath71 would be referred to as _ hadamard - regularization
_ scales entering the computation of divergent poisson integrals [ see section [ hadamardpoisson ] above ] , while the ( finite ) scales @xmath892 and @xmath893 would be referred to as the arbitrary _ renormalization _ scales entering the final , renormalized harmonic - coordinates equations of motion . in the present paper , in order to remain close to the notation used in @xcite , we shall keep the notation @xmath70 and @xmath71 , but interpret them as arbitrary finite constants , which means that we shall _ identify _ them with the finite renormalization length scales @xmath892 and @xmath893 .
in other words , the scales @xmath70 , @xmath71 used in the present section should in principle be distinguished from the scales @xmath70 , @xmath71 used in section [ hadamardpoisson ] above .
[ remember , in this respect , that the regularization scales @xmath70 , @xmath71 have disappeared when computing the dim .
equations of motion . ] with our notation , and still focussing on the static contributions to the equations of motion , the `` parametric '' equations of motion ( [ eq5.53 ] ) imply the following structure for the static coefficients @xmath894 : [ cstaticlambda ] @xmath895 it will be convenient to replace the parameter @xmath4 by the parameter @xmath9 , using ( [ eq5.54 ] ) as a defining one - to - one map between @xmath4 and @xmath9 . with this change of notation the static coefficients become [ cstaticomega ] @xmath896 note that there are two combinations of the three coefficients @xmath897 which do not depend on @xmath669 , namely @xmath898 , and the combination @xmath899 , or even better the combination @xmath900 which depends neither on @xmath669 nor on @xmath9 ( or @xmath4 ) , and which contains , like for @xmath898 , simpler looking rational numbers .
we now come back to the effect of the general shift ( [ eq5.51 ] ) on the dim .
equations of motion .
let us first focus on the static terms .
we recall that the ( dim .
_ renormalized _ equations of motion had necessarily the form ( [ eq5.49 ] ) . by projecting the latter equation along the static terms @xmath894 , with @xmath901 [ recalling eq .
( [ eq5.5 ] ) ] , it will induce a result for the _ renormalized _ static coefficients of the form @xmath902 where the @xmath903 s are the static coefficients corresponding to @xmath904 , eq . ( [ eq5.50 ] ) .
when choosing @xmath905 of the form ( [ eq5.51 ] ) , we see from eq . ( [ eq5.52 ] ) that @xmath906 is simply obtained by projecting the first term on the r.h.s . of ( [ eq5.52 ] ) . remembering that @xmath907 and @xmath908 , we see that the latter term contains the factor @xmath909 . therefore , without doing any further calculation , we see that the shifts @xmath910 have the special properties : @xmath911 and @xmath912 . in other words ,
a shift of the world - lines of the type ( [ eq5.51 ] ) leaves invariant both @xmath700 and the combination @xmath913 ( as well therefore as the combination @xmath914 considered above ) . as a consequence
, we can compute without effort from our previous regularized ( but unrenormalized ) dim .
reg . results ( [ eq5.29 ] ) the following two combinations of the @xmath915-_renormalized _ static coefficients : @xmath916 by comparing ( [ eq5.63 ] ) with eq .
( [ eq5.59 ] ) we discover that our present calculation using dimensional regularization in harmonic coordinates necessarily implies that @xmath917 this nicely confirms the previous determination of @xmath9 by a dim . reg .
calculation in adm - type coordinates @xcite .
we think that our present harmonic - coordinates dim .
result calculation is important in proving the consistency of dimensional regularization , and thereby in confirming the physical significance of the result ( [ eq5.65 ] ) .
a recent calculation @xcite has also confirmed independently the result ( [ eq5.65 ] ) by means of a completely different method based on surface integrals , and aimed at describing compact ( strongly - gravitating ) objects . by comparing ( [ eq5.64 ] ) with ( [ eq5.61 ] )
, we further see that @xmath918}\,.\ ] ] contrary to ( [ eq5.65 ] ) which represents the determination of a physical parameter ( having an invariant meaning ) , the result ( [ eq5.66 ] ) has no invariant physical significance .
( [ eq5.66 ] ) is simply a consequence of our particular choice for the shift vector ( [ eq5.51 ] ) , in which we assumed that @xmath880 is a purely numerical coefficient , independent on any properties indexed by the particles labels 1 and 2 . in summary
the particular shift ( [ eq5.51 ] ) yields some equations of motion which are physically equivalent to a subclass of the general equations of motion considered in @xcite , characterized by the constraint ( [ eq5.66 ] ) .
next we relate the common length scale ( [ eq5.66 ] ) to the basic length scale @xmath90 entering dimensional regularization . for doing this
we need to fully specify the value of the shift , _
i.e. _ , to choose a specific coefficient @xmath919 in eq .
( [ eq5.51 ] ) .
we already know from eq .
( [ eq5.40 ] ) that the coefficient @xmath880 in eq .
( [ eq5.51 ] ) must tend to @xmath920 when @xmath32 , if the @xmath921-shift is to remove the poles in the bulk metric . as in quantum field theory
( qft ) we could then define the _ minimal subtraction _ ( ms ) shift as @xmath922 however , as is well - known in qft , such a ms subtraction has the unpleasant feature of leaving some logarithms of @xmath923 and the euler constant in the renormalized results .
these numbers come from the expansion of the gamma function and the associated dimension - dependent powers of @xmath923 entering the @xmath1-dimensional green function . in our context
, these numbers showed up in eq .
( [ eq5.29 ] ) in the guise of the combination @xmath924 like in qft , this leads us to consider the following _ modified minimal subtraction _ ( @xmath925 ) shift ,
@xmath926 which differs from the ms shift ( [ eq5.67 ] ) by the explicit factor of @xmath927 it contains , and by the use of the @xmath1-dimensional ( newtonian ) acceleration given by eq .
( [ eq5.46 ] ) .
the inclusion of two explicit powers of @xmath928 in the coefficient @xmath880 entering eq .
( [ eq5.51 ] ) , _ i.e. _ the definition @xmath929 , means , when remembering that @xmath824 , eq .
( [ eq5.46 ] ) , contains one power of @xmath928 , that the static terms in eq .
( [ eq5.52 ] ) will have four powers of @xmath928 and the kinetic terms three .
the overall factor @xmath930 in the static terms is natural because these terms are of order @xmath931 and the @xmath241-space gravitational propagator in @xmath1 dimensions always includes the combination @xmath932 . finally , using the fact that the expansion of @xmath933 near @xmath42 reads @xmath934 it is easy to see that the @xmath925 shift defined by ( [ eq5.68 ] ) will cancel the @xmath935 in the bare dim .
results of eq .
( [ eq5.29 ] ) . finally , we find that the evaluation of eq .
( [ eq5.52 ] ) for the specific @xmath925 shift , given by @xmath936 , yields for the @xmath925-renormalized static coefficients [ eq5.70abc]@xmath937 where the reader can note that the @xmath938 entering the bare dim .
result ( [ eq5.29 ] ) have been transformed into @xmath939 through the @xmath36-expansion of the factor @xmath940 present in @xmath904 .
we already discussed above the comparison of two simple combinations of the dim .
results ( [ eq5.70abc ] ) with the hadamard - regularization results ( [ cstaticlambda ] ) .
it is easy to see that the remaining independent combination , say ( [ eq5.70 ] ) , is fully consistent with its counterpart ( [ eq5.55 ] ) , and allows one to relate the basic renormalization length scale @xmath90 entering dim .
reg . to the common length scale ( [ eq5.66 ] )
entering the general equations of motion of @xcite : @xmath941 evidently , the precise values one gets for @xmath61 and @xmath62 depend on the precise choice of the compensating shift .
let us now remark that in fact one can recover exactly , provided of course that the crucial result ( [ eq5.65 ] ) holds , the general `` dissymmetric '' class of equations of motion of @xcite , _
i.e. _ , the general parametric result ( [ cstaticlambda ] ) or ( [ cstaticomega ] ) with @xmath942
. for this purpose it suffices to consider a slightly more general shift than the one assumed in the @xmath925 regularization .
namely , consider a shift of the same form as ( [ eq5.51 ] ) , but in which one allows the @xmath1-dependent coefficient @xmath880 to depend on the label of the particle in question , that is @xmath943 where now @xmath944 and @xmath945 are allowed to be different from each other . the most general way of parametrizing such dissymmetric @xmath946 [ however constrained by @xmath947 is @xmath948\ , , \label{ead}\ ] ] with two independent numerical coefficients @xmath949 .
it is then easily checked that the shift ( [ shifteta ] ) defined by the particular choice @xmath950 transforms the dim .
equations of motion into the general ( @xmath951-dependent ) family of solutions obtained in @xcite .
if we suppose that the constraints ( [ eq5.73 ] ) hold , then @xmath952 and we recover the shift assumed in the @xmath925 regularization . on the other hand , note than one can reach even more general classes of renormalized harmonic equations of motion in dim . reg .
( as one could have also done in hadamard regularization ) .
indeed , we could use the freedom indicated in eq .
( [ eq5.44 ] ) of adding _ arbitrary _ finite parts to the shifts .
anyway , the result that the shift ( [ shifteta])-([rhoa ] ) gives equivalence between the dim . reg . and
the extended - hadamard 3pn accelerations [ we check in section [ kinetic ] below that the kinetic terms work also ] , constitutes the main result of the present paper ( theorem 2 in the introduction ) .
the two approaches we have discussed here are of course equivalent : choosing some dim .
basic length scale @xmath90 and some specific , simplifying dim .
reg . shift ( such as the @xmath925 one ) , and then determining the values of the scales @xmath951 for which the dim .
results match with the hadamard reg .
ones ; or arbitrarily choosing some hadamard scales @xmath951 and then determining the corresponding general dissymmetric dim .
reg . shift ( [ shifteta])-([rhoa ] ) , in terms of the chosen @xmath951 s . what is important is that we have checked that the _ three _ renormalized dim .
static coefficients ( [ eq5.29 ] ) are fully compatible with the _ three _ extended - hadamard reg .
static coefficients ( [ cstaticlambda ] ) or ( [ cstaticomega ] ) , and that their comparison yields _ one and only one _ physical result , namely : @xmath114 . up to now we have verified the following aspects of the consistency of a dim .
treatment of the 3pn dynamics of two point particles : ( 1 ) consistency between the shift ( [ eq5.40 ] ) needed to renormalize the bulk metric and the shift ( [ eq5.68 ] ) [ or ( [ shifteta])-([ead ] ) ] needed to renormalize the equations of motion ; ( 2 ) consistency between the three finite , renormalized dim .
static coefficients ( [ eq5.70abc ] ) and the general three - dimensional ones ( [ cstaticlambda ] ) , @xcite ; and ( 3 ) consistency between the present dim . reg .
value of @xmath4 and the previously derived dim .
value of @xmath9 in the adm - hamiltonian @xcite .
it remains , however , to check that the velocity - dependent terms in the renormalized dim .
equations of motion do agree with their analogs in the harmonic - coordinates equations of motion of @xcite .
this will in particular prove that the dim .
equations of motion are lorentz invariant . in the notation of eq .
( [ eq5.5 ] ) above , we need to consider the values of the velocity - dependent coefficients @xmath953 , @xmath954 , @xmath955 , @xmath956 , @xmath957 , and @xmath958 . in ref .
@xcite , they were shown to take the following parametric forms , which actually depend only on the regularization scale @xmath61 but not on @xmath62 nor @xmath4 : [ eq5.78 ] @xmath959 \ln\left(\frac{r_{12}}{r'_1}\right ) + \frac{48197}{840}\ , v_1 ^ 2 -\frac{36227}{420 } ( v_1 v_2 ) + \frac{36227}{840}\ , v_2 ^ 2 \nonumber\\ & & - \frac{45887}{168 } ( n_{12}v_1)^2 + \frac{24025}{42 } ( n_{12}v_1)(n_{12}v_2 ) -\frac{10469}{42 } ( n_{12}v_2)^2 \,,\qquad\label{eq5.78a}\\ c'^{\text{bf}}_{21}(r'_1 ) = c''^{\text{bf}}_{21}(r'_1)&= & -44 ( n_{12}v_{12 } ) \ln\left(\frac{r_{12}}{r'_1}\right ) + \frac{31397}{420}(n_{12}v_1 ) - \frac{36227}{420 } ( n_{12}v_2 ) \,,\label{eq5.788b}\\ c_{03}^{\text{bf}}&=&18 ( v_1 v_2 ) - 9\ , v_2 ^ 2 - ( n_{12}v_1)^2 + 2(n_{12}v_1)(n_{12}v_2 ) + \frac{43}{2}(n_{12}v_2)^2\,,\label{eq5.78c}\\ c'^{\text{bf}}_{03 } = c''^{\text{bf}}_{03}&=&4(n_{12}v_1 ) + 5 ( n_{12}v_2)\ , .
\label{eq5.78d}\end{aligned}\ ] ] as explained in eq .
( [ eq5.1 ] ) , the dim .
expressions of these coefficients can be computed as the sum of pure hadamard - schwartz ( phs ) contributions , @xmath694 and @xmath960 , and the `` @xmath961 '' differences , @xmath703 , @xmath962 .
the calculation of the phs contributions has been explained in section [ additional ] above , and we get the following results from eqs .
( [ defph ] ) and ( [ eq5.78 ] ) : [ eq5.74 ] @xmath963 \ln\left(\frac{r_{12}}{r'_1 } \right ) + \frac{10639}{168}\ , v_1 ^ 2 - \frac{5879}{60 } ( v_1 v_2 ) + \frac{5843}{120}\ , v_2 ^ 2 \nonumber\\ & & - \frac{50885}{168 } ( n_{12}v_1)^2 + \frac{1892}{3 } ( n_{12}v_1)(n_{12}v_2 ) - \frac{3325}{12 } ( n_{12}v_2)^2 \,,\label{eq5.74a}\\ c'^{\text{phs}}_{21}&=&- 44 ( n_{12}v_{12 } ) \ln\left(\frac{r_{12}}{r'_1}\right ) + \frac{7279}{84 } ( n_{12}v_1 ) - \frac{5879}{60 } ( n_{12}v_2 ) \,,\label{eq5.74b}\\ c''^{\text{phs}}_{21}&=&- 44 ( n_{12}v_{12 } ) \ln\left(\frac{r_{12}}{r'_1}\right ) + \frac{5189}{60 } ( n_{12}v_1 ) - \frac{5843}{60 } ( n_{12}v_2 ) \,,\label{eq5.74c}\\ c_{03}^\text{phs}&=&18 ( v_1 v_2 ) - \frac{64}{7}\ , v_2 ^ 2 - ( n_{12}v_1)^2 + 2 ( n_{12}v_1)(n_{12}v_2 ) + \frac{311}{14}(n_{12}v_2)^2\,,\label{eq5.74d}\\ c'^{\text{phs}}_{03}&=&4(n_{12}v_1 ) + 5 ( n_{12}v_2)\,,\label{eq5.74e}\\ c''^{\text{phs}}_{03}&=&4(n_{12}v_1 ) + \frac{37}{7 } ( n_{12}v_2 ) \,.\label{eq5.74f}\end{aligned}\ ] ] @xmath176 secondly , the method explained in sections [ diff ] and [ dimregstat ] for computing the differences @xmath964 is found ( after doing calculations similar to those reported in the tables of `` static '' contributions above ) to lead to [ eq5.75 ] @xmath965 \left[v_{12}^2 -5 ( n_{12}v_{12})^2 \right ] + \frac{499}{42}\ , v_1 ^ 2 - \frac{359}{15}(v_1 v_2 ) \nonumber\\ & & + \frac{184}{15}\ , v_2 ^ 2 - \frac{2957}{42 } ( n_{12}v_1)^2 + \frac{425}{3 } ( n_{12}v_1 ) ( n_{12}v_2 ) - \frac{217}{3 } ( n_{12}v_2)^2 + { \mathcal o } ( \varepsilon)\,,\qquad \label{eq5.75a}\\ { \mathcal d } c'_{21}&=&\left [ \frac{22}{\varepsilon } -33\ln\left(\overline q\ , r'^{4/3}_1 r_{12}^{2/3}\right ) \right](n_{12}v_{12 } ) + \frac{499}{21}(n_{12}v_1 ) - \frac{359}{15}(n_{12}v_2 ) + { \mathcal o } ( \varepsilon)\ , , \label{eq5.75b}\\ { \mathcal d } c''_{21}&=&\left [ \frac{22}{\varepsilon } -33\ln\left(\overline q\ , r'^{4/3}_1 r_{12}^{2/3}\right ) \right](n_{12}v_{12 } ) + \frac{359}{15}(n_{12}v_1 ) - \frac{368}{15}(n_{12}v_2 ) + { \mathcal o } ( \varepsilon)\ , , \label{eq5.75c}\\ { \mathcal d } c_{03}&=&\frac{1}{7}\ , v_2 ^ 2 - \frac{5}{7 } ( n_{12}v_2)^2\,,\label{eq5.75d}\\ { \mathcal d } c'_{03}&=&0\,,\label{eq5.75e}\\ { \mathcal d } c''_{03}&=&-\frac{2}{7}(n_{12}v_2 ) \,.\label{eq5.75f}\end{aligned}\ ] ] @xmath176 together with eqs . ( [ eq5.28 ] ) above , these equations give the full difference between the dimensionally regularized and the pure hadamard accelerations , and they constitute the main new input of the present work .
the bare dim .
results , @xmath966 , read therefore [ eq5.76 ] @xmath967 \left[v_{12}^2 - 5 ( n_{12}v_{12})^2\right ] + \frac{1805}{24}\ , v_1 ^ 2 - \frac{1463}{12 } ( v_1 v_2 ) \nonumber\\ & & + \frac{1463}{24}\ , v_2 ^ 2 -\frac{8959}{24 } ( n_{12}v_1)^2 + \frac{2317}{3 } ( n_{12}v_1 ) ( n_{12}v_2 ) - \frac{4193}{12 } ( n_{12}v_2)^2 \,,\qquad\label{eq5.76a}\\ c'^{\text{dr}}_{21 } = c''^{\text{dr}}_{21}&=&\left [ \frac{22}{\varepsilon } - 33 \ln\left(\overline q\ , r_{12}^2\right ) \right](n_{12}v_{12 } ) + \frac{1325}{12}(n_{12}v_1 ) - \frac{1463}{12 } ( n_{12}v_2 ) \,,\label{eq5.76b}\\ c_{03}^\text{dr}&=&18 ( v_1 v_2 ) - 9\ , v_2 ^ 2 - ( n_{12}v_1)^2 + 2(n_{12}v_1)(n_{12}v_2 ) + \frac{43}{2}(n_{12}v_2)^2\,,\label{eq5.76c}\\ c'^{\text{dr}}_{03 } = c''^{\text{dr}}_{03}&=&4(n_{12}v_1 ) + 5 ( n_{12}v_2)\,.\label{eq5.76d}\end{aligned}\ ] ] @xmath176 note that in the final result the equality between @xmath968 and @xmath969 , _
i.e. _ , between @xmath658 and @xmath659 in eq .
( [ eq5.3 ] ) , is recovered .
the bare dim .
kinetic coefficients ( [ eq5.76 ] ) contain poles @xmath93 but do not depend anymore of the arbitrary hadamard regularization scale @xmath61 which appeared in ( [ eq5.74 ] ) . as in the case discussed above of the static coefficients
the previous kinetic coefficients do not involve any adimensionalizing length scales in the logarithms of @xmath663 they contain .
this is consistent with the fact that it is the combinations @xmath970 and @xmath971 which have the same physical dimension as their @xmath42 counterparts .
finally , given a specific choice of shift , say the @xmath925 one , eq .
( [ eq5.68 ] ) , the _
kinetic coefficients are obtained by adding to ( [ eq5.76 ] ) the velocity - dependent part of the effect of the shift , _
i.e. _ , the second term on the r.h.s . of eq .
( [ eq5.52 ] ) [ with , say , @xmath972 .
our final @xmath973 results for the renormalized kinetic coefficients are found to be [ eq5.77 ] @xmath974
\ln\left(\frac{r_{12}}{\ell_0}\right ) + \frac{1321}{24}\ , v_1 ^ 2 -\frac{979}{12 } ( v_1 v_2 ) \nonumber\\ & & + \frac{979}{24}\ , v_2 ^ 2 - \frac{6275}{24 } ( n_{12}v_1)^2 + \frac{1646}{3 } ( n_{12}v_1)(n_{12}v_2 ) -\frac{2851}{12 } ( n_{12}v_2)^2 \,,\qquad\label{eq5.77a}\\ c'^{\overline{\text{ms}}}_{21 } = c''^{\overline{\text{ms}}}_{21}&= & -44 ( n_{12}v_{12 } ) \ln\left(\frac{r_{12}}{\ell_0}\right ) + \frac{841}{12}(n_{12}v_1 ) - \frac{979}{12 } ( n_{12}v_2 ) \,,\label{eq5.77b}\\ c_{03}^{\overline{\text{ms}}}&=&18 ( v_1 v_2 ) - 9\ , v_2 ^ 2 - ( n_{12}v_1)^2 + 2(n_{12}v_1)(n_{12}v_2 ) + \frac{43}{2}(n_{12}v_2)^2\,,\label{eq5.77c}\\ c'^{\overline{\text{ms}}}_{03 } = c''^{\overline{\text{ms}}}_{03}&=&4(n_{12}v_1 ) + 5 ( n_{12}v_2 ) \,.\label{eq5.77d}\end{aligned}\ ] ] when comparing these results with the ones of ref .
@xcite , eqs .
( [ eq5.78 ] ) above , one remarkably finds that our previously derived link ( [ eq5.73 ] ) is necessary and sufficient for ensuring the full compatibility between the renormalized dim .
results and the corresponding hadamard reg
note that the rational coefficients entering the dim .
results are often simpler than the coefficients entering the equations of motion of @xcite . the results eq .
( [ eq5.77 ] ) complete our check of the full consistency of the dim .
evaluation of the 3pn equations of motion , and the proof of theorems 1 and 2 stated in the introduction .
we have used dimensional regularization ( _ i.e. _ analytic continuation in the spatial dimension @xmath1 ) to determine the spacetime metric and the equations of motion ( eom ) in _ harmonic coordinates _ , of two , gravitationally interacting , point masses , at the third post - newtonian ( 3pn ) order of general relativity .
our starting point consisted in writing the 3pn - accurate metric @xmath975 in terms of a certain number of `` elementary potentials '' @xmath976 , satisfying a hierarchy of inhomogeneous dalembert equations of the form @xmath977 .
the sources of the latter equations contain both `` compact '' terms , _
i.e. _ , in the present case _ contact _
terms of the form @xmath978 \delta^{(d)}(\mathbf{x}-\mathbf{y}_1)$ ] , and nonlinearly generated `` non compact '' terms of the typical form , say , @xmath979 .
this representation of the 3pn metric , as well as the associated iterative way of solving for the potentials [ using the time - symmetric green s function @xmath980 is a direct generalization of the one used in ref .
@xcite . however , it has been crucial for our work to determine ( in section [ fieldeq ] ) the dependence upon the dimension @xmath1 of the coefficients appearing in this representation , as well as the @xmath1-dependence of the kernels expressing the operators @xmath759 and @xmath981 in @xmath241-space . by studying the structure of the iterative solution for the metric , and that of the corresponding eom ( which are conveniently pictured by means of diagrams , see figs . [ fig1]-[fig8 ] ) , we determined , in the form of a laurent expansion in @xmath35 , the pole part of the metric @xmath94 , and the pole and finite parts of the eom , namely @xmath982 where @xmath983 and @xmath984 .
[ see , however , appendix [ littleg ] where the basic quadratically non - linear kernel @xmath43 is computed in any @xmath1 dimensions , not necessarily close to 3 . ]
our calculations relied heavily on previous work in @xmath42 @xcite , and were technically implemented in two steps ( at least for the determination of the eom , which are more delicate ; the determination of the pole part of the metric uses only the second step ) .
\(i ) a first step consisted of subtracting from the final , published results for the eom @xcite , seven contributions that were specific consequences of the use of an extension of the hadamard regularization method @xcite ( which included an extension of the schwartz notion of distributional derivative ) .
the result of this first step is referred to as the `` pure hadamard - schwartz '' evaluation of the eom .
\(ii ) the second step was the evaluation of the _ difference _ between the dimensional regularization of each contribution to the eom ( written in terms of the iterative solutions for the various potentials @xmath976 ) , and the corresponding `` pure hadamard - schwartz '' contribution obtained in the first step .
this difference is obtained , similarly to the method used in @xcite , by splitting the @xmath1-dimensional integral into several pieces , and by carefully analyzing the terms due to the neighborhoods of the two singular points @xmath985 ( including possible @xmath1-dimensional distributional contributions ) .
concerning the `` bulk '' metric @xmath94 , at a field point away from the singular particle world - lines @xmath986 , we derived only the pole part , that is the coefficient of @xmath778 in the laurent expansion of @xmath987 .
we found that at the 3pn order only the time - time component of the metric @xmath988 contained a pole [ see eq .
( [ eq5.39 ] ) ] . for the eom we derived both the pole part and the finite part , _
i.e. _ @xmath989 .
the parts of the eom for which the regularization was delicate are given by the nine coefficients @xmath990 defined in eq .
( [ eq5.5 ] ) .
our complete results for the dimensionally - regularized values of these nine `` delicate '' coefficients are given in eqs .
( [ eq5.29 ] ) and ( [ eq5.76 ] ) .
we proved that the pole parts of both the metric and the eom can be `` renormalized away '' by suitable _
shifts of the world - lines _ of the form @xmath991 , where @xmath841 is the original world - line on which are initially concentrated the @xmath992-function sources representing the point masses , where the shifts @xmath993 are of the 3pn order , and where the eom of the renormalized world - line @xmath842 is _ finite _ as @xmath104 .
the general form of the needed shifts is given by eq .
( [ eq5.44 ] ) with ( [ eq5.40 ] ) .
the renormalized eom corresponding to the `` modified minimal subtraction '' scheme ( [ eq5.68 ] ) are given by eqs .
( [ eq5.70abc ] ) and ( [ eq5.77 ] ) .
the finite renormalized 3pn - accurate eom obtained by using the general ( two - parameter ) renormalization shift ( [ shifteta])-([rhoa ] ) were shown to be _ equivalent _ to the final ( three - parameter ) eom of @xcite if and only if the hadamard - undetermined dimensionless parameter @xmath4 which entered the latter equations takes the unique value @xmath994 .
this value is in agreement with the result of a previous dimensional - regularization determination of the arnowitt - deser - misner hamiltonian ( in adm - like coordinates ) @xcite , which led to the unique determination of the adm analogue of @xmath4 , namely @xmath6 .
the value for @xmath4 or @xmath9 is also in agreement with the recent work @xcite which derived the 3pn equations of motion in harmonic gauge using a surface - integral approach .
our result provides an important check of the consistency of dimensional regularization because our calculations are very different from the ones of @xcite , notably we use a different coordinate system and a different method for iterating einstein s field equations .
however , the applicability of our general approach to higher post - newtonian orders remains unexplored .
finally , the present work opens the way to a dimensional - regularization determination of the several unknown dimensionless parameters that were shown to enter the hadamard - regularization of the 3pn binary s energy flux ( in harmonic coordinates ) @xcite .
the completion of the 3pn energy flux is urgent in view of its importance in determining the gravitational waveforms emitted by inspiralling black hole binaries , which are primary targets for the international network of interferometric gravitational wave detectors ligo / virgo / geo .
most of the algebraic calculations reported in this paper were done with the help of the software _
mathematica_. t.d . would like to thank the kavli institute for theoretical physics for hospitality ( under the partial support of the national science foundation grant no .
phy99 - 07949 ) while this work was completed .
we give in this appendix several expanded expressions which are too lengthy to be included in the body of the article .
the expanded form of the metric ( [ metric ] ) is easier to compare with the literature , and notably with eqs .
( 3.24 ) of ref .
@xcite : [ metricexp ] @xmath995 + \frac{8}{c^6}\left[\hat x + v_i v_i + \frac{1}{6}\ , v^3 + \left(\frac{d-3}{d-2}\right ) v k\right]\nonumber\\ & & + \frac{32}{c^8}\left [ \hat t -\frac{1}{2}\ , v \hat x + \hat r_i v_i -\frac{1}{2}\ , v v_i v_i -\frac{1}{48}\ , v^4 + \frac{1}{4}\left(\frac{d-3}{d-2}\right ) k v^2 -\frac{1}{4}\left(\frac{d-3}{d-2}\right)^2k^2\right]\nonumber\\ & & + { \mathcal{o}}\left(\frac{1}{c^{10}}\right ) , \label{g00exp}\\ g_{0i}&=&-\frac{4}{c^3}\ , v_i -\frac{8}{c^5}\left[\hat r_i -\frac{1}{2}\left(\frac{d-3}{d-2}\right)v v_i\right ] -\frac{16}{c^7}\biggl [ \hat y_i+\frac{1}{2}\ , \hat w_{ij}v_j\nonumber\\ & & + \frac{1}{4}\left(1+\frac{1}{(d-2)^2}\right)v^2 v_i -\frac{1}{2}\left(\frac{d-3}{d-2}\right ) v \hat r_i + \frac{1}{2}\left(\frac{d-3}{d-2}\right)^2 k v_i\biggr ] + { \mathcal{o}}\left(\frac{1}{c^{9}}\right ) , \label{g0iexp}\\ g_{ij}&=&\delta_{ij}\biggl\ { 1 + \frac{2}{(d-2)c^2}\ , v + \frac{2}{(d-2)^2 c^4}\left[v^2 -2 ( d-3 ) k\right]\nonumber\\ & & + \frac{8}{c^6}\left [ \frac{\hat x}{d-2 } + \frac{v_k v_k}{d-2 } + \frac{v^3}{6(d-2)^3 } -\frac{(d-3)}{(d-2)^3}\ , v k\right ] \biggr\}\nonumber\\ & & + \frac{4}{c^4}\ , \hat w_{ij } + \frac{16}{c^6}\left [ \hat z_{ij } + \frac{v\hat w_{ij}}{2(d-2 ) } -v_i v_j\right ] + { \mathcal{o}}\left(\frac{1}{c^8}\right ) . \label{gijexp}\end{aligned}\ ] ] @xmath176 the inverse metric is such that @xmath996 in @xmath120 space - time dimensions . in terms of the modified newtonian potential @xmath198 defined in eq .
( [ calv ] ) above , it reads : [ invmetric ] @xmath997 + \frac{8\hat r_i}{c^5 } + \frac{16}{c^7}\left[\hat y_i -\frac{1}{2}\hat w_{ij}v_j\right]\right\ } + { \mathcal{o}}\left(\frac{1}{c^9}\right),\nonumber\\ \label{gup0i}\\ g^{ij}&=&e^{-\frac{2\mathcal{v}}{(d-2)c^2}}\left\{\delta_{ij } -\frac{4}{c^4 }
\hat w_{ij } -\frac{16}{c^6 } \left[\hat z_{ij } + \frac{1}{2(d-2)}\ , \delta_{ij } v_k v_k \right]\right\ } + { \mathcal{o}}\left(\frac{1}{c^8}\right ) .
\label{gupij}\end{aligned}\ ] ] note the change of signs in the exponentials [ with respect to the covariant metric ( [ metric ] ) ] , in front of @xmath998 in eq .
( [ gup0i ] ) , as well as for the @xmath999 and @xmath193 terms in eq .
( [ gupij ] ) .
note also that the @xmath1000 contribution to @xmath170 has disappeared in the inverse spatial metric @xmath1001 .
the full post - newtonian expansion of this inverse metric reads : [ invmetricexp ] @xmath1002 -\frac{8}{c^6}\left[\hat x -v_i v_i + \frac{v^3}{6 } -\left(\frac{d-3}{d-2}\right ) v k\right]\nonumber\\ & & -\frac{32}{c^8}\left [ \hat t + \frac{1}{2}\ , v \hat x -\hat r_i v_i -\frac{1}{2}\ , v v_i v_i + \frac{v^4}{48 } -\frac{1}{4}\left(\frac{d-3}{d-2}\right ) k v^2 + \frac{1}{4}\left(\frac{d-3}{d-2}\right)^2k^2\right]\nonumber\\ & & + { \mathcal{o}}\left(\frac{1}{c^{10}}\right ) , \label{gup00exp}\\ g^{0i}&=&-\frac{4}{c^3}\ , v_i -\frac{8}{c^5}\left[\hat r_i + \frac{1}{2}\left(\frac{d-3}{d-2}\right)v v_i\right ] -\frac{16}{c^7}\biggl [ \hat y_i-\frac{1}{2}\ , \hat w_{ij}v_j\nonumber\\ & & + \frac{1}{4}\left(1+\frac{1}{(d-2)^2}\right)v^2 v_i + \frac{1}{2}\left(\frac{d-3}{d-2}\right ) v \hat r_i -\frac{1}{2}\left(\frac{d-3}{d-2}\right)^2 k v_i\biggr ] + { \mathcal{o}}\left(\frac{1}{c^{9}}\right ) , \label{gup0iexp}\\ g^{ij}&=&\delta_{ij}\biggl\ { 1 -\frac{2}{(d-2)}\ , \frac{v}{c^2 } + \frac{2}{(d-2)^2 c^4}\left[v^2 + 2 ( d-3 ) k\right]\nonumber\\ & & -\frac{8}{c^6}\left [ \frac{\hat x}{d-2 } + \frac{v_k v_k}{d-2 } + \frac{v^3}{6(d-2)^3 } + \frac{d-3}{(d-2)^3}\ , v k\right ] \biggr\}\nonumber\\ & & -\frac{4}{c^4}\ , \hat w_{ij } -\frac{16}{c^6}\left [ \hat z_{ij } -\frac{1}{2(d-2)}\ , v\hat w_{ij}\right ] + { \mathcal{o}}\left(\frac{1}{c^8}\right ) .
\label{gupijexp}\end{aligned}\ ] ] @xmath176 the determinant @xmath1003 of the metric is a useful quantity , notably to compute the `` gothic '' metric @xmath127 , which is the natural variable when using the harmonic - coordinate system .
the simplest way to compute it is to use the exponential form ( [ metric ] ) of the metric , and to perform a cofactor expansion across both the first line and the first column : @xmath1004 since @xmath1005 \times\left[\delta_{ij } + { \mathcal{o}}(1/c^4)\right]$ ] , the determinant of the @xmath1006 matrix @xmath1007 reads @xmath1008 therefore , the determinant of the full metric is given by @xmath1009 where we have used the fact that @xmath1010 . note that this formula suffices to compute @xmath167 up to order @xmath200 included if one knows the spatial metric @xmath170 up to this same order . at the 3pn order
, we get easily @xmath1011\nonumber\\ & & \hphantom{-e^\frac{4\mathcal{v}}{(d-2)c^2}\biggl\{}+ \frac{16}{c^6}\ , v_i v_i \biggr\ } + { \mathcal{o}}\left(\frac{1}{c^8}\right)\nonumber\\ & = & -e^\frac{4\mathcal{v}}{(d-2)c^2 } \left[1 + \frac{4}{c^4}\ , \hat w + \frac{16}{c^6}\left(\hat z + \frac{1}{d-2}\ , v_i v_i\right ) \right]+{\mathcal{o}}\left(\frac{1}{c^8}\right ) , \label{g}\end{aligned}\ ] ] where we used the expansion @xmath1012 , valid for any matrix @xmath1013 whose entries are small with respect to 1 .
we can now compute the square root of this determinant , and give its full post - newtonian expansion : [ sqrtg ] @xmath1014+{\mathcal{o}}\left(\frac{1}{c^8}\right ) \label{sqrtgcompact}\\ & = & 1 + \frac{2}{(d-2)}\ , \frac{v}{c^2 } + \frac{2}{c^4}\left [ \hat w + \frac{v^2}{(d-2)^2 } -\frac{2(d-3)}{(d-2)^2}\ , k \right ] + \frac{8}{c^6}\biggl [ \hat z + \frac{v_i v_i}{d-2}\nonumber\\ & & + \frac{\hat x}{d-2 } + \frac{v\hat w}{2(d-2 ) } + \frac{v^3}{6(d-2)^3 } -\frac{d-3}{(d-2)^3}\ , v k \biggr ] + { \mathcal{o}}\left(\frac{1}{c^8}\right ) .
\label{sqrtgexpanded}\end{aligned}\ ] ] the gothic metric @xmath127 can now be written easily by combining eqs .
( [ invmetric ] ) or ( [ invmetricexp ] ) with ( [ sqrtgcompact ] ) or ( [ sqrtgexpanded ] ) .
we shall not display here the explicit results , since they were not directly useful for the present article .
let us however quote the expression of the ricci tensor in terms of the gothic metric , in @xmath120 space - time dimensions and in any gauge : @xmath1015 as usual , a comma denotes partial derivation , and @xmath1016 is the inverse of @xmath1017 . in terms of the gothic metric , the harmonic gauge condition ( [ harmgauge ] ) takes a particularly simple form : @xmath1018 this is the reason why this gothic metric can be useful to write the field equations . note that several terms of eq .
( [ riccigoth ] ) vanish in this gauge , namely the first two ( involving second derivatives ) and those proportional to @xmath1019 in the second line . nevertheless , this expression for @xmath1020 is slightly more complicated than the one we used in section [ fieldeq ] above , eq .
( [ ricciharm ] ) , which does not depend explicitly on the spatial dimension @xmath1 .
it should be noted that many equations given in the book @xcite are erroneous when @xmath165 ( _ i.e. _ , when @xmath1021 , in this book s notation ) , including eq .
( i , 14 , 30 ) in @xcite which gives the ricci tensor in terms of the gothic metric .
let us end this appendix by displaying the full expansion of the geodesic equation ( [ eqgeod ] ) , or more precisely of the vectors @xmath1022 and @xmath1023 , quickly illustrated in eqs .
( [ smallexppifi ] ) .
the following expressions are @xmath1-dimensional generalizations of eqs .
( 3.35 ) of ref .
@xcite , and we keep the same writing and order of the terms to ease the comparison .
the `` linear momentum '' @xmath1022 reads @xmath1024\nonumber\\ & & + \frac{1}{c^6}\biggl [ \frac{5}{16}\ , v^6 v^i + \frac{3(5 d -4)}{8(d-2)}\ , v v^4 v^i -\frac{3}{2}\ , v_i v^4 -6 v_j v^i v^j v^2 + \frac{(3 d -2)^2}{4(d-2)^2}\ , v^2 v^2 v^i\nonumber\\ & & + 2\hat w_{ij } v^j v^2 + 2\hat w_{jk } v^i v^j v^k -\frac{2(2d-1)}{d-2}\ , v v_i v^2 -\frac{4(2d-1)}{d-2}\ , v v_j v^i v^j -4 \hat r_i v^2\nonumber\\ & & -8 \hat r_j v^i v^j + \frac{d^3}{6(d-2)^3}\ , v^3 v^i + \frac{4 d}{d-2}\ , v_j v_j v^i + \frac{4 d}{d-2}\ , \hat w_{ij } v v^j + \frac{4 d}{d-2}\ , \hat x v^i\nonumber\\ & & + 16 \hat z_{ij } v^j -2\,\frac{d(d-2)+2}{(d-2)^2}\ , v^2 v_i -8 \hat
w_{ij } v^j -\frac{8}{d-2 } v \hat r_i -16 \hat y_i\nonumber\\ & & -\frac{(3d-2)(d-3)}{(d-2)^2}\ , k v^2 v^i -\frac{2 d^2(d-3)}{(d-2)^3}\ , k v v^i + \frac{8(d-3)}{(d-2)^2}\ , k v_i \biggr ] + { \mathcal{o}}\left(\frac{1}{c^8}\right ) .
\label{exppi}\end{aligned}\]]@xmath176 this @xmath1-dimensional expression actually allows us to understand better some of the numerical coefficients found for @xmath42 in ref .
for instance , we find that a factor 33 comes from the expression @xmath1025 , and that a factor 49 comes from @xmath1026 .
the full post - newtonian expansion of the `` force '' @xmath1023 is given by an even longer formula : @xmath1027 \nonumber\\ & & + \frac{1}{c^4}\biggl [ \frac{3 d -2}{8(d-2)}\ , \partial_i v v^4 -2 \partial_i v_j v^j v^2 + \frac{d^2}{2(d-2)^2}\ , v\partial_i v v^2 + 2 \partial_i\hat w_{jk } v^j v^k \nonumber\\ & & -\frac{4}{d-2}\left(v_j\ , \partial_i v v^j + v\partial_i v_j v^j\right ) -8 \partial_i\hat r_j v^j + \frac{1}{2}\ , v^2 \partial_i v + 8 v_j\ , \partial_i v_j \nonumber\\ & & + 4\partial_i\hat x + 2\left(\frac{d-3}{d-2}\right)\left ( k \partial_i v + v \partial_i k\right ) -\frac{d(d-3)}{(d-2)^2}\ , \partial_i k v^2 \biggr ] \nonumber\\ & & + \frac{1}{c^6}\biggl [ \frac{1}{16}\left(\frac{5d-4}{d-2}\right)v^6\partial_i v -\frac{3}{2}\ , \partial_i v_j v^j v^4 + \frac{1}{8}\left(\frac{3d-2}{d-2}\right)^2 v\partial_i v v^4 + \partial_i\hat w_{jk } v^2 v^j v^k \nonumber\\ & & -2\left(\frac{2d-1}{d-2}\right ) v_j\ , \partial_i v v^2 v^j -2\left(\frac{2d-1}{d-2}\right ) v \partial_i v_j v^2 v^j -4 \partial_i
\hat r_j v^2 v^j \nonumber\\ & & + \frac{1}{4}\left(\frac{d}{d-2}\right)^3 v^2 \partial_i v v^2 + \frac{4d}{d-2}\ , v_j \partial_i v_j v^2 + \frac{2d}{d-2}\ , \hat w_{jk}\ , \partial_i v v^j v^k \nonumber\\ & & + \frac{2d}{d-2}\ , v \partial_i \hat w_{jk } v^j v^k + \frac{2d}{d-2}\ , \partial_i\hat x v^2 + 8\partial_i \hat z_{jk } v^j v^k -4\,\frac{d(d-2)+2}{(d-2)^2}\ , v_j v \partial_i v v^j \nonumber\\ & & -2\,\frac{d(d-2)+2}{(d-2)^2}\ , v^2 \partial_i v_j v^j -8 v_k \partial_i\hat w_{jk } v^j -8 \hat w_{jk } \partial_i v_k v^j -\frac{8}{d-2}\ , \hat r_j \partial_i v v^j \nonumber\\ & & -\frac{8}{d-2}\ , v\partial_i\hat r_j v^j -16 \partial_i \hat y_j v^j -\frac{1}{6}\ , v^3 \partial_i v -4 v_j v_j \partial_i v + 16\hat r_j \partial_i v_j + 16 v_j \partial_i\hat r_j \nonumber\\ & & -8 v v_j \partial_i v_j -4\hat x\partial_i v -4 v \partial_i\hat x + 16\partial_i\hat t -\frac{d^2(d-3)}{(d-2)^3}\ , k \partial_i v v^2 -2\left(\frac{d-3}{d-2}\right ) k v \partial_i v \nonumber\\ & & + \frac{8(d-3)}{(d-2)^2}\ , k \partial_i v_j v^j -\frac{(3d-2)(d-3)}{4(d-2)^2}\ ,
\partial_i k v^4 -\frac{d^2(d-3)}{(d-2)^3}\ , \partial_i k v v^2 \nonumber\\ & & -\frac{d-3}{d-2}\ , v^2\partial_i k + \frac{8(d-3)}{(d-2)^2}\ , \partial_i k v_j v^j -4\left(\frac{d-3}{d-2}\right)^2 k \partial_i k \biggr ] + { \mathcal{o}}\left(\frac{1}{c^8}\right ) . \label{expfi}\end{aligned}\]]@xmath176
this appendix is intended to provide a compendium of ( mostly well - known ) formulae for working in a space with @xmath1 dimensions . as usual , though we shall motivate some formulae below by writing some intermediate expressions which make complete sense only when @xmath1 is a strictly positive integer , our final formulae are to be interpreted , by complex analytic continuation , for a general complex dimension , @xmath1028 .
actually one of the main sources of the power of dimensional regularization is its ability to prove many results by invoking complex analytic continuation in @xmath1 .
we discuss first the volume of the sphere having @xmath580 dimensions ( _ i.e. _ , embedded into euclidean @xmath1-dimensional space ) .
we separate out the infinitesimal volume element in @xmath1 dimensions into radial and angular parts , @xmath1029 where @xmath1030 denotes the radial variable ( _ i.e. _ , the euclidean norm of @xmath1031 and @xmath578 is the infinitesimal solid angle sustained by the unit sphere with @xmath580 dimensional surface . to compute the volume of the sphere , @xmath1032 ,
one notices that the following @xmath1-dimensional integral can be computed both in cartesian coordinates , where it reduces simply to a gaussian integral , and also , using ( [ ddx ] ) , in spherical coordinates : @xmath1033 where @xmath470 in the last equation denotes the eulerian function .
this leads to the well known result @xmath1034 for instance one recovers the standard results @xmath1035 and @xmath1036 , but also @xmath1037 , which can be interpreted by remarking that the sphere with 0 dimension is actually made of two points .
if we parametrize the sphere @xmath581 in @xmath580 dimensions by means of @xmath580 spherical coordinates @xmath1038 , @xmath1039 , @xmath313 , which are such that the sphere @xmath1040 in @xmath1041 dimensions is then parametrized by @xmath1039 , @xmath1042 , @xmath313 , and so on for the lower - dimensional spheres , then we find that the differential volume elements on each of the successive spheres obey the recursive relation @xmath1043 note that this implies @xmath1044 which can also be checked directly by using the explicit expression ( [ omegad1 ] ) .
next we consider the dirac delta - function @xmath1045 in @xmath1 dimensions , which is formally defined , as in ordinary distribution theory @xcite , by the following linear form acting on the set @xmath384 of smooth functions @xmath1046 with compact support : @xmath1047 , @xmath1048 where the brackets refer to the action of a distribution on @xmath1049 .
let us now check that the function defined by @xmath1050 [ where @xmath1051 is the radial coordinate in @xmath1 dimensions , such that @xmath1052 is the `` green s function '' of the poisson operator , namely that it obeys the distributional equation @xmath1053 for any @xmath1054 we have @xmath1055 , thus we see that @xmath1056 in the sense of functions .
let us formally compute its value in the sense of distributions in @xmath241-space .
[ the usual verification of ( [ deltau ] ) is done in fourier space .
] we apply the distribution @xmath1057 on some test function @xmath1049 , use the definition of the distributional derivative to shift the laplace operator from @xmath1058 to @xmath1059 , compute the value of the @xmath1-dimensional integral by removing a ball of small radius @xmath289 surrounding the origin [ say @xmath1060 , apply the fact that @xmath1061 in the exterior of @xmath1062 , use the gauss theorem to transform the result into a surface integral , and finally compute that integral by inserting the taylor expansion of @xmath1059 around the origin .
the proof of eq .
( [ deltau ] ) is thus summarized in the following steps : @xmath1063 \nonumber\\ & = & \lim_{s\rightarrow 0}\int s^{d-1}d\omega_{d-1}(-n_i ) \left[u\partial_i\varphi-\partial_iu\,\varphi\right ] \nonumber\\ & = & \lim_{s\rightarrow 0}\int s^{d-1}d\omega_{d-1}(-n_i)\left[-\tilde{k}\,(2-d)s^{1-d}n_i\varphi ( \mathbf{0})\right ] \nonumber\\ & = & \omega_{d-1}\tilde{k}\,(2-d)\varphi(\mathbf{0 } ) \nonumber\\ & = & -4\pi\varphi(\mathbf{0})\ , .
\label{deltauphi}\end{aligned}\ ] ] in the last step we used the relation between @xmath469 and the volume of the sphere , which is @xmath1064 from @xmath1065 one can next find the solution @xmath1066 satisfying the equation @xmath1067 ( in a distributional sense ) , namely @xmath1068 from ( [ v ] ) we can then define a whole `` hierarchy '' of higher - order functions @xmath1069 , @xmath313 satisfying the poisson equations @xmath1070 , @xmath313 in a distributional sense .
however , the latter hierarchy of functions @xmath1058 , @xmath1066 , @xmath313 is better displayed using some different , more systematic notation .
this leads to the famous riesz kernels , here denoted @xmath1071 , in @xmath1-dimensional euclidean space @xcite .
[ these euclidean kernels differ from the minkowski kernels @xmath1072 , also introduced by riesz , and alluded to in the introduction . ]
these kernels depend on a complex parameter @xmath1054 .
they are defined by @xmath1073 for any @xmath1054 , and also for any @xmath1028 , the riesz kernels satisfy the recursive relations @xmath1074 furthermore , they obey also an interesting convolution relation , which reads simply , with the chosen normalization of the coefficients @xmath1075 , as @xmath1076 when @xmath1077 we recover the dirac distribution in @xmath1 dimensions , @xmath1078 ( the coefficient vanishes in this case , @xmath1079 ) , and we have @xmath1080 , @xmath1081 , @xmath313 . the convolution relation ( [ convolutiondeltaalpha ] ) is nothing but an elegant formulation of the riesz formula in @xmath1 dimensions . to check it
let us consider the fourier transform of @xmath1082 in @xmath1 dimensions , @xmath1083 using ( [ ddx ] ) we can rewrite it as @xmath1084 in which the angular integration can be performed as an application of eq .
( [ recursiondomega ] ) .
this yields an expression depending on the usual bessel function , ] to obtain eq .
( [ fourier ] ) we employ the integration formula @xmath1085 ] @xmath1086 where @xmath1087 . the radial integration in eq .
( [ radialang ] ) is then readily done from using the previous expression , and we obtain @xmath1088 where the factor in front of the power @xmath1089 , say @xmath1090 , is checked from the parseval theorem for the inverse fourier transform , which necessitates that @xmath1091 .
finally we can check the riesz formula by going to the fourier domain , using the previous relations .
the result , @xmath1092 is equivalent to eq .
( [ convolutiondeltaalpha ] ) .
a set of formulae concerning symmetric - trace - free ( stf ) multipole expansions in @xmath1 dimensions is presented next , without proofs .
we use the multi - index notation @xmath1093 ; more generally the notation is the same as in appendix a of @xcite . in particular
@xmath1094 denotes the stf projection of @xmath1095 , @xmath1096 $ ] means the integer part of @xmath1097 , @xmath1098 denotes the ( unnormalized , minimal ) sum of @xmath1099 where the @xmath143 s are permutations of the indices such that @xmath1098 is fully symmetric in @xmath399 ( for convenience we do not normalize the latter sum , for instance @xmath1100 ) .
@xmath1101}a_\ell^k\delta_{\{i_1i_2}\cdots \delta_{i_{2k-1}i_{2k}}\hat{n}_{l-2k\}}\,,\\ \hat{n}_l&=&\sum_{k=0}^{[\frac{\ell}{2 } ] } b_\ell^k\delta_{\{i_1i_2}\cdots \delta_{i_{2k-1}i_{2k}}n_{l-2k\}}\ , , \label{nl}\end{aligned}\ ] ] where the coefficients are @xmath1102 in particular ( the brackets @xmath1103 surrounding the indices mean the stf projection ) @xmath1104 spherical averages : @xmath1105 stf decomposition of a scalar function : @xmath1106 decomposition of a function @xmath1107 in terms of gegenbauer polynomials : is the coefficient of @xmath1108 in the expansion @xmath1109 the particular polynomial @xmath1110 represents an appropriate generalization of the legendre polynomial in @xmath1 dimensions [ indeed @xmath1111 . ]
@xmath1112 integration formulae : @xmath1113
a very important technical fact which allowed one to compute analytically the @xmath42 equations of motion is the possibility to obtain explicitly the quadratically non linear potentials , _
i.e. _ , to evaluate in closed form the integrals appearing in the pn expansion of the cubic - vertex diagram of fig .
[ fig7 ] . at the lowest approximation in the @xmath1114 expansion , the diagram of fig .
[ fig7 ] leads , in @xmath42 , to the integral @xmath1115 which was ( probably ) first evaluated by fock in 1939 ( `` sur le mouvement des masses finies daprs la thorie de la gravitation einsteinienne '' @xcite ) , with the simple result @xmath1116 remembering that @xmath1117 , @xmath1118 and @xmath1119 the combination @xmath1120 entering the logarithm in eq .
( [ eqd2 ] ) is simply seen to be the perimeter of the triangle joining the three spatial points @xmath241 , @xmath75 and @xmath1121 entering the ( newtonian approximation of the ) diagram of fig . [ fig7 ] . at the @xmath1122 level of the pn expansion of fig .
[ fig7 ] , there enter several new integrals which can be reduced to @xmath1123 together with @xmath1124 the explicit evaluation of the integrals ( [ eqd3 ] ) , ( [ eqd4 ] ) is also possible , as was shown in refs . @xcite,@xcite ( drawing on earlier works @xcite ) . in this appendix
we shall explicitly evaluate the @xmath1-dimensional generalization of ( [ eqd1 ] ) .
it will be clear , however , that our method can be rather straightforwardly generalized to the @xmath1122 diagrams contained in fig .
[ fig7 ] , _ i.e. _ , to the @xmath1-dimensional generalizations of ( [ eqd3])-([eqd4 ] ) . for our present purpose
it will be more convenient not to include the two factors of @xmath928 that accompany the two propagators issued from 1 and 2 in fig .
we shall therefore define @xmath1125 the method we present here consists of four basic steps : ( i ) expand the integrand in series and construct a corresponding series for a _
solution @xmath1126 , ( ii ) resum the series to get an explicit line - integral form of @xmath1127 , ( iii ) compute @xmath1128 in a distributional sense to discover that it satisfies @xmath1129 where @xmath1130 is a distributional source ( localized along a line ) , and finally ( iv ) subtract @xmath1131 ( which is given by another line - integral ) from @xmath1127 to get @xmath498 as a sum of line - integrals ( which are expressible in terms of one special function of one argument ) . what is crucial in the argument is the uniqueness of the global solution ( decaying at infinity ) of any ( distributional ) poisson equation @xmath1132 when the ( distributional ) source decays fast enough ( or , at least , does not grow too fast ) at infinity . in our case
, the sources @xmath143 involved will have fast - enough decay at infinity if we analytically continue @xmath1 toward large enough real parts ( say @xmath1133 > 3 $ ] ) .
there are several ways of implementing our method . for instance
, we could start by expanding @xmath1134 in the source of ( [ eqd5 ] ) in powers of @xmath37 , such an expansion being valid only in a neighborhood of @xmath75 .
namely , we have the @xmath1-dimensional generalization of the familiar @xmath42 legendre - polynomial expansion of @xmath1135 near @xmath255 ( more precisely in the ball @xmath1136 ) @xmath1137 here , we denoted for visual clarity @xmath1138 , where @xmath1139 is a gegenbauer polynomial such that @xmath1140 is the usual @xmath42 legendre polynomial [ see also appendix [ formulae ] above ] .
the quantity @xmath1141 in ( [ eqd6 ] ) denotes the cosine of the angle @xmath1142 between @xmath506 and @xmath1143 .
the notation we shall use is summarized in fig .
[ fig8 ] . when inserting the _ local _ expansion ( [ eqd6 ] ) into the source of ( [ eqd5 ] ) we are led to solving ( locally ) an equation of the form @xmath1144 .
however , using the general formula @xmath1145 we know a particular solution of @xmath1146 , namely @xmath1147 the formulae ( [ eqd7])-([eqd8 ] ) apply to any source with fixed multipolarity @xmath1148 and a power law dependence on a radius .
in particular , they apply when @xmath1149 , @xmath1150 and @xmath1151 ( because a generalized legendre polynomial is just proportional to the contraction of an stf - projected multi unit vector @xmath1152 onto a fixed `` @xmath1153 '' direction ; see appendix [ formulae ] ) .
this leads to a corresponding expansion of a _ local _ solution @xmath1154 ( near @xmath75 ) of @xmath1155 of the form @xmath1156 in order to proceed further , we now need to resum the expansion ( [ eqd9 ] ) .
this is done by a trick introduced , in a similar context of resummation of multipolar expansions containing extra @xmath363-dependent denominators , by ref .
one introduces some radial - integration operators @xmath1157 \ , ( \mathbf{r } ) = \int_0 ^ 1 d\lambda \ , \lambda^{\alpha } \ , \phi ( \lambda \ , \mathbf{r})$ ] .
for instance , in the context of ( [ eqd9 ] ) , one replaces @xmath1158 by @xmath1159 = \int_0 ^ 1 d\lambda \ , ( \lambda \ , r_1)^{\ell}$ ] or equivalently @xmath1160 by @xmath1161 .
this transforms back the multipolar series appearing in ( [ eqd9 ] ) into the original `` legendre '' series entering eq .
( [ eqd6 ] ) .
this allows one to write @xmath1154 as a simple line - integral : @xmath1162 here , @xmath1163 is a point on the segment joining @xmath75 to @xmath241 , located a distance @xmath1164 away from @xmath75 .
it is more convenient to replace the line - integration over the dimensionful length @xmath1164 ( @xmath1165 ) into an integration over the dimensionless parameter @xmath1166 ( @xmath1167 ) .
this leads to the explicit expression @xmath1168 the resummed line - integral expression ( [ eqd10 ] ) allows one to define everywhere @xmath1154 , _ including in the domain _
@xmath1169 where the original series ( [ eqd9 ] ) was _ not _ convergent .
having in hands such a global definition of @xmath1154 then allows one to compute its laplacian , _ in the sense of distributions _ , and to see how it differs from @xmath1170 .
the calculation of @xmath1171 is done by techniques similar to the ones used in ref .
one needs to rewrite some terms in the form of @xmath1172-derivatives .
for instance , several of the terms appearing in @xmath1171 can be rewritten as the line - integral @xmath1173 \nonumber \\ & & \qquad = \int_0 ^ 1 d\alpha \ , \frac{\partial}{\partial \alpha } \bigl [ \alpha \ , r_1^{2-d } \ , \vert\mathbf{y}_{\alpha } - \mathbf{y}_2 \vert^{2-d } \bigr ] = r_1^{2-d } \ , r_2^{2-d}\,,\end{aligned}\ ] ] where the last line - integral gave only the end contribution @xmath1174 corresponding to @xmath1175 . besides the terms yielding ( [ eqd12 ] ) , _ i.e. _ , the looked - for `` source '' of the complete @xmath498 , the calculation of @xmath1176 yields also the distributional source ( where @xmath1177 entered through @xmath1178 ) @xmath1179 this is conveniently transformed by introducing @xmath1180 ( with @xmath1181 ) and @xmath1182 which varies along a semi - infinite line going from @xmath76 to infinity along the direction @xmath1143 , _
i.e. _ , _ away _ from @xmath75 .
this transformation allows one to rewrite ( [ eqd13 ] ) in the more transparent form @xmath1183 at this stage , we recognize in ( [ eqd15 ] ) a very simple source , namely a _ uniform _
distribution of `` mass '' along the half - line along which @xmath1184 runs .
this allows one to easily compute the unique , global ( decaying at infinity ) solution of the poisson equation with source ( [ eqd15 ] ) and to subtract it from @xmath1185 to get the unique , global @xmath498 in the form of two line - integrals : @xmath1186 where @xmath1187 and @xmath1188 . in other words , ( [ eqd16 ] ) expresses @xmath498 as , essentially , the difference between the newtonian potentials generated by two uniform line distributions : a segment joining @xmath75 to @xmath241 and the half - line starting from @xmath76 in the direction away from @xmath75 .
it is easily seen ( modulo the slight delicacy of the logarithmic divergence of the potential of a semi - infinite line when @xmath32 , _ i.e. _ , the occurrence of a @xmath778 pole ; see below ) that the result ( [ eqd16 ] ) yields , when @xmath32 , the well - known result ( [ eqd2 ] ) .
[ actually , this was the way one of us ( t.d . ) had derived long ago for himself ( [ eqd2 ] ) , unaware of its derivations in the literature . ]
the expression ( [ eqd16 ] ) has the advantage of being explicitly regular ( except at the point @xmath255 ) in the ball @xmath1136 .
however , it has the default of treating dissymmetrically the two points @xmath75 and @xmath76 ( in spite of the fact that the result ( [ eqd16 ] ) for @xmath498 _ is _ , actually , symmetric under @xmath542 ) .
one can derive an exchange - symmetric expression for @xmath498 by modifying the first step of our method . instead of expanding the source @xmath1170 in the neighborhood of @xmath245 , _
i.e. _ , in a series of positive powers of @xmath37 , we can expand it in the _ neighborhood of infinity _ , _ i.e. _ , in a series of negative powers of @xmath37 .
such an expansion is directly related to the expansions used in @xcite , which led to the decomposition of @xmath1189 in two pieces denoted @xmath472 and @xmath1190 there , where the source of @xmath1190 was a uniform mass distribution along the segment joining @xmath75 and @xmath76 .
let us briefly indicate the successive steps of this new calculation of @xmath498 . instead of the `` local '' expansion ( [ eqd6 ] ) ( valid for @xmath1136 ) ,
one expands @xmath1134 near infinity ( @xmath1169 ) as @xmath1191 solving term by term @xmath1192 `` near infinity '' by means of ( [ eqd8 ] ) , and transforming away the appearing @xmath363-dependent denominators by means of @xmath1193 , leads to the following resummation of @xmath1194 : @xmath1195 here , @xmath1196 is still defined by ( [ eqd11 ] ) , but the parameter @xmath1172 now varies in @xmath1197 so that ( [ eqd18 ] ) is the potential of a semi - infinite line . computing the distributional laplacian of the particular solution @xmath1194 , eq .
( [ eqd18 ] ) , leads to the presence , besides the looked - for source @xmath1170 , of an additional distributional source localized now along the segment joining @xmath75 to @xmath76 , namely @xmath1198 where @xmath1199 varies between 0 and 1 and where @xmath1200 is again defined by ( [ eqd14 ] ) .
it is then easy to subtract from @xmath1194 ( which tends , in @xmath42 , to the function @xmath1201 of @xcite ) the poisson integral of the source ( [ eqd19 ] ) ( which is a uniform distribution along the segment @xmath75-@xmath76 and which tends , in @xmath42 , to the function @xmath1202 of @xcite ) to get the following alternative expression for @xmath498 , @xmath1203 or , equivalently , @xmath1204 the form ( [ eqd20])-([eqd20 ] ) still does not look quite symmetric between 1 and 2 but a moment of reflection will show that it is .
the two methods above have expressed @xmath498 in terms of the newtonian potentials generated by half - lines or segments , _
i.e. _ , integrals of the type @xmath1205 where @xmath1206 varies along a straight line ( but where @xmath74 might be @xmath241 or @xmath76 ) .
clearly , any such potential can be reduced ( through linear decompositions ) to the newtonian potential generated by a _
half - line_. let us then consider a generic half - line starting at the point @xmath1207 and going to infinity in the direction @xmath1208 , and let us consider the newtonian potential generated by this half - line at the origin of the coordinate system ( not located on the half - line ) .
let us denote @xmath1209 , @xmath1210 , @xmath1211 , @xmath1212 ( cosine of the angle @xmath11 between the radius vector from the origin , _ i.e. _ , the `` field point '' , towards the beginning point of the half - line and the direction of the half - line , away from its beginning ) .
then it is easy to find that @xmath1213 where the function @xmath1214 is given by the integral @xmath1215 the integral ( [ eqd22 ] ) converges for @xmath1216 , has a pole @xmath1217 as @xmath32 , and can be expressed in terms of hypergeometric functions , _ e.g. _ @xmath1218 $ ] .
it is , however , simpler to keep the form ( [ eqd22 ] ) .
reads @xmath1219 on this expression one sees clearly the occurrence of the simple pole of @xmath1214 when @xmath1220 , which is given by the `` monopolar '' term @xmath401 as @xmath1221 ] finally , using the half - line potentials ( [ eqd21 ] ) as building blocks one can write our result ( [ eqd20 ] ) in the final , @xmath292 symmetric , form @xmath1222 \nonumber \\ & - & \frac{r_1^{3-d } \ ,
r_2^{3-d}}{2(4-d ) } \ , \varphi ( c_{12})\,.\end{aligned}\ ] ] the quantities entering ( [ eqd23 ] ) are those defined in fig .
[ fig8 ] , notably @xmath1223 , @xmath1224 , @xmath1225 , with @xmath1226 being the orthogonal distance between the field point @xmath241 and the segment joining @xmath75 to @xmath76 [ with associated argument @xmath1227 in @xmath1214 ] .
note the following properties of the function @xmath1214 , the simplest way to prove ( [ eqd24 ] ) is to notice that the newtonian potential of an _ infinite _ line can be written either as twice that of two half - lines beginning at the orthogonal projection of the field point on the original line , so that @xmath1230 or as that of two other half - lines obtained by a more arbitrary cut ( under an angle @xmath1231 and @xmath1232 ) .
we can verify the @xmath32 limit of eqs .
( [ eqd16 ] ) and ( [ eqd23 ] ) by using the following @xmath104 expansion of the elementary function @xmath1214 , namely @xmath1233 to obtain ( [ phic ] ) we notice that the finite part of @xmath1214 when @xmath104 , which is @xmath1234\ , , \label{phi0c}\ ] ] can be re - expressed in the form of the following sum of two convergent integrals , @xmath1235 combining the expansion ( [ phic ] ) with the basic relations we have @xmath1236 the perpendicular distance @xmath1237 is given by @xmath1238 ] associated with the triangle of fig .
[ fig8 ] , our @xmath1-dimensional expressions ( [ eqd16 ] ) or ( [ eqd23 ] ) are found to admit the expansion @xmath1239 which indeed reduces to the three - dimensional result ( [ eqd2 ] ) modulo an additive constant linked to the @xmath778 pole .
nice as it is to have in hand an analytic expression for the @xmath1-dimensional basic non linear potential @xmath498 , its practical utility in explicit computations of the @xmath1-dimensional equations of motion is not evident because , contrary to the 3-dimensional expression ( [ eqd2 ] ) , the expression is not _
explicitly _ regular along the @xmath1240 segment .
[ the regularity of eq .
( [ eqd23 ] ) as @xmath1241 comes by compensations between the three terms in the bracket , using ( [ eqd24 ] ) . ]
it would need some transforming [ using ( [ eqd24 ] ) , and/or using the other expressions derived from the previous form ( [ eqd16 ] ) , which are regular along the @xmath1240 segment , but singular somewhere else ] to write an explicit expression which is regular everywhere , except at the two isolated points @xmath75 and @xmath76 .
finally , let us just mention that the method explained above can , in principle , be straightforwardly generalized to the computation of the higher post - newtonian potentials contained in the diagram of fig .
[ fig7 ] . for instance , in computing the @xmath1-dimensional analog of ( [ eqd3 ] ) , say @xmath1242 , it is easy [ by iterating ( [ eqd8 ] ) ] to get the analog of ( [ eqd9 ] ) . then a more complicated radial - integration operator ( see , _
e.g. _ , @xcite ) will allow one to resum the series to get a line - integral expression for @xmath1243 or @xmath1244 .
we anticipate that a somewhat more delicate application of either @xmath1245 ( to go back to @xmath498 ) or @xmath1246 ( to go back to @xmath1247 ) will yield additional line - distributed sources .
it should then be a simple matter to compute the poisson , or iterated poisson , integral of these line - distributed sources .
we leave an explicit study of these details to future work . | dimensional regularization is used to derive the equations of motion of two point masses in harmonic coordinates . at the third post - newtonian ( 3pn ) approximation
, it is found that the dimensionally regularized equations of motion contain a pole part [ proportional to @xmath0 which diverges as the space dimension @xmath1 tends to @xmath2 .
it is proven that the pole part can be renormalized away by introducing suitable shifts of the two world - lines representing the point masses , and that the same shifts renormalize away the pole part of the `` bulk '' metric tensor @xmath3 .
the ensuing , finite renormalized equations of motion are then found to belong to the general parametric equations of motion derived by an extended hadamard regularization method , and to uniquely determine the 3pn ambiguity parameter @xmath4 to be : @xmath5 .
this value is fully consistent with the recent determination of the equivalent 3pn `` static ambiguity '' parameter , @xmath6 , by a dimensional - regularization derivation of the hamiltonian in arnowitt - deser - misner coordinates .
our work provides a new , powerful check of the consistency of the dimensional regularization method within the context of the classical gravitational interaction of point particles . |
a large class of dilute magnetic semiconductors ( dms ) is based on manganese doping of iii - v or group - iv hosts having the zincblende , wurzite , or diamond structure .
formally , mn is either a single or double acceptor , depending on whether it substitutes for a group - iii or -iv atom , respectively.@xcite aside from this difference , there are no simple guidelines for predicting how the magnetic behavior of the resulting dms depends on the choice of host semiconductor . elucidating such guidelines would be of great practical interest for efforts to control and optimize the magnetic properties of dms materials . in particular
, it would be very helpful to understand how the resulting curie temperatures depend on the choice of host semiconductor .
an important and early contribution was made by , who used a mean - field solution of the zener model to predict the curie temperatures for mn doping of a wide range of ii - vi , iii - v , and group - iv host semiconductors.@xcite mn ions with localized spin @xmath0=5/2 were assumed to substitute on the cation site , and to interact by an indirect exchange interaction mediated by holes .
curie temperatures were calculated for fixed mn content @xmath1 and hole concentration @xmath2 @xmath3 using the mean - field result @xmath4 in this equation , the most direct dependence of @xmath5 on the host semiconductor comes from the density per unit volume of cation sites @xmath6 , which varies with the host lattice constant as @xmath7 . a much weaker dependence on the host arises from the density - of - states effective mass @xmath8 and the fermi wavevector @xmath9 .
the remaining variables , i.e. the @xmath10-@xmath11 exchange integral @xmath12 and the fermi - liquid parameter @xmath13 , were assumed not to depend on the host . the full numerical evaluation of eq .
( 1 ) for different host semiconductors@xcite leads to values of @xmath5 that indeed track the prefactor @xmath7 quite closely.@xcite for example , the hosts with the largest lattice constants ( cdte , inas , ge ) lead to the smallest predicted curie temperatures ( below @xmath1475 k ) .
likewise , the hosts with the smallest lattice constants ( zno , gan , c ) lead to the largest predicted values of @xmath5 ( well above room temperature ) . diamond has the smallest lattice constant of all the semiconductors ( @xmath15 ) and , accordingly , the highest predicted curie temperature ( @xmath16 470 k ) .
these extreme values make diamond a logical choice for examining some of the assumptions underlying the predictions of high curie temperatures for hosts having small lattice constants . in this paper
, we use first - principles theoretical methods to test the assumption that mn occupies the substitutional site , with spin @xmath17 , in the diamond lattice .
we find that substitutional mn has a low - spin @xmath0=1/2 ground state in diamond .
moreover , because of diamond s small lattice constant , the substitutional site is energetically unfavorable relative to the much larger `` divacancy site , '' denoted @xmath18 , which we identify as the more likely mn impurity location .
we analyze the electronic and magnetic structure of this divacancy mn impurity in diamond , and show that it is unlikely to lead to magnetic ordering at any reasonable temperature .
most of the results reported in this article are for isolated mn impurities , which we simulate using a supercell of 64 carbon atoms in the diamond structure .
all atomic positions were relaxed using total energies and forces calculated within the generalized - gradient approximation to density - functional theory ( dft ) , as implemented in vasp.@xcite carbon and mn ultrasoft pseudopotentials were used with a cutoff energy of 286 ev . for total energies , we used 2@xmath192@xmath192 monkhorst - pack sampling of the brillouin zone , and 4@xmath194@xmath194 sampling for convergence checks . in the discussion of the electronic structure of the isolated mn impurity at a divacancy site , eigenvalues at the zone center of a 128-atom supercell
are reported in order that degeneracies are properly represented . within the supercell formalism , the formation energy of a mn impurity is given by @xmath20 = e_t[{\rm mn}^q ] - n_{\rm c } \mu_{\rm c } - \mu_{\rm mn } + q e_f,\ ] ] where @xmath21 $ ] is the total energy of a supercell containing @xmath22 carbon atoms and one mn impurity , with chemical potentials @xmath23 and @xmath24 , @xmath25 is the charge state of the mn impurity , and @xmath26 is the fermi level . since the host crystal is elemental , the chemical potentials are simply the energy per atom in the diamond phase of carbon and the ground - state @xmath27mn .
the electrical activity of a mn impurity is determined by its formation energy as a function of charge state .
electrically active defects will have more than one stable charge state within the host band gap ; the value of @xmath26 for which two such charge states have equal formation energies is referred to as a `` charge transition level . '' the charge transition level between the neutral and @xmath28 states is the donor ionization energy , while the transition level between neutral and @xmath29 states is the acceptor ionization energy . in practice
, the total energy of a charged supercell must be calculated by adding a uniform compensating background charge to the supercell .
the long - range nature of the coulomb interaction then gives rise to a spurious madelung - like contribution to the total energy , which must be subtracted from the calculated total energy .
this was done using a standard approach in which a multipole expansion of the defect charge ( up to quadrupole order ) was used to estimate the interaction energy analytically and then subtract it off.@xcite the magnetic interaction between mn impurities in nearby divacancy sites was calculated using larger supercells of 128 carbon atoms .
no relaxations were performed for these calculations ; tests for a few configurations confirmed that this approximation did not change the results significantly .
the interactions are represented by numerically mapping the dft total energies into the heisenberg form . in practice
this amounts to computing the difference in total energy between the parallel and antiparallel spin alignment of two mn impurities in a supercell .
the mn impurity enjoys a special status as a dopant in dilute magnetic semiconductors such as mn@xmath30ga@xmath31as and mn@xmath30ge@xmath31 .
it preferentially occupies the substitutional ( cation ) site , where it serves two roles : it is electrically active as an acceptor , and it contributes a localized spin @xmath17 . at sufficient mn concentrations ,
holes are created in the host valence band ; these delocalized carriers mediate the ferromagnetic interaction between the localized spins.@xcite gaas and ge have similar lattice constants , and since strain contributes to the impurity formation energy , it is instructive to note that the local strain around a substitutional mn impurity , denoted mn@xmath32 , is very small for both hosts the nearest neighbors distort by less than 2% from their ideal positions .
tetrahedrally coordinated interstitial mn impurities , denoted mn@xmath33 , have higher formation energies ( by 12 ev for neutral impurities ) ; interestingly , the atomic volume available at this site is the same as for the substitutional site , and consequently very little distortion is created by the mn interstitial in either host.@xcite the lattice constant of diamond is 37% smaller than that of ge , and so it is not surprising that a substitutional mn impurity creates a much larger local strain in the lattice within dft
the nearest - neighbor distortion is 12% . because of the extreme stiffness of diamond
, such a large strain is expected to lead to a relatively high formation energy .
a similar distortion occurs for mn in the tetrahedral interstitial site .
a simple way to relieve this strain is to place the mn impurity at a divacancy site i.e . to remove two neighboring carbon atoms and place the mn at their midpoint , as shown in fig .
this doubles the atomic volume available , leading to a much reduced local strain
only 2% distortion of the nearest neighbors . even in its ground - state configuration , which includes a slight off - center relaxation of the mn ion as shown in fig .
1(b ) , the local lattice distortion around this mn@xmath34 impurity is very small .
it is already widely believed that other large impurities including si , p , ni , and co preferentially occupy the divacancy site in diamond;@xcite hence our prediction that mn also prefers this site appears plausible . to make the comparison between different sites quantitative we turn now to the formation energies calculated within dft . for neutral mn impurities ,
we find the divacancy site is strongly preferred ( by 0.6 ev ) to the substitutional site ; the interstitial site is much less favorable than either ( by @xmath1410 ev ) .
figure 2 shows the formation energies for divacancy and substitutional mn as a function of magnetic moment , calculated using the fixed - spin - moment method within dft . even in the preferred divacancy site ,
the absolute formation is almost an order of magnitude larger for substitutional mn in gaas or ge , suggesting that standard methods for attaining high concentrations of magnetic dopants will probably not work for diamond .
the equilibrium magnetic moments @xmath35 in the substitutional and divacancy sites are also quite different compared to substitutional mn in gaas and ge ( @xmath36 and 3 @xmath37 , respectively ) . in the substitutional site , mn adopts a low - spin @xmath38 state , with an extremely small energy gain ( 0.1 ev ) relative to the spin - unpolarized state .
hence , we anticipate that even if mn could be forced into the substitutional site , its magnetic moment may not be sufficiently stable to facilitate magnetic ordering . in the divacancy site
two nearly degenerate moments are found ( @xmath38 and 5 @xmath37 ) with an energy difference of only 0.05 ev ; the formation energy vs. magnetic moment is symmetrical about @xmath39 3 @xmath37 .
this peculiar feature of the mn@xmath34 impurity will be explained in section iv .
the mn@xmath34 impurity has an energy - level structure that arises from the relatively weak interaction of atomic mn levels and host divacancy levels , as shown in fig .
the host divacancy defect itself has been extensively studied both theoretically and experimentally , and is well understood.@xcite the removal of two carbon atoms from the host lattice leads to a defect with @xmath40 symmetry , and thus the six available dangling - bond orbitals must form linear combinations that transform under the irreducible representations @xmath41 , @xmath42 , @xmath43 , and @xmath44.@xcite for the neutral divacancy , six electrons are available to fill these levels . in an early theoretical study , coulson and larkins@xcite predicted the ground state one - electron configuration
to be @xmath45 .
later calculations confirmed this one - electron level ordering for both silicon and diamond , and we find it here too , as shown . in fig .
3 we consider separately the interaction of the majority and minority mn levels with the four divacancy levels . under the @xmath40 crystal field ,
the 3@xmath11 level of mn splits into one @xmath41 and two @xmath44 levels .
hence , hybridization is allowed only between mn and @xmath18 levels both having either @xmath41 or @xmath44 symmetry , but no mixing is allowed between mn levels and @xmath18 levels having either @xmath42 or @xmath43 symmetry . among the calculated majority mn@xmath46 levels we find very little mixing within any level , and therefore each has a clear parentage from either mn or @xmath18 orbitals , as indicated in fig .
3 . for the minority mn@xmath47 levels
there is strong mixing between the empty @xmath18 @xmath44 level and the empty mn @xmath48 level , so that the occupied bonding combination falls well below the diamond valence - band maximum ( vbm ) . considering for the moment
only those levels that sit below the vbm , we find that seven electrons occupy majority levels and four electrons occupy minority levels .
hence , these low - lying levels carry a net `` core '' magnetic moment @xmath49 .
the highest occupied molecular orbital ( homo ) of the mn@xmath34 impurity has @xmath43 symmetry and , because it can not mix with mn states , is purely derived from the @xmath18 dangling - bond orbitals .
the two electrons in this fourfold level form a spin triplet ( @xmath50 1 @xmath37 ) as shown , but the system is energetically indifferent to whether they are aligned parallel or antiparallel to the core spin ( fig . 3 shows the parallel alignment ; when the alignment is made antiparallel the spin - up and spin - down homo levels simply switch positions ) . because of this weak coupling between the homo spin and the core spin , the total magnetic moment of mn@xmath34 can assume two possible low - energy values : either @xmath51 5 @xmath37 or @xmath52 1 @xmath37 for parallel or antiparallel alignment , respectively .
moreover , there is an energy penalty for violating hund s rule by forcing the homo into a spin - singlet state ; this is the origin of the energy increase that occurs when the total moment is constrained to 3 @xmath37 ( the core moment alone ) .
these two features the weak core - homo spin coupling , and the hund s rule penalty for constraining the homo to a spin singlet explain the unusual appearance , shown earlier in fig . 2 , of the mn@xmath34 formation energy vs. total magnetic moment .
4 we show the calculated formation energies of mn@xmath34 for charge states that are stable within the experimental 5.4-ev band gap of diamond . there are three such stable charge states ( @xmath53 ) , making mn@xmath34 formally a double acceptor . in principle
this is similar to the case of substitutional mn in ge , which is also a double acceptor , and might suggest similar electrical activity for mn@xmath34 .
quantitatively , however , the two cases are quite different .
the first two acceptor ionization energies for mn in ge have been measured experimentally to be 160 and 370 mev , relative to the valence - band edge.@xcite for mn concentrations of several percent , the width of the resulting impurity band is sufficient to create overlap with the valence band and thus to allow the holes to delocalize .
likewise , mn in gaas is a single acceptor with a measured acceptor level at 113 mev,@xcite which is accurately given by dft calculations as 100 mev .
@xcite on the other hand , the predicted first acceptor ionization energy for mn@xmath34 in diamond is an order of magnitude larger , @xmath54 1.2 ev , suggesting that the holes will remain localized on the impurity sites for any reasonable mn concentration .
for this reason alone , it appears very unlikely that there will be sufficient carriers to mediate a ferromagnetic interaction between the localized spins of mn@xmath34 impurities .
the weak coupling between core and homo spins in the isolated mn@xmath34 impurity suggests that the effective coupling between spins of two impurities will also be weak .
this expectation is confirmed in fig . 5 , where we show the results of our dft calculations for ( twice ) the heisenberg coupling constant between two mn@xmath34 impurities in a 128-atom diamond supercell .
for comparison , we show the results of similar calculations for substitutional mn in ge and si .
the results for substitutional mn in ge and si show a similar trend . when the two mn are nearest neighbors , there is a strong preference for antiparallel spin alignment . at larger separations
the interactions vary in sign ( ge ) and/or magnitude ( si ) , a consequence of the crystallographic anisotropy that arises from the strong @xmath11 character of the acceptor wavefunction in substitutional mn .
the largest ferromagnetic coupling strengths are in the range [email protected] ev .
( for si there is a shift of all the couplings toward more favorable ferromagnetic interactions ; since mn - doped ge is known to be ferromagnetic this suggests that mn - doped si might also be perhaps with a higher curie temperature ) .
the coupling between two mn@xmath34 impurities in diamond is substantially weaker : the largest ferromagnetic couplings are of order 1 mev , two orders of magnitude smaller than for mn in ge or si .
hence , any ferromagnetically ordered phase involving mn@xmath34 impurities would be characterized by a curie temperature at most on this scale , i.e. of order 10 k.
we have shown theoretically that the mn impurity in diamond is energetically more favorable in the divacancy site than in the substitutional site .
the magnetic properties of the divacancy mn impurity show nearly degenerate low - spin and high - spin configurations ; the degeneracy arises from the weak coupling of separate core and valence moments .
divacancy mn is predicted to be a double acceptor with two hyperdeep levels , suggesting poor prospects for creating the delocalized holes in the valence band that are required to mediate a ferromagnetic interaction within the zener model
. explicit calculation of the effective heisenberg coupling between mn in nearby divacancy sites confirms that mn - doped diamond shows little promise of ferromagnetism above a few kelvins
. our findings may not be unique to diamond : other semiconductors with predicted high curie temperatures have similarly small atomic volumes , which may likewise make the substitutional site unfavorable for mn doping .
comprehensive calculations of the site energetics for gan , zno , and other host semiconductors would be of great interest .
this work was supported by the office of naval research and the darpa spins in semiconductors program .
computations were performed at the dod major shared resource centers at asc and navo . | we use density - functional theoretical methods to examine the recent prediction , based on a mean - field solution of the zener model , that diamond doped by mn ( with spin @xmath0=5/2 ) would be a dilute magnetic semiconductor that remains ferromagnetic well above room temperature .
our findings suggest this to be unlikely , for four reasons : ( 1 ) substitutional mn in diamond has a low - spin @xmath0=1/2 ground state ; ( 2 ) the substitutional site is energetically unfavorable relative to the much larger `` divacancy '' site ; 3 ) mn in the divacancy site is an acceptor , but with only hyperdeep levels , and hence the holes are likely to remain localized ; ( 4 ) the calculated heisenberg couplings between mn in nearby divacancy sites are two orders of magnitude smaller than for substitutional mn in germanium . |
most formation scenarios of globular clusters commence with giant molecular clouds ( gmcs ) undergoing a phase of rapid star formation .
this star formation can be triggered by several processes , e.g. a thermal instability ( fall & rees 1985 ; murray & lin 1990 ) , a radiative shock ( kang et al . 1990 ; shapiro 1993 ) or other perturbations like collisions of clouds ( fujimoto & kumai 1997 ; lee , schramm , & mathews 1995 ) or galaxy interactions ( ashman & zepf 1992 ) .
a common characteristic of all these scenarios is , that globulars are formed from smooth gaseous distributions ( if we neglect the clumpy structure of the gmcs for the moment ) which are transformed into stars .
this assumption leads to several difficulties : first , the gravitational binding energy of a homogeneous gmc with @xmath0 and a radius of 30 pc is about @xmath1 ergs , whereas it decreases for a @xmath2 cloud of 10 pc to less than @xmath3 ergs .
thus , already a single supernova injects sufficient energy to destroy a small cloud completely , and a few ob stars can even disrupt a @xmath0 cloud .
therefore , the formation of the globular must have been finished within a few myrs , before the first ob stars explode .
a second problem is related to the star formation efficiency ( sfe ) .
assuming that the gmc is in virial equilibrium prior to the cluster formation and that the newly born stars keep the velocity of their parent gas packages , the total energy @xmath4 of the stellar system can be estimated by @xmath5 ( with the gmc mass @xmath6 , its radius @xmath7 and the gravitational constant @xmath8 ) .
thus , the system is only gravitationally bound ( i.e. @xmath4 becomes negative ) , if the sfe @xmath9 , which is the mass fraction of the gmc transformed into stars , is larger than 50% .
though mass redistribution in a violent relaxation stage and a detailed treatment of the energy injection can reduce the critical level down to 20% ( goodwin 1997 ) , the required sfe still exceeds the typical observed values for gmcs by at least one order of magnitude ( blitz 1993 ) .
with respect to these problems , an alternative suggestion by brown , burkert , & truran ( 1991 ) is very interesting : they suggest that cluster formation starts with an ob - association undergoing typeii supernova events near the center of a molecular cloud .
the expanding supernova remnant sweeps up the cloud material , decelerates and might almost be stopped by the external pressure of the ambient hot gas .
meanwhile the shell breaks into fragments and forms stars .
if the total energy of this stellar shell is negative , the stars will recollapse and form a bound system . for a simplified spherical configuration burkert , brown , & truran ( 1993 ) demonstrated that the binding energy of an isolated shell always becomes negative _ independent _ of the sfe , provided the star formation process does not start too early .
thus , the sfe efficiency problem is less severe for this scenario . in case of an homogeneous ambient medium
( e.g. in the core of a gmc ) ehlerov et al .
( 1997 ) demonstrated that fragmentation in an expanding shell takes sufficient time to prevent too rapid star formation .
this result holds also for non - homogeneous power - law density profiles , if the density distribution in the gmc is not steeper than isothermal ( theis et al .
1998 ) .
according to the shell - scenario the dynamics of its last stage , i.e. the collapse of a thin stellar shell , is studied and compared with the collapse of homogeneous spheres representing the standard scenarios .
n - body simulations are performed for isolated configurations as well as for collapses within a galactic tidal field .
the main question addressed here is , whether we can discern between different formation scenarios by means of their collapse dynamics .
* initial conditions and numerical scheme . * the shell is modelled by a unit mass which is homogeneously distributed within the radial range [ 0.9,1.0 ] giving a shell thickness of 10% of the shell s radius @xmath7 .
initially the shell is at rest .
the individual velocities of the equal mass particles are chosen from an isotropic velocity distribution resulting in a virial coefficient @xmath10 ( @xmath11 is the kinetic and @xmath12 the potential energy . ) .
the simulations are performed with @xmath13 particles adopting a softening length @xmath14 of 0.01 .
the equations of motion are integrated with a leap - frog scheme using a fixed timestep @xmath15 of @xmath16 .
this gives an energy conservation of typically 0.1 - 0.2% or better over the whole integration time .
the simulations were performed either with a direct summation code ( using a grape3af board ) or a tree - code .
* results . *
the dynamics of the shell shows three stages . during the first stage ( @xmath17 ) the shell is slowly
contracting and small inhomogeneities start to grow ( cf .
also upper left diagram in fig .
1 ; @xmath18^{1/2}$ ] is the free - fall time corresponding to the mass density @xmath19 ) .
already during this early stage the particles in the shell are strongly mixed because of the radially _ decreasing _ ( global ) free - fall time in the shell . in the second phase ( @xmath20 ) the inhomogeneities become bound clumps .
they merge after the shell s free - fall time @xmath21 which exceeds the free - fall time of the corresponding sphere by 40% .
finally , a radially anisotropic triaxial system is formed .
the final configuration of the shell simulations is characterized by a more flattened shape , a mass - loss of only 8% ( i.e. a reduction by a factor of 3.5 ) , a smaller anisotropy @xmath22 , an increased half - mass radius ( by 50% ) and a decreased 90% lagrange radius ( factor of 6 ) compared to the corresponding collapse of a sphere .
thus , violent relaxation is less efficient in case of collapsing shells .
a realistic model should also include the galactic tidal field .
therefore , the gravitational force of a static isothermal halo with a circular velocity of 220 kms@xmath23 is added to the force derived from the self - gravity of the n - body configurations . in all following models
the galactic orbits have an apogalacticon of 5 kpc and the clusters start with a mass of @xmath2 and a size of 30 pc . at apogalacticon
this corresponds to a tidal radius of 42 pc , i.e. systems on circular orbits are not expected to be tidally disrupted .
* circular orbit .
* though circular orbits are not very realistic for globular clusters , they keep the tidal field almost time - independent which allows a more direct investigation of the influence of the host galaxy .
1 shows the evolution of the system projected onto the orbital plane ( and normalized to the center of mass of the particles in the collapsing system ) for both , a shell and a sphere . similarly to the isolated case , we find a strong fragmentation prior to the collapse of the whole system in case of the shell , whereas almost no substructure is seen during the collapse of the sphere .
lateron , the sphere forms a single bound object which shows immediately after the collapse some elongation caused by the tidal field .
after one revolution around the galactic center , however , the system is almost spherical . in case of the shell ,
the influence of the tidal field is already obvious in the distortion prior to the collapse .
the delayed collapse , the fragmentation and also the reduced violent relaxation leads to a less dense system which is much stronger affected by the galactic tidal field : the system does not merge into a single object , but into two ( twin ) stellar systems of almost identical total mass and density distribution .
however , in their kinematical properties both objects differ : one of the clusters shows a counter - rotating core which is neither found in the second twin nor in the product of the collapsing sphere
. * eccentric orbit . * on eccentric orbits the evolution additionally depends on the initial phase ( fig . 2 ) .
a collapse of a shell starting at apogalacticon ends up again in a twin cluster ( a in fig . 2 ) .
if the phase is chosen such that maximum compression is reached at perigalacticon ( b ) , the system breaks up into many small systems , but no large cluster is formed .
when the collapse starts at perigalacticon even the formation of these small clusters is prevented ( c ) .
if the collapse starts well after perigalacticon a multiple system of clusters ( with two larger ones ) is built up ( d ) .
the final fate of the twins is less clear : the simulations show preferably surviving twins which separate after a few orbital revolutions to distances of several 100 pc up to kpc scale or more .
however , also merging twins have been found .
the low number statistics of the simulations performed so far does not allow to give more accurate rates here .
contrary to the shells the spheres do not form any multiple clusters : either the systems are completely destroyed or single bound objects are formed .
in case of isolated evolution the simulations show several systematic differences between shells and spheres : the collapse of the shell is delayed , the resulting configurations are flatter and ( more ) triaxial , the mass - loss is strongly reduced and the system is less dense . all of these effects suggest that _ violent relaxation _ is less efficient for collapsing shells . another difference is the strongly pronounced fragmentation of the shell which is not seen during the collapse of the sphere .
unfortunately , all of these features seem to be very difficult to be used for an observational discrimination between the different formation scenarios .
e.g. the easy accessible flattening of globular clusters is not only strongly influenced by the formation scenario , but also by the secular evolution , the galactic tidal field , and by initial conditions like the mass distribution in the parent gmc and the virial ratio of the collapsing stellar system .
much more promising are the simulations including a galactic tidal field : they demonstrated that shells tend to form multiple stellar systems , preferably twin ( but not binary ) clusters . in case
these twins are not destroyed nor merged , one should be able to find pairs of globular clusters which are characterized by the properties of their common birth place , i.e. identical metallicities and evolutionary stage or similar orbital characteristics .
another observational feature is the rotation profile of globulars .
the merging of fragments in the shell scenario would give a natural explanation for counter - rotation in stellar clusters similar to the counter - rotation found in n - body simulations of merging galaxies .
ashman , k.m .
, zepf , s.e .
1992 , , 384 , 50 blitz , l. 1993 , in protostars and planets iii , e.h . levy & j.i .
lunine ( eds . ) , 125 brown , j.h . , burkert , a. , & truran , j.w .
1991 , , 376 , 115 burkert , a. , brown , j. , & truran , j.w .
1993 , in the globular cluster - galaxy connection , asp conf .
48 , graeme h. smith , jean p. brodie ( eds . ) , 656 ehlerov , s. , palou , j. , theis , ch . , & hensler , g. 1997 , , 328 , 121 fall , s.m . , rees , m.j .
1985 , , 298 , 18 fujimoto , m. , kumai , y. 1997 , , 113 , 249 goodwin , s. 1997 , , 284 , 785 kang , h. , shapiro , p.r . , fall , m. , & rees , m.j .
1990 , , 363 , 488 lee , s. , schramm , d.n .
, & mathews , g.j .
1995 , , 449 , 616 murray , s.d . ,
lin , d.n.c .
1990 , , 363 , 50 shapiro , p.r .
1993 , in the globular cluster - galaxy connection , asp conf .
48 , graeme h. smith , jean p. brodie ( eds . ) , 664 theis , ch . , ehlerov , s. , palou , j. , & hensler , g. 1998 , in the local bubble and beyond , iau coll . 166 , d. breitschwerdt , m.j .
freyberg , j. trmper ( eds . ) , 409 \1 ) the velocity distributions of both , spheres and shells , are isotropic in the central regions . in the outer region a radial anisotropy ( defined here as @xmath22 ) evolves reaching about 0.8 in case of the shell .
it is systematically below the anisotropy of the sphere by about 0.15 .
2 ) the animations are available via my home - page http://www.astrophysik.uni-kiel.de/pershome/theis \1 ) the initial virial coefficient here is close to , but exceeding the limit for the onset of the radial orbit instability for my simulations .
starting from a lower virial ratio of e.g. 0.02 shows the roi for the collapsing spheres as expected .
+ 2 ) the fragmentation depends partly on the number of particles and and also on the initial setup of the particle configuration ( here a random realization , i.e. white noise ) .
the total number of particles was @xmath25 .
however , the most important parameters for the onset of fragmentation are the velocity dispersion in the shell and its thickness .
e.g. a thickness of 50% or an initial virial ratio of 0.2 strongly suppresses fragmentation . | though it is generally assumed that massive molecular clouds are the progenitors of globular clusters , their detailed formation mechanism is still unclear .
standard scenarios based on the collapse of a smooth matter distribution suffer from strong requirements with respect to cluster formation time scale , binding energy and star formation efficiency . an alternative model assuming cluster formation due to the recollapse of
a supernova - induced , fragmented shell can relax these difficulties . in this paper the final collapse stages of the different scenarios
are compared by n - body simulations for shells and spheres .
it is shown that fragmentation is much more pronounced for shells . taking a galactic tidal field into account shells preferably
form twin ( or multiple ) systems , whereas spheres end up as single clusters .
the twins are characterized by identical metallicities , and stellar mass functions ; some of them show counter - rotating cores .
their orbital evolution can result in both , a final merger or well separated twins sharing a common galactic orbit . |
fig . [ fig1 ] .
comparison of experimental and model results for coherence length variation with observation angle for ( a ) a tuned rcled and ( b ) a detuned rcled . fig . [ fig2 ] .
coherence length variation as a function of numerical aperture generated using empirical model data in the coherence model for tuned and detuned rcleds .
markers at na extrema denote experimental points determine by measurements at low na and from an integrating sphere .
predictions of maximum and minimum coherence as a function of microcavity finesse and detuning with respect to the emission source across the useful emission na ( see text ) .
the diagram in the top left corner shows highlights the definition of useful na . | an analytical expression for the self coherence function of a microcavity and partially coherent source is derived from first principles in terms of the component self coherence functions .
excellent agreement between the model and experimental measurements of two resonant cavity leds ( rcleds ) is evident .
the variation of coherence length as a function of numerical aperture is also described by the model .
this is explained by a microcavity s angular sensitivity in filtering out statistical fluctuations of the underlying light source .
it is further demonstrated that the variable coherence properties of planar microcavities can be designed by controlling the underlying coherences of microcavity and emitter whereby coherence lengths ranging over nearly an order of magnitude could be achieved .
the last two decades have seen widespread use of optical microcavities , both for experimental physics and commercial applications .
microcavities redistribute emission from an underlying source and depending on the ensuing radiation pattern allow light collection for use elsewhere . commercially available microcavity devices such as
resonant cavity light emitting diodes ( rcleds ) use the planar microcavity geometry to increase the extraction efficiency of spontaneous emission from materials with high dielectric constants .
@xcite more recently microcavities have been used to spectrally and spatially isolate quantum dot emitters to increase the efficiency of single photon production .
@xcite recent work on planar microcavities has identified the dependence of numerical aperture ( na ) on emission properties such as spectral linewidth @xcite and coherence length,@xcite the latter of which is the focus of the following paper . note that , migration of these results to the spectral domain are trivial due to the implicit fourier relationship with the coherence domain .
in addition to the choice of domain , the name _ coherence _ has been used , instead of _
time_. coherence highlights the statistical properties of light and not necessarily to the time dependence of the emission process that produces the light .
this distinction between the light and the emission event that produced it is necessary for the descriptions presented in this letter .
the coherence length is a relevant attribute of any optical device as it defines the length scales over which mutual interference occurs .
applications such as low coherence interferometry , for non invasive medical imaging @xcite ( also known as optical coherence tomography ) and optical time domain reflectometry ( otdr ) , for ranging measurements in optical components @xcite and surface mapping in integrated circuits @xcite , rely on the coherence of source being both small enough to eliminate coherent reflections from distant objects and large enough to examine detail on the relevant length scale .
the correlations of low coherence sources are also applicable to schemes of all - optical routing.@xcite the variable coherence properties of rcleds @xcite make them ideal as a light source for applications where a range of length scales need to be analysed where previously , multiple light sources would have been required .
this letter examines the physics of variable emission coherence from planar microcavities and indicates methods by which this aspect of commercial devices such as rcleds could be engineered .
although experimental results of rcleds are used in this paper , the results are applicable to planar microcavities in general .
the coherence domain picture of microcavity emission requires the definition of the self coherence properties of a light source .
the self coherence function of an emitter , @xmath0 , defined in eqn .
( [ eqn0 ] ) is synonymous with the spectral distribution function by the wienner - kinchin theorem .
@xmath1 a derivation for @xmath2 , the self coherence function of a combined emitter and microcavity system , is too involved to be described here from first principles . instead
, a key result from the spectral domain is used in combination with the wienner - kinchin theorem . the emission intensity from an emitter microcavity system is given by the spectral overlap of the optical transfer function of the bare cavity , @xmath3 , which is a function of the angle relative to the emission region , @xmath4 , and the field distribution of the underlying source @xmath5 , which is assumed to be isotropic .
@xcite @xmath6 therefore the coherence function for the microcavity can be written as a convolution expression : @xmath7 @xmath8 is the maximum microcavity enhancement for an emitter tuned to the cavity resonance at an angle @xmath9 , @xmath10 here @xmath11 and @xmath12 are the transmission and reflectivity of the front cavity mirror and @xmath13 is the reflectivity of the back mirror ; all function of @xmath4 , the internal emission angle .
@xmath2 can be greatly simplified by evaluating the definite integral . to do this ,
the underlying spectral distribution of emission is assumed to be lorentzian in nature corresponding to exponential decay in the time domain .
the transfer function of the cavity , @xmath3 can also be approximated by lorentzians in the immediate vicinity of the cavity modes , @xmath14 as the free spectral range of longitudinal modes in microcavities is usually much greater than the spectral width of the source .
@xmath15 here @xmath16 and @xmath14 are the frequencies of the emission and cavity resonances respectively and @xmath17 and @xmath18 are the coherence times of the underlying source and cavity resonances respectively .
note that @xmath17 and @xmath16 are assumed to be constant with emission angle . since @xmath19 is not necessarily symmetric ,
consider the expansion of the integral in eqn .
( [ eqn9 ] ) for @xmath20 : @xmath21 here , @xmath22 and @xmath23 . evaluating the integrals of eqn .
( [ eqn10 ] ) gives the self coherence function for the microcavity at an internal angle @xmath4 , @xmath24 the real and imaginary parts of this expression are evaluated to be : @xmath25 \right\ } \end{aligned}\ ] ] the expansion of the integral in eqn .
( [ eqn9 ] ) for @xmath26 gives @xmath27 .
such a symmetry is necessary in this case so that the fourier transform gives the real - valued spectral density function for cavity emission .
examination of eqns .
( [ eqn11 ] ) and ( [ eqn12 ] ) highlight the correct asymptotic response : if @xmath28 then @xmath29 , the coherence function of the cavity . for this condition
@xmath30 such that : @xmath31 similarly , if @xmath32 then @xmath33 , the coherence function for the underlying emission .
for this condition @xmath34 such that : @xmath35 these two limiting cases are also of particular interest to commercial applications . for @xmath28
is typically required for lasing applications , here , notice that the self coherence function , @xmath36 is a function of angle , however is solely dependent on the cavity .
engineering the cavity allows direct control of the lasing mode .
the second limit , @xmath32 , has only recently become applicable in semiconductor devices with the advent of devices that have isolated single quantum dot light sources .
@xcite the spectrally pure emission from single quantum dot sources within a microcavity will have coherence properties that vary weakly with angle since @xmath37 .
the control over emission direction is still available by engineering the cavity s tuning . for rcleds @xmath38 , however , the difference is not large enough to reduce the dynamics to one of the cases described above by eqns .
( [ eqn13 ] ) and ( [ eqn14 ] ) .
the coherence model has been compared with experimental results for tuned and detuned rcleds operating near @xmath39 nm , the details of which can be found in refs . and .
angle resolved spectra were sampled through a pin hole subtending a half angle @xmath40 at the device using a monochromator with a resolution @xmath41 nm . these were converted to the coherence domain by inverse fourier transform . the coherence length , @xmath42 , measured with respect to the angle , @xmath43 in air , is defined in eqn .
( [ eqn15 ] ) is used to analyse the correspondence of spectral domain measurements and the coherence model developed here .
@xmath44 the fitting parameters for the coherence model are the coherence times of the cavity mode and underlying source , @xmath45 and @xmath17 respectively and the relative detuning , @xmath46 . here
, it is assumed that the underlying cavity coherence is approximately constant with angle , a fair approximation for high index devices .
figure [ fig1 ] shows the variation of coherence length , @xmath42 for the tuned ( fig .
[ fig1 ] .
( a ) ) and detuned ( fig .
[ fig1 ] .
( b ) ) samples .
the broken lines represent the limiting case of @xmath47 which is clearly an insufficient description for both rcled samples .
the solid line is a least squares fit of the model and the three unknown model parameters .
the correspondence of experiment and theory is excellent for the lower angles . at large angles
, it should be noted that optical loss from the doped regions of the rcleds may cause the deviation from the theoretical trend .
this is supported by the observation that the deviations occur at the same spectral position , i.e. a relative shift in the deviation s angular position of @xmath48 due to the relative device tunings .
table [ tab1 ] shows the fitting parameter values for the rcled samples in terms of coherence time ( and spectral wavelength ) .
this has been done as it is more natural to specify the detuning in terms of wavelength .
.table of fitting parameter values given for both coherence ( time ) and spectral domain . [ cols="<,>,>",options="header " , ] good agreement is apparent between parameters of the tuned and detuned rcleds .
in addition , the values for the detuning closely match the design values of @xmath49 nm and @xmath50 nm for tuned and detuned devices respectively .
the trend itself is also of significant importance ; clearly interference between the cavity mode and underlying emission only resolves to a peak in coherence at the tuning angle .
this is only visible in the case of the detuned rcled at approximately @xmath51 , corresponding to a detuning of @xmath52 nm . in the previous study , the density of emission states within the numerical aperture of emission was taken into account . here
, the density of off - axis states increases as @xmath53 .
in addition , the differential change in solid angle with respect to air @xmath54 and the cavity @xmath55 must be taken into account . by integrating eqn .
( [ eqn12 ] ) over solid angle , the variation of the self coherence function as a function of na is given by : @xmath56 figure [ fig2 ] .
shows the variation of coherence length as a function of na for the two rcleds under investigation evaluated using the coherence model with the parameters shown in tab .
[ tab1 ] .
although experimental results for these trends are not available , they do follow the generic trend observed recently in ref . .
the square markers show the coherence length extrema calculated from measured spectra at normal incidence and @xmath57 in an integrating sphere offering further confirmation of the result .
the coherence properties of planar microcavities are modelled well by the coherence formula of eqn .
( [ eqn12 ] ) .
consider therefore , using this formula to predict how large a range of coherence length variation can be engineered over useful numerical apertures .
the reader will notice , that a low nas , when only a few of the transverse cavity modes are sampled , the coherence is near to that of the cavity .
statistical fluctuations within the cavity are small since the cavity samples the emission source over a long time than its coherence time resulting in a sharper spectral line and enhanced coherence properties .
the behaviour is also seen for all other emission angles into air as the cavity lifetime is approximately constant , however , the emission frequency does change with angle . at large nas ,
each cavity mode still samples the underlying emission over a longer time scale than the coherence time of the source , however , the sampled emission is reconstituted spectrally , reproducing to some extent the underlying statistical fluctuations .
this appears to be a fair description given the observations .
therefore , in order to create a larger range of coherence variation in these device , the finesse of the cavity must be increased and the coherence of the emission source must be minimised . most relevant , however , will be to increase the cavity finesse , over which most control can be achieved . to test this concept ,
consider varying the cavity finesse in the model for the tuned device discussed above .
fig .
[ fig3 ] shows the maximum and minimum coherence lengths for useful nas for the rcled device as a function of finesse and for a range of device tunings . here ,
_ useful _
na must be stressed : the maximum coherence corresponds to the na through which at least @xmath58 % of the device power emitted into air is sampled while the minimum coherence is limited by the limitation of coupling optics set here at an angle of @xmath59 .
these limits are illustrated at the top left corner of fig .
[ fig3 ] . despite the restrictions on useful na ,
a large variation in coherence length is apparent in this extrapolated example as is evident from fig .
[ fig3 ] . increasing the finesse by an order of magnitude , increases the maximum coherence length by just over a factor of @xmath60 .
the minimum na is increased by nearly a factor of @xmath61 .
the range of coherence lengths , on the other hand goes from @xmath62 m to @xmath63 m such that it almost spans an entire order of magnitude .
it was evident in the result of figs .
[ fig1 ] a ) and [ fig1 ] b ) that the maximum coherence length occurs at the tuning angle or wavelength .
this behaviour is also seen in the results of fig .
[ fig3 ] where the maximum coherence length occurs for a tuned sample and indeed gives the greatest variation over useful nas .
quantitatively , this amounts to a @xmath61 fold increase in maximum coherence for a tuned device compared with a @xmath64 nm detuned , which would be used for greater power extraction into large nas .
clearly , high finesse , tuned microcavity devices provide the greatest coherence variation over useful nas .
the self coherence function for planar microcavities has been modelled using an analytical formula based on the underlying self coherence functions of the emitter and microcavity .
the correspondence of the measured coherence of two rcled samples and the coherence model described by eqns .
( [ eqn12 ] ) and ( [ eqn16 ] ) is compelling .
firstly , a good function fit to the coherence lengths of fourier transformed spectra sampled at discrete angles was found .
furthermore , the same fitting parameters reproduced closely , the coherence variation as a function of na , by integrating over the whole set of emission states.@xcite the central result of this analysis is the observation that the coherence of the device is dependent on the underlying coherence of the cavity and emitter .
indeed the underlying statistical fluctuation of the light source are exposed when varying na . at low nas ,
the coherence length is close to that of the microcavity . at large nas ,
the coherence length is closer to that of the underlying emitter .
a key aspect of the modelling shows that the coherence properties of planar microcavities can be engineered through microcavity design and selection of an emitter with suitable coherence properties . here
, the model was used to predict a larger range of coherence variation by increasing the finesse of the microcavity .
coherence lengths across an order of magnitude could be achieved if these design consideration were considered carefully .
these observations suggest that variable coherence is a result of a filtering mechanism ; this is also the spectral domain picture of microcavity emission .
however , in the coherence domain , it is the statistical fluctuations of the light source that are filtered by the slow microcavity sampling time .
this is the only robust interpretation that can reconcile the difference in coherence between planar microcavity emission viewed through small and large nas and raises questions about the relationship between the emission lifetime and coherence time in any type of microcavity device .
h. benisty et al , ieee j. quantum electron . *
34 * , 9 , 1612 - 1631 ( 1998 ) .
e. moreau et al , appl .
phys .
lett . ,
* 79 * , 18 , 2865 - 2867 ( 2001 ) .
c. santori et al , nature , * 419 * , 10 , 594 - 597 ( 2002 ) .
p. n. stavrinou et al , j. appl
. phys .
* 86 * , 6 , 3475 - 3477 ( 1999 ) .
j.w .
gray et al , angular emission profiles and coherence length measurements of highly efficient , low - voltage resonant - cavity light - emitting diodes operating around 650 nm . " in _ light - emitting diodes : research , manufacturing , and applications v _ , h. walter yao , e. f. schubert , eds . , proc .
spie * 4278 * , 81 - 89 ( 2001 ) .
r. f. oulton et al , optics comms .
* 195 * , 5 - 6 , 327 - 338 ( 2001 ) .
r. c. coutinho et al , j. lightwave technol .
* 21 * , 1 , 149 - 154 ( 2003 ) .
j.m .
schmitt , ieee j. sel .
topics in quant .
electron . * 5 * , 4 , 1205 - 1215 ( 1999 ) .
r. c. youngquist et al , opt .
lett . , * 12 * , 158 ( 1987 ) g. s. kino et al , appl
. opt .
, * 29 * , 3775 , ( 1990 ) d. d. sampson et al , electron .
lett .
* 26 * , 19 ,
1550 - 1551 ( 1990 ) . m. born & e. wolf , _ principles of optics , 6@xmath65 ( corrected ) edition _ , ( cambridge , 1999 ) .
k. h. drexhage , progress in optics " ed .
e. wolf , vol .
12 , ch . 6 , ( 1974 ) .
h. g. deppe et al , j. mod . opt . * 41 * , 2 , 325 - 344 ( 1994 ) .
c. santori et al , nature * 419 * , 594 - 597 ( 2002 ) .
z. yuan et al , science * 295 * , 102 - 105 ( 2002 ) . |
the habitable zone is defined as the region around a star where a planet with co@xmath0 and h@xmath0o as its main greenhouse gases can support liquid water at its surface @xcite .
the habitable zone is relatively wide because of the silicate - weathering feedback @xcite .
silicate - weathering is a geological process that removes co@xmath0 from the atmosphere and a negative ( stabilizing ) feedback is possible because this process is believed to run faster at higher temperatures and higher co@xmath0 partial pressures @xcite .
the inner edge of the habitable zone is set by the moist or runaway greenhouse @xcite , which should not be influenced by the details of the silicate - weathering feedback since the co@xmath0 should have been drawn down to low levels when they occur .
on the other hand , calculations of the outer edge of the habitable zone generally assume that the silicate - weathering feedback can maintain co@xmath0 at arbitrarily high levels . the outer edge
is then marked by some threshold where adding co@xmath0 to the atmosphere no longer provides additional warming , for example if additional co@xmath0 increases rayleigh scattering more than it increases greenhouse warming @xcite .
this picture essentially considers the asymptotic limit of unlimited co@xmath0 outgassing capacity .
more recently it has been recognized that for finite co@xmath0 outgassing rates the maximum co@xmath0 that a planet can achieve may be lower than the co@xmath0 needed to keep that planet habitable at the outer edge of the habitable zone @xcite .
this leads to what i will call the `` effective outer edge of the habitable zone , '' which will depend on the co@xmath0 outgassing rate and the functioning of silicate weathering on the planet . beyond the effective outer edge of the habitable zone , a planet may experience cycles @xcite between a globally frozen , snowball earth climate @xcite and a habitable climate .
understanding and constraining the effective outer edge of the habitable zone is critical because it determines our estimate of the fraction of stars that host an earth - like planet @xcite , which is essential for planning future telescopes that would observe such planets .
the purpose of this paper is to investigate what determines the position of the effective outer edge of the habitable zone and what happens beyond the effective outer edge .
the behavior of a planet in this regime is largely determined by its uncertain silicate - weathering behavior . instead of viewing this an an obstacle
, i will exploit it to make progress on the problem .
the importance and uncertainty of weathering justifies the use of a simple climate model since the errors associated with making the grave assumptions that such a model entails pale in comparison to the uncertainty in weathering .
it also allows me to use a relatively simple weathering parameterization that captures the expected qualitative behavior .
i will use the simplifying assumptions to derive useful formulae .
this will allow us to understand issues such as which processes determine whether the effective outer edge of the habitable zone is inside the traditional outer edge of the habitable zone and how much the parameters associated with these processes would need to be changed from our best estimates of their values to change the qualitative behavior of the system .
moreover , i am able to determine the conditions for climate cycles to occur beyond the effective outer edge of the habitable zone , an analytical solution for these cycles , and a formula for the period of the cycles . using more complicated climate and weathering models will alter quantitative results , but is unlikely to alter the qualitative dependencies on parameters that my formulae give .
this work therefore complements recent work by @xcite and @xcite , who used more complicated radiative and climate models , but only explored a few values of weathering parameters .
i will neglect sophisticated radiative @xcite and 3d calculations @xcite and consider the following linearized , zero - dimensional model of planetary climate , similar to that used by @xcite : @xmath1 where @xmath2 and @xmath3 are constants , @xmath4 is the global mean temperature , @xmath5 is the temperature of the reference state , @xmath6 is the atmospheric partial pressure of co@xmath0 , @xmath7 is the atmospheric partial pressure of co@xmath0 in the reference state , @xmath8 is the stellar flux , @xmath9 is the stellar flux in the reference state , @xmath10 is the heat capacity in units of j m@xmath11
@xmath12c@xmath13 , @xmath14 is the temperature - dependent planetary albedo ( reflectivity ) , and @xmath15 is the albedo of the reference state .
table [ tab : params ] contains a list model parameters and their standard values .
the values i use here are drawn from @xcite , @xcite , and @xcite , but it is important to emphasize that the arguments in this paper do not depend on the exact values chosen .
i will let the albedo be specified by @xmath16 where @xmath17 is the albedo of the warm climate state ; @xmath18 is the albedo of the cold and icy , `` snowball '' climate state ; and @xmath19 is the temperature at which the planet transitions between the two climate states .
equation ( [ eq : albedo ] ) assumes a step - wise transition in albedo , which neglects the effects of spatial resolution @xcite .
nevertheless , it allows us to make easier analytical progress and does not alter the qualitative behavior of the system unless multiple snowball - like climate states possible @xcite , which is a possibility we will not consider here . as we will see below , introducing a smoothed albedo transition only introduces a repulsing fixed point in some situations that the climate could not exist stably in . finally , if we assume that the reference climate state is warm , then @xmath20 .
the partial pressure of co@xmath0 is determined by outgassing and weathering @xmath21 where @xmath22 is the co@xmath0 outgassing rate , @xmath23 is the rate of removal of co@xmath0 from the atmosphere by silicate weathering in the reference climate state , @xmath24 is a rate constant for the increase in weathering with temperature , and @xmath25 is an exponent that determines how strongly weathering depends on atmospheric co@xmath0 partial pressure . @xcite used an analysis of river catchments to estimate that @xmath26 @xmath12c@xmath13 ( 1-@xmath27 error ) .
based on that study , the plausible range for @xmath24 is roughly 00.2 .
i will use a standard value of @xmath24=0.1 and vary @xmath24 over this range .
@xmath25 could be zero if land plants concentrate co@xmath0 in the soil at the same level regardless of the atmospheric co@xmath0 concentration @xcite , and theoretical arguments suggest @xmath25 has a maximum of 1 @xcite .
i will use a standard value of @xmath25=0.5 @xcite , but consider variations within the plausible range .
depressurization caused by glacial unloading can cause temporary increases in the co@xmath0 outgassing rate @xcite .
i neglect this affect and take the co@xmath0 outgassing rate to be independent climate state here , which is appropriate for longterm average behavior .
note also that in section [ sec : cycles ] we will consider @xmath23=0 when @xmath28 . for the purposes of this paper
i am assuming that weathering follows a similar parameterization on land and at the seafloor , but it should be understood that the weathering behavior of an ocean planet would likely be quite different from that of a planet with an earth - like land fraction @xcite .
the weathering parameterization in equation ( [ eq : co2_bal ] ) is similar to that used by other authors @xcite , but i have dropped the relatively weak dependence on temperature that is often included to represent changes in precipitation with temperature .
the qualitative behavior of weathering as a function of temperature is captured by equation ( [ eq : co2_bal ] ) without this additional complication .
given the ad hoc nature of all weathering parameterizations , it is reasonable to choose the simplest parameterization that gives the expected qualitative behavior of weathering processes given the objectives of this paper . equations ( [ eq : climate_model])-([eq : co2_bal ] ) define the system .
the plan for analyzing them is as follows .
first we will consider the warm ( habitable ) state ( section [ sec : warm ] ) .
i will set the albedo ( equation ( [ eq : albedo ] ) ) to its warm state value , set the time derivatives in equations ( [ eq : climate_model ] ) and ( [ eq : co2_bal ] ) to zero , and solve the system for conditions when the temperature is high enough for the warm state to exist .
this will allow us to put bounds on the existence of the warm state , that is , define an effective outer edge of the habitable zone that may be more restrictive than the traditional outer edge .
next i will find nullclines of the system defined by equations ( [ eq : climate_model ] ) and ( [ eq : co2_bal ] ) ( that is , lines where the time derivatives equal zero ) , find their intersections ( fixed points , or solutions of the system ) , and determine the stability of these fixed points .
physically , this will reveal that if equation ( [ eq : co2_bal ] ) is followed as is , enough weathering occurs that the system settles into a stable snowball state when the warm climate state ceases to exist , rather than into climate cycles between snowball and warm conditions .
i will then show that climate cycles do occur if we set weathering to zero in the snowball state in equation ( [ eq : co2_bal ] ) ( section [ sec : cycles ] ) .
i will find analytical solutions for the components of these cycles and derive an analytical formula for their period that can predict well results from the more intricate model of @xcite .
i will then discuss these results in section [ sec : discussion ] and conclude in section [ sec : conclusions ] .
.a list of the model parameters , their descriptions , and the standard values i use for them .
the parameter values in this table are taken from @xcite , @xcite , and @xcite .
note that , following @xcite , i use a value of @xmath23 ten times higher than @xcite .
i perform sensitivity analyses where i vary @xmath24 and @xmath25 .
[ cols="<,<,<",options="header " , ]
in this section we will consider the conditions that allow the carbon cycle to maintain a planet in the warm climate state , which is necessary for the planet to be considered habitable in the traditional sense .
this requires a steady - state solution , so we can set the time derivatives in equations ( [ eq : climate_model ] ) and ( [ eq : co2_bal ] ) to zero .
we can solve this system to find @xmath29 where @xmath30 is the warm state temperature .
we could rewrite equation ( [ eq : t_w ] ) in non - dimensional form , but i will leave it , and subsequent equations , in dimensionful form to make them more accessible physically . from equation ( [ eq : t_w ] ) we can see that increasing the co@xmath0 outgassing rate logarithmically increases the warm state temperature , and that increased outgassing warms the warm state more if the climate is more sensitive to co@xmath0 ( higher @xmath3 ) , provided @xmath31 .
we can also see that the influence of stellar flux on the warm state temperature depends on @xmath25 , the weathering - co@xmath0 power - law exponent . if @xmath32 , then changing the stellar flux has no effect on the warm state temperature as long as the stellar flux is high enough that @xmath33 , so that the warm state can exist .
instead , the warm state temperature is determined by the outgassing rate . as @xmath25 increases
, the warm climate state becomes increasingly sensitive to changes in stellar flux .
we can solve for the warm state co@xmath0 partial pressure ( @xmath34 ) as follows @xmath35 the warm state co@xmath0 partial pressure has a power law dependence on the co@xmath0 outgassing rate .
the exponent depends most strongly on @xmath2 , which determines the increase in outgoing longwave radiation with temperature .
the more that increasing the warm state temperature by a given amount increases longwave cooling , the higher the warm state co@xmath0 must be to maintain that temperature at a given co@xmath0 outgassing rate . as expected , the warm state co@xmath0 partial pressure is exponentially lower if the stellar flux is higher .
this effect is mediated mainly by @xmath24 , the parameter that determines the increase in weathering with temperature .
if @xmath24 is higher , then a given warming from an increase in stellar flux causes the weathering to increase more , and draws down the co@xmath0 more .
using equation ( [ eq : t_w ] ) , we can solve for the coldest possible warm state by setting @xmath36 .
this is a very important condition because habitability would be lost if anything were to cool the climate when it is in the coldest possible warm state .
we can solve for the stellar flux at which the coldest possible warm state exists ( @xmath37 ) , which we can think as the effective outer edge of the habitable zone , as follows @xmath38 a lower value of @xmath37 means that the stellar flux must be decreased more to reach the coldest possible warm state , which means that the warm climate state exists in more of the habitable zone .
the ideal situation for habitability is when @xmath37 is decreased enough that it is smaller than the outer edge of the habitable zone , the point at which some other process , such as rayleigh scattering , prevents co@xmath0 from warming the planetary surface . if this is the case then the silicate - weathering feedback can keep a planet habitable throughout the entire habitable zone .
the first thing we can note from equation ( [ eq : s_star ] ) is that if the albedo of the warm state is smaller , the effective outer edge of the habitable zone will be further out .
since absorption of stellar flux by atmospheric water vapor should be larger for planets orbiting smaller , redder stars @xcite , the effective outer edge of the habitable zone is less likely to restrict the traditional habitable zone for m - stars than for g - stars and more likely for f - stars , which is consistent with what has been recently found by @xcite .
equation ( [ eq : s_star ] ) also tells us that the effective outer edge of the habitable zone depends only logarithmically on the co@xmath0 outgassing rate .
this is important because it tells us that large changes in the co@xmath0 outgassing rate are necessary to significantly change the effective outer edge .
for example , increasing the co@xmath0 outgassing rate by a factor of ten only decreases the effective outer edge from 794 w m@xmath11 to 530 w m@xmath11 , for our standard parameters ( figure [ fig : outer_outgassing2 ] ) .
changing the weathering parameters can have a much larger effect , which we can investigate by doing a sensitivity analysis in which we vary them .
for example , either increasing the weathering - temperature rate constant ( @xmath24 ) by a factor of two or decreasing the weathering - co@xmath0 power law exponent ( @xmath25 ) by a factor of two causes a similar reduction in the effective outer edge as increasing the co@xmath0 outgassing by a factor of ten ( figure [ fig : outer_outgassing2 ] ) . changing @xmath24 and changing @xmath25 have qualitatively different effects on the response of @xmath37 to changes in the co@xmath0 outgassing rate ( figure [ fig : outer_outgassing2 ] ) . decreasing @xmath25 increases the slope of @xmath37 as a function of @xmath39 , which means that increasing the co@xmath0 outgassing rate by a given amount extends the effective habitable zone further ( equation ( [ eq : s_star ] ) ) .
this is because the temperature becomes more sensitive to the co@xmath0 outgassing rate when @xmath25 is smaller ( equation ( [ eq : t_w ] ) ) . alternatively , changing @xmath24 changes the offset of the @xmath37 versus @xmath39 line , but not the slope .
a higher value of @xmath24 means that the temperature has to change less to cause the same change in weathering rate .
this allows the stellar flux to be dropped to a lower value before the warm state temperature reaches the temperature at which the warm state is lost ( @xmath36 ) . for reference
i have plotted the estimated positions of modern earth , earth 3.8 gyr ago , and mars 3.8 gyr on figure [ fig : outer_outgassing2 ] .
the simple model used here indicates that early and modern earth are safely inside the effective outer edge of the habitable zone , whereas early mars would have been beyond the effective outer edge of the habitable zone even if it were within the traditional habitable zone .
this is consistent with the histories of the two planets if we assume that fluvial features on early mars were episodic @xcite .
that said , it should be understood that specific conclusions like these depend on details of weathering parameterizations , and the purpose of this paper is to expose the qualitative effects of weathering , rather than to try to answer detailed questions .
another way we can think about this is to consider the set of weathering parameters that would make the effective outer edge of the habitable zone correspond to the traditional outer edge of the habitable zone .
if we denote the traditional outer edge of the habitable zone as @xmath40 , we can rewrite equation ( [ eq : s_star ] ) as @xmath41 equation ( [ eq : outer_edge ] ) represents a series of lines of @xmath24 as a function of @xmath25 for different values of the co@xmath0 outgassing rate .
if @xmath24 is larger or @xmath25 is smaller than the line for a particular co@xmath0 outgassing rate , then the effective habitable zone is just as large as the traditional habitable zone for this co@xmath0 outgassing rate .
figure [ fig : outer_outgassing ] shows three such curves for different co@xmath0 outgassing rates . for the default values of @xmath24 and @xmath25 , some of the habitable zone would be lost even if the co@xmath0 outgassing rate were ten times higher than modern earth s
( @xmath42 would be required for the effective outer edge to equal the traditional outer edge ) . however , relatively small changes to @xmath24 , @xmath25 , or some combination of the two can save the habitable zone even for modern earth s co@xmath0 outgassing rate ( figure [ fig : outer_outgassing ] ) .
the parameters @xmath24 and @xmath25 are uncertain and may vary between planets , but the expressions in this section could be used for a probabilistic estimate of planetary habitability if appropriate priors are used for @xmath24 and @xmath25 .
if the stellar flux or co@xmath0 outgassing rate is lowered enough that the warm climate state no longer exists , the planet enters a snowball state and the albedo increases according to equation ( [ eq : albedo ] ) .
it is an open question how weathering would behave in a snowball state .
@xcite assumed that weathering would go to zero due to lack of rain ( liquid water ) , and we will consider this case in section [ sec : cycles ] .
the possibility remains , however , that weathering could continue to occur under wet - based ice sheets or at the seafloor @xcite .
for illustrative purposes , we will continue to use equation ( [ eq : co2_bal ] ) to characterize weathering in the snowball state , but it should be understood that either subglacial or seafloor weathering could lead to different parameterizations .
as we will see , however , the important point is that if some co@xmath0-dependent weathering can occur in the snowball state , and it causes weathering to increase enough to balance co@xmath0 outgassing before the snowball deglaciates , then it is possible for this state to be a stable solution of the system .
we can solve for the temperature tendency nullcline ( the line on which @xmath43=0 in equation ( [ eq : climate_model ] ) ) , which results in @xmath44 equation ( [ eq : t_nullcline ] ) describes two lines of @xmath45 as a function of @xmath46 , each with a slope of @xmath47 .
the colder solution has a larger vertical offset .
this is because more co@xmath0 would be required to keep the cold state at a given temperature than the warm state because the albedo is higher in the cold state ( although they can never actually exist at the same temperature ) .
similarly , we solve for the co@xmath0 partial pressure tendency nullcline ( the line on which @xmath48=0 in equation ( [ eq : co2_bal ] ) ) to find @xmath49 equation ( [ eq : p_nullcline ] ) describes a line of @xmath45 as a function of @xmath46 with a slope of @xmath50 .
since the slope is negative , there will always be at least one intersection of the two nullclines , which will be a steady - state of the system .
as the co@xmath0 outgassing rate ( @xmath22 ) increases , the intercept of equation ( [ eq : p_nullcline ] ) is increased and the solution becomes warmer . for high values of @xmath22 only the warm state is a steady state , and for low values only the cold state is a steady state . for intermediate values of @xmath22
it is possible to have both states to be a steady state of the system .
we already solved for the warm state temperature and co@xmath0 in equations ( [ eq : t_w ] ) and ( [ eq : p_w ] ) .
we can now solve for the cold state temperature ( @xmath51 ) and co@xmath0 ( @xmath52 ) using the cold branch of equation ( [ eq : t_nullcline ] ) @xmath53 the final point of interest is to determine the stability of the fixed points described by equations ( [ eq : t_w ] ) , ( [ eq : p_w ] ) , ( [ eq : t_c ] ) , and ( [ eq : p_c ] ) .
we can do this by evaluating the jacobian ( @xmath54 ) of the system at the fixed points @xmath55 for both the warm and cold states , the trace of the jacobian ( @xmath56 ) is less than zero and its determinant ( @xmath57 ) is greater than zero .
this means that the warm and cold states are always attracting @xcite .
@xmath58 is generally positive , which means that the fixed points will usually be stable nodes , although it is possible for them to be stable spirals for some parameter combinations .
i have plotted what we have learned about the nullclines and their intersections for a representative set of parameters in figure [ fig : nullclines ] .
this plot shows how intersections of the nullclines lead to steady states , and how at least one intersection will always occur since the co@xmath0 partial pressure tendency nullcline has a negative slope and the two temperature tendency nullclines have a positive slope .
note that although the simplicity of the model we are considering constrains the nullclines to be linear , we would expect similar , but potentially nonlinear , behavior from a more complicated model .
for example , if we used a more sophisticated radiative transfer model for the climate calculations ( e.g. , * ? ? ?
* ) it would lead to curvature in the temperature tendency nullclines , but no change in the topology of the system . because we have assumed a discontinous albedo transition ( equation ( [ eq : albedo ] ) ) stable fixed points appear and disappear in isolation in figure [ fig : nullclines ] .
if we had instead assumed a smoothed albedo transition , for example smoothed with a hyperbolic tangent function , the two temperature tendency nullclines would smoothly join together . in this case
, there would always be an unstable saddle fixed point between the two attracting fixed points representing the warm and snowball climate states when both attracting states exist at the same co@xmath0 outgassing rate .
if the co@xmath0 outgassing rate were changed sufficiently , the saddle fixed point would merge with either of the attracting fixed points in a saddle node bifurcation , rather than the attracting fixed point just disappearing as it does in the discontinous albedo system .
alternatively , we can consider the situation where the weathering rate is smaller than the co@xmath0 outgassing rate for temperatures less than the temperature at which the planet transitions between the two climate states ( @xmath28 ) . for simplicity , i will set the weathering rate to zero for @xmath28 , following @xcite . in this case
no snowball steady state is possible because there is no weathering term to balance co@xmath0 outgassing when the planet is experiencing a snowball ( equation ( [ eq : co2_bal ] ) ) .
instead co@xmath0 simply accumulates in the snowball state , warming it , until the temperature @xmath19 is reached .
this causes the albedo to decrease ( physically the ice melts ) and the planet abruptly jumps into the warm climate state . if we assume that the co@xmath0 outgassing rate is low enough that no warm climate steady state exists , then the warm climate leads to the rapid removal of co@xmath0 by weathering until @xmath19 is again reached , then the climate abruptly jumps into the snowball state , and the cycle repeats
. figure [ fig : cycles ] shows a diagram of this cycle in phase space and figure [ fig : cycles2 ] shows timeseries of co@xmath0 partial pressure and temperature through the cycle .
the two transitions between the snowball and warm states occur on the timescale of relaxation back to the temperature tendency nullcline .
this timescale is given by @xmath59 years , which is essentially instantaneous for our purposes ( the system is extremely stiff ) .
this allows us to make the approximation that as the co@xmath0 changes in either the warm or snowball state , the climate exists along the temperature tendency nullcline ( @xmath43=0 in equation ( [ eq : climate_model ] ) ) .
the co@xmath0 partial pressure in the warm state ( @xmath60 ) as a function of the temperature in the warm state ( @xmath61 ) is therefore @xmath62 and the co@xmath0 partial pressure in the cold state ( @xmath63 ) as a function of the temperature in the cold state ( @xmath64 ) is @xmath65 i have used a tilde for the co@xmath0 partial pressure and temperature variables in these equations because they are not true solutions of the system , since weathering never balances co@xmath0 outgassing during the cycles .
@xmath48 is constant during the snowball phase if the weathering rate is zero ( equal to @xmath22 ) , so it is easy to calculate the time spent in the snowball phase ( @xmath66 ) as @xmath67 where we have used the fact that in general @xmath68 .
similarly , if we assume that silicate weathering is limited by the supply of silicate cations from erosion @xcite , then we can approximate the weathering rate as a constant during the warm phase . using similar logic as we used to get equation ( [ eq : snow_time ] ) , we arrive at a first estimate for the time spent in the warm state ( @xmath69 ) of @xmath70 where @xmath71 is the factor by which the supply - limited maximum weathering rate exceeds modern earth s weathering rate ( @xmath722.5 is a good guess for an earth - like planet , * ? ? ? * ) and @xmath73 is the fractional reduction in warm state relative to cold state lifetime due to the fact that the weathering is higher in the warm state . using @xmath23=20@xmath22 , as in figure [ fig : cycles ] , we get @xmath73=@xmath74 .
this would indicate that a negligible amount of the time in the cycle is spent in the warm state relative to the cold state .
it is more difficult to calculate the time spent in the warm climate state if we let @xmath48 vary .
substituting into equation ( [ eq : co2_bal ] ) we find the following differential equation for the co@xmath0 partial pressure in the warm state ( @xmath60 ) @xmath75 equation ( [ eq : co2_bal2 ] ) has a simple analytical solution for @xmath76=1 , which happens to be the case for our default parameters ( table [ tab : params ] ) .
i will use this limit to illustrate the behavior of the warm state co@xmath0 drawdown .
the initial condition is @xmath77 , so that equation ( [ eq : co2_bal2 ] ) is solved by @xmath78 we are seeking the time , @xmath79 , such that @xmath80 , so plugging into equation ( [ eq : co2_bal2_soln ] ) we find @xmath81 in general @xmath82 , as we would expect because the temperatures are high and the weathering is fast in the warm state ( leading to a low @xmath83 ) and we limit the weathering rate in our other warm state timescale estimate ( @xmath69 ) . for example , for the parameters used in figure ( [ fig : cycles ] ) ,
@xmath66=250 myr ( slow ) , the two estimates for the time spent in the warm state are @xmath69=5 myr and @xmath83=0.5 myr ( fast ) , and the time to transition between the warm and snowball states ( @xmath84 ) is about 3 years ( very fast ) .
since the other components of the cycle take are short , @xmath66 ( equation ( [ eq : snow_time ] ) ) yields a good approximation of the period of the total cycle , which we will call @xmath56 .
we can drop the constants associated with the reference state in equation ( [ eq : snow_time ] ) as follows to think about the variable dependencies of @xmath56 @xmath85 as one might expect , the period of the cycles is inversely proportional to the co@xmath0 outgassing rate .
the more interesting aspect of equation ( [ eq : snow_time2 ] ) is that it gives us a functional form for the dependence of the period of the cycle on the stellar flux and albedo of the cold state .
if we start in the warm state and decrease the stellar flux ( say by moving the planet away from the star ) , a limit cycle will suddenly appear when the threshold for existence of the warm state is crossed ( equation ( [ eq : s_star ] ) ) .
if we continue to decrease the stellar flux , the period of this cycle will grow exponentially as the stellar flux is decreased .
the exponential functional dependence ultimately derives from the logarithmic effect of co@xmath0 on infrared emission to space ( equation ( [ eq : climate_model ] ) ) .
this means that the period of the cycle for planets near the outer edge of the habitable zone will be very long , so the difference between having a permanent snowball state as in section [ sec : stable_snow ] and having very slow cycles through the snowball and warm states as described in this section will be marginal
. similarly , the timescales will be exponentially shorter as we decrease the snowball albedo .
since ice and snow have a much lower albedo for an m - star spectrum @xcite , this implies that the climate cycles for planets in m - star systems would have much shorter periods .
finally , increasing @xmath3 decreases the effect of changing the stellar flux and snowball albedo .
it is important to note that since we have set the weathering rate to zero in the snowball , which is the part of the cycle that determines its period , none of the uncertain weathering parameters affect the period of the cycle .
the dependence of @xmath56 on @xmath19 ( equation ( [ eq : snow_time2 ] ) can significantly affect the comparison of different models .
for example , i implemented a smoothed version of the albedo transition ( equation ( [ eq : albedo ] ) ) and integrated the system numerically .
i found that the period of the cycle was strongly dependent on the temperature smoothing of the albedo parameterization because this affected the effective temperature at which the transition from the snowball to the warm climate occurred .
using @xmath86 , a decrease in @xmath19 by about 10 k leads to a decrease in @xmath56 by about an order of magnitude .
this sensitivity will affect the comparison of cycle periods between models , however , the dependencies shown in equation ( [ eq : snow_time2 ] ) should hold within a given model .
we can use equation ( [ eq : snow_time2 ] ) to understand simulations in more complex models .
for example , @xcite performed six different simulations of climate cycles for planets at different orbital distances ( corresponding to different stellar fluxes ) , co@xmath0 outgassing rates , and values of @xmath25 .
two of these simulations have a co@xmath0 outgassing rate three times higher than the others , and , according to equation ( [ eq : snow_time2 ] ) , i have adjusted their period by multiplying it by three .
i have plotted the logarithm of the adjusted period as a function of stellar flux in figure [ fig : menou ] .
i have also plotted the time spent in the snowball state for each simulation , which we can calculate because @xcite gives the percentage of the cycle spent in the warm state .
the first thing to note is that the logarithm of the period is fairly linear in stellar flux , consistent with equation ( [ eq : snow_time2 ] ) .
if we assume that the snowball albedo is 0.7 , the slope of the line corresponds to a value of @xmath3 of about 27 w m@xmath11 , which is the right order of magnitude .
two of the simulations have periods 1020% longer than expected .
these simulations have a lower value of @xmath25 , and therefore spend a higher fraction of their cycle in the warm state @xcite .
we should also note that even the time spent in the snowball state does not fall exactly on a straight line , and it is not a single - valued function of stellar flux .
the reasons for this are : ( 1 ) @xcite uses a much more complex radiation model in which infrared emission to space is not simply a linear function of the logarithm of co@xmath0 , ( 2 ) @xcite calculates the snowball albedo including the effect of the amount of co@xmath0 , so it is not a constant , and ( 3 ) the weathering parameterization that @xcite uses allows some weathering for temperatures slightly below the snowball deglaciation temperature threshold , which can delay the deglaciation and causes different results for different values of @xmath25 . despite these differences ,
equation ( [ eq : snow_time2 ] ) does an excellent job of describing the qualitative behavior of period of the cycles from the simulations of @xcite .
a major advance of this paper is to derive an explicit formula for the effective outer edge of the habitable zone ( equation [ eq : s_star ] ) .
although the weathering parameters in this formula are uncertain , it could be incorporated into future probabilistic estimates of habitability of discovered exoplanets , with appropriate prior distributions placed on weathering parameters . once a planet is beyond this limit , i have found that it will either experience a permanent snowball state or long cycles between a snowball and warm climate , depending on whether weathering goes completely to zero during the snowball or note . either way the habitability of the planet would be greatly reduced .
also , the fact that the snowball episodes earth has experienced did end does not imply that the weathering was zero during them @xcite , since the stellar flux was relatively high during these episodes .
@xcite found that the outer edge of the traditional habitable zone has only a small dependence on planet size , but the effective outer edge of the habitable zone could strongly depend on planet size .
a larger planet will tend to have a higher rate of volcanism , and presumably co@xmath0 outgassing , yet it will also have a larger overburden pressure for a given volatile inventory @xcite . these competing effects will determine how planetary size affects co@xmath0 outgassing rate , and consequently susceptibility to loss of the warm climate state inside the traditional habitable zone .
moreover , the co@xmath0 outgassing rate should decrease strongly with time @xcite , which indicates that planets near the outer edge of the traditional habitable zone are more likely to actually be habitable in younger systems .
the albedo and thermal phase curve of an earth - like planet could be interrogated to determine whether it was in a snowball or warm climate state @xcite .
this might be possible for a planet near the outer edge of the habitable zone with the _ james webb space telescope _
@xcite , and would certainly be possible with a future mission such as the _ high definition space telescope _
new geochronological data @xcite suggest that the sturtian and marinoan snowball earth episodes had a combined duration of about 80 myr , which is about 10% of the time since they occurred , implying that a roughly 10% snowball duty cycle could be realistic for an earth - like planet .
planets near the outer edge of the traditional habitable zone that have a stable warm state but are perturbed away from it and into a snowball could take longer to warm up via co@xmath0 outgassing and will therefore spend a somewhat higher percentage of their time as a snowball .
if these planets are anything like earth , however , they should still spend the vast majority of their time in the warm climate state .
planets outside the effective outer edge of the habitable zone , on the other hand , should spend most of their time as a snowball . therefore a measurement of the fraction of earth - like planets in a warm state as a function of position in the habitable zone would tell us whether co@xmath0-outgassing limitations on the habitable zone are important .
a large increase in the average number of earth - like planets in a snowball state near the outer edge of the habitable zone would indicate that co@xmath0-outgassing limitations are important on average . in this way
astronomical measurements could increase our understanding of weathering . similarly , continued study of river catchments , paleoclimate , and laboratory weathering analogs should help improve our understanding of weathering , and therefore the effective outer edge of the habitable zone , which will inform the astronomical search for habitable exoplanets .
in this paper i have not included the effect of co@xmath0 on planetary albedo .
this means that if the weathering is set to zero in a snowball state , then the co@xmath0 can always build up to high enough values to cause deglaciation and lead to climate cycles if the warm state does not exist ( section [ sec : cycles ] ) . for snowball states that require tens of bars of co@xmath0 to deglaciate ,
increased shortwave scattering by co@xmath0 could prevent deglaciation from ever occurring .
if this were the case , the snowball state could be stable even if the co@xmath0 cycle is not balanced .
if deglaciaton does occur at a very high co@xmath0 level , then the atmospheric albedo might be so high that the change in surface albedo has a minimal effect on the planetary albedo @xcite , so the warming associated with deglaciation is minimal . in this case co@xmath0
could still be drawn down due to high co@xmath0 concentrations ( equation ( [ eq : co2_bal ] ) ) , leading to a climate cycle , but the planet would likely spend more time in the warm state . finally , we should note that even if the effective outer edge of the habitable zone occurs at a significantly higher stellar flux than the traditional outer edge , there should still be plenty of habitats for life in the universe . if the habitable zone were cut in half , the proportion of sun - like stars hosting earth - like potentially habitable planets would go from @xmath725% to @xmath722.5% @xcite , which might have been considered an optimistic estimate before the _ kepler _ mission .
moreover , recent work on h@xmath0-greenhouse planets suggests that habitable planets can exist even outside of the traditional habitable zone @xcite .
additionally , life , and maybe even animal life , seems to have survived the snowball earth episodes , and could potentially survive permanent or cyclical snowball climates near the outer edge of the habitable zone , although such conditions would certainly be less favorable to complex life than modern earth . finally , if simple life can exist in subglacial oceans on distant or unbound earth - like planets @xcite , then the snowball planets beyond the effective habitable zone would still be viable hosts for some sort of life .
the main conclusions of this paper are : 1
. the stellar flux at the effective outer edge of the habitable zone can be approximated by the following formula : @xmath87 larger values of @xmath24 , the weathering - temperature rate constant , linearly decrease the stellar flux of the effective outer edge of the habitable zone , moving it farther from the star and providing more habitable space in the system .
smaller values of @xmath25 , the weathering - co@xmath0 power law exponent , directly decrease the stellar flux of the effective outer edge of the habitable zone and also leverage the effect of increases in the co@xmath0 outgassing rate .
if @xmath24 is increased by about a factor of two or @xmath25 is decreased by a factor of two , or some smaller combination of the two , then the effective outer edge of the habitable zone equals the traditional outer edge of the habitable zone even for modern earth s co@xmath0 outgassing rate , and none of the habitable zone would be lost .
these changes are within the uncertainty in the values of @xmath24 and @xmath25 .
this equation also tells us that m - star planets should tend to have less of a reduction in their habitable zone due to limited co@xmath0 outgassing , since @xmath17 , the warm state albedo , will tend to be smaller for m - star planets ( making @xmath37 smaller ) .
+ the formula for @xmath37 could be incorporated into probabilistic estimates of whether a discovered exoplanet is habitable , using appropriate priors on @xmath24 and @xmath25 .
it could similarly be used to estimate the fraction of stars that host an earth - like planet given statistics of exoplanet occurrences .
both of these uses would aid in the planning of telescopes that would observe earth - like planets and search for biosignatures .
2 . beyond the effective outer edge of the habitable zone ( but inside the traditional outer edge )
cycles between a snowball and warm climate are possible if weathering is weak enough that the co@xmath0 needed to deglaciate a snowball is reached before weathering can balance co@xmath0 outgassing ( for example if the weathering rate is simply set to zero in a snowball climate ) . if weathering occurs either subglacially or at the seafloor , it is possible to have a stable , attracting snowball climate state .
if climate cycles between a snowball and warm state occur , then the period of these cycles scales as @xmath88 this formula comes from the time spent in the snowball state , which dominates the total period and can be calculated by dividing the co@xmath0 needed to deglaciate the snowabll by the co@xmath0 outgassing rate ( which explains why the period of the cycles is inversely proportional to the co@xmath0 outgassing rate ) . the exponential dependence on the temperature at which the snowball state deglaciates and the negative exponential dependence on the stellar flux ultimately derive from the fact that co@xmath0 has a logarithmic effect on infrared emission to space and the greenhouse warming of a planet . the negative exponential dependence on stellar flux indicates that cycles near the outer edge of the habitable zone will have very long periods , and may be hard to distinguish from permanent snowball states .
the exponential dependence on the planetary albedo of the cold state indicates that climate cycles will have a shorter period for planets orbiting m - stars because the albedo of ice and snow is lower for an m - star spectrum .
finally , it is important to note that none of the uncertain weathering parameters appear in this scaling .
i acknowledge support from the nasa astrobiology institute s virtual planetary laboratory , which is supported by nasa under cooperative agreement nnh05zda001c .
i thank navah farahat for helping me learn python plotting routines .
i thank cael berry , edwin kite , daniel koll , mary silber , and robin wordsworth for reading an early draft of this paper and providing detailed and insightful suggestions . | the habitable zone concept is important because it focuses the scientific search for extraterrestrial life and aids the planning of future telescopes .
recent work has shown that planets near the outer edge of the habitable zone might not actually be able to stay warm and habitable if co@xmath0 outgassing rates are not large enough to maintain high co@xmath0 partial pressures against removal by silicate weathering . in this paper
i use simple equations for the climate and co@xmath0 budget of a planet in the habitable zone that can capture the qualitative behavior of the system . with these equations
i derive an analytical formula for an effective outer edge of the habitable zone , including limitations imposed by the co@xmath0 outgassing rate .
i then show that climate cycles between a snowball state and a warm climate are only possible beyond this limit if the weathering rate in the snowball climate is smaller than the co@xmath0 outgassing rate ( otherwise stable snowball states result ) .
i derive an analytical solution for the climate cycles including a formula for their period in this limit .
this work allows us to explore the qualitative effects of weathering processes on the effective outer edge of the habitable zone , which is important because weathering parameterizations are uncertain . |
solar flares and coronal mass ejections ( cmes ) involve a sudden release of magnetic energy stored in complex active regions through magnetic reconnection @xcite .
the several mechanisms may compel the energy build - up in the flaring and eruptive regions that later released in form of bulk mass motion , heating , as well as acceleration of the energetic particles .
these processes may involve magnetic instabilities , flux and helicity emergence , building of the magnetic field complexity , etc .
* ; * ? ? ?
* ; * ? ? ?
* and references cited therein ) .
the two main space weather consequences of cmes are solar energetic particle ( sep ) events and geomagnetic storms ( gmss ) .
both flares and cmes may contribute to sep events . according to @xcite , there are two types of sep events , impulsive and gradual .
impulsive sep events , as characterized by enhanced @xmath4he , heavy ions and electrons and high charge states , are attributed to impulsive solar flares .
the solar sources of impulsive sep events are identified with coronal jets and narrow cmes @xcite .
gradual sep events , which are far more important in the context of space weather , are due to acceleration of particles at cme - driven shocks @xcite . fast moving shocks associated with cmes
accelerate seps in the interplanetary medium near the sun @xcite and at 1 au in energetic storm particle ( esp ) events @xcite .
cme interaction has been reported to increase the sep acceleration efficiency @xcite .
the interaction between cmes results in a new cme with the magnetic field and plasma from both the cmes .
the process is described as `` cme cannibalism '' @xcite .
recently , @xcite have reported on the implications of cme - cme interaction for shock propagation , particle acceleration and space weather forecasting .
such interactions have been modeled by three - dimensional mhd simulation @xcite .
some authors have made statistical argument questioning the importance of cme interaction for sep production @xcite . however
, if we consider cmes from the same active region , the physical interaction is inevitable , so there must be some influence of the cme interaction on the sep intensity profile @xcite .
intense geomagnetic storms occur when the southward field in the interplanetary cmes ( icmes ) reconnect with the earth s magnetic field @xcite .
intense storms can also occur when the sheath region between the icme and the shock has southward component of the magnetic field .
it has been found that nearly 90% of intense storms ( dst @xmath5 -100 nt ) result from earth - directed halo cmes @xcite . in this paper
we study multiple flares and cmes from the same active region on january 23 , 2012 that resulted in one of the largest sep events of solar cycle 24 .
the sequence of eruptions also resulted in a moderate gms .
we study the interacting cmes for the first time using multiple view points and extended field of view ( fov ) available from the solar terrestrial relations observatory ( stereo ) and solar and heliospheric observatory ( soho ) . in sec . 2
, we describe the data set used .
we report the observational results in sec . 3 .
in the last section ( sec . 4 ) , we summarize the results .
we use imaging observations from the stereo secchi suites @xcite , including coronagraphs cor1 and cor2 , and the extreme ultraviolet imager ( euvi , @xcite ) , the large angle and spectrometric coronagraph ( lasco , @xcite ) on board soho , the atmospheric imaging assembly ( aia , @xcite ) and the helioseismic magnetic imager ( hmi , @xcite ) on board the solar dynamics observatory ( sdo , @xcite ) . radio spectral observations from wind / waves @xcite are used to study interplanetary radio bursts .
we use h@xmath6 images from aryabhatta research institute of observational sciences ( aries ) @xcite and global oscillation network group ( gong ) for chromospheric flare information . for the radio emission analysis , the nobeyama radioheliograph ( norh ) images at 17 and 34 ghz
have been used @xcite .
radio spectral data from the hiraiso radio spectrograph ( hiras ) are used for metric type ii radio burst information .
interplanetary scintillation ( ips ) observations from the ooty radio telescope ( ort ) , rac , ncra - tifr has been used to study the interplanetary ( ip ) propagation of cmes @xcite .
we use ace and wind spacecraft data available at omni web ( http://omniweb.gsfc.nasa.gov ) for solar wind ( in - situ ) plasma ( temperature , density , speeds , plasma beta ) and magnetic field information .
information on cmes observed by lasco is basically taken from the online cme catalog ( http://cdaw.gsfc.nasa.gov , @xcite ) .
soft x - ray flare and sep information is obtained from geostationary operational environmental satellite ( goes ) .
the dst ( in nt ) index has been obtained from the world data center for geomagnetism , kyoto university , japan ( e.g. , http://wdc.kugi.kyoto-u.ac.jp/dstdir/ ) .
on january 23 , 2012 three solar flares ( c2.5 , m1.1 and m8.7 ) occurred , and the last two were associated with two major cmes .
the time line of the whole event is presented in table 1 .
the upper panel of fig .
1 shows the white light image and magnetogram of active region ar 11402 , which is the source region of the eruptions .
the active region was located at @xmath0 n28 w36 , and had a simple @xmath7 magnetic field configuration .
the bottom panel of fig . 1 shows the goes soft x - ray plot in the @xmath8 and @xmath9 wavelength bands during 00:00 ut to 09:30 ut .
the soft x - ray flux shows three flares ( indicated by the arrows ) : first flare ( c2.5 ) peaked around 01:44 ut , the second flare ( m1.1 ) around 03:13 ut and the third one ( m8.7 ) around 03:59 ut . since all three flares occurred in quick succession , it was not possible to determine the exact onset time of the m1.1 and m8.7 flares .
for the onset of the m1.1 flare we have used the time of the dip between the c2.5 and m1.1 flares . for the onset time of the m8.7 flare
we have used the time when the d@xmath10/d@xmath11 turns positive ( i is the soft x - ray intensity of the flare ) .
the m1.1 and m8.7 are each associated with a cme . the second cme is associated with large filament eruption . in the upcoming subsections we discuss the details on the flares , associated cmes and their interplanetary consequences .
2 shows images during the m1.1 flare in sdo / aia 304 , 171 and 94 wavelengths .
the 304 channel gives information about the upper transition region , while 171 and 94 provides information about the coronal region and the flaring region , respectively .
the charterstic log temperature ( in mk ) for 304 , 171 and 94 lines are around 4.7 , 5.8 and 6.8 .
the three columns show the images during the rise , maximum and the decay phase respectively .
it is clear from the observations that the flare starts with two bright sources in 304 wavelength ; one lies near the sunspot and other away from the sunspot .
overplotted magnetogram ( fig .
2e ) shows that the kernels lie on the opposite magnetic polarity regions .
we have overplotted the microwave sources ( i.e. , 17 and 34 ghz ) over the sdo / aia 304 images ( see figs . 2a , 2b and 2c ) .
these sources lie over the bright kernels of the flare , which are part of the flare ribbons .
post flare loops connecting the flare ribbons have been observed in aia 171 and 94 images around 03:29 ut ( figs .
2f and 2i ) .
the m1.1 flare was associated with the first cme .
the cme has an extended hot core ( @xmath126.3 mk ) which will be discussed later .
the long duration m8.7 flare starts with the eruption of the northern part of the active region filament .
3(a - c ) show the eruption of the filament in aia 304 wavelength .
3(d - f ) show the same filament in stereo - a euvi 304 wavelength .
we show in fig .
3 g a height - time measurement of the filament on aia 304 images .
the filament shows slow rise with an average speed of @xmath06 km s@xmath1 between 03:00 - 03:31 ut followed by a fast rise with an asymptotic speed of @xmath0590 km s@xmath1 during 03:40 - 03:46 ut . the average speed calculated from images in fig .
3(d - f ) is @xmath13280 km s@xmath1 , consistent with speed derived form sdo / aia observations of the filament corrected for projection effects during 03:26 ut to 03:46 ut .
g clearly shows that the rise of the filament triggers the soft x - ray flare .
4 shows the evolution of the m8.7 flare in the sdo / aia 304 , 171 and 94 wavelengths .
the first , second and third columns of this figure show the images from the rise , the maximum and the decay phases of flare respectively .
the flare starts with two bright kernels .
we have overplotted the 17 ghz and 34 ghz microwave flare sources over the aia 304 flare arcade during the rise , maximum and decay phases ( figs . 4a , 4b and 4c ) .
the microwave emission is from a subset of the flare loops observed in aia 304 .
the hmi magnetogram contours plotted over the sdo / aia 171 image ( fig .
4d ) , show that the flare ribbons are located at the opposite magnetic polarity regions of the active region .
the flare ribbons and loops can also be seen in gong and aries h@xmath6 images ( fig .
5 ) . in summary , both the m1.1 and m8.7 flares show typical flare arcades .
the filament eruption is clear in the m8.7 flare , whereas in the m1.1 flare there was a hot ejecta which became the cme core in coronagraphic images ( see details below ) .
the m1.1 and m8.7 flares were accompanied by cmes observed by soho and stereo coronagraphs .
stereo b and a were located at -114 degrees and + 108 degrees , respectively relative to earth .
stereo - a cor1 and cor2 observed the two cmes above the north - east limb .
we used stereo a data for height - time measurements , since the cmes were limb events in stereo - a view ( less projection effects ) .
6(a - c ) show the two cmes ( cme1 and cme2 ) from stereo - a cor1 and euvi 195 at 03:05 ut , 03:45 ut and 03:50 ut . we were able to image the cmes in their early phase because of the extended fov of stereo cor1 ( 1.4 - 4 r@xmath14 ) .
the height time plots of the cmes obtained using stereo - a cor1 and cor2 data give speeds of cme1 and cme2 as @xmath01400 km s@xmath1 and @xmath02000 km s@xmath1 , respectively ( fig .
the cmes occurred in quick succession : cme2 entered into the aftermath of cme1 at the first appearance of cme2 ( fig .
the cme interaction took place due to the difference in speed , their origin from the same active region and their propagation in roughly the same direction . the interaction completed at a height of 11 - 12 @xmath2 at @xmath13 04:48 ut ( fig .
6 g ) . a sequence of lasco c2 and c3 difference images in fig . 7 , shows the further evolution of the two cmes .
cme1 appeared in the lasco c2 fov at @xmath1303:12 ut at a position angle of 329@xmath15 with a width of 221@xmath15 .
the cme1 had a linear speed of 685 km s@xmath1 , which is much smaller than the speed estimated from the stereo measurements at lower heights .
this is likely to be due to projection effects because the solar source at n28 w36 in earth view , whereas the source was at the limb in stereo - a view .
cme1 had an extended core ( see fig .
7b ) . in cor1
fov the core was also seen as a bright feature , which can be traced in euv images as a hot ejecta .
this is somewhat unusual because normally the cme core is a cooler filament .
8a shows the hot ejecta ( dotted line ) in the aia 94 image at 02:22 ut . the hot ejecta was surrounded by cme1 ( solid line ) as can be seen in the aia 193 difference image .
this can also be seen in the stereo - a euvi 195 image at 02:35:30 ut ( fig .
cme2 appeared at a height of 2.01 @xmath2 in the cor1 fov at around 03:40 ut . at this time , the leading edge of the cme1 was at 5 @xmath2 .
when cme2 appeared in the lasco / c2 fov at 04:00 ut its leading edge had already reached the core of cme1 ( core1 ) ( see fig .
the linear speed of cme2 was @xmath02175 km s@xmath1 in the lasco fov , which is comparable to the stereo - a speed ( the speed difference is within measurement error ) .
the lasco cme also seems to be closer to the sky plane .
7(c - f ) shows that the two cmes merged at 04:12 ut and thereafter moved out as a single compound cme .
this can also be seen in the three frames of stereo - a cor2 shown in fig .
6(d - f ) .
even though a metric type ii bursts was not reported in the solar geophysical data our examination of the dynamic spectrum from the hiras radio spectrograph , shows a type ii burst feature between 03:40 and 03:50 ut .
given the high speed of cme2 it is unusual that the type ii feature is very weak .
this may be due to the fact that cme2 is running into the moving material of cme1 making it difficult to form a strong shock .
furthermore , the starting frequency of the type ii burst was at 200 mhz . assuming that the radio emission is at harmonic of the local plasma frequency , we see that the 200 mhz plasma level is at an unusually large height of 2.0 @xmath2 .
this may be due to the fact that the shock is formed in the body of cme1 .
the interplanetary counterpart of the type ii burst can be seen in the wind / waves dynamic spectrum shown in fig .
. an intense type iii burst , which is due to electrons accelerated at the flare site , can also be seen .
the dh type ii burst stats around @xmath1304:00 ut , which is roughly the time the leading edge of cme2 reaches the core of cme1 ( fig .
6 ( d - f ) and fig . 7(b ) ) .
at this time the shock seems to have become very strong due to the decline of the alfven speed upstream of the shock .
the type ii is observed as a broad band feature at lower frequencies ( fig .
9(a - b ) ) .
the same shock accelerated protons as indicated by the sep onset at @xmath1304:20 ut , which is about 40 minutes after the metric type ii start ( see fig .
however , if we consider the fact that @xmath12100 mev protons take about 15 minutes to reach earth , the sep release at the sun is close to the onset of the dh type ii bursts . as noted before cme2 has already interacting with the hot core of cme1 .
there is also a second increase in the sep intensity around 05:30 ut , which is right after the formation of compound cme .
it is possible that some particles were trapped in the interaction region and released after the interaction ends . in the @xmath1610 mev channel ,
the sep intensity reaches a plateau of about @xmath03000 pfu by 12:00 ut on january 23 , 2012 until the shock arrival at 1 au ( on january 24 , 2012 at 14:33 ut ) and causing an esp event ( see fig .
the esp event attained a peak proton flux of @xmath06263 pfu at 15:30 ut on january 24 , 2012 .
the sep event was the largest in solar cycle 24 @xcite as of this writing . the formation of shock in the near - sun region ( as shown by the type ii radio emission ) and shock signatures observed at 1 au imply that the compound cme was able to drive a shock and accelerate particles in the entire sun - earth distance . from the onset of cme2 in cor1 fov (
i.e. , 03:40 ut on january 23 , 2012 ) to the shock arrival at 1 au ( @xmath1314:33 ut on january 24 , 2012 ) , we see that the shock transit time is @xmath035 hour .
10 shows interplanetary scintillation ( ips ) data of the cme event on january 23 , 2012 observed by the ort .
the ips technique provides information on the turbulent plasma at the front of the moving cme .
the ips observations describe the evolution of the compound cme beyond 50 r@xmath14 . in the ips fov , between 100 and 225 @xmath2 ,
the cme decelerates from @xmath0900 to @xmath0700 km s@xmath1 .
the cme deceleration is mostly due to the interaction with the background solar wind flow .
11 shows the variation of interplanetary field parameters ( total magnetic field ( b ) , z component of magnetic field ( bz ) , electric field ( ey ) ) , plasma parameters ( solar wind speed ( v ) , proton density ( n ) , proton temperature ( t ) , plasma beta ( @xmath7 ) ) and dst index .
the shock arrival is at @xmath014:33 ut on january 24 , 2012 , which is followed by an extended sheath and a narrow interval of icme material .
the shock speed ( @xmath13700 km s@xmath1 ) is consistent with the speed derived from ips observations .
the plasma beta and magnetic field plots do not show the classical signatures of a magnetic cloud .
the narrow icme suggests that its impact on earth is at an angle rather than direct .
the dst index shows that the icme resulted in a moderate gms ( dst @xmath13 -73 nt ) .
a peak negative dst of -73 nt was reached at @xmath011:00 ut on january 25 , 2012 which corresponds to the main phase of the moderate gms .
actually there are two dips in dst indicating two different interactions with the earth magnetic field . first is due to the interaction of the sheath region with negative bz with earth s magnetic field ; the second one is due to the negative bz within the icme .
in this paper we presented the case study of the january 23 , 2012 solar eruptions from noaa ar 11402 ( n28 w36 ) that resulted in a large sep event and a moderate gms .
the eruptions consisted of two fast interacting cmes associated with m1.1 and m8.7 flares . we have shown that the interaction of the fast primary cme with the preceding cme resulted in the huge sep event .
the results presented in this study are consistent with the earlier suggestions that the interaction between cmes has important implications for large sep events @xcite .
the january 23 , 2012 events were observed by the stereo mission taking advantage of its extended fov much closer to the sun , capturing cme interaction in more detail .
it is significant that the metric type ii burst was very weak but the interplanetary type ii burst was very intense .
this may be attributed to different combinations of cme speed , alfven speed and the physical conditions of the ambient medium .
the temporal coincidence of the cme interaction with the onset of the intense interplanetary type ii burst and the large sep event is significant and suggests the plausible increase in the efficiency of particle acceleration .
one of the interesting features of these eruptions is the fact that the core of cme1 was unusually hot because it was observed only in the hot plasma images ( 94 and 193 ) .
the temperature exceeded 6 mk which is a few times the average temperature of the corona .
the sep event started when the shock of cme2 propagated through the hot ejecta .
the particles in the hot plasma were already energized by virtue of the high temperature , which makes it easy for the shock to accelerate these particles .
the cmes presented in this paper resulted in a huge sep event but not a strong gms .
one of the possible causes for the moderate gms may be that only a small section ( about 6 hours ) of the icme interacted with earth .
following the cme motion using movies of stereo cor2 images confirms that most of the compound cme propagated above the ecliptic .
this is also seen in the stereo and lasco images in figs . 6 and 7 .
in this paper we have described multi - wavelength observations of flares and cmes that are important in understanding how solar eruptions cause space weather .
deeper insights into the link between the solar origin of cmes and their interplanetary and geo - space consequences can be obtained by a detailed modeling of cme initiation and propagation through extreme ambient conditions .
+ the key results of this study can be summarized as follows : + 1 .
this study demonstrated that interaction between fast cmes from the same active region can occur very close to the sun as revealed by stereo observations .
the gms was of double - dip nature because the shock sheath and the icme contained negative bz values and caused the two dips in the dst index . +
* acknowledgments * we thank the anonymoud referees for their valuable comments and suggestions .
we thank iusstf / jc - solar eruptive phenomena/99 - 2010/2011 - 2012 project on `` multiwavelength study of solar eruptive phenomena and their interplanetary responses '' for its support to this study during our bilateral collaboration .
we acknowledge the wind / waves and sdo s aia and hmi teams for providing their data .
soho is a project of international cooperation between esa and nasa .
the stereo science center made available the data used in this work .
n.c.j thanks aryabhatta research institute of observational sciences ( aries ) , nainital for providing post doctoral grant .
this work was partly supported by nasa lws program .
rc , aks and wu acknowledge isro / respond project no .
isro / res/2/379/12 - 13 .
kaiser , m. l. , kellogg , p. j. , manning , r. , goetz , k. , monson , s. j. , monge , n. , friel , l. , meetre , c. a. , perche , c. , sitruk , l. , hoang , s. , waves : the radio and plasma wave investigation on the wind spacecraft .
_ space sci .
rev . _ 71 , 231263 , 1995 .
, g. e. , howard , r. a. , koomen , m. j. , korendyke , c. m. , michels , d. j. , moses , j. d. , socker , d. g. , dere , k. p. , lamy , p. l. , llebaria , a. , bout , m. v. , schwenn , r. , simnett , g. m. , bedford , d. k. , eyles , c. j. , the large angle spectroscopic coronagraph ( lasco ) .
_ solar phys .
_ 162 , 357402 , 1995 . , d. a. , cline , t. l. , desai , u. d. , mcdonald , f. b. , explorer 12 observations of solar cosmic rays and energetic storm particles after the solar flare of september 28 , 1961 .
_ j. geophys .
_ 67 , 49835000 , 1962 . , n. , energetic particle and other space weather events of solar cycle 24 . in : hu , q. ,
li , g. , zank , g. p. , ao , x. , verkhoglyadova , o. , adams , j. h. ( eds . ) , american institute of physics conference series .
1500 of american institute of physics conference series .
1419 , 2012 . , n. , yashiro , s. , akiyama , s. , mkel , p. , xie , h. , kaiser , m. l. , howard , r. a. , bougeret , j. l. , coronal mass ejections , type ii radio bursts , and solar energetic particle events in the soho era .
annales geophysicae 26 , 30333047 , 2008 .
, n. , yashiro , s. , michaek , g. , kaiser , m. l. , howard , r. a. , reames , d. v. , leske , r. , von rosenvinge , t. , interacting coronal mass ejections and solar energetic particles .
_ astrophys .
_ 572 , l103l107 , 2002 . , r. a. , moses , j. d. , vourlidas , a. , newmark , j. s. , socker , d. g. , plunkett , s. p. , korendyke , c. m. , cook , j. w. , hurley , a. , davila , j. m. , thompson , w. t. , st cyr , o. c. , mentzell , e. , mehalick , k. , lemen , j. r. , wuelser , j. p. , duncan , d. w. , tarbell , t. d. , wolfson , c. j. , moore , a. , harrison , r. a. , waltham , n. r. , lang , j. , davis , c. j. , eyles , c. j. , mapson - menard , h. , simnett , g. m. , halain , j. p. , defise , j. m. , mazy , e. , rochus , p. , mercier , r. , ravet , m. f. , delmotte , f. , auchere , f. , delaboudiniere , j. p. , bothmer , v. , deutsch , w. , wang , d. , rich , n. , cooper , s. , stephens , v. , maahs , g. , baugh , r. , mcmullin , d. , carter , t. , sun earth connection coronal and heliospheric investigation ( secchi ) .
_ space sci .
_ 136 , 67115 , 2008 . , n. c. , bankoti , n. s. , pande , s. , pande , b. , pandey , k. , relationship between interplanetary field / plasma parameters with geomagnetic indices and their behavior during intense geomagnetic storms .
_ new astronomy _ 16 , 366385 , 2011 . , s. w. , sheeley , jr . , n. r. , howard , r. a. , michels , d. j. , koomen , m. j. , mcguire , r. e. , von rosenvinge , t. t. , reames , d. v. , associations between coronal mass ejections and solar energetic proton events .
_ j. geophys .
_ 89 , 96839693 , 1984 . , j. r. , title , a. m. , akin , d. j. , boerner , p. f. , chou , c. , drake , j. f. , duncan , d. w. , edwards , c. g. , friedlaender , f. m. , heyman , g. f. , hurlburt , n. e. , katz , n. l. , kushner , g. d. , levay , m. , lindgren , r. w. , mathur , d. p. , mcfeaters , e. l. , mitchell , s. , rehse , r. a. , schrijver , c. j. , springer , l. a. , stern , r. a. , tarbell , t. d. , wuelser , j .-
wolfson , c. j. , yanari , c. , bookbinder , j. a. , cheimets , p. n. , caldwell , d. , deluca , e. e. , gates , r. , golub , l. , park , s. , podgorski , w. a. , bush , r. i. , scherrer , p. h. , gummin , m. a. , smith , p. , auker , g. , jerram , p. , pool , p. , soufli , r. , windt , d. l. , beardsley , s. , clapp , m. , lang , j. , waltham , n. , the atmospheric imaging assembly ( aia ) on the solar dynamics observatory ( sdo ) .
_ solar phys .
_ 275 , 1740 , 2012 . , g. , zank , g. p. , some aspects of particle acceleration and transport at cme - driven shocks . in : dere , k. , wang , j. , yan , y. ( eds . ) ,
coronal and stellar mass ejections .
226 of iau symposium .
332337 , 2005 . , y. d. , luhmann , j. g. , mstl , c. , martinez - oliveros , j. c. , bale , s. d. , lin , r. p. , harrison , r. a. , temmer , m. , webb , d. f. , odstrcil , d. , interactions between coronal mass ejections viewed in coordinated imaging and in situ observations .
_ astrophys .
_ 746 , l15 , 2012 . , h. , nishio , m. , enome , s. , shibasaki , k. , takano , t. , hanaoka , y. , torii , c. , sekiguchi , h. , bushimata , t. , kawashima , s. , shinohara , n. , irimajiri , y. , koshiishi , h. , kosugi , t. , shiomi , y. , sawa , m. , kai , k. , the nobeyama radioheliograph
ieee proceedings 82 , 705713 , 1994 . , n. v. , reames , d. v. , de rosa , m. l. , liu , y. , yashiro , s. , gopalswamy , n. , solar sources of impulsive solar energetic particle events and their magnetic field connection to the earth .
_ astrophys .
j. _ 650 , 438450 , 2006 .
, i. g. , lawrence , g. r. , haggerty , d. k. , kucera , t. a. , szabo , a. , are cme `` interactions '' really important for accelerating major solar energetic particle events ?
_ geophys .
_ 30 ( 12 ) , 1200001 , 2003 . ,
j. , scherrer , p. h. , bush , r. i. , wachter , r. , couvidat , s. , rabello - soares , m. c. , bogart , r. s. , hoeksema , j. t. , liu , y. , duvall , t. l. , akin , d. j. , allard , b. a. , miles , j. w. , rairden , r. , shine , r. a. , tarbell , t. d. , title , a. m. , wolfson , c. j. , elmore , d. f. , norton , a. a. , tomczyk , s. , design and ground calibration of the helioseismic and magnetic imager ( hmi ) instrument on the solar dynamics observatory ( sdo ) .
_ solar phys .
_ 275 , 229259 , 2012 . , f. , feng , x. s. , wang , y. , wu , s. t. , song , w. b. , guo , j. p. , zhou , y. f. , three - dimensional mhd simulation of two coronal mass ejections propagation and interaction using a successive magnetized plasma blobs model .
_ j. geophys .
_ 116 , 9103 , 2011 . , b. t. , smith , e. j. , gonzalez , w. d. , tang , f. , akasofu , s. i. , origin of interplanetary southward magnetic fields responsible for major magnetic storms near solar maximum ( 1978 - 1979 ) .
_ j. geophys .
_ 93 , 85198531 , 1988 . ,
b. , rudjak , d. , sudar , d. , gopalswamy , n. , kinematics of coronal mass ejections between 2 and 30 solar radii . what can be learned about forces governing the eruption ?
_ 423 , 717728 , 2004 . , s. , gopalswamy , n. , cliver , e. w. , reames , d. v. , kaiser , m. l. , howard , r. a. , association of coronal mass ejections and type ii radio bursts with impulsive solar energetic particle events . in : sakurai , t. , sekii , t. ( eds . ) ,
the solar - b mission and the forefront of solar physics .
325 of astronomical society of the pacific conference series .
, 2004 . , j .-
, lemen , j. r. , tarbell , t. d. , wolfson , c. j. , cannon , j. c. , carpenter , b. a. , duncan , d. w. , gradwohl , g. s. , meyer , s. b. , moore , a. s. , navarro , r. l. , pearson , j. d. , rossi , g. r. , springer , l. a. , howard , r. a. , moses , j. d. , newmark , j. s. , delaboudiniere , j .-
p . , artzner , g. e. , auchere , f. , bougnet , m. , bouyries , p. , bridou , f. , clotaire , j .- y . ,
colas , g. , delmotte , f. , jerome , a. , lamare , m. , mercier , r. , mullot , m. , ravet , m .- f . , song , x. , bothmer , v. , deutsch , w. , euvi : the stereo - secchi extreme ultraviolet imager . in : fineschi , s. , gummin , m. a. ( eds . ) , society of photo - optical instrumentation engineers ( spie ) conference series .
5171 of society of photo - optical instrumentation engineers ( spie ) conference series .
111122 , 2004 . , j. , richardson , i. g. , webb , d. f. , gopalswamy , n. , huttunen , e. , kasper , j. c. , nitta , n. v. , poomvises , w. , thompson , b. j. , wu , c .- c . , yashiro , s. , zhukov , a. n. , solar and interplanetary sources of major geomagnetic storms ( dst @xmath18 -100 nt ) during 1996 - 2005 .
_ j. geophys .
_ 112 , 10102 , 2007 . | we use multiwavelength data from space and ground based instruments to study the solar flares and coronal mass ejections ( cmes ) on january 23 , 2012 that were responsible for one of the largest solar energetic particle ( sep ) events of solar cycle 24 .
the eruptions consisting of two fast cmes ( @xmath01400 km s@xmath1 and @xmath02000 km s@xmath1 ) and m - class flares that occurred in active region 11402 located at @xmath0n28 w36 .
the two cmes occurred in quick successions , so they interacted very close to the sun .
the second cme caught up with the first one at a distance of @xmath011 - 12 @xmath2 .
the cme interaction may be responsible for the elevated sep flux and significant changes in the intensity profile of the sep event .
the compound cme resulted in a double - dip moderate geomagnetic storm ( @xmath3 ) .
the two dips are due to the southward component of the interplanetary magnetic field in the shock sheath and the icme intervals .
one possible reason for the lack of a stronger geomagnetic storm may be that the icme delivered a glancing blow to earth .
sun , solar flares , coronal mass ejections , solar energetic particles , geomagnetic storm |
in the first realizations of bose - einstein condensation ( bec ) in the laboratory @xcite and in many experiments ever since , the bose gas is trapped in a potential that can be considered as parabolic to a very good approximation . in the thermodynamic limit , within the ideal gas approximation , the critical temperature for such a system is given by @xmath0 , where @xmath1 is the geometric mean of the trap frequencies and all the other symbols have their usual meaning ( see e.g. @xcite ) .
soon after the first experiments , corrections to this expression , @xmath2 , were found . on the one hand , experiments do not take place in the thermodynamic limit .
hence , finite - size corrections are required . on the other hand ,
the gases are not ideal , having a non - vanishing scattering length .
hence , interaction effects must be taken into account .
the first order shift @xmath3 due to interactions was determined analytically early on in @xcite within a mean - field approximation , in the form of a linear term in the scattering length .
higher order corrections followed in several works @xcite , both numerical and analytical .
expansions for @xmath3 in powers of the scattering length down to second order were determined , both within a mean - field approach @xcite and accounting for critical correlations @xcite . the first order finite - size induced shift was given in the isotropic case in @xcite and in the general anisotropic case in @xcite as @xmath4 where @xmath5 .
more recently , a higher order result was given in @xcite ( see also @xcite ) .
this result relies on the local density approximation , in which the discrete energy levels of the finite system are approximated by a continuum , therefore requiring that the typical thermal energy at the transition be much greater than the typical inter - level spacing ( for example , in the case of an isotropic harmonic trap , @xmath6 ) , i.e. , the thermodynamic limit .
moreover , in order to overcome the vagueness ( or non - point - like character ) associated with the critical temperature of the finite system , we believe it would be useful to consider an explicit physical criterion for this critical temperature , related for example to the condensate fraction or the specific heat , when going to the level of detail of higher - order corrections @xcite . strictly speaking , a finite - size correction to @xmath7 is an ill - defined concept when taken on its own because the effect of finite size is to spread out the phase transition from a point to a narrow temperature interval .
the first order correction ( [ classic ] ) is typically extracted from a high temperature finite - size expansion of the number of particles , which takes into due account the discreteness of the energy levels and which can be obtained in several ways @xcite .
if one attempts to find a second order correction from this expansion , the absence of a true critical temperature makes itself noticed : the next order term in the expansion is divergent at the critical point , ultimately implying the non - existence of bec as a sharp , mathematically defined phase transition in finite systems .
it follows that the first order corrected @xmath8 must not be taken too seriously .
it merely provides a reference value for signaling the transition . in experimental work where the bec critical temperature is measured @xcite , the expression generally quoted for purposes of comparison with theory , namely for splitting off finite - size effects from interaction effects ,
is the one in ( [ classic ] ) .
now , as mentioned above , this expression should not be taken at face value .
thus , there is the possibility that a misinterpretation of the finite - size related shift can lead to a bias in the reported values of the interaction induced shift .
it would be of interest to make this matter clearer .
what is actually measured in experiments is the number of particles , ground state fraction , trap frequencies and temperature .
it is by performing some polynomial fit to a plot involving these quantities that an experimental value for @xmath8 is usually extracted @xcite . in the landmark experiment reported in @xcite
the fit is performed in the region where the condensate fraction `` noticeably starts to increase '' .
condensate fractions as low as about @xmath9 could be measured in this experiment .
if lower condensate fractions could be measured , higher critical temperatures would have been obtained , even rising above @xmath7 for sufficiently small condensate fractions .
this is because for finite systems the condensate fraction is not zero for temperatures above the critical region .
it is just very small .
this fact becomes more conspicuous for low particle numbers .
another major experiment in what concerns high precision measurements of @xmath8 is reported in @xcite . here , very much the same comments apply . in this case ,
condensate fractions as low as @xmath10 could be detected .
the authors overcome the problem of isolating interaction from finite - size corrections by performing differential measurements with reference to a standard value of the scattering length .
nevertheless , as recently pointed out @xcite , this assumes that finite size and interaction effects are independent . at second order
, it might not be the case .
our aim in the present work is to obtain higher - order finite - size corrections to the critical temperature of a bose gas in a general harmonic trap . to do this in a meaningful way , which at the same time can connect to experimental procedures
, we overcome the non - existence of a true critical temperature by asking instead for the temperature @xmath11 at which the condensate fraction has a given small value @xmath12 , @xmath13 .
other criteria could be used , like defining @xmath8 by the maximum of the specific heat or the inflection point of the @xmath14 curve ; but the one we adopt here is probably the most useful because it uses the condensate fraction and it is very simple . from the well known bulk behaviour of the condensate fraction in the bec regime , @xmath15
, we have in the thermodynamic limit @xmath16 . for @xmath17 ,
this yields @xmath18
. we will provide finite - size corrections to @xmath11 down to third order .
stopping at second order is not accurate enough in some circumstances , as detailed below .
our approach preserves all the finite - size characteristics of the system , with no approximations involved .
the information on the discrete structure of the energy levels is carried in the expansions ( [ condfraction ] ) and ( [ t_k ] ) below .
finally , we note that our expressions are also valid ( and highly accurate ) for @xmath19 not small , i.e. , deep into the bec regime .
let @xmath20 and @xmath21 .
@xmath22 is a rescaled inverse temperature and @xmath23 can be looked at as a rescaled chemical potential . we define the anisotropy vector @xmath24 . using grand - canonical statistics ,
the number of particles @xmath25 of an ideal bose gas in this trap is given by @xmath26 ^{-1 } = \sum_{\mathbf{n}}\sum_{k=1}^{\infty}e^{-kx(\bm{\lambda}\cdot\mathbf{n}+\epsilon ) } \ ; .
\label{nnew}\ ] ] the sum in @xmath27 is over all single particle states , of energy @xmath28 , @xmath29 .
let @xmath30 .
the usual bulk result for @xmath25 , which is exact in the thermodynamic limit , reads in our variable @xmath31 if @xmath32 and @xmath33 if @xmath34 ( where @xmath35 is the quantity that remains finite in the thermodynamic limit , as opposed to @xmath25 ) .
@xmath36 is the polylogarithm of index 3 , with the property @xmath37 .
define @xmath38 . as we approach the thermodynamic limit in the usual way ( @xmath39 kept fixed ) we have @xmath40 , or for any fixed temperature , @xmath41 .
@xmath22 and eventually @xmath42 will be our expansion parameters . in the bec regime , we have in addition ( still in the thermodynamic limit ) @xmath43 , from where we see that @xmath23 scales as @xmath44 .
what we need is an expansion for @xmath25 that contains the finite - size corrections and that is valid _ throughout _ the critical region .
this can be achieved by applying a mellin - barnes transform to the exponential inside the @xmath45 summation in ( [ nnew ] ) , as indeed was done before in @xcite .
the same procedure was also applied to a bose gas subject to other confinements @xcite .
an expansion is obtained by solving a contour integral in the complex plane using the theorem of residues . in this case , the riemann and three - dimensional barnes zeta functions , here denoted @xmath46 and @xmath47 respectively , make their appearance .
knowledge of the residues at the poles of these functions is required .
we refer the reader to @xcite for details of the procedure .
@xmath48 is a multi - dimensional generalization of the hurwitz zeta function , which was studied in depth by barnes in @xcite ( see also @xcite ) . in @xcite
the expansion for @xmath25 was calculated to subleading order . however , for our purposes we need also the third and fourth terms .
the calculation of the third term , in particular , is more involved due to the existence of a double pole , requiring the knowledge of the finite part at the @xmath49 pole of @xmath47 , not only its residue .
specifically , below we need the quantity @xmath50 defined in the following way .
let @xmath51 be the finite part at the @xmath52 pole of @xmath47 .
then @xmath53 , i.e. , @xmath54 .
@xmath55 is a function of @xmath56 only .
we obtain the expansion @xmath57
x^{-1 } -\frac{1}{2}{\zeta}_b(0,{\epsilon}|\bm{\lambda } ) + \mathcal{o}(x ) \ ; , \label{nexpansion}\end{gathered}\ ] ] where we have adopted the following notational conventions : @xmath58 and @xmath59 .
the first two terms in ( [ nexpansion ] ) were given in @xcite .
from @xcite we have that @xmath60 . the full asymptotic expansion for @xmath25 could easily be given , but it is not needed .
define the rescaled temperature @xmath61 . in ( [ nexpansion ] ) ,
change from the variables @xmath25 , @xmath22 and @xmath62 to @xmath42 , @xmath63 and @xmath23 by performing the substitutions @xmath64 and @xmath65 .
equation ( [ nexpansion ] ) gives us @xmath23 implicitly as a function of @xmath42 and @xmath63 .
since @xmath66 for @xmath67 , we solve for @xmath23 perturbatively by letting @xmath68 and find the coefficients @xmath69 , @xmath70 next we use the expression for the condensate fraction @xmath71 . in this expression
, we change again to the variables @xmath42 , @xmath63 and @xmath23 and substitute the newly found expansion for @xmath23 . expanding the resulting expression in powers of @xmath42 yields @xmath72 \\
\times x_0 ^ 2 + \frac{\lambda^3}{{\zeta}(3)}\left [ \frac{{\zeta}(2)}{{\zeta}(3)}\frac{t^3}{1-t^3}+\frac{(\lambda_i^2\lambda_j)}{48\lambda^3}-\frac{7}{16}\right]x_0 ^ 3+\cdots\ ; . \label{condfraction}\end{gathered}\ ] ] this equation gives us the condensate fraction as a function of @xmath63 and @xmath25 ( or @xmath63 and @xmath42 ) .
it is valid throughout the bec regime and critical region .
note that the first two terms , @xmath73 , are just the bulk result for @xmath74 in the condensate region .
we then set the condensate fraction at @xmath75 and solve ( [ condfraction ] ) perturbatively , this time to find @xmath76 as a function of @xmath19 and @xmath42 .
this finally yields @xmath77 with the coefficients being given by @xmath78\\ c'({\kappa})&=\frac{9+(\lambda_i\lambda_j)}{(1-{\kappa})^{1/3}36{\zeta}(3)}\\ d({\kappa})&=(1-{\kappa})^{-2/3}\left [ -\frac{{\zeta}(2)^3}{12{\zeta}(3)^3}+\frac{(\lambda_i^2\lambda_j)-21\lambda^3}{144{\zeta}(3 ) } + \frac{{\zeta}(2)}{6{\zeta}(3)^2}\right.\\ & \left .
\times \frac{9+(\lambda_i\lambda_j)}{12 } \left ( 1+\frac{1}{3}\ln ( 1-{\kappa } ) \right ) + \frac{{\zeta}(2)\lambda^3}{6{\zeta}(3)^2}\left ( b_0 - 2+\frac{2}{{\kappa}}\right ) \right ] \\
d'({\kappa})&=-\frac{{\zeta}(2)(9+(\lambda_i\lambda_j))}{36{\zeta}(3)^2(1-{\kappa})^{2/3}}\ ; .\end{aligned}\ ] ] note that since @xmath79 this is an expansion in powers of @xmath80 .
the leading term is just the bulk result for @xmath81 .
the subleading term is the well known first order finite - size correction to the critical temperature given in ( [ classic ] ) : @xmath82 .
the higher order terms are new .
it is interesting to note that when @xmath83 the coefficient @xmath84 , unlike the other coefficients above , diverges ( due to the @xmath85 in the denominator of the very last term ) .
we return to this point in the next section . as mentioned above , this expansion is valid not only in the critical region ( @xmath86 ) , but also throughout the bec regime ( @xmath85 not small ) . in the isotropic case , @xmath87 , @xmath88 , @xmath89 and , by writing @xmath90 in terms of hurwitz zeta - functions , it is easily seen that @xmath91 , where @xmath92 is euler s constant .
the @xmath93 and @xmath94 coefficients in ( [ t_k ] ) are then given more simply as @xmath95 where we have used the numerical values of @xmath96 and @xmath97 .
the subscript @xmath98 stands for``isotropic '' . in order to use ( [ t_k ] ) ( or for that matter ,
any of the previous expansions to more than subleading order ) in the case of an anisotropic trap , we must be able to find @xmath55 in the general case . from barnes
s work @xcite ( pp.398 and 404 ) ,
it is easy to arrive at @xmath99 , where the @xmath100 are gamma modular forms , which barnes gives quite generally in terms of contour integrals in the complex plane . application to our case yields @xmath101 \ ; , \label{b0integral}\end{gathered}\ ] ] where we have made use of the fact that @xmath102 , from the definition of @xmath56 .
table [ b0list ] presents values of @xmath55 in a few illustrative cases . in the experiment by gerbier _
et al _ @xcite , the trap is cigar shaped with aspect ratio @xmath103 .
the widely used trap of ensher _ et al _ @xcite is disc shaped with aspect ratio @xmath104 .
we include these two shapes in the table . for a more complete table ,
see the supplementary material , where we give values of @xmath55 for axially symmetric traps with integer aspect ratios ranging from @xmath105 to @xmath106 .
c@d .. 4d .. 4 & & + 2 & 0.1991 & 0.2433 + 3 & 0.9605 & 1.1317 + 5 & 3.0403 & 3.5180 + 10 & 10.8558 & 11.4467 + 30 & 73.7461 & 55.4503 + gerbier _ et al _ & & 102.0169 + ensher _ et al _ & 0.8147 &
in fig . [ fig : isotropic ] we plot @xmath107 for @xmath108 and @xmath109 in the isotropic case .
we plot both purely numerical results and our analytical results from ( [ t_k ] ) . since ( [ t_k ] ) is an expansion in powers of @xmath80 , its accuracy increases for larger @xmath25 . for small @xmath25
, the value of @xmath85 can not be chosen too small .
this is because in this situation we will have @xmath110 due to the spreading out of the phase transition , while we require @xmath111 .
this can also be seen by looking at the coefficient of the @xmath94 term in ( [ t_k ] ) , @xmath84 .
it contains a term with factor @xmath112 which becomes large if @xmath85 is very small .
in fact , if we carry on with the expansion , it is seen that at every third term a new factor @xmath112 will appear .
thus , the terms in @xmath94 , @xmath113 and @xmath114 contain the factor @xmath112 , the terms in @xmath115 , @xmath116 and @xmath117 contain the factor @xmath118 and so on .
it follows that this expansion is valid only if @xmath119 , or equivalently , @xmath120 .
this is a very reasonable condition .
it corresponds exactly to temperatures very close to @xmath7 or below it , confirming in this way the original requirement @xmath111 for the validity of the expansion .
if @xmath121 just holds but the tighter condition @xmath122 ( equivalently , @xmath123 ) does not , then the expansion is valid but the term in @xmath94 is of the same order as the @xmath42 term ( while the terms in @xmath124 , @xmath125 , will be of smaller order ) .
hence , it is very important to include it .
this limiting situation happens for example for @xmath126 and @xmath127 or for @xmath128 and @xmath129 .
all this is very well corroborated by comparing with numerical results as can be seen in the figure . for medium to large @xmath25 ,
the analytical and numerical curves are superimposed or hardly distinguishable , due to the high accuracy of ( [ t_k ] ) .
for example , for @xmath130 and @xmath108 or @xmath131 and @xmath109 the error in @xmath81 from ( [ t_k ] ) is less than @xmath132 ( in the isotropic case ) .
as we lower @xmath25 , the accuracy slowly decreases . as expected
, this happens more quickly for @xmath109 .
the broadening of the phase transition is observed in the rise of @xmath107 for low @xmath25 .
our approximation captures this behaviour . for comparison , we plot the usual first order result , given by eq .
( [ classic ] ) .
we see it provides a useful reference value for the transition for @xmath133 , even though it does not have a precise meaning . for lower values of @xmath25 ,
the new corrections are particularly important .
roughly , in most situations , the new corrections should be important for @xmath134 .
in such cases , if the interaction strength is small , it is quite possible that these effects become of the same order or even dominant over the interaction effects .
we plot @xmath81 for @xmath135 and axially symmetric disc shaped and cigar shaped traps .
@xmath81 is plotted as a function of the aspect ratio @xmath136 . for both shapes ,
the anisotropy causes a decrease in @xmath81 .
this effect is more pronounced for the disc shape , which can be understood from the fact that the dependency of @xmath137 ( and hence of @xmath42 ) on the aspect ratio @xmath136 is stronger in the disc case .
the effect seen in fig .
[ fig : isotropic ] where @xmath81 rises above @xmath105 for low @xmath25 is not seen in fig .
[ fig : anisotropic ] because anisotropy lowers the critical temperature without spreading out the phase transition , which remains sharp .
it can also be seen that the accuracy of ( [ t_k ] ) slightly decreases with increasing anisotropy .
this is expected and is due to the fact that @xmath137 decreases with increasing @xmath136 , which causes @xmath42 to increase . naturally , this is more noticeable for @xmath129 .
it is also more noticeable in the cigar shape case .
this should be due to the exact nature of the higher order terms ( @xmath113 and higher ) left out of our expansion ( [ t_k ] ) .
all in all , it can be seen that ( [ t_k ] ) is still quite accurate in most anisotropic situations ( the exception being the highly anisotropic cigar shaped trap with very small @xmath85 ) . as in the isotropic case ,
higher values of @xmath25 lead to higher accuracy .
we also plot the first order result from ( [ classic ] ) in fig .
[ fig : anisotropic ] .
again , we see that for this number of particles , eq . ( [ classic ] ) works well for providing a reference value for the transition .
let us now take the experimental conditions of smith _ et al _ @xcite : @xmath138 and a nearly isotropic trap .
condensate fractions as low as @xmath139 could reliably be measured in this experiment . for @xmath135 and @xmath109
we have @xmath140 from numerical calculations .
( [ t_k ] ) yields @xmath141 whereas the result from ( [ classic ] ) yields @xmath142 , which is lower than @xmath81 by @xmath143 of @xmath7 . in gerbier _
et al _ @xcite , the trap is cigar shaped with aspect ratio @xmath144 and @xmath145 at the transition .
condensate fractions as low as @xmath146 could be measured . for this trap shape with @xmath135 and @xmath108 , we have @xmath147 whereas ( [ t_k ] ) yields @xmath148 . the first order result ( [ classic ] )
yields @xmath149 , which is higher than @xmath107 by @xmath9 of @xmath7 . for @xmath150
this difference reduces to @xmath151 of @xmath7 .
if the trap had a larger anisotropy , the difference would be larger . in order to obtain the shift @xmath3 due to interactions
, the authors in @xcite subtracted the finite - size correction as given in ( [ classic ] ) from their experimental values for @xmath8 .
if the authors had used the lowest reliably detected condensate fraction ( @xmath146 ) for defining an experimental critical temperature , the finite - size shift that should be subtracted would be @xmath152 and not the one given in ( [ classic ] ) .
however , the procedure for obtaining this temperature was actually more complex than that , involving linear fits to the plots of @xmath153 , @xmath25 and @xmath154 as functions of the trap depth . the main feature of this idea can perhaps be understood by thinking of a linear fit applied to the experimental points near the @xmath108 region of the familiar @xmath155 curve . in this simplified version , the experimental @xmath8
would then be given by the point of intersection of the straight line of the linear fit with the horizontal ( @xmath154 ) axes .
due to the complexity of the whole procedure , it is not clear exactly what finite - size correction should be subtracted .
nevertheless , we see that an error of the order of 1% of @xmath7 ( not more ) could be involved in the determination of the interaction induced @xmath3 due to the use of eq .
( [ classic ] ) instead of a more precise expression .
, an accuracy of about @xmath151 or less in the reported values of @xmath63 should suffice .
this requirement seems to be quite realistic and , in particular , it seems to hold in the experiment of @xcite . ]
these examples illustrate the relevance of accurate analytical finite - size corrections , while suggesting the usefulness of better defined criteria for measuring the critical temperature , whenever finite - size effects play a role . in particular , in such cases it would seem to us perhaps more relevant to talk about @xmath107 , the temperature at which the condensate fraction is @xmath19 , rather than _ the _ critical temperature . for lower particle numbers ,
these considerations are even more important , as can be seen in fig .
[ fig : isotropic ] . on another note
, the use of bec experiments to probe planck - scale physics has been suggested in the last few years ( see @xcite and references therein ) .
the idea is that a quantum gravity effect could alter the single particle energy spectrum of the atoms in a harmonic trap , with a consequent shift in @xmath8 . for this effect to show
, we would ideally have a very weakly interacting gas and relatively small atom numbers due to finite size is estimated to be about @xmath156 , while the shift due to interactions is less than @xmath157 , but thermal equilibrium can not be assumed . ) ] .
finite - size effects would be crucial in such an experiment . in this context
, the need has been recognized @xcite for higher order finite - size corrections .
we have given these corrections in the present work .
this work took as a starting point a certain criterion for the bec critical temperature of a finite system .
as mentioned above , other physical criteria could be adopted . in principle
, it should be possible to apply the techniques of the present work to these alternative criteria .
finally , it would be interesting to study analytically the interplay between finite size and interaction effects .
it would not be surprising if there is a second order correction cross term containing a dependency on both the finite size and the interaction strength .
39 natexlab#1#1[1]`#1 ` [ 2]#2 [ 1]#1 [ 1]http://dx.doi.org/#1 [ ] [ 1]pmid:#1 [ ] [ 2]#2 , , , , , ( ) . . , , , ,
( ) . . , , , , , , , ( ) . . , , , , ( ) . , , , ( ) . . , , , ( ) .
, , ( ) . .
, , , ( ) . . , ( ) . , , , ( ) .
, , ( ) . ,
( ) . , , , , ( ) . , , , , ( )
, , , , ( ) .
, , , , ( ) .
, , , , , ( ) .
, , , ( ) . , ( ) .
, , , ( ) . , , ( ) .
, , , ( ) . , , , ( ) .
, , , , , ( ) . , , , , , , ( ) .
, , , , ( ) . , , , , , , ( ) .
, , ( ) . ,
( ) . , , ( ) . , ( ) . , , , , . , , ( ) . , , ( ) . . , ( ) . , , , , , , , , ( ) . | we obtain second and higher order corrections to the shift of the bose - einstein critical temperature due to finite - size effects .
the confinement is that of a harmonic trap with general anisotropy .
numerical work shows the high accuracy of our expressions .
we draw attention to a subtlety involved in the consideration of experimental values of the critical temperature in connection with analytical expressions for the finite - size corrections .
bose - einstein condensation , bose gas , finite - size effects 03.75.hh , 05.30.jp |
in the present paper , we consider the following cauchy problem for the derivative nonlinear schrdinger equation @xmath4 where @xmath5 is a complex - valued wave function , both @xmath6 and @xmath2 are real numbers . a great deal of interesting research has been devoted to the mathematical analysis for the derivative nonlinear schrdinger equations @xcite . in @xcite , c. e. kenig , g. ponce and l. vega studied the local existence theory for the cauchy problem of the derivative nonlinear schrdinger equations @xmath7 with small data @xmath8 in @xmath9 where @xmath10 is a polynomial having no constant or linear terms with the lowest order term of degree being greater than or equal to @xmath11 .
subsequently , it was improved to @xmath12 by n. hayashi and t. ozawa @xcite .
if the nonlinearity consists mostly of the conjugate wave @xmath13 , then it can be done much better . in the case
@xmath14 , a. grenrock , in @xcite , obtained local well - posedness when @xmath15 , @xmath16 , and @xmath17 was an integer .
in particular , the global well - posedness in @xmath18 is obtained when @xmath19 with the help of the bourgain spaces ( cf .
@xcite ) . in @xcite ,
h. takaoka discussed the derivative nonlinear schrdinger equation of the form @xmath20 and obtained the local well - posedness in @xmath21 for @xmath22 by performing a fixed point argument in an adapted bourgain space @xmath23 which yields a @xmath24-solution map .
a very similar equation to is the generalized benjamin - ono equation @xmath25 where @xmath26 is a real - valued function , @xmath27 is the hilbert transformation defined by @xmath28 and @xmath17 is an integer , the symbol @xmath29 ( or @xmath30 ) denotes the spatial fourier transform . for this equation ,
l. molinet and f. ribaud @xcite obtained the local well - posedness in the sobolev space @xmath21 for @xmath31 if @xmath32 , @xmath33 if @xmath34 and @xmath22 if @xmath2 by using tao s gauge transformation . in @xcite , c. e. kenig and h. takaoka have shown the global well - posedness for the case @xmath35 in @xmath21 for @xmath22 by combining the gauge transformation with a littlewood - paley decomposition and following the compactness argument with a priori estimates with the help of the preservation of the hamiltonian and the @xmath36-mass .
in the present paper , we shall generalize the above results to the derivative nonlinear schrdinger equation with @xmath2 by using some ideas in @xcite .
however , we have to reconstruct new and complicated estimates for the case @xmath2 which is quite different from the case @xmath35 .
we first state the main result of this paper as follows , though we shall define later the function space @xmath37 at the end of this section .
[ thm ] for any @xmath38 , there exist a @xmath39 and a unique solution @xmath26 of - satisfying @xmath40;h^{1/2})\cap x_t.\end{aligned}\ ] ] for convenience , we now introduce some notations . for nonnegative real numbers @xmath41 , @xmath42
, we use @xmath43 to denote @xmath44 for some @xmath45 which is independent of @xmath41 and @xmath42 .
@xmath46 means @xmath47 , and @xmath48 denotes @xmath44 for some small @xmath45 which is also independent of @xmath41 and @xmath42 . to give the littlewood - paley decomposition ,
let @xmath49 be a fixed even @xmath24 function with a compact support , @xmath50 , and @xmath51 for @xmath52 .
define @xmath53 .
let @xmath54 be a dyadic number of the form @xmath55 , @xmath56 or @xmath57 .
writing @xmath58 for @xmath59 , we define the convolution operator @xmath60 by @xmath61 , where the symbol @xmath62 ( or @xmath63 ) denotes the spatial fourier inverse transform .
we define the function @xmath64 by @xmath65 and denote @xmath66 .
then we introduce a spatial littlewood - paley decomposition @xcite @xmath67 throughout this paper , we often use the littlewood - paley theorem ( cf .
@xcite ) @xmath68 for @xmath69 .
we also use more general operators @xmath70 and @xmath71 which are defined by @xmath72 and @xmath73 , @xmath74 and @xmath75 which can be defined in a similar way .
the littlewood - paley operators commute with derivative operators ( including @xmath76 and @xmath77 ) , the propagator @xmath78 , and conjugation operations , are self - adjoint , and are bounded on every lebesgue space @xmath79 and homogeneous sobolev space @xmath80 if @xmath81 .
furthermore , they obey the following sobolev and bernstein estimates for @xmath82 with @xmath83 and @xmath81 ( which is similar to those of three dimensions @xcite ) : @xmath84 which can be verified by combining the bernstein multiplier theorem @xcite and the interpolation theorem of sobolev spaces .
we define the lebesgue spaces @xmath85 and @xmath86 by the norms @xmath87 ) } , \quad { \left\|f\right\|}_{l_x^pl_t^q}={\left\|{\left\|f\right\|}_{l_t^q([0,t])}\right\|}_{l_x^p({\mathbb r})}.\end{aligned}\ ] ] in particular , we abbreviate @xmath85 or @xmath86 as @xmath88 in the case @xmath89 .
we also use the elementary inequality @xcite @xmath90 for all @xmath91 and arbitrary functions @xmath92 , and the dual version @xmath93 where @xmath94 is the conjugate number of @xmath95 given by @xmath96 .
it is easy to verify that they also hold if we replace the norm @xmath85 by the norm @xmath97 in both side of the above inequalities .
let @xmath98 .
we use the fractional differential operators @xmath99 and @xmath100 defined by @xmath101 thus , we can introduce the resolution space . for @xmath102 , we define the function space @xmath37 in a similar way as in @xcite by @xmath103 where @xmath104
we transform the equation by introducing the following complex - valued function @xmath105 for a dyadic number @xmath54 given by @xmath106 by computation , we have @xmath107{\nonumber}\\ & -\frac{i{\lambda}}{2 } { e^{-\frac{i{\lambda}}{2}\int_{-\infty}^x{\left|{p_{\ll n}u}\right|}^k dy}}{p _
{ n}u}(i\partial_t-\partial_x^2)\int_{-\infty}^x{\left|{p_{\ll n}u}(t , y)\right|}^k dy{\nonumber}\\ & + \frac{{\lambda}^2}{4}{e^{-\frac{i{\lambda}}{2}\int_{-\infty}^x{\left|{p_{\ll n}u}\right|}^k dy}}{\left|{p_{\ll n}u}\right|}^{2k}{p _ { n}u}.\end{aligned}\ ] ] for the second term , we integrate by parts and have @xmath108dy\\ & -\int_{-\infty}^x\frac{i{\lambda}k}{2}{\left|{p_{\ll n}u}\right|}^{k-2}p_{\ll n}{\left|u\right|}^k\left(u_x+\bar{u}_x \right)dy -k{\left|{p_{\ll n}u}\right|}^{k-2}{p_{\ll n}u}\overline{{p_{\ll n}u}_x}\end{aligned}\ ] ] thus , @xmath109 obeys the following differential - integral equation @xmath110{\nonumber}\\ & -\frac{i{\lambda}k(k-2)}{8}{e^{-\frac{i{\lambda}}{2}\int_{-\infty}^x{\left|{p_{\ll n}u}\right|}^k dy}}{p _ { n}u}\int_{-\infty}^x { \left|{p_{\ll n}u}\right|}^{k-4}\cdot{\nonumber}\\ & \qquad\qquad\left[(\overline{{p_{\ll n}u}_x}{p_{\ll n}u})^2-({p_{\ll n}u}_x\overline{{p_{\ll n}u}})^2\right]dy{\nonumber}\\ & -\frac{{\lambda}^2 k}{4}{e^{-\frac{i{\lambda}}{2}\int_{-\infty}^x{\left|{p_{\ll n}u}\right|}^k dy}}{p _ { n}u}\int_{-\infty}^x{\left|{p_{\ll n}u}\right|}^{k-2}p_{\ll n } { \left|u\right|}^k\left(u_x+\bar{u}_x \right)dy{\nonumber}\\ & + \frac{i{\lambda}k}{2}{e^{-\frac{i{\lambda}}{2}\int_{-\infty}^x{\left|{p_{\ll n}u}\right|}^k dy}}{\left|{p_{\ll n}u}\right|}^{k-2}{p _ { n}u}{p_{\ll n}u}\overline{{p_{\ll n}u}_x}{\nonumber}\\ & + \frac{{\lambda}^2}{4}{e^{-\frac{i{\lambda}}{2}\int_{-\infty}^x{\left|{p_{\ll n}u}\right|}^k dy}}{\left|{p_{\ll n}u}\right|}^{2k}{p _ { n}u}{\nonumber}\\ \equiv & i_{n,1}(t , x)+i_{n,2}(t , x)+i_{n,3}(t , x)+i_{n,4}(t , x)+i_{n,5}(t , x).\end{aligned}\ ] ] the equivalent integral equation reads @xmath111(\tau ) d\tau.\end{aligned}\ ] ]
in order to prove the a priori estimate for the equation of @xmath109 , we need the linear estimates associated with the one - dimensional schrdinger equation .
we first recall the strichartz estimates , smoothing effects and maximal function estimates . for the proofs
, one can see @xcite .
[ lem4 ] for all @xmath112 , @xmath113 $ ] and @xmath114 , @xmath115 we also need the @xmath85 and @xmath86 estimates for the linear operator @xmath116 . for the proofs , one can see @xcite .
[ lem5 ] for @xmath117 , @xmath113 $ ] and @xmath118 , @xmath119 where @xmath94 is the conjugate number of @xmath120 $ ] , i.e. @xmath96 , and @xmath121 next , we recall the leibniz rule for a product of the form @xmath122 where @xmath123 is the spatial primitive of some function @xmath10 . for the proof , we refer to @xcite .
[ lem1 ] let @xmath124 , @xmath95 , @xmath125 , @xmath126 , @xmath127 , @xmath128 , @xmath129 $ ] with @xmath130 , @xmath131 , and let @xmath132 , with real - valued function @xmath10
. then @xmath133
in this section , we prove the following space - time estimate which is crucial to the proof of the nonlinear estimates .
[ prop1 ] let @xmath134 and @xmath135 be a real number .
then we have @xmath136 by the littlewood - paley decomposition , we can write @xmath137 now , we derive the estimates for @xmath138 , @xmath139 and @xmath140 , respectively . from the hlder inequality ,
the bernstein type inequalities and the real interpolation theorem , we have @xmath141 for @xmath139 or @xmath140 , it is suffice to consider one of them , e.g. @xmath139 , in view of symmetry . for @xmath142 , we have @xmath143 where @xmath144 .
we split these into three cases , i.e. @xmath145 , @xmath146 and @xmath147 . for the case
@xmath145 , from the hlder inequality and the littlewood - paley theorem , we can get @xmath148 for the case @xmath146 , we have , by the hlder inequality and the littlewood - paley theorem , that @xmath149 for @xmath150 , we have , by the sobolev embedding theorem , that @xmath151 where @xmath152 .
thus , can be bounded by @xmath153 now , we turn to the case @xmath147 . from the littlewood - paley theorem
, we have for @xmath135 @xmath154 noticing that @xmath155 and for @xmath156 , @xmath157 and @xmath135 @xmath158 we can bound by @xmath159 in view of the hlder inequality .
thus , we have obtained @xmath160 therefore , we have the desired result for any real number @xmath135 .
to state the estimates for the nonlinearities @xmath161 , we define the function space @xmath162 equipped with the following norm : @xmath163 we have the following proposition for the nonlinearities .
[ prop2 ] let @xmath26 be a @xmath164-solution to - .
then , @xmath165\\ & + t^{\frac{1}{2}}\left({\left\|u\right\|}_{x_t}^{2k-1 } + { \left\|u\right\|}_{x_t}^{(5k-2)/2}\right){\left\|p_{\gg 1}u\right\|}_{x_t}\\ & + t^{\frac{1}{4}}(1+t^{\frac{1}{4}}{\left\|u\right\|}_{x_t})^{\frac{1}{2}}{\left\|u\right\|}_{x_t}^{2k-1}{\left\|p_{\gg 1}u\right\|}_{x_t}^{\frac{3}{2}}\\ & + ( 1+{\left\|u\right\|}_{x_t}^k)\left[t^{\frac{1}{2 } } { \left\|u\right\|}_{x_t}^k{\left\|p_{\gg
1}u\right\|}_{x_t}+(1+t^{\frac{1}{4}}{\left\|u\right\|}_{x_t})^2{\left\|u\right\|}_{x_t}^{k-1}{\left\|p_{\gg 1}u\right\|}_{x_t}^2\right]\\ & + t^{\frac{1}{2}}{\left\|u\right\|}_{x_t}^{3k}{\left\|p_{\gg 1 } u\right\|}_{x_t},\end{aligned}\ ] ] where @xmath166 denotes the maximal integer that is less than @xmath167 ( i.e. @xmath168 $ ] if @xmath167 is not an integer and @xmath169 if @xmath167 is an integer where @xmath170 $ ] denotes the maximal integer that is less than or equal to @xmath167 ) .
we consider each nonlinearity separately . noting that the term @xmath172 has fourier support in @xmath173 , we have @xmath174 where @xmath175 . for the second term in , we have the following estimate .
[ lem2 ] let @xmath26 be a solution of - .
then , we have for any @xmath176 @xmath177 to shift a derivative from the high - frequency function @xmath178 to the low - frequency function @xmath179 , we require the following leibniz rule for @xmath60 from @xcite : @xmath180 thus , we have @xmath181 decomposing @xmath182 for @xmath183 , we have @xmath184 for the first term in , we have @xmath185 by the littlewood - paley theorem , we can obtain @xmath186 in the similar way , we have @xmath187 thus , @xmath188 for the last two term in , in a similar way as in the proof of proposition [ prop1 ] , we can obtain the following bound : @xmath189 from the sobolev embedding theorem and - , we obtain that can be bounded by @xmath190 thus , we can bound by @xmath191 which yields the desired result .
for the first term in , we have the following estimate : [ lem3 ] let @xmath26 be a solution of - .
then , we have for any @xmath176 @xmath192 we split into several terms for @xmath183 and @xmath176 @xmath193 notice that @xmath194 where @xmath195 is defined by @xmath196 .
thus , for the first term , from the fact @xmath197 and proposition [ prop1 ] , we have for @xmath176 @xmath198.\end{aligned}\ ] ] therefore , we obtain , for any @xmath176 , that @xmath199 for , in the same way as the case , we have @xmath200 now , we derive the estimate for by using the induction argument in @xmath167 . for @xmath201 , we have @xmath202 from the young inequality , the hlder inequality , and proposition [ prop1 ] , we can get for @xmath201 @xmath203.\end{aligned}\ ] ] from the triangle inequality for complex number , i.e. @xmath204 for @xmath205 , we can get @xmath206 for any @xmath207 $ ] . for @xmath208 $ ]
, we have @xmath209 then @xmath210{\left\|\bar{u}u_x\right\|}_{l_x^k l_t^2}{\left\|p_{\ll n}u\right\|}_{l_x^{2}l_t^\infty}\\ \lesssim & n^{-{\varepsilon}_k}\left[t^{\frac{1}{2}}{\left\|u\right\|}_{x_t}^4{\left\|p_{\gg 1}u\right\|}_{x_t}^{k-3 } + ( 1+t^{\frac{1}{4}}{\left\|u\right\|}_{x_t}){\left\|u\right\|}_{x_t}^3{\left\|p_{\gg 1}u\right\|}_{x_t}^{k-2}\right],\end{aligned}\ ] ] where @xmath211 for @xmath208 $ ] . by the same procedure
, we can obtain for any @xmath176 @xmath212,\end{aligned}\ ] ] where @xmath213 .
therefore , we have for any @xmath176 @xmath214 thus , we have proved the desired result . for the case
@xmath35 , the estimate also holds true .
in fact , we can follow the proof of ( * ? ? ?
* lemma 5.1 ) step by step and get @xmath215 from the proof of lemma [ lem3 ] , we can see that @xmath216 holds for any @xmath217 in view of proposition [ prop1 ] .
we turn to the proof of proposition [ prop2 ] for the nonlinearity @xmath171 .
we also consider the decomposition in . for convenience ,
we denote @xmath218 . from , , and , we have @xmath219 by lemma [ lem2 ] , the first term can be bounded by @xmath220 for the second term , we split the sum @xmath221 into three parts @xmath222 @xmath223 as in @xcite . for the part of @xmath224
, it is the same as lemma [ lem2 ] by summing in @xmath225 such that @xmath224 . for the part @xmath226
, we can add the projection operator @xmath75 to @xmath227 since @xmath228 has fourier support in @xmath173 .
thus , by the hlder inequality , we have @xmath229 where @xmath230 . by the bernstein inequality , we have @xmath231 and from and the hlder inequality
, we can get , as a similar way as in , that @xmath232 thus , can be bounded by @xmath233 for the part @xmath234 , we can add the projection operator @xmath235 to @xmath227 . in a similar way with the part @xmath226 , we have @xmath236 for the first term in , we denote it by @xmath237 , i.e. @xmath238 . similarly , from , , and , we can get @xmath239 from lemma [ lem3 ] , the first term is bounded by @xmath240 noticing that , and in the same way as in dealing with the second term of , we can bound the second term of by @xmath241 therefore , we have obtained @xmath242.\end{aligned}\ ] ] from , , and , we have @xmath244 where @xmath245dy$ ] . for the first term , from lemma [ lem1 ] and the hlder inequality ,
it can be bounded by @xmath246 by the hlder inequality , we have for @xmath2 @xmath247\right\|}_{l_{x , t}^1}{\nonumber}\\ \lesssim & { \left\|\overline{{p_{\ll n}u}}{p_{\ll n}u}_x\right\|}_{l_x^4l_t^2}^2{\left\|{p_{\ll n}u}\right\|}_{l_x^{2(k-4)}l_t^\infty}^{k-4}{\nonumber}\\ \lesssim & t { \left\|u\right\|}_{x_t}^{k}+(1+t^{\frac{1}{4}}{\left\|u\right\|}_{x_t})^2{\left\|u\right\|}_{x_t}^{k-2}{\left\|p_{\gg 1}u\right\|}_{x_t}^2,\label{eq21}\end{aligned}\ ] ] and from proposition [ prop1 ] and the proof of lemma [ lem2 ] , @xmath248\right\|}_{l_{x , t}^2}\\ \lesssim & { \left\|p_n u_x\right\|}_{l_x^\infty l_t^2}{\left\|{p_{\ll n}u}_x p_{\ll
n}\bar{u}\right\|}_{l_x^{2(k-1)}l_t^2}{\left\|{p_{\ll n}u}\right\|}_{l_x^{2(k-1)}l_t^\infty}^{k-2}\\ \lesssim & { \left\|p_n u_x\right\|}_{l_x^\infty l_t^2}\left[t^{\frac{1}{2 } } { \left\|u\right\|}_{x_t}^k+(1+t^{\frac{1}{4}}{\left\|u\right\|}_{x_t}){\left\|u\right\|}_{x_t}^{k-1}{\left\|p_{\gg 1}u\right\|}_{x_t}\right].\end{aligned}\ ] ] thus , we can bound by @xmath249 for , we split the sum @xmath221 into two parts @xmath250 , which gives the bound by @xmath251 for the first term , noticing that @xmath252 and , we can bound it by @xmath253 for the second term , in a similar way with , we bound it by @xmath254 therefore , we obtain @xmath255 from , , and , we have @xmath257 where @xmath258 . by hlder inequality , we get @xmath259 from the hlder inequality and proposition [ prop1 ] , we have @xmath260 in addition , for @xmath183 , we have @xmath261 . thus , in the same way as in the case @xmath243 , we can bound by @xmath262 from , , and , we have @xmath264 where @xmath265 . by the hlder inequality
, we have @xmath266{\left\|p_n u\right\|}_{l_x^{k}l_t^\infty}.\end{aligned}\ ] ] thus , the first term in can be bounded by @xmath267 by the hlder inequality , we get @xmath268{\left\|p_n u\right\|}_{l_x^{k}l_t^\infty}.\end{aligned}\ ] ] noticing that @xmath269 has fourier support in @xmath270 , we can repeat the procedure which we use to deal with the second term in , and obtain that the second term in can be bounded by @xmath271 therefore , we obtain @xmath272.\end{aligned}\ ] ] from , , and , we have @xmath274 where @xmath275 . by the hlder inequality , we have @xmath276 and @xmath277 thus , in a similar way as dealing with @xmath171 and @xmath263 , and noticing that @xmath278 has fourier support in @xmath270 , we can bound by @xmath279
by the scaling argument @xmath280 we have @xmath281 thus , we may rescale @xmath282 where we choose @xmath283 , and take the time interval @xmath284 depending on @xmath285 later .
we now drop the writing of the scaling parameter @xmath285 and assume @xmath286 we now apply this to the norms @xmath37 and @xmath3 , and define new version of the norms of @xmath37 and @xmath3 , given by with the decomposition @xmath287 , @xmath288 and @xmath289 which implies that @xmath290 . for the low frequency part
, we have the following estimates . [ lem6 ]
let @xmath26 be a solution of - . then @xmath291 using the integral equation of @xmath292 and by , , , , , and the hlder inequality , we have @xmath293 which is the desired result . for the high frequency part , we have [ lem7 ] let @xmath26 and @xmath109 be given in
. then @xmath294 by , we have @xmath295 for @xmath296-norm , by the interpolation theorem , we obtain for @xmath183 , @xmath297 which yields the desired estimate by summing on @xmath298 . for the @xmath299-norm , noticing that @xmath300 we have @xmath301 to estimate the second term , we split the sum @xmath302 . for @xmath303 , from the bernstein inequality ,
we bound by @xmath304 for the part @xmath305 , we split it as @xmath306 . noticing that for @xmath307 , @xmath308 and for @xmath309 , @xmath310 we can bound , in view of the bernstein inequality and the hlder inequality , by @xmath311 therefore , summing on @xmath298 , we complete the proof for the @xmath299-norm .
for the @xmath312-norm , it is easy to obtain the desired result since @xmath313 .
we turn to estimate the @xmath314-norm .
it is similar with the proof for the @xmath299-norm , since @xmath315 for @xmath183 .
in fact , we have @xmath316 we also split @xmath302 . for @xmath303 , we bound by @xmath317 for the part @xmath305 , we split it as @xmath306 . noticing that for @xmath307 , @xmath318 and for @xmath309 , @xmath319 we can bound , in view of the bernstein inequality and the hlder inequality , by @xmath320 which yields the desired estimate by applying @xmath298-sum .
thus , we complete the proof of this lemma .
of course , we need the following estimate of the data .
[ lem8 ] for any @xmath38 , we have @xmath321 from , and , we bound the left hand side of by @xmath322 from lemma [ lem1 ] , we have @xmath323\right)^{1/2}\\ \lesssim & ( 1+{\left\|u_0\right\|}_{h^{1/2}}^2){\left\|p_{\gg 1}u_0\right\|}_{h^{1/2}}.\end{aligned}\ ] ] for the second term , it is similar with .
we split the sum @xmath324 . by the bernstein inequality , the hlder inequality and the sobolev embedding theorem , we bound by @xmath325 which yields the desired result . with the help of the above lemmas
, we can prove the following proposition which yields the a priori estimate .
[ prop3 ] let @xmath26 be a smooth solution to - and @xmath326 .
then we have @xmath327 noticing that @xmath328 and from lemmas [ lem6 ] , [ lem7 ] , [ lem8 ] and proposition [ prop2 ] , we obtain through a complicated computation @xmath329\\ & + t^{\frac{1}{2}}\left({\left\|u\right\|}_{x_t}^{2k-1 } + { \left\|u\right\|}_{x_t}^{(5k-2)/2}\right){\left\|p_{\gg 1}u\right\|}_{x_t}\\ & + t^{\frac{1}{4}}(1+t^{\frac{1}{4}}{\left\|u\right\|}_{x_t})^{\frac{1}{2}}{\left\|u\right\|}_{x_t}^{2k-1}{\left\|p_{\gg 1}u\right\|}_{x_t}^{\frac{3}{2}}\\ & + ( 1+{\left\|u\right\|}_{x_t}^k)\left[t^{\frac{1}{2 } } { \left\|u\right\|}_{x_t}^k{\left\|p_{\gg 1}u\right\|}_{x_t}+(1+t^{\frac{1}{4}}{\left\|u\right\|}_{x_t})^2{\left\|u\right\|}_{x_t}^{k-1}{\left\|p_{\gg 1}u\right\|}_{x_t}^2\right]\big\}\\ { \leqslant}&c+c(c_{low})t^{\frac{1}{2}}{\left\|u\right\|}_{\tilde{x}_t}^{k+1 } + ( 1+{\left\|u\right\|}_{l_t^\infty h_x^{1/2}}^{2k})(1+{\left\|u_0\right\|}_{h^{1/2}}^k)\\ & + c(c_{low})(1+{\left\|u\right\|}_{l_t^\infty h_x^{1/2}}^{2k})\big\{t^{\frac{1}{4}}{\left\|u\right\|}_{\tilde{x}_t}^{3k+1}+t^{\frac{1}{4 } } ( 1+{\left\|u\right\|}_{\tilde{x}_t}^k){\left\|u\right\|}_{\tilde{x}_t}^{k+1}\\ & + t^{\frac{1}{2}}({\left\|u\right\|}_{\tilde{x}_t}^{k+1 } + { \left\|u\right\|}_{\tilde{x}_t}^{5k/2 } ) + c_{high}^{k-\tilde{k}}(1+t^{\frac{1}{4}}{\left\|u\right\|}_{\tilde{x}_t})(1+{\left\|u\right\|}_{\tilde{x}_t}^k ) { \left\|u\right\|}_{\tilde{x}_t}^{k+1}\\ & + c_{high}^{3/2}(1+t^{\frac{1}{4}}{\left\|u\right\|}_{\tilde{x}_t}){\left\|u\right\|}_{\tilde{x}_t}^{2k } + c_{high}^{2}(1+t^{\frac{1}{4}}{\left\|u\right\|}_{\tilde{x}_t})^2(1+{\left\|u\right\|}_{\tilde{x}_t}^k ) { \left\|u\right\|}_{\tilde{x}_t}^{k+1}\big\}.\end{aligned}\ ] ] notice that @xmath330 the high frequency part @xmath331 can be absorbed into the @xmath332-norm . then substituting lemma [ lem6 ] again in estimating the low frequency part of the norm @xmath333
, we complete the proof of proposition [ prop3 ] . from proposition [ prop3
] , we have the following a priori estimate for the solution of - if we take @xmath284 and @xmath334 small enough .
let @xmath26 be a smooth solution to - .
we have @xmath335 for @xmath284 and @xmath334 small enough . for the proof of theorem [ thm ]
, we can follow the compactness argument with the a priori estimate . since the proof is standard
, we omit the details and refer to the papers @xcite .
the author would like to thank the referees for valuable comments and suggestions on the original manuscript and prof .
l. hsiao for her frequent encouragement .
j. colliander , m. keel , g. staffilani , h. takaoka , t. tao , _ global well - posedness and scattering for the energy - critical nonlinear schrdinger equation in @xmath336 _ , to appear in annals math . ,
` arxiv : math.ap/0402129 ` . | in this paper , we investigate the one - dimensional derivative nonlinear schrdinger equations of the form @xmath0 with non - zero @xmath1 and any real number @xmath2 .
we establish the local well - posedness of the cauchy problem with any initial data in @xmath3 by using the gauge transformation and the littlewood - paley decomposition .
chengchun hao ( communicated by gigliola staffilani ) |
gotthilf heinrich ludwig hagen is most renowned for his contributions to the study on laminar flow in pipes ; his measurements published in 1839 studied what is now well - known as the ( hagen-)poiseuille law @xcite .
less well - known is hagen s work on granular systems . while janssen , with his 1895 paper , typically receives credit for the saturation effect in granular silos @xcite , it was hagen in his paper _ ber den druck und die bewegung des trocknen sandes _
@xcite who measured this effect earlier but also not for the first time , cf .
@xcite and offered a first model that provided a qualitative understanding of the effect .
hagen proposes a quadratic law ( with some cutoff ) for the pressure instead of the exponential form put forward by janssen more than 40 years later @xcite .
in addition to the discussion of a static pile of sand , hagen examines in considerable detail the flow through an opening of the container , and discovers that the rate of discharge is proportional to the diameter of the opening raised to the power @xmath0 .
this result is elegantly derived from dimensional analysis ( * ? ? ?
* ch 10.2 ) ; but it fits the data best if instead of the real diameter some effective diameter is used the resulting law is known as the beverloo correlation @xcite .
hagen finds an effective opening diameter that is smaller by twice the particle diameter , which is consistent with more recent measurements .
hagen s work lays out the basis of the so - called hourglass theory where the flow of granular material is found to be independent of the filling height in the container , thus allowing the measure of time with an hourglass ( * ? ? ?
* sec 10.4 ) .
later work both confirms and extends hagen s early analysis @xcite .
more recently , a lot of work has been devoted to understanding the fundamental differences between fluid and granular flows @xcite . while on the level of the individual grains the probability of arching at the opening is a matter of current investigations @xcite
, hagen provides a successful route to a continuum description of the flow by the rescaling of the diameter of the opening .
consider a container with a horizontal bottom that includes a circular opening of radius @xmath1 . in this opening
is placed a disk which is easily movable but seals tightly ; on top of it is an extended filling of sand up to a height @xmath2 . as a result , there is a pressure exerted on the disk created by the weight of the cylinder of sand above it less the friction which is experienced by that cylinder from the sand surrounding it .
the friction is proportional to the horizontal pressure , or the square of the height .
let @xmath3 denote a friction dependent constant and @xmath4 be the weight of a unit volume of sand , then the pressure against the disk equals for a growing @xmath2 , this expression will increase in the beginning , reach a maximum , and subsequently decrease afterwards ; it will become not only zero but even negative .
however , the sand cylinder is not rigidly connected throughout , and therefore the axial pressure , which its lower part exerts on the bottom disk , can not be compensated by the strong friction acting on the cylinder as a whole . hence , the pressure on the disk in fact remains unchanged for fillings higher than that height at which the pressure on the disk reaches its maximum value .
for the case of pressures below the maximum , we seek to represent that pressure by the weight of a free - standing body of sand on the disk .
the body is bounded by the surface of the filling .
it is a conoid that is formed by the rotation of a parabola around its axis . here
the parameter of the parabola is @xmath6 , while the height of the paraboloid is @xmath7 .
the latter body joins the circumference of the opening . to compare these results with the real phenomenon , i created openings of radii of 0.3791 inch @xcite . ] and 0.7271 inch , respectively , in two brass plates used as the the bottom of the sand - filled container ; these openings were closed with suitable disks that were supported from underneath by hooks that were connected to one arm of a balance , while the other arm carried the counterweight . to reduce the counterweight to the pressure on the disk slowly and without concussion , sand was allowed to discharge through a small hole in the bottom of the plate .
the constant error of this method could be found easily by measurement of the excess weight of the disk and the hook compared to the plate both by the discharge of sand and by direct weighing .
while repeating measurements of the pressure due to the the sand against the disks multiple times , considerable deviations among the measurements were observed ; this was apparently due to different ways of settling .
when the settling was as loose as possible , the weight of a cubic inch of the sand , a crude ferrous grit , was 2.9 loth @xcite . ] .
however , the weight increased to 3 loth when there was modest agitation during filling and rose to 3.25 loth as soon as a fairly compact settling was generated by severe agitation or by pushing a wire into the container . in so doing , the friction increased by even more than the specific weight . hence , the pressure against the disk became remarkably smaller for the more compact settling . with the larger disk
the pressure maximum was reached at a filling height of around 1 inch : for larger height the pressure decreased somewhat , since despite great care the sand settled in a slightly denser state . for the loosest fillings i found @xmath8 to 0.175 .
in contrast , the deviations were less when the sand was dropped from a height of several inches in a narrow stream , and when the sand flow was cone - shaped : the value for @xmath3 was limited to between 0.21 and 0.22 .
the influence of different settling states was also clearly noticeable as the sand discharged through the openings of various radii .
when the settled filling was somewhat denser , there was less sand flowing within a second .
as the discharge duration increased , the sand became agitated , particularly in the vicinity of the opening , resulting in an increased sensitivity , which in turn led to a diminished flow .
incidentally , the height of the filling had no influence , as was already recognized by huber - burnand some time ago @xcite .
to reduce the mentioned irregularities as much as possible , i limited the duration of each observation to 30 to 200 seconds and , additionally , tried to make the fillings rather uniform . to this end
, the container was placed in a metal - sheet cylinder with a sieve - like bottom , filled with sand , and lifted slowly thereafter ; in this way the sand poured into the container in several hundred thin streams , each from a very small height .
the radius reduction @xmath9 for the discharge opening is found to be close to the diameter of a grain of sand . from the constant @xmath10
can be determined the average distance from the opening at which the sand begins its free fall .
assuming that the sand forms a compact mass until reaching the opening , the amount of sand discharged per second is @xmath11 where @xmath2 designates the mentioned height of fall and @xmath12 the effective opening .
on the other hand , the observations yield @xmath13 equating both expressions and setting @xmath4 to 2.93 as obtained from the average loose packings , one finds @xmath14 if , in contrast , one assumes that for each unit of time a layer of sand of the same vertical height separates from the whole inner area of the paraboloid mentioned above in a free fall , then one can easily find the average velocity of this layer while passing the opening , and from the latter the average height of fall of the entire mass .
this height is @xmath15 but from the data for loose packings it was found that @xmath16 and hence @xmath17 if instead one introduces the value @xmath18 which is valid for packings where sand is flowing sideways , which really happens during the discharge , then @xmath19 the result derived from the observations is in between the two when @xmath1 is interchanged with @xmath12 .
this interchange is necessary because the sand only hits the edge of the opening when in motion ; in contrast , when at rest all the sand grains encountering the movable disk also load it .
this confirms the assumption from above that the free fall of the sand starts on the surface of the paraboloid ; it also explains that the amount of sand flowing through the opening is proportional to the power @xmath20 of the effective radius of the opening .
in the four inch wide and ten inch tall container , above the outlet , the entire surface of the sand packing subsided uniformly in the beginning . only gradually did a dip form vertically above the outlet .
the dip grew continually , and above its sides sand fell down .
concurrently , at the rim of the container a ring - shaped , almost horizontal surface remained .
this surface also subsided , but without the granules of sand experiencing strong sideways motion .
the flat ring gradually assumed a smaller width and disappeared completely when the funnel - shaped dip reached the outlet . from this
it follows that the sand flows not only vertically towards the outlet but also along concentric inclined trajectories , and that the motion extends up to a slope , which a free surface of sand can exhibit .
the motion in the inner part of the sand mass revealed itself very explicitly when i filled sand in a container having side walls made from a glass panel .
since this glass panel touched the outlet , one could follow the motion of single grains of sand down to the opening .
the strongest flow formed vertically above the outlet ; in fact the sand granules approached it with increasing yet moderate speed until , directly above , they were accelerated in a way that they could no longer be seen .
nevertheless , the sand also flowed inwards from the side of the outlet , but this motion was interrupted frequently and only occurred periodically , presumably due to friction at the glass . underneath the opening ,
the stream of sand was not nearly as sharply bounded as a water - jet ; rather , it was surrounded by single granules that from time to time departed as far as several lines . because of this the streams flowing from the larger outlets showed a significant reduction in their diameters , extending about 2 inches deep .
in addition , the measurement showed that even immediately under the disk the stream is already much weaker than at the outlet .
the outlet was 0.335 inch in diameter while the stream was only 0.29 inch at a distance of @xmath21 lines , and contracted to 0.27 inch at greater depth .
the reason for this effect is not the effective reduction of the opening mentioned above , since this would only explain the weakening of the stream by about 2% of an inch ; instead , the sand flowing sideways continues its motion towards the axis even after having passed the outlet , and the granules hitting the rim of the opening are also reflected towards the axis .
the stream of sand hence experiences a contraction similar to the stream of a liquid ; and when one compares the diameter of the opening with the smallest diameter of the stream , the ratio appears as @xmath22 or for the ratio of cross sections @xmath23 this agrees closely with the known contraction ratios for liquid streams leaving openings in thin walls . | in a remarkable paper from 1852 , gotthilf heinrich ludwig hagen measured and explained two fundamental aspects of granular matter : the first effect is the saturation of pressure with depth in a static granular system confined by silo walls generally known as the _ janssen effect_. the second part of his paper describes the dynamics observed during the flow out of the container today often called the _ beverloo law _ and forms the foundation of the _ hourglass theory_. the following is a translation of the original german paper from 1852 . |
spin - glasses are simple models for disordered systems .
they can be defined very easily in mathematical terms ; on the hand , the properties of such simple models usually are quite rich .
statistical physics of spin - glasses has been studied for more than thirty years since the concept of spin - glasses was first presented by edwards and anderson @xcite in 1975 , but there are still many unsolved and heavily debated issues . in the last decades
there have been a lot of theoretical investigations concerning models defined on a finite - connectivity random graph .
these later models are more realistic than conventional spin - glass models ( e.g. , the sherrington - kirkpatrick model @xcite ) on a complete graph , in the sense that each spin interacts only with a finite number of other spins . as direct analytical studies of spin - glass models on three - dimensional ( regular ) lattices are still beyond reach , people hope that a deep understanding of ( mean - field ) models on finite - connectivity random graphs will shed much light on the properties of 3d systems . for a spin - glass system of very large size
, it is generally believed that ergodicity is broken when the environmental temperature @xmath0 becomes lower than a certain value @xmath5 , the spin - glass transition temperature . in this low - temperature spin - glass phase
, the time average of a given physical quantity no longer equals to the ensemble average .
the whole configurational phase space of the system breaks into many separated domains .
ergodicity is still preserved within each such configurational domains .
if the system initially is in a spin configuration that belongs to domain @xmath1 of the configurational space , it will eventually visit all other configurations in this domain as time elapses . on the other hand , due to the existence of very high free energy barriers , the system is unable to transit spontaneously from one configurational space domain to another different domain . in this ergodicity - broken situation ,
each such configurational phase space domain is regarded as a _
macroscopic state _ or _
thermodynamic state_. the free energy @xmath6 of a macroscopic state @xmath1 is related to the microscopic configurations in @xmath1 through the following fundamental formula of statistical mechanics , @xmath7 where @xmath8 is the total number of spins in the system ; @xmath9 is the inverse temperature ( we set boltzmann s constant to unity throughout this paper ) ; and @xmath10 is the partition function for the macroscopic state @xmath1 as defined by @xmath11 \ .\ ] ] in eq .
( [ eq : partitionfunctionalpha ] ) , @xmath12 denotes a microscopic spin configuration , and @xmath13 is the total energy of this configuration .
the summation in eq .
( [ eq : partitionfunctionalpha ] ) is over all those microscopic configurations belonging to macroscopic state @xmath1 .
for a given spin - glass system , although we can formally write down the expressions of the partition function and free energy for a macroscopic state @xmath1 , the principal difficulty of spin - glass statistical physics is that we do not know _ a priori _ how the configurational space of the system is organized and which are the constituent spin configurations of each macroscopic state . to overcome this difficulty , one possibility is to first assume certain structural organization of the system s configurational space and then try to derive a self - consistent theory . for the sherrington - kirkpatrick model ,
the full - step replica - symmetry - broken ( frsb ) theory of parisi with an ultrametric organization of macroscopic states @xcite has met with great success . for finite - connectivity mean - field spin - glasses ,
a cavity approach was also developed by mzard and parisi @xcite .
this cavity approach combined the bethe - peierls approximation @xcite for a ferromagnetic ising model with the physical picture that there is a proliferation of macroscopic states in a spin - glass system .
this approach , which has been shown @xcite to be equivalent to the first - step replica - symmetry - broken ( 1rsb ) replica theory , can give very good predictions concerning the low - temperature free energy density of a system on a random graph .
later on , this cavity approach was extended by mzard and parisi @xcite to the limiting case of zero temperature .
the zero - temperature cavity method was applied to some hard combinatorial optimization problems with good performances ( see , e.g. , refs .
these interdisciplinary applications in return also call for further understanding on the mean - field cavity approach and its possible extensions ( see , e.g. , refs
. @xcite ) . for a spin - glass system at a low temperature @xmath0
, the total number of macroscopic states with a given free energy @xmath14 is denoted as @xmath15 .
although it is natural to anticipate that the logarithm of @xmath15 should be an extensive quantity , the exact relationship between @xmath16 and @xmath14 is unknown and is system - dependent .
inspired by parisi s frsb solution , in the cavity approach of mzard and parisi @xcite one assumes that the total number @xmath17 of macroscopic states with free energy @xmath14 diverges exponentially with @xmath14 with respect to a reference free energy @xmath18 @xcite , i.e. , @xmath19 where @xmath20 is a dimensionless constant to be determined self - consistently .
the parisi parameter @xmath21 stems from the replica theory of infinite - connectivity spin - glasses @xcite , its physical meaning in the cavity framework is not transparent . for some spin - glass systems with many - body interactions , even if the exponential form of eq .
( [ eq : mp01 ] ) is valid in certain range of free energy values , the parameter @xmath21 in this equation may exceed unity . the original cavity iterative equations of ref .
@xcite diverge for this situation of @xmath22 , where the statistical physical property of the spin - glass system is not determined by those macroscopic states with the globally minimal free energy density .
since the cavity approach of mzard and parisi is a very useful theoretical tool in studying the low - temperature properties of many spin - glass systems , it might be helpful for us to interpret the approach from an another slightly different angle .
complementary interpretations of the 1rsb cavity theory will also facilitate its further development . in this paper
we demonstrate that , the 1rsb cavity formalism of mzard and parisi can be derived in an alternative way without using eq .
( [ eq : mp01 ] ) .
this slightly revised cavity theory is based on the gedanken experiment of repeated temperature heating - annealing .
the final macroscopic state reached at the end of an annealing process is anticipated to follow the boltzmann distribution of free energies .
this theoretical approach is applied to the @xmath3 spin - glass model on a random regular graph of degree @xmath4 to test its validity .
the results of the present work are in close agreement with earlier simulational and numerical results .
the mathematical format of the present cavity theory is the same both for non - zero temperatures and for zero temperature .
if the macroscopic states of a spin - glass system further organize into clusters of macroscopic states , it can be easily extended to take into account this situation .
this paper is organized as follows .
the next section introduces the @xmath3 spin - glass model and the ensemble of random regular graphs of degree @xmath23 .
section [ sec : mfep ] describes the free energy boltzmann distribution of macroscopic states . in sec .
[ sec : cavity - field - distribution ] the concepts of cavity field and cavity magnetization are re - introduced , and the distribution of a vertex s cavity magnetization among all the macroscopic states is calculated .
various thermodynamic quantities , including the grand free energy density of the whole system and the free energy density of a macroscopic state , are calculated in sec .
[ sec : densities ] and sec .
[ sec : single - instance ] for an ensemble of systems and for a single system , respectively . the numerical results for the @xmath3 spin - glass model
are reported and analyzed in sec .
[ sec : numerics ] .
we conclude the present work and discuss its possible extensions in sec .
[ sec : discussion ] .
the appendix gives an explicit expression for the mean energy density .
in this paper we focus on just a single example , the @xmath3 spin - glass model @xcite on a random regular graph of degree @xmath23 .
this model was also studied in ref .
@xcite , lending us the opportunity to directly compare the results of both treatments .
let us consider the graph @xmath24 which is obtained by randomly choosing with equal probability a graph from the set of all regular graphs of size @xmath8 and vertex - degree @xmath23 .
@xmath24 is a random regular graph of degree @xmath23 .
each of the @xmath8 vertices of @xmath24 is connected to @xmath23 other vertices , but there is no structure in the connection pattern of @xmath24 . for each edge @xmath25 of graph @xmath24 ,
we assign a coupling constant @xmath26 or @xmath27 with equal probability .
once the coupling @xmath28 for an edge is assigned , it no longer changes .
therefore , we have a _ network _ with random quenched connection pattern and random quenched coupling constants .
( in the remaining part of this paper , when we use the term @xmath29network , we mean the connection pattern of the graph plus the quenched couplings ; when we use the term @xmath29graph , we mean only the connection pattern . ) on top of such a network we define the following energy function @xmath30 where @xmath31 ( @xmath32 ) is the spin variable of vertex @xmath33 , @xmath31 can take two values , @xmath34 and @xmath35 . equation ( [ eq : energy01 ] ) is the @xmath3 spin - glass model .
the total number of microscopic configurations @xmath36 in such a system is simply @xmath37 .
the vertex degree of a random regular graph of degree @xmath23 is equal to that of a regular square lattice in @xmath38 dimensions .
one hopes that some statistical physical properties of a spin - glass model on a random regular graph will also hold for the same model on a regular lattice . on the other hand , in a regular lattice , there exist many short loops , which make analytical calculations extremely difficult . in a random regular graph of very large size
there is no such short loops .
the graph is locally tree - like ; and the typical loop length in the graph scales as @xmath39 .
one can exploit this absence of short loops to construct a self - consistent mean - field theory . for the benefit of later discussions
, we also mention the concept of a random regular cavity graph @xmath40 @xcite . in this graph of size @xmath8 , there are @xmath41 ( @xmath42 ) vertices of degree @xmath43 and @xmath44 vertices of degree @xmath23 .
a vertex of degree @xmath43 in graph @xmath40 is referred to as a _ cavity vertex_. like @xmath24 , the connection pattern of @xmath40 is also completely random ; and the quenched coupling constants for the edges of @xmath40 are also independently and equally distributed over @xmath3 . the spin - glass hamiltonian eq .
( [ eq : energy01 ] ) can also be defined for this cavity network .
consider the spin - glass system eq .
( [ eq : energy01 ] ) . at high temperatures ,
the system is in the paramagnetic phase .
each vertex of the network does not have any spontaneous magnetization , its spin fluctuates over the positive and negative directions and stays in each orientation with equal probability .
the configurational phase space of the system is ergodic ( see left panel of fig . [
fig : ergodicitybreaking01 ] ) .
when the temperature @xmath0 is decreased below a spin - glass transition temperature @xmath5 , the system is in the spin - glass phase . in this phase
, a vertex of the system might favor one spin orientation over the opposite orientation , but this orientation preference is vertex - dependent .
when the number @xmath8 of vertices is sufficiently large , the configurational phase space of this spin - glass system splits into many separated domains ( see right panel of fig . [
fig : ergodicitybreaking01 ] ) .
each domain of the phase space , which corresponds to a macroscopic state or thermodynamic state of the system , contains a set of microscopic spin configurations .
the system is ergodic within each macroscopic state , while ergodicity is broken at the level of macroscopic states .
however , the system is able to transit from one macroscopic state to another different macroscopic state if the following temperature heating - annealing process is performed : the system is first heated to a high temperature beyond @xmath5 and waited for a long time till it reaches equilibrium .
the system now stays in a high - temperature ergodic phase .
afterwords , it is cooled infinitely slowly till the final low temperature @xmath0 is reached .
( a simulated - annealing idea was previously explored by kirkpatrick and co - workers @xcite to tackle hard optimization problems . ) since in the high - temperature phase the system loses memory about its prior history , at the end of the annealing process it may reach any a macroscopic state .
the probability @xmath45 of the system end up being in a particular macroscopic state @xmath1 , however , is in general different for different macroscopic states .
we argue in the following that , if an infinite number of the above - mentioned annealing experiment are performed , @xmath45 should depend only on the free energy @xmath6 of the macroscopic state @xmath1 . a schematic picture of ergodicity breaking in a spin - glass system .
( left ) when the temperature is high , all the microscopic configurations ( denoted by dots ) which contribute to the free energy of the system are in a single domain of the configurational phase space .
( right ) when the temperature is low enough , the microscopic configurations which contribute to the free energy of the system are contained in different domains in the configurational phase space .
each domain contains a number of microscopic configurations .
ergodicity is still preserved within each phase space domain , while the configuration of the system is unable to jump spontaneously from one domain to another different domain of the phase space . ]
the total partition function of the system is defined as @xmath46 where @xmath13 is the total energy expression as given by eq .
( [ eq : energy01 ] ) . when ergodicity is broken , the partition function @xmath47 can be re - written as a sum over all the macroscopic states @xmath1 : @xmath48 in eq .
( [ eq : partitiondomain02 ] ) , @xmath10 and @xmath49 are the partition function and free energy of the macroscopic state @xmath1 as defined in eq .
( [ eq : partitionfunctionalpha ] ) and eq .
( [ eq : falpha ] ) , respectively . by way of repeated temperature annealing ,
the whole configurational space of a spin - glass system is explored ( and ergodicity is recovered ! ) . the total partition function eq .
( [ eq : partitiontotal ] ) can therefore be understood as containing all the information of the system as measured by an infinite number of temperature annealing experiments . from eq .
( [ eq : partitiondomain02 ] ) we see that each macroscopic state @xmath1 contributes a term @xmath50 to the total partition function of the system .
therefore we can anticipate that , at temperature @xmath0 , the probability of the system being in macroscopic state @xmath1 at the end of an annealing experiment is given by the following distribution @xmath51 this is a free energy boltzmann distribution over all the macroscopic states @xmath1 . at this point
, we introduce an artificial inverse temperature @xmath52 at the level of macroscopic states .
( such an artificial inverse temperature @xmath52 was first introduced in the early work of ref .
@xcite for @xmath53 spin - glasses . )
this is better explained by constructing the following artificial single - particle system : the particle has a set of @xmath29energy levels , the @xmath29energy of level @xmath1 is equal to the free energy @xmath54 of the macroscopic state @xmath1 of the actual spin - glass system ; this one - particle system is in a heat bath with inverse temperature @xmath52 .
then the partition function of the artificial system is @xmath55 the probability of this artificial system being in energy level @xmath1 is given by @xmath56 when the inverse temperature @xmath52 of the artificial system is set to @xmath57 , then the partition function @xmath58 reduces to the total partition function @xmath59 of the original spin - glass system . the partition function @xmath58 of eq .
( [ eq : newpartitionfunction ] ) can be written in another form as @xmath60 \ .
\label{eq : zyb03}\ ] ] in eq .
( [ eq : zyb03 ] ) , @xmath61 is the free energy density of a macroscopic state ; the function @xmath62 is called the complexity @xcite , which is related to the total number @xmath15 of macroscopic states by the following equation @xmath63 the complexity @xmath62 , which should be non - negative , is a measure of the total number of macroscopic states with free energy density @xmath64
. when the spin - glass system is very large ( @xmath65 ) , eq .
( [ eq : zyb03 ] ) indicates that the partition function @xmath58 is contributed exclusively by those macroscopic states whose free energy density @xmath64 satisfying the equation @xmath66 equation ( [ eq : y ] ) gives an implicit relationship between the inverse temperature @xmath52 and the observed mean free energy density @xmath64 of the system . for the benefit of later discussions ,
we define a grand free energy @xmath67 by the following equation @xmath68\ .\ ] ] from eq .
( [ eq : zyb03 ] ) we see that the grand free energy density @xmath69 is @xmath70 where @xmath71 is the solution of eq .
( [ eq : y ] ) . in the next three sections we will study the statistical physical properties of the @xmath3 spin - glass system using @xmath72 and @xmath52 as a pair of control parameters .
the free energy at the microscopic level ( within a macroscopic state ) will be calculated with the inverse temperature @xmath72 , while the grand free energy at the macroscopic state level will be calculated with the inverse temperature @xmath52 .
although in the actual spin - glass system , the two inverse temperatures are identical ( @xmath73 ) , they are decoupled in the following analytical theory .
this decoupling gives us extra freedom in the theoretical development .
finally , to go from the artificial system to the original system , we will choose the largest value of @xmath52 in the interval of @xmath74 which satisfies the requirement that the complexity @xmath75 is non - negative . to analytically study the statistical physical property of a spin - glass model on a random graph
, there are basically two different but equivalent approaches .
the first one is the replica method @xcite and the second one is the cavity method @xcite .
the present paper exploits the cavity approach .
it corresponds to the 1rsb approximation of the replica method .
basically , one assumes that macroscopic states of the spin - glass system are distributed evenly in the whole configurational phase space , and there is no further organization of the macroscopic states or further structures within each macroscopic state .
consider the spin - glass system eq .
( [ eq : energy01 ] ) on a cavity network @xmath76 of @xmath8 vertices and @xmath43 cavity vertices ( see fig . [
fig : cavity01m1]a ) .
let us suppose that the configurations of such a cavity system are in a macroscopic state @xmath1 .
cavity vertices in a random regular cavity graph of vertex degree @xmath23 ( here @xmath77 ) .
( a ) the neighborhoods for three cavity vertices @xmath78 , @xmath79 and @xmath80 .
these three vertices have vertex degree @xmath43 ( @xmath81 ) while all the other @xmath82 vertices in the graph have vertex degree @xmath23 . in a macroscopic state @xmath1 ,
the three cavity vertices feel the cavity fields @xmath83 , @xmath84 and @xmath85 , respectively .
( b ) a new cavity graph with @xmath86 vertices is generated by adding a new vertex @xmath33 and connecting it to @xmath43 old cavity vertices . the cavity vertex @xmath33 of the new graph feels the cavity field @xmath87 in the corresponding macroscopic state @xmath1 of the new cavity system . ] at given inverse temperature @xmath72 , the spin value @xmath88 of a cavity vertex @xmath78 fluctuates over time around certain mean value @xmath89 which depends on the macroscopic state @xmath1 .
since @xmath90 is a binary variable , the marginal distribution @xmath91 of @xmath88 can be expressed in the following form @xmath92 the magnetization of the cavity vertex @xmath78 is @xmath93 from eq .
( [ eq : magnetization01 ] ) we know that @xmath83 is the magnetic field experienced by the cavity vertex @xmath78 due to the spin - spin interactions between vertex @xmath78 and its @xmath43 nearest - neighbors .
we call @xmath83 the _ cavity field _ on vertex @xmath78 and @xmath89 the _ cavity magnetization _ of vertex @xmath78 .
we emphasize again that @xmath83 and @xmath89 depends on the macroscopic state @xmath1 .
since the connection pattern in a cavity graph with @xmath43 cavity vertices is completely random , the typical length @xmath94 of a shortest - distance path between two cavity vertices @xmath78 and @xmath79 in the cavity graph @xmath76 is a large value .
actually , it can easily be shown @xcite that this distance scales logarithmically with the cavity graph size @xmath8 : @xmath95 denote @xmath96 as the joint probability distribution of the spin values @xmath97 of a group of cavity vertices in a cavity network . in a large random graph , according to eq .
( [ eq : lengthscale ] ) the shortest path length between any two randomly chosen vertices is long .
therefore , as the zeroth - order approximation , one may assume that this joint probability distribution can be written as the following factorized form @xmath98 where @xmath99 is vertex @xmath78 s marginal spin value distribution as given by eq .
( [ eq : cavityfield01 ] ) .
equation ( [ eq : factorization ] ) is called the bethe - peierls approximation @xcite in the literature .
it assumes statistical independence among the spin states of the cavity vertices within a macroscopic state @xmath1 .
( for recent references on extensions of the bethe - peierls approximation , see refs .
@xcite . )
the cavity magnetization @xmath89 of the cavity vertex @xmath78 depends on the identity of the macroscopic state @xmath1 . its value may be different in different macroscopic states .
let us denote @xmath100 as the fraction of macroscopic states in which the cavity magnetization of vertex @xmath78 takes the value @xmath89 , and denote @xmath101 as the fraction of macroscopic states in which the cavity magnetization of vertex @xmath102 take the value @xmath103 , respectively .
we extend the bethe - peierls approximation eq .
( [ eq : factorization ] ) to the level of macroscopic states and assume that @xmath104 equation ( [ eq : factorization - magnetization ] ) is equivalent to saying that , the fluctuations ( among all the macroscopic states ) of the cavity magnetization of two different cavity vertices are mutually independent of each other . due to the absence of short loops in a random graph , eq .
( [ eq : factorization - magnetization ] ) turns out to be a rather good approximation . to obtain a self - consistent equation for the marginal probabilities @xmath105
, we add a new vertex @xmath33 and connect it to the @xmath43 cavity vertices of @xmath76 .
the quenched coupling constant @xmath28 of each newly added edge is set to be @xmath106 with equal probability .
this results in a new cavity network @xmath107 of @xmath86 vertices and one single cavity vertex @xmath33 ( see fig . [
fig : cavity01m1]b ) . in the corresponding macroscopic state @xmath1 of the new cavity system @xmath107
, the cavity vertex @xmath33 feels a cavity field @xmath87 . to calculate the cavity field @xmath87 in the new cavity system
, we first notice that the energy difference between the cavity network @xmath107 and the old cavity network @xmath76 is @xmath108 where @xmath109 denotes the set of @xmath43 nearest - neighbors of vertex @xmath33 in the cavity graph @xmath107 . in the macroscopic state @xmath1 ,
the partition function for the cavity system @xmath107 is @xmath110 \sum\limits_{\sigma_i } \exp\bigl ( \beta h_i \sigma_i \bigr ) \label{eq : zacv03m1 } \\ & = & e^{-\beta f_\alpha \bigl ( { \cal g}_k(n , k-1)\bigr ) } \prod\limits_{j\in \partial^\prime i } \biggl[\frac { \cosh \beta j_{i j } } { \cosh \beta u(j_{i j } , h_{j } ) } \biggr ] ( 2 \cosh \beta h_{i } ) \ .
\label{eq : zacv03}\end{aligned}\ ] ] in going from eq .
( [ eq : zacv03m3 ] ) to eq .
( [ eq : zacv03m2 ] ) , we have used the bethe - peierls approximation eq .
( [ eq : factorization ] ) .
@xmath111 is the free energy of the cavity system @xmath76 in its macroscopic state @xmath1 . in eq .
( [ eq : zacv03m1 ] ) , the quantity @xmath112 is defined as @xmath113 and the quantity @xmath87 is calculated according to @xmath114 from eq .
( [ eq : zacv03m1 ] ) we know that @xmath87 as expressed by eq .
( [ eq : hia ] ) is just the cavity field felt by vertex @xmath33 in the cavity network @xmath107 . if we know all the cavity fields on the nearest - neighbors of vertex @xmath33 , then we obtain the cavity field on vertex @xmath33 through the iterative equation ( [ eq : hia ] ) .
consequently , the magnetization of the cavity vertex @xmath33 in the macroscopic state @xmath1 of the cavity system @xmath107 is calculated as @xmath115 - \prod\limits_{j\in \partial^\prime i } \bigl [ 1 - v_{i j } m_j \bigr ] } { \prod\limits_{j\in \partial^\prime i } \bigl [ 1 + v_{i j } m_j \bigr ] + \prod\limits_{j\in \partial^\prime i } \bigl [ 1 - v_{i j } m_j ] } \ , \ ] ] where @xmath116 is a shorthand notation for @xmath117 , i.e. , @xmath118 equation ( [ eq : mi ] ) is an iterative equation for the cavity magnetization @xmath119 _ within one macroscopic state @xmath1_. as we have emphasized in sec . [
subsec : single01 ] , the input cavity magnetization @xmath89 in eq .
( [ eq : mi ] ) may be different in different macroscopic states . as a consequence , the cavity magnetization @xmath119 of vertex @xmath33
does not necessarily take the same value in different macroscopic states . on the contrary
, its value may fluctuate a lot among different macroscopic states of the new cavity system @xmath107 .
the task is to obtain an expression for the marginal distribution @xmath120 of @xmath119 among different macroscopic states . from eq .
( [ eq : zacv03 ] ) , we know that , after the addition of the vertex @xmath33 , the free energy difference @xmath121 between the macroscopic state @xmath1 of the system @xmath107 and that of the system @xmath76 is @xmath122 + \prod\limits_{j\in \partial^\prime i } \bigl[1- v_{i j } m_j \bigr ] \biggr ) \ .
\label{eq : deltaf102}\end{aligned}\ ] ] according to the free energy boltzmann distribution eq .
( [ eq : boltzmanndistribution02 ] ) , each macroscopic state is weighted by the boltzmann factor @xmath123 .
after the addition of vertex @xmath33 , the total partition function of the system @xmath107 is @xmath124 \prod\limits_{j\in \partial^\prime i } \biggl [ \int { \rm d } m_j { \cal p}_j(m_j ) \biggr ] \exp(-y \delta f_1 ) \ . \label{eq : rho - m1}\end{aligned}\ ] ] we have used the factorization approximation eq . ( [ eq : factorization - magnetization ] ) in going from eq .
( [ eq : rho - m1a ] ) to eq .
( [ eq : rho - m1 ] ) . similarly , the total weight of those macroscopic states of the system @xmath107 in which the cavity magnetization of vertex @xmath33 being equal to @xmath119 is expressed as @xmath125 \times \nonumber \\ & & \prod\limits_{j\in \partial^\prime i } \biggl [ \int { \rm d } m_j { \cal p}_j(m_j ) \biggr ] \exp(-y \delta f_1 ) \delta\biggl(m_i-\frac { \prod\limits_{j\in \partial^\prime i } \bigl [ 1 + v_{i j } m_j \bigr ] - \prod\limits_{j\in \partial^\prime i } \bigl [ 1 - v_{i j } m_j \bigr ] } { \prod\limits_{j\in \partial^\prime i } \bigl [ 1 + v_{i j } m_j \bigr ] + \prod\limits_{j\in \partial^\prime i } \bigl [ 1 - v_{i j } m_j ] } \biggr ) \ . \label{eq : rho - m2}\end{aligned}\ ] ] from eq .
( [ eq : rho - m1 ] ) and eq .
( [ eq : rho - m2 ] ) we realize that , in the new system , the fraction of macroscopic states in which vertex @xmath33 bearing a cavity magnetization @xmath119 is equal to @xmath126 e^{-y \delta f_1 } \delta\biggl ( m_i - \frac { \prod\limits_{j\in \partial^\prime i } \bigl [ 1 + v_{i j } m_j \bigr ] - \prod\limits_{j\in \partial^\prime i } \bigl [ 1 - v_{i j } m_j \bigr ] } { \prod\limits_{j\in \partial^\prime i } \bigl [ 1 + v_{i j } m_j \bigr ] + \prod\limits_{j\in \partial^\prime i } \bigl [ 1 - v_{i j } m_j ] } \biggr ) \ .\ ] ] equation ( [ eq : rho ] ) is a self - consistent iterative equation for the cavity distributions @xmath127 . a steady state solution of eq .
( [ eq : rho ] ) can be obtained by population dynamics @xcite .
an array of @xmath128 probability distributions @xmath129 are stored . at each step of the population dynamics , @xmath43 probability distributions
are randomly chosen from this array of @xmath128 stored distributions , and a new probability distribution is generated by using eq .
( [ eq : rho ] ) .
this new distribution then replaces a randomly chosen old probability distribution in the array .
this iteration process is repeated many times until the population dynamics reaches a steady state .
the free energy of a macroscopic state @xmath1 is defined formally by eq .
( [ eq : falpha ] ) . as we noted
before , this expression is not directly applicable since we do not know what are the microscopic configurations of state @xmath1 .
the cavity approach @xcite circumvents this problem by calculating the grand free energy difference between a system of @xmath8 vertices and an enlarged system of @xmath130 vertices . as demonstrated in fig .
[ fig : cavity01 ] , a random network @xmath24 can be constructed from a random cavity network @xmath131 by adding @xmath43 new edges .
the grand free energy difference between these two systems can be calculated .
similarly , one can add two new vertices and @xmath132 new edges to change the same random cavity network @xmath133 into a random regular network @xmath134 . the grand free energy difference between these two systems
can also be calculated . from the two grand free energy differences ,
one can obtain the grand free energy density @xmath135 for a random regular system @xmath24 .
the mean free energy density of a macroscopic state and the complexity of the system can then be obtained from @xmath135 .
the cavity approach to the spin - glass model eq .
( [ eq : energy01 ] ) on a random regular graph of connectivity @xmath136 .
( a ) : part of a cavity graph with @xmath137 cavity vertices .
( b ) and ( c ) : construction of a random regular graph of size @xmath8 ( b ) and @xmath130 ( c ) from the cavity graph with @xmath8 vertices and @xmath137 cavity vertices .
, title="fig : " ] the cavity approach to the spin - glass model eq .
( [ eq : energy01 ] ) on a random regular graph of connectivity @xmath136 .
( a ) : part of a cavity graph with @xmath137 cavity vertices .
( b ) and ( c ) : construction of a random regular graph of size @xmath8 ( b ) and @xmath130 ( c ) from the cavity graph with @xmath8 vertices and @xmath137 cavity vertices .
, title="fig : " ] the cavity approach to the spin - glass model eq .
( [ eq : energy01 ] ) on a random regular graph of connectivity @xmath136 . ( a ) : part of a cavity graph with @xmath137 cavity vertices .
( b ) and ( c ) : construction of a random regular graph of size @xmath8 ( b ) and @xmath130 ( c ) from the cavity graph with @xmath8 vertices and @xmath137 cavity vertices .
, title="fig : " ] the difference @xmath139 between the configurational energy of the system on the random graph @xmath24 in fig .
[ fig : cavity01]b and that of the system on the random cavity graph @xmath140 in fig .
[ fig : cavity01]a is @xmath141 where @xmath142 denotes the set of @xmath43 newly added edges in going from @xmath140 to @xmath24 .
each edge in set @xmath142 connects two cavity vertices of @xmath133 . following the analytical procedure of sec .
[ subsec : field - distribution ] , we know that difference between the free energy of a macroscopic state @xmath1 of the system @xmath24 and that of the same macroscopic state @xmath1 of the cavity system @xmath131 is @xmath143 where @xmath144 can be understood as the free energy increase caused by the addition of an edge ( with coupling @xmath28 ) between two cavity vertices @xmath33 and @xmath78 . the grand free energy @xmath145 of the new system @xmath24 is related to the grand free energy @xmath146 of the old cavity system @xmath147 by the following equation @xmath148 where @xmath149 \\ ] ] is the increase to the grand free energy caused by adding an edge @xmath25 . the random network @xmath134 in fig . [ fig : cavity01]c is constructed from the random cavity network @xmath140 of fig . [ fig : cavity01]a by adding two vertices ( @xmath151 and @xmath152 ) .
vertex @xmath151 is connected to @xmath43 cavity vertices ( @xmath153 ) of @xmath133 , and vertex @xmath152 is connected to the remaining @xmath43 cavity vertices ( @xmath154 ) of @xmath147 .
vertex @xmath151 and @xmath152 are directly connected by a new edge @xmath155 , so that every vertex in the graph @xmath134 has degree @xmath23 . the energy difference between the system @xmath134 and the cavity system @xmath147 is @xmath156 the increase @xmath157 in the free energy of macroscopic state @xmath1 due to the addition of two new vertices and @xmath158 new edges
can be obtained following the same procedure as given in sec .
[ subsec : field - distribution ] .
we find that @xmath159 in eq .
( [ eq : freeenergy03 ] ) , @xmath160 is the free energy increase caused by adding vertex @xmath151 and connecting it to the set @xmath161 of @xmath23 vertices .
the explicit expression for @xmath160 is @xmath162 + \prod\limits_{j\in \partial i_0 } \bigl[1- v_{i_0 j } m_j \bigr ] \biggr ) \ .
\label{eq : deltafi0}\ ] ] the expression for @xmath163 has the same form as eq . ( [ eq : deltafi0 ] ) .
let us emphasize that , in eq .
( [ eq : deltafi0 ] ) , @xmath164 is the cavity magnetization of vertex @xmath152 when vertex @xmath151 is not added , i.e. , @xmath165- \prod\limits_{j_s\in \partial i_0 \backslash j_0 } \bigl [ 1 - v_{i_0 j_s } m_{j_s } \bigr ] } { \prod\limits_{j_s\in \partial i_0 \backslash j_0 } \bigl [ 1 + v_{i_0 j_s } m_{j_s } \bigr ] + \prod\limits_{j_s\in \partial i_0 \backslash j_0 } \bigl [ 1 - v_{i_0 j_s } m_{j_s } \bigr ] } \ , \label{eq : m0val } \\ m_{j_0 } & = & \frac { \prod\limits_{i_s \in \partial j_0 \backslash i_0 } \bigl [ 1 + v_{j_0 i_s } m_{i_s } \bigr]- \prod\limits_{i_s \in \partial j_0 \backslash i_0 } \bigl [ 1 - v_{j_0 i_s } m_{i_s } \bigr ] } { \prod\limits_{i_s \in \partial j_0 \backslash i_0 } \bigl [ 1 + v_{j_0 i_s } m_{i_s } \bigr ] + \prod\limits_{i_s\in \partial j_0 \backslash i_0 } \bigl [ 1 - v_{j_0 i_s } m_{i_s } \bigr ] } \ .
\label{eq : m0pval}\end{aligned}\ ] ] the term @xmath166 of eq .
( [ eq : freeenergy03 ] ) is the free energy increase caused by setting up an edge between vertex @xmath151 and vertex @xmath152 ; its expression is given by eq .
( [ eq : freeenergy02 ] ) , with @xmath167 and @xmath164 being determined by eq .
( [ eq : m0val ] ) and eq . ( [ eq : m0pval ] ) , respectively . the free energy increase eq .
( [ eq : freeenergy03 ] ) can be intuitively understood as follows : since the contribution of the edge @xmath168 is counted twice in @xmath160 and @xmath163 , the free energy increase should be corrected with an edge term . with these preparations
, we can calculate the total grand free energy @xmath169 of the system @xmath134 .
similar to eq .
( [ eq : gfd - i ] ) , we find that @xmath170 in eq .
( [ eq : delta - g-2 - 01 ] ) , @xmath171 is the increase to the grand free energy caused by adding vertex @xmath151 and connecting it to @xmath23 vertices , with @xmath172 where @xmath173 is the cavity magnetization distribution of vertex @xmath152 in the absence of vertex @xmath151 .
@xmath174 in eq .
( [ eq : delta - g-2 - 01 ] ) is calculated through eq .
( [ eq : dgij ] ) using @xmath175 and @xmath176 .
the grand free energy density of the spin - glass system is @xmath177 an explicit expression for @xmath135 can be written down by applying eq .
( [ eq : gfd - i ] ) and eq .
( [ eq : delta - g-2 - 01 ] ) , which is a function of the quenched randomness in the system .
the grand free energy density has the nice property of self - averaging @xcite , namely the value of @xmath69 as calculated for a typical system is equal to the averaged value of @xmath69 over many systems with different realizations of the quenched randomness in the graph connection pattern and in the edge coupling constants .
when the quenched randomness is averaged out , we obtain that @xmath178 where @xmath179 and @xmath180 are calculated through eq .
( [ eq : dgi0 ] ) and eq .
( [ eq : dgij ] ) , respectively ; and an overline indicates averaging over the quenched randomness of the spin - glass system .
the mean free energy density of a macroscopic state of the system is related to @xmath135 by @xmath181 and the complexity of the system at given value of the reweighting parameter @xmath52 is @xmath182
the discussion in sec .
[ sec : densities ] was concerned with the typical properties of an ensemble of spin - glass systems governed by a given distribution of quenched randomness .
one important advantage of the cavity approach is that , for a single instance of the quenched randomness , the thermodynamic properties can also be calculated . under the bethe - peierls approximation , the total grand free energy of a spin - glass system eq .
( [ eq : energy01 ] ) on a graph @xmath183 with couplings @xmath184 is equal to @xmath185 it is easy to see that the above equation is consistent with eq .
( [ eq : g - final ] ) . in eq .
( [ eq : gfe - single-01 ] ) , @xmath186 denotes the contribution of vertex @xmath33 and its associated edges to the total grand free energy of the system ; its expression has the same form as eq .
( [ eq : dgi0 ] ) : @xmath187 e^{-y \delta f^{(i ) } } \biggr ) \ , \ ] ] where @xmath188 is the cavity magnetization of vertex @xmath78 ( with respect to vertex @xmath33 ) in a macroscopic state @xmath1 , and @xmath189 is the probability distribution of this cavity magnetization among all the macroscopic states ; @xmath190 is the free energy contribution ( in a given macroscopic state ) of vertex @xmath33 and its associated edges , which has the same form as eq .
( [ eq : deltafi0 ] ) but with @xmath89 replaced by @xmath188 . similarly , the term @xmath191 in eq .
( [ eq : gfe - single-01 ] ) is the contribution of an edge @xmath25 to the total grand free energy of the system ; it is expressed as @xmath192 where @xmath193 is given by eq .
( [ eq : freeenergy02 ] ) but with @xmath119 replaced by @xmath194 and @xmath89 replaced by @xmath188 . the set of @xmath195 ( @xmath196 being the total number of edges in the graph @xmath183 ) probability distributions @xmath189 in eq .
( [ eq : gfe - single-01 ] ) should be carefully chosen such that the total grand free energy achieves a minimal value at given @xmath197 values .
in other words , for any edge @xmath25 of the graph @xmath183 , the variation of @xmath198 with respect to both @xmath199 and @xmath200 should vanish : @xmath201 equation ( [ eq : gfe - single-04 ] ) results in the following self - consistent bethe - peierls equation for the @xmath202 s : @xmath203 e^{-y \delta f_j^{(i ) } } \delta\biggl ( m_{i\to j}-\frac { \prod\limits_{k\in \partial i \backslash j } [ 1 + v_{i k } m_{k\to i } ] - \prod\limits_{k\in \partial i \backslash j } [ 1 - v_{i k } m_{k\to i } ] } { \prod\limits_{k\in \partial i\backslash j } [ 1 + v_{i k } m_{k\to i}]+ \prod\limits_{k\in \partial i\backslash j } [ 1 - v_{i k } m_{k\to i } ] } \biggr ) \ , \ ] ] where @xmath204 + \prod\limits_{k\in \partial i\backslash j } \bigl[1- v_{i k } m_{k\to i } \bigr ] \biggr ) \ .\ ] ] @xmath205 has the same physical meaning as @xmath121 of sec .
[ subsec : field - distribution ] . equation ( [ eq : gfe - single-06 ] ) is consistent with eq .
( [ eq : rho ] ) of sec .
[ subsec : field - distribution ] . similar to eq .
( [ eq : f - final ] ) and eq .
( [ eq : sigma - final ] ) , the mean free energy of a macroscopic state of the sample is expressed as @xmath206 and the complexity of the system is calculated through @xmath207 because of eq .
( [ eq : gfe - single-04 ] ) , the first derivative of @xmath208 with respect to @xmath52 can easily be expressed .
we find that eq .
( [ eq : fe - single-02 ] ) can be re - written as @xmath209 where @xmath210 e^{-y \delta f^{(i ) } } \delta f^{(i ) } } { \prod\limits_{j\in \partial i } \bigl[\int { \rm d } m_{j\to i } { \cal p}_{j\to i}(m_{j\to i } ) \bigr ] e^{-y \delta f^{(i ) } } } \ , \label{eq : average - dfi } \\
\bigl\langle \delta f^{(i j ) } \bigr\rangle & = & \frac{\int { \rm d } m_{j\to i } \int { \rm d } m_{i\to j } { \cal p}_{j\to i}(m_{j\to i } ) { \cal p}_{i\to j}(m_{i\to j } ) e^{-y \delta f^{(i j ) } } \delta f^{(i j ) } } { \int { \rm d } m_{j\to i } \int { \rm d } m_{i\to j } { \cal p}_{j\to i}(m_{j\to i}){\cal p}_{i\to j}(m_{i\to j } ) e^{-y \delta
f^{(i j ) } } } \ .
\label{eq : average - dfij}\end{aligned}\ ] ] the mean free energy of a macroscopic state is also decomposed into vertex contributions and edge contributions .
according to the discussion in sec .
[ sec : mfep ] , in the mean - field theory , the reweighting parameter @xmath52 should be chosen such that its value is closest to the physical inverse temperature @xmath72 .
therefore , at a given value of @xmath72 , the mean free energy @xmath211 of the system is obtained by setting @xmath57 in eq .
( [ eq : fe - single-02 ] ) or eq .
( [ eq : free - energy-1rsb ] ) , provided that the complexity @xmath212 is always non - negative in the range of @xmath213 .
if , on the other hand , @xmath212 changes from being positive to being negative at a point of @xmath214 , we should set @xmath52 to @xmath215 . in this later situation ,
the mean free energy of the system is equal to the maximal value of the grand free energy , @xmath216 .
after the mean free energy @xmath211 of the spin - glass system is obtained , the mean energy of a macroscopic state is calculated through @xmath217 and the mean entropy of a macroscopic state is calculated through @xmath218 ensemble - averaged densities of mean energy and mean entropy can also be obtained from eqs .
( [ eq : mean - energy ] ) and ( [ eq : mean - entropy ] ) by averaging out the quenched randomness .
an alternative derivation for the mean energy density is given in the appendix .
we have applied the mean - field treatments of sec . [
sec : cavity - field - distribution ] and sec .
[ sec : densities ] to the @xmath3 spin - glass model eq .
( [ eq : energy01 ] ) on a random regular graph of vertex degree @xmath4 .
( in what follows , energy is in units of @xmath219 . ) at temperature @xmath220 ( i.e. , @xmath221 ) , carrus , marinari and zuliani ( as cited in ref .
@xcite ) estimated that the energy density of this system is @xmath222 .
the mean - field theory of ref .
@xcite also predicted the same result .
this reference also predicted a mean - free energy density of @xmath223 . in the present population dynamics simulation , each probability profile @xmath120 in eq .
( [ eq : rho ] ) was represented by an vector of @xmath224 elements . a total number of @xmath128 such probability profiles were stored in the computer and they were updated using eq .
( [ eq : rho ] ) . in our work ,
various @xmath128 values ranging from @xmath225 to @xmath226 were used . in each elementary update of the probability distribution @xmath120 , a metropolis importance sampling @xcite with inverse temperature @xmath52
was exploited to generate a total number of @xmath227 magnetizations @xmath119 .
these magnetizations were stored in an ordered set .
the averaged value of the first ( second , third , @xmath228 ) @xmath229 magnetizations in this set is assigned to the first ( second , third,@xmath228 ) element of the new @xmath129 .
we chose @xmath230 in our simulations . the grand free energy density @xmath69 of the @xmath3 spin - glass model on a random regular graph of vertex degree @xmath4 .
the inverse temperature is fixed to @xmath221 .
symbols are results obtained by population dynamics simulations using a population size of @xmath231 .
the dashed line is a fit to the data using @xmath232 , with the fitting parameters @xmath233 and @xmath234 . ]
mean free energy density of the @xmath3 spin - glass model on a random regular graph of vertex degree @xmath4 at @xmath221 as obtained by population dynamics with different population size @xmath128 .
( inset ) the optimal reweighting parameter @xmath235 . ] the energy density , free energy density , entropy density , and complexity as a function of the reweighting parameter @xmath52 for the @xmath3 spin - glass model on a random regular graph of vertex degree @xmath4 at inverse temperature @xmath221 . energy unit is @xmath219 .
the symbols represent numerical population dynamics results using @xmath236 .
the dashed line in the bottom panel is a fitting to the numerical data with @xmath237 in the range of @xmath238 . to reduce computational time , for the simulations reported in this figure , a faster but
less precise numerical recipe was used to generate a new probability profile @xmath120 using eq .
( [ eq : rho ] ) .
this might be the reason for the relatively large fluctuations of the mean densities with the reweighting parameter @xmath52 . ] the grand free energy density @xmath69 of the system as calculated according to eq .
( [ eq : g - final ] ) is shown in fig .
[ fig : gy ] . the grand free energy density @xmath239 first increases with the reweighting parameter @xmath52 till @xmath52 reaches @xmath234 , at which point @xmath135 attains a maximal value of @xmath240 .
this maximal value of @xmath135 corresponds to the best estimate of the mean free energy density of the system by the present mean - field method .
it is in close agreement with the prediction of ref .
we have checked that the estimated mean free energy density value is not sensitive to the population size @xmath128 ( see fig . [
fig : mfd ] ) . we have also calculated the mean energy density and mean entropy density for the model system at @xmath221 using a population size of @xmath241 , with the mean energy density being @xmath242 and the mean entropy density being @xmath243 ( see fig . [
fig : result01 ] ) .
these predictions are also in good agreement with the numerical results of ref .
@xcite and with the simulational work of carrus , marinari and zuliani .
these good agreements of the present theoretical results with earlier simulational and numerical work suggest that the present cavity approach based on the concept of cyclic heating and annealing is feasible .
as a summary , in this paper we have calculated the thermodynamic properties of a spin - glass model by combining the physical idea of repeated heating - annealing and the cavity approach of mzard and parisi @xcite . we have assumed that , during a cyclic annealing experiment , all the thermodynamic ( or macroscopic ) states of the spin - glass system at a given low temperature @xmath0 will be reachable , but with different frequencies which decrease exponentially with the free energy values of the macroscopic states . by using this free energy boltzmann distribution and by using the bethe - peierls approximation ,
the grand free energy of the spin - glass system can be calculated as a function of a reweighting parameter @xmath52 ; and from the knowledge of the grand free energy , the mean free energy , energy , and entropy of a macroscopic state of the system can also be obtained . for the @xmath3 spin - glass model on a random regular graph of vertex degree @xmath4 ,
the theoretical predictions of the present work are in good agreement with the results of earlier simulational and numerical calculations . for the @xmath3 model at @xmath221 , we found that the complexity @xmath75 becomes negative before @xmath52 reaches @xmath72 .
therefore , the thermodynamic properties of the system are contributed by those macroscopic states which have the global minimal free energy density .
however , it may exist other systems for which the complexity is still positive even when the reweighting parameter @xmath52 reaches @xmath72 .
for such systems , the thermodynamic properties are not determined by those macroscopic states of the minimal free energy density , but by a set of @xmath29metastable macroscopic states . on the one hand ,
these metastable macroscopic states have a higher free energy density ; on the other hand , the number of such macroscopic states greatly exceeds the number of macroscopic states of the global minimal free energy density . in the competition between these two factors
, the @xmath29metastable macroscopic states may win .
then the system will reach one of these metastable macroscopic states almost surely in each round of the temperature annealing experiment .
we will work on a model system with many - body interactions to check whether this is really the case .
as in the work of ref .
@xcite , the present theoretical treatment also assumes that the configurational space of the spin - glass system breaks into exponentially many macroscopic states , but there are no further organizations of these macroscopic states ; and it is assumed that each macroscopic state is ergodic . as pointed out in ref .
@xcite , this first - step replica - symmetry - broken ( 1rsb ) cavity solution might be unstable .
two types of instability is conceivable .
the type i instability concerns with possible clustering of macroscopic states into @xmath29super - macroscopic states ; the type ii instability is caused by splitting of each macroscopic state into many @xmath29sub - macroscopic states . to account for
the further clustering of macroscopic states appears to be relatively easy in the present mean - field framework : for each cluster of macroscopic states , a vertex has a probability profile concerning its cavity magnetization ; this probability profile is different in different clusters , and we can introduce a probability distribution of probability profiles to characterize this variation [ h. zhou , submitted to comm .
( beijing ) ] . on the hand , to take into consideration the possibility of splitting of a macroscopic state , maybe one has to introduce other reweighting parameters in the mean - field theory .
this is an important issue waiting for further explorations .
another important issue is related to the updating of the probability profiles through the iterative equation ( [ eq : rho ] ) or ( [ eq : gfe - single-05 ] ) . in the present paper , we used metropolis importance sampling technique to get an updated probability profile of cavity magnetization .
this method appears to be rather precise but time - consuming .
if faster algorithms with comparable precision could be constructed , it will be highly desirable .
the numerical simulations reported in this paper were carried out at the pc clusters of the state key laboratory for scientific and engineering computing , cas , china .
in the main text we did not give an explicit formula for the mean energy density @xmath244 of a spin - glass system .
instead , the mean energy density was calculated from first knowing the mean free energy density of the system [ see , e.g. , eq .
( [ eq : mean - energy ] ) ] . in this appendix
, we derive an explicit expression for the mean energy density @xmath244 .
the assumptions used in this derivation will be clearly pointed out .
consider again the two systems , @xmath133 in fig .
[ fig : cavity01]a and @xmath24 in fig .
[ fig : cavity01]b .
( for notational simplicity , let us hereafter referr to these two systems as system a and b , respectively . ) as was mentioned in sec .
[ subsec : nton ] , for a given configuration the energy difference between these two systems is @xmath139 . when averaged over all macroscopic states , the mean total energy of the system a is @xmath245 where @xmath246 is the total free energy of macroscopic state @xmath1 of system a , and @xmath247 is the the configurational energy of system a. similarly , the mean total energy of the system b is @xmath248 \exp\bigl(-\beta h^a(\vec{\sigma } ) - \beta \delta h_2 \bigr ) } { \sum_{\vec{\sigma}\in \alpha } \exp\bigl(-\beta h^a(\vec{\sigma } ) - \beta \delta h_2 \bigr ) } } { \sum\limits_\alpha \exp\bigl(-y f_\alpha^a - y \delta f_2 \bigr ) } \ , \ ] ] where @xmath249 is defined by eq .
( [ eq : freeenergy01 ] ) .
the difference of these two mean energy expressions is @xmath250 in the above equation , @xmath251 where @xmath252 in going from eq .
( [ eq : v1m1 ] ) to eq .
( [ eq : v1 ] ) we have used the bethe - peierls approximation both at the level of microscopic configurations [ eq . ( [ eq : factorization ] ) ] and at the level of macroscopic states [ eq . ( [ eq : factorization - magnetization ] ) ] . 1 . the distribution of the configurational energy of the cavity system a and that of the energy increase can be regarded as mutually independent within a macroscopic state .
2 . the distributions of the free energy of the cavity system a and that of the free energy increase can also be regarded as mutually independent . now consider the two systems , @xmath133 in fig .
[ fig : cavity01]a ( system a ) and @xmath134 in fig .
[ fig : cavity01]c ( system c ) .
following the same analytical procedure as given in the preceding subsection , we find that , if the two assumptions ( i ) and ( ii ) listed above are valid when going from system a to system c , then the mean total energy of system c is related to that of system a by the following simple equation @xmath257 ( notice that in calculating @xmath258 through eq .
( [ eq : deij ] ) , @xmath167 should be the cavity magnetization of vertex @xmath151 in the absence of vertex @xmath259 . ) .
the same analytical procedure can be applied to calculate the mean free energy density . under the assumption ( ii ) of the preceding subsection
, we obtain that the mean free energy density @xmath64 is equal to @xmath261 which is just the same as eq .
( [ eq : free - energy-1rsb ] ) . based on the assumption of exponentially many macroscopic states ,
it was argued in ref .
@xcite that , the assumption ( ii ) of the preceding subsection should be valid under the bethe - peierls approximation .
in other words , the joint distribution of the free energy of the old cavity system @xmath14 and the free energy increase @xmath262 is factorized .
the consistence between eq .
( [ eq : f - another - final ] ) and eq .
( [ eq : free - energy-1rsb ] ) also confirmed this point .
since each macroscopic state ifself contains an exponential number of microscopic configurations , we believe that the assumption ( i ) is also valid in the bethe - peierls mean - field framework . the mean energy density of the @xmath3 spin - glass model as shown in fig .
[ fig : result01 ] was calculated using eq .
( [ eq : epsilon - final ] ) , and the output was in good agreement with earlier simulation result .
maybe one can further check this point by working on some simple models which are exactly solvable both analytically and computationally .
the maximum matching problem as studied in refs .
@xcite could be a good candidate model system . | at sufficiently low temperatures , the configurational phase space of a large spin - glass system breaks into many separated domains , each of which is referred to as a macroscopic state .
the system is able to visit all spin configurations of the same macroscopic state , while it can not spontaneously jump between two different macroscopic states .
ergodicity of the whole configurational phase space of the system , however , can be recovered if a temperature - annealing process is repeated an infinite number of times . in a heating - annealing cycle , the environmental temperature is first elevated to a high level and then decreased extremely slowly until a final low temperature @xmath0 is reached .
different macroscopic states may be reached in different rounds of the annealing experiment ; while the probability of finding the system in macroscopic state @xmath1 decreases exponentially with the free energy @xmath2 of this state . for finite - connectivity spin glass systems , we use this free energy boltzmann distribution to formulate the cavity approach of mzard and parisi [ eur .
phys .
j. b * 20 * , 217 ( 2001 ) ] in a slightly different form . for the @xmath3 spin - glass model on a random regular graph of degree @xmath4
, the predictions of the present work agree with earlier simulational and theoretical results . |
the field of flavour physics is experiencing rapid experimental progress , e.g. increasingly precise measurements of the leptonic mixing angles and perspectives to measure currently unknown parameters , such as the leptonic dirac cp phase and the neutrino mass ordering .
this progress provides both a challenge for some of the existing models explaining the flavour structure of the standard model , and an opportunity to constrain and test such models with increasing precision .
for example , the measurement of @xmath5 by t2k @xcite , double chooz @xcite , reno @xcite , and in particular daya bay @xcite , has put tensions on popular models that predict tribimaximal mixing , and has encouraged the search for new models that are compatible with the experimental values @xcite .
a sufficiently complete particle physics model should also provide a consistent cosmological evolution .
many processes in the early universe happen at high energy scales , which makes these processes very sensitive to particle physics beyond the standard model .
inflation @xcite and baryogenesis are obvious examples .
inflation is driven by the potential of some scalar field , which can be identified with a new field of the particle theory , and baryogenesis is particularly sensitive to the flavour structure and cp violating interactions at high energies .
precision cosmology could therefore provide interesting hints to the high energy particle theory .
several scenarios have been proposed for realising inflation with observable matter fields , e.g. higgs inflation with non - minimal coupling to gravity @xcite , mssm inflation near inflection points @xcite , and tribrid inflation @xcite .
the inflaton in these models can be an observable gauge non - singlet field , like a higgs field or some d - flat mssm direction .
in such models , the inflationary dynamics and the reheating phase depend on the couplings of these observable matter fields . therefore , they have good prospects for connecting cosmological and particle physics observables , leading to more predictive , testable models . in this paper
, we discuss how tribrid inflation can be realised in flavour models . in the first section ,
we briefly review tribrid inflation and discuss the general procedure how it can be implemented in a given particle physics model .
afterwards , we provide an explicit example model with an @xmath0 family symmetry to demonstrate how this works in practice .
the resulting model is predictive and phenomenologically viable . in particular
, it connects cosmological and particle physics parameters , predicting a relationship between the running of the spectral index @xmath1 and the neutrino yukawa coupling @xmath6 or the right - handed neutrino mass @xmath7 . in the last section
we show that the model can automatically avoid the production of potentially dangerous topological defects at the end of inflation , because the waterfall field receives a small shift already during inflation .
we then finish with a summary of our results .
in this section , we discuss tribrid inflation @xcite and how it can be implemented in supersymmetric particle physics models .
we start with a brief review and then describe the requirements and the strategy for realising tribrid inflation in a given model .
we will later apply this general strategy to an explicit @xmath0 model in section [ sec : a4model ] as an illustrative example .
tribrid inflation is a variant of supersymmetric hybrid inflation @xcite , which uses a superpotential of the schematic form @xmath8 with @xmath9 .
we use natural units with @xmath10 to keep the notation simple . with this superpotential , tribrid inflation is defined as single - field inflation along the @xmath11 direction while @xmath12 , @xmath13 .
inflation ends with a waterfall transition in @xmath12 , which is triggered when @xmath14 .
the singlet field @xmath15 is merely an auxiliary field which is approximately zero during and after inflation .
usually gets a hubble - size mass from the khler potential during inflation . for @xmath16 , the auxiliary field
is then stabilized at @xmath17 . for @xmath18 , @xmath19 along the inflationary trajectory , but as long as it is heavy enough , we still have @xmath20 and @xmath21 , which can be roughly approximated by @xmath13 . @xcite ] the main difference between tribrid inflation and conventional susy hybrid inflation , which uses the singlet @xmath15 as the inflaton , is that the tribrid inflaton @xmath11 can be charged under symmetries .
we can therefore use matter fields , e.g. d - flat mssm directions , as inflaton directions .
this connects inflation and particle physics more closely : inflation is most sensitive to the inflaton couplings , so if the inflaton is composed of observable matter fields , then the particle theory can more easily constrain inflationary predictions .
it turns out that depending on @xmath22 and @xmath23 , the effective inflaton potential @xmath24 is dominated by different contributions , leading to qualitatively different regimes : * @xmath25 ( `` loop - driven regime '' ) @xcite : in this case , @xmath24 is usually dominated by one - loop quantum corrections to the effective potential .
the inflaton potential can receive dangerously large supergravity contributions from the khler potential , but these can be avoided either by the use of symmetries in the khler potential @xcite or by a percent - level tuning of one khler potential parameter . * @xmath26 ( `` khler - driven regime '' ) @xcite : in this case , @xmath24 is generated mostly from planck - suppressed operators in the khler potential .
, the loop potential is usually dominant ( loop - driven regime ) , except for some specific khler potentials , where even @xmath27 can be khler - driven . for @xmath28 ,
loop corrections are strongly suppressed , and the khler - driven regime is the generic case . ] for @xmath29 , where the effective field theory ( eft ) is valid , the resulting inflaton potential usually has a hilltop - type shape with only three relevant parameters .
khler - driven inflation can produce an observably large @xmath30 , which if it should be measured could fix the most important model parameter , and also rule out the other regimes of tribrid inflation .
* @xmath31 ( `` pseudosmooth regime '' ) @xcite : in this case , @xmath24 is generated from the superpotential coupling @xmath32 , with @xmath33 already during inflation .
inflation proceeds similar to smooth hybrid inflation , but ends with a waterfall .
note that because @xmath33 already during inflation , no topological defects are formed during the waterfall transition .
the predictions have so far been calculated for @xmath34 only , so it is not yet clear whether models with @xmath35 can generate the observed cmb spectrum .
the generic predictions for tribrid inflation within the eft regime are @xmath36 , @xmath37 and @xmath38 .
the small tensor - to - scalar ratio is typical for small - field models and the mass scale @xmath39 is typical for hybrid inflation .
some of these are falsifiable predictions : an observable @xmath30 would exclude both the loop - driven and the pseudosmooth regimes , and an observable @xmath40 or @xmath41 would exclude all three regimes of tribrid inflation .
in addition to these rough predictions , tribrid inflation provides relations between cmb observables and model parameters .
these detailed predictions depend on the regime .
for example , the khler - driven regime links the running of the spectral index @xmath1 to the superpotential coupling @xmath42 and the symmetry breaking scale @xmath43 .
these relations can connect cosmology and particle physics in explicit models . in this paper
, we focus on tribrid inflation in the khler - driven regime : @xmath44 with @xmath45 , @xmath46 . for model building purposes
, one should realize that @xmath47 , @xmath48 and @xmath49 can be d - flat combinations of different fields : @xmath50 , @xmath51 .
we can , for example , use some d - flat mssm direction as the inflaton by replacing @xmath52 or @xmath53 . note that @xmath47 and @xmath54 can be partially composed of different fields , as long as at least one of the fields in @xmath47 and @xmath54 is the same .
for example , one could have @xmath55 where @xmath56 is the component of the waterfall field that appears both in @xmath47 and @xmath54 .
as long as @xmath57 and @xmath58 get a large mass from the superpotential or the khler potential , the waterfall can only start along the @xmath56 direction , which is stabilized by @xmath59 during inflation . to avoid massless directions in the @xmath60 hypersurface after inflation .
an explicit example is given in section [ sec : a4model ] . ] the field @xmath61 which appears only in @xmath54 and not in @xmath47 will not get a vacuum expectation value after inflation , and should therefore not be identified with a higgs field , but with a conventional matter field like a right - handed neutrino .
@xmath62{graphics / h3.pdf } & \includegraphics[width=0.48\textwidth]{graphics / m3.pdf } \\
\includegraphics[width=0.48\textwidth]{graphics / y3.pdf } & \includegraphics[width=0.48\textwidth]{graphics / yx3.pdf } \end{array}$ ] after inflation , tribrid inflation predicts either a mass or a yukawa coupling for the inflaton direction , depending on @xmath63 and on whether we use @xmath64 or @xmath65 : example application + ' '' '' @xmath66 & @xmath67 & @xmath68 & neutrino majorana mass term + ' '' '' @xmath69 & @xmath70 & @xmath71 & quark or electron yukawa coupling + ' '' '' @xmath72 & @xmath67 & @xmath73 & neutrino yukawa coupling + the examples mentioned in the table are * generating a neutrino majorana mass term from @xmath74 , where the right - handed sneutrino @xmath75 is the inflaton direction * generating a quark or charged lepton yukawa coupling from e.g. @xmath76 , where @xmath77 is the inflaton direction * generating a neutrino yukawa coupling from e.g. @xmath78 , where @xmath79 is the inflaton direction and @xmath75 is zero during and after inflation . the factor 2 in the table for the generated mass appears if a majorana mass is generated . if one uses a composite inflaton direction @xmath80 , one instead generates a dirac mass term , and the factor 2 is absent .
slow - roll inflation links the model parameters to observables in the cosmic microwave background ( cmb ) , so the parameters can be constrained from measurements of the spectral index @xmath81 and the running of the spectral index @xmath1 .
the resulting constraints on @xmath82 and @xmath83 after inflation are shown in fig .
[ fig : l3predict ] . is given as a prediction depending on the inflaton s khler potential coupling @xmath84 and the spectral index @xmath81 .
for this paper , we inverted @xmath85 and fixed the coupling @xmath84 by the observables @xmath1 and @xmath81 , which makes the plots more convenient for comparing tribrid models with observations . ]
inflation also constrains the symmetry breaking scale of the waterfall transition by constraining the value of @xmath86 .
this translates into bounds on @xmath43 , which are also shown in fig .
[ fig : l3predict ] . note that @xmath87 is a free parameter of the model .
the plots are shown for @xmath88 and @xmath89 . for smaller @xmath87 and larger @xmath22 , the predicted masses and yukawa couplings are larger due to @xmath90 , with @xmath91 fixed by inflation .
similar plots for @xmath92 can be found in appendix [ appendix : kahlerplots ] .
we now want to illustrate how tribrid inflation can be implemented in an explicit particle physics flavour model .
we present an example model based on a discrete @xmath0 family symmetry family symmetry , see e.g. @xcite . ] to generate the leptonic yukawa couplings and neutrino masses .
we will assume that in addition to the family symmetry , there is also a cp symmetry which is broken spontaneously only by the vacuum expectation values ( vevs ) of the flavons .
one can easily realize tribrid inflation in this model using either the d - flat @xmath93 direction or a sneutrino direction .
the model then predicts a relationship between the running of the spectral index @xmath1 and the neutrino yukawa coupling or the heavy neutrino mass .
interestingly , the production of topological defects during the waterfall transition may be avoided , because the waterfall field can be slightly shifted already during inflation .
this is related to our choice of the flavon alignment potential , as we will discuss later in section [ sec : topologicalmechanism ] .
our model is an alternative to flavon inflation @xcite , where one of the @xmath0 symmetry breaking flavons is the inflaton .
both models use similar types of superpotentials and relate the flavour symmetry breaking scale to the amplitude of the cmb fluctuations .
however , in our tribrid model the inflaton is a combination of matter fields .
since the cmb spectrum is related to the inflaton couplings , this can lead to particularly close relations between observables from inflation and particle physics .
one advantage of using flavons as the inflaton @xcite is that topological defects from the @xmath0 symmetry breaking are automatically inflated away , because the symmetry breaking occurs at the beginning and not at the end of inflation .
our model instead requires a mechanism to avoid or remove topological defects from the waterfall transition , like the one discussed in section [ sec : shiftedtribrida4 ] .
the model is based on the superpotential @xmath94 where summation over @xmath95 , @xmath96 , @xmath67 is implied .
the lepton doublets @xmath97 and the flavons @xmath98 are @xmath0 triplets , while the flavon @xmath99 is an @xmath0 singlet . in this expression ,
the first terms are the lepton yukawa couplings and the right - handed neutrino masses .
@xmath100 is the flavon alignment superpotential that enforces the correct breaking of the @xmath0 flavour symmetry via vacuum expectation values @xmath101 .
@xmath102 contains the quark yukawa couplings and possibly additional interactions which are not relevant for our purpose .
how this model relates to the simple tribrid superpotential from eq . will be explained in section [ sec : inflatondirections ] .
the flavon alignment superpotential of our model has the following form : @xmath103 ^ 6 - \mu_5 ^ 2 \right ) + d_5 \left\ { ( \theta_5 ^ 2 ) ^2 + k_5 ( \theta_5 \star \theta_5 ) ^2 \right\ } + p_{35 } \theta_3 \theta_5 \notag \\ & + s_1 \left ( \theta_1^{n_1 } - \mu_1 ^ 2 \right ) + p_{16 } \theta_1 \theta_6 \notag \\ & + \kappa s_2 \left ( \theta_2 ^3 - \mu_2 ^ 2 \right ) + \alpha_1 d_2 ' ( \theta_2 ^ 2 ) _ { 1 '' } + \alpha_2 d_2 '' ( \theta_2 ^ 2 ) _ { 1 ' } + p_{12 } \theta_1 \theta_2 .
\label{eq : flavonw}\end{aligned}\ ] ] terms @xmath104^n$ ] stand for all possible @xmath0 contractions of the field .
representations , we use the ma - rajasekaran ( `` @xmath105-like '' ) basis , as introduced in the first reference of @xcite . ]
@xmath106 is the symmetric @xmath0 product of three triplets .
the symmetries and charge assignments for the entire superpotential are discussed in appendix [ appendix : symmetries ] .
prefactors which are not needed in the later calculations are dropped for brevity .
@xmath87 , @xmath107 , @xmath108 and @xmath109 are real in a suitable basis due to the assumed cp symmetry ( before family symmetry breaking ) .
@xmath110 , @xmath111 , @xmath112 and @xmath113 can take even integer values and @xmath114 any integer value greater than two . the flavon alignment potential fixes the vacuum expectation values in the true vacuum to be @xmath115 where the @xmath116 can be chosen independently from each other with the constraint that @xmath117 , and @xmath118 , with global phases of the @xmath119 absorbed by redefinitions of the fields . for @xmath120 in eq . , the terms for @xmath121 lead to a phase difference of @xmath122 between the two components of the flavon vev
the relative size of the components , parameterized by @xmath123 , can be adjusted by the value of @xmath124 ( cf .
@xcite ) . for @xmath125
, this vacuum alignment will generate leptonic yukawa matrices of the form @xmath126 and a diagonal right - handed neutrino mass matrix vanishes because @xmath127 . ]
@xmath128 @xmath129 and @xmath130 generate the light neutrino masses via the seesaw mechanism and realize so - called `` constrained sequential dominance '' @xcite where the mixing in the neutrino sector is tri - bimaximal @xcite . the total leptonic mixing receives a contribution from the 1 - 2 mixing in @xmath131 .
this induces a non - zero @xmath132 @xmath133 furthermore , the leptonic dirac cp phase @xmath134 is maximal , @xmath135 due to the phase difference between the components of @xmath136 . plugging this into the lepton mixing sum rule @xcite , @xmath137 , which is applicable here since
the 1 - 3 mixings in the mass matrix of the light neutrinos and in @xmath131 both vanish , we can further see that @xmath138 is not affected by charged lepton 1 - 2 mixing , and therefore is predicted to be close to the tri - bimaximal value . also @xmath139 does not receive a contribution from charged lepton mixing , and we obtain @xmath140 thus , @xmath141 , @xmath142 and @xmath134 are predicted by the model , while @xmath143 can be fitted by choosing a suitable @xmath123 , which in turn depends on the parameter @xmath124 in the flavon alignment potential .
is linked to the quark sector , which can lead to a prediction also for @xmath132 . ]
the first step for implementing tribrid inflation in any model is to find suitable inflaton and waterfall directions in the superpotential . in our @xmath0 flavour model ,
possible waterfall directions are given by the flavon directions @xmath144 and possible inflaton directions are given by d - flat combinations of matter fields . as we have mentioned in section
[ sec : tribridmodelbuilding ] , khler - driven tribrid inflation requires a superpotential term of the form @xmath145 or @xmath146 , where @xmath12 is the waterfall field , @xmath49 is a d - flat inflaton direction and @xmath61 is some matter field with @xmath147 during and after inflation . in the superpotential ,
the candidate terms are 1 .
@xmath148 , with @xmath149 , @xmath53 and @xmath150 , 2 .
@xmath151 , with @xmath152 , @xmath153 , for any @xmath154 , without implied summation over @xmath155 . in the first case ,
@xmath156 during inflation , and tribrid inflation predicts the yukawa coupling @xmath157 as a function of @xmath1 ( fig .
[ fig : l3predict ] ) . in the second case ,
the right - handed sneutrino @xmath158 is the inflaton and tribrid inflation predicts its mass @xmath159 as a function of @xmath1 ( fig .
[ fig : l3predict ] ) .
we assume that @xmath99 and all @xmath144 ( @xmath160 ) except @xmath161 have already settled at their minimum 60 e - folds before the end of inflation , so that only the dynamics of the inflaton @xmath49 and the chosen waterfall field @xmath161 are important . using @xmath162 as
the inflaton direction is straightforward , as @xmath158 does not have any @xmath0 structure , and the operator @xmath163 stabilizes all components of the @xmath0 triplet flavon @xmath161 during inflation .
the relevant terms during inflation exactly reduce to the simple structure of eq . , and we expect that the usual results for khler - driven tribrid inflation apply . can be ignored during inflation because @xmath164 is enforced by the flavon alignment potential . ] using @xmath93 as the inflaton direction , we must account for the possibility that the @xmath165 , @xmath166 , can induce a large f - term potential for the inflaton : @xmath167 during inflation , these terms can generate a large inflaton potential except if @xmath168 .. this ambiguity disappears if we include hubble - scale mass terms for the @xmath169 , which can be naturally generated from the khler potential .
] this means that the inflaton component with @xmath170 will be quickly driven to zero .
if inflation lasts more than the bare minimum of 60 e - folds , we can therefore assume that @xmath168 throughout the last 60 e - folds of inflation . for @xmath171 or @xmath172 , the model
then again reduces to eq . , with @xmath173 , where @xmath174 is the inflaton direction in @xmath0 space , normalized to @xmath175 .
the case @xmath176 is more complicated and will be discussed below .
many models of hybrid inflation can produce dangerous topological defects at the end of inflation , and the above model is no exception . in our case , the waterfall transition for the @xmath161 usually generates @xmath177 domain walls .
the model therefore needs some mechanism to either prevent their production or to remove the domain walls after they have formed . in tribrid inflation ,
the problem is alleviated by the fact that the inflaton can be charged under the symmetries of the waterfall field , and therefore those symmetries are usually broken already during inflation .
if the symmetry breaking is transmitted to the waterfall field , e.g. via additional inflaton - waterfall field couplings , the formation of topological defects at the end of inflation may be avoided .
@xcite however , in this paper s @xmath0 model there is one unique inflaton direction for which the production of domain walls is avoided automatically : the @xmath93 direction with @xmath178 as the waterfall field ( @xmath176 ) .
we want to discuss this case in more detail in section [ sec : shiftedtribrida4 ] .
in this section , we have presented an explicit flavour model based on an @xmath0 symmetry , which makes testable predictions for @xmath179 , @xmath139 and for the dirac cp phase @xmath134 .
we have shown that tribrid inflation can be easily realised in these models : we found six possible inflaton trajectories which predict a relation between the running of the spectral index @xmath1 and either one of the neutrino yukawa couplings @xmath129 or one of the right - handed neutrino masses @xmath7 .
the analysis of this section can be easily repeated for other flavour models based on different family symmetries and flavon alignment potentials to realise tribrid inflation in various different models that are motivated from particle physics .
as we have discussed in section [ sec : inflatondirectionstopological ] , one can use different mechanisms to avoid the overproduction of topological defects at the end of inflation . in this section ,
we discuss a particular inflaton trajectory for which topological defects are avoided automatically , because the waterfall field has a small shift during inflation . in our model , this naturally occurs for inflation along the @xmath93 direction with @xmath178 as the waterfall field .
we restrict our analysis to that special case throughout this section
. we will need to keep track of the different @xmath0 triplet components of both the inflaton and the waterfall field , so it will be useful to introduce the notation @xmath180 where we chose an inflaton direction which satisfies the condition @xmath181 which we have explained in section [ sec : lhudirections ] .. we focus on one specific direction , which suffices to demonstrate how the proposed mechanism works . ] we should also note that during inflation @xmath182 can have any value @xmath183 the flavon alignment potential only fixes @xmath184 and @xmath185 when @xmath178 has a sufficiently large vacuum expectation value , i.e. after the waterfall transition . during inflation , @xmath184 and @xmath185
are instead determined by their initial conditions . to see that @xmath178 is non - zero during inflation , we need to minimize its scalar potential during inflation .
the relevant superpotential terms are @xmath186 we use the @xmath0 contractions , with @xmath187 : @xmath188 when calculating the f - term potential , we must include the khler potential coupling @xmath189 the f - term potential then becomes @xmath190 the last term enforces @xmath164 , which implies that @xmath191 and which also guarantees that we can ignore the operator @xmath192 throughout this calculation . to find the shifts in @xmath193 and @xmath194 , we approximate @xmath195 and @xmath196 .
we will later see under which conditions this approximation is valid . with this approximation and eq .
, the potential simplifies to @xmath197 minimizing the potential with respect to @xmath198 gives @xmath199 with @xmath200 for @xmath201 , one can show that eq
. implies @xmath202 .
note that the potential is not minimized with respect to @xmath203 . instead , @xmath184 and @xmath185 are constants given by the initial conditions , because in general their potential is very flat compared to the inflaton potential , and therefore the fields can be treated as constant during inflation . and @xmath185 initially minimize the f - term potential for @xmath204 , so the only term which induces a potential for @xmath184 and @xmath185 is the interaction term @xmath205 , which creates a potential for @xmath182 and @xmath178 when they are not perpendicular to each other
. however , this potential for each field is proportional to the vacuum expectation value of the other field , and due to @xmath206 , the potential will generally be much greater for @xmath178 and thus be minimized by an adjustment in @xmath178 . ] using the shifts in @xmath193 and @xmath194 , but still assuming that @xmath208 , we can find the minimum for @xmath207 by minimizing eq . :
@xmath209 the two dominant terms are @xmath210 and the linear term in @xmath207 .
the other mass terms for @xmath207 are negligible during inflation because @xmath211 , and @xmath212 , @xmath213 .
the @xmath214 term is also negligible , as we will find that @xmath215 .
we can therefore minimize the simple potential @xmath216 with respect to @xmath207 .
the result is @xmath217 during inflation , we have @xmath218 , and therefore @xmath219 , so @xmath207 is suppressed with respect to @xmath193 , @xmath220 .
this justifies our assumption in section [ sec : shiftyz ] , where we set @xmath195 for the calculation of the shifts in @xmath193 and @xmath194 . near the critical point
, we find that @xmath207 is no longer suppressed with respect to @xmath193 and @xmath194 : @xmath221 . in this case ,
@xmath193 and @xmath194 are no longer stabilized by @xmath222 and @xmath223 , and we expect @xmath178 to roll down the potential along the direction @xmath224 . therefore , near the critical inflaton value , we could have a smooth transition towards the minimum of the waterfall field instead of a sudden waterfall at @xmath225 .
anyway , as the waterfall potential is steep , we expect inflation to end quickly near @xmath226 , so that the model still approximately acts as a model of khler - driven tribrid inflation .
we expect that the predictions for @xmath1 and @xmath43 are not influenced by the shift in @xmath178 , as they depend only on the inflaton potential far away from the critical point .
however , @xmath227 ( and the derived @xmath228 ) might be changed by approximately up to a factor of @xmath229 , where @xmath230 is the number of e - folds before the critical point at which the shift in @xmath178 becomes relevant . happens if the shift in @xmath178 either accelerates or delays the phase transition , because @xmath227 must be adjusted to get the correct number of e - folds between @xmath231 and the end of inflation . a precise prediction of @xmath227 therefore requires an analysis of the complete multi - field model near @xmath226 including all components of @xmath178 , and possibly a quantitative treatment of the tachyonic preheating phase .
] @xmath232 and @xmath233 also get small shifts due to the shift in @xmath207 .
however , these are completely negligible if @xmath232 and @xmath233 also get positive hubble - size masses from the khler potential : in that case , one can easily show that @xmath234 .
we therefore set @xmath232 , @xmath196 in this paper .
the shift in @xmath178 induces a small correction to the inflaton potential from @xmath235 for @xmath236 , the extra inflaton potential that is induced by @xmath237 is therefore much smaller than the inflaton potential @xmath238 .
we conclude that despite the shift in @xmath178 , inflation is mostly khler - driven , except possibly near @xmath226 around which we expect inflation to end anyway .
therefore we can approximately use the predictions for @xmath239 and @xmath240 from fig .
[ fig : l3predict ] , which have been calculated for purely khler - driven tribrid inflation .
a simple variant of this model can be obtained by replacing the vacuum energy term for @xmath178 in the superpotential : @xmath242 the replacement suppresses the shift in @xmath207 during inflation by another power of @xmath243 , which makes it completely negligible .
this improves the quality of our approximations and we expect that the model will behave more exactly like purely khler - driven tribrid inflation , even very close to @xmath226 .
nevertheless , this setup may still remove domain walls during the waterfall transition , as the potential still prefers a specific phase for @xmath207 when @xmath207 , @xmath193 and @xmath194 grow during the waterfall transition . however , it is not clear whether the preference is sufficiently large to either prevent the production of domain walls or to efficiently remove them if they are produced . to answer this question ,
the preheating phase should be studied in detail , which is beyond the scope of this paper .
we have seen that if the flavon @xmath178 is identified with the waterfall field and @xmath93 with the inflaton , the waterfall field obtains a small shift during inflation , which can help with preventing domain walls .
however , if we had used @xmath182 or @xmath244 as the waterfall field or @xmath233 as the inflaton , no such shift would have occured during inflation . in this section ,
we briefly want to sketch which characteristics of the superpotential determine whether the waterfall field obtains a shift during khler - driven tribrid inflation or not .
the main condition can be split in three parts : 1 .
the waterfall field @xmath12 consists of several fields , e.g. it is a multiplet under some symmetry .
2 . at least one of the multiplet components of @xmath12 does not receive a mass from the inflaton .
3 . the same multiplet component of @xmath12
does not receive a mass term from the alignment potential , but is instead stabilized by higher - order terms only .
let us apply these considerations to our example model : the first condition restricts the range of waterfall fields for which this mechanism can work . in our model it is satisfied because we use an @xmath0 triplet flavon as the waterfall field .
the second condition explains why we do not get a shifted waterfall field when we use @xmath233 as the inflaton : the superpotential term @xmath245 provides masses to all three @xmath0 components of @xmath178 as long as @xmath246 , so the waterfall field is stabilized exactly at zero during inflation .
the third condition explains why using @xmath182 or @xmath244 does not lead to shifted waterfall trajectories .
although an @xmath93 inflaton only provides masses to one of the three @xmath0 components of @xmath161 , the other two components acquire mass terms from the flavon alignment potential via the terms @xmath247 and @xmath248 .
therefore , during inflation all three @xmath0 directions of the waterfall field are stabilized at zero , and no shift can occur .
we have discussed how khler - driven tribrid inflation can be implemented in realistic particle physics models , using a d - flat combination of charged matter fields as the inflaton and a higgs field as the waterfall field to terminate inflation .
we explained how one can identify suitable inflaton and waterfall fields and which predictions tribrid inflation makes for the model parameters .
our strategy can be easily applied to a wide range of models based on different symmetries , leading to predictive models of inflation in motivated particle physics models . as an explicit example
, we have analysed a particular flavour model based on a spontaneously broken @xmath0 family symmetry .
the inflaton can be either a right - handed sneutrino or a d - flat @xmath93 direction , and tribrid inflation predicts a relation between the running of the spectral index @xmath1 and either the right - handed neutrino mass @xmath7 or the neutrino yukawa coupling @xmath129 .
the model also predicts @xmath249 and a relation between the mixing angles and the leptonic dirac cp phase .
further predictions could be obtained from a detailed study of the reheating phase @xcite .
as the inflaton is composed of mssm fields , its couplings to the visible sector are known ( or could be measured with a collider ) , so one could calculate extra bounds from leptogenesis or from possible overproduction of gravitinos or dark matter .
though such an analysis is beyond the scope of our paper , it is important to realize that by realizing inflation in the visible matter sector , models of tribrid inflation can be very predictive in principle .
we have also shown that in our @xmath0 model , the formation of topological defects at the end of inflation can be avoided automatically , because the waterfall field can have a small shift .
this is related to our choice of the flavon alignment potential .
we expect that such a shift could be generated in models with other discrete family symmetries as well .
this work was supported by the swiss national science foundation .
we thank vinzenz maurer and constantin sluka for useful discussions .
in this section , we provide plots for the predictions for khler - driven tribrid inflation for different superpotential parameters @xmath22 .
the plots are based on the calculations in @xcite and normalized to the experimental data from the planck satellite @xcite .
@xmath62{graphics / h4.pdf } & \includegraphics[width=0.48\textwidth]{graphics / m4.pdf } \\
\includegraphics[width=0.48\textwidth]{graphics / y4.pdf } & \includegraphics[width=0.48\textwidth]{graphics / yx4.pdf } \end{array}$ ] for @xmath45 , loop corrections are generally suppressed , and the results should be fair approximations to the exact slow - roll results .
more details can be found in @xcite . for @xmath27 ,
tribrid inflation can be successfully realized , but the predictions depend on both the khler potential and the one - loop effective potential . in some cases ,
the loop potential dominates , e.g. if the khler potential contributions vanish due to some symmetry @xcite ; in other cases , the khler potential is dominant and the loop potential can be neglected . in the more general mixed case , however , one should include both contributions to the inflaton potential for calculating the slow - roll predictions .
@xmath62{graphics / h5.pdf } & \includegraphics[width=0.48\textwidth]{graphics / m5.pdf } \\
\includegraphics[width=0.48\textwidth]{graphics / y5.pdf } & \includegraphics[width=0.48\textwidth]{graphics / yx5.pdf } \end{array}$ ] @xmath62{graphics / h6.pdf } & \includegraphics[width=0.48\textwidth]{graphics / m6.pdf } \\
\includegraphics[width=0.48\textwidth]{graphics / y6.pdf } & \includegraphics[width=0.48\textwidth]{graphics / yx6.pdf } \end{array}$ ]
when we compose the inflaton as @xmath250 , e.g. @xmath251 , the inflaton potential @xmath252 depends on all @xmath253 independently . in this section , we show that the inflaton trajectory during khler - driven tribrid inflation nevertheless has the same inflaton potential as if it was a single scalar field .
this ensures that the predictions for the single - field case still apply even when we replace @xmath254 . , which shifts @xmath42 by a factor of @xmath255 .
] we assume that the khler potential depends only on the modulus squared of the superfields : @xmath256 the inflaton potential , expressed in terms of canonically normalized real scalar fields @xmath257 , then turns out to be : @xmath258 with the coefficients .
@xcite ] @xmath259 these coefficients depend entirely on the khler potential , and if the cutoff scale is around the planck scale , we find that @xmath260 , @xmath261 . in this section , we work with only two fields @xmath262 and @xmath263 to keep the discussion simple : @xmath264 from this function , we want to derive the one - dimensional inflaton potential @xmath265 along the inflaton trajectory .
the d - term will generally depend on the gauge charges of the @xmath253 . in our example @xmath262 and @xmath263 : @xmath266 ^ 2 + \frac{g_2 ^ 2}{8 } \sum\limits_{i=1}^{3 } \left [ \left ( l^\dagger , h_u^\dagger \right ) \begin{pmatrix } \sigma^i & 0 \\ 0 & \sigma^i \end{pmatrix } \begin{pmatrix }
h_u \end{pmatrix } \right]^2 \notag \\ & = \ , \frac{g_1 ^ 2}{8 } \left ( \lvert l \rvert^2 - \lvert h_u \rvert^2 \right)^2 + \frac{g_2 ^ 2}{8 } \sum\limits_{i=1}^{3 } \left ( l^\dagger \sigma^i
l + h_u^\dagger \sigma^i h_u \right)^2.\end{aligned}\ ] ] the second term enforces that @xmath97 and @xmath267 have opposite weak isospin . as the inflaton potential does not otherwise depend on the @xmath268 structure of @xmath97 and @xmath267 , this condition will be satisfied during inflation , and the term will be zero .
the first term enters the inflaton potential @xmath269 : @xmath270 where we defined @xmath271 .
the terms with @xmath272 dominate over the small coefficients @xmath273 , @xmath261 .
therefore , we expect that the inflaton trajectory will be along a nearly d - flat direction .
looking at trajectories near the d - flat trajectory @xmath274 , we define @xmath275 if we insert this in eq . , we get @xmath276 with @xmath277 and @xmath278 \delta^2 .
\label{eq : vdelta}\end{aligned}\ ] ] where @xmath84 , @xmath279 describe the averaged potential for @xmath280 and @xmath281 , while @xmath282 , @xmath283 quantify the asymmetries between the potentials for @xmath280 and @xmath281 : @xmath284 in these equations , @xmath285 contains the couplings along the perfectly d - flat direction , while @xmath286 contains the potential due to the inflaton s deviation from d - flatness
. the minimum of @xmath287 can be found by minimizing eq . :
@xmath288 if we assume that @xmath287 tracks its minimum perfectly , then the effective inflaton potential is given by @xmath289 the inflaton potential is changed by @xmath290 with @xmath291 and @xmath292 , @xmath293 , this correction is negligible compared to @xmath285 , so assuming d - flatness is a very good approximation .
we should still check that @xmath294 to justify our assumption that @xmath287 tracks its minimum : @xmath295 for khler - driven tribrid inflation , the inflaton field value during inflation is bounded by @xmath296 , which we can insert in eq . to find @xmath297 therefore , @xmath298 , and our assumption that @xmath299 is warranted .
we conclude that any deviation from d - flatness during khler - driven tribrid inflation is completely negligible , and the inflaton potential indeed reduces completely to the simple single - field form which has been previously studied in @xcite .
the superpotential can be fixed by a set of suitable shaping symmetries in addition to the @xmath0 family symmetry .
in particular , one can use an @xmath300 symmetry , under which the superpotential has charge 2 , and one @xmath177 symmetry for each flavon field .
the @xmath300 charge assignment is straightforward .
the mass parameters @xmath301 in the superpotential can not be charged , so the @xmath302 must have 2 units of @xmath300 charge .
then the @xmath144 can not have any @xmath300 charge . knowing this
, we also find that the other auxiliary fields @xmath303 , @xmath304 and @xmath305 need 2 units of @xmath300 charge .
the only freedom we have is in the lepton sector , where we can distribute the @xmath300 charge between the higgs doublets @xmath267 , @xmath306 and the lepton fields @xmath97 , @xmath169 , @xmath307 .
we choose to keep the higgs fields uncharged and give 1 unit of @xmath300 charge to each lepton . to construct the @xmath308 symmetries , we start with the observation that we want to keep the different flavons separate in the @xmath309 terms . for this reason ,
we start with one @xmath310 symmetry for each flavon , with any set of positive integers @xmath311 , under which the @xmath95-th flavon has charge @xmath311 . to simplify the notation , we instead work with @xmath312 and allow fractional charges for the other fields , except for the @xmath0 singlet flavon @xmath99 , which we normalize to 2 units of charge . with the flavon charges fixed
, we can uniquely determine the @xmath308 charge assignments for the auxiliary fields @xmath302 , @xmath303 , @xmath304 and @xmath305 by demanding that all terms in the superpotential must be allowed by the @xmath308 symmetries .
now the only remaining task is to make sure that the neutrino yukawa coupling and mass terms are allowed by the @xmath308 symmetries .
there are several possible solutions , of which we have chosen a particularly simple one for table [ tab : symmetrycharges ] .
k. abe _ et al . _
[ t2k collaboration ] , _ indication of electron neutrino appearance from an accelerator - produced off - axis muon neutrino beam _ , phys .
lett . * 107 * ( 2011 ) 041801 [ arxiv:1106.2822 [ hep - ex ] ] . y. abe _ et al . _
[ double - chooz collaboration ] , _ indication for the disappearance of reactor electron antineutrinos in the double chooz experiment _ ,
* 108 * ( 2012 ) 131801 [ arxiv:1112.6353 [ hep - ex ] ] .
j. k. ahn _
_ [ reno collaboration ] , _ observation of reactor electron antineutrino disappearance in the reno experiment _
* 108 * ( 2012 ) 191802 [ arxiv:1204.0626 [ hep - ex ] ] .
f. p. an _ et al .
_ [ daya - bay collaboration ] , _ observation of electron - antineutrino disappearance at daya bay _ ,
* 108 * ( 2012 ) 171803 [ arxiv:1203.1669 [ hep - ex ] ] ; f. p. an _ et al . _
[ daya - bay collaboration ] , _ improved measurement of electron antineutrino disappearance at daya bay _ , chin .
c * 37 * ( 2013 ) 011001 [ arxiv:1210.6327 [ hep - ex ] ] . for recent reviews on flavour models ,
see e.g. : s. f. king and c. luhn , _ neutrino mass and mixing with discrete symmetry _
* 76 * ( 2013 ) 056201 [ arxiv:1301.1340 [ hep - ph ] ] ; g. altarelli , f. feruglio and l. merlo , _ tri - bimaximal neutrino mixing and discrete flavour symmetries _ , arxiv:1205.5133 [ hep - ph ] .
a. h. guth , _ the inflationary universe : a possible solution to the horizon and flatness problems _ , phys .
* d23 * ( 1981 ) , 347356 .
a. d. linde , _ a new inflationary universe scenario : a possible solution of the horizon , flatness , homogeneity , isotropy and primordial monopole problems _ , phys . lett .
* b108 * ( 1982 ) , 389393 . for
a review containing an extensive list of references , see e.g. : d. h. lyth and a. riotto , _ particle physics models of inflation and the cosmological density perturbation _
* 314 * ( 1999 ) 1 [ hep - ph/9807278 ] .
i. masina and a. notari , _ standard model false vacuum inflation : correlating the tensor - to - scalar ratio to the top quark and higgs boson masses _ , phys .
* 108 * ( 2012 ) 191302 , arxiv:1112.5430 [ hep - ph ] .
r. allahverdi , k. enqvist , j. garcia - bellido and a. mazumdar , _ gauge invariant mssm inflaton _ , phys .
* 97 * ( 2006 ) 191304 [ hep - ph/0605035 ] . c. boehm , j. da silva , a. mazumdar and e. pukartas , _ probing the supersymmetric inflaton and dark matter link via the cmb , lhc and xenon1 t experiments _ , arxiv:1205.2815 [ hep - ph ] .
s. antusch , m. bastero - gil , k. dutta , s. f. king and p. m. kostka , _ solving the @xmath314-problem in hybrid inflation with heisenberg symmetry and stabilized modulus _ ,
jcap * 0901 * ( 2009 ) 040 [ arxiv:0808.2425 [ hep - ph ] ] .
e. ma and g. rajasekaran , _ softly broken a(4 ) symmetry for nearly degenerate neutrino masses _ , phys .
d * 64 * ( 2001 ) 113012 [ hep - ph/0106291 ] ; k. s. babu , e. ma and j. w. f. valle , _ underlying a(4 ) symmetry for the neutrino mass matrix and the quark mixing matrix _ ,
b * 552 * ( 2003 ) 207 [ arxiv : hep - ph/0206292 ] ; g. altarelli and f. feruglio , _ tri - bimaximal neutrino mixing , a(4 ) and the modular symmetry _ , nucl .
b * 741 * ( 2006 ) 215 [ hep - ph/0512103 ] .
s. antusch , c. gross , v. maurer and c. sluka , _ a flavour gut model with @xmath315 _ , arxiv:1305.6612 [ hep - ph ] ; s. antusch , c. gross , v. maurer and c. sluka , _ inverse neutrino mass hierarchy in a flavour gut model _ , arxiv:1306.3984 [ hep - ph ] .
p. f. harrison , d. h. perkins and w. g. scott , _ tri - bimaximal mixing and the neutrino oscillation data _ , phys .
b * 530 * ( 2002 ) 167 [ hep - ph/0202074 ] .
s. f. king , _ predicting neutrino parameters from so(3 ) family symmetry and quark - lepton unification _ , jhep * 0508 * ( 2005 ) 105 [ arxiv : hep - ph/0506297 ] ; i. masina , _ a maximal atmospheric mixing from a maximal cp violating phase _ , phys .
b * 633 * ( 2006 ) 134 [ arxiv : hep - ph/0508031 ] ; s. antusch and s. f. king , _ charged lepton corrections to neutrino mixing angles and cp phases revisited _ , phys .
b * 631 * ( 2005 ) 42 [ hep - ph/0508044 ] ; for a review , see e.g. : r. allahverdi , r. brandenberger , f. -y .
cyr - racine and a. mazumdar , _ reheating in inflationary cosmology : theory and applications _ , ann .
nucl . part .
* 60 * ( 2010 ) 27 [ arxiv:1001.2600 [ hep - th ] ] . | we discuss model building in tribrid inflation , which is a framework for realising inflation in the matter sector of supersymmetric particle physics models .
the inflaton is a d - flat combination of matter fields , and inflation ends by a phase transition in which some higgs field obtains a vacuum expectation value .
we first describe the general procedure for implementing tribrid inflation in realistic models of particle physics that can be applied to a wide variety of bsm particle physics models around the gut scale .
we then demonstrate how the procedure works for an explicit lepton flavour model based on an @xmath0 family symmetry .
the model is both predictive and phenomenologically viable , and illustrates how tribrid inflation connects cosmological and particle physics parameters . in particular
, it predicts a relation between the neutrino yukawa coupling and the running of the spectral index @xmath1 .
we also show how topological defects from the flavour symmetry breaking can be avoided automatically . + stefan antusch@xmath2 and david nolde@xmath3 @xmath3 _ department of physics , university of basel ,
_ _ klingelbergstr .
82 , ch-4056 basel , switzerland _
@xmath4 _ max - planck - institut fr physik ( werner - heisenberg - institut ) , _ _ fhringer ring 6 , d-80805 mnchen , germany _ |
the @xmath5 galaxy fsc 10214 + 4724 was one of the most remarkable objects detected by the survey .
originally proposed to be the most luminous galaxy known ( rowan - robinson 1991 ) , multi - wavelength observations subsequently showed that it is lensed by an intervening @xmath6 galaxy , boosting its intrinsic emission by a factor of @xmath7 10100 ( depending on the location and extent of the unlensed emission with respect to the caustic ; e.g.,broadhurst & lehar 1995 ; downes 1995 ; trentham 1995 ; eisenhardt 1996 ; evans 1999 ) .
optical and near - ir spectroscopic / polarimetric observations have unambiguously shown that fsc 10214 + 4724 hosts an obscured active galactic nucleus ( agn ; e.g.,elston 1994 ; soifer 1995 ; goodrich 1996 ) .
multi - wavelength analyses have suggested that the agn is powerful ( e.g. , goodrich 1996 ; granato 1996 ; green & rowan - robinson 1996 ) , although it is generally accepted that star - formation activity dominates the bolometric output ( e.g.,rowan - robinson 1993 ; rowan - robinson 2000 ) .
the lensing - corrected properties of fsc 10214 + 4724 are similar to those of galaxies ( e.g. , blain 2002 ; ivison 2002 ; smail 2002 ; chapman 2003 ; neri 2003 ) : it lies at @xmath8 , is optically faint with @xmath9 25 , has an 850@xmath10 m flux density of a few mjy and a 1.4 ghz flux density of @xmath0 25 @xmath10jy , is massive ( a molecular gas mass of @xmath11@xmath12 @xmath13 ) , and has a bolometric luminosity of @xmath14 l@xmath15 ( e.g. , rowan - robinson 1993 ; downes 1995 ; eisenhardt 1996 ) .
galaxies appear to be massive galaxies undergoing intense star - formation activity and are likely to be the progenitors of the local @xmath7 @xmath16 early - type galaxy population ( e.g. , blain 2004 ; chapman 2004 ) .
deep x - ray observations have shown that a large fraction ( at least @xmath0 40% ; alexander 2003b , 2004 ) of the galaxy population is actively fuelling its black holes during this period of intense star formation , which can account for a significant fraction of the black - hole growth in massive galaxies ( alexander 2004 ) .
the agn in fsc 10214 + 4724 was identified via optical / near - ir observations , and the large lensing boost of the agn emission provides interesting insight into agn activity in the galaxy population . in this letter we present the results from a @xmath0 20 ks observation of fsc 10214 + 4724 with the _ chandra x - ray observatory _ ( hereafter _ chandra _ ; weisskopf et al .
2000 ) . due to the large lensing boost of fsc 10214 + 4724
, these observations provide the equivalent sensitivity of an up - to @xmath0 4 ms exposure .
the galactic column density toward fsc 10214 + 4724 is ( stark et al .
@xmath17 km s@xmath3 mpc@xmath3 , @xmath18 , and @xmath19 are adopted .
fsc 10214 + 4724 was observed with _ chandra _ on 2004 march 4 ( observation i d 4807 ) .
the advanced ccd imaging spectrometer ( acis ; garmire et al . 2003 ) with the ccd s3 at the aim point was used for the observation ( the ccds s1s4 , and i2i3 were also turned on ) ; the optical position of fsc 10214 + 4724 is @xmath20 10@xmath21 24@xmath22 3454 , @xmath23 @xmath2409@xmath25 .
faint mode was used for the event telemetry format , and the data were initially processed by the _ chandra _ x - ray center ( cxc ) using version 7.1.1 of the pipeline software . [ cols="^,^,^,^,^,^,^,^,^,^,^ " , ] notes : @xmath26offset between the _ chandra _ full - band source position and the _ hst _ source position ( component 1 ) of eisenhardt et al .
@xmath27source counts and errors .
`` fb '' indicates full band , `` sb '' indicates soft band , and `` hb '' indicates hard band .
the source counts are determined with wavdetect .
the errors correspond to 1 @xmath28 and are taken from gehrels ( 1986 ) .
@xmath29ratio of the count rates in the 2.08.0 kev and 0.52.0 kev bands .
the errors were calculated following the numerical method " described in 1.7.3 of lyons ( 1991 ) .
@xmath30effective photon index for the 0.58.0 kev band , calculated from the band ratio .
the photon index is related to the energy index by @xmath31 = @xmath32 where @xmath33 .
@xmath34fluxes in units of @xmath35 erg @xmath36 s@xmath3 .
these fluxes have been calculated from the count rate in each band using the cxc s portable , interactive , multi - mission simulator ( pimms ) assuming @xmath37 ; they have been corrected for galactic absorption . + the reduction and analysis of the data used _
chandra _ interactive analysis of observations ( ciao ) version 3.0.2 tools .
the ciao tool acis_process_events was used to remove the standard pixel randomisation .
the data were then corrected for the radiation damage sustained by the ccds during the first few months of _ chandra _ operations using the charge transfer inefficiency ( cti ) correction procedure of townsley et al .
all bad columns , bad pixels , and cosmic ray afterglows were removed using the `` status '' information in the event files , and we only used data taken during times within the cxc - generated good - time intervals .
the background light curve was analysed to search for periods of heightened background activity using the contributed ciao tool analyze_ltcrv with the data binned into 200 s intervals ; there were no periods of high background ( i.e. , a factor of @xmath7 2 above the median level ) .
the net exposure time for the observation is 21.26 ks .
the _ asca _
grade 0 , 2 , 3 , 4 , and 6 events were used in all subsequent analyses .
the pointing accuracy of _ chandra _ is excellent ( the 90% uncertainty is @xmath38 ) .
however , for our observation we wanted to improve the source positions to provide an unambiguous distinction between fsc 10214 + 4724 and nearby objects ( e.g. , the @xmath39 lensing galaxy lies @xmath0 @xmath40 from fsc 10214 + 4724 ; eisenhardt et al .
1996 ) . to achieve this we matched sources detected in the observation to sources found in the sloan digital sky survey ( sdss ) , which has a positional accuracy of @xmath41 ( rms ; pier et al .
source searching was performed using wavdetect ( freeman 2002 ) with a false - positive threshold of @xmath42 in the full ( fb ; 0.58.0 kev ) , soft ( sb ; 0.52.0 kev ) , and hard ( hb ; 28 kev ) bands ; we used wavelet scale sizes of 1 , 1.44 , 2 , 2.88 , 4 , 5.66 , and 8 pixels .
the resulting source lists were then merged with a @xmath43 matching radius , producing a catalog of 37 sources .
the three brightest x - ray sources in this catalog ( those with @xmath4420 counts in the full band ) that lay within @xmath45 of the aim point were matched to sdss sources in the data release 2 catalog ( dr2 ; abazajian 2004 ) with a @xmath43 search radius .
the mean sdss-_chandra _ positional offset of the matched sources ( excluding fsc 10214 + 4724 ) was ( right ascension ; ra ) and ( declination ; dec ) .
these corrections were applied to the x - ray source positions .
fsc 10214 + 4724 is detected in all three bands ( see table 1 ) . the x - ray source position ( taken from the full band )
lies @xmath46 from the position measured by the _ hubble space telescope _
( hereafter ) in the f814w band ( eisenhardt et al . 1996 ; see table 1 & figure 1 ) ; the x - ray position is also @xmath47 offset from both the co(32 ) position ( downes 1995 ) and the 1.49 ghz radio position ( lawrence 1993 ) . the x - ray source is clearly identified with fsc 10214 + 4724 rather than the @xmath39 lensing galaxy ( see figure 1 ) .
the x - ray properties of fsc 10214 + 4724 are shown in table 1 .
although this observation achieves the equivalent sensitivity of an up - to @xmath0 4 ms exposure ( e.g. , a 10-count source has a full - band flux of @xmath7 @xmath48 erg @xmath36 s@xmath3 for a lensing boost of @xmath49 ; compare to table 9 in alexander 2003a ) , only a few x - ray counts are detected .
the @xmath0 10 counts in the soft band correspond to a significant detection in the rest - frame band while the @xmath0 4 counts in the hard band correspond to a weak detection in the rest - frame 6.626.3 kev band .
the soft - band flux is @xmath0 10 times below the 2 @xmath28 constraint reported in lawrence ( 1994 ) and the hard - band flux is @xmath0 20 times below the upper limit reported in iwasawa ( 2001 ) .
we re - examined the pspc image and could not find unambiguous evidence of x - ray emission at the location of fsc 10214 + 4724 . from our analyses of the pspc image we determine a 3 @xmath28 0.52.0 kev upper limit of @xmath50 erg @xmath36 s@xmath3 . the observed ( uncorrected for gravitational lensing )
full - band luminosity is @xmath51 2.4 @xmath52 erg s@xmath3 .
the band ratio ( i.e. , the ratio of the hard to soft - band count rate ) of fsc 10214 + 4724 implies an effective photon index of @xmath53 ( see table 1 ) .
this is generally consistent with that of an unobscured agn ( i.e. , @xmath54 2.0 ; e.g.,nandra & pounds 1994 ; george 2000 ) ; however , due to the large uncertainties and comparatively high redshift of fsc 10214 + 4724 , this could also be consistent with a column density of @xmath55 @xmath36 at @xmath5 ( for an intrinsic x - ray spectral slope of @xmath56 ) .
the latter would be more consistent with the obscured agn classification of fsc 10214 + 4724 than the former ( e.g. , elston 1994 ; soifer 1995 ) .
see 3.1 for further obscuration constraints .
the gravitational lensing boost of fsc 10214 + 4724 is unknown in the x - ray band .
since the lensing boost is a function of the source size , basic constraints can be placed from the extent of the x - ray emission ( e.g. , see 2 of broadhurst & lehar 1995 ) . in figure
2 we show the full - band profiles ( s - n and e - w orientations ) of fsc 10214 + 4724 and compare them to the on - axis acis - s point spread function ( psf ) . while this analysis is limited by small - number statistics , the extent of fsc 10214 + 4724 is consistent with that of an unresolved x - ray source ( @xmath0 1@xmath57 ) .
the half - power radius of fsc 10214 + 4724 ( i.e. , the radius over which the central seven counts are distributed ; ) is also consistent with that of an unresolved x - ray source .
although somewhat uncertain , this suggests that the magnification in the x - ray band is @xmath7 25 ( e.g. , compare to the extent and magnification found in the 2.05 @xmath10 m nicmos observations of evans 1999 ) .
we can compare this estimate to the expected lensing boost from other observations .
the _ hst _
observations and source model of nguyen ( 1999 ) suggest that the central source ( i.e. , the emitting agn ) is @xmath0 100 pc from the caustic , indicating that the magnification of the agn emission is likely to be @xmath0 100 .
the strong optical polarisation and prominent high - excitation emission lines also indicate that the caustic lies close to the central source ( e.g. , broadhurst & lehar 1995 ; lacy 1998 ; simpson 2004 ) .
-axis corresponds to the offset from the wavdetect - determined position , and the @xmath58-axis error bars correspond to 1 @xmath28 uncertainties ( gehrels 1986 ) .
the psf was simulated using the ciao tool mkpsf and has been normalised to the peak of the s - n orientation profile .
although the signal - to - noise ratio of the data is low , the profiles and the half - power radius ( @xmath0 05075 ) are consistent with those of an unresolved source , suggesting a lensing boost in the x - ray band of @xmath7 25 ; see
with a @xmath0 20 ks acis - s observation we have shown that fsc 10214 + 4724 is comparatively weak at x - ray energies ( @xmath59 erg s@xmath3 , @xmath60 erg s@xmath3 , and @xmath61 erg s@xmath3 for a lensing boost of @xmath62 ) .
previous studies have suggested that fsc 10214 + 4724 hosts both a powerful starburst and a powerful agn ( goodrich 1996 ; granato 1996 ; green & rowan - robinson 1996 ) . in this final section
we predict the expected x - ray emission from both star formation and agn activity in fsc 10214 + 4724 and compare it to the observed x - ray emission .
we also compare the x - ray properties of fsc 10214 + 4724 to those of high - redshift galaxies and discuss the x - ray identification of compton - thick agns at high redshift .
@xmath635007 luminosity of fsc 10214 + 4724 for a lensing boost of 100 ( simpson 2004 ; see 3.1 ) ; the light shaded region shows the variance in the @xmath64}$]/@xmath65 relationship ( mulchaey 1994 ) .
the filled circles indicate the x - ray detected galaxies from alexander ( 2003b ) ; the crosses indicate sources classified as agns , and the `` u ' 's indicate sources with unknown classifications but with x - ray properties consistent with those of starburst galaxies .
the dark shaded region denotes the 1 @xmath28 dispersion in the locally determined x - ray - radio correlation for star - forming galaxies ( see figure 6 of shapley , fabbiano , & eskridge 2001 ; bauer 2002 ; ranalli 2003 ) .
the rest - frame @xmath66 8 kev emission from fsc 10214 + 4724 is consistent with that expected from star - formation activity ; see
3.1.,width=321 ] many studies have shown a correlation between the 1.4 ghz radio luminosity density and the x - ray luminosity of star - forming galaxies ( e.g. , shapley 2001 ; bauer 2002 ; ranalli 2003 ) .
since the radio emission from fsc 10214 + 4724 is consistent with star - formation activity ( e.g. , lawrence 1993 ; rowan - robinson 1993 ; eisenhardt 1996 ) , we can use this correlation to predict the expected x - ray emission from star formation . the radio extent and morphology of fsc 10214 + 4724
are similar to those found in the rest - frame ultra - violet , suggesting that the radio emission is lensed by a factor of ( eisenhardt 1996 ) . in figure
3 we show the rest - frame 1.4 ghz radio luminosity density versus the rest - frame 0.58.0 kev luminosity for fsc 10214 + 4724 ; the rest - frame 0.58.0 kev luminosity was calculated from the rest - frame 1.66.6 kev luminosity ( observed soft band ) assuming @xmath37 and is shown for a range of lensing boosts . the rest - frame emission from fsc 10214 + 4724 is entirely consistent with that expected from star formation for the range of probable lensing boosts at radio wavelengths . the x - ray - to - optical flux ratio is also concordant with that expected from star formation [ @xmath67 ; see 4.1.1 of bauer 2004 ] when appropriate @xmath68 corrections and lensing boosts are applied [ @xmath37 with lensing boosts of 25100 in the x - ray band , and the host galaxy templates of mannucci ( 2001 ) with a lensing boost of 100 in the @xmath69-band ] .
these results imply that the agn in fsc 10214 + 4724 is either heavily obscured or intrinsically weak
. we can estimate the instrinsic luminosity of the agn in fsc 10214 + 4724 using the [ oiii]@xmath635007 luminosity ( e.g. , mulchaey 1994 ; bassani 1999 ) .
taking the [ oiii]@xmath635007 luminosity from serjeant ( 1998 ) , the [ oiii]@xmath635007 to x - ray correlation of mulchaey ( 1994 ) , and assuming the lensing boost to the [ oiii]@xmath635007 emission - line region is @xmath70 ( simpson 2004 ) , the predicted rest - frame 0.58.0 kev luminosity is @xmath71 erg s@xmath3 ( with a variance of @xmath52 erg s@xmath3 ) ; see figure 3 . , a typical intrinsic x - ray spectral slope for agns .
] these predicted x - ray luminosities are within the range expected for quasars and would be even higher if the [ oiii]@xmath635007 emission - line region suffers from reddening ( e.g. , elston 1994 ; soifer 1995 ; cf serjeant 1998 ) .
since the lensing boost to the [ oiii]@xmath635007 emission - line region is a lower limit , these constraints should be considered upper limits . however , given that the [ oiii]@xmath635007 emission - line region is likely to be more extended than the central source , the [ oiii]@xmath635007 emission is unlikely to be much more magnified than the x - ray emission .
these results suggest that the agn in fsc 10214 + 4724 is powerful . however , the rest - frame 1.626.3 kev luminosity is approximately 12 orders of magnitude below the constraint estimated from the [ oiii]@xmath635007 luminosity .
this is significant since rest - frame @xmath72 kev emission is not easily attenuated [ e.g. , one order of magnitude of extinction at @xmath72 kev requires compton - thick obscuration ( @xmath73 @xmath36 ) ; see appendix b in deluit & courvoisier 2003 ] . hence
, if fsc 10214 + 4724 hosts a quasar , as previously suggested , then it must be obscured by compton - thick material ; these general conclusions are consistent with those found for other _ iras _ galaxies of similar luminosity ( e.g. , iwasawa 2001 ; wilman 2003 ) . under this assumption ,
the observed x - ray emission from the agn would be due to reflection and scattering and a strong fe k@xmath31 emission line should be detected . with only @xmath0 14 x - ray counts , the current x - ray observations can not provide good constraints on the presence of fe k@xmath31 ; however , a scheduled @xmath0 50 ks observation ( pi : k. iwasawa ) may be able to place some constraints .
the lensing - corrected properties of fsc 10214 + 4724 are similar to those of galaxies ( see 1 ) , sources that are probably the progenitors of massive galaxies ( and hence massive black holes ) in the local universe .
the steep x - ray spectral slope and moderately luminous x - ray emission of fsc 10214 + 4724 contrasts with the x - ray properties of the five x - ray detected galaxies classified as agns in alexander ( 2003b ) ; however , the x - ray properties of fsc 10214 + 4724 are similar to those of the two x - ray detected galaxies classified as unknown ( see figure 3 ) . since we know that an agn is present in fsc 10214 + 4724 ( possibly powerful and compton thick ) , this suggests that further unidentified agns may be present in the galaxy population and the @xmath0 40% agn fraction ( alexander 2003b , 2004 ) should be considered a lower limit .
many compton - thick agns may be present in ultra - deep x - ray surveys ( e.g. , fabian 2002 ) .
however , the presence of vigorous star formation may make them difficult to identify on the basis of their x - ray properties alone .
we acknowledge support from the royal society ( dma ) , pparc ( feb ; cs ) , cxc grant go4 - 5105x ( wnb ) , nsf career award ast-9983783 ( wnb ) , and miur ( cofin grant 03 - 02 - 23 ; cv ) .
we thank p. eisenhardt , d. hogg , k. iwasawa , and n. trentham for useful discussions . | we present a @xmath0 20 ks _ chandra _ acis - s observation of the strongly lensed ultra - luminous infrared galaxy fsc 10214 + 4724 .
although this observation achieves the equivalent sensitivity of an up - to @xmath0 4 ms _ chandra _ exposure ( when corrected for gravitational lensing ) , the rest - frame 1.626.3 kev emission from fsc 10214 + 4724 is weak ( @xmath1 2@xmath2 erg s@xmath3 for a lensing boost of @xmath4 ) ; a significant fraction of this x - ray emission appears to be due to vigorous star - formation activity . if fsc 10214 + 4724 hosts a quasar , as previously suggested , then it must be obscured by compton - thick material .
we compare fsc 10214 + 4724 to high - redshift galaxies and discuss the x - ray identification of compton - thick agns at high redshift .
[ firstpage ] x - rays : individual : fsc 10214 + 4724 galaxies : active gravitational lensing |
rarefied gas flows have recently attracted significant research interest due to the rapid development of micro / nano - fluidic technologies . gaseous transport in micro / nano devices is often found to be non - equilibrium , and non - equilibrium phenomena have not yet been well understood @xcite . the conventional theory to describe gas flows is the navier stokes equations , which assume that the fluid is in a quasi - equilibrium state .
however , for non - equilibrium flows , the navier stokes equations break down because that the molecular nature of the gas strongly affects the bulk flow behavior i.e. the gas can no longer be regarded as a fluid continuum . whether gas flows are in local equilibrium or not can be classified by the non - dimensional knudsen number , kn , defined as the ratio of mean free path and the device characteristic length scale .
the navier stokes equations with no - velocity - slip wall boundary condition are only appropriate when @xmath0 . however , gas flows in micro / nano - fluidic devices are often in the slip flow regime ( @xmath1 ) or the transition flow regime ( @xmath2 ) . in these regimes ,
the gas flow can not properly be described as a continuous flow , nor as a free molecular flow . in practice , most devices operate with a range of knudsen numbers in different parts of the device ; this makes it even more difficult to develop a generalized flow model .
direct simulation monte carlo ( dsmc ) methods and direct numerical simulation of the boltzmann equation can provide accurate solutions for rarefied gas flows . however , these are computationally intractable for 3d flow systems , and impractical with the current computer technology , especially for the low speed gas flows usually encountered in micro / nano - systems .
statistically , one needs to take significantly more samples of the flow field at any point for the dsmc method to resolve flows with low mach numbers .
the direct simulations based on the boltzmann equation requires significant computational resources for integrating the velocity space ranging from @xmath3 to @xmath4 .
in addition , it is usually difficult to solve the full boltzmann equation directly via either numerical or analytical methods .
meanwhile , the continuum methods beyond the navier stokes level have failed to produce satisfactory results for gas flows in the transition flow regime @xcite .
it is well - known that continuum expressions for the viscous stress and heat flux in gases may be derived from the fundamental boltzmann equation via either a kn - series solution ( known as the chapman - enskog approach ) or by an expansion of the distribution function as a series of hermite tensor polynomials @xcite .
to first order ( i.e. for near - equilibrium flows ) both approaches yield the navier stokes equations .
however , the solution methods can be continued to second and higher orders , incorporating more and more of the salient characteristics of a non - equilibrium flow .
the classical second - order stress and heat flux expressions are the burnett equations ( from the chapman - enskog approach ) , and the grad 13-moment equations ( from the hermite polynomial method ) @xcite .
these can be seen as corrections to the navier stokes constitutive relations to make them more appropriate to continuum - transition flows .
however , different physical interpretations of the solution methods at second and higher orders have recently led to a variety of sets of equations , including the bhatnagar - gross - krook ( bgk)-burnett@xcite , eu @xcite , augmented burnett @xcite , and regularized moment ( r13)@xcite equations . while each purports to be the proper high - order correction to the stress and heat flux ( there is no disagreement about the form of the navier stokes equations at first - order ) , none of these models are satisfactory @xcite .
in addition , these models suffer unknown additional boundary conditions at solid walls to appropriately reflect gas molecule / wall surface interactions .
the lattice boltzmann ( lb ) approach offers an alternative method for rarefied gas flow simulations . historically , the lb model was evolved from the lattice - gas automata ( lga ) for mimicking the navier stokes hydrodynamics ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
over the past two decades , the lb method has been developed to provide accurate and efficient solutions for continuum flow simulations as the validity of the model can be ensured by the chapman - enskog expansion . due to its kinetic nature , the lb model has distinct advantages over the continuum computational methods , including easy implementation of multi - physical mechanisms and the boundary conditions for fluid / wall interactions .
the potential of lb models for simulating rarefied gas flows have been demonstrated ( e.g. * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
recently , the lb models were shown to be able to be derived systematically from the boltzmann - bgk equation based on the hermite expansion ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this creates another theoretical foundation different from the chapman - enskog expansion , so that higher - order lb approximations to the boltzmann - bgk equation beyond the navier stokes level can be constructed by using the high - order hermite expansion with appropriate quadratures .
this indicates that high - order lb models have the potential to capture non - equilibrium effects in rarefied flows .
in addition to the systemic framework of constructing lb models , @xcite also established the link between the orders of hermite polynomials and chapman - enskog expansion .
the authors concluded that the order of hermite expansion is responsible for obtaining correct velocity moments .
the precise relation among the orders of hermite expansion , chapman - enskog expansion and velocity moments was described by eq.(4.7 ) in @xcite .
for instance , the third - order expansion is required for accurate pressure tensor and momentum at the navier stokes level .
these conclusions are key to constructing appropriate lb models for non - equilibrium gas flows .
however , the numerical simulations do not support these conclusions .
in contrast , the simulation data showed that the higher order terms in the equilibrium distribution function have negligible influence for low speed rarefied flows .
this indicates that the hermite expansion order is not related to the order of chapman - enksog expansion , in contrary to the theoretical conclusions drawn by @xcite . in this work
, we aim to answer this question whether the hermite expansion order is important for the lb method , as it is for the grad s moment method , to capture non - equilibrium effects in rarefied flows , especially at micro / nano - scales .
furthermore , we will analyze theoretically and numerically the mechanisms that are important in constructing high - order lb models for rarefied gas dynamics .
we will discuss the differences between the approaches of @xcite and grad s moment method . to help us to understand the modeling capability of the lb method for rarefied gas dynamics
, we will also analyze the similarities and differences between the lb method and the discrete velocity method ( dvm ) of solving the bgk equation .
in particular , we will prove that the hermite expansion order is not important for the flows that the linearized bgk equation can accurately describe .
since the important nonlinear constitutive relations in the knudsen layer are still not captured satisfactorily @xcite , our numerical analysis will be based on a standing - shear - wave problem specifically designed by @xcite to exclude the effect of gas molecule / wall interactions , so we can concentrate on the model capabilities .
although the lb models were originally developed from lga , the link to the kinetic theory has late been established by @xcite .
consequently , the lb approach may be considered as a special finite difference scheme of solving the boltzmann - bgk equation @xcite .
this theoretical link indicates that the lb methodology may provide a reasonable approximation to the boltzmann - bgk equation .
the central question is how accurate the lb models can capture non - equilibrium effects in rarefied gas dynamics . to answer this question
, we will revisit the derivation process of lb models from the boltzmann - bgk equation proposed by @xcite and we will emphasize on the model capability in describing rarefied gas flows .
the original boltzmann - bgk equation can be written as : @xmath5 where @xmath6 denotes the distribution function , @xmath7 the phase velocity , @xmath8 the pressure , @xmath9 the body force and @xmath10 the gas viscosity . using the well - known chapman - enskog expansion ,
the collision frequency can be represented by the ratio of pressure and gas viscosity , which is convenient to obtain the knudsen number definition consistent with that of hydrodynamic models . without losing generality ,
we define the following non - dimensional variables : @xmath11 where @xmath12 is the macroscopic velocity , @xmath13 the gas constant , @xmath14 the gas temperature , @xmath15 the reference temperature , @xmath16 the spatial position and @xmath17 the inverse of the characteristic length of the flow system . the symbol _ hat _ , which denotes a dimensionless value , will hereinafter be omitted .
we define the knudsen number using macroscopic properties as below : @xmath18 by using these non - dimensional variables , the non - dimensional form of the boltzmann - bgk equation becomes @xmath19 where the maxwell distribution in @xmath20-dimensional cartesian coordinates can be written as @xmath21.\ ] ] from the non - dimensional format of eq.([bgk ] ) , we can clearly see the relationship between the relaxation time and the mean free path ( i.e. knudsen number ) , which plays a key role in lb simulation of rarefied gas flows ( e.g. * ? ? ?
to discretize the velocity space , we project the distribution function onto a functional space spanned by the orthogonal hermite basis : @xmath22 where @xmath23 is the @xmath24th order hermite polynomial . the weight function @xmath25 is given by @xmath26 and the coefficients @xmath27 are @xmath28 the coefficient @xmath29 for the equilibrium distribution is @xmath30 where @xmath31 and @xmath32 , @xmath33 , are the weights and abscissae of a gauss - hermite quadrature of degree @xmath34 respectively .
herein , the distribution function is approximated by the first @xmath35 hermite polynomial . using the derivation relation ,
the body force term @xmath36 can be approximated as @xmath37 as an example , the second order approximation of the equilibrium distribution and the body force are : @xmath38 \right\},\ ] ] @xmath39 where @xmath14 should be set to unity for isothermal problems and @xmath40 is constant for incompressible problems .
an appropriate gauss - hermite quadrature , see the appendix in @xcite for a list of quadratures , can be chosen to evaluate the integral to obtain @xmath41 .
consequently , eq.([bgk ] ) can be discretized as @xmath42 where @xmath43 , @xmath44 and @xmath45 .
we have obtained the lattice boltzmann equation , i.e. eq.([lbgk ] ) , by discretizing eq.([bgk ] ) in the velocity space . an appropriate numerical scheme is now required to solve eq.([lbgk ] ) .
if a finite difference scheme is chosen , one can obtain the so - called finite difference lattice boltzmann model .
in particular , when the first - order upwind finite - difference scheme is chosen , one can obtain the standard form of lb model : @xmath46 where the relationship between the relaxation time @xmath47 and the knudsen number is established naturally i.e. @xmath48 . for continuum
flows where the navier stokes equations are valid , the above first - order scheme can become effectively second - order accurate in both space and time by simply replacing the non - dimensional relaxation time @xmath47 with @xmath49 ( see * ? ? ? * ; * ? ? ?
in doing so , the second order discretization error can be absorbed into an artificial viscosity .
therefore , this simple but accurate scheme has been widely used to simulate flows at the navier stokes level . since any lb model intended to simulate rarefied gas dynamics beyond the navier stokes level needs to recover the navier stokes equation at small knudsen number , i.e. @xmath50
, this first - order scheme with correction has been commonly used in lb simulation of rarefied gas flows ( e.g. * ? ? ?
* ; * ? ? ?
* ) for rarefied gas problems
. however , the artificial viscosity has only corrected the momentum transfer to the second order , which is only appropriate for the navier - stokes hydrodynamics .
this correction will lead to inconsistency for the transfer of the other higher - order moments , which are essential for capturing non - equilibrium effects in rarefied gas flows .
therefore , the dilemma is that we need correction on the relaxation time to recover the navier stokes hydrodynamics appropriately when the knudsen number is close to zero where the high - order moments are not important .
meanwhile , we should not have this correction for the higher - order moments which are more important to rarefied flows . in deed
, the simulation will diverge when the knudsen number is approaching to 0.5 if no relaxation time correction is introduced .
the reason is that it goes beyond the stability regime of the relaxation scheme .
@xcite suggested to use the correction when the knudsen number is less than 0.5 and switch to no correction when the knudsen number is larger than 0.5 .
however , it will lead to inconsistency at the knudsen number around 0.5 which is the most important flow regime in micro / nano - fluidic devices .
the above first - order upwind scheme should not be used for simulating the gas flows with finite knudsen numbers . to resolve this problem inherited from the standard lb method
, we should not use the artificial viscosity to achieve correct physics at the navier stokes level .
we propose to discretize eq.([lbgk ] ) using a numerical scheme with second - order accuracy , which was first used by @xcite for thermal flow simulation at the navier stokes level : @xmath51 \nonumber \\ & & -\frac{\delta t}{2kn}\left[f_\alpha(\bm r , t)-f_\alpha^{eq}(\bm r , t)\right ] \nonumber \\ & & + \frac{\delta t}{2 } \left[g_\alpha(\bm r+\bm \xi_\alpha \delta t , t+\delta
t)+ g_\alpha(\bm r , t)\right ] . \end{aligned}\ ] ] by introducing @xmath52 the above implicit scheme can be written as @xmath53 + \frac{kn g_\alpha \delta_t}{kn+0.5\delta_t},\ ] ] with @xmath54 @xmath55 therefore , the viscosity is now @xmath56 rather than @xmath57 .
most importantly , the same relation between the relaxation time and the mean free path can be used for the transfer of any order moments .
although the construction of lb models based on the hermite polynomials is straightforward , the hermite polynomials higher than the third order give irrational roots . the integer stream velocity is an essential feature of lb models , i.e. the simple and efficient `` stream - collision '' mechanism .
so high - order lb models , which have non - integer discrete velocities , will need additional effort , such as point - wise interpolation @xcite .
therefore , they essentially become off - lattice discrete velocity method for solving the kinetic boltzmann equation , which will increase the computational cost dramatically and introduce extra numerical error .
@xcite suggested a method for searching abscissae on the grid points of cartesian coordinates to construct high - order lb models with integer discrete velocities .
the examples are d2q17 and d2q21 models given by @xcite and @xcite ( note , we follow the conventional terminology for the lb models as first introduced by @xcite dubbed as dnqm model i.e. n dimensional model with m discrete velocities ) . furthermore , @xcite proposed an alternative method to seek rational - number approximation to the rations of the hermite roots based on the relation between the entropy and the roots of hermite polynomials .
they also proposed the higher - order lb models with integer discrete velocity , such as d2q16 and d2q25 models .
the above high - order lb models with integer stream velocities will be numerically examined in this work and the details are listed in table.([model ] ) .
.the quadratures of five lb models where k is the number of discrete velocities with the same velocity magnitude , the subscript @xmath58 denotes a fully symmetric set of points , and @xmath31 are the weights .
the quadrature accuracy is fifth - order for the d2q9 model , seventh - order for the d2q16 , d2q17 and d2q21 models , and ninth - order for the d2q25 model .
the details of d2q17 and d2q21 models can be found in @xcite while the d2q16 and d2q25 models are discussed in @xcite . [
cols="<,<,<,<,<",options="header " , ] based on the above model construction procedure , the accuracy of lb models depends on three level of approximations .
firstly , it depends on the accuracy of the numerical scheme for solving eq.([lbgk ] ) .
as we have demonstrated , the commonly used first - order upwind scheme will lead to incorrect physics for rarefied flow**s**. our second - order numerical scheme given by eq .
( [ 2order ] ) is essential to capture non - equilibrium effects accurately .
secondly , the order of the hermite expansion was considered to be important to obtain the correct moments @xcite .
thirdly , the gauss - hermite quadrature accuracy should be sufficiently high so that the integration of eq.([an ] ) can be evaluated accurately .
therefore , the term _ higher - order _ lb models here refer to the lb models with high - order of hermite expansion and gauss - hermite quadrature in comparison with the standard lb model .
similar to the grad s method for deriving higher order continuum systems ( e.g , grad 13-moment equations ) , using the hermite expansion to approximate the boltzmann - bgk equation can lead to the lb equation , i.e. eq.([lbgk ] ) .
however , the major difference is that lb models are always staying at the kinetic level , i.e. solving the kinetic equation - eq.([lbgk ] ) , while the grad s method will produce a set of continuum equations .
the basic idea of grad s method is to use the truncated hermite polynomials to approximate the full boltzmann ( or boltzmann - bgk ) equation . due to the unique feature of hermite polynomial
, the moments of up to the chosen truncation order can be described accurately by the derived macroscopic moments systems . in contrary , the only explicit effect of the truncation on the lb models is on the approximation of the equilibrium distribution function and the body force , while the grad s moment equations do not approximate the equilibrium distribution function .
although the order of hermite expansion determines the accuracy level of the moment model , which is not the same for the lb models . essentially , the lb equation i.e. eq.([lbgk ] ) is similar to any model equation which is to simplify the full boltzmann equation .
the kinetic process , i.e. gas molecules relaxing to the equilibrium state through collisions , is still the same .
therefore , the lb method is very close to the discrete velocity method solving the boltzmann - bgk equation ( especially the linearized - bgk equation ) , which we will discuss in the section below .
the above procedure of establishing lb models is similar to the problem solving process of the discrete velocity method , which directly solves the boltzmann - bgk equation . since dvm has been proved to be able to provide accurate results for rarefied gas dynamics , it is helpful to compare two numerical methods in depth .
the discrete velocity method is to discretize the velocity space based on quadratures e.g. gauss - hermite quadrature and newton - cotes quadrature ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the first step is to non - dimensionalize the boltzmann - bgk equation and obtain the reduced functions , e.g. @xmath59 and @xmath60 in eqs.([ga ] ) and ( [ gb ] ) , which are important to reduce computational costs . the second step is to apply an appropriate discretization method for the velocity space , which is important but difficult because the velocity space ranges from @xmath3 to @xmath4 and the properties of conservation and dissipation of the entropy should be kept .
a typical choice is the gauss - hermite quadrature , which is to be adopted in our simulations . in order to reduce the velocity components which need to be integrated from @xmath3 to @xmath4
, curvilinear coordinates including the polar coordinates for 2d systems may be used for the velocity space .
afterwards , the continuous maxwell equilibrium should also be discretized .
the last step is to adopt an appropriate numerical scheme for the space and time discretization .
therefore , we can see that lb methodology closely resembles the dvm problem solving process .
@xcite noticed this similarity and stated * * * * the lb equation is essentially dvm with finite discrete velocities and fully discretized space and time tied to the discrete velocity set**"**. for simulating rarefied gas flows , this similarity is important as we have shown how the lb framework is developed from the boltzmann - bgk equation . for both dvm and lb methods ,
the most critical task is to discretize the velocity space .
when the gauss - hermite quadrature is used in dvm , the discretization of the velocity space in these two methods are the same , which may indicate that the lb models with sufficiently accurate gauss - hermite quadrature can capture the higher - order non - equilibrium effects in the rarefied gas flows .
this in deed is confirmed by the simulation results presented in fig.[exm ] , which we will discuss in detail in section 4 . however , an important advantage of the lb models is the `` stream - collision '' mechanism which is mainly inherited from the lattice gas automata .
this `` stream - collision '' mechanism makes the lb method easy to understand and simple for computer programming .
therefore , the `` stream - collision '' mechanism is an important feature of the lb models which distinguishes them from dvm .
the coupled time step and physical space in the lb models will dramatically reduce the computational cost .
in addition , dvm relies heavily on mathematical techniques which depend on specific problem , while the lb methodology is straightforward and more suitable for developing a generic simulation package for engineering design . by introducing @xmath61 to denote the unknown perturbed distribution function and assuming the flow is weakly non - equilibrium , @xmath6 can be approximated by @xmath62 where @xmath63 which is the global ( absolute ) equilibrium distribution function . using the taylor series to expand the local equilibrium distribution function and keeping the terms up to the first order
, one can obtain the following equation @xmath64 \right\},\ ] ] where we assume the flow is incompressible . using eq.([lsum ] ) , we can obtain the linearized bgk equation : @xmath65= -\frac{1}{kn}\left\{\psi-\left[\bm \xi \cdot \bm u + \frac{1}{2}(t-1)(\xi^2-d)\right ] \right\}.\ ] ] for lattice boltzmann models , one can rewrite eq.([approxf ] ) as @xmath66,\ ] ] where @xmath67 substituting eq.([approxfhermite ] ) into the boltzmann - bgk equation and keeping the first- and second - order expansions of the equilibrium distribution , one can obtain @xmath68= -\frac{1}{kn}\left ( \varphi-\bm \xi \cdot \bm u \right),\ ] ] and @xmath69= -\frac{1}{kn}\left \{\varphi-\bm \xi \cdot \bm u - \frac{1}{2}\left[(\bm \xi \cdot \bm u)^2-u^2+(t-1)(\xi^2-d)\right ] \right\}.\ ] ] because @xmath25 is equal to @xmath70 , we can observe the following interesting facts by comparing eqs.([lbm1 ] ) and ( [ lbm2 ] ) with eq.([generallbgk ] ) .
first of all , by keeping the first order hermite expansion , the essential lb model equation is the same with that of the isothermal ( @xmath71 ) linearized bgk equation except the body force term .
_ this implies that @xmath72 is indeed equivalent to @xmath61 though @xmath72 is prescribed to include only the finite order terms of the hermite polynomials ( cf .
eq.([approxf ] ) and eq.([approxfhermite ] ) ) .
therefore , the lb equation with the first order terms should be as good as the linearized bgk equation for isothermal flows_. this indicates that high - order hermite expansion is not necessary for rarefied gas flows .
secondly , with the second order hermite expansion , there is an extra velocity term @xmath73 $ ] for the lb equation in comparison to the linearized bgk equation .
however , for flows with low mach number , this term is a higher - order small quantity which can be ignored .
this is the reason why the hermite expansion order is reported to make negligible difference on the simulation results ( see * ? ? ?
in fact , the first order expansion is sufficient to obtain the accurate results for isothermal rarefied flows with low speed .
furthermore , the lb equation with the second order expansion can in principle describe thermal problems since the temperature information is included in eq.([lbm2 ] ) , which at least has the same capability as the linearized bgk equation , though the bgk kinetic model gives wrong @xmath74 number .
thirdly , the treatment of the body force makes the difference between the lb model and the linearized bgk equation .
it is because that the linearized bgk model keeps the full information while the lb model uses the hermite expansion to approximate @xmath75 , i.e. @xmath76 .
however , for the problem is not far from equilibrium state , this difference is not important , which will be confirmed by the numerical simulations in section 4 . from the above analysis , we can see that the hermite expansion order does not determine the accuracy of lb models for rarefied gas flows as described by eq.(4.7 ) in .
the hermite expansion provides a means to approximate the equilibrium distribution and the body force in the kinetic equation .
therefore , the lb equation , similar to the linearized bgk equation , is an approximation of the boltzmann - bgk equation .
in contrast to the grad s moment method , lb models include the information of any order moment though it may not be accurate .
for instance , with the first order expansion , the lb model equation is as the same as the isothermal linearized bgk equation in the incompressible limit , which will give accurate results for any order velocity moment . when the mach number of flow increases , high - order terms in the hermite expansion become important .
therefore , the order of hermite expansion is important to simulate compressible flows rather than rarefied flows . to capture non - equilibrium effects in rarefied flows ,
the gauss - hermite quadrature is the key as it determines the discretization accuracy to the model equation . with sufficiently high order of the gauss - hermite quadrature
, lb models can give excellent numerical results , e.g. the results presented in fig.([exm ] ) where 400 discrete velocities are used are identical to the dvm solution . considering the similarity of the lb equation and linearized bgk equation , insufficient quadrature order should be responsible for the failure on capturing the constitutive relations in the knudsen layer because the kinetic boundary condition have been well accepted in solving the linearized bgk equation . in summary ,
the lb method is essentially a special discrete velocity model , which approximates the boltzmann - bgk equation with finite discrete velocities and fully discretized space and time tied to the discrete velocity set .
the capability of lb equation is similar to the linearized bgk equation for simulating rarefied gas flows .
the hermite expansion order determines the model equation and is important for compressible flows .
it has no direct effect on the accuracy of capturing high - order non - equilibrium effects .
meanwhile , the gauss - hermite quadrature as a discretization technique for the velocity space directly determines whether the lb models can describe rarefied flows accurately .
in addition to the above theoretical analysis , we will numerically evaluate the lb models . to exclude the boundary condition effect , we choose the standing - shear - wave problem as the benchmark case , which was specially designed for assessing the accuracy of various models @xcite .
it is a shear flow driven by a temporally and spatially oscillating body force , which can be written as the following form : @xmath77 where @xmath78 is the body force in the direction @xmath79 ( which is perpendicular to the @xmath80 direction ) , @xmath81 is the amplitude , and @xmath17 is the wave number and @xmath82 is the frequency .
this isothermal problem is sufficiently simple because the flow direction is perpendicular to the space variation but it is intended to capture the shear - dominated characteristic of microscale flows .
furthermore , the distinct advantage is that the boundary is not important here so that one can focus on the model itself without the interference from gas molecule / wall interactions . with eq.([nondime ] ) , the body force becomes : @xmath83 where @xmath17 is considered as a measure of the characteristic length , and @xmath84 .
another distinctive advantage for using this benchmark problem is that analytical solutions can be obtained for many hydrodynamic models , such as the navier stokes equation and the regularized 13-moment model ( r13 ) . for convenience ,
the r13 solution is listed as below : @xmath85 where @xmath86 denotes the velocity amplitude .
one can refer to @xcite for the detail of hydrodynamics models .
the discrete velocity method of solving the linearized bgk equation has already been served as a benchmark for the standing shear wave problem by @xcite , where the linearized bgk model eq.([generallbgk ] ) can be simply written in the scalar form for isothermal flows : @xmath87= \frac{1}{kn}\left(\xi_x u_x - \psi \right).\ ] ] since the problem is essentially one - dimensional , one can eliminate @xmath88 by multiplying the above equation with @xmath89 and @xmath90 respectively .
integrating over @xmath88 , the resulting equations are : @xmath91 @xmath92 where the reduced unknown functions @xmath59 and @xmath60 are defined as @xmath93 @xmath94 the macroscopic velocity can be expressed as @xmath95 to solve eqs.([ga ] ) and ( [ gb ] ) , the essential task is to choose an appropriate quadrature to discretize the velocity space which ranges from @xmath3 to @xmath96 .
the typical highly accurate choice for low speed rarefied gas flows is the gauss - hermite quadrature , which is used here .
based on the discretization of the phase space , the integration operation over the velocity space is converted to sum operation and then a series of equations like eq.([lbgk ] ) are obtained .
naturally , the discretized maxwell equilibrium distribution can also be obtained by directly using its value on the grid of the velocity space .
one can then use typical numerical methods such as finite difference scheme ( e.g. , the lax - wendroff scheme ) to solve these equations respectively . when the same gauss - hermite quadrature with 400 discrete velocities are used in the dvm solution of the linearized bgk equation and our lb model , fig.[exm ]
shows that the results for both velocity and shear pressure amplitude are nearly identical for a broad range of knudsen numbers from 0.1 to 1.5 .
even with the first order hermite expansion , the lb model can predict shear pressure accurately , which confirms that the hermite expansion order does not directly affect accuracy of the lb models in capturing non - equilibrium effects measured by the knudsen number .
although the standard lb model ( d2q9 ) is not sufficiently accurate in comparison with the dvm solution , high - order lb model ( d2q16 ) with minimal increase of the discrete velocity set can produce good results .
fig.[exm ] shows that the lb model with increasing order of the gauss - hermite quadrature can closely approximate the linearized bgk equation .
therefore , in comparison with the dvm simulation , lb method can provide a practical engineering design simulation tool which can produce reasonably accurate results with significantly reduced computational cost .
as discussed in section 2.3 , at least three factors will influence the problem - solving process , i.e. the numerical scheme for solving eq.([lbgk ] ) , the order of hermite expansion and gauss - hermite quadrature . for the numerical scheme ,
our second - order scheme is essential as discussed in section 2.2 . regarding the role of hermite expansion and gauss - hermite quadrature
, we have theoretically proved that the gauss - hermite quadrature rather than the order of hermite expansion is key to capturing non - equilibrium effects accurately .
the numerical simulations have also performed to testify our theoretical analysis . in figs.[normal ] and [ lag ] , the simulation results of the three lb models are compared with the solutions of directly solving the linearized bgk equation and the navier stokes equation .
the expansion of the equilibrium distribution function and the forcing term is second - order for the d2q9 model , third - order for the d2q16 and forth - order for d2q25 .
the results in fig.[normal ] show that the prediction for velocity amplitude of the d2q25 model are in excellent agreement with the dvm solution of the linearized bgk equation across a broad range of knudsen number ( @xmath97 $ ] ) for the quasi - steady and time - varying problems with @xmath17 up to @xmath98 .
meanwhile , the results of the d2q9 model deviate from the dvm solution of the linearized bgk equation significantly .
surprisingly , the d2q9 model does not agree with the results predicted by the navier stokes equation .
fig.[lag ] shows the velocity wave phase lag , which suggests that high - order lb models perform better in the transition flow regime .
although figs.[normal ] and [ lag ] demonstrate that increasing order of lb model in terms of the hermite expansion and gauss - hermite quadrature will lead to more accurate results , we still do not know the exact role the orders of the hermite expansion and the gauss - hermite quadrature play .
therefore , we single out the effect of the hermite expansion in fig.[diff ] , where the results of the lb models with the same quadrature but different hermite expansion order are compared . the results clearly show that the hermite expansion order for both the force and the equilibrium distribution function make negligible difference to the simulation results .
even the first order expansion is sufficient to obtain accurate velocity for the d2q25 model .
the simulation results support our theoretical analysis that the hermite expansion has no direct influence on model accuracy for capturing non - equilibrium effects . specifically ,
the lb model equation determined by the first order hermite expansion is sufficient for a typical gas flow in micro - devices where the mach number is usually small .
in contrast , the gauss - hermite quadrature determines the model accuracy as the higher - order quadratures give better results .
not only the order of quadrature but also the abscissae may influence the model accuracy . therefore , the simulation results of the three lb models with the same order quadrature but different abscissae are compared in fig.[three ] . although increasing quadrature order will lead to improved accuracy , more discrete velocities may not improve the model performance if the quadratures are the same order .
for example , the quadratures of the d2q16 , d2q17 and d2q21 models are the same order .
surprisingly , the d2q16 model produces the results better than the other two models with more discrete velocities .
the reason may be attributed to that the abscissae of the d2q16 model has better symmetry .
in addition , all these models are better than the d2q9 model which has low order quadrature .
therefore , appropriate abscissae may improve the model accuracy and reduce the computational costs with smaller number of discrete velocities . since @xcite has shown that the r13 equation gives the best performance among the extended hydrodynamic models , we compare the lb models with the r13 model here
fig.[r13e ] shows that , in comparison with the data obtained from directly solving the linearized bgk equation , the high - order lb models including the d2q16 and d2q25 models can give better results than the r13 equation over a broad range of knudsen numbers .
therefore , the high - order lb models with modest number of discrete velocity set , such as the d2q16 and d2q25 models , can offer close approximation to the linearized bgk equation .
most importantly , these high - order lb models achieve such degree of accuracy at a fraction of computational costs associated with directly solving the linearized bgk equation .
we have theoretically and numerically analyzed the high - order lb models for rarefied gas flows .
the lattice boltzmann equation is shown to be equivalent to the linearized bgk equation in the incompressible limit . when the same gauss - hermite quadrature is used , both lb and dvm simulations produce results in excellent agreement across a broad range of the knudsen numbers .
this suggests the importance of the gauss - hermite quadrature and the great potential of the lb method for modeling rarefied gas flows .
while the gauss - hermite quadrature is of the most importance to capturing non - equilibrium effects , the first - order hermite expansion on the equilibrium distribution function is sufficient to obtain the correct moments for isothermal flows e.g. increasing the hermite expansion order further will not improve the model accuracy .
for the same order gauss - hermite quadratures , the chosen abscissae will influence the model accuracy and more discrete velocities may not lead to improved model accuracy .
overall , we have demonstrated that lb method offers a computationally efficient approach to solve the bgk equation .
we can choose a suitable lb model to meet different requirement on model accuracy and computational efficiency , which offers an ideal flexible engineering design simulation tool to be able to simulate flows in the continuum and transition regimes .
the authors would like to thank jason reese , xiaojun gu and guihua tang for many informative discussions .
this work was financially supported by the engineering and physical sciences research council u.k . under grants no .
ep / d07455x/1 and no .
ep/ f028865/1 .
1991 numerical analysis of steady flows of a gas condensing on or evaporating from its plane condensed phase on the basis of kinetic theory : effect of gas motion along the condensed phase .
fluids a _ * 3 * ( 9 ) , 22602275 . ) here .
the results show that the lb model with sufficiently large discrete velocity sets can be very accurate.,title="fig:",width=453,height=264 ] ) here .
the results show that the lb model with sufficiently large discrete velocity sets can be very accurate.,title="fig:",width=453,height=264 ] which is 2 , 3 , 4 for the d2q9 , d2q16 and d2q25 models respectively , and the order of gauss - hermite quadrature is @xmath99.,title="fig:",width=453,height=264 ] + which is 2 , 3 , 4 for the d2q9 , d2q16 and d2q25 models respectively , and the order of gauss - hermite quadrature is @xmath99.,title="fig:",width=453,height=264 ] + which is 2 , 3 , 4 for the d2q9 , d2q16 and d2q25 models respectively , and the order of gauss - hermite quadrature is @xmath99.,title="fig:",width=453,height=264 ] | in this work , we have theoretically analyzed and numerically evaluated the accuracy of high - order lattice boltzmann ( lb ) models for capturing non - equilibrium effects in rarefied gas flows . in the incompressible limit ,
the lb equation is proved to be equivalent to the linearized bhatnagar - gross - krook ( bgk ) equation .
therefore , when the same gauss - hermite quadrature is used , lb method closely assembles the discrete velocity method ( dvm ) . in addition
, the order of hermite expansion for the equilibrium distribution function is found not to be correlated with the approximation order in terms of the knudsen number to the bgk equation , which was previously suggested by @xcite .
furthermore , we have numerically evaluated the lb models for a standing - shear - wave problem , which is designed specifically for assessing model accuracy by excluding the influence of gas molecule / surface interactions at wall boundaries .
the numerical simulation results confirm that the high - order terms in the discrete equilibrium distribution function play a negligible role .
meanwhile , appropriate gauss - hermite quadrature has the most significant effect on whether lb models can describe the essential flow physics of rarefied gas accurately .
for the same order of the gauss - hermite quadrature , the exact abscissae will also modestly influence numerical accuracy .
using the same gauss - hermite quadrature , the numerical results of both lb and dvm methods are in excellent agreement for flows across a broad range of the knudsen numbers , which confirms that the lb simulation is similar to the dvm process .
therefore , lb method can offer flexible models suitable for simulating continuum flows at navier stokes level and rarefied gas flows at the linearized boltzmann equation level .
[ firstpage ] |
the monte carlo method for classical and quantum lattice models has been improved dramatically since the proposal of the metropolis method@xcite .
the introduction of the extended ensemble , being one of the ideas that enhanced the monte carlo method , is characterized by a random walker traveling in the energy space .
the most well - known methods among the ones based on the extended ensemble is the multicanonical method @xcite . in this method ,
one can obtain the density of states ( dos ) from the histogram of visiting frequency at each value of the energy . since the random walk in the extended ensemble methods
is typically biased in a complicated way , the scaling property of the the traveled distance of the walker deviates from that of the free random walk , i.e. , @xmath0 .
this means that the random - walk nature of the method does not necessarily guarantee a quick equilibration .
however , in many important application , it turns out that the random walker can still travel the whole range of the energy within a reasonable computational time , i.e. , a computational time bounded by some polynomial in the system size .
therefore , the method is quite useful for models that has a first - order phase transition and/or ones that have local minima in the free energy .
in addition , the direct calculation of the dos makes it possible to calculate the absolute value of the free energy of a given system with high precision .
another advantage of this type of approach is that thermal averages of various quantities can be obtained for the whole range of the temperature from only a single run of simulation . while the early versions of extended ensemble methods still suffered from a slow diffusion , even this problem was removed recently by the acceleration method proposed by wang and landau ( wl ) @xcite , in which the random walker is forced to move away whenever it lingers about a place for a long time . for many important models , classical or quantum
, it was found that the loop / cluster algorithm @xcite is so efficient that the simulation can be done with no ( or negligible ) difficulty of slowing down at any temperature .
however , since the algorithm is originally designed for a simulation at a fixed temperature , one can not obtain from a single run the whole temperature dependence of various quantities nor can one estimate the free energy .
therefore , it is natural to wonder if the loop / cluster algorithm can be used together with the extended ensemble methods .
an apparent obstacle is that the energy in the quantum case is not defined so that constructing a histogram on it is straightforward and meaningful .
janke and kappler proposed@xcite that in the cluster algorithm for classical spin systems one can use the number of graph elements ( such as bonds in the swendsen - wang algorithm ) instead of the energy for constructing the histogram , showing a way for combining the extended ensemble method and the cluster algorithm for classical spin systems .
furthermore , two of the present authors demonstrated@xcite the enhancement of the efficiency achieved with the use of the wl method , and also suggested that the resulting framework can be generalized to quantum systems .
later , the application to the quantum systems was done with a different formulation by troyer , wessel and alet @xcite , who also showed the efficiency of the resulting algorithm . in the present paper
, we propose that a further improvement can be achieved for quantum models by using the broad - histogram ( bh ) relation for estimating the dos . the idea has been partially described in our previous paper @xcite for the classical models . for the ising models ,
it was pointed out @xcite that a ratio of the densities of states at two neighboring values of the energy can be expressed in terms of the microcanonical averages of the numbers of distinct potential moves . in the application to the ising model ,
the number of potential moves is nothing but the number of spins such that flipping one of them causes a change in the energy by a given amount . since this is a macroscopic quantity , whereas the number of the visits at each value of the energy is not , we can estimate the dos more precisely by using the relation than we do so directly from the histogram . in the case that we discuss here , our dos is not simply the total number of states .
therefore we have to use the generalized form of the transition matrix monte carlo method @xcite . in what follows ,
we describe the method specialized for quantum systems .
the simulation consists of two stages . in the first stage
we perform short runs using the wl method for obtaining a working estimate of the dos . as explained below , each run in the first stage
is characterized by a parameter called the modification factor , which controls the adoptive change of the fictitious weight of the random walk .
a stronger control is imposed on earlier runs whereas a weaker control for the latter runs . because of this adoptively changing weight , the microcanonical ensemble obtained in each run of the first stage deviates from the correct one .
this leads to some systematic error .
however , we need unbiased estimates of microcanonical averages@xcite because we use them for obtaining the dos from the bh relation .
therefore , in the second stage we perform a single long run with the weight fixed to the one obtained in the first stage .
this is in contrast to the ordinary wl method where the outcome of the first stage is taken as the final estimate of the dos .
the organization of the whole simulation is schematically depicted in fig . [
fig : simulation ] .
we will show this two - stage procedure gives more precise results for quantum systems than the one - stage procedure of the wl method described in @xcite , within the same amount of the total cpu time .
a little more specifically , the whole simulation can be described as follows .
we construct a histogram of the expansion order or the number of active vertices , which we denote as @xmath1 , instead of the histogram of the energy . accordingly
, the fictitious weight @xmath2 is introduced as a function of @xmath1 , rather than the energy .
the monte carlo dynamics is defined so that the detailed balance holds for this weight .
we start the first run of the first stage by setting the weight @xmath2 to be constant , i.e. , @xmath3 for all @xmath1 . at the beginning of each run , the counter ( or the histogram ) , @xmath4 , of visits to the value @xmath1 is set to be 0 while @xmath2 is not reset except at the beginning of the first run . for the @xmath5-th run we use the modification factor @xmath6 . every time the random walker visit a state with @xmath7 , @xmath2 is replaced by @xmath8 in order to penalize the random walker staying at this value of @xmath1 . at the same time
, @xmath4 is increased by one .
the @xmath5-th run is terminated when the histogram becomes flat . to be specific ,
the flatness condition is @xmath9 for all @xmath1 .
since the first stage is only for obtaining a working estimate of the dos , we stop when @xmath10 whereas in the original wl method @xmath11 was used instead . at the end of the first stage
, @xmath12 is approximately proportional to the generalized dos discussed below . in the second stage
, we do the same as the first stage but with a longer run and with @xmath2 fixed to be the result of the first stage . in this way
, we can compute non - biased microcanonical averages of various quantities in the second stage . in both the first and the second stages , the update of the configuration
is done using the loop update if the update does not involve a change in @xmath1 .
other similar algorithms such as the directed loop algorithm @xcite can be used as well , and the generalization is straight - forward .
the choice , however , is only of the secondary importance for the applications to the models without the magnetic field as considered in the present paper . for the loop algorithm
, there are two formulations ; one is based on the continuous imaginary time and the other is the discrete imaginary time . although the difference in the efficiencies of two formulations is minor , in the present paper we have used the discrete time formulation following troyer , wessel and alet@xcite . in what follows , we reformulate their algorithm , use it for deriving the bh relation , and describe how the relation is used in the second stage of the simulation .
the systematic error due to the discretization in the discrete time formulation can be removed by starting from the high - temperature series expansion @xcite rather than the path - integral representation .
therefore , we start with expressing the partition function @xmath13 as a high - temperature series expansion truncated at the @xmath14-th order , @xmath15 where @xmath16 is inverse temperature @xmath17 , @xmath18 is boltzmann s constant , and @xmath19 is a variable that takes on @xmath20 or @xmath21 .
the high - temperature expansion coefficient , @xmath22 , is here regarded as the generalized dos for the series order @xmath1 . when the hamiltonian is expressed as a sum of @xmath23 local interaction terms as @xmath24 ,
the generalized dos is given by @xmath25 we can simplify the notation by defining a _
vertex _ , @xmath26 , as a pair of indices @xmath27 and @xmath28 , i.e. , @xmath29 .
several other symbols are also introduced for simplification of the notation : @xmath30 , @xmath31 , @xmath32 , @xmath33 , @xmath34 , and @xmath35 . as a result
, we can rewrite the generalized dos as @xmath36 where @xmath37 here , @xmath38 . by comparing these expressions to those in
@xcite it is natural to identify @xmath1 as the number of graph elements .
here we call a vertex @xmath26 for which @xmath39 an _ active _ vertex .
we then replace the factor @xmath40 in ( [ partition ] ) by an adjustable function @xmath2 , which makes the simulated partition function to be @xmath41 with the weight of the state @xmath42 being @xmath43 in the first stage , the function @xmath2 is adaptively modified so that the probability of having @xmath44 to be independent of @xmath1 .
in other words , in the first stage we try to make @xmath45 constant by adjusting @xmath2 .
this goal is achieved by the use of the wl method . as a result
, we obtain @xmath22 as the inverse of @xmath2 .
the algorithm used for updating @xmath42 can be viewed as the dual monte carlo @xcite .
namely , by regarding the variable @xmath46 as a graph element , analogous to the bond in the swendsen - wang algorithm , one can establish a correspondence between the present algorithm and the swendsen - wang algorithm .
one step in a dual monte carlo simulation consists of two phases ; the `` graph '' @xmath47 is updated in the first phase with `` spin configuration '' @xmath48 being fixed and the @xmath48 is updated in the second phase with fixed @xmath47 . while the first phase changes @xmath1 , the latter does not . in what follows ,
we describe updating of @xmath47 and show how to estimate @xmath22 based on estimates of macroscopic quantities .
the second phase , i.e. , how the update @xmath48 with fixed @xmath1 , is described in the next section . in order to relate @xmath22 to the microcanonical averages of some macroscopic quantities , we consider an arbitrary function @xmath49 that satisfies @xmath50 while @xmath49 could be the transition probability from @xmath42 to @xmath51 , in which case the above equation is nothing but the detailed balance condition
, it does not have to be so for the derivation of the bh relation .
we sum up each side of the above equation over such @xmath42 and @xmath51 that @xmath7 and @xmath52 , yielding @xmath53 where @xmath54 and @xmath55 is the microcanonical average , i.e. , @xmath56 for the derivation of the bh relation , we only have to consider the case where @xmath51 and @xmath42 have the same spin configuration @xmath48 in common .
then , the procedure for generating @xmath57 from @xmath58 is as follows : 1 . : : generate an integer @xmath27 ( @xmath59 ) uniform - randomly .
2 . : : if @xmath60 , generate an integer @xmath28 ( @xmath61 ) uniform - randomly , and change @xmath47 into @xmath62 ( i.e. , activate @xmath26 ) with the probability @xmath63 where @xmath64 .
2. : : if @xmath65 , change @xmath47 into @xmath66 ( i.e. , inactivate @xmath26 ) with the probability @xmath67 where @xmath26 is the vertex @xmath68 such that @xmath69 .
3 . : : repeat 1 and 2 ( or 2 ) @xmath70 times . the overall transition probability corresponding to one repetition of 1 and 2 ( or 2 ) can be written in the form of the transition matrix as @xmath71 where some kronecker s delta s have been written as @xmath72 instead of @xmath73 for clarity . in order for this transition probability to satisfy ( [ eq : detailedbalance ] ) with ( [ eq : w ] ) and ( [ eq : u ] )
, @xmath74 must satisfy @xmath75 where @xmath76 and @xmath26 is the vertex at which @xmath42 and @xmath51 differ from each other .
the simplest choice is @xmath77 note here that we do not have to make @xmath74 smaller than 1 if we only need to make @xmath49 satisfy ( [ eq : detailedbalance ] ) and if @xmath78 does not have to be a real transition matrix . since this is the case for obtaining the bh relation , we shall forget about the condition @xmath79 for a time being . with ( [ eq : activationprobability ] ) , @xmath80 and @xmath81 become @xmath82 and @xmath83 where @xmath84 is the number of kinks , i.e. , the vertices at which @xmath85 .
thus we obtain from ( [ eq : generalbhrelation ] ) for @xmath86 , @xmath87 we call this relation the bh relation for quantum systems . when multiplied @xmath88 for @xmath89 , this yields @xmath90 where @xmath91 has been used .
note that the validity of ( [ eq : bhrelation ] ) and ( [ eq : theestimator ] ) does not depend on what transition probability we use in the simulation although we have used a particular form of it in deriving the relation ( [ eq : bhrelation ] ) .
therefore , we can use any algorithm to estimate the dos as long as it produces the correct microcanonical ensemble .
in fact , since we have not guaranteed that @xmath92 above in ( [ eq : activationprobability ] ) , @xmath49 mentioned above may not serve as the transition probability . to obtain the real transition probability , we simply replace ( [ eq : activationprobability ] ) by @xmath93 with @xmath94
this is the transition probability that is actually used in the simulation presented below .
we now turn to the second phase in which the state @xmath95 is updated to @xmath96 with @xmath47 fixed .
this can be done by either of the loop algorithm or the worm - type algorithm .
we here adopt the loop type update @xcite .
the weight with which @xmath48 must be sampled is @xmath97 . following the general framework of the loop algorithm we decompose the weight at every active vertex @xmath26 in @xmath47 as @xmath98 where @xmath99 is a graph and @xmath100 is the function that takes on @xmath20 or @xmath21 depending on the matching between @xmath101 and @xmath99 .
the readers are referred to ref.@xcite for the details of the graphs and the @xmath102-functions .
here we only show an example for the compactness of the description . in the case of the @xmath103 antiferromagnetic heisenberg model ,
the decomposition consists of a single term since the local hamiltonian can be expressed as a single delta function . @xmath104 where @xmath105 is a _ horizontal graph _ that consists of two horizontal lines connecting @xmath106 to @xmath107 and @xmath108 to @xmath109 . the function @xmath110 is defined for @xmath111 as @xmath112 here we have assumed that the state @xmath113 is the simultaneous eigenstate of all the @xmath114-components of spins .
namely , @xmath115 @xmath116 denotes the eigenvalue of @xmath117 of the eigenstate @xmath113 .
therefore , the update is done by placing horizontal graphs on all the vertices in @xmath47 , and connect the end points of horizontal lines by vertical lines to form loops , then finally flip every loop with probability 1/2 .
the generalization to the case where the expansion of the hamiltonian ( [ eq : decomposition ] ) consists of multiple terms is straight - forward .
in such a case , we simply choose a graph from the ones for which @xmath118 with the probability proportional to @xmath119 .
in order to see if the present procedure yields better statistics than previous methods , we calculate the spin-1/2 antiferromagnetic heisenberg chain of ten sites .
we set the initial value of the modification factor to @xmath120 for the whole calculation in the paper .
( for an appropriate initial value of the modification factor , see the discussions in the references@xcite . )
we set the cut off @xmath14 to be 500 .
this limits the accessible temperature range to @xmath121 . in the first stage , it takes approximately @xmath122 monte carlo steps ( mcs ) before the last run with @xmath10 terminates . in the second stage of simulation , we fix @xmath2 and let the random walker travel for @xmath123 mcs . the final estimate of the dos is shown in fig .
[ fig : fig1 ] . by using the dos in fig .
[ fig : fig1 ] , canonical ( fixed - temperature ) averages of the energy , the specific heat , the free energy , and the entropy are calculated as functions of temperature . in fig .
[ fig : fig2 ] , we show the relative error in the free energy , @xmath124 the top solid line is the one that is obtained based on the working estimate of @xmath22 resulting from the first stage that takes about @xmath122 sweeps .
the bottom dashed line represent the one based on the final estimate of @xmath22 obtained from the second stage using the estimator ( [ eq : theestimator ] ) .
the middle dotted line also represent the free energy based on the second stage data but without using the estimator ( [ eq : theestimator ] ) .
this corresponds to the standard procedure of the multicanonical simulation @xcite .
namely , the middle curve is based on the naive estimate of the dos using the histogram itself , i.e. , the estimate using the relation @xmath125 where @xmath126 is the trial dos estimated in the first stage and @xmath4 is the number of visits to the value @xmath1 counted during the second stage simulation .
the difference between the middle curve and the bottom one is as much as 1 to 2 orders of magnitude , and it clearly shows the utility of the estimator ( [ eq : theestimator ] ) .
the horizontal dashed line indicates the level of the largest error of the previous simulation by troyer , wessel and alet@xcite for the same system with the same size .
the number of monte carlo sweeps performed in their simulation was the same as the second stage of our simulation .
it is natural that their result is about the same position as the worst ( i.e. , the highest ) point in the middle curve , since the last few runs in the wl method , where the modification factor is very close to one , are almost equivalent to the ordinary multicanonical simulation with fixed weight . in order to see the system - size dependence of the efficiency , we calculate the standard deviation @xmath127 of the specific heat per site at temperature @xmath128 for various system sizes . again , we compare three results : ( 1 ) the one based on a wl method only with the termination condition @xmath11 ( denoted as `` wang - landau '' in the figure ) , ( 2 ) the one based on the two stage simulation with the termination condition @xmath10 for the first stage and with the naive estimation of the dos using the histogram itself in the second stage ( denoted as `` multicanonical '' ) , and ( 3 ) the same as the above but with the estimator ( [ eq : theestimator ] ) ( denoted as `` broad histogram '' ) . in ( 2 ) and ( 3 ) , the total number of monte carlo sweeps are chosen to be the same as that is spent in ( 1 ) in order to make the comparison fair .
we choose @xmath14 and @xmath23 so that @xmath129 is fixed . with this choice
the accessible temperature range is @xmath121 .
40 independent sets are performed for @xmath130 whereas 12 for @xmath131 .
the results are shown in fig .
[ fig : fig3 ] .
the system size dependence is small .
the results of the wl algorithm and the naive estimate of the dos based on the two - stage simulation are close to each other again .
the proposed method outperforms the other method by more than an order of magnitude in all the cases studied here .
generalizing the histogram methods@xcite based on the graph representation , we have presented a monte carlo method for quantum lattice models that consists of two stages . for the method s performance , the generalized broad histogram relation ( [ eq : bhrelation ] ) , proved for quantum systems , is crucial . in the first stage ,
the wl method is employed as in @xcite for obtaining the working estimate of the dos with a short cpu time . in the second stage ,
microcanonical averages are computed and the final estimate of the dos is obtained from them through the use of the generalized bh relation . in both the stages , the update of the states are performed by loop algorithm .
we have demonstrated that the proposed method yields very accurate estimates of the dos in the case of the @xmath103 antiferromagnetic heisenberg model , which we believe to be a typical case that represents many other cases . in ref .
@xcite , an alternative usage of the extended ensemble method was proposed in which one takes the series expansion not in @xmath16 but in other coupling constants .
this type of extension is also available in the present method .
we thank m. troyer , s. wessel and f. alet for comments and the exact diagonalization results that were useful for the error estimation .
we are also grateful to k. harada , h. otsuka , y. tomita , and t. surungan for useful comments .
c.y . thanks t. horiguchi , y. fukui , k. tanaka , k. katayama , n. yoshiike , and t. omori for useful comments .
n.k.s work was supported by the grant - in - aid ( program no.14540361 ) from monka - sho , japan .
999 n. metropolis , a. rosenbluth , m. rosenbluth , a. teller , and e. teller , j. chem . phys .
* 21 * , 1087 ( 1953 ) .
b. a. berg and t. neuhaus , phys .
b * 267 * , 249 ( 1991 ) .
j. lee , phys .
lett . * 71 * , 211 ( 1993 ) .
f. wang and d. p. landau , phys .
. lett . * 86 * , 2050 ( 2001 ) ; phys .
e * 64 * , 056101 ( 2001 ) .
h. g. evertz , in _ numerical methods for lattice quantum many - body problems _ , edited by d. j. scalapino ( perseus books , cambridge , 2001 ) .
r. swendsen and j .- s .
wang , phys .
* 58 * , 86 ( 1987 ) .
w. janke and s. kappler , phys .
* 74 * , 212 ( 1995 ) . c. yamaguchi , n. kawashima , phys .
e * 65 * , 056710 ( 2002 ) .
m. troyer , s. wessel , and f. alet phys .
lett . * 90 * , 120201 ( 2003 ) . c. yamaguchi , n. kawashima , and y. okabe , phys .
e * 66 * , 036704 ( 2002 ) .
m. c. de oliveira , t. j. p. penna , and h. j. herrmann , braz .
* 26 * , 677 ( 1996 ) ; eur .
. j. b * 1 * , 205 ( 1998 ) .
wang , t. k. tay and r. h. swendsen , phys .
* 82 * , 476 ( 1999 ) . throughout this paper
, we refer to the average with a fixed number of graphs as well as the one with a fixed value of the energy as a microcanonical average . o. f. syljusen and a. w. sandvik , phys .
e * 66 * , 046701 ( 2002 ) .
a. w. sandvik and j. kurkijarvi , phys .
b * 43 * , 5950 ( 1991 ) ; a. w. sandvik , j. phys .
a * 25 * , 3667 ( 1992 ) .
d. kandel and e. domany , phys .
b , * 43 * , 8539 ( 1991 ) . n. kawashima and j. gubernatis , phys .
e * 51 * , 1527 ( 1995 ) .
a. w. sandvik , phys .
b * 59 * , r14157 ( 1999 )
. n. kawashima and j. gubernatis , j. stat .
phys . * 80 * , 169 ( 1995 ) ; n. kawashima , j. stat .
phys . * 82 * , 131 ( 1996 ) . | we propose an efficient method for monte carlo simulation of quantum lattice models .
unlike most other quantum monte carlo methods , a single run of the proposed method yields the free energy and the entropy with high precision for the whole range of temperature .
the method is based on several recent findings in monte carlo techniques , such as the loop algorithm and the transition matrix monte carlo method . in particular , we derive an exact relation between the dos and the expectation value of the transition probability for quantum systems , which turns out to be useful in reducing the statistical errors in various estimates . |
velocity dispersion profiles and surface density profiles are among the most basic tools for investigating the structure of star clusters .
however , such investigations indicate that the region around the tidal radius , at which the internal acceleration of a star cluster is similar to the tidal acceleration due to the galactic tidal field , is particularly poorly understood .
+ velocity dispersion profiles sometimes show peculiarities which have been discussed in the literature .
@xcite , for example , observed a flattening in the outer parts of the velocity dispersion profile of the galactic globular cluster m15 which they interpreted as an effect of tidal heating by the general galactic tide or by tidal shocks .
@xcite also found a flattening of the velocity dispersion profile for @xmath3 cen and more recently for other galactic globular clusters like ngc6171 , ngc7099 and ngc288 @xcite .
the deviation from an expected keplerian fall - off in the velocity dispersion profile occurred in all clusters at radii where the internal gravitational acceleration is about @xmath4 and was therefore interpreted by scarpa et al .
as a hint of modified newtonian dynamics @xcite in globular clusters .
alternatively , they briefly discussed the possible effect of tidal heating or a dark matter halo on the cluster stars .
+ on the contrary , @xcite found that by fitting a @xcite model , which has a less sharp cut - off at the tidal radius than the commonly used @xcite model , to the re - analyzed @xmath3 cen data , its velocity dispersion profile could be explained without modifying newtonian gravity and without adding dark matter .
similar investigations of cluster profiles , e.g. @xcite and @xcite , were also not in favour of mond . + but it is not only the velocity dispersion profiles of star clusters which behave strangely at the tidal boundary ; their surface density profiles also show peculiarities and sometimes controversial behaviour .
+ @xcite showed that the surface density profiles of many globular clusters can be fitted by a simple analytical formula having a sharp cut - off radius , which could be interpreted as the tidal radius of the cluster .
later he derived a set of physically motivated models , with a similar cut - off radius corresponding to an energy cut - off in the energy distribution function of the cluster stars , which provided an even better fit to cluster profiles @xcite .
+ in contrast to that , @xcite , and more recently @xcite , found that young massive clusters in the lmc show an exponential surface density profile without any tidal truncation at the expected tidal radius .
they interpreted these findings as being due to tidal debris which was expelled at birth from the clusters and has not had time to disperse yet .
+ furthermore , @xcite , @xcite and @xcite find the outer halo milky - way globular clusters palomar 13 , whiting 1 and am 4 , respectively , to have a significant excess of stars at the tidal boundary , which makes any fit to the surface density data very inaccurate .
for all clusters they find the radial surface density profile , @xmath5 , to be well represented by a power - law @xmath6 with slopes of about @xmath7 .
this excess of stars is interpreted by the authors as heavy mass loss in a final stage of dissolution .
the same was found for palomar 5 , which is a well studied mw globular cluster close to the apogalacticon of its orbit @xcite .
+ moreover , @xcite showed that most globular clusters of the milky way , the lmc and smc , as well as of the fornax dwarf spheroidal , are more extended than can be explained by @xcite models , and therefore are better represented by @xcite models . in this context
they emphasised the lack of physical understanding of this phenomenon . + on the contrary , @xcite found that of 23 young massive clusters in m31 most were better fitted by @xcite models as these clusters do not show extended haloes . + as a consequence of this lack of understanding , @xcite asked in their investigation how robust the physical parameters are which were derived in such analyses .
this they tried to estimate by analyzing artificial clusters in the same way as the real observations .
a similar analysis has also been performed by @xcite , although they tested the analytic formula of @xcite with artificial observations under various limiting conditions .
both investigations came to the conclusion that physical parameters can in principle be well recovered from such idealized mock observations .
+ but probably the idealized nature of these investigations is misleading , since both tests were performed with dynamically unevolved clusters .
however , a self - consistent test of deriving physical parameters from a set of numerical computations of star clusters with a range of initial parameters has not been performed yet .
just a few investigations have touched this topic so far by means of numerical computations ( e.g. @xcite , @xcite , @xcite ) .
this is due to the fact that fast codes for globular cluster integration like fokker - planck or monte - carlo codes are not able to address this problem properly , as they can not follow the evolution of the tidal debris and are restricted to cut - off criteria in energy or angular momentum space .
moreover , limits in computational power prevented us from carrying out such investigations by means of collisional @xmath1-body codes .
but the recent improvements in computational speed of @xmath1-body codes and the availability of accelerator hardware ( grapes , gpus ) now allows us to study the dynamical evolution of star clusters with masses of up to several times @xmath8 . + thus , by computing various star clusters in a range of tidal conditions over several gyr we investigate the structure of star clusters , and in particular the evolution of the region around the tidal radius - the transition region between the star cluster and the tidal tails - in terms of velocity dispersion and surface density profiles .
the paper is organized as follows : first , a brief introduction to the topic of potential escapers is given in sec .
[ sec : pesc ] as these stars play a key role in this investigation .
then a few methodological remarks will be made in sec .
[ sec : method ] before we come to the velocity dispersion profiles in sec .
[ sec : vdp ] and then to the surface density profiles in sec .
[ sec : sdp ] . in the last section
we will give a summary with a brief discussion .
a major problem in producing velocity dispersion and surface density profiles from observations is how to disentangle cluster members from background / foreground stars . by only looking at stars which are inside
a projected , estimated jacobi radius with similar radial velocities and colours it is at least possible to distinguish between the cluster population and field stars . even here
, however , the question arises whether the choice of the cut in radial velocity significantly influences the results , especially of the velocity dispersion profile ( see @xcite ) .
moreover , one has to be careful with how the jacobi radius was estimated , i.e. was it evaluated using a mass estimate and was the mass estimated through the velocity dispersion ? or was the jacobi radius estimated by a cut - off radius in the profile and is the assumption that this cut - off radius is equal to the jacobi radius reasonable ?
what if the cluster is on an eccentric orbit about the galaxy ; how does this influence the cut - off radius ?
+ besides , the observer does not know if the stars in the sample which was extracted in this way are actually members of the cluster or if they are already evaporated from the cluster and are now part of its tidal debris , and just lie in projection within the cluster . + but even if the observer could distinguish between stars inside the jacobi radius and those that are beyond this radius , it would still be unclear if the stars in the sample are bound to the cluster or have already gained enough energy to leave the cluster but just havent done so yet .
these , so - called , potential escapers will have a significant influence on both the surface density and the velocity dispersion profile , if there is a non - negligible fraction of them in the cluster , as these stars will make the profiles deviate from any theory which does not take them into account - which most theoretical approaches do not .
+ in numerical modelling of star clusters we have detailed phase space information on every single star in the computation . mainly by numerical investigations it has been found that the escape process through which stars escape from a cluster in a constant tidal field , e.g. on a circular orbit about a galaxy , is divided into two steps .
first the stars get unbound via two - body relaxation , thus on a relaxation time scale ( ignoring constants ) , @xmath9 where @xmath1 is the number of stars in the cluster and @xmath10 is the crossing time of the cluster ( note that we neglected the slowly varying coulomb logarithm in eq .
[ eq : trel ] , for further details see e.g. @xcite ) . in the second step , these energetically unbound stars , or potential escapers , with specific energy @xmath11 higher than some critical escape energy @xmath12 , escape from the cluster on an escape time scale which is given by ( @xcite , equation 9 therein ) @xmath13 where @xmath12 is the critical energy at the lagrange points , i.e. at the jacobi radius . from this
it follows that the excess energy can be written as @xmath14 on the other hand , the relation @xmath15 holds , whence we find that @xmath16 hence , a fraction of stars with excess energy @xmath17 escapes on a time scale @xmath18 and the time scale of mass loss , e.g. the time on which a cluster dissolves , @xmath19 , is given by @xmath20 which was also found by @xcite .
this result is in contrast to the ` classic ' picture where the dissolution times of clusters scale with the relaxation time ( see e.g. @xcite ) , i.e. @xmath21 and demonstrates the importance of potential escapers on the dissolution process of star clusters .
the remaining question is whether the population of potential escapers is large enough that it also has an influence on the velocity dispersion and surface density profiles
. for clusters of a few @xmath2 stars in a constant tidal field @xcite finds about 10 - 20% of all stars within the jacobi radius to be potential escapers , @xcite even find the fraction of potential escapers for similar clusters to be 1/3 of the cluster population .
moreover , @xcite found that for clusters in a constant tidal field the fraction of potential escapers varies with the number of stars within a cluster approximately as @xmath22 .
thus , this fraction should still be significant for high-@xmath1 globular clusters .
+ star clusters in time - dependent tidal fields have not been investigated in this respect yet , but as tidal perturbations tend to increase the energy of the stars in a cluster ( e.g. @xcite ) , the population of potential escapers should be even larger in such clusters . thus , potential escapers should have a significant effect on both the velocity dispersion and the surface density profiles of all kinds of star clusters .
when analyzing velocity dispersion and surface density data of star clusters , confusion over nomenclature often arises .
therefore we strictly stick to the following three terms : _ profiles _ , _ models _ and _ templates_. the first will be used for the data , which in our case comes from @xmath1-body computations but might also originate from observations .
the term _ models _ will be used for physically motivated distribution functions such as those proposed by @xcite or @xcite . the word _
templates _ covers the analytic expressions which have been found empirically and which do not originate from a physical derivation by being solutions of the collisionless boltzmann equation ( e.g. @xcite ) .
the word template hereby emphasises their artificial nature .
+ further confusion arises about the term _ tidal radius_. again there is a physically motivated and an empirical version .
the former is the radius at which the internal gravitational acceleration equals the tidal acceleration from the host galaxy ( see e.g. @xcite ) . this will be named _
jacobi radius _ in our text .
the empirical tidal radius is defined through the cut - off radius of the models and templates which are fitted to the profiles .
we will refer to this cut - off radius as _ edge radius _ here and would like to emphasise that jacobi radius and edge radius are two different concepts and therefore are not necessarily equal .
in fact , we will show that the two are often significantly different and that conclusions from a fitted edge radius should be drawn with caution .
+ note also that the results of this investigation can not be easily scaled to more massive globular clusters , as the relative importance of tidal features depends on the mass of the cluster , since the mass in the tidal debris is proportional to the mass - loss rate of the cluster and the mass - loss rate does not scale linearly with cluster mass , @xmath23 , but rather with @xmath24 @xcite .
+ moreover , even though we have some concentrated clusters in our sample this investigation focuses on less concentrated clusters , with a ratio of projected half mass radius to jacobi radius between @xmath25 because @xcite showed that clusters with ratios of @xmath26 can not be properly fitted by either @xcite templates or @xcite models as their ratio of @xmath27 is too high . therefore
, our conclusions mainly hold for less concentrated clusters which @xcite named the _ extended cluster population _ , and which are likely to be on the main sequence of cluster evolution @xcite , the final and universal stage of cluster evolution after core collapse , where the ratio @xmath27 only depends on the mass of the cluster . + in the following sections we will first give some details about our data set before we explain how we extract the velocity dispersion and surface density profiles .
thereafter we briefly describe the analytical templates which are used in this investigation and specify our fitting method .
.overview of all computed models .
@xmath28 gives the apogalactic radius of the cluster orbit and @xmath29 the corresponding eccentricity .
the latter is defined as @xmath30 with @xmath28 being the apogalactic radius of the cluster orbit , and @xmath31 being the perigalactic radius .
@xmath27 gives the initial concentration of the model , whereas @xmath32 is the initial half - mass radius .
the clusters are divided into three categories : extended clusters in a constant tidal field ( a0-d0 ) , extended clusters in time - dependent tidal fields ( a1-d3 ) and compact clusters ( a0c - d0c ) .
note that each model has initially @xmath33 stars , a lower stellar mass limit of @xmath34 and an upper limit of @xmath35 . [ cols="<,^,^,^,^",options="header " , ] | based on our recent work on tidal tails of star clusters @xcite we investigate star clusters of a few @xmath0 by means of velocity dispersion profiles and surface density profiles .
we use a comprehensive set of @xmath1-body computations of star clusters on various orbits within a realistic tidal field to study the evolution of these profiles with time , and ongoing cluster dissolution .
+ from the velocity dispersion profiles we find that the population of potential escapers , i.e. energetically unbound stars inside the jacobi radius , dominates clusters at radii above about 50% of the jacobi radius . beyond 70% of the jacobi radius
nearly all stars are energetically unbound .
the velocity dispersion therefore significantly deviates from the predictions of simple equilibrium models in this regime .
we furthermore argue that for this reason this part of a cluster can not be used to detect a dark matter halo or deviations from newtonian gravity . + by fitting templates to the about @xmath2 computed surface density profiles we estimate the accuracy which can be achieved in reconstructing the jacobi radius of a cluster in this way .
we find that the template of king ( 1962 ) works well for extended clusters on nearly circular orbits , but shows significant flaws in the case of eccentric cluster orbits .
this we fix by extending this template with 3 more free parameters . our template can reconstruct the tidal radius over all fitted ranges with an accuracy of about 10% , and is especially useful in the case of cluster data with a wide radial coverage and for clusters showing significant extra - tidal stellar populations .
no other template that we have tried can yield comparable results over this range of cluster conditions .
all templates fail to reconstruct tidal parameters of concentrated clusters , however .
+ moreover , we find that the bulk of a cluster adjusts to the mean tidal field which it experiences and not to the tidal field at perigalacticon as has often been assumed in other investigations , i.e. a fitted tidal radius is a cluster s time average mean tidal radius and not its perigalactic one .
+ furthermore , we study the tidal debris in the vicinity of the clusters and find it to be well represented by a power - law with a slope of -4 to -5 .
this steep slope we ascribe to the epicyclic motion of escaped stars in the tidal tails .
star clusters close to apogalacticon show a significantly shallower slope of up to -1 , however .
we suggest that clusters at apogalacticon can be identified by measuring this slope .
[ firstpage ] galaxies : kinematics and dynamics galaxies : star clusters methods : analytical methods : @xmath1-body simulations |
shear flows provide one of the paradigms of transport processes @xcite .
the importance of chaos in the equations underlying macroscopic shearing has recently been addressed by various numerical studies @xcite , which to some extent were supported by kinetic theory @xcite and rigorous mathematical work @xcite .
in contrast , however , a simple , exactly solvable model based on a low - dimensional chaotic dynamics whose mixing property would be the cause of irreversibility has not yet been established . for material and heat transport such type of models
have helped to understand the physical content of thermostating schemes used in numerical simulations @xcite ( cf .
however @xcite for open questions ) . in the present article
we introduce a similar model for shear flows in the hope that it can also serve such a purpose .
the approach will be based on multibaker maps .
previous works in this spirit successfully described the phenomena of diffusion @xcite , conduction in an external field @xcite , chemical reactions @xcite , thermal conduction @xcite , and cross effects due to the simultaneous presence of an external field and heat conduction @xcite .
our aim is to model a sheared fluid confined between two parallel walls at the coordinates @xmath0 and @xmath1 ( fig .
[ fig : geometry ] ) . the flow is assumed to be two - dimensional in the @xmath2 plane .
shear is induced by prescribing different @xmath3 components of the average velocities @xmath4 of particles close to the respective walls . in order to make the calculations more transparent
, we confine the discussion to cases where the driving is sufficiently weak to induce only a laminar flow , i.e. , to cases where the velocity of particles is always directed in the vertical direction such that @xmath5 . for this system we establish a local entropy balance that covers time dependent effects and does not rely on the implementation of boundary conditions .
three different boundary conditions for dealing with the dissipated heat are considered : * ( i ) * in the simplest case the system is isolated .
a stationary linear velocity profile emerges in that case , and the temperature becomes uniform .
no steady state is reached , however , due to a constant increase of temperature in response to the viscous heating .
in addition , we consider systems where * ( ii ) * there is a bulk thermostat uniformly taking out the viscous heat , and * ( iii ) * the temperature is fixed to the same value at both boundaries so that the asymptotic temperature profile is stationary , but no longer uniform . @xmath6{fig1}\ ] ] the paper is organized as follows : in sect .
[ sec : thermodynamics ] we recall basic notions of irreversible thermodynamics that are to be recovered in a suitable continuum limit of the multibaker dynamics . in sect .
[ sec : multibaker ] we introduce the model , and establish the evolution equations for the velocity and the temperature field .
this allows us to address the entropy dynamics and its balance equation ( sect .
[ sec : entropy ] ) .
subsequently , in sect . [
sec : limit ] the macroscopic limit of the resulting equations is worked out . the global behavior at different boundary conditions
is compared to irreversible thermodynamics in sect .
[ sec : macflow ] , and conclusions are drawn in sect .
[ sec : conclude ] .
in this section we recall the thermodynamic description of shear flows accompanied by viscous heating . the picture is simplified by considering an incompressible fluid at constant pressure . for a system with constant density and pressure
the temperature @xmath7 is the only relevant state variable , and for a complete description one also has to specify the velocity field @xmath4 of the fluid @xcite .
the thermodynamic state variables are the velocity field @xmath4 and the temperature @xmath7 @xcite .
mass conservation is expressed by a continuity equation . for incompressible fluids the uniform mass density
@xmath8 implies that the flow is divergence free , i.e. , @xmath9 here @xmath10 labels the components of the local flow velocity @xmath11 , and we adopted the einstein convention , i.e. , summation over repeated indices .
the equation of motion for the velocity components is given by the navier - stokes equation .
for the case of a negligible pressure gradient it reads @xmath12 where @xmath13 is the kinematic viscosity , and @xmath14 is the total time derivative .
the system of equations is closed @xcite by the equation @xmath15 for the temperature evolution . here
, @xmath16 is the specific heat at constant volume , and @xmath17 is the thermal conductivity .
the balance equation of the entropy density @xmath18 is @xmath19 where @xmath20 is the irreversible entropy production reflecting the viscous heating of the flow , @xmath21 the entropy flux , and @xmath22 is the entropy - current density .
the term @xmath23 models an entropy flux let into a heat bath . in the bulk of a hydrodynamic system
this term typically vanishes , but it takes non - vanishing values wherever there is a flux into the environment ( i.e. , for instance at the boundaries when there is a heat flux through the walls ) . in simulations
where one does not desire to focus on the effects of temperature gradients , and applies a thermostating algorithm , a non - vanishing @xmath23 can appear even in the bulk of the system .
we say that a system is subjected to an _ ideal thermostat _ , when the dissipated heat is directly let into the heat bath . in that case
the entropy current vanishes in the steady state , @xmath24 , while @xmath23 is non - zero in the bulk , and counterbalances the steady - state entropy production . for an incompressible fluid in local equilibrium
the gibbs relation @xmath25 applies locally .
the evolution equations can be used to evaluate the terms in eq .
( [ eq : s - dot ] ) . a straightforward calculation yields for the rate of irreversible entropy production ( cf .
for instance ( * ? ? ?
( 23 ) ) ) @xmath26 the associated entropy current takes the form ( * ? ? ?
( 22 ) , ( 24 ) ) @xmath27 it depends only on the local temperature and its gradient .
the flow velocity @xmath4 does not enter . for a laminar flow driven by prescribed non - trivial @xmath3 components of the velocity at the two walls ,
the velocity field at any position @xmath2 takes the form @xmath28 ( see fig .
[ fig : geometry ] ) .
the @xmath29 component of the velocity vanishes , and the profile is translational invariant in the @xmath3 direction ( parallel to the walls ) . we restrict our investigation to cases where the same holds for the temperature such that @xmath30 .
consequently , for the laminar flow the transport equations take the form @xmath31 while the rate of irreversible entropy production @xmath20 and the entropy current @xmath32 can be written as @xmath33 the scalar current @xmath32 denotes the @xmath29 component of the entropy current , and an analogous convention is adopted for all currents .
the @xmath3 components of the currents vanish in the considered setting .
@xmath34{fig2a}\ ] ] @xmath34{fig2b}\ ] ]
in order to model the laminar flow we divide the plane into square cells of size @xmath35 that are large enough , on the one hand , to admit thermodynamically meaningful averages , but , on the other hand , sufficiently small to neglect gradients across cells . the cells are considered as the regions used in irreversible thermodynamics to define local equilibrium variables . the system shown in fig .
[ fig : geometry ] is represented by a rectangular array of @xmath36 cells in horizontal direction , and an infinite number of cells vertically ( fig . [ fig : mbaker2d]a ) .
cells are labeled by the indices @xmath37 , and @xmath38 .
all cells have the same dynamics except the outermost ones where it is modified to implement boundary conditions . after each time unit @xmath39
every cell is divided into five columns ( fig .
[ fig : mbaker2d]b ) . the rightmost ( r ) and leftmost ( l ) column have width @xmath40 .
they are mapped onto a strip of height @xmath40 in cell @xmath41 and cell @xmath42 , respectively . the left column ( u ) of widths @xmath43
is mapped upwards onto a strip of height @xmath43 in cell @xmath44 , and the right one labeled by ( d ) downwards into @xmath45 , respectively .
region @xmath46 stays within the cell . in all cases
the area of the strips is preserved .
this dynamics is driving two fields . denoting the composite index @xmath47 as @xmath48 ,
these fields are + ( i ) the velocity field @xmath49 describing the mean flow velocity @xmath50 in cell @xmath48 + ( ii ) an energy field @xmath51 that represents the kinetic energy of the fluid in cell @xmath48 . the dynamics on this two - dimensional lattice can be considered as a model of the velocity and energy transport in the configuration space . in order to obtain a faithful representation of the entropy balance , however , one has to consider the phase - space dynamics . to define the dynamics in the analog of a @xmath52-space , we take into account the translation invariance of the problem .
the velocity and energy fields can take on different values in the columns @xmath53 of fig .
[ fig : mbaker2d]a , but the fields have to be uniform within every column . in this respect
it is not necessary to follow the dynamics in the @xmath3 direction , and one achieves a quasi one - dimensional dynamics in an @xmath54 space , where @xmath55 represents the non - trivial momentum - like variable of the model .
the domain of the multibaker in the @xmath54 space is shown in fig .
[ fig : mbaker]a .
it comprises a chain of @xmath56 cells of size @xmath57 , that for sake of more condensed notations are labeled by the index @xmath58 .
the middle @xmath59 cells represent the bulk , and two additional ones are used to implement boundary conditions ( cf . section [ sec : bc ] ) . the parameter @xmath60 sets the momentum scale .
it will not play any role in thermodynamic considerations . in order to maintain the same dynamics in the transport direction in the two representations of the flow , each cell
is divided into three columns of size @xmath40 , @xmath61 and @xmath40 , where @xmath62 is the same as above and @xmath63 .
the left and right columns of cell @xmath58 are mapped into a strip of height @xmath40 in cell @xmath64 and cell @xmath65 , respectively , as shown in fig .
[ fig : mbaker]b .
the middle column of size @xmath61 is squeezed and stretched onto a strip of height @xmath61 and remains in the same cell .
this multibaker dynamics drives the velocity and energy fields .
their values might depend on the phase - space coordinate .
hence , we are dealing with the bivariate distribution of the velocity @xmath66 and kinetic energy @xmath67 within each cell @xmath58 .
only the cell averages , @xmath68 and @xmath69 appear in the transport equations .
the dependence of @xmath66 and @xmath67 on the phase - space coordinates contributes essentially to the entropy dynamics , however .
@xmath34{fig3}\ ] ] the mass density @xmath70 is proportional to the average phase - space density @xmath71 , in cell @xmath58 , i.e. , @xmath72 where @xmath73 is a constant of dimension mass .
when interpreting the mass density or associated quantities from the point of view of the two - dimensional flow , @xmath70 is to be understood as a mass density in the @xmath29 coordinate per unit length in @xmath3 direction .
hence , @xmath74 represents the preserved momentum density ( per unit length ) of the hydrodynamic flow .
its evolution expresses momentum conservation . as a consequence ,
the updated values for the velocity on the strips @xmath75 after one time unit are ( cf . the right side of fig .
[ fig : mbaker]b ) @xmath76 where , @xmath77 , @xmath78 , and @xmath79 stand for the velocities of the flow in the regions @xmath75 of cell @xmath58 , respectively .
the prime indicates that the updated values of the field are considered .
the average momentum @xmath74 of the full cell is the average of the contributions on the different strips , i.e. , @xmath80 at any instant of time .
if the velocity is initially uniform in the full cell ( i.e. , the values of @xmath81 in all strips agree ) , then due to eqs .
( [ eq : v - average ] ) and ( [ eq : v_update ] ) the average velocity of cell @xmath58 becomes after one time step @xmath82 observing that the mass density @xmath8 is uniform and constant in time , the updated value of the cell velocity @xmath68 is found to be @xmath83 this evolution can be written in the form of a discrete balance equation @xmath84 with the discrete current @xmath85 in the boundary columns @xmath86 and @xmath87 the dynamics will be augmented by force terms ( sect . [
sec : bc ] ) in order to fix the velocity to the constant values @xmath88 and @xmath89 , respectively , irrespective of the momentum flowing into these cells .
the energy of cell @xmath58 is obtained by integrating @xmath90 over the volume @xmath91 of the cell . at any instant of time
it is the spatial average of the energies @xmath92 , @xmath93 and @xmath94 of the different strips , @xmath95 the difference between the energy @xmath96 and the translational specific kinetic energy @xmath97 of the flow defines the specific kinetic - energy density @xmath98 , whose macroscopic limit will be proportional to the local temperature .
therefore , on the strip @xmath99 in cell @xmath58 one observes kinetic - energy densities @xmath100 that fulfill @xmath101 while the coarse - grained kinetic energy @xmath102 obeys at the same time @xmath103 using these definitions and the averaging rule ( [ eq : e - average ] ) for the energy one finds @xmath104 .
\label{eq : w - average}\ ] ] 2 this shows that the average kinetic energy @xmath69 is not the straightforward spatial average of the values @xmath100 on the strips
. rather intra - cell variations of the velocity field ( i.e. , a non - trivial distribution of the @xmath105 ) also contribute to @xmath69 . for a thermally closed system the specific kinetic energy @xmath106
is advected by the flow .
the values on the strips after one time unit are then @xmath107 from this and the update ( [ eq : v_update ] ) of the velocities the updated kinetic energy is found to be @xmath108 \nonumber \\ & - & \frac{m}{2 } { \left [ a^2 g \frac{(v_{m-1}+v_{m+1}-2v_m)}{a^2 } \right]}^2 .
\label{eq : tpn0}\end{aligned}\ ] ] the terms proportional to @xmath73 describe the effect of viscous heating in this discrete setting .
in order to model the action of a thermostat , that leads to local changes of the specific kinetic energy due to a heat flux into the environment , an additional _ thermostat heat source _ @xmath109 is incorporated into the update of @xmath69 by multiplying the right - hand side of eq .
( [ eq : tpn0 ] ) by a factor @xmath110 $ ] , @xmath111 \nonumber \\ & - & \frac{m}{2 } g^2 { \left [ v_{m-1}+v_{m+1}-2v_m \right]}^2 \biggr\ } \ ; [ 1 + \tau q_m ] .
\label{eq : tpn}\end{aligned}\ ] ] this equation can be rewritten as @xmath112 with the _ full heat source _
@xmath113 \nonumber \\ & - & \frac{\tau}{2 } m { \left [ \frac{a^2 g}{\tau } \frac{(v_{m-1}+v_{m+1}-2v_m)}{a^2 } \right]}^2 .
\label{eq : qmmicr}\end{aligned}\ ] ] the equation for the update of @xmath106 can be rearranged into a balance equation for the heat per unit volume @xmath114 which comprises the divergence of the discrete `` heat '' current @xmath115 the first contribution to @xmath116 in ( [ eq : qmmicr ] ) reflects the action of the thermostat , and the latter two the effect of viscous heating of the fluid . a steady state with a uniform @xmath106 profile can be found for @xmath117 , which thus mimics an ideal thermostat .
in contrast , in a bulk system in the sense of conventional irreversible thermodynamics @xmath118 , and @xmath116 only vanishes when the discrete velocity gradients @xmath119 and @xmath120 across the cell vanish .
the gibbs entropy for cell @xmath58 of a multibaker system with a density field @xmath121 ) and specific kinetic - energy field @xmath67 takes the form @xcite
@xmath122 \label{eq : defsg}\ ] ] where @xmath123 is a free exponent .
since the phase - space density @xmath124 is constant , one can identify the reference density @xmath125 with @xmath124 , and obtains @xmath126 here , we have already replaced the phase - space density by the mass density ( [ eq : rhomm ] ) .
the coarse - grained entropy of cell @xmath58 is based on the cell - averaged coarse - grained value @xmath69 of the kinetic energy @xmath67 . for the form ( [ eq : smg ] ) of the gibbs entropy , one obtains @xmath127 as the coarse - grained entropy . in order to discuss the thermodynamic time evolution of entropies , one conveniently starts with uniform densities in every cell @xcite . in that case
the coarse - grained and the gibbs entropy initially coincide .
after one time step , however , they typically differ .
the gibbs entropy has changed due to the fact that the @xmath106 field takes different values on the strips @xmath128 , leading to the new value @xmath129 + ( 1 - 2 g ) \ ; \ln [ w_m \ ; ( 1+\tau q_m ) ] + g \ ;
\ln [ w_{m-1 } \ ; ( 1+\tau q_m ) ] \right\ } \nonumber \\ & = & k_b \gamma \varrho a b \left\ { \ ; \ln [ w_m \ ; ( 1+\tau q_m ) ] + \ , g \ ; \ln\frac{w_{m-1}}{w_m } - \ , g \ ; \ln\frac{w_m}{w_{m+1 } } \right\ } .\end{aligned}\ ] ] 2 on the other hand , the coarse - grained entropy after one time step is @xmath130 the coarse - grained entropy only depends on the average kinetic - energy density @xmath69 , and is therefore considered as the multibaker analogue of the thermodynamic entropy .
its temporal change can be decomposed as @xmath131 where information - theoretic arguments @xcite lead one to identify @xmath132 with the rate of entropy production , and @xmath133 [ eq : deltasdef ] with the entropy flux .
note that the second term of the numerator of @xmath134 vanishes due to the initial condition of uniform fields in the cells .
inserting the values for @xmath135 and @xmath136 into the definition ( [ eq : deltasidef ] ) yields @xmath137 .
\nonumber \end{aligned}\ ] ] remarkably , this expression for the entropy production does not depend on the source term @xmath116 , but only on the values of the coarse - grained field @xmath106 in cell @xmath58 and its neighbors .
the shear flow enters only indirectly through the update @xmath138 of @xmath69 .
similarly to the other balance equations , the entropy flux can be written as @xmath139 where the discrete entropy current @xmath140 takes the form @xmath141 and @xmath142 describes the direct flux into the thermostat .
such a flux is encountered whenever there is a non - vanishing source term @xmath109 . in view of eq .
( [ eq : qmmicr ] ) this finding further supports the interpretation of @xmath109 and @xmath116 given at the end of sect .
[ ssec : efield ] .
in this section we evaluate the expressions of the different quantities considered in sects . [ sec : multibaker ] and [ sec : entropy ] , and identify conditions for consistency with the thermodynamic results described in sect .
[ sec : thermodynamics ] .
the macroscopic limit implies that @xmath143 ( i.e. , @xmath144 ) , and @xmath39 is much smaller than typical macroscopic time scales .
formally it can be defined as @xmath145 such that the spatial coordinate @xmath146 is finite . as mentioned earlier ,
the field @xmath106 is assumed to go over into the local temperature @xmath147 in the macroscopic limit , i.e. , @xmath148 where @xmath149 is a dimensionless constant .
the macroscopic form of the velocity current ( [ eq : jmv ] ) becomes @xmath150 in order to achieve a meaningful thermodynamic result the ratio @xmath151 has to be finite in the limit .
indeed , comparison with eq .
( [ eq : ptv ] ) shows that @xmath152 is the kinematic viscosity .
equations ( [ eq : jmv - limit ] ) and ( [ eq : ga2 ] ) imply that eq .
( [ eq : vn ] ) is the discrete form of the navier - stokes equation ( [ eq : ptv ] ) for the laminar flow .
moreover , with this choice for @xmath62 one obtains for the heat current ( [ eq : jmw ] ) @xmath153 hence , fick s law of heat conduction is recovered with the heat conduction coefficient @xmath154 due to eq .
( [ eq : lambda1 ] ) , the macroscopic limit of eq .
( [ eq : tpn ] ) reduces to eq .
( [ eq : pttshort ] ) for the temperature evolution : @xmath155 comparing the coefficients in this equation with the ones in eq .
( [ eq : pttshort ] ) , one obtains @xmath156 the proportionality constant @xmath149 introduced in eq .
( [ eq : ckbt ] ) corresponds thus to the specific heat at constant volume ( measured in units of @xmath157 ) . in the macroscopic limit
the rate of irreversible entropy production ( [ eq : dism ] ) becomes @xmath158 it fully agrees with the thermodynamic form of the entropy production ( [ eq : sirrshort ] ) when the coefficient @xmath159 in front of the first term is the heat conductivity . in view of ( [ eq : lambda1 ] )
we thus conclude , that in our model @xmath160 , i.e. , the exponent @xmath123 appearing in the definition ( [ eq : defsg ] ) of the entropy is proportional to the specific heat .
hence , the final form of the heat conductivity @xmath17 can be settled to @xmath161 working out the expression ( [ eq : jmsdef ] ) of the entropy current @xmath140 , one finds in the macroscopic limit @xmath162 in view of ( [ eq : lambda2 ] ) this relation also fully agrees with its thermodynamic counterpart ( [ eq : jsshort ] ) .
in this section we demonstrate how boundary conditions for the shear flow can be implemented in the multibaker dynamics .
@xmath163{fig4}\ ] ] the multibaker dynamics for cell @xmath164 is shown in fig .
[ fig : bc ] .
( an analogous
prescription holds in cell @xmath87 . )
the action of the strip mapped from cell @xmath164 to @xmath165 fully agrees with the one of the corresponding strips in the bulk such that the dynamics of cell @xmath165 agrees with the bulk dynamics of fig .
[ fig : mbaker]b .
the important difference about the dynamics at the boundary is that + ( i ) there is no column leaving cell @xmath164 to the left , and + ( ii ) the momentum of the particles entering cell @xmath164 is not preserved .
+ condition ( i ) reflects that no particles can penetrate the walls of the system , and ( ii ) reflects that there is a force exerted by the boundary on the fluid . in strip @xmath166 entering cell @xmath164 ( and analogously strip @xmath167 entering cell @xmath87 ) the update of the velocity consequently contains the contribution of the external force @xmath168 ( @xmath169 ) , leading to @xmath170 the update of the coarse - grained velocity of the leftmost cell is @xmath171 and an analogous relation holds for cell @xmath87 .
a shear flow with a prescribed shear rate is enforced by adjusting the forces in such a way that @xmath172 and @xmath173 take the respective values @xmath174 and @xmath175 of the velocities at the walls at any time . in view of the time evolution of @xmath68 , in a steady state [ eq .
( [ eq : vn ] ) with @xmath176 , the asymptotic velocity profile becomes linear @xmath177 irrespective of the temperature distribution in the system .
the modification of the dynamics of the velocities at the boundary also implies changes of the thermostat source terms .
since the velocity in cell @xmath164 is uniform by definition , the update of @xmath178 can be found from eq .
( [ eq : w - average ] ) where the terms containing @xmath179 and @xmath180 are not present since no particles enter cell @xmath164 from the left .
consequently , @xmath181 ( 1 + \tau q_0 ) .\ ] ] this leads to the discrete temporal change @xmath182 instead of eq .
( [ eq : wmbalance0 ] ) . for a steady state @xmath183
vanishes , and the equation implies [ cf .
( [ eq : jmw ] ) ] @xmath184 } .
\label{eq : q0}\ ] ] the right - hand side of this expression is finite in the macroscopic limit , where @xmath185 since , however , @xmath186 in the macroscopic limit , the thermostat heat source @xmath187 can not be interpreted as a density when there is a finite heat current @xmath188 .
its integral over the cell , @xmath189 , stays however finite . as a consequence ,
the heat current has to vanish in a steady state when @xmath190 .
we have seen that in the macroscopic limit the local momentum , energy and entropy balances for the multibaker map coincide with their thermodynamic forms .
this result was achieved by inspecting the local time evolution of the densities without referring to particular boundary conditions , and it is not restricted to steady states . in order to underline these features , we discuss the global transport pictures for three different settings of transport .
we emphasize , that the considerations immediately generalize to arbitrary time - dependent states although the particular calculations are carried out for stationary coarse - grained velocity fields . in thermodynamics sheared systems
are often considered to be thermally isolated . in that case @xmath109
identically vanishes for all cells @xmath191 .
( [ eq : w0update ] ) implies for the current ( [ eq : jmw ] ) @xmath192 assuming that the time derivative of @xmath178 is finite in the macroscopic limit , the current @xmath193 approaches zero as @xmath194 .
an analogous arguments applies also to @xmath195 .
thus , for the asymptotic state , the boundary conditions imply that @xmath178 and @xmath196 practically coincide with @xmath197 and @xmath198 , respectively .
moreover , the viscous heating term @xmath116 in the bulk only depends on the square of @xmath119 .
the heating is therefore spatially uniform in the long time limit . with this input
, one immediately verifies that @xmath69 approaches a spatially uniform value @xmath199 .
however , in view of the heat source @xmath116 in ( [ eq : qmmicr ] ) , it constantly grows in time , @xmath200 this _ temporal _ evolution of @xmath106 reflects the rise of temperature due to the entropy production in the system .
after all , @xmath201 which in the macroscopic limit reduces to the thermodynamically expected value @xmath202 where @xmath203 is the spatially uniform , but temporally increasing asymptotic temperature distribution .
note that the flux @xmath23 and the entropy current vanishes in this setting , implying an ever increasing entropy . in a system subjected to an ideal thermostat the viscous heating is instantaneously released into a heat bath .
accordingly , @xmath116 vanishes for all cells @xmath204 such that @xmath109 can be determined from eq .
( [ eq : qmmicr ] ) . moreover , according to the results obtained in sect .
[ ssec : bcw ] we are again dealing with a system with vanishing heat currents through its boundaries , i.e. , for the asymptotic state the boundary conditions imply @xmath205 and @xmath206 , and there is no source term @xmath207 for the heat . consequently , @xmath69 approaches a spatially uniform state @xmath199 that is _ stationary _ in this case .
entropy production arises due to the nontrivial form of @xmath109 [ cf .
( [ eq : qmmicr ] ) with @xmath208 @xmath209 in the bulk [ cf . the steady - state velocity profile ( [ eq : vm - distr ] ) ] .
hence , the irreversible entropy change is @xmath210 and in the macroscopic limit it reduces to the thermodynamically expected value @xmath211 where @xmath212 is the steady - state temperature .
also in this case there is no entropy current in the steady state .
however , at every location in the system there is a non - vanishing entropy flux into the thermostat @xmath213 which exactly compensates @xmath20 such that the entropy indeed is stationarity .
finally , we discuss a steady state in a hydrodynamic bulk system that generates heat flux into the walls due to the prescribed temperatures at the boundaries .
there are no thermostat heat sources in the bulk , i.e. , @xmath118 for @xmath204 . on the other hand ,
there are source terms @xmath187 and @xmath214 acting in the two outermost cells that fix the values of @xmath178 and @xmath196 ( i.e. , the temperature ) to the same constant value @xmath178 . to this end
the sources counterbalance a macroscopic heat current releasing the viscous heat into the bath at the boundaries [ cf .
( [ eq : q0 ] ) ] .
in contrast to the previous cases , the discrete heat current does not vanish at the boundaries . in view of eqs .
( [ eq : jmw ] ) and ( [ eq : w0update ] ) the heat current through the left boundary of the steady state system is @xmath215 since the dynamics is symmetrical and @xmath216 , the current at the right boundary takes the same value up to a change of the sign @xmath217 in a steady state the cells in the bulk consequently fulfill @xmath218 leading to a linear profile of the heat current , @xmath219 in view of the definition eq .
( [ eq : jmw ] ) of @xmath220 , this implies a parabolic profile of @xmath106 in the steady state , @xmath221 \ ; m \ ; \left ( \frac{\delta v}{a(n+1 ) } \right)^2 , \ ] ] as expected for the temperature profile in a shear flow subjected to identical temperatures at the two ends .
the entropy production in the bulk is related to the spatial variation of the steady - state distribution @xmath222 [ cf .
( [ eq : dism ] ) ] , @xmath223 \nonumber\\ & = & \frac{k_b \gamma}{m } \left [ \frac{1}{w^*_m } \ ; \frac{j_{m+1}^{(w ) } - j_{m}^{(w)}}{a } + \frac{{j_{m+1}^{(w)}}^2 + { j_{m}^{(w)}}^2}{2 \rho \nu w_m^{*2 } } \right ] , \end{aligned}\ ] ] where the second equation was obtained by expanding the logarithms to second order in the differences @xmath224 , and rearranging terms . using eqs .
( [ eq : wmbalance ] ) , ( [ eq : qmmicr ] ) and ( [ eq : jmwlin ] ) , the first term can be related to the applied shear rate , while the second one represents the contribution to the entropy production arising from the heat flow .
indeed , in the macroscopic limit one recovers in this case both contributions to eq .
( [ eq : deltasmirr ] ) .
( for small shear rates the contribution from the temperature change is , however , negligible ) . correspondingly , there is a finite entropy current @xmath32 at every point in the system , but only at the boundaries there is a flux into the thermostat . according to eq .
( [ eq : q0limit ] ) the macroscopic limit of the full entropy flux through the boundaries becomes @xmath225 since the stationary temperature profile is obtained from ( [ eq : wmparab ] ) in the form of @xmath226 ( \delta v /l )
^2 $ ] , the derivative at the origin is @xmath227 .
the contribution at the right boundary is the same .
therefore , we find for the integrated entropy flux @xmath228 where we used @xmath229 . for weak shear
this is essentially the same as the integral of the constant flux ( [ eq : phithisol ] ) over the chain .
thus , we conclude that for sufficiently small heat currents thermostating in the bulk and in the boundaries can lead to the _ same _ global behavior .
the work per unit time done by the external force densities @xmath168 and @xmath169 can be evaluated using their form eq .
( [ eq : wallveloc ] ) and the value of the constant velocity gradient @xmath230 this expression is exactly @xmath231 , such that in a steady state the work done by the external forces equals the total heat flux into the thermostat .
we have enlarged the family of multibaker maps by a model for momentum , energy and entropy transport in viscous hydrodynamic flows .
although the macroscopic problem is of strongly dissipative nature , the proposed multibaker dynamics is area preserving ( hamiltonian ) .
this is to be contrasted with previous multibaker models of electric transport @xcite and thermoelectric cross effects @xcite where the inclusion of a reversible dissipation mechanism was necessary in order to simulate the effect of thermostating on the particle dynamics .
the form in which thermostating appears in the present model is via a heat - source term in the microscopic energy dynamics ( which was already present in the model of cross effects @xcite ) .
this term , however , does not give rise to phase - space contraction .
the model has the following basic features : + ( a ) the time evolution of the system can be interpreted as that of weakly interacting particles . the resulting `` multi - baker '' gas obeys the classical ideal - gas equation of state .
the particle and heat diffusion coefficient , and the kinematic viscosity are proportional to each other , as in the kinetic theory for classical ideal gases @xcite .
+ ( b ) the distribution of a macroscopic velocity does not enter the entropy explicitly .
rather the shear rate appears in the entropy balance via its influence on the temperature dynamics only .
+ ( c ) the connection to a thermodynamic description of transport is achieved by considering fields , which are coarse grained in regions of small spatial extension .
their properties are to be contrasted to those of fields characterizing the microscopic evolution .
+ ( d ) comparing these two levels of description allows us to identify all contributions to the _ local _ entropy balance , in full consistency with thermodynamics .
they apply to both stationary and transient states .
+ ( e ) a source term is implemented in the evolution equation of the microscopic kinetic energy .
it provides the possibility to implement local irreversible cooling of the system , i.e. , to extract heat such that states which are permanently warmed up by viscous heating can become stationary . in traditional thermodynamics
these terms are only present at the boundaries , and they vanish in the bulk . + ( f ) it is indicated how the velocity and the kinetic - energy dynamics can be implemented at the walls in order to achieve correspondence with different macroscopic boundary conditions .
+ ( g ) when source terms are present in the bulk , the local entropy balance of nonequilibrium thermodynamics is generalized by introducing at every location an instantaneous flux of entropy ( i.e. , of heat ) into a thermostat ( the entropy flux is then no longer the divergence of the entropy current ) .
the dynamics is in that case reminiscent of numerical algorithms based on gaussian thermostats .
+ ( h ) the global entropy balance was worked out , in order to demonstrate that the total heat flux into a thermostat is independent to a large extend of whether thermostating is applied in the bulk or at the walls .
the major interest of the present model lies in the light it sheds on the origin of viscous heating in deterministic models of transport .
it was pointed out how fractal structures emerge in multibaker models for a variety of physical settings of shear flow . in all cases
the structures arise from the mixing of regions with different local temperatures and flow velocities whose differences are exponentially proliferating to smaller and smaller scales for a nonequilibrium system .
one can explicitly follow this redistribution of the kinetic energy , until it reaches the scale of the domains used to define local thermodynamic averages .
motion at that scale has to be considered as contributing to the non - directional motion , hence leading to viscous heating .
it is _ only _ due to this coarse - graining mechanism that the macroscopic shear rate appears in the expression of the irreversible entropy production .
we would like to thank bob dorfman , christan gruber , and gary morris for illuminating discussions .
support from the hungarian science foundation ( otka grand no .
032423 ) is acknowledged .
evans and g.p .
morriss , _ statistical mechanics of nonequilibrium liquids _ ( academic press , london , 1990 ) ; w.g .
hoover , _ computational statistical mechanics _ ( elsevier , amsterdam , 1991 ) .
dorfman , _ an introduction to chaos in non - equilibrium statistical mechanics _
( cambridge univ . press , cambridge , 1999 ) .
dellago and h. posch , j. stat .
phys . * 88 * , 825 ( 1997 ) ; ch .
dellago , w.g .
hoover , and h. posch , phys .
e * 57 * , 4969 ( 1998 ) ; g. ayton , o.g .
jepps , and d.j .
evans , molecular physics * 96 * , 915 ( 1999 ) . | a consistent description of shear flow and the accompanied viscous heating as well the associated entropy balance is given in the framework of a deterministic dynamical system .
a laminar shear flow is modeled by a hamiltonian multibaker map which drives velocity and temperature fields . in an appropriate macroscopic limit one
recovers the navier - stokes and heat conduction equations along with the associated entropy balance .
this indicates that results of nonequilibrium thermodynamics can be described by means of an abstract , sufficiently chaotic and mixing dynamics .
a thermostating algorithm can also be incorporated into this framework . 2 |
considerable uncertainties remain as to what controls the apparent link between the activity of active galactic nuclei ( agn ) during the time that a black hole is fed and star formation .
the accretion disk surrounding a super - massive black hole ( smbh ) emits highly energetic radiation and particles , and can form powerful winds and/or collimated , relativistic jets ( e.g. , * ? ? ?
. the radiation and outflows might then affect the interstellar medium , triggering star formation which might be detectable as an excess of blue light from the central regions of galaxies .
alternatively , the activity of the agn may be sparked by specific events in the galaxy s past .
for example , there is morphological evidence that activity in radio galaxies might be triggered by mergers and galaxy interactions ( e.g. , * ? ? ? * ) , which in turn could contribute to central blue light through enhanced star formation . in either case
, we might expect an observable association between agn activity , for which in this work we use radio loudness as a tracer , with bluer stellar continuum in the central regions of the galaxy .
previous work has hinted at such a relationship .
for example , @xcite examined the central light of 30 radio galaxies as compared to 30 normal galaxies from the molongo reference catalogue and found excess blue light in the inner regions of the radio galaxies as compared with the control sample .
this central excess is unrelated to the conventional colour gradient of the broader distribution of stars .
@xcite studied colour variations within galaxies from the sloan digital sky survey ( sdss ; * ? ? ?
* ) dr7 , finding marginally steeper colour gradients in massive galaxies with nuclear activity and concluded that this is due to a higher fraction of young stars in their central regions . in the present paper
we expand the work of @xcite , by combining optical information from sdss with the first and nvss radio surveys @xcite to construct a local ( @xmath2 ) sample of powerful radio - loud agn host galaxies ( r - agn ) , and a control sample of ` normal ' early - type ellipticals .
we define @xmath10 as the ratio of sdss @xmath6 to @xmath12 band de vaucouleurs effective radii and compare its value between the samples .
the paper is structured as follows . in section [ sec : surveys ] we briefly describe the surveys used to construct our samples .
separation into the radio - loud and control samples is discussed in section [ sec : sampleselection ] .
section [ sec : analysis ] outlines the maximum likelihood analysis used to compare the average values of @xmath10 for the two samples . in section [ sec : discussion ] we present a discussion of our findings in the context of radio - agn fuelling mechanisms and the properties of the host galaxies . throughout
, we use a flat @xmath13cdm cosmology with @xmath14 and @xmath15 .
we adopt @xmath16 kms@xmath1mpc@xmath1 .
sdss is an imaging and spectroscopic survey covering @xmath1710,000 deg@xmath18 , primarily in the northern hemisphere .
its 7th data release ( sdss - dr7 ; * ? ? ?
* ) contains over 350 million entries , of which about one million are confirmed galaxies with spectroscopic follow up . in the photometric sample ,
flux densities are measured simultaneously in five broadband filters ( @xmath19 ) , with effective wavelengths @xmath20 , @xmath21 , @xmath22 , @xmath23 and @xmath24 @xcite .
all magnitudes are given on the ab@xmath25 system @xcite , and are determined using petrosian apertures ( see * ? ? ?
* ; * ? ? ?
* for a complete definition of the petrosian system ) .
in common with @xcite , we refer to sdss measured magnitudes as @xmath26 and @xmath27 due to the uncertainty in the absolute calibration of the sdss photometric system , which has been assessed at @xmath280.03mag @xcite . for
a complete guide to the photometric system , see @xcite , @xcite and @xcite .
we use the main galaxy survey spectroscopic sample ( mgs , * ? ? ?
the mgs algorithm selected extended sources brighter than @xmath29mag with petrosian half - light surface brightness @xmath30mag arcsec@xmath31 , providing @xmath17 90 galaxies deg@xmath31 @xcite .
candidate galaxies were then observed spectroscopically .
@xmath32 of the resulting redshifts have velocity errors @xmath3330kms@xmath1 .
the survey is unbiased , except for galaxies with close companions . around @xmath34 of galaxies satisfying the photometric target criteria
were not included for spectroscopic follow up due to a companion galaxy within the 55 minimum fibre separation , although some of these galaxies have been subsequently observed .
the mgs contains @xmath17680,000 spectroscopically confirmed galaxies .
the full target selection is described in @xcite .
we corrected for galactic extinction using the sdss ` reddening ' parameter , which was derived from maps of the infrared emission from dust across the sky in accordance with @xcite .
we applied @xmath35-corrections using the sdss - derived ` kcorr ' parameter as detailed in @xcite .
typical corrections @xmath36 were @xmath37 , @xmath38 , and @xmath39mag .
the first survey @xcite utilized the vla ( at 1.4ghz in b array ) to map the radio sky over @xmath179000deg@xmath18 in the northern hemisphere and in a @xmath40 wide strip along the celestial equator .
it has @xmath41deg@xmath18 overlap with sdss , contains @xmath42 sources deg@xmath31 at the 1mjy survey threshold and reaches an rms sensitivity of @xmath170.15mjybeam@xmath1 . in this configuration , the vla has a synthesized beam of 5.4 fwhm , providing accurate flux densities for small - scale radio structures , but underestimating the flux densities of sources extended to several arcminutes . at the 1mjy survey threshold ,
individual sources have 90% confidence astrometric errors @xmath43 .
we adopted a mean spectral index of @xmath44 ( where @xmath45 ) and obtained rest - frame 1.4ghz power densities by applying k - corrections assuming the usual form @xmath46 where @xmath47 is the luminosity distance .
the nrao vla sky survey ( nvss , * ? ? ?
* ) mapped the radio sky ( at 1.4ghz in d array ) north of @xmath48 declination .
the survey is complete down to a point source flux density of @xmath172.5mjy .
the synthesized beam of 45 is much larger than first , providing more accurate flux measurements for highly extended sources .
astrometric accuracy ranges from 1 for bright sources to 7 for the faintest detections .
the entire survey contains over 1.8 million unique sources brighter than 2.5mjy .
@xmath35-corrections are applied as in [ subsec : firstsurvey ] .
we initially derived a first - sdss sample of radio galaxies and a comparison control sample with no nearby detectable radio counterpart .
figure [ fig : idlsep ] shows the distribution of angular separations between sdss objects and their nearest first counterpart .
we find that at @xmath175 the distribution becomes dominated by random matches .
hence we adopt a match criterion of 2 angular separation to avoid such false matches .
our cross - matched catalogue contains @xmath49 galaxies and we denote this as our preliminary ` radio ' sample .
all unmatched sources are initial candidates for our control sample of radio - quiet galaxies .
we define our efficiency as the fraction of matches in our radio sample which are physically real ( including contamination from line of sight false matches , which can not be accounted for here ) . in order to evaluate the number of random false matches in our sample ,
we offset the ra and dec of the first sources by @xmath50 and re - matched to our set of @xmath51 sdss galaxies , selecting objects within a 2 radius .
we found on average @xmath52 random matches , corresponding to 0.3% contamination of the first - sdss match sample ( @xmath5399% efficiency , figure [ fig : idlsep ] , red dotted line ) . and
sdss is shown as a red dotted line .
this indicates the number of likely false matches .
the net distribution of ` real ' matches is shown by the blue dashed line , and our 2 cut indicates 99.9% efficiency of the matched sample . ] we also need to know the completeness of our matching criteria - the fraction of correct matches we recover . a factor affecting
this is the possible exclusion of large lobe - dominated radio sources @xcite .
both lobes will be included in first but may be excluded in the cross - matched subset for lobe - core distances @xmath532 , if the core is weak .
we estimated @xmath178@xmath54 of real matches between the sdss subset and first that are excluded from our radio sample , by assuming the radio lobes are outside a 2 radius from the optical core , but within a 5 radius ( see figure [ fig : idlsep ] ) .
our matching algorithm efficiency and completeness are similar to @xcite , who match first to sdss - dr6 positions within a 2 radius for sources with @xmath55 2 ( 95% efficiency and 98% completeness ) .
however , we could not account for sources with first lobes @xmath56 from the optical core or the population of first double - lobed sources with no detectable radio core .
@xcite cross - correlated all first sources with sdss , then identified potential double - lobed radio sources with undetected cores and accepted all matches with offsets @xmath57 .
] , estimating these contribute less than @xmath58 of all radio sources
* estimate @xmath59 of their radio - loud agn sample had no first detection ) .
@xcite estimated the first catalogue to be @xmath60 complete at 2mjy and @xmath61 complete down to the survey limit of 1mjy .
figure [ fig : radiopower_z ] shows integrated radio power against redshift for the preliminary radio sample ( 25,931 sources ) .
the solid black lines trace the 1mjy and 2mjy peak flux density thresholds , the dotted and dashed lines show cuts we applied in redshift and radio power ( see [ subsec : redshiftrange ] and [ sec : radioagn ] ) , and the red points are the resultant selection of radio sources .
it is noted that for sources that are not well - described by an elliptical gaussian model , the integrated flux density as derived by first may be an inaccurate measure of the true value .
such sources with radio powers corresponding to flux densities below the 1mjy threshold can be seen in figure [ fig : radiopower_z ] .
we note that although the two cuts improve the sample completeness ( all red points are above 1mjy ) , many of our sources are below the 2mjy line , and so our sample completeness can not be @xmath62 . combining this with our estimation of completeness in our selection criteria ( @xmath63 ) , we estimate the actual completeness of our preliminary radio sample to be @xmath64 for sources brighter than 1mjy .
the incompleteness is dominated by statistical effects rather than the @xmath58 effect of sources missing due to weak cores .
and the dotted line denotes the radio power cut imposed in [ sec : radioagn ] .
the red points are the remaining radio sample after these cuts are applied.,scaledwidth=45.0% ] to produce a preliminary control sample , we select sources from the sdss subset which fall within the same spatial region as first ( @xmath65 deg@xmath18 ) . we selected all objects from this subset which are not matched to a first source within 2 of their optical core , providing @xmath17625,664 radio - quiet galaxies .
the efficiency of this sample , based on figure [ fig : idlsep ] is @xmath66 ( i.e. , @xmath67 contamination by radio sources detectable at the first flux density limit ) and the sample completeness is @xmath68 . for first sources substantially larger than @xmath69 some flux density is resolved out , leading to an increase in the survey threshold for extended objects and flux density underestimates for larger objects @xcite .
our first - sdss radio sample is large due to the number of first candidate sources near the 1mjy survey limit .
its limitations are potential flux density underestimates , and incompleteness to extended radio sources , notably extended , double - lobed sources with no detectable radio core within several arcseconds of its optical counterpart .
to attempt to quantify these limitations , we created a complementary sample of radio - loud agn host galaxies using both nvss and first .
@xcite note that the 45 resolution of nvss is large enough that @xmath70 of all radio sources are contained within a single component , allowing for a higher sample completeness .
however , nvss is less deep than first and consequently the matched sample contains fewer galaxies .
the nvss - first - sdss matched sample therefore does not benefit from such good statistics .
we used the unified radio catalogue constructed by @xcite ( see sample ` c ' in their table 8) , selecting sources which are detected by both first and nvss , matched to within 25 .
this selection yields a radio flux density limited ( @xmath71mjy ) catalogue ( hereafter first - nvss ) containing 141,881 sources .
we cross - matched these to the @xmath72 mgs sources adopting the same 2 match criterion used in [ subsec : crosscorrelation ] .
our cross - matched catalogue contains 5,719 galaxies and we denote this as the preliminary ` comparison radio ' ( hereafter cr ) sample .
integrated flux densities as derived by @xcite are adopted in our analysis .
we estimated our matching criterion of 2 to be @xmath11 complete , and @xmath11 efficient ( 8 random matches were found when the first - nvss ra and dec were offset by @xmath50 ) .
the first - nvss catalogue used to create the cr sample is 99@xmath54 complete and matched with @xmath73 efficiency ( see table 2 of * ? ? ?
we therefore estimated the comparison radio sample to be @xmath11 complete and @xmath74 efficient ( see table [ tab : comeff ] ) .
5 of the 5,719 galaxies in the cr sample reside in the control sample derived in [ subsec : crosscorrelation ] and these were removed .
4,935 of the cr sample ( 86@xmath54 ) are also in the first - sdss sample . in the following sections ( [ subsec : redshiftrange]-[subsec : redshiftdist ] )
we discuss the secondary selection criteria applied to the first - sdss sample .
the cr sample follows the same path , and a summary of the final cr sample properties are presented alongside those of the final radio sample in [ subsec : crprop ] .
we imposed a redshift cut of @xmath75 0.18 to ensure that the galaxy spatial structure could be well examined . for @xmath76 ,
redshift is not a reliable distance indicator .
figure [ fig : lumredradio ] shows less than 2% of matched sources have redshift @xmath77 .
the mgs magnitude limit of @xmath7817.77 corresponds to l@xmath79whz@xmath1 at @xmath80 , and our sample should be complete to this optical luminosity density .
we restricted the sdss - derived redshift confidence ( zconfidence ) to be greater than @xmath60 , which cuts out 12% of objects from the preliminary radio sample and 10% of objects from the preliminary control sample .
after redshift selection , @xmath81 radio - quiet galaxies remained in the control sample , @xmath1770% of the parent mgs , whilst the radio sample contained @xmath82 galaxies .
the average error in redshift for both samples is @xmath83 and the redshift distributions of these two samples were indistinguishable at the 1% level on the basis of a two - tailed k - s test .
p@xmath84 luminosity bins .
we cut at @xmath2 , providing a high sample completeness ( @xmath85 ) down to p@xmath86whz@xmath1 . ]
@xcite suggested that typical radio - loud agn have powers above @xmath87whz@xmath1 at 1.4ghz and @xcite shows that p@xmath88whz@xmath1 separates the spiral starburst population from the agn - e / s0 population in local galaxy fields .
we therefore restrict the radio sample to galaxies harbouring radio sources with p@xmath89whz@xmath1 , which selects 5,119 galaxies ( 30% ) from the radio sample .
figure [ fig : lumredradio ] shows that this threshold in radio power is well matched to our redshift selection .
we name this set of 5,119 galaxies the ` radio - loud ' sample .
the mean redshift of the radio sample increased from @xmath90 to @xmath91 when the radio power cut was applied .
agn can be split on the basis of large - scale radio structure into fri type , where radio brightness decreases outwards from the centre and frii type , with edge - brightened lobes and hot spots @xcite .
friis are more luminous p@xmath92whz@xmath1 and are rare in a low - redshift sample such as ours ( figure [ fig : lumredradio ] ) , since this 178mhz power corresponds to p@xmath93whz@xmath1 for a typical spectral index of @xmath44 . in unified agn models ( e.g , *
* ) the appearance of the central black hole and associated continuum of an agn differ only in the viewing angle at which it is observed .
sources viewed face on ( type 1 ) show broad emission lines that are absent in those observed edge on and where the broad emission line region is obscured by a dusty torus ( type 2 ) .
type 1 agn are excluded from this study , as the optical continuum can be dominated by relativistically boosted non - thermal emission , which may overwhelm measurements of the host galaxy s properties .
the sdss spectral classification pipeline automatically flags and excludes quasars from the mgs , but we chose to verify its reliability and the relative numbers of type 1 agn remaining in our sample .
we followed the method outlined by @xcite to identify broad line emission .
h@xmath94 or h@xmath95 emission lines exceeding 1000kms@xmath1 ( fwhm ) with a @xmath96 and h@xmath94/h@xmath95 ew @xmath56 were classified as broad line .
galaxies with both broad h@xmath94 and h@xmath95 emission were classified as type 1 agn .
3 galaxies of the 5,119 radio - loud sample have broad h@xmath94 and h@xmath95 emission lines and were removed from the sample .
184 galaxies ( 4@xmath54 ) of the remaining 5,116 radio - loud sample have broad h@xmath94 emission , but do not have broad @xmath97 emission .
@xcite classified these objects as type ` 1.9 ' agn , which have substantial but not complete obscuration of the central continuum source .
@xcite determine that the contribution to the observed continuum is not significant in these sources , so we retain these within our sample .
the control sample contained 10 type 1 agn and 1835 type ` 1.9 ' agn .
we removed the type 1 agn to leave 499,408 galaxies within the control sample .
type 2 agn have narrow permitted and forbidden lines and their stellar continuum is often similar to normal starforming galaxies .
@xcite show a tight correlation between far infra - red luminosity ( indicative of star formation ) and p@xmath84 .
this will cause a level of contamination by star - forming galaxies if p@xmath98 is used as the sole tracer of galaxies hosting a radio - loud agn .
we should therefore remove the small subset of radio - loud galaxies in which the radio power arises from star formation ( sf ) and not from an agn .
agn separation from sf galaxies in the local universe can be achieved via optical emission line ratio diagnostics ( * ? ? ?
* herein bpt ) .
emission - line ratios probe the ionizing source : for agn , non - thermal continuum from the accretion disc around a black hole and in star - forming galaxies ( sfgs ) , photoionization via hot massive stars .
however , @xcite find no correlation between a galaxy being radio - loud and whether it is optically classified as an agn . in agreement with this
, we found no correlation0.06 or less between the radio - flux and optical emission - line flux ratios for our radio galaxy sample . ] between radio flux at 1.4ghz and the optical line ratios [ nii]/h@xmath94 or [ oiii]/h@xmath95 in our sample .
hence , a substantial fraction of radio - loud agn would not be selected using bpt diagnostics , and were we to apply them to the radio sample it would be biased towards radio galaxies with particularly strong optical emission lines .
we instead remove galaxies which are strongly identified as non - agn , i.e. star - forming . despite this method leaving a small fraction of star - forming galaxies which are faint in optical line emission , all radio - loud agn
, whether optically bright or otherwise , will remain in the sample .
this decrease in efficiency of the sample is preferential to a drastic decrease in completeness .
a similar problem was identified by @xcite , who defined a sample of radio - loud agn from the 2dfgrs catalogue .
they found approximately half of the sample have absorption spectra similar to those of inactive giant ellipticals , and therefore would be mostly missed by optical agn emission - line selection . in classifying galaxies as agn or star - forming
, we utilized the demarcation criterion of @xcite @xmath99}/\mathrm{h}\beta ) > 0.61/\{\mathrm{log}([\mathrm{nii}]/\mathrm{h}\alpha ) - 0.05\ } + 1.3 % \\\ ] ] plotted as the dashed line on figure [ fig : bpt_sf ] . )
indicates the demarcation given by @xcite between optical agn ( above the line ) and sf galaxies ( below the line ) .
766 galaxies plotted have emission line ratios with @xmath5 significance .
the density of these galaxies is shown by contours , 85% of the 766 galaxies lie above the line as optically classified agn .
the remaining 15% ( 118 galaxies ) lie below the curve ( red circles ) and we classified these as star - forming .
we also show the maximum uncertainty for a point on the kauffmann separator with s / n of 3 ( dotted lines ) .
69 objects detected at s / n@xmath100 ( black crosses ) lie between the kauffmann separator and the upper maximum uncertainty line . ]
figure [ fig : bpt_sf ] shows the standard line ratio diagnostics for galaxies from the radio - loud sample with all four emission lines catalogued in the sdss ( grey crosses , 1,069 objects of the 5,116 ) .
766 of these 1,069 galaxies have both optical emission - line ratios at s / n @xmath101 and their density on the bpt diagram is shown as contours .
118 ( 11% ) lie below the demarcation line and are marked with red circles ( 112 out of these 118 have all four emission - lines with s / n @xmath102 ) .
we removed these 118 optically selected sf galaxies from the radio - loud sample , to leave a sample of 4,998 predominantly radio - loud agn hosts .
we then estimated the residual contamination expected in the radio agn sample from star - forming galaxies by plotting the maximum uncertainty for a point on the kauffmann separator with s / n of 3 ( fig [ fig : bpt_sf ] , dotted lines ) .
69 objects detected at s / n@xmath100 ( black crosses ) lie between the kauffmann separator and the upper maximum uncertainty line .
therefore the contamination expected in the radio agn sample from star - forming galaxies is @xmath103 .
our radio sample contains predominantly fri galaxies , which are usually hosted by giant elliptical galaxies and on average have weak or no optical nuclear emission lines ( * ? ? ?
* and references therein ) . within our sample of radio - loud agn
, 79% of objects do not have optical emission line fluxes .
we estimate @xmath104 contamination of non - agn ( e.g. sfg ) if galaxies without sdss emission lines are similar to galaxies with bright line emission .
5 galaxies without emission line fluxes @xmath53 3@xmath105 lie below the demarcation line in figure [ fig : bpt_sf ] and are potentially star forming , but without reliable line information we retain these within our sample .
as discussed by @xcite , a potential shortfall of spectral classification of emission - line radio - loud agn is that emission - line agn activity is often accompanied by star formation ( e.g. , * ? ? ?
. this star formation will give rise to radio emission , even if the agn itself is radio quiet .
for these sources , the optical spectrum could still be dominated by a ( radio - quiet ) agn leading to classification as an emission - line radio - loud agn .
@xcite ( hereafter m10 ) derive a population of sources matched from all nvss galaxies in the unified radio catalog @xcite , the sdss - mgs and iras data .
the matched sources are divided into star - forming , composite and agn using standard bpt diagnostics .
star formation rates are derived via broad - band spectral fitting to the nuv - nir sdss photometry , and the _ average _ fractional star formation / agn contribution to the radio power is estimated ( see their table 2 ) .
they find that in 203 composite galaxies , 81.3@xmath54 of the total radio power is due to star formation .
the variation of this fraction with radio power is not specified .
following m10 , we defined 206 composite galaxies in our own radio sample ( 27@xmath54 of the emission - line galaxies confirmed at s / n @xmath101 ) , using the diagnostics of @xcite and @xcite .
if the average fractional contribution is independent of total radio luminosity , then we estimated an upper limit of @xmath106 of our radio sample may have radio power boosted by star formation but possess a radio - quiet agn .
however , @xmath107 of the composite sample defined by mori have log[p@xmath108(whz@xmath1 ) ] @xmath109 .
@xcite find sf galaxies tend to have median log[p@xmath108(whz@xmath1 ) ] = 22.13 , whereas agn have a median log[p@xmath108(whz@xmath1 ) ] = 23.04 .
therefore , our luminosity cut will have significantly reduced the numbers of galaxies where star - formation is the principal contributer to the total radio power , and we expect far less than @xmath110 contamination from radio - quiet , optically - loud agn .
we also can not account for the population of ` composite ' radio - loud host galaxies with ongoing star formation that have been lost from our sample through this selection technique .
@xcite estimate that @xmath111 of local radio sources may have a starforming spectrum but have radio flux densities dominated by a radio - loud agn . as low - redshift radio - loud agn
are hosted predominantly by elliptical galaxies , we attempted to constrain the control sample to contain only early - type galaxies .
the colour distributions of galaxies have been shown to be highly bimodal @xcite .
figure [ fig : colourcolour ] shows a colour - colour plot in @xmath112 * -@xmath7 * against @xmath8 * -@xmath112 * as explored by @xcite who define an optimal colour separator98% completeness and @xmath11383% reliability . ] between early and late types of @xmath114mag .
early - types ( e , s0 , sa ) populate the upper right region , and late types ( sb , sc , irr ) occupy the lower left
. contours ) .
the dashed line is @xmath114mag from @xcite and defines a morphological colour separator .
80% of the galaxy zoo - mgs sample lie above the line .
the @xmath12 , @xmath115 and @xmath6 magnitudes are derived from sdss columns : petrocounts @xmath116 reddening @xmath116 kcorr . ] in order to test the completeness of this demarcation , we used the 62,190 galaxies visually classified as ellipticals in galaxy zoo data @xcite .
we cross - matched these to the control sample , selecting objects within 1 radius .
34,606 matches were found , and 27,569 of these lie above @xmath8 * -@xmath7*@xmath532.22mag .
therefore , this demarcation gives @xmath117 reliability ( fig [ fig : colourcolour ] , contours ) , suggesting @xmath118 of the early - type galaxy population may be excluded through applying this criterion .
147,275 visually classified ` spiral ' galaxies from the galaxy zoo data are also present in our control sample .
22% of these lie above @xmath8 * -@xmath7*@xmath532.22 , i.e. we expect @xmath119 contamination in the early - type sample . of the 499,408 galaxies in the control sample ,
@xmath120 were classified as early - types , using the @xcite colour selection , to leave 199,391 early - type galaxies in the control sample , morphologically selected in colour space at @xmath121 completeness and @xmath122 efficiency based on galaxy zoo visual analysis .
we explore the 4,998 radio - loud agn selected galaxies in colour space .
figure [ fig : radiocc ] shows the radio - loud agn are predominately early types , with @xmath123 lying above @xmath8 * -@xmath7*@xmath532.22 , i.e. the radio - loud agn sources are predominantly hosted by ellipticals as defined in colour space .
the distribution in colour space is similar to that of the control sample , despite the agn colours possibly influencing the light . in agreement with this result , @xcite found radio - selected agn have a high incidence of being hosted by early - type galaxies .
the 118 galaxies we flagged as star forming and removed from the radio - loud sample in [ sec : sf ] are shown as crosses in figure [ fig : radiocc ] , and 98% of these sit below the line .
we select the radio - loud agn above the colour cut , to give 3,516 ` early - type ' radio - loud agn hosts . mag , 4 equally spaced levels , max density contour = 400/bin ) , and sf galaxies as defined in [ sec : sf ] ( crosses ) the dashed line is @xmath114 from @xcite and defines a morphological colour separator . ] at this stage we re - examined the redshift distributions of the samples , using the ks test .
the probability of the control sample and the radio - loud agn being drawn from the same parent sample in redshift was @xmath124 : the distributions in redshift differ ( figure [ fig : idl_zscale_lum ] ) .
evolutionary differences in host galaxy properties should be small over this entire redshift range , however we chose to re - sample the control sample redshift distribution to match that of the radio - loud agn sample .
we also examined the intrinsic @xmath6 band luminosity of the two samples .
figure [ fig : idl_lum_zscale ] shows the radio sample is globally brighter in @xmath6 than the control sample .
it has been well established that radio - loud agn are hosted preferentially in the brightest and most massive elliptical galaxies @xcite . since we wanted to test whether radio - loud agn harbour an excess of blue emission in their centres , we chose to re - sample the control sample to have the same @xmath6 band optical properties so as to avoid any possible correlation of colour gradient with galaxy luminosity .
we generated a matched control sample by selecting the 10 galaxies from the whole sample that lie closest to each radio - loud agn in magnitude - redshift space .
the final control sample contains 35,160 galaxies , whose magnitude and redshift distributions are shown as dotted lines in figure [ fig : idl_scaled ] .
this final selection should allow the distribution of observables in the radio and control samples to be compared directly .
the criteria applied to the ` radio ' sample in [ subsec : redshiftrange]-[subsec : redshiftdist ] were also applied to the 5,719 objects in the cr sample ( see [ subsec : crossnvssfirst ] ) .
we imposed a redshift cut of @xmath2 and restricted the redshift confidence parameter given by sdss to @xmath74 , leaving 3,727 sources .
we examined the cr sample radio powers based on nvss and first .
figure [ fig : lumfn ] shows the nvss and first luminosities for the 3,727 cr sources ( grey crosses ) .
those with p@xmath125whz@xmath1 are highlighted by blue squares ( 50@xmath54 ) .
we restricted the cr sample to galaxies harbouring radio sources with p@xmath125whz@xmath1 , which selects 1,847 ( 50@xmath54 ) of the sample .
of those 1,847 sources , 251 have first derived powers @xmath126whz@xmath1 , and were therefore excluded in the radio sample selection in [ sec : radioagn ] .
whz@xmath1 are highlighted by blue squares ( 1,847 ) .
of those , 251 sources with p@xmath127whz@xmath1 are shown as red crosses .
these are the sources excluded in the radio sample s luminosity cut ( [ sec : radioagn ] ) . ]
we removed 2 type 1 agn and 48 star - forming galaxies confirmed at s / n @xmath128 from the cr sample , leaving 1,797 radio - loud narrow - line agn hosts . 4@xmath54 of the sample are classified as type ` 1.9 ' agn , with substaintial but not full obscuration of the central source .
we applied the strateva et al .
colour cut ( @xmath8 * -@xmath7*@xmath532.22mag ) detailed in
[ subsec : colour ] to select 1,295 early - type galaxies from the cr sample .
the control sample was cross - matched with the first - nvss catalogue to within 2and the 5 galaxies found were removed to ensure the comparison control sample is comprised solely of ` radio - quiet ' agn elliptical hosts .
199,386 galaxies remain in the cr - control sample .
we selected the 10 galaxies from the control sample that lie closest to each cr radio - loud agn in magnitude - redshift space .
the final cr - control sample contains 12,950 sources , matched to the cr sample in @xmath6-band magnitude and redshift distributions .
* radio - loud ` early type ' agn ( r - agn ) - 3,516 sources * after matching first to the mgs ( @xmath3mag , @xmath2 ) within 2 , we classified as radio - loud agn hosts with p@xmath89whz@xmath1 .
type 1 agn exhibiting both broad h@xmath94 and h@xmath95 emission lines were removed from the sample .
optical emission line ratios ( if confirmed at 3@xmath105 ) were used to remove star - forming galaxies using the demarcation of @xcite .
we then created a subset with @xmath12-@xmath129mag i.e. ` early - types ' to remain consistent with the colour cut applied to the control sample .
* control sample - 35,160 sources * we removed all galaxies from sdss - mgs ( @xmath3mag , @xmath2 ) with first and nvss counterparts .
type 1 agn exhibiting both broad h@xmath94 and h@xmath95 emission lines were removed from the sample .
galaxies were constrained in colour to bias towards early - type morphology where a comparison with results of the galaxy zoo program suggests we have @xmath130 completeness and @xmath131 efficiency .
the resultant sample comprises radio - quiet , ` normal ' , predominantly elliptical galaxies .
we then selected a subsample to match the redshift and optical @xmath6-band magnitude distributions of the r - agn sample , thus removing any redshift bias and potential correlation of colour gradient with galaxy luminosity .
* comparison radio ( cr ) - 1,295 sources * we cross - matched the first - nvss catalogue derived by @xcite with the mgs ( @xmath3mag , @xmath2 ) to within 2 , and classified radio - loud agn hosts as those sources exhibiting p@xmath132whz@xmath1 .
type 1 agn exhibiting both broad h@xmath94 and h@xmath95 emission lines were removed from the sample .
optical emission line ratios ( if confirmed at 3@xmath105 ) were used to remove star - forming galaxies using the demarcation of @xcite .
we then created a subset with @[email protected] i.e. ` early - types ' to remain consistent with the colour cut applied to the control sample .
* cr - control sample - 12,950 sources * all selection criteria as above for the control sample , we also removed 5 galaxies within the control sample that were also present in the cr sample .
we then selected a subsample to match the redshift and optical @xmath6-band magnitude distributions of the cr sample .
we therefore have two radio - loud agn samples , both with a comparison control sample .
the cr sample is derived using both radio catalogues , and is a brighter radio sample ( the nvss flux density completeness limit is @xmath134mjy ) for comparison with our deeper ( @xmath135mjy ) , larger but more incomplete r - agn sample derived solely from the first catalogue .
the surface brightness profile of an early - type galaxy can be described by the srsic equation @xmath136}\ ] ] where @xmath137 is the effective radius ( scale length ) and @xmath138 is the corresponding effective surface brightness .
@xmath139 is chosen so @xmath137 contains half the light in the galaxy .
n @xmath140 4 for bright ellipticals decreasing to n @xmath140 2 as luminosity decreases @xcite .
sdss chooses to fix @xmath141 , fitting the ` de vaucouleurs profile ' , truncating the profile beyond 7@xmath137 .
the model fitted by sdss has an arbitrary axis ratio and position angle and is convolved with a double - gaussian representation of the psf .
the fitting then yields , among other properties ; the effective radius , @xmath137 and error , @xmath142 . in general , the error for the @xmath8 model is greater than other bands .
it is noted that sdss s fitting algorithm generates some weak discretization of model parameters , especially in @xmath143 and @xmath144 .
objects with scale lengths lying in discrete bands in @xmath137 tend to have poorer goodness of fit .
this is not included in the error estimates , @xmath142 , which were derived from count statistics on the image . in our work with the @xmath137 values , we consider @xmath142 rather than dev_l ( the likelihood of the model fit ) due to our reluctance to remove data with poor de vaucouleurs @xmath8 band fitting
. a poor goodness of fit to the devaucouleurs profile may be indicative of a cuspy @xmath8 band and removing such objects could bias results .
sdss denotes the fracdev parameter as the fraction of the total flux contributed by the de vaucouleurs component in a linear combination of a de vaucouleurs and an exponential model to find the best fit .
it is noted that the likelihood values in the @xmath6-band are intrinsically poor ( all samples have mean values @xmath145 ) , but the likelihood ratios still pick out reliable best - fit parameters .
the fracdev parameter ( @xmath146 hereafter ) is correlated with the srsic index ; n=1 corresponds to @xmath146=0 , n=4 corresponds to @xmath146=1 ( * ? ? ? * and references therein ) .
following @xcite , galaxies with @xmath147 ( @xmath148n@xmath149 ) are labeled de / ex " galaxies and galaxies with @xmath146@xmath150 ( n@xmath151 ) are labeled de " galaxies .
we chose not to remove objects with @xmath146@xmath152 , but instead created subsets of each sample with @xmath146@xmath153 for comparison .
we investigate the colour structure of agn host galaxies principally through the distribution of @xmath10 , which we define as @xmath154 , which is the ratio of @xmath7 to @xmath8 de vaucouleurs effective radii .
this should avoid any intrinsic scale differences in @xmath7 and @xmath8 between individual galaxies .
the distributions of @xmath10 for the r - agn and control sample are shown in figure [ fig : dist1 ] .
both samples have a few ( @xmath155 ) strong outliers ( @xmath156 ) which significantly disturb the mean .
the medians of the r - agn and control distributions are 0.837 and 0.793 respectively . using a two - tailed kolmogorov - smirnov test we find the probability of the control and the r - agn sample being drawn from the same distribution to be small ( @xmath124 ) .
similarly , the cr sample s @xmath10 distribution also shows a higher median compared to the cr - control sample ( 0.851 and 0.797 respectively ) .
the ks test also finds the probability of the cr and cr - control sample being drawn from the same distribution in @xmath10 is vanishingly small ( @xmath124 ) .
for the radio - loud agn ( r - agn ) sample ( solid ) and the control sample ( red dashed).,scaledwidth=45.0% ] the fractional errors in individual values of @xmath144 are relatively large , which makes proving an intrinsic difference between the radio - loud agn samples and their respective control sample distributions challenging .
we used a maximum likelihood analysis to compare the average values of @xmath10 for the agn and control samples , taking into account measurement errors according to @xcite and @xcite , in order to quantify the difference between the distributions of @xmath10 . under the assumption that both the agn and control samples have normally - distributed intrinsic values of @xmath10 ,
then the population mean , @xmath157 , and intrinsic dispersion , @xmath105 , of either can be determined from individual values @xmath158 by minimizing the function @xmath159 } .
\label{eq : smin}\ ] ] figure [ fig : rsig ] shows the results of minimization of eq [ eq : smin ] across @xmath157 and @xmath105 parameter space , for the two samples .
the 90% and 99% joint - confidence contours for @xmath157 and @xmath105 are given by s@xmath160s for @xmath161s@xmath162 and 9.21 respectively .
table [ tab : smin ] shows the best fit @xmath163 planes for both distributions . a clear separation between radio - loud `
early type ' agn and ` normal ' control galaxies can be seen for both the r - agn and cr samples , and we conclude the two are not drawn from the same population at @xmath164 confidence .
we applied the cut @xmath165 in the @xmath6-band , and re - examine the results for both radio - agn samples .
this cut will ensure the reliability of the de vaucouleur s scale lengths used to derive @xmath10 , and this subset is used as a comparison to the full samples to verify whether galaxies with shapes that differ significantly from the de vaucouleurs profile perturb our main result .
table [ tab : smin ] also shows the fraction of each sample with ` good ' de vaucouleurs @xmath6-band fits and their @xmath157 values ( see figure [ fig : rsigf ] ) .
although the values do not significantly differ from the full sample , the confidence contours now overlap in the smaller , but more complete , cr sample .
however , the r - agn and control samples are still seen to come from distinct populations .
l ccccc sample & @xmath157 & @xmath105 & @xmath165 & @xmath166 & @xmath167 + [ 0.5ex ] r - agn & [email protected] & [email protected] & 53@xmath54 & [email protected] & [email protected] + control & [email protected] & [email protected] & 49@xmath54 & [email protected] & [email protected] + cr & [email protected] & [email protected] & 44@xmath54 & [email protected] & [email protected] + cr - control & [email protected] & [email protected] & 49@xmath54 & [email protected] & [email protected] + [ 0.5ex ]
our results confirm that radio - loud agn are hosted by brighter , bigger galaxies on average ( see figure [ fig : idl_scaled ] ) .
our technique has shown the presence of an agn is associated with an increase of the scale size of a galaxy in red light relative to blue light .
this suggests that radio galaxies have a bluer central bulge and/or more diffusely distributed red light compared to their radio - quiet counterparts . with the analysis
so far , we can not tell whether this difference is from agn driven or diffuse star - like light within the bulge , or more distributed red light .
the cr sample is also significantly different in @xmath157 relative to the cr - control sample , so that our r - agn sample s lower completeness in extended sources does not significantly affect the result .
the cr sample dispersion is smaller than that of the r - agn sample , whereas all four control samples have values of @xmath105 that are similar .
we applied a flux cut of @xmath169mjy to the r - agn sample ( 61@xmath54 ) , and found @xmath170 and @xmath171 .
the increased scatter about @xmath10 is caused by fainter radio sources , which tend to be more heterogenous , and/or have underestimates of errors on @xmath137 values in more distant objects .
the subset of cr galaxies constrained to have good @xmath6-band de vaucouleurs fits ( @xmath165 ) does not contain enough radio galaxies to show strongly a significant separation from the cr - control sample ( see figure [ fig : rsigf](b ) ) .
we present discussion predominantly on the r - agn and control sample , which display a clear separation in both figure [ fig : rsig](a ) and [ fig : rsigf](a ) .
we concluded the two are not drawn from the same population at @xmath164 confidence .
this sample is incomplete to radio sources with lobes @xmath172 and the population of galaxies with a weak radio agn core but powerful lobes .
however it is large due to the depth of the first survey .
the cr results for all figures in the discussion are similar to the results presented for the r - agn sample .
we find no significant deviation between the two in any subsequent parameter examined , except in @xmath105 , as discussed above . for the cr sample .
we found no correlation ( @xmath11 certainty ) between the two.,scaledwidth=45.0% ] ( * ? ? ? * hereafter mkm99 ) used a complementary ratio ] to demonstrate the ubiquity of inner blue components in a sample of only 30 radio galaxies and 30 control galaxies , and argued that the blue light is due to star formation associated with the presence of a radio source ( e.g. * ? ? ? * ) .
figure 4 of mkm99 ( power at 408mhz vs log [ @xmath173 ) shows a trend for more powerful radio galaxies to have steeper colour gradients as indicated by the ratio of scale lengths .
in contrast to this , when considering only the brightest ( p@xmath174whz@xmath1 ) radio sources in the cr sample ( for which the powers are more accurate ) we find no correlation between radio power and @xmath10 .
we also find no correlation over the full range of radio powers and @xmath10 as shown in figure [ fig : lumr ] : the spearman s rank correlation result is @xmath175 or less between p@xmath84 and @xmath10 .
we infer that a high value of @xmath10 ( and possible blue central bulge ) is not associated with the power of the radio source .
there are several possible scenarios to be considered to explain the larger @xmath10 in the r - agn sample .
central point - like @xmath12 light may be indicative of a blue agn .
we investigated the nature of the r - agn excess @xmath10 , using the goodness of the devaucouleurs fit .
the sdss devaucouleurs fitting procedure returns the likelihood ( between @xmath176 and @xmath177 ) associated with the model from the @xmath178 fit . if the @xmath12-band goodness of fit for the r - agn sample approaches 1 for low @xmath10 and 0 for high @xmath10 then this is indicative of an agn / core component ( as the agn point - like component would perturb a devaucouleurs fit ) driving a blue excess and high @xmath10 .
if no structure is seen in @xmath10 vs devaucouleurs @xmath12-band goodness of fit in either sample , then any blue excess is more likely driven by some sort of diffuse starlight in the central bulge . the @xmath12-band data for both the r - agn and control samples are generally well fit by a devaucouleurs profile ( the r - agn / control median likelihood values are 0.86 and 0.83 respectively ) , and there is no trend for the @xmath12-band or @xmath6-band fits to be worse at large @xmath10 in the r - agn than the control sample .
we can conclude that the high @xmath10 in the r - agn sample is not due to any point - like agn component .
it is generally agreed that agn feedback regulates the supply of cold gas for star formation by supplying energy to the ism and igm .
some radio - loud star - forming agn show a connection between the suppression of star formation and the strength of the radio jets , by heating and expelling the surrounding gas @xcite .
@xcite looked at a sample of low redshift sdss early - type galaxies for which late - time star formation is being quenched .
they found that molecular gas disappears less than 100myr after the onset of accretion onto the central black hole .
these galaxies were not associated with radio jets , but show that low - luminosity agn episodes are sufficient to suppress residual star - formation in early - type galaxies .
-band and @xmath6-band scale radii .
there is a higher fraction of blue cores in the r - agn sample as compared to the control sample at @xmath179kpc .
this plot does not demonstrate the correlation between the red and blue scale radii ( see figure [ fig : rdev ] ) .
the measurement errors are larger in the @xmath12 band which may account for the higher variability in the residuals.,scaledwidth=45.0% ] the residual fractional number density distribution of scale lengths are shown in figure [ fig : residual ] , for the control sample subtracted from the r - agn sample .
there is a clear excess of blue cores ( @xmath180kpc ) in the r - agn sample as compared to the control sample .
there are also more small @xmath181 in the r - agn sample , ranging from @xmath182kpc .
this plot does not demonstrate the correlation between the red and blue emission , which is shown in figure [ fig : rdev ] , but it does demonstrate that the presence of a radio agn seems not to have suppressed star - formation in the central regions of its host galaxy . and @xmath181 for the r - agn ( solid ) and control sample ( red dashed ) .
the dotted - dashed line is the locus @xmath183 , galaxies above this line become bluer inward .
contours show the fractional number densities of each sample with levels at 0.004 , 0.008 and 0.011 .
there is a larger fraction of the r - agn sample above the line , possibly denoting a tendency to become bluer inward more often than the control sample.,scaledwidth=45.0% ] figure [ fig : rdev ] plots @xmath184 against @xmath181 for the r - agn sample ( black ) and the control sample ( red dashed ) .
the dotted - dashed line denotes the locus @xmath185 as defined by the control sample ( see table [ tab : smin ] ) . at @xmath186 ,
the surface brightness in @xmath12 increases more rapidly toward the centre of a galaxy than in @xmath6 as compared to the mean of the control sample , i.e. half the blue light from the galaxy is contained in a smaller region than half the red light , implying the galaxy becomes bluer inward .
the distribution of number densities in figure [ fig : rdev ] shows radio galaxies appear to become bluer inwards more often than the control galaxies ( 66% of the r - agn sample are above the locus @xmath187 , compared to 60% of the control sample ) .
there are fewer blue smaller cores in the control sample as compared with the r - agn sample ( see also figure [ fig : residual ] ) , implying there may be a blue excess near the agn core for some of the radio galaxies .
however , the higher @xmath10 values of the r - agn sample also seems to be driven by a higher scale length in @xmath6 , seen in the difference between the inner two contour levels of each sample ( see also [ subsubsec : redenv ] ) , so we can not definitively conclude that the increased @xmath10 in the r - agn sample is due to star formation within the central few kpc .
star - formation within the central few kpc would contradict feedback models which predict the suppression of star formation near an agn .
@xcite found that radio sources in massive hosts are re - triggered more frequently than their less massive counterparts , suggesting that the onset of an agn quiescent phase is due to fuel depletion .
agn activity is therefore promoted by an increase in gas in the centres of the galaxies , which may imply a link between the agn radio phase and star formation in the bulge ( @xmath188kpc ) of the host galaxy .
it has been suggested that agn activity and a major episode of star formation in radio - loud galaxies is triggered by the accretion of gas during major mergers and/or tidal interactions .
however agn activity is initiated later in the merger event than the starburst ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
our spectroscopic selection rules out major merger events ( see [ subsubsec : mgs ] ) , but residual star formation may still be present on global scales .
observationally , there is a strong link between agn and starbursts ( * ? ? ? * and references therein ) .
@xcite found powerful optical agn ( as classified by the strength of the [ o ] emission @xmath189}}>10 ^ 7 l_{\odot}$ ] ) predominantly reside in ` young bulges ' .
recent star formation can provide up to @xmath190 of the optical / uv continuum in radio galaxies at low and intermediate z e.g. ( e.g. , * ? ? ?
* ; * ? ? ?
we found a higher fraction of r - agn galaxies have @xmath191kpc as compared to the control sample ( see figure [ fig : residual ] ) , perhaps indicative of star formation in the outskirts of the bulge , fuelled by gas expelled from the central regions by the agn .
an alternative interpretation of a larger @xmath10 is that the r - agn sample has more distributed red light .
radio - loud ` early - type ' galaxies predominantly reside in elliptical galaxies , which are well known to be redder than late - types ( e.g. , * ? ? ?
diffuse red emission would cause a larger devaucouleurs scale radius in the @xmath6 band and an increased @xmath10 .
figure [ fig : rdev ] suggests the increased @xmath10 in the r - agn sample may also be attributed to more diffuse red light ; between @xmath192kpc , @xmath181 is on average higher for the radio population .
this is also seen in the lower panel of figure [ fig : residual ] .
our results show a difference in the distributions of red and blue light in r - agn and normal ` early - type ' galaxy populations , though a high value of @xmath10 is not a property of every active galaxy .
we found the higher @xmath10 in the r - agn sample is contributed to by a higher fraction of radio galaxies harbouring small blue cores but also an increase in the numbers of radio galaxies with more diffuse red light as compared to the control sample . given that star formation proceeds over a longer timescale than radio activity , this disfavours the idea that all galaxies undergo short bursts of radio activity , but rather implies that a subset have the predisposition to become radio - loud .
we cross - matched low - redshift ( @xmath2 ) data from the sdss mgs ( @xmath3mag ) and first to within 2 , deriving a radio sample of galaxies at @xmath11 efficiency and @xmath193 completeness .
radio luminosities were in the range @xmath0whz@xmath1 .
type 1 agn were removed from the sample ( identified via @xmath194 and @xmath97 broad line characteristics ) , 4@xmath54 of the sample are classified as type ` 1.9 ' agn , with substantial but not full obscuration of the central source .
contamination from sfgs ( identified via optical emission line ratios ) in the radio sample is expected at @xmath103 .
a control sample was defined from sdss sources with no match to a first source within 2 of their optical core , providing a sample of @xmath11 efficiency and @xmath68 completeness . at 80% reliability , the demarcation @xmath4mag selected ` early - type ' galaxies in both samples .
samples were matched in @xmath6-band magnitude and redshift distributions and final sample sizes were 3,516 radio - loud agn galaxies ( r - agn ) and 35,160 control galaxies .
we also created a complementary flux - limited sample through cross - matching with nvss ( the cr sample ) .
the same cuts were applied to derive a radio - loud ` early - type ' agn sample , except we used the more all - encompassing nvss flux estimates to cut in radio luminosity , thereby retaining the populations of galaxies with weak / no core radio emission but bright , extended radio lobes .
this sample had higher completeness for comparison with the r - agn sample .
a control sample was defined from the sdss sources with no match to a nvss - first source within 2 of their optical cores .
we further considered only sources where fracdev(f ) @xmath195 to restrict all four samples to having good @xmath6-band devaucouleurs fits .
we investigated the colour structure of agn host galaxies through @xmath10 , the ratio of @xmath6 to @xmath12 de vaucouleurs effective radii and used maximum likelihood analysis to quantify the degree of difference in the distribution of @xmath10 between samples .
we concluded the radio ( r - agn / cr ) samples are not drawn from the same population as their radio - faint control samples at @xmath196 confidence : the presence of an agn increases the scale size of a galaxy in red light relative to blue light , on average .
our result does not appear to be driven by the presence of blue agn in the radio - loud samples since the goodness of the de vaucouleurs fits does not become worse as @xmath10 increases .
we found no structure in @xmath10 vs @xmath12-band goodness of fit in either radio sample .
we found an excess of blue cores in radio - loud galaxies as compared to radio - quiet , ` early - type ' galaxies , implying the increased @xmath10 may be due to star formation in the central few kpc , in contrast with feedback models which predict the suppression of star formation near an agn .
spectroscopic selection of our samples rules out major merger events , as starbursts can be triggered by the accretion of gas / tidal interactions .
we also found radio agn hosts to have larger red scale lengths in relation to their blue light and note this to be a contributing factor in an increased @xmath10 .
we can not definitively discern whether a small blue core or larger distribution of red light is the driving factor in this result .
given the longer timescale for star formation than radio activity , our results imply a subset of galaxies have the predisposition to become radio - loud , rather than all galaxies undergoing bursts of radio activity at some stage in their lifetimes .
em gratefully acknowledges support from the uk science and technology facilities council and thanks luke davies and james price for helpful comments . we thank the anonymous referee for helpful and insightful comments that have resulted in an improved analysis . in undertaking this research
, we made extensive use of the topcat software @xcite .
funding for the sdss 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 .
this research makes use of the nvss and first radio surveys , carried out using the national radio astronomy observatory very large array : nrao is operated by associated universities inc .
, under cooperative agreement with the national science foundation .
w. , fragile c. , croft s. , de vries w. , anninos p. , murray s. , 2004 , in storchi - bergmann t. , ho l. c. , schmitt h. r. , eds , the interplay among black holes , stars and ism in galactic nuclei vol .
222 of iau symposium , jet - induced star formation : good news from big , bad black holes .
pp 485488 d. m. , 1989 , in j. hunt & b. battrick ed . , two topics in x - ray astronomy , volume 1 : x ray binaries .
volume 2 : agn and the x ray background vol . 296 of esa special publication , population properties of agn in the soft x - ray band . | we construct a sample of 3,516 radio - loud host galaxies of active galactic nuclei ( agn ) from the optical sloan digital sky survey ( sdss ) and faint images of the radio sky at twenty cm ( first ) .
these have 1.4ghz luminosities in the range @xmath0whz@xmath1 , span redshifts @xmath2 , are brighter than @xmath3mag and are constrained to ` early - type ' morphology in colour space ( @xmath4mag ) . optical emission line ratios ( at @xmath5 ) are used to remove type 1 agn and star - forming galaxies from the radio sample using bpt diagnostics . for comparison
, we select a sample of 35,160 radio - quiet galaxies with the same @xmath6-band magnitude - redshift distribution as the radio sample .
we also create comparison radio and control samples derived by adding the nrao vla sky survey ( nvss ) to quantify the effect of completeness on our results .
we investigate the effective radii of the surface brightness profiles in the sdss @xmath7 and @xmath8 bands in order to quantify any excess of blue colour in the inner region of radio galaxies .
we define a ratio @xmath9 and use maximum likelihood analysis to compare the average value of @xmath10 and its intrinsic dispersion between both samples .
@xmath10 is larger for the radio - loud agn sample as compared to its control counterpart , and we conclude that the two samples are not drawn from the same population at @xmath11 significance .
given that star formation proceeds over a longer time than radio activity , the difference suggests that a subset of galaxies has the predisposition to become radio loud .
we discuss host galaxy features that cause the presence of a radio - loud agn to increase the scale size of a galaxy in red relative to blue light , including excess central blue emission , point - like blue emission from the agn itself , and/or diffuse red emission .
we favour an explanation that arises from the stellar rather than the agn light .
[ firstpage ] |
one of eddington s famous discoveries is that the isotropic distribution function ( df ) of a spherical stellar system can be calculated from the density using an abel transform pair @xcite : @xmath6 here , @xmath7 is the binding energy per unit mass , and @xmath8 is the relative potential . however , galaxy halos produced in cosmological simulations are held up by anisotropic velocity dispersions ( see e.g. , * ? ? ?
the recovery of an anisotropic df for a spherical system is a more difficult problem @xcite .
now , the steady state df can depend not only on @xmath9 but also on the magnitude of the specific angular momentum @xmath10 through jeans theorem . here
@xmath11 is the two - dimensional velocity component projected on the tangential plane . at least in the comparison of observational data with models ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , a widely used _
ansatz _ for the form of df is @xmath12 where @xmath13 is a function of the binding energy alone .
it is easy to show that the models generated by this df exhibit the property that @xmath14 here , @xmath15 and @xmath16 are the radial and the tangential velocity second moments . the general inversion formula for the unknown function @xmath13 ( * ?
* ; * ? ? ?
* ; * ? ? ?
* see also appendix [ sec : con ] ) is no more difficult than eddington s formula ( eq . [ eq : eddington ] ) .
we note that it is common practice to characterize the velocity anisotropy of a spherically symmetric stellar system by the parameter @xmath17 defined in equation ( [ eq : beta ] ) , which is sometimes referred to as the anisotropy parameter @xcite . if @xmath18 , the velocity dispersion ellipsoid is a prolate spheroid with the major axis along the radial direction ( radially anisotropic ) , while it is an oblate spheroid with the tangential plane being the plane of symmetry ( tangentially anisotropic ) if @xmath19 .
in particular , @xmath20 and @xmath21 indicate that every star is in a radial or circular orbit , respectively . on the other hand , systems with isotropic velocity dispersions
have @xmath22 . in general , the anisotropy parameter varies radially , but for the system with a df of the form of equation ( [ eq : ansatz ] ) , it is constant everywhere .
despite the attractiveness of a simple inversion of the df in equation ( [ eq : ansatz ] ) , the assumption of uniform anisotropy is not always desirable .
for example , it is useful to be able to build models that are isotropic in the center and radially anisotropic in the outer parts , as seems to be the case for dark matter halos @xcite .
dfs that are tangentially anisotropic in the outer parts are helpful for understanding the substructure and the satellite galaxy populations in dark matter halos .
in this paper we study algorithms for building galaxies with varying anisotropy , using sums of dfs of the form of equation([eq : ansatz ] ) .
[ sec : gis ] introduces a new family of cusped halo models , the generalized isochrones , and [ sec : gp ] examines the generalized plummer models .
every member of the family has analytic potential - density pairs , velocity dispersions , and dfs .
specific examples are provided with the cosmologically important @xmath0 , @xmath2 , and @xmath1 density cusps .
[ sec : app ] discusses a specific application of our theoretical models to the problem of estimating the mass of a host galaxy from the motion of its satellites .
finally , [ sec : conc ] concludes with a brief comparison between our methods of producing varying - anisotropy dfs and other suggestions in the literature ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the isochrone sphere has the potential - density pair @xcite @xmath23 where @xmath24 is the total mass of the system , @xmath25 , and @xmath26 is a constant defining the scale length .
like the @xcite model , the density is finite and well behaved at the center . unlike the plummer model
, the density falls off as @xmath27 at large radii , which is a better fit to elliptical galaxies than the @xmath4 fall - off of the plummer model .
there are many available studies regarding its dfs ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , some of which are essentially numerical in nature . inspired by this ,
let us consider the potential - density pair @xmath28 where @xmath29 here , @xmath30 is a new parameter used for the generalization of the isochrone sphere ( eq . [ eq : iso ] ) .
the condition that the density decreases monotonically with increasing radius restricts @xmath31 , with the limit ( @xmath32 ) being the cored classical isochrone sphere .
in fact , all the halo models with @xmath33 are cusped . in particular
, the density behaves as @xmath34 as @xmath35 . on the other hand , we find that @xmath36 where @xmath37 as @xmath38
. the density profiles of some members of this family are shown in figure [ fig : geniso ] .
we note that the case @xmath39 , for which @xmath40 , is actually recognized as the @xcite model , albeit with a slightly unusual identification for the scale length ( @xmath41 ) .
models with @xmath42 and @xmath43 possess the cosmological cusps suggested by @xcite and @xcite . the enclosed mass and the circular speed
are found to be @xmath44^{-2 } \,;\qquad v_\mathrm c^2=\frac{gm_r}r=\frac{gmr^p}{s^{p-1}(b+s)^2}.\ ] ] the potential @xmath45 of the generalized isochrone sphere is simple enough that the inverse , @xmath46 , can be easily found ( with @xmath47 without any loss of generality ) : @xmath48^{1/p}}\psi.\ ] ] using this , in principle , one can construct infinitely many different expressions for @xmath49 , each of which can be inverted to yield the df by means of integral transforms @xcite .
however , for most cases this procedure is analytically intractable .
rather , it is often more productive to look for a specific form of @xmath49 that is associated with a plausible _ ansatz _ for the df , whose inversion reduces to elementary integrals .
one such possible ( @xmath50,@xmath8)-split with an analytically tractable inversion is given by @xmath51 we note that a df of the form of equation ( [ eq : ansatz ] ) can be inverted from an ( @xmath50,@xmath8)-split that is a monomial of @xmath50 , with its power index linearly related to @xmath17 . by having the ( @xmath50,@xmath8)-split given by a linear combination of two monomials of @xmath50 , the inversion of equation ( [ eq : expone ] )
can be achieved by simply extending the known procedure , while the system is no longer restricted to have uniform velocity anisotropy .
inspired by this , let us suppose that the df is of the form @xmath52 where @xmath53 and @xmath54 are functions of @xmath9 to be determined .
then , the density is found to be @xmath55 where @xmath56 is the gamma function . by comparing this to equation ( [ eq : expone ] ) ,
a possible df may be found by inverting the integral equations @xmath57 the inversion procedure is essentially identical to the one that leads to equations ( [ eq : disint ] ) and ( [ eq : dishalf ] ) .
after this inversion , we find that @xmath53 and @xmath54 are @xmath58 both of which are always nonnegative for all accessible @xmath9-values ( @xmath59 ) . while we have given the result in complete generality in terms of hypergeometric functions @xmath60 , the dfs reduce to entirely elementary and analytic functions if @xmath42 , 1 , or 2 . in fact , if @xmath42 , both hypergeometric functions are just unity , whereas if @xmath61 is an integer , both reduce to elementary functions .
if @xmath61 is a half integer other than @xmath62 , only @xmath63 reduces to an elementary function . of particular interest
is the case in which @xmath42 , for which the model possesses a central cusp like @xmath64 , as found by cosmological simulations @xcite .
the corresponding df is very simple , namely , @xmath65 if @xmath39 ( the hernquist model with a rescaled scale length ) , @xmath66 + \frac1{4\pi^3l}\frac{e^2(3 - 2e)}{(1-e)^2},\ ] ] or if @xmath32 ( the classical isochrone sphere ) ; @xmath67 \nonumber\\&+ & \frac3{2^{15/2}\pi^3(1-e)^4 } \left[(16e^2 - 12e+3)\,\frac{15\arcsin\!\sqrt e}{\sqrt{1-e } } -\left(45 - 150e+144e^2 - 208e^3 + 64e^4\right)\sqrt e\right].\end{aligned}\ ] ] both of these dfs are no more complicated than the isotropic dfs deduced by @xcite and @xcite , respectively . to find the velocity dispersion of these models , we exploit the splitting @xmath68 , where @xmath69 then , @xmath70 where @xmath71 the anisotropy parameter @xmath72 \label{eq : gicvi}\ ] ] is now no longer constant
. figure [ fig : anisoc ] shows the behavior of @xmath17 for some of these models .
the curves can be understood on examination of equation ( [ eq : gicvi ] ) .
this reveals that whereas @xmath73 at @xmath74 for any @xmath61 , @xmath75 if @xmath76 , or @xmath77 if @xmath78 . at the critical value @xmath39 , we find that the anisotropy parameter monotonically decreases from @xmath79 at @xmath74 to @xmath80 as @xmath38 .
hence , the dfs discussed in this section build the generalized isochrone spheres with either decreasing @xmath17 in the inner region and increasing @xmath17 in the outer region ( @xmath78 ) , or decreasing @xmath17 throughout ( @xmath81 ) .
the particular case of the latter also includes the hernquist model ( @xmath39 ) and the classical isochrone sphere ( @xmath32 ) .
the simplest case is again @xmath42 , which leads to @xmath82 then we find that the virial theorem holds locally for this case : @xmath83 and the anisotropy parameter @xmath84 varies from @xmath85 at @xmath74 , to @xmath86 at @xmath87 , and back to @xmath88 as @xmath38 .
this model in fact corresponds to one of the generalized hypervirial models studied by @xcite .
that is , @xmath89^{-1}=(2 + 2r^{1/2}+r)^{-1}$ ] so that @xmath90 and @xmath91 in their parameters .
we note that both @xmath92 and @xmath93 in fact reduce to the incomplete beta functions , which may be useful for the purpose of numerical calculations . furthermore , if @xmath94 is an integer , both @xmath92 and @xmath93 reduce to elementary functions of @xmath8 , and so do the velocity second moments .
in particular , if @xmath39 [ for which @xmath95 , @xmath96 @xmath97
next , let us consider the one - parameter family of potential - density pairs , @xmath98 which was originally introduced by @xcite .
this includes the @xcite model ( @xmath39 ) and the @xcite model ( @xmath32 ) as particular cases .
@xcite recently found that this family can be constructed from power - law dfs of the form of @xmath99 .
here we construct slightly more complicated dfs that build this family of models .
we first find every df of the form of equation ( [ eq : ansatz ] ) , and then use them to build models with a more general variation of anisotropy . with @xmath100 , we find that @xmath101 and that @xmath102 for @xmath103
. then , equation ( [ eq : disint ] ) reduces to @xmath104 where @xmath105 is the integer floor of @xmath106 , and @xmath107 .
if @xmath108 , it is at least formally possible to derive the series expression of the df for arbitrary @xmath61 and @xmath17 from the integral form , namely , @xmath109 which becomes the generalized hypergeometric series if @xmath61 is a rational number . on the other hand ,
if @xmath110 , we find that @xmath111 so that the corresponding df is unphysical . hence , for the potential - density pair of equation ( [ eq : pdpair ] ) , the constant anisotropy df is physical only if @xmath112 .
this means that the hypervirial models of @xcite , which have an anisotropy parameter @xmath73 , have the maximally radially biased velocity dispersions for a given @xmath61 and a constant @xmath17 .
although there exist models with @xmath17 locally exceeding @xmath113 for the generalized plummer sphere of given @xmath61 ( see e.g. , * ? ? ?
* ) , @xmath17 at the center can not be greater than @xmath113 for physically valid dfs .
this is in agreement with the cusp slope
central anisotropy theorem derived by @xcite .
the corresponding velocity dispersions can be found either from equation ( [ eq : vint ] ) or by solving jeans equation .
the results can be generally expressed using the incomplete beta functions .
the equivalent results expressed in terms of hypergeometric functions are also found in equation ( 9 ) of @xcite . for particular values of @xmath61 or @xmath17 ,
the df of equation ( [ eq : gpdf ] ) ( or equivalently , eq . [ eq : gpdfs ] ) reduces to a simpler form .
for example , if @xmath79 , the df is ( see eq .
[ eq : dishalf ] ) @xmath114,\ ] ] where the positive definiteness of the df implies that @xmath115 .
similarly , for @xmath116 , the df is @xmath117,\ ] ] which is nonnegative for @xmath118 if @xmath119 . on the other hand , the df of the hernquist model with constant @xmath17 reduces to the hypergeometric function ( see e.g. , * ? ? ?
* ) @xmath120 which is nonnegative everywhere if @xmath121 , whereas that for the constant-@xmath17 plummer model reduces to @xmath122 which is physical if @xmath123 .
since equation ( [ eq : gpd ] ) is valid for any constant @xmath17 with common @xmath124 , it is , in general , also possible to express the density profile of the generalized plummer model by @xmath125 where the parameters @xmath126 are normalized weights .
consequently , the superposition of the constant anisotropy dfs , @xmath127 , still builds the generalized plummer sphere . here
each @xmath128 is given by equation ( [ eq : gpdf ] ) ( or equivalently , eq . [ eq : gpdfs ] ) in which @xmath17 is replaced by @xmath129 .
the velocity second moments for these models are found to be @xmath130 where @xmath131 and therefore the anisotropy parameter @xmath17 is no longer constant . for the particular case in which the sum only contains two terms ( such that @xmath132 , @xmath133 , and @xmath134 for all other @xmath135 values )
, the anisotropy parameter is @xmath136 which is a decreasing function of @xmath50 , provided that @xmath137 and @xmath138 .
if we assume @xmath139 and @xmath137 , the limiting values are found to be @xmath140 on the other hand , @xmath141 or @xmath142 the simplest example of the dfs of this kind is obtained when we choose @xmath143 and @xmath144 : @xmath145 , \label{eq : sdf } \ ] ] which is physical if @xmath146 .
the behavior of the anisotropy parameter @xmath17 for the model given by dfs of equation ( [ eq : sdf ] ) is shown in figure [ fig : aniso ] .
again we note that equation ( [ eq : hypg ] ) in general reduces to the incomplete beta function while , for this case , it further reduces to an expression involving only elementary functions of @xmath8 if @xmath147 is an integer , and so and so do @xmath17 and the velocity second moments .
as an application of the preceding theory , let us consider the problem of estimating the mass of the milky way from the radial velocities of its satellites
. there are 27 distant globular clusters and dwarf galaxies at galactocentric distance @xmath50 greater than 20 kpc ( see e.g. , tables 2 and 3 of * ? ? ?
these range from arp 2 at @xmath148 out to leo i at @xmath149 .
the median distance of the satellite population is @xmath150 .
the number density of the satellites falls off roughly like @xmath151 .
this dataset is often used to estimate the mass of the milky way @xcite , as the h1 rotation curve can not be traced beyond @xmath520 kpc .
the satellites are therefore the only available probe of the gravitational field of the distant halo .
all the satellites have accurate radial velocities , but proper motions are available only for a handful of the closest objects . of course , the heliocentric radial velocity and the galactocentric radial velocity are basically the same ( once the motion of the sun has been taken out ) , as the galactocentric distance of the sun is much smaller than those of the satellites .
the kinematic data therefore constrains the mean kinetic energy in the radial direction @xmath152 . again using the data in tables 2 and 3 of @xcite
, we find that @xmath153 . to model the milky way halo
, we assume that its potential and density take the form of either the generalized isochrone or the generalized plummer models .
furthermore , we assume that the number density of satellites shadows the total density ( e.g. , * ? ? ?
* ; * ? ? ?
this means that the satellites are drawn from dfs of the form given by either equation ( [ eq : diffiso ] ) or equation ( [ eq : sdf ] ) .
we set the scale length so that the half - mass radius of the assumed density profile corresponds to the observed median radius of the population .
the mass of the milky way @xmath24 enters into the potential and the df .
it can therefore be constrained by requiring that the mean kinetic energy in the radial direction be consistent with the data , viz , @xmath154 the parameter @xmath61 in the dfs controls the variation of the velocity anisotropy with radius .
hence , equation ( [ eq : crux ] ) defines a line in the plane @xmath155 , which is the _ mass - anisotropy degeneracy curve _ for the model .
figure [ fig : massaniso ] , inferred from numerical integrations , shows such curves computed for a number of the generalized isochrones and the generalized plummer models . in this problem ,
the mean kinetic energy in the radial direction is fixed .
models with tangential anisotropy at large radii therefore lead to a higher inferred mass for the milky way galaxy .
the relative location of the curves in figure [ fig : massaniso ] can be understood by examining the behavior of the anisotropy parameter for the models .
more importantly , as the velocity anisotropy changes , the inferred mass of the milky way ranges from @xmath156 to @xmath157 .
now , the velocity anisotropy of the satellite galaxy population is unknown , so the mass - anisotropy degeneracy _ by itself _ enforces a factor of @xmath53 uncertainty in the mass of the milky way .
this fundamental limitation can only be overcome by proper - motion data .
of course , there are other uncertainties ; for example , the half - mass radius of the satellite population used to set the scale length in these calculations is also not well known ( note that the locations of the satellites are likely to be biased toward their apogalacticon ) and has an uncertainty of @xmath520% .
this carries over into an additional @xmath520% uncertainty in the inferred mass .
this is small in comparison with the mass - anisotropy degeneracy , but not insignificantly so .
this paper provides a number of flexible and simple galaxy models extended from two classical cored profiles : the isochrone sphere and plummer model .
we also derive analytic distribution functions ( dfs ) with varying velocity anisotropy for all the models including new dfs for the classical isochrone sphere and the hernquist model .
dfs of the form @xmath158 build a spherical system with a constant anisotropy parameter @xmath17 , where @xmath9 is the binding energy and @xmath159 is the specific angular momentum . the assumption of constant anisotropy is not enough to provide realistic models , as both observational data and numerical simulations suggest that dark halos and their constituents ( such as satellites ) have velocity anisotropies that vary with radius ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
here we have explored models whose dfs are superpositions of two or more such terms , namely , @xmath160 this can be done in two ways .
first , as illustrated by the generalized isochrones in [ sec : gis ] , we can sometimes find a splitting of the density in the form @xmath161 and apply the inversion separately to each component .
for the generalized isochrones , this provides models with either decreasing anisotropy parameter @xmath17 in the inner region and increasing @xmath17 in the outer parts , or decreasing @xmath17 throughout .
second , as illustrated by the generalized plummer models in
[ sec : gp ] , we can always add together weighted sums of constant anisotropy dfs , each of which individually reproduces the required density .
this appears to provide models with decreasing anisotropy parameter @xmath17 throughout . at large radii ,
the df is dominated by the component of smallest @xmath17 , and the models can be either radially or tangentially anisotropic in the outer parts , according to choice .
for both cases , it appears that the anisotropy parameter always decreases on beginning to move outward from the very center .
there have been a number of previous algorithms suggested for building anisotropic dfs with varying anisotropy .
for example , the osipkov - merritt algorithm @xcite builds models that are isotropic in the inner parts and tend to extreme radial anisotropy ( @xmath162 ) in the outer parts .
@xcite provided a clever algorithm for building separable dfs of the form @xmath163 where @xmath164 is the angular momentum of a circular orbit with energy @xmath9 .
the analytic examples provided by @xcite all have increasing radial anisotropy in the outer parts .
his method can be used to generate tangentially anisotropic models , but usually at the cost of a numerical inversion .
the methods developed in this paper therefore provide a complement to the existing inversions .
tangential anisotropy often arises in accreted populations , such as satellite galaxies . as an application of our models
, we have shown that the mass - anisotropy degeneracy by itself provides a factor @xmath53 uncertainty in the mass of the milky way galaxy as deduced from the kinematics of its satellites .
the density profiles of models studied in this paper typically fall off faster than @xmath165 and slower than @xmath166 at large radii , which are in good agreement with both observations of spheroidal components of galaxies and results from numerical simulations . at small radii
, they are typically cusped , with cusp indices in the range suggested by numerical simulations . in particular
, there are members with density cusps like @xmath167 , @xmath168 , and @xmath169 , which have been suggested as important on cosmogonic grounds @xcite , although observational evidence seems to favor the presence of a constant - density core @xcite
. however , even if real galaxies are cored , the core would be such a small fraction of the halo that it is not practically important for studies of the whole .
our families of new models therefore should find widespread application in the modeling of galaxies and dark halos .
we thank c. hunter , who made a number of interesting comments on the various incarnations of this paper .
an , j. , & evans , n. w. 2005a , , 444 , 45 ( astro - ph/0508419 ) an , j. h. , & evans , n. w. 2005b , , submitted ( astro - ph/0511686 ) baes , m. , & dejonghe , h. 2002 , , 393 , 485 bertin , g. , leeuwin , f. , pegoraro , f. , & rubini , f. 1997 , , 321 , 703 binney , j. , & tremaine , s. 1987 , galactic dynamics ( princeton nj : princeton univ . press ) cuddeford , p. 1991
, , 253 , 414 de blok , w. j. g. , mcgaugh , s. s. , bosma , a. , & rubin , v. c. 2001 , , 552 , l23 dejonghe , h. 1987 , , 224 , 13 donato , f. , gentile , g. , & salucci , p. 2004
, , 353 , l17 eddington , a. s. 1916 , , 76 , 572 eggen , o. j. , lynden - bell , d. , & sandage , a. r. 1962 , , 136 , 748 evans , n. w. , & an , j. 2005 , , 360 , 492 ( astro - ph/0501091 ) evans , n. w. & an , j. h. 2006 , , submitted ( astro - ph/0511687 ) evans , n. w. , & collett j. l. 1997 , , 480 , l103 evans , n. w. , de zeeuw , p. t. , & lynden - bell , d. 1990 , , 244 , 111 gerhard , o. e. 1991 , , 250 , 812 hansen , s. h. , & moore b. 2006 , new a , 11 , 333 ( astro - ph/0411473 ) hnon , m. 1959 , ann . dastrophys . , 22 , 126 hnon , m. 1960 , ann .
dastrophys . , 23 , 668 hernquist , l. 1990 , , 356 , 359 hunter , c. , & qian , e. 1993 , , 262 , 401 kleyna , j. t. , wilkinson , m. i. , gilmore , g. , & evans , n. w. 2003 , , 588 , l21 kochanek , c. s. 1996 , , 457 , 228 lynden - bell , d. 1962 , , 123 , 447 merritt , d. 1985 , , 90 , 1027 moore , b. , governato , f. , quinn , t. , stadel , j. , & lake , g. 1998 , , 499 , l5 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1995 , , 275 , 720 osipkov , l. p. 1979 , pisma astron .
, 5 , 77 ( english trans . in soviet astron .
lett . , 5 , 42 ) palunas , p. , & williams , t. b. 2000 , , 120 , 2884 plummer , h. c. 1911 , , 71 , 460 treu , t. , & koopmans , l. 2004 , , 611 , 738 tyson , j. , a. , kochanski , g , p. , & dellantonio , i. p. , 1998
, , 498 , 107 van der marel , r. p. 1994
, , 270 , 271 van der marel , r. p. , magorrian , j. , carlberg , r. g. , yee , h. k. c. , & ellingson , e. 2000 , , 119 , 2038 veltmann , .-
k. 1979 , , 56 , 976 ( english trans . in , 23 , 551 )
white , s. d. m. 1985 , , 294 , l99 wilkinson , m. i. , & evans , n. w. 1999 , , 310 , 645 wilkinson , m. i. , kleyna , j. t. , evans , n. w. , gilmore , g. , irwin , m. j. , & grebel , e. k. 2004 , , 611 , 738
in this appendix we summarize some results regarding constant - anisotropy dfs used in the main body of the paper . for the df given in equation ( [ eq : ansatz ] ) ,
direct integration over velocity space gives @xmath170 the formula for inversion for the unknown function @xmath13 is given by @xcite @xmath171 , \label{eq : disint}\end{aligned}\ ] ] where @xmath172 is expressed as a function of @xmath8 , and @xmath105 and @xmath173 are the integer floor and the fractional part of @xmath106 , respectively .
this includes equation ( [ eq : eddington ] ) as a particular case of @xmath22 ( @xmath174 & @xmath175 ) . if @xmath17 is a half - integer constant ( i.e. , @xmath176 @xmath62 , @xmath177 , @xmath178 , and so on ) , the above inversion simplifies to @xmath179 that is , this only involves differentiations in the process @xcite .
in addition , the expression for the differential energy distribution ( ded ) is found to be @xmath180 @xmath181 where @xmath182 , and @xmath183 and @xmath184 are the dirac delta ` function ' and the heaviside unit step function , respectively . in particular , if @xmath79 , the last integral does not explicitly involve the potential , and so the ded may be found simply as @xmath185 .
it is also straightforward to find the integral for the second velocity moments ; @xmath186 in fact , the velocity dispersion of any constant anisotropy models may be found from @xmath187 without explicitly evaluating the df @xcite . on differentiating equation ( [ eq : diffy ] ) with respect to @xmath8 and noting that the right - hand side is the same as equation ( [ eq : den ] ) , that is , @xmath188
, it follows that @xmath189 we note that equation ( [ eq : vint ] ) can also be derived from the integral solution of jeans equation with a constant @xmath17 ( see eq .
9 of * ? ? ?
this also indicates that equation ( [ eq : vint ] ) is actually a general result for a constant @xmath17 model , regardless of the specific form of the df that generates the model . as an example , let us think of the dfs that build the generalized isochrone sphere with a constant anisotropy parameter .
first , the df and the ded for the generalized isochrone sphere with @xmath79 can be found from equation ( [ eq : dishalf ] ) ; @xmath190^{1/p } } \,;\qquad \frac{dm}{de}= \frac{\left[(1-e)^p - e^p\right]^{1/p}}{2(1-e)^{2p}}\ g(e)\ ] ] @xmath191 + ( 1-e)^pe^{p-1}\left[(p+1)(p+2)+12e^2 - 6(p+2)e\right ] + 6(1-e)^{2p+1}. \ ] ] here , because @xmath59 , the nonnegativity of the df is satisfied only if @xmath115 .
in addition , the integration involved in the inversion ( eq . [ eq : disint ] ) is also analytically tractable if @xmath73 .
that is , for the density profile of the generalized isochrone sphere , we find that @xmath192.\ ] ] this leads to the df of the form of equation ( [ eq : ansatz ] ) @xmath193,\qquad\end{aligned}\ ] ] which build a constant anisotropy model with @xmath73 .
if @xmath39 ( corresponding to the hernquist model ) and @xmath79 , the df and the ded further reduce to a particularly simple form ; @xmath194 which was originally found by @xcite for the @xcite model . for @xmath32 and @xmath22 , the well - known expression for the df of the isotropic isochrone sphere ( see e.g. , * ? ? ?
* ; * ? ? ?
* ) is recovered ; @xmath195.\ ] ] finally , let us think of the ded of the generalized plummer models corresponding to the df given in equation ( [ eq : gpdf ] ) .
for this case , equation ( [ eq : ded ] ) reduces to a similar integral as equation ( [ eq : gpdf ] ) @xmath196 like the df , equation ( [ eq : dedgp ] ) reduces to a closed form function of @xmath9 for particular values of @xmath61 or @xmath17 . among them are for @xmath79 @xmath197e^{p-1 } , \label{eq : dedpp}\ ] ] for @xmath116 , @xmath198e^{p-1 } { \ensuremath{\mbox{}_2f_1\!\left(1,\frac{1}{p}+1;\frac{4}{p}+2;1-e^p\right)}},\qquad \label{eq : dedpn}\end{aligned}\ ] ] and for @xmath39 , @xmath199 this suggests that @xmath200 diverges as @xmath201 for @xmath78 , that it vanishes for @xmath76 , and that it has a finite limiting value of @xmath202 for @xmath39 .
we believe that this is because the behavior of @xmath200 near @xmath203 is dominated by the stars at large radii and the @xmath27 fall - off marks a critical point @xcite . | this paper provides two families of flexible and simple galaxy models .
many representatives of these families possess important cosmological cusps , with the density behaving like @xmath0 , @xmath1 , or @xmath2 at small radii .
the density falls off between @xmath3 and @xmath4 at large radii .
we provide analytic and anisotropic distribution functions for all the models . unlike many existing methods ,
our algorithm can yield tangentially anisotropic velocity dispersions in the outer parts , and so is useful for modeling populations of satellite galaxies and substructure in host galaxy halos . as an application , we demonstrate the degeneracy between mass and anisotropy for the satellite galaxy population of the milky way .
this can introduce a factor of @xmath53 uncertainty in the mass of the milky way as inferred from the kinematics of the satellite population . |
very recently , the d collaboration and the cdf collaboration at the fermilab tevatron reported their updated measurements of the mass difference between @xmath7 and @xmath8 mesons ( @xmath9 ) .
the new bounds on the mass difference are @xcite : @xmath10 it was the first time that both the lower bound and the upper bound for the @xmath0 mixing are presented .
especially the cdf result has reached an accuracy of about 1% .
the new results are important for the precision test of the standard model ( sm ) , especially for the determination of the unitary triangle . moreover ,
if the sm predictions are consistent with the above results , these data will put severe constraints on the flavor structure of the possible new physics models beyond the sm . in the literature , there have already been many discussions about the implications of the new measurements . in ref .
@xcite , the authors carried out model - independent analysis of the constraints on extensions of the sm .
there are also model - dependent calculations in some new physics models beyond the sm @xcite . in this paper
, we further investigate the effects of the minimal supersymmetric standard model with r - parity violation ( mssm - rpv ) @xcite on the neutral meson mixing .
r - parity is a discrete symmetry defined by @xmath11 , where b is the baryon number , l is the lepton number and s is the spin of the particle . in a supersymmetric extension of the sm ,
all the particles in the sm have @xmath12 , while all the superpartners have @xmath13 .
r - parity conservation is imposed in the minimal supersymmetric standard model ( mssm ) to keep proton stable . however , this requirement is not necessary for a fundamental theory , and one can always introduce lepton number or baryon number violating terms in the lagrangian .
for the neutral meson mixing , the relevant terms in the lagrangian are @xmath14 these interaction terms can induce @xmath0 mixing through tree level sneutrino exchange , and thus , probably , the relevant combination of couplings @xmath15 will be severely constrained by the recent measurements@xcite .
similarly , the terms shown in eq .
( [ largrangian ] ) also can induce the @xmath2 and @xmath3 mixing , which were discussed in ref .
@xcite at the leading - order , and the constraints on the relevant combinations of couplings through comparing with data were given by @xcite @xmath16 where @xmath17 , @xmath18 is the mass of the sneutrino of the @xmath19-th generation .
however , authors of ref .
@xcite did not include the contributions of the sm , since they believed that both theoretical and experimental results involved considerable uncertainties at that time .
recently , both sm theoretical predictions and experimental results has been improved significantly @xcite .
thus , besides investigating of @xmath20-@xmath21 mixing , it is also worthwhile to reinvestigate the constraints on the combinations of r - parity violating ( rpv ) couplings with the updated data on @xmath22-@xmath23 and @xmath24-@xmath25 mixing including the sm contributions .
moreover , in general , the next - to - leading - order ( nlo ) qcd corrections are significant , so we also calculate the nlo qcd effects on the above neutral meson mixing in this paper .
in addition to the above mass differences , the r - parity violating couplings can also contribute to the cp asymmetries in the meson decay processes , when the couplings are complex .
so the combinations of rpv couplings discussed above can affect observables related to time - dependent cp violation in processes such as @xmath26 and @xmath27 . in this paper
, we also consider the constraints on the relevant combinations of mssm - rpv couplings from these cp violation observables .
we organize our paper as following .
section [ sec : basic ] is a brief summary of the formalism for the calculation of the neutral meson mixing , cp asymmetry in b physics and the results in the sm . in section [ sec : nlo ] , we calculate the contributions from the mssm - rpv at leading - order and next - to - leading - order in qcd . in section [ sec
: num ] , we present our numerical results and discussions .
in order to make our paper self - contained , we first illustrate the basic formalism for the calculation of the @xmath28-@xmath29 and @xmath24-@xmath25 mass differences @xmath30 , the time - dependent cp asymmetry in @xmath31 decay , @xmath5 and the time - dependent cp asymmetry in @xmath7 decay , @xmath32 .
we start from the @xmath33 effective hamiltonian @xcite @xmath34 the relevant operators for our concerning are @xmath35 where @xmath36 for @xmath31 or @xmath7 meson mixing , respectively .
similar expressions for @xmath3 mixing can be obtained by replacing @xmath37 with @xmath38 and setting @xmath39 . with this effective hamiltonian , the @xmath40 mass difference can be expressed as @xmath41 @xmath42 , time - dependent cp asymmetry in @xmath43 decays , can be expressed as @xmath44 where @xmath45 , and @xmath46 . in the framework of sm ,
@xmath47 with @xmath48 in the sm , only @xmath49 is nonzero , which has been calculated to the nlo in qcd @xcite and is given by @xmath50,\end{aligned}\ ] ] where @xmath51 @xcite , @xmath52 in the naive dimensional regularization scheme ( ndr ) , @xmath53 and @xmath54 is the inami - lim function @xcite . the scale @xmath55 is usually taken to be @xmath56 in @xmath57 physics .
the expression for the @xmath3 mass difference is a little different , which is given by @xmath58 the nonzero nlo wilson coefficient in the sm is @xmath59 \left ( \alpha_s^{(3)}(\mu ) \right)^{-2/9 } \left [ 1 + \frac{\alpha_s^{(3)}(\mu)}{4\pi } j_3 \right],\end{aligned}\ ] ] where @xmath60 , @xmath61 in the ndr scheme and the parameters @xmath62 are @xmath63 , @xmath64 , @xmath65 @xcite , respectively .
@xmath66 functions can be found in ref.@xcite .
the matrix elements of the operators @xmath67 ( i=0,1,2 ) between two hadronic states can be obtained by using the vacuum insertion approximation ( via ) and partial axial current conservation ( pcac ) .
we refer the readers to ref .
@xcite for details .
the results are @xmath68 where @xmath69 and @xmath70 is the mass and decay constant of the @xmath7 mesons , respectively .
clearly , the expressions for @xmath31 and @xmath24 mesons can be obtained by simple substitutions .
@xmath71 is the non - perturbative parameters and can be calculated with lattice method @xcite .
we list their values to be used in our numerical calculations in table [ bpara ] : .non - perturbative parameters used in our numerical calculations . [ cols="^,^,^,^",options="header " , ]
in the mssm - rpv , there are tree - level contributions to the @xmath72 and @xmath3 mixing through sneutrino exchange .
the tree diagrams contribute through the operator @xmath73 , and the corresponding wilson coefficient is @xmath74 where @xmath75 for @xmath7 , @xmath31 and @xmath24 mesons , respectively . we assume universal sneutrino masses for simplicity , i.e. , @xmath76 , @xmath77 . at nlo in qcd , both @xmath73 and @xmath78 contribute due to color exchange . to calculate the nlo wilson coefficients
, we match the full theory onto the effective theory at the susy scale , and then run the coefficients down to the hadronic scales using the renormalization group equations ( rge ) . in our calculations , we use dimensional regularization in @xmath79 dimensions to regulate , and use @xmath80 scheme to renormalize the ultraviolet ( uv ) divergences .
we keep the heaviest quark mass @xmath81 ( @xmath82 for @xmath57 mesons and @xmath83 for @xmath84 mesons ) and set all other quark masses to be zero in the wilson coefficients .
however , we will keep the lighter quark mass @xmath85 in the intermediate stages of our calculation in order to regulate the infrared ( ir ) divergences .
we also set all external momenta to zero , since the coefficients should not depend on them .
the nlo diagrams in the full theory are shown in fig . [ full ] in the full theory ,
all the uv divergences should be removed by the renormalization of the quark wave functions and the coupling constants .
the renormalized amplitude in the full theory is @xmath86
\braket{q_1}_{\text{tree } } \right .
\nonumber \\ & \qquad \left . - \frac{\alpha_s}{4\pi } \frac{1}{n }
\frac { m_{\tilde{\nu}}^2 \ln\frac{m_q^2}{m_q^2 } + m_q^2 \ln\frac{m_q^2}{m_{\tilde{\nu}}^2 } } { m_{\tilde{\nu}}^2-m_q^2 } \braket{q_2}_{\text{tree } } \right\},\end{aligned}\ ] ] where @xmath87 , @xmath88 is the number of colors , and @xmath89 is the tree level matrix elements of the operators . in the effective theory ,
there are remaining uv divergences after taking into account the quark field renormalization , which must be canceled by the renormalization of the effective operators . calculating the diagrams in fig .
[ effective ] with the insertion of the operators @xmath73 and @xmath78 , we get the following amplitudes after quark field renormaliztion : @xmath90 \braket{q_1}_{\text{tree } } - \frac{\alpha_s}{4\pi } \frac{1}{n } \ln\frac{m_q^2}{m_q^2 } \braket{q_2}_{\text{tree } } , \\
\braket{q_2 ^ 0 } & = \frac{\alpha_s}{4\pi } \left ( \frac{3}{\hat{\epsilon } } + 1 - 3\ln\frac{m_q^2}{\mu^2 } - \frac{1}{n } \ln\frac{m_q^2}{m_q^2 } \right ) \braket{q_1}_{\text{tree } } \nonumber \\ & \qquad + \left [ 1 - \frac{\alpha_s}{4\pi } \frac{1}{n } \left ( \frac{3}{\hat{\epsilon } } + 1 - 3\ln\frac{m_q^2}{\mu^2 } \right ) + \frac{\alpha_s}{2\pi } c_f \left ( 1 - \ln\frac{m_q^2}{\mu^2 } \right ) + \frac{\alpha_s}{4\pi } \ln\frac{m_q^2}{m_q^2 } \right ] \braket{q_2}_{\text{tree}},\end{aligned}\ ] ] where @xmath91 .
the renormalization constant matrix for the two operators is @xmath92 from which we obtain the anomalous dimension matrix : @xmath93 our results of anomalous dimensions agree with those in ref .
@xcite . matching the results in the full theory and the effective theory ,
we extract the wilson coefficients : @xmath94 where @xmath95 .
these coefficients satisfy the renormalization group equations @xmath96 from which we can solve the wilson coefficients for arbitrary scale @xmath55 .
@xmath97 , @xmath98 are defined in terms of the matrix element @xmath99 , which can be written as @xmath100 where @xmath101 and @xmath102 denote matrix elements of the sm and the mssm - rpv effective hamiltanian , respectively . with the above matrix element , @xmath97 and @xmath98
can be expressed @xmath103 where @xmath104 and @xmath105 denotes the sm contributions , respectively . in eq.([basicformula ] ) , hadronic uncertainties arising from hadron decay constants cancel between @xmath101 and @xmath102 , and hadronic uncertainties remain only in @xmath104 and @xmath105 . as for @xmath3 mixing , @xmath106 can be obtained straightforwardly .
in this section we present our numerical results .
the susy scale is taken to be the mass of the sneutrino @xmath107 .
the ckm matrix elements are parametrized in the wolfenstein convention with four parameters @xmath108 , @xmath109 , @xmath110 and @xmath111 .
the other standard model parameters are taken to be @xmath112 gev@xmath113 , @xmath114 , @xmath115 gev @xcite .
the mass of b meson and k meson are @xmath116 mev , @xmath117 mev and @xmath118 mev @xcite .
the time - dependent cp asymmetry in @xmath31 decay @xmath119 @xcite .
the recent experimental values of the mass differences of the @xmath31 and @xmath24 mesons mixing are @xmath120 ps@xmath121 , @xmath122 ps@xmath121 @xcite , and @xmath123 is shown in eq .
( [ fbsmixing ] ) . in our numerical calculations
, we first neglect the uncertainties from the hadronic parameters in the sm and assume that the predictions of the sm can reproduce the central values of the experimental data .
thus we demand that the rpv contributions must not exceed the experimental upper and lower bounds of the corresponding data .
first , we consider the new measurement of @xmath0 mixing , which can constrain the combination @xmath124 once the sneutrino mass is given .
[ diagram1 ] shows the allowed region in the @xmath125 plane , where the gray area and the dark area is the allowed region extracted from the leading - order amplitude .
we find that with the new data , the couplings are constrained to the level of @xmath1 . after including the nlo qcd corrections ,
the gray area is excluded and only the dark area survives .
one can see that since the nlo qcd effects increase the resulting mass difference , the bound is more stringent ( about 50% lower ) than that from lo calculations .
if setting @xmath126 gev , the bounds on @xmath127 from the calculations of lo and nlo are @xmath128 respectively . the bounds for @xmath129 vary with the mass of the sneutrino . the gray area and the dark area are the allow region extracted from the lo amplitude
after nlo corrections , the gray area is excluded and only the dark area survives . ] fig .
[ diagram2 ] and fig .
[ diagram3 ] show the constraints on the combination @xmath130 from @xmath131 mixing and @xmath132 from @xmath133 mixing .
the situation here is similar to that in @xmath0 mixing : the nlo qcd corrections increase the mass differences and thus give more strong constraints on the combinations of couplings of mssm - rpv . at the reference point
@xmath126 gev , the bounds are @xmath134 above bounds are two and four orders of magnitude stronger than ones given in ref .
@xcite , as shown in eq .
( [ eq : old ] ) , which is due to the fact that the updated data of the measurements and the contributions from the sm have been considered in our calculations .
we further consider the situation that the hadronic uncertainties are involved .
in this case , the bounds on the combinations of couplings in mssm - rpv are looser than those in eqs .
( [ constraintsbs ] ) , ( [ constraintsbd ] ) and ( [ constraintsk ] )
. the new bounds from the three observables are @xmath137 respectively .
we also discuss the situation that @xmath138 is complex , and parametrize the combinations of relevant couplings as @xmath139 , where @xmath140 can be @xmath141 and @xmath142 . after assuming arbitrary phases @xmath143
, we can obtain constraints on the magnitude @xmath144 .
[ diagrambsc ] and fig .
[ diagrambsd ] shows the the allowed range of @xmath127 and @xmath145 for @xmath146gev , taking into account the constraints from @xmath9 and @xmath147 . for the @xmath0 mixing
, @xmath127 is about @xmath148 , while for the @xmath4 mixing , @xmath145 is about @xmath149 . in both figures
, there are additional areas with large @xmath144 besides ordinary areas with small @xmath144 , which correspond to ones where contributions from the mssm - rpv are larger than those from the sm , roughly twice the sm contributions but with different signs .
we do not discuss complex couplings in k system due to the fact that eq .
( [ k ] ) holds only under the condition that imaginary part of @xmath150 is far less than the real part . the allowed range of the magnitude and phase of @xmath129 constrained by the updated experimental data of @xmath9 . the gray area accounts for the tree level results .
when the qcd corrections are added , the allowed area is reduced to the dark area . ] the combinations of complex couplings can introduce new cp - violation origins , so we further investigate how the cp asymmetries in b decays constrain these combinations of couplings . using the experimental data of @xmath152 @xcite , we plot the allowed areas of the magnitude @xmath153 and the phase @xmath154 in fig.[sinbeta ] , which shows that the upper bound for @xmath153 is generally about @xmath155 , except for some special values of @xmath154 .
those special values correspond to @xmath156 , and we have @xmath157 .
in fact , another area with @xmath158 also survives from constraint of experiment data of @xmath5 , however , this possibility has been excluded by an angular analysis of @xmath159 and a time - dependent dalitz plot analysis of @xmath160@xcite . for the cp asymmetry in @xmath7
decays , although there is no experimental data of @xmath161 currently , future lhcb experiment will provide enough data to reach @xmath162 in the first year @xcite , which would provide a strong constraint on new physics .
so we assume @xmath163 and plot constraints on @xmath164 and @xmath165 in fig.[sinbetas ] , which shows that the upper bound for @xmath164 is about @xmath166 except for some special values of @xmath165 similar to the case of @xmath5 . in conclusion , using the updated data , we have calculated contributions to @xmath0 mixing , @xmath4 mixing and @xmath3 mixing in the framework of the minimal supersymmetric standard model with r - parity violation including nlo qcd corrections , and presented new constraints on the relevant couplings of mssm - rpv .
our results show that upper bound on the relevant combination of couplings of @xmath0 mixing is of the order @xmath1 , and upper bounds on the relevant combinations of couplings of @xmath4 mixing and @xmath3 mixing are two and four orders of magnitude stronger than ones reported in ref .
@xcite , respectively .
we also discussed the case of complex couplings and showed that how the relevant combinations of couplings are constrained by the updated experiment data of @xmath0 , @xmath4 mixing and time - dependent cp asymmetry @xmath5 , and future possible observations of @xmath6 at lhcb , respectively . @xmath168 . while preparing this manuscript the paper of @xcite appeared where
the same coupling combination from @xmath0 mixing is also discussed .
however , the authors of @xcite mainly dealt with contributions through the box diagram . our results induced by
the tree - level diagram is a few orders of magnitude stronger than theirs .
this work was supported in part by the national natural science foundation of china , under grant no . 10421503 , no .
10575001 and no .
10635030 , and the key grant project of chinese ministry of education under grant no .
305001 and the specialized research fund for the doctoral program of higher education .
b. aubert _
et al_. , [ babar colaboration ] , phys .
d71 , 032005 ( 2005 ) ; r. itoh _
et al_. , [ belle colaboration ] , phys .
95 , 091601 ( 2005 ) ; k abe _
et al_. , [ belle colaboration ] , hep - ex/0507065 ; for a comprehensive review , see robert fleischer , hep - ph/0608010 .
o. schneider , talk given at the workshop on `` flavour in the era of lhc '' first meeting , cern , 2005 , http://lhcb-doc.web.cern.ch/lhcb-doc/presentations/conferencetalks/ postscript/2005presentations / schneider_epfl.pdf . | we calculate contributions to @xmath0 mixing through tree - level sneutrino exchange in the framework of the minimal supersymmetric standard model with r - parity violation , including the next - to - leading - order qcd corrections .
we compare our results with the updated bounds on the @xmath0 mass difference reported by cdf collaborations , and present new constraints on the relevant combinations of parameters of the minimal supersymmetric standard model with r - parity violation .
our results show that upper bound on the relevant combination of couplings of @xmath0 mixing is of the order @xmath1 .
we also calculate the @xmath2 and @xmath3 mass differences , and show that the upper bounds on the relevant combinations of couplings are two and four orders of magnitude stronger than ones reported in the literatures , respectively .
we also discuss the case of complex couplings and show that how the relevant combinations of couplings are constrained by the updated experiment data of @xmath0 , @xmath4 mixing and time - dependent cp asymmetry @xmath5 , and future possible observations of @xmath6 at lhcb , respectively . |
in @xcite aerts constructed an remarkable macroscopic model in which he could operationally define coincidence experiments violating ( the chsh version of ) bell s inequality ( bi ) , with exactly the same numerical value @xmath0 as the one obtained in typical coincidence experiments with entangled microscopic entities in a singlet state @xcite ( which corresponds to the maximal violation obtainable in quantum mechanics @xcite ) .
subsequently , the model was generalized in @xcite , with the introduction of two parameters , @xmath1 $ ] , quantifying the degree of indeterminism and of non - locality present in the model , respectively .
more precisely , @xmath2 corresponds to the classical situation of absence of indeterminism , whereas @xmath3 to the situation of maximum indeterminism , typical of pure quantum systems . on the other hand
, @xmath4 corresponds to the situation of maximum locality , when the two pairs forming the double - system are totally disconnected , whereas @xmath5 corresponds to the opposite situation of perfect correlation .
the authors of @xcite obtain that bi can only be violated if @xmath6 , and that the violation takes its maximal numerical value @xmath7 when @xmath5 ( maximum correlation ) and @xmath2 ( minimum indeterminism ) . also , they find that for any @xmath8 , it is always possible to restore the validity of bi by increasing @xmath9 , whereas this is not any more possible if @xmath10 . to sum up , the study of the model described in @xcite has showed that the source of the violation of bi is the existence of a non zero correlation between the two pairs forming the double - system ( @xmath6 ) , whereas the only effect of increasing the level of indeterminism ( increasing @xmath9 ) is to decrease the value the inequality can take .
the purpose of the present paper is to analyze a different macroscopic model in which two parameters @xmath9 and @xmath11 will also be introduced , as in @xcite , to continuously vary the level of indeterminism and non - locality ( correlation ) .
this will allow us to confirm that a non zero correlation ( @xmath6 ) is the necessary condition for the violation of the inequality .
however , we will also show that by increasing the indeterminism ( increasing @xmath9 ) we can either increase or decrease the value bi can take , depending on the state in which the system is prepared before the coincidence experiments .
before describing our macroscopic model , we briefly recall the expression of bell s inequality ( bi ) . @xcite . on a given physical entity
, we assume that four different experiments can be performed : @xmath12 , @xmath13 , @xmath14 and @xmath15 .
we call @xmath16 , @xmath17 , @xmath18 and @xmath19 the outcomes associated to these four experiments , which can only take the values @xmath20 or @xmath21 .
we also assume that experiments @xmath12 and @xmath13 can be performed together with either of experiments @xmath14 and @xmath15 , so defining additional _ coincidence _ experiments : @xmath22 , @xmath23 , @xmath24 and @xmath25 . to every coincidence experiment @xmath26 , @xmath27 , @xmath28 , we can then associate the expectation value @xmath29 of the product of outcomes @xmath30 , by : @xmath31 where @xmath32 is the probability that the coincidence experiment @xmath26 yields the outcomes @xmath33 . under certain hypothesis ( usually referred to as _ bell locality _ , which have to do with the existence of hidden variables independently determining the experiments outcomes ) , one can prove the following ( bell ) inequality @xcite : @xmath34 inequality ( [ bell inequalities ] ) is generally violated by quantum systems in entangled states , like for instance those formed by two spin-@xmath35 entities in a _ singlet ( zero ) spin state _ , for which one can show that @xmath36 .
@xcite this means that entangled quantum systems generally violate bell s locality assumption , a fact which remains true even when the two subsystems are separated by a very large spatial distance . in other terms , no local physical theory in the sense specified by bell can agree with all statistical implications of quantum mechanics , and spatial separation is not a sufficient condition for _ experimental separation_.
let us consider a solid object , of homogeneous density , shaped as a rectangular @xmath37-prism ( @xmath37 even ) , i.e. , as a polyhedron formed by @xmath37 identical parallelogram - faces , and two lateral regular polygon - faces with @xmath37 sides ( see fig .
[ esagono singolo ] for the case @xmath38 , of an hexagonal prism ) .
we consider the prism as a special kind of die with @xmath37 faces , which can be rolled on a flat surface ( perpendicular to the gravitational field ) , along a given `` roll - direction , '' parallel to the two short sides of the rectangular upper face ( see fig . [ esagono singolo ] ) .
each rectangular face of the prism shows either a symbol `` @xmath39 , '' or a symbol `` @xmath40 . ''
the `` + '' symbol is only printed on @xmath41 parallel ( opposed ) rectangular faces of the prism , whereas the symbol `` - '' is printed on all the remaining @xmath42 rectangular faces .
we also assume that the `` + '' and `` - '' faces are made of different materials .
the two `` + '' faces can slide with an extremely low friction on the flat surface , whereas the @xmath42 `` - '' faces present a very high coefficient of friction with respect to it .
considering that the upper and lower faces of the prism always present the same symbol ( and therefore are made of the same material ) , we have that when the prism presents a `` + '' upper face , it can slide with almost no friction , whereas it can not do so if the upper face has a `` - '' symbol . in the following , when the prism presents a `` + '' ( respectively , `` - '' ) upper face , we shall simply say that it is in state ( + ) [ ( respectively , state ( - ) ] .
let us now describe what we shall call a _ rolling experiment _ with the prism .
it consists in placing a specific shooter ( similar to a `` flipper ball shooter '' ) behind the prism , along the roll - direction , pulling firmly its knob and then releasing it , thus communicating to the prism an a priori unpredictable impulsion ( see fig .
[ shooter esagono ] ) .
one then waits until the prism stops completely , and read the symbol on its upper face .
if it is a `` + , '' the outcome of the experiment is the value @xmath20 , otherwise the value @xmath21 .
if the prism is in state @xmath43 , then , because of the very low friction of the face in contact with the surface , it will not roll , but only glide on it , until all translational kinetic energy will be converted into heat .
therefore , the rolling experiment will not change the prism s upper face , and the outcome will be @xmath20 , with certainty
. on the other hand , if the prism is in state @xmath44 , then the face in contact with the surface will present an extremely high coefficient of friction with it , so that the prism will not anymore slide but roll ( i.e. , rotate around its longitudinal axis ) .
typically , most of the energy communicated to it by the shooter will be initially transformed into rotational kinetic energy , then , because of the positive work performed by the friction forces , the rotational energy will be gradually transformed into translational kinetic energy and heat , and of course in the end the prism will stop and show a specific upper face .
however , the prism will roll only for as long as the non - elastic effects associated with the _ rolling frictions _
remain lower than the _ sliding frictions _
, since in this case the prism requires less energy to be moved by rolling than by sliding .
but since two of the @xmath37 faces involved in the rolling movement ( those with the `` + '' symbol ) present a very low sliding friction , it is highly probable that the prism will conclude its run sliding on one of them , before it will ultimately totally stops . in other terms ,
apart from exceptional circumstances , which we can simply ignore not to complicate unnecessarily our discussion , we can _ ideally _ assume that , following a rolling experiment , if the initial state is @xmath44 then the final state will be @xmath43 , with certainty .
so , our prism is so conceived that , independently of its initial state [ @xmath43 or @xmath44 ] , the outcome of a rolling experiment will always be @xmath20 [ i.e. , its final state will always be @xmath43 ] .
we now consider two identical @xmath37-prisms , and use them to construct an entangled system by connecting them through space , by means of a rigid rod ( the length of the rod is arbitrary , but we assume it is made of an extremely light and rigid material ) whose two ends are glued at the center of the two opposed polygon - faces of the two prisms , as indicated in fig .
[ esagoni - connected ]
. clearly , the connecting rod creates correlations between the @xmath37 different rectangular faces of the two prisms .
here we assume that the two prisms have been connected in such a way that the correlations in question are those described in fig .
[ n - prism - correspondence ] .
we assume that the glue used to connect the rod to the two prisms is strong enough , so that if the two prisms are hit together , simultaneously ( in the same direction ) they will be able to maintain their connection while rolling , i.e. , to remain a whole entity .
but we also assume that the glue , although strong , is not as strong as to allow the two prisms to remain connected if only one prism is hit at a time . in other terms ,
if we hit with the shooter only one of the two prism , then , because of the inertia of the other prism , the impact will cause the rod to suddenly detach and fall , thus disconnecting the two solids ( one of which will then move in the roll - direction , whereas the other one will remain still ) . on the other hand ,
if two experimenters hit with two shooters both prisms at the same time , then the torque experienced by the rod will be much lower , so that the latter will not detach and the two prisms will be able to roll together on the surface , as a one piece entity .
on the double - prism system we have just defined , we perform four different coincidence experiments and show that they produce a violation of bi . for this
, we assume that two experimenters ( who we shall call experimenter @xmath47 and experimenter @xmath48 ) are placed close to each one of the two prisms .
experimenter @xmath47 can perform on his prism ( say , the left one ) two different experiments , @xmath12 and @xmath13 , which are defined as follow . experiment @xmath12 is the rolling experiment we have defined in sec . [ a rolling prism ] : it consists in hitting the left prism with a shooter along the roll - direction , then reading the symbol marked on the obtained upper face , producing in this way one of the two outcomes : @xmath49 , or @xmath50 .
experiment @xmath13 is much simpler : it simply consists in looking at the prism s upper face and check whether it is flat or not .
if it is so , then the outcome is @xmath51 , otherwise it is @xmath52 .
experimenters @xmath48 can perform on its prism ( the right one ) the same two experiments as experimenters @xmath47 . in other terms , @xmath14 is defined as @xmath12 , and @xmath15 as @xmath13 . clearly ,
all of the four above mentioned experiments , when singly performed , can only produce the outcome @xmath20 . also , when the coincidence experiments @xmath23 , @xmath24 and @xmath25 are performed by the two experimenters , the only possible outcome for them is @xmath53 , so that according to ( [ expectation value ] ) , @xmath54 .
the situation is however more articulate for experiment @xmath22 , which creates correlations between upper faces .
indeed , if the two experimenters hit simultaneously the two prisms , then , as we explained , the rod will not separate and they will remain connected as they roll ( the double - prism system can not slide , but only roll , as one of its two lower faces is always a high - friction face ) .
therefore , according to fig .
[ n - prism - correspondence ] , we have that the probability for the outcome @xmath55 is zero , the probability for each one of the two outcomes @xmath56 and @xmath57 is @xmath58 , and the probability for the outcome @xmath55 is @xmath59 .
thus , @xmath60 = 1-(8/n)$ ] . inserting all this in ( [ bell inequalities ] ) ,
we obtain : @xmath61 eq .
( [ violation - n ] ) clearly violates bi ( [ bell inequalities ] ) .
it does so in a maximal way ( @xmath62 ) for the case @xmath63 of two tetragonal prisms , and with a value which is very close to the quantum mechanical maximum of @xmath64 for the case @xmath65 of two decagonal prisms ( @xmath66 ) .
before continuing in our analysis , let us mention that it was diederik aerts who was the first , in the early eighties of last century , to conceive an explicit macroscopic model with non - local correlations violating bi in a maximal way , @xcite , as our double - prism system with @xmath63 can do ( see also ref .
@xcite and the references cited therein ) .
an important difference between our double - prism model and the historical ( connected vessels of water ) model of aerts , is that in the latter it is the fact that the system is broken which is the mechanism responsible for the creation of correlations , whereas in our model it is exactly the contrary : it is only when the double system is _ not _ broken that the coincidence experiment @xmath22 can produce correlations between upper faces ( it is important to distinguish `` faces '' from `` upper faces , '' as only the latter are created in a rolling experiment : if we do nt create upper faces , by rolling the prisms , bi can not be violated @xcite ) .
we want now to vary the amount of indeterminism and non - locality in our model , to see how this can affect the degree of violation of bi , thus elucidating their role in the violation .
first of all , let us observe that if the number @xmath37 of rectangular faces of the two prisms tends to infinity , the probability @xmath67 tends to @xmath68 , so that in this limit , according to ( [ violation - n ] ) , @xmath69 . one could be tempted then to affirm that the level of indeterminism decreases as @xmath70 , and that this produces a corresponding decrease of the degree of violation of bi , contrary to what was obtained by aerts and collaborators , who showed instead that a decrease of indeterminism produces a stronger violation of bi .
@xcite this however would be a wrong conclusion .
indeed , if by increasing @xmath37 we can reduce the value taken by bi , this is so because for each different @xmath37 we have a different physical system , and not because we are decreasing the level of indeterminism in a given physical system .
what we have to do , instead , is to vary such level within a same double - prism system , characterized by a fixed value of @xmath37 .
there are of course different ways to do so . a simple one is to consider that the two experimenters , who have to hit the prisms in the rolling experiments , may not just do it aimlessly , but trying to obtain the specific effect of having the two prisms rolling over a predetermined distance , thus producing a predetermined total angle of rotation .
of course , this will make a difference only when the two experimenters are executing the coincidence experiment @xmath22 , as is clear from the fact that when they do nt hit the two prisms together ( in experiments @xmath23 and @xmath24 ) , then because of the extreme low friction of the `` + '' faces ( and the fact that , according to the protocol , they have to pull the knob firmly , i.e. , hit the prism strongly ) they will not succeed in altering the predetermined @xmath71 outcome .
now , let us assume that if the two players are successful in producing the chosen rotation in the coincidence experiment @xmath23 , then , considering the state in which the system was prepared , the outcome of the rolling experiment will be @xmath72 ( or @xmath73 ) . in general terms , we can characterize the ability of the two players in obtaining the desired effect by means of a continuous parameter @xmath74 $ ] , such that @xmath75 corresponds to the totally random situation , of maximum indeterminism , when the experimenters hit the two prisms without trying to obtain any specific result , and @xmath2 corresponds to the opposite situation where the two players are able to perfectly control their shot and therefore produce the @xmath72 ( or @xmath73 ) outcome , without fail . in other terms , by varying @xmath9 from @xmath68 to @xmath76 , we can decrease the degree of indeterminism in the measurement processes .
following ref .
@xcite , we also want to introduce , in addition to the indeterminism parameter @xmath9 , an additional continuous parameter @xmath77 $ ] , characterizing the connectedness of the two prisms ( i.e. , the strength of the correlations between the rectangular faces of the two prisms )
. a simple and natural way to do so is to consider the possibility that , sometimes , the rod can also detach and fall during the execution of the @xmath22 experiment .
let us simply assume that @xmath11 corresponds to the probability for the rod of remaining duly glued to the two prisms , when @xmath22 is executed .
this means that @xmath5 corresponds to the case of maximum correlation , whereas @xmath4 to the case of absence of correlation . in other terms , by varying @xmath11 from @xmath68 to @xmath76 , we can decrease the degree of connectedness ( non - locality ) of the two entangled prism - entities .
clearly , when the rod detaches , the two prism will necessarily end their respective roll by showing an upper face with the `` + '' symbol , and this means that the outcome @xmath53 is now also possible for the experiment @xmath22 , with probability @xmath78 .
assuming for simplicity a linear variation of the probabilities as a function of the ability parameter @xmath9 , we can write for the four different outcomes of experiment @xmath22 : @xmath79 then , observing that the probabilities associated with the other three coincidence experiments are not directly affected by the values taken by @xmath9 and @xmath11 ( this because , when the two prisms are rolled independently from one another , they will inevitably end their run by showing a `` + '' upper face ) , we obtain , after a short calculation : @xmath80.\ ] ] visibly , ( [ i - epsilon - rho ] ) generalizes ( [ violation - n ] ) , and in accordance with the analysis of the model in @xcite , we can observe the following . only for the value @xmath4 , corresponding to two totally disconnected prisms , bi is obeyed , and this means that it is the correlation between the two subsystems , i.e. , the non - locality ingredient , which is really responsible for the violation . also , we can see that when we increase @xmath9 , i.e. , the amount of uncertainty in the outcome of the experiments , we clearly also diminish the value of ( [ i - epsilon - rho ] ) , i.e. , the degree of violation of bi , which means that , in accordance with @xcite , not only indeterminism ( associated here to parameter @xmath9 ) does nt contribute to the violation , but actually tends to reduce it .
clearly , the maximum violation , @xmath62 , is obtained for @xmath81 and @xmath2 , i.e. , for the situation of perfect correlation and full predictability .
it is however important to observe that although inequality ( [ violation - n ] ) is independent of the choice of the initial state of the double - prism system ( which can either be @xmath82 , @xmath83 or @xmath84 ) , this is not anymore the case when the outcomes of the experiments are affected by the ability parameter @xmath9 . indeed , for a given rolling distance ( i.e. , for a given total angle of rotation ) that the two experimenters will try to obtain , if we change the initial state of the system , then we will also change the expected outcome of the experiment , and therefore the value taken by bi . to see this ,
let us assume this time that the system is prepared in a state such that when the two experimenters can successfully impart to the double - prism the chosen total rotation angle , the final outcome is now @xmath55 , instead of the previous @xmath72 ( or @xmath73 ) .
then , we obtain the following probabilities for the four outcomes of experiment @xmath22 : @xmath85.\end{aligned}\ ] ] observing once more that the probabilities associated to the other three coincidence experiments are not affected by the specific values taken by @xmath9 and @xmath11 , we find after a short calculation : @xmath86 as we can see , ( [ i - epsilon - rho - bis ] ) differs sensibly from ( [ i - epsilon - rho ] ) , and the effect produced by a variation of parameter @xmath9 is now exactly opposite : an increase of the level of indeterminism ( i.e. , an increase of @xmath9 ) produces a strengthening of the violation of bi , and not a weakening of it .
also , the situation of full predictability ( @xmath2 ) is not anymore associated to a maximal violation of the inequality , but to the non - violation of it !
we conclude with a few comments .
first of all , it could be objected that , contrary to the model explored in @xcite , @xmath9 is not the only source of indeterminism in our model , considering that we have defined the parameter @xmath11 as a probability , and that the outcomes of the coincidence experiments @xmath22 clearly depend on whether the rod will detach or not during their execution . in other terms
, @xmath11 also contributes to the degree of unpredictability of the outcomes of @xmath22 .
this is obviously true , but can not alter our conclusion . indeed ,
if we keep @xmath11 fixed , then an increase of @xmath9 does actually correspond to a global increase of the level of indeterminism in our model , which , according to ( [ i - epsilon - rho ] ) and ( [ i - epsilon - rho - bis ] ) , can either reinforce or weaken the degree of the violation of bi , depending on the state in which the system was prepared .
in fact , our probabilistic description of non - locality in the model highlights an additional mechanism through which a variation of the amount of indeterminism can affect the value taken by bi , in a way that is not unique . indeed ,
as regards the randomness incorporated in @xmath11 , we can consider that the situation @xmath87 corresponds to the one of maximum uncertainty .
then , when from that value of maximum uncertainty we increase @xmath11 , and therefore reduce the uncertainty , according to ( [ i - epsilon - rho ] ) and ( [ i - epsilon - rho - bis ] ) we increase the violation of bi , and this independently of the initial state of the system .
but when starting from the same value @xmath87 we decrease @xmath11 , also in this case we reduce the uncertainty , yet this time we decrease the violation of bi ( which for the limit value @xmath4 is obeyed ) , independently of the initial state of the system .
having said that , let us now explain why , when the level of indeterminism is increased in our model , by increasing @xmath9 , we can either increase or decrease the violation of bi , depending on the initial state of the system . for simplicity ,
let us set @xmath5 ( perfect correlation ) .
considering that in our model @xmath88 , independently of the value taken by @xmath9 , we can write : @xmath89 .
now , having assumed perfect correlation , it is clear that only three outcomes are possible for experiment @xmath22 : @xmath84 , @xmath83 and @xmath82 . according to ( [ expectation value ] )
, outcome @xmath84 contributes positively to @xmath90 , whereas outcomes @xmath83 and @xmath82 contribute negatively to it .
therefore , if the outcome of experiment @xmath22 is predetermined , and corresponds to @xmath84 , @xmath91 and bi is obeyed ( @xmath92 ) . on the other hand , if the predetermined outcome is @xmath83 , or @xmath82 , @xmath93 and bi is maximally violated ( @xmath62 ) .
assuming that we are in the situation where @xmath91 , for @xmath94 , then by increasing @xmath9 the system will start sometimes to also explore the outcomes @xmath83 , or @xmath82 , and since the latter contribute negatively to @xmath95 , their possible selection will cause its value to diminish , thus producing an increase in the violation of bi . and of course , the situation is reversed when @xmath93 , for @xmath94
. the reason why aerts et al . could nt highlight in @xcite this double role played by indeterminism in coincidence experiments , is because the statistics of outcomes of their model does not depend on the specific state in which the system is prepared , but only on the relative orientation of the measuring apparatus ( as is the case in spin measurements on singlet states , provided the direction of flight of the entangled pair is orthogonal to the directions of orientation of the stern - gerlach filters ) .
therefore , although they have studied a model which is more elaborated than ours ( as meant to reproduce the same statistics as spin measurements on singlet states ) , it was actually too specific to fully elucidate the question of the role played by indeterminism in the violation of bi . to conclude
, we briefly summarize our results . in agreement with @xcite
, we have found that it is the aspect of non - locality , expressed by the connecting rod in our model , which produces the violation of bi , through the _ creation of correlations _ between outcomes of experiments performed in coincidence . in agreement with @xcite
, we have also found that by increasing the indeterminism ( increasing @xmath9 ) , we can decrease the value taken by bi .
however , we have also shown that this is not a general fact : depending on the state in which the system is prepared , an increase of the level of indeterminism is also capable of increasing the value taken by bi .
an additional source of indeterminism can also be envisioned , that was not considered in @xcite , associated to the possibility of actualizing different degrees of non - locality ( connectedness ) in a coincidence experiment .
this possibility was described in our model by assuming that not only correlations between `` upper faces '' had to be considered as potential before a coincidence experiment , but also correlations between `` faces . '' when we do so , we find that if the amount of indeterminism associated to this additional level of potentiality is decreased , it can either increase or decrease the value of bi , depending on whether the decrease produces a strengthening or a weakening of the non - local aspect , respectively . as far as this author
can judge , it is still an open experimental question to know if a quantum microscopic system in an entangled state , like a singlet state , has also a probability of `` breaking '' during a coincidence experiment ( in the same way the rod in our macroscopic model has a probability of detaching ) , and produce in this way uncorrelated outcomes , instead of correlated ones .
@xcite a last remark is in order .
the description of the rolling / sliding behavior of the @xmath37-prisms must be understood , as we already said , in an idealized sense .
we have never performed real experiments with systems of this sort , and therefore can not guarantee that the way we have theoretically described their behavior , although reasonable , would be perfectly adequate from the perspective of a real experiment .
but of course , this is not an essential point in the present analysis : what is important is that the idealized system we have considered behaves in a logical manner , according to coherent mechanisms .
the question of how to exactly implement such behavior in real models , which can be subject to real experiments , is a technological issue which goes beyond the purely conceptual scope of the present note .
d. aerts , s. aerts , j. broekaert and l. gabora , the violation of bell inequalities in the macroworld , found .
phys . , 30 , pp . 1387 - 1414 ( 2000 ) .
d. aerts , a mechanistic classical laboratory situation violating the bell inequalities with @xmath0 , exactly ` in the same way ' as its violations by the epr experiments , helv .
acta , 64 , 1 - 23 .
( 1991 ) b. s. cirelson , quantum generalizations of bi , lett .
phys . , 4 , 93 ( 1980 ) .
j.s . bell , in : b. despagnat ( ed . ) , proceedings of the international school of physics `` enrico fermi , '' course xlix ( academic press , new york , 1971 ) 171 .
j. s. bell , physics ( long island city , n.y . ) 1 , 195 ( 1964 ) . reproduced as ch . 2 of j. s. bell , speakable and unspeakable in quantum mechanics ( cambridge university press , 1987 ) .
a. aspect et al . , experimental realization of einstein - podolsky - rosen - bohm gedankenexperiment : a new violation of bell s inequalities , phys .
lett . , 49 , p. 91
a. aspect , bi test : more ideal than ever , nature ( london ) , 398 , p. 189
d. aerts , example of a macroscopical situation that violates bell inequalities , lettere al nuovo cimento , 34 , pp . 107 - 111 ( 1982 ) .
d. aerts , the missing element of reality in the description of quantum mechanics of the epr paradox situation , helv .
acta , 57 , 421428 ( 1984 ) .
m. sassoli de bianchi , using simple elastic bands to explain quantum mechanics : a conceptual review of two of aerts machine - models , to appear in : central european journal of physics , doi : 10.2478/s11534 - 012 - 0164 - 2 .
d. aerts , an attempt to imagine parts of the reality of the micro - world , 325 , in : problems in quantum physics ii ; gdansk 89 , edited by : j. mizerski et al . ,
world scientific publishing company , singapore ( 1990 ) .
m. sassoli de bianchi , quantum dice , arxiv:1301.3038 [ quant - ph ] ( 2013 ) . | some years ago aerts _ et al . _
@xcite presented a macroscopic model in which the amount of non - locality and indeterminism could be continuously varied , and used it to show that by increasing non - locality one increases , as expected , the degree of violation of bell s inequality ( bi ) , whereas , more surprisingly , by increasing indeterminism one decreases the degree of the violation of bi . in this note
we propose a different macroscopic model in which the amount of non - locality and indeterminism can also be parameterized , and therefore varied , and we find that , in accordance with the model of aerts _ et al .
_ , an increase of non - locality produces a stronger violation of bi .
however , differently from their model , we also find that , depending on the initial state in which the system is prepared , an increase of indeterminism can either strengthen or weaken the degree of violation of bi . |
the basic questions faced by extreme value analysis consist in estimating the probability of exceeding a threshold that is larger than the sample maximum and estimating a quantile of an order that is larger than 1 minus the reciprocal of the sample size . in words , they consist in making inferences on regions that lie outside the support of the empirical distribution . in order to face these challenges in a sensible framework , extreme value theory ( ) assumes that the sampling distribution @xmath0 satisfies a regularity condition .
indeed , in heavy - tail analysis , the tail function @xmath1 is supposed to be regularly varying that is , @xmath2 exists for all @xmath3 .
this amounts to assume the existence of some @xmath4 such that the limit is @xmath5 for all @xmath6 .
in other words , if we define the _ excess distribution above the threshold @xmath7 _ by its survival function : @xmath8 for @xmath9 , then @xmath10 is regularly varying if and only if @xmath11 converges weakly towards a pareto distribution .
the sampling distribution @xmath0 is then said to belong to the _ max - domain of attraction _ of a frchet distribution with index @xmath12 ( abbreviated in @xmath13 ) and @xmath14 is called the _ extreme value index_. the main impediment to large exceedance and large quantile estimation problems alluded above turns out to be the estimation of the extreme value index . since the inception of extreme value analysis , many estimators have been defined , analysed and implemented into software .
@xcite introduced a simple , yet remarkable , collection of estimators : for @xmath15 , @xmath16 where @xmath17 are the _ order statistics _ of the sample @xmath18 ( the non - increasing rearrangement of the sample ) .
an integer sequence @xmath19 is said to be _ intermediate _ if @xmath20 while @xmath21 .
it is well known that @xmath0 belongs to @xmath22 for some @xmath4 if and only if , for all intermediate sequences @xmath19 , @xmath23 converges in probability towards @xmath14 @xcite . under mildly stronger conditions
, it can be shown that @xmath24 is asymptotically gaussian with variance @xmath25 this suggests that , in order to minimise the quadratic risk @xmath26 $ ] or the absolute risk @xmath27 , an appropriate choice for @xmath28 has to be made . if @xmath28 is too large , the hill estimator @xmath23 suffers a large bias and , if @xmath28 is too small , @xmath23 suffers erratic fluctuations .
as all estimators of the extreme value index face this dilemma ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) , during the last three decades , a variety of data - driven selection methods for @xmath28 has been proposed in the literature ( see @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite to name a few ) . a related but distinct problem is considered by @xcite : constructing uniform and adaptive confidence intervals for the extreme value index .
the rationale for investigating adaptive hill estimation stems from computational simplicity and variance optimality of properly chosen hill estimators @xcite .
the hallmark of our approach is to combine techniques of with tools from concentration of measure theory . as up to our knowledge
, the impact of the concentration of measure phenomenon in has received little attention , we comment and motivate the use of concentration arguments .
talagrand s concentration phenomenon for products of exponential distributions is one instance of a general phenomenon : concentration of measure in product spaces @xcite .
the phenomenon may be summarised in a simple quote : functions of independent random variables that do not depend too much on any of them are almost constant @xcite . the concentration approach helps to split the investigation in two steps : the first step consists in bounding the fluctuations of the random variable under concern around its median or its expectation , while the second step focuses on the expectation .
this approach has seriously simplified the investigation of suprema of empirical processes and made the life of many statisticians easier @xcite .
to point out the potential uses of concentration inequalities in the field of is one purpose of this paper . in statistics
, concentration inequalities have proved very useful when dealing with estimator selection and adaptivity issues : sharp , non - asymptotic tail bounds can be combined with simple union bounds in order to obtain uniform guarantees of the risk of collection of estimators . using concentration inequalities to investigate adaptive choice of the number of order statistics to be used in tail index estimation is a natural thing to do . in the present
setting , tail index estimators are functions of independent random variables .
talagrand s quote raises a first question : in which way are these tail functionals smooth functions of independent random variables ?
we do not attempt here to revisit the asymptotic approach described by @xcite which equates smoothness with hadamard differentiability .
our approach is non - asymptotic and our conception of smoothness somewhat circular , smooth functionals are these functionals for which we can obtain good concentration inequalities . in this paper , we combine talagrand s concentration inequality for smooth functions of independent exponentially distributed random variables ( theorem [ bernstein : expo ] ) with three traditional tools of : the quantile transform , karamata s representation for slowly varying functions , and rnyi s characterisation of the joint distribution of order statistics of exponential samples .
this allows us to establish concentration inequalities for the hill process @xmath29 ( theorem [ prp : hill : concentration ] ) in section [ sec : conc - ineq - hill ] . in section [ sec
: adapt - hill - estim ] , we build on these concentration inequalities to analyse the performance of a variant of lepki s rule defined in sections [ sec : lepsk - meth - adapt ] and [ sec : adapt - hill - estim ] : theorem [ thm : adapt - hill - estim ] describes an oracle inequality and corollary [ cor : adapt - hill - estim-2ndrv ] assesses the performance of this simple selection rule under a mild assumption on the so - called von mises function .
note that the condition is less demanding than the regular variation condition on the von mises function that has often been assumed when looking for adaptive tail index estimators ( notable exceptions being @xcite and @xcite ) .
it reveals that the performance of hill estimators selected by lepski s method matches known lower bounds ( see section [ sec : lower - bound ] ) that is , they suffer the loss of efficiency which is inherent to this problem , but not more .
proofs are given in section [ sec : proofs ] .
finally , in section [ sec : simulations ] , we examine the performance of this adaptive hill estimator for finite sample sizes using monte - carlo simulations .
the quantile function @xmath30 is the generalised inverse of the distribution function @xmath0 .
the _ tail quantile function _ of @xmath0 is a non - decreasing function defined on @xmath31 by @xmath32 , or by @xmath33 in this text , we use a variation of the quantile transform that fits : if @xmath34 is exponentially distributed , then @xmath35 is distributed according to @xmath0 .
moreover , by the same argument , the order statistics @xmath17 are distributed as a monotone transformation of the order statistics @xmath36 of a sample of @xmath37 independent standard exponential random variables .
@xmath38 thanks to rnyi s representation for order statistics of exponential samples , agreeing on @xmath39 , the rescaled exponential spacings @xmath40 are independent and exponentially distributed . the quantile transform and rnyi s representation are complemented by karamata s representation for slowly varying functions .
recall that a function @xmath41 is _ slowly varying at infinity _ if for all @xmath3 , @xmath42 .
the von mises condition specifies the form of karamata s representation ( see * ? ? ?
* corollary 2.1 ) of the slowly varying component @xmath43 of @xmath44 .
[ dfn : vmises : cond ] a distribution function @xmath0 belonging to @xmath45 , satisfies the von mises condition if there exist a constant @xmath46 , a constant @xmath47 and a measurable function @xmath48 on @xmath31 such that , for @xmath49 @xmath50 with @xmath51 .
the function @xmath48 is called the _
von mises function_. in the sequel , we assume that the sampling distribution @xmath52 , @xmath4 , satisfies the von mises condition with @xmath53 , von mises function @xmath48 and define the non - increasing function @xmath54 from @xmath55 to @xmath56 by @xmath57 . in the text , we assume that @xmath58 . combining the quantile transformation , rnyi s and karamata s representations ,
it is straightforward that , under the von mises condition , the sequence of hill estimators is distributed as a function of the largest order statistics of a standard exponential sample . [ hill : rep ] the vector of hill estimators @xmath59 is distributed as the random vector @xmath60 where @xmath61 are independent standard exponential random variables while , for @xmath62 , @xmath63 is distributed like the @xmath64th order statistic of an @xmath37-sample of the exponential distribution . for a fixed @xmath65 , a second distributional representation is available , @xmath66 where @xmath67 and @xmath68 are defined as in proposition [ hill : rep ] .
this second , simpler , distributional representation stresses the fact that , conditionally on @xmath68 , @xmath69 is distributed as a mixture of sums of independent random variables approximately distributed as exponential random variables with scale @xmath14 .
this distributional identity suggests that the variance of @xmath70 scales like @xmath71 , an intuition that is corroborated by analysis , see section [ sec : conc - ineq - hill ] .
the bias of @xmath70 is connected with the von mises function @xmath48 by the next formula @xmath72= { \ensuremath{\mathbb{e}}}\left[\int_1^\infty \frac{\eta\left({\mathrm{e}}^{y_{(k+1 ) } } v\right)}{v^2 } \mathrm{d}v\right]\ , .\ ] ] henceforth , let @xmath73 be defined on @xmath31 by @xmath74 the quantity @xmath75 is the bias of the hill estimator @xmath70 given @xmath76 .
the second expression for @xmath73 shows that @xmath73 is differentiable with respect to @xmath77 ( even though @xmath48 might be nowhere differentiable ) and that @xmath78 the von mises function governs both the rate of convergence of @xmath79 towards @xmath80 , or equivalently of @xmath81 towards @xmath5 , and the rate of convergence of @xmath82 towards @xmath83 .
the difficulty in extreme value index estimation stems from the fact that , for any collection of estimators , for any intermediate sequence @xmath19 , and for any @xmath4 , there is a distribution function @xmath84 such that the bias @xmath85 decays at an arbitrarily slow rate .
this has led authors to put conditions on the rate of convergence of @xmath79 towards @xmath80 as @xmath77 tends to infinity while @xmath3 , or equivalently , on the rate of convergence of @xmath81 towards @xmath5 .
these conditions have then to be translated into conditions on the rate of decay of the bias of estimators . as we focus on hill estimators ,
the connection between the rate of convergence of @xmath79 towards @xmath80 and the rate of decay of the bias is transparent and well - understood @xcite : the theory of @xmath86-regular variation provides an adequate setting for describing both rates of convergence @xcite . in words , if a positive function @xmath87 defined over @xmath55 is such that , for some @xmath88 , for all @xmath89 , @xmath90 } g(tx)/g(t ) < \infty$ ] , @xmath87 is said to have _ bounded increase_. if @xmath87 has bounded increase , the class @xmath91 is the class of measurable functions @xmath92 on some interval @xmath93 , such that as @xmath94 @xmath95 for all @xmath96 for example , the analysis carried out by @xcite rests on the condition that , if @xmath84 , for some @xmath97 , @xmath98 and @xmath99 , @xmath100 this condition implies that @xmath101 with @xmath102 @xcite .
thus , under the von mises condition , condition implies that the function @xmath103 belongs to @xmath91 with @xmath104 moreover , the abelian and tauberian theorems from @xcite assert that @xmath105 if and only if @xmath106 for any intermediate sequence @xmath19 . in this text , we are ready to assume that if @xmath84 and satisfies the von mises condition , then , for some @xmath97 and @xmath99 and @xmath107 , @xmath108 this condition is arguably more stringent than ( [ eq : kimcarpcond ] ) .
however , we do not want to assume that @xmath48 satisfies a regular variation property .
this would imply that @xmath109 is @xmath110-regularly varying . indeed , assuming as in @xcite and several subsequent papers that @xmath0 satisfies @xmath111 where @xmath112 are constants and @xmath113 , or equivalently,@xcite that @xmath114 satisfies @xmath115 ( which entails that @xmath48 is regularly varying ) makes the problem of extreme value index estimation easier ( but not easy ) .
these conditions entail that , for any intermediate sequence @xmath19 , the ratio @xmath116|/(n / k_n)^\rho$ ] converges towards a finite limit as @xmath37 tends to @xmath117 @xcite .
this makes the estimation of the second - order parameter a very natural intermediate objective ( see for example * ? ? ?
the necessity of developing data - driven index selection methods is illustrated in figure [ fig : riskcomp - student ] , which displays the estimated standardised root mean squared error ( rmse ) of hill estimators @xmath118^{1/2}\ ] ] as a function of @xmath119 for four related sampling distributions which all satisfy the second - order condition with different values of the second - order parameters .
for samples of size @xmath120 from student s distributions with different degrees of freedom @xmath121 .
all four distributions satisfy condition with @xmath122 .
the increasing parts of the lines reflect the values of @xmath110 .
rmse is estimated by averaging over @xmath123 monte - carlo simulations . ] under this second - order condition , @xcite proved that the asymptotic mean squared error of the hill estimator is minimal for sequences @xmath124 satisfying @xmath125 with @xmath126 . since @xmath97 , @xmath98 and the second - order parameter @xmath99
are usually unknown , many authors have been interested in the construction of data - driven selection procedures for @xmath28 under conditions such as .
a great deal of ingenuity has been dedicated to the estimation of the second - order parameters and to the use of such estimates when estimating first order parameters .
as we do not want to assume a second - order condition such as condition , we resort to lepski s method which is a general attempt to balance bias and variance . since its introduction @xcite , this general method for model selection has been proved to achieve adaptivity and to provide one with oracle inequalities in a variety of inferential contexts ranging from density estimation to inverse problems and classification @xcite .
very readable introductions to lepski s method and its connections with penalised contrast methods can be found in @xcite . in
, we are aware of three papers that explicitly rely on this methodology : @xcite , @xcite and @xcite .
the selection rule analysed in the present paper ( see section [ sec : adapt - hill - estim ] for a precise definition ) is a variant of the preliminary selection rule introduced in @xcite @xmath127 where @xmath128 is a sequence of thresholds such that @xmath129 and @xmath130 , and @xmath131 is the hill estimator computed from the @xmath132 largest order statistics .
the definition of this stopping time " is motivated by lemma 1 from @xcite which asserts that , under the von mises condition , @xmath133 | = o_p \left ( \sqrt{\ln \ln n } \right ) { \enspace .}\ ] ] in words , this selection rule almost picks out the largest index @xmath119 such that , for all @xmath64 smaller than @xmath119 , @xmath70 differs from @xmath131 by a quantity that is not much larger than the typical fluctuations of @xmath131 .
this index selection rule can be performed graphically by interpreting an alternative hill plot as shown on figure [ fig : cauchy - lepski ] ( see * ? ? ?
* ; * ? ? ?
* for a discussion on the merits of alt - hill plots ) . computed on a pseudo - random sample of size @xmath134 from student distribution with 1 degree of freedom ( cauchy distribution ) .
hill estimators are computed from the positive order statistics .
the grey ribbon around the plain line provides a graphic illustration of lepski s method . for a given value of @xmath64 ,
the width of the ribbon is @xmath135 .
a point @xmath136 on the plain line corresponds to an eligible index if the horizontal segment between this point and the vertical axis lies inside the ribbon that is , if for all @xmath137 , @xmath138 .
if @xmath139 were replaced by an appropriate quantile of the gaussian distribution , the grey ribbon would just represent the confidence tube that is usually added on hill plots .
the triangle represents the selected index with @xmath140 .
the cross represents the oracle index estimated from monte - carlo simulations , see table [ tab : ratios:1 ] . ]
the goal of @xcite is not to investigate the performance of the preliminary selection rule defined in display but to design a selection rule @xmath141 , based on @xmath142 , that would asymptotically mimic the optimal selection rule @xmath143 under second - order conditions .
our goal , as in @xcite , is to derive non - asymptotic risk bounds without making a second - order assumption . in both papers
, the rationale for working with some special collection of estimators seems to be the ability to derive non - asymptotic deviation inequalities for @xmath70 either from exponential inequalities for log - likelihood ratio statistics or from simple binomial tail inequalities such as bernstein s inequality ( see * ? ? ?
* section 2.8 ) . in models satisfying condition , the estimators from @xcite achieve the optimal rate up to a @xmath144 factor .
@xcite prove that the risk of their data - driven estimator decays at the optimal rate @xmath145 up to a factor @xmath146 in models satisfying condition .
we aim at achieving optimal risk bounds under condition using a simple estimation method requiring almost no calibration effort and based on mainstream extreme value index estimators . before describing the keystone of our approach in section [ sec : talagr - conc - phen ] , we recall the recent lower risk bound for adaptive extreme value index estimation
. one of the key results in @xcite is a lower bound on the accuracy of adaptive tail index estimation .
this lower bound reveals that , just as for estimating a density at a point @xcite , or point estimation in sobolev spaces @xcite , as far as tail index estimation is concerned , adaptivity has a price . using fano s lemma , and a bayesian game that extends cleanly in frameworks of @xcite and @xcite , @xcite were able to prove the next minimax lower bound .
[ thm - kc - lower - bound ] let @xmath147 and @xmath148 $ ] .
then , for any tail index estimator @xmath149 and any sample size @xmath37 such that @xmath150 , there exists a probability distribution @xmath151 such that a. @xmath152 with @xmath4 , b. @xmath151 meets the von mises condition with von mises function @xmath48 satisfying @xmath153 for some @xmath154 , c. @xmath155 + and @xmath156 \geq \frac{c_\rho}{4(1 + 2{\mathrm{e } } ) } \left(\frac{v\ln\ln n}{n}\right)^{|\rho|/(1 + 2|\rho| ) } \ , , \ ] ] with @xmath157 .
using birg s lemma instead of fano s lemma , we provide a simpler , shorter proof of this theorem ( see appendix [ proof : lower : bound ] )
. the lower rate of convergence provided by theorem [ thm - kc - lower - bound ] is another incentive to revisit the preliminary tail index estimator from @xcite .
however , instead of using a sequence @xmath128 of order larger than @xmath158 in order to calibrate pairwise tests and ultimately to design estimators of the second - order parameter ( if there are any ) , it is worth investigating a minimal sequence where @xmath139 is of order @xmath159 , and check whether the corresponding adaptive estimator achieves the carpentier - kim lower bound ( theorem [ thm - kc - lower - bound ] ) . in this paper
, we focus on @xmath139 of the order @xmath158 . the rationale for imposing @xmath139 of the order @xmath158
can be understood by the fact that , even if the sampling distribution is a pure pareto distribution with shape parameter @xmath14 ( @xmath160 for @xmath161 ) , if @xmath162 the preliminary selection rule will , with high probability , select a small value of @xmath119 and thus pick out a suboptimal estimator .
this can be justified using results from @xcite ( see appendix [ sec : calibr - prel - select ] for details ) .
such an endeavour requires sharp probabilistic tools .
they are the topic of the next section . deriving authentic concentration inequalities for hill estimators
is not straightforward .
fortunately , the construction of such inequalities turns out to be possible thanks to general functional inequalities that hold for functions of independent exponentially distributed random variables .
we recall these inequalities ( proposition [ poincare : expo ] and theorem [ bernstein : expo ] ) which have been largely overlooked in statistics .
a thorough and readable presentation of these inequalities can be found in @xcite .
we start by the easiest result , a variance bound that pertains to the family of poincar inequalities .
[ poincare : expo ] if @xmath87 is a differentiable function over @xmath163 and @xmath164 where @xmath165 are independent standard exponential random variables , then @xmath166 \ , .\ ] ] the constant @xmath167 can not be improved .
the next corollary is stated in order to point the relevance of this poincar inequality to the analysis of general order statistics and their functionals .
recall that the _ hazard rate _ of an absolutely continuous probability distribution with distribution @xmath0 is : @xmath168 where @xmath92 and @xmath1 are the density and the survival function associated with @xmath0 , respectively .
[ var : os ] assume the distribution of @xmath169 has a positive density , then the @xmath119th order statistic @xmath170 satisfies @xmath171 \le \frac{c}{k } \left(1 + \frac{1}{k}\right){\ensuremath{\mathbb{e}}}\left [ \frac{1}{h(x_{(k)})^2 } \right]\ ] ] where @xmath172 can be chosen as @xmath167 . by smirnov s lemma @xcite , @xmath172 can not be smaller than @xmath173 .
if the distribution of @xmath169 has a non - decreasing hazard rate , the factor of @xmath167 can be improved into a factor @xmath174 @xcite .
@xcite show that smooth functions of independent exponential random variables satisfy bernstein type concentration inequalities .
the next result is extracted from the derivation of talagrand s concentration phenomenon for product of exponential random variables in @xcite .
the definition of sub - gamma random variables will be used in the formulation of the theorem and in many arguments .
[ dfn : sub - gamma ] a real - valued centred random variable @xmath169 is said to be _ sub - gamma _ on the right tail with variance factor @xmath175 and scale parameter @xmath176 if @xmath177 we denote the collection of such random variables by @xmath178 .
similarly , @xmath169 is said to be sub - gamma on the left tail with variance factor @xmath175 and scale parameter @xmath176 if @xmath179 is sub - gamma on the right tail with variance factor @xmath175 and tail parameter @xmath176 .
we denote the collection of such random variables by @xmath180 and @xmath181 by @xmath182 . if @xmath183 , then for all @xmath184 , with probability larger than @xmath185 @xmath186 the entropy of a non - negative random variable @xmath169 is defined by @xmath187= { \mathbb{e}}[x \ln x]-{\mathbb{e}}x \ln{\mathbb{e}}x$ ] .
[ bernstein : expo ] assume that @xmath87 is a differentiable function on @xmath163 with @xmath188 .
let @xmath164 where @xmath61 are @xmath37 independent standard exponential random variables and @xmath189 .
then , for all @xmath190 such that @xmath191 , @xmath192 \leq \frac{2\lambda^2}{1-c } { \ensuremath{\mathbb{e}}}\left [ { \mathrm{e}}^{\lambda ( z-{\ensuremath{\mathbb{e}}}z ) } \| \nabla g\|^2\right ] \ , .\ ] ] let @xmath175 be the essential supremum of @xmath193 , then @xmath194 is sub - gamma on both tails with variance factor @xmath195 and scale factor @xmath196 .
again , we illustrate the relevance of these versatile tools on the analysis of general order statistics .
this general theorem implies that if the sampling distribution has non - decreasing hazard rate , then the order statistics @xmath170 satisfy bernstein type inequalities ( see * ? ? ?
* section 2.8 ) with variance factor @xmath197 $ ] ( the poincar estimate of variance ) and scale parameter @xmath198 ) .
starting back from the efron - stein - steele inequality , the authors derived a somewhat sharper inequality @xcite .
[ prp : hazard : conc : ineg ] assume the distribution function @xmath0 has non - decreasing hazard rate @xmath199 that is , @xmath200 is @xmath201 and concave .
let @xmath202 be distributed as the @xmath119th order statistic of a sample distributed according to @xmath0 .
then , @xmath194 is sub - gamma on both tails with variance factor @xmath203 $ ] and scale factor @xmath204 .
this corollary describes in which way central , intermediate and extreme order statistics can be portrayed as smooth functions of independent exponential random variables .
this possibility should not be taken for granted as it is non trivial to capture in a non - asymptotic way the tail behaviour of maxima of independent gaussians @xcite . in the next section , we show in which way the hill estimator can fit into this picture .
in this section , the sampling distribution @xmath0 is assumed to belong to @xmath22 with @xmath4 and to satisfy the von mises condition ( definition [ dfn : vmises : cond ] ) with bounded von mises function @xmath48 .
it is well known that , under the von mises condition , if @xmath19 is an intermediate sequence , the sequence @xmath205 converges in distribution towards @xmath206 , suggesting that the variance of @xmath23 scales like @xmath207 ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
proposition [ prop : bound - vari - hill ] provides us with handy non - asymptotic bounds on @xmath208 - \gamma^2/k $ ] using the von mises function .
[ prop : bound - vari - hill ] let @xmath70 be the hill estimator computed from the @xmath209 largest order statistics of an @xmath37-sample from @xmath0 .
then , @xmath210 \leq { \operatorname{var } } [ \widehat{\gamma}(k ) ] - \frac{\gamma^2}{k } \leq \frac{2\gamma}{k } { \ensuremath{\mathbb{e}}}\left [ \overline{\eta}\left({\mathrm{e}}^{y_{(k+1 ) } } \right)\right ] + \frac{5}{k}{\ensuremath{\mathbb{e}}}\left [ \overline{\eta}\left({\mathrm{e}}^{y_{(k+1 ) } } \right)^2\right ] \ , .\ ] ] the next abelian result might help in appreciating these variance bounds .
[ sec : abel - vari - hill ] assuming that @xmath48 is @xmath110-regularly varying with @xmath99 , then , for any intermediate sequence @xmath19 , @xmath211 we may now move to genuine concentration inequalities for the hill estimator .
the exponential representation suggests that the rescaled hill estimator @xmath212 should be approximately distributed according to a @xmath213 distribution where @xmath119 is the shape parameter and @xmath14 the scale parameter .
therefore , we expect the hill estimators to satisfy bernstein type concentration inequalities that is , to be sub - gamma on both tails with variance factors connected to the tail index @xmath14 and to the von mises function .
representation actually suggests more .
following @xcite , we actually expect the sequence @xmath214 to behave like normalized partial sums of independent square integrable random variables that is , we believe @xmath215 to scale like @xmath158 and to be sub - gamma on both tails ( see appendix [ sec : calibr - prel - select ] ) . the purpose of this section is to meet these expectations in a non - asymptotic way .
proofs use the markov property of order statistics : conditionally on the @xmath216th order statistic , the first largest @xmath217 order statistics are distributed as the order statistics of a sample of size @xmath217 of the excess distribution .
they consist of appropriate invocations of talagrand s concentration inequality ( theorem [ bernstein : expo ] ) .
however , this theorem generally requires a uniform bound on the gradient of the relevant function .
when hill estimators are analysed as functions of independent exponential random variables , the partial derivatives depend on the points at which the von mises function is evaluated . in order to get interesting bounds
, it is worth conditioning on an intermediate order statistic . throughout this subsection ,
let @xmath218 be an integer larger than @xmath219 and @xmath217 an integer not larger than @xmath37 .
we denote @xmath220 , @xmath37 independent standard exponential random variables and we work on the probability space where all @xmath221 are defined , and therefore consider the hill estimators defined by representation . as we use the exponential representation of order statistics , besides hill estimators , the random variables that appear in the main statements are order statistics of exponential samples . as before , @xmath222 will denote the @xmath119th order statistic of a standard exponential sample of size @xmath37 ( we agree on @xmath39 ) .
the first theorem provides an exponential refinement of the variance bound stated in proposition [ prop : bound - vari - hill ] .
however , as announced , there is a price to pay : statements hold conditionally on some order statistic .
this is not an impediment to analyse lepski s rule using this theorem . indeed , when analysing lepki s rule it is sufficient to control the hill process @xmath223)\right)_i$ ] for indices @xmath64 ranging between @xmath224 ( that should not be smaller than @xmath144 ) and some upper bound @xmath28 that achieves a certain balance between bias and standard deviation ( the bias of @xmath23 should be of order @xmath139 times the standard deviation , that is approximately @xmath225 where @xmath226 ) .
the second clause of next theorem is the cornerstone in the derivation of the risk bounds presented in the next section . in the sequel ,
let @xmath227 where @xmath228 may be chosen not larger than @xmath167 and @xmath229 not larger than @xmath230 .
[ prp : hill : concentration ] let @xmath231 be a shorthand for @xmath232 . for some @xmath119 such that @xmath233 where @xmath234 , let @xmath235 \right|\ , .\ ] ] then , conditionally on @xmath231 , a. for @xmath236 , @xmath237\right ) \in \gamma_\pm\left(4 \left(\gamma + 3\overline{\eta}(t ) \right)^2,\left(\gamma+2\overline{\eta}(t)\right)\right ) \ , .\ ] ] b. let @xmath238 be such that @xmath239 where @xmath240 with @xmath241 .
assume that @xmath242 , then @xmath243 and @xmath244 \leq \gamma \xi_n \left(1 + \frac{3r_n}{\sqrt{j } } \right)\ , .\ ] ] if @xmath0 is a pure pareto distribution with shape parameter @xmath4 , then @xmath245 is distributed according to a gamma distribution with shape parameter @xmath119 and scale parameter @xmath173 .
tight and well - known tail bounds for gamma distributed random variables assert that @xmath246 \right| \geq \frac{\gamma}{\sqrt{k } } \left ( \sqrt{2\ln\left(2/\delta \right ) } + \frac{\ln \left(2/\delta \right)}{\sqrt{k}}\right ) \right\ } \leq 2 \delta \ , .\ ] ] first part of statement ii ) reads as : conditionally on @xmath247 , with probability larger than @xmath248 , @xmath249\right| \leq \gamma ( 1 + 3r_n /\sqrt{j } ) \left ( \sqrt{8 \ln \left(2/\delta \right ) } + \frac{\ln \left(2/\delta \right)}{\sqrt{\ell}}\right ) \ , . \,\ ] ] combining both parts of statement ii ) , we also get that , conditionally on @xmath247 , with probability larger than @xmath248 , @xmath250 the reader may wonder whether resorting to the exponential representation and usual chernoff bounding would not provide a simpler argument .
the straightforward approach leads to the following conditional bound on the logarithmic moment generating function , @xmath251 \right ) \right ) \mid y_{(k+1)}\right]}\\ & \leq & \frac{\left(\gamma + \overline{\eta}({\mathrm{e}}^{y_{(k+1)}})\right)^2}{2k\left(1-\lambda(\gamma + \overline{\eta}({\mathrm{e}}^{y_{(k+1 ) } } ) ) \right ) } + \lambda \left(\overline{\eta}({\mathrm{e}}^{y_{(k+1)}})-b\left ( { \mathrm{e}}^{y_{(k+1)}}\right )
\right ) \ , . \end{aligned}\ ] ] a similar statement holds for the lower tail .
this leads to exponential bounds for deviations of the hill estimator above @xmath252 + \overline{\eta}({\mathrm{e}}^{y_{(k+1)}})-b\left ( { \mathrm{e}}^{y_{(k+1)}}\right)$ ] that is , to control deviations of the hill estimator above its expectation plus a term that may be of the order of magnitude of the bias .
attempts to rewrite @xmath253 $ ] as a sum of martingale increments @xmath254-{\ensuremath{\mathbb{e}}}[\widehat{\gamma}(k ) \mid y_{(i+1)}]$ ] , for @xmath255 , and to exhibit an exponential supermartingale met the same impediments . at the expense of inflating the variance factor , theorem [ bernstein : expo ]
provides a genuine ( conditional ) concentration inequality for hill estimators .
as we will deal with values of @xmath119 for which bias exceeds the typical order of magnitudes of fluctuations , this is relevant to our purpose .
we are now able to characterise the performance of the variant of the selection rule defined by @xcite with @xmath256 where @xmath257 .
let @xmath258 where @xmath259 is a constant to be defined below .
the deterministic sequence of indices @xmath260 is defined ( for @xmath37 large enough ) by @xmath261 where @xmath262 the sequence @xmath263 is defined by choosing @xmath264 . the deterministic sequences @xmath265 and @xmath260 achieve specific balances between bias and variance . in full generality , because @xmath266 is just an upper bound on the conditional bias @xmath75 , it is difficult to precisely connect @xmath265 and @xmath260 with the oracle sequence @xmath267 .
we call these two sequences the pivotal sequences . in the sequel , @xmath28 stands for @xmath268 . if the context is not clear , we specify @xmath269 or @xmath270
. let @xmath271 .
recall , from section [ sec : conc - ineq - hill ] , that @xmath272 and agree on the shorthands @xmath273 and @xmath274 which is defined by replacing @xmath28 by @xmath224 in the definition of @xmath275 ( @xmath274 depends on @xmath276 but not on the sampling distribution ) . in the sequel
, @xmath259 is assumed to be chosen so that @xmath277 for @xmath278 and @xmath279 ( @xmath259 may be chosen not larger than 100 ) . the index @xmath280 is selected according to the following rule : @xmath281 where @xmath282 .
the quantity @xmath283 scales like @xmath284 .
the tail index estimator is @xmath285 . as tail adaptivity has a price ( see theorem [ thm - kc - lower - bound ] )
, the ratio between the risk of the data - driven estimator @xmath285 and the risk of the pivotal index @xmath286 can not be upper bounded by a constant factor , let alone by a factor close to @xmath173 .
this is why in the next theorem , we compare the risk of the empirically selected index @xmath285 with the risk of the pivotal index @xmath23 .
recall , from section [ sec : conc - ineq - hill ] , that @xmath287 [ thm : adapt - hill - estim ] assume the sampling distribution @xmath288 satisfies the von mises condition with bounded von mises function @xmath48 , and @xmath289 let @xmath290 is large enough so that @xmath28 ( definition [ def : kn ] ) is well defined .
then , for @xmath291 , with probability larger than @xmath292 , @xmath293 and , with probability larger than @xmath294 , @xmath295 where @xmath296 for @xmath297 , @xmath298 if the bias @xmath73 is @xmath110-regularly varying ( or equivalently , if the von mises function @xmath48 or even @xmath54 are regularly varying ) , then , elaborating on proposition 1 from @xcite , sequences @xmath299 and @xmath265 are connected by @xmath300 and their quadratic risk are related by @xmath301}{{\ensuremath{\mathbb{e}}}[(\gamma-\widehat{\gamma}(k_n^*))^2 ] } = \frac{2}{2|\rho|+1}(2|\rho|)^{2|\rho|/(1 + 2|\rho| ) } \ , .\ ] ] moreover , under the second - order assumption , the two pivotal sequences @xmath265 and @xmath260 are also connected .
thus , if the bias is @xmath110-regularly varying , theorem [ thm : adapt - hill - estim ] provides us with a connection between the performance of the simple selection rule and the performance of the ( asymptotically ) optimal choice .
recall that one of the main aims of this paper is to derive performance guarantees for the data - driven index selection method @xmath280 without resorting to second - order assumptions that is , without assuming that the von mises function is regularly varying .
the next corollary upper bounds the risk of the preliminary estimator when we just have an upper bound on the bias .
[ cor : adapt - hill - estim-2ndrv ] assume that , for some @xmath97 and @xmath99 , for all @xmath302 , @xmath303 then , there exists a constant @xmath304 depending on @xmath305 and @xmath110 such that , with probability larger than @xmath294 , @xmath306 where @xmath307 is defined in theorem [ thm : adapt - hill - estim ] .
this meets the information - theoretic lower bound of theorem [ thm - kc - lower - bound ] .
this proposition is a straightforward consequence of rnyi s representation of order statistics of standard exponential samples .
as @xmath0 belongs to @xmath22 and meets the von mises condition , there exists a function @xmath48 on @xmath31 with @xmath308 such that @xmath309 and @xmath310 then , @xmath311 let @xmath312 . by the pythagorean relation , @xmath313 + { \operatorname{var}}\left ( { \ensuremath{\mathbb{e}}}[z \mid y_{(k+1)}]\right ) \ , .\ ] ] representation
asserts that , conditionally on @xmath68 , @xmath194 is distributed as a sum of independent , exponentially distributed random variables .
let @xmath34 be an exponentially distributed random variable .
@xmath314 where we have used the cauchy - schwarz inequality and @xmath315 taking expectation with respect to @xmath68 leads to @xmath316 \leq k{\gamma^2 } + 2k\gamma{\ensuremath{\mathbb{e}}}\left [ \overline{\eta}\left({\mathrm{e}}^{y_{(k+1 ) } } \right)\right ] + k{{\ensuremath{\mathbb{e}}}\left [ \overline{\eta}\left({\mathrm{e}}^{y_{(k+1 ) } } \right)^2\right ] } \ , .\ ] ] the last term in the pythagorean decomposition is also handled using elementary arguments .
@xmath317 = k \gamma + k\int_0^\infty{\mathrm{e}}^{-u } \eta\left({\mathrm{e}}^{u + y_{(k+1)}}\right ) \mathrm{d}u \ , .\end{aligned}\ ] ] as @xmath68 is a function of independent exponential random variables ( @xmath318 ) , the variance of @xmath319 $ ] may be upper bounded using poincar inequality ( proposition [ poincare : expo ] ) @xmath320\right ) \leq 4k { \ensuremath{\mathbb{e}}}\left [ \overline{\eta}\left({\mathrm{e}}^{y_{(k+1 ) } } \right)^2\right]\ , .\ ] ] in order to derive the lower bound , we first observe that @xmath321 \ , .\ ] ] now , using cauchy - schwarz inequality again , @xmath322 in the proof of theorem [ prp : hill : concentration ] , we will use the next maximal inequality ( see * ? ? ?
* corollary 2.6 ) . recall the definition of @xmath178 ( definition [ dfn : sub - gamma ] ) .
[ zlapcor ] let @xmath323 be real - valued random variables belonging to @xmath178 .
then @xmath324 \leq \sqrt{2v\ln n}+ c\ln n~.\ ] ] proofs follow a common pattern . in order to check that some random variable is sub - gamma ,
we rely on its representation as a function of independent exponential variables and compute partial derivatives , derive convenient upper bounds on the squared euclidean norm and the supremum norm of the gradient and then invoke theorem [ bernstein : expo ] . at some point
, we will use the next corollary of theorem [ bernstein : expo ] .
[ sec : proof - prop - refprp : technique ] if @xmath87 is an almost everywhere differentiable function on @xmath325 with uniformly bounded derivative @xmath326 , then @xmath327 is sub - gamma with variance factor @xmath328 and scale factor @xmath329 we start from the exponential representation of hill estimators ( proposition [ hill : rep ] ) and represent all @xmath131 as functions of independent random variables @xmath330 where the @xmath331 , are standard exponentially distributed and @xmath332 is distributed like the @xmath216th largest order statistic of an @xmath37-sample of the standard exponential distribution .
we consistently use the notation @xmath333 , for @xmath334 . @xmath335 let @xmath336 be such that @xmath337 , let us agree on @xmath338 let @xmath339 for @xmath340 , as @xmath341 for @xmath342 and @xmath83 otherwise , @xmath343 this entails that , for @xmath344 , @xmath345 for @xmath346 , @xmath347 this is enough to entail that , for @xmath348 , @xmath349 all in all , for @xmath350 , @xmath351 [ [ proof - of - i ] ] proof of i ) + + + + + + + + + + + an upper bound on the variance factor for @xmath352 , conditionally on @xmath231 , is obtained by specialising to the case @xmath353 and using and as well as the monotonicity of @xmath54 , @xmath354 using theorem [ bernstein : expo ] conditionally on @xmath355 , we realise that @xmath356)$ ] is sub - gamma on both sides with variance factor not larger than @xmath357 and scale factor not larger than @xmath358 .
this yields & \ { ( i)-@xmath359 ( + ) t } 2 ^-s . & \ { ( i)-@xmath359 ( + ) } 2 ^-s .
[ [ par : proof_of_ii _ ] ] proof of ii ) + + + + + + + + + + + + the proof of the upper bound on @xmath360 $ ] in statement ii ) from theorem [ prp : hill : concentration ] relies on standard chaining techniques from the theory of empirical processes and uses repeatedly the concentration theorem [ bernstein : expo ] for smooth functions of independent exponential random variables and the maximal inequality for sub - gamma random variables ( proposition [ zlapcor ] ) . for general
@xmath336 , the variance factor for @xmath361 is upper bounded by @xmath362 let @xmath238 be such that @xmath363 where @xmath364 with @xmath257 .
now , as we assume , in the sequel , that @xmath365 , we may use the next upper bound for the variance factor of @xmath366 ( conditionally on @xmath367),@xmath368 recall that @xmath369 \right|\ , .\ ] ] as it is commonplace in the analysis of normalised empirical processes ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) , we peel the index set over which the maximum is computed .
let @xmath370 and , for all @xmath371 , @xmath372 .
define @xmath373 as @xmath374 \right|\ , .\ ] ] then , @xmath375 & = & { \ensuremath{\mathbb{e}}}[\max_{j\in \mathcal{l}_n } z^a_j \mid y_{(k+1 ) } ] \\
& \leq & { \ensuremath{\mathbb{e}}}[\max_{j\in \mathcal{l}_n } ( z^a_j-{\ensuremath{\mathbb{e}}}[z^a_j\mid y_{(k+1 ) } ] ) \mid y_{(k+1 ) } ] ] + \max_{j\in \mathcal{l}_n } { \ensuremath{\mathbb{e } } } [ z^a_j\mid y_{(k+1 ) } ] ] \ , .
\end{aligned}\ ] ] we now derive upper bounds on both summands by resorting to the maximum inequality for sub - gamma random variables ( proposition [ zlapcor ] ) .
we first bound @xmath376 $ ] , for @xmath377 .
note that direct invocation of lemma [ zlapcor ] and statement i ) shows that @xmath378 \leq 2 \gamma(1 + 3r/\sqrt{j})\left ( \sqrt{8 j \ln(2 ) } + j \ln(2)\right ) \ , .\ ] ] this bound will be useful for handling small values of @xmath379 .
for @xmath380 , @xmath381 .
we now handle generic @xmath379 using chaining .
fix @xmath377 , @xmath382 \right| \leq \frac{1}{2^{j/2 } } \max_{i \in \mathcal{s}_j } \ , i\left| \widehat{\gamma}(i ) - { \ensuremath{\mathbb{e}}}[\widehat{\gamma}(i ) \mid y_{(k+1 )
} ] \right| \ , .\ ] ] in order to alleviate notation , let @xmath383 \right)$ ] , for @xmath384 . for @xmath384 ,
let @xmath385 be the binary expansion of @xmath64 .
then , for @xmath386 , let @xmath387 be defined by @xmath388 so that @xmath389 , @xmath390 and @xmath391 .
using the fact that @xmath392 does not depend on @xmath64 and that @xmath393=0\ , , \ ] ] we obtain @xmath394 \right)\mid y_{(k+1 ) } \right]}\\ & = & { \ensuremath{\mathbb{e}}}\left [ \max_{i \in \mathcal{s}_j } w(i ) \mid y_{(k+1 ) } \right ] \\ & = & { \ensuremath{\mathbb{e}}}\left [ \max_{i \in \mathcal{s}_j } w(\pi_j(i ) ) -w(\pi_0(i))\mid y_{(k+1 ) } \right ] \\ & = & { \ensuremath{\mathbb{e}}}\left [ \max_{i \in \mathcal{s}_j } \sum_{h=0}^{j-1 } ( w(\pi_{h+1}(i ) ) -w(\pi_h(i)))\mid y_{(k+1 ) } \right ] \\ & \leq & \sum_{h=0}^{j-1 } { \ensuremath{\mathbb{e}}}\left [ \max_{i \in \mathcal{s}_j } ( w(\pi_{h+1}(i ) ) -w(\pi_h(i ) ) ) \mid y_{(k+1 ) } \right ] \ , .\end{aligned}\ ] ] now , for each @xmath395 , the maximum is taken over @xmath396 random variables which are sub - gamma with variance factor @xmath397 and scale factor @xmath398 . by proposition [ zlapcor ] , since @xmath399 , @xmath394 \right)\mid y_{(k+1 ) } \right]}\\ & \leq & \gamma \sum_{h=0}^{j-1 } \left(\left(1 + \frac{r_n}{\sqrt{j}}\right ) \sqrt{8 h 2^{(j - h-1)}\ln 2 } + \sqrt{32 h \ln 2 } r_n+ \left(1 + \frac{2r_n}{\sqrt{j}}\right ) h \ln 2\right ) \\ & \leq & \gamma \left(1 + \frac{2r_n}{\sqrt{j}}\right ) \left ( 2^{(j-1)/2 } 4.15 \sqrt{8 \ln 2 } + \frac{2}{3 } \sqrt{32 c_2 } \ln(2 ) j^{2}+ \frac{j(j-1)}{2 } \ln 2\right ) \end{aligned}\ ] ] where we have used @xmath400 for @xmath401 ,
@xmath402 \right)\mid y_{(k+1 ) } \right ] } & \leq & 17 \gamma \ , 2^{j/2 } \left(1 + \frac{2r_n}{\sqrt{j}}\right)\ , . \end{aligned}\ ] ] finally , for all @xmath403 , @xmath404 \leq 34 \ , \gamma \ , \left(1 + \frac{3r_n}{\sqrt{j}}\right ) \,.\ ] ] in order to prove statement ii ) , we check that , for each @xmath403 , @xmath373 is sub - gamma on the right - tail with variance factor at most @xmath405 and scale factor not larger than @xmath406 . under the von mises condition ( definition [ dfn : vmises : cond ] ) ,
the sampling distribution is absolutely continuous with respect to lebesgue measure . for almost every sample ,
the maximum defining @xmath373 is attained at a single index @xmath384 .
starting again from the exponential representation and repeating the computation of partial derivatives , we obtain the desired bounds . by proposition [ zlapcor ] , @xmath407 ) \mid y_{(k+1)}\right ] } \\ & \leq & \left ( \sqrt{8\ln |\mathcal{l}_n| } + \frac { \ln |\mathcal{l}_n|}{\sqrt{\ell}}\right)\left(\gamma + 3 \overline{\eta}(t ) \right ) \\ & \leq & 4 \sqrt{\ln |\mathcal{l}_n| } \left(\gamma + 3 \overline{\eta}(t ) \right ) \\ & \leq & 4 \ , \gamma\ , \sqrt{\ln |\mathcal{l}_n| } \left(1 + \frac{3r_n}{\sqrt{j } } \right ) \ , \end{aligned}\ ] ] where we have used @xmath408 , for @xmath409 . combining the different bounds leads to the upper bound on @xmath360 $ ] . throughout this proof ,
let @xmath410 let us define the events @xmath411 and @xmath412 as @xmath413\right|\leq \gamma z_\delta \big\}\ , , \\
e_2 & = & \big\ { t_n \geq \frac{n}{k_n^\delta } \big\ } \text { with } k_n^\delta=
k_n+2{\ln \left(1/\delta \right)}+\sqrt{2k_n \ln ( 1/\delta)}{\enspace .}\end{aligned}\ ] ] the fact that @xmath414 follows from the following reformulation of proposition 4.3 from @xcite ( a proof is given in appendix [ proof : prop : right : tail : order : stat ] ) .
[ prop : right : tail : order : stat ] for @xmath184 , with probability larger that @xmath248 , @xmath415 where @xmath68 is the @xmath209th largest order statistic of an exponential sample of size @xmath37 . by theorem
[ prp : hill : concentration ] , @xmath416 .
hence , the event @xmath417 has probability at least @xmath418 . under @xmath419 , a. @xmath420 .
b. for all @xmath421 , @xmath422|\leq \overline{\eta}(t_n).$ ] the first step of the proof consists in checking that under @xmath417 , the selected index is not smaller than @xmath28 .
it suffices to check that for all @xmath423 , @xmath424 for all @xmath425 , @xmath426\right| + \left|\widehat{\gamma}(i ) -{\ensuremath{\mathbb{e}}}[\widehat{\gamma}(i)\mid t_n]\right|\\ & \leq & \overline{\eta}(t_n ) + \frac{\gamma z_\delta}{\sqrt{i } } \\ & \leq & \frac{\gamma r_n}{\sqrt{k_n } } + \frac{\gamma z_\delta}{\sqrt{i } } \end{aligned}\ ] ] so that @xmath427 meanwhile , for all @xmath428 , @xmath429\big|}_{\textsc{(i)}}+ \underbrace{\left| { \ensuremath{\mathbb{e}}}[\widehat{\gamma}(i)-\widehat{\gamma}(k)\mid t_n]\right|}_{\textsc{(ii ) } } + \underbrace{\left| \widehat{\gamma}(k)-{\ensuremath{\mathbb{e}}}[\widehat{\gamma}(k)\mid t_n]\right|}_{\textsc{(iii ) } } \ , .
\end{aligned}\ ] ] under @xmath417 , for @xmath430 , @xmath431 under @xmath412 , @xmath432\right| + \left| { \ensuremath{\mathbb{e } } } [ \gamma-\widehat{\gamma}(k)\mid t_n]\right| \\ & \leq & 2 \overline{\eta}(t_n ) \\ & \leq & 2 \gamma r_n /\sqrt{k_n } \ , .
\end{aligned}\ ] ] plugging upper bounds on ( i ) , ( ii ) and ( iii ) , it comes that , under @xmath417 , for all @xmath433 and for all @xmath434 , @xmath435 in order to warrant that , under @xmath417 , for all @xmath436 and for all @xmath64 such that @xmath437 , @xmath438 , it is enough to have @xmath439 the last inequality holds because @xmath440 by definition of @xmath283 .
hence , with probability larger than @xmath441 , @xmath417 is realised , and under @xmath417 , @xmath442 .
we now check that if @xmath443 , the risk of @xmath285 is not much larger than the risk of @xmath23 .
@xmath444 therefore , under @xmath445 , @xmath446 now , consider the event @xmath447 with @xmath448\right| \leq \left(\gamma+3 \overline{\eta}(t_n ) \right ) \big ( \sqrt{8 \ln \left(2/\delta \right ) } + \frac{\ln \left(2/\delta \right)}{\sqrt{k_n}}\big)\bigg\ } { \enspace .}\ ] ] since , @xmath449 , thanks to statement i ) from theorem [ prp : hill : concentration ] , the event @xmath447 has probability at least @xmath450 . then , by definition of @xmath28 , under @xmath412 , @xmath451\right|\leq \overline{\eta}(t_n ) \leq
\gamma r_n/\sqrt{k_n } \ , .\ ] ] hence , under @xmath452 , @xmath453|+ | \widehat{\gamma}({k}_n ) -\mathbb{e}[\widehat{\gamma}({k}_n)\mid t_n]| \\ & \leq & \frac{\gamma}{\sqrt{k_n}}\left(r_n + \left(1 + \frac{3r_n}{\sqrt{k_n}}\right)\big ( \sqrt{8 \ln \left(2/\delta \right ) } + \frac{\ln \left(2/\delta \right)}{\sqrt{k_n}}\big )
\right ) \ , . \end{aligned}\ ] ] therefore , plugging this bound into , with probability larger than @xmath292 , @xmath454 where @xmath455 if , for some @xmath97 and @xmath99 , @xmath456 then , by the definition of @xmath28 , @xmath457 which entails that @xmath458 solving this inequality leads to @xmath459 and finally to @xmath460 thus , for sufficiently large @xmath37 , there exists a constant @xmath176 depending on @xmath461 such that @xmath462 starting from equation of theorem [ thm : adapt - hill - estim ] , with probability @xmath292 , @xmath463 and , there exists a constant @xmath464 , depending on @xmath305 and @xmath110 , such that @xmath465 hence , with probability larger than @xmath294 , @xmath466
risk bounds like theorem [ thm : adapt - hill - estim ] and corollary [ cor : adapt - hill - estim-2ndrv ] are conservative . for all practical purposes ,
they are just meant to be reassuring guidelines . in this numerical section ,
we intend to shed some light on the following issues : 1 .
is there a reasonable way to calibrate the threshold @xmath283 used in the definition of @xmath280 ?
how does the method perform if we choose @xmath283 close to @xmath467 ?
how large is the ratio between the risk of @xmath285 and the risk of @xmath468 for moderate sample sizes ?
the finite - sample performance of the data - driven index selection method described and analysed in section [ sec : adapt - hill - estim ] has been assessed by monte - carlo simulations .
computations have been carried out in ` r ` using packages ` ggplot2 ` @xcite , ` knitr ` , ` foreach ` , ` iterators ` , ` xtable ` and ` dplyr ` ( see * ? ? ? * for a modern account of the r environment ) . to get into the details , we investigated the performance of index selection methods on samples of sizes @xmath469 and @xmath470 from the collection of distributions listed in table [ tab - risk - profiles ] .
the list comprises the following distributions a. frchet distributions @xmath471 for @xmath3 and @xmath472 . b. student distributions @xmath473 with @xmath474 degrees of freedom .
c. the log - gamma distribution with density proportional to @xmath475 , which means @xmath476 and @xmath477 .
d. the lvy distribution with density @xmath478 , @xmath479 and @xmath480 ( this is the distribution of @xmath481 when @xmath482 ) .
e. the @xmath483 distribution is defined by @xmath484 and von mises function equal to @xmath485 .
this distribution satisfies the second - order regular variation condition with @xmath486 but does not satisfy condition .
f. two pareto change point distributions with distribution functions @xmath487 and @xmath488 , @xmath489 , and thresholds @xmath7 adjusted in such a way that they correspond to quantiles of order @xmath490 and @xmath491 , respectively .
frchet , student , log - gamma distributions were used as benchmarks by @xcite , @xcite and @xcite .
table [ tab - risk - profiles ] , which is complemented by figure [ fig : risk - comp ] , describes the difficulty of tail index estimation from samples of the different distributions .
monte - carlo estimates of the standardised root mean square error ( rmse ) of hill estimators @xmath492^{1/2}\ ] ] are represented as functions of the number of order statistics @xmath119 for samples of size @xmath470 from the sampling distributions .
all curves exhibit a common pattern : for small values of @xmath119 , the rmse is dominated by the variance term and scales like @xmath493 . above a threshold that depends on the sampling distribution but that is not completely characterised by the second - order regular variation index , the rmse grows at a rate that may reflect the second - order regular variation property ( if any ) of the distribution .
not too surprisingly , the three frchet distributions exhibit the same risk profile .
the three curves are almost undistinguishable .
the student distributions illustrate the impact of the second - order parameter on the difficulty of the index selection problem . for sample size @xmath494 ,
the optimal index for @xmath495 is smaller than @xmath496 , it is smaller than the usual recommendations . for such moderate sample sizes ,
distribution @xmath495 seems as hard to handle as the @xmath497-gamma distribution which usually fits in the horror hill plot gallery .
the @xmath498-stable lvy distribution and the @xmath483-distribution behave very differently .
even though they both have second - order parameter @xmath110 equal to @xmath499 , the @xmath483 distribution seems almost as challenging as the @xmath500 distribution while the lvy distribution looks much easier than the frchet distributions .
the pareto change point distributions exhibit an abrupt transition .
.estimated oracle index @xmath501 and standardised rmse @xmath502^{1/2}/\gamma$ ] for benchmark distributions .
estimates were computed from @xmath123 replicated experiments on samples of size @xmath503 [ cols="<,^,^,^,^",options="header " , ] figure [ fig : pcp - risk - plot ] concisely describes the behaviour of the two index selection methods on samples from the pareto change point distribution with parameters @xmath504 and threshold @xmath7 corresponding to the @xmath490 quantile .
the plain line represents the standardised rmse of hill estimators as a function of selected index .
this figure contains the superposition of two density plots corresponding to @xmath505 and @xmath506 .
the density plots were generated from @xmath123 points with coordinates latexmath:[$(\widehat{k}(r_n ) , points with coordinates latexmath:[$(\widehat{k}^{\textsc{dk}}_n , contoured and well - concentrated density plot corresponds to the performance of @xmath285 .
the diffuse tiled density plot corresponds to the performance of @xmath505 .
facing pareto change point samples , the two selection methods behave differently .
lepski s rule detects correctly an abrupt change at some point and selects an index slightly above that point .
as the conditional bias varies sharply around the change point , this slight over estimation of the correct index still results in a significant loss as far as rmse is concerned .
the drees - kaufmann rule , fed with an a priori estimate of the second - order parameter , picks out a much smaller index , and suffers a larger excess risk . from the pareto change point distribution with parameters @xmath504 and threshold @xmath7 corresponding to the @xmath490 quantile .
the concentrated density plot corresponds to points @xmath508 . ]
the authors are thankful to the editor and the referees for their careful reading and valuable suggestions , which led to detect an error and to a improved version of the paper .
50 [ 1]#1 [ 1]`#1 ` urlstyle [ 1]doi : # 1 j. beirlant , y. goegebeur , j. teugels , and j. segers . _
statistics of extremes_. john wiley & sons , ltd . , 2004 .
j. beirlant , c. bouquiaux , and b. werker .
semiparametric lower bounds for tail index estimation
. _ journal of statistical planning and inference _ , 1360 ( 3):0 705729 , 2006 .
n. bingham , c. goldie , and j. teugels .
_ regular variation_. cambridge university press , 1987 .
l. birg .
an alternative point of view on lepski s method .
in _ state of the art in probability and statistics ( leiden , 1999 ) _ , volume 36 of _ ims lecture notes monogr .
_ , pages 113133 .
, 2001 .
l. birg .
a new lower bound for multiple hypothesis testing . _
ieee trans .
inform . theory _ , 51:0 16111615 , 2005 .
s. bobkov and m. ledoux .
poincar s inequalities and talagrand s concentration phenomenon for the exponential distribution .
. fields _ , 107:0 383400 , 1997 .
s. boucheron and m. thomas .
concentration inequalities for order statistics .
_ , 17:0 112 , 2012 .
s. boucheron , g. lugosi , and p. massart .
_ concentration inequalities_. oxford university press , 2013 .
a. carpentier and a. kim .
adaptive confidence intervals for the tail coefficient in a wide second order class of pareto models .
_ , 8:0 20662110 , 2014 .
a. carpentier and a. kim .
adaptive and minimax optimal estimation of the tail coefficient .
_ statistica sinica _ , 25:0 11331144 , 2015 .
s. chatterjee .
_ superconcentration and related topics_. springer - verlag , 2014 .
t. cover and j. thomas .
_ elements of information theory_. john wiley , 1991 .
s. csrg , p. deheuvels , and d. mason .
kernel estimates of the tail index of a distribution .
_ , 130 ( 3):0 10501077 , 1985 .
j. danielsson , l. de haan , l. peng , and c. g. de vries . using a bootstrap method to choose the sample fraction in tail index estimation .
_ j. multivariate anal .
_ , 760 ( 2):0 226248 , 2001 . d. darling and p. erds . a limit theorem for the maximum of normalized sums of independent random variables
_ duke math .
j _ , 23:0 143155 , 1956 .
l. de haan and a. ferreira .
_ extreme value theory_. springer - verlag , 2006 .
g. draisma , l. de haan , l. peng , and t. pereira . a bootstrap - based method to achieve optimally in estimating the extreme value index .
_ extremes _ , 2:0 367404 , 1999 .
h. drees .
optimal rates of convergence for estimates of the extreme value index .
_ , 260 ( 1):0 434448 , 1998 .
h. drees . on smooth statistical tail functionals .
j. statist .
_ , 250 ( 1):0 187210 , 1998 .
h. drees .
minimax risk bounds in extreme value theory .
_ , 290 ( 1):0 266294 , 2001 .
h. drees and e. kaufmann . selecting the optimal sample fraction in univariate extreme value estimation .
_ stochastic process .
_ , 750 ( 2):0 149172 , 1998 .
h. drees , l. de haan , and s. resnick . .
_ , 280 ( 1):0 254274 , 2000 .
j. geluk , l. de haan , s. resnick , and c. stric .
second - order regular variation , convolution and the central limit theorem .
_ stochastic process .
_ , 690 ( 2):0 139159 , 1997 . e. gin and v. koltchinskii .
concentration inequalities and asymptotic results for ratio type empirical processes .
_ , 340 ( 3):0 11431216 , 2006 .
i. grama and v. spokoiny .
statistics of extremes by oracle estimation .
_ , 360 ( 4):0 16191648 , 2008 .
p. hall and i. weissman . on the estimation of extreme tail probabilities .
_ , 250 ( 3):0 13111326 , 1997 .
p. hall and a. welsh .
adaptive estimates of parameters of regular variation .
_ , 130 ( 1):0 331341 , 1985
. b. hill . a simple general approach to inference about the tail of a distribution . _ ann .
_ , 3:0 11631174 , 1975 .
v. koltchinskii . _ oracle inequalities in empirical risk minimization and sparse recovery problems .
ecole det de probabilit de saint - flour xxxviii _ ,
volume 2033 of _ lecture notes in math._. springer - verlag , 2008 .
m. ledoux . _
the concentration of measure phenomenon_. american mathematical society , 2001 .
m. ledoux and m. talagrand .
_ probability in banach space_. springer - verlag , 1991 .
o. lepski . a problem of adaptive estimation in gaussian white noise .
_ teoriya veroyatnosteui
i ee primeneniya _ , 350 ( 3):0 459470 , 1990 . o. lepski .
asymptotically minimax adaptive estimation . i. upper bounds .
optimally adaptive estimates . _ teoriya veroyatnosteui
i ee primeneniya _ , 360 ( 4):0 645659 , 1991 .
o. lepski .
asymptotically minimax adaptive estimation .
schemes without optimal adaptation .
adaptive estimates .
_ teoriya veroyatnosteui i ee primeneniya _ , 370 ( 3):0 468481 , 1992 .
o. lepski and a. tsybakov .
asymptotic exact nonparametric hypothesis testing in sup - norm and at a fixed point .
theory rel .
fields _ , 1170 ( 1):0 1748 , 2000 . d. mason
. laws of large numbers for sums of extreme values .
_ , 10:0 754764 , 1982 .
p. massart .
_ concentration inequalities and model selection .
ecole det de probabilit de saint - flour xxxiv _ , volume 1896 of _ lecture notes in math._. springer - verlag , 2007 .
p. math .
the lepski principle revisited . _ inverse problems _ , 220 ( 3):0 l11l15 , 2006 .
b. maurey .
some deviation inequalities . _ geometric and functional analysis _ , 10 ( 2):0 188197 , 1991 .
s. novak .
lower bounds to the accuracy of inference on heavy tails .
_ bernoulli _ , 200 ( 2):0 979989 , 2014 .
s. resnick .
_ heavy - tail phenomena : probabilistic and statistical modeling _
, springer - verlag , 2007 .
j. segers .
abelian and tauberian theorems on the bias of the hill estimator .
j. statist .
_ , 290 ( 3):0 461483 , 2002 .
m. talagrand . a new isoperimetric inequality and the concentration of measure phenomenon . in _ geometric aspects of functional analysis ( 198990 )
_ , volume 1469 of _ lecture notes in math .
_ , pages 94124 .
springer - verlag , 1991 .
m. talagrand . a new look at independence .
_ , 24:0 134 , 1996 .
m. talagrand .
new concentration inequalities in product spaces . _ inventiones mathematicae _ , 126:0 505563 , 1996 .
m. talagrand . _ the generic chaining_. springer - verlag , 2005 .
a. b. tsybakov .
pointwise and sup - norm sharp adaptive estimation of functions on the sobolev classes .
_ , 260 ( 6):0 24202469 , 1998
. s. van de geer .
_ applications of empirical process theory_. cambridge university press , 2000 .
h. wickham .
_ ggplot2 : elegant graphics for data analysis_. springer - verlag , 2009 .
h. wickham .
_ advanced r_. chapman & hall / crc , 2014 .
let @xmath509 . then , @xmath510 and @xmath511 let @xmath189 , then for all @xmath512 , @xmath513\lambda^2 } { 2(1-c ) } \ , .\ ] ] now , start from the first statement in theorem [ bernstein : expo ] , @xmath514 & \leq & \frac{2\lambda^2}{1-c } { \ensuremath{\mathbb{e}}}\left [ { \mathrm{e}}^{\lambda ( z-{\ensuremath{\mathbb{e}}}z ) } \| \nabla f\|^2\right ] \\ & = & \frac{4\lambda^2}{2(1-c ) } \frac{1}{k } \left ( 1 + \frac{1}{k}\right ) { \ensuremath{\mathbb{e}}}\left [ \frac{{\mathrm{e}}^{\lambda ( z-{\ensuremath{\mathbb{e}}}z)}}{h(z)^2}\right ] \\ & \leq & \frac{4\lambda^2}{2(1-c ) } \frac{1}{k } \left ( 1 + \frac{1}{k}\right ) { \ensuremath{\mathbb{e}}}\left [ { { \mathrm{e}}^{\lambda ( z-{\ensuremath{\mathbb{e}}}z)}}\right ] { \ensuremath{\mathbb{e}}}\left [ \frac{1}{h(z)^2}\right]\end{aligned}\ ] ] where the last inequality follows from chebychev negative association inequality .
hence , @xmath515 = \frac{\operatorname{ent } \left [ { \mathrm{e}}^{\lambda ( z-{\ensuremath{\mathbb{e}}}z)}\right]}{\lambda^2 { \ensuremath{\mathbb{e}}}\left [ { \mathrm{e}}^{\lambda ( z-{\ensuremath{\mathbb{e}}}z)}\right ] } \leq \frac{1}{2(1-c ) } \frac{4}{k } \left ( 1+\frac{1}{k}\right ) { \ensuremath{\mathbb{e}}}\left [ \frac{1}{h(z)^2}\right ] \ , .\ ] ] this differential inequality is readily solved and leads to the corollary .
the proof proceeds by classical arguments . in the sequel
, we use the almost sure representation argument . without loss of generality
, we assume that all the random variables live on the same probability space , and that , for any intermediate sequence @xmath19 , @xmath516 converges almost surely towards a standard gaussian random variable . complemented with dominated convergence arguments
, the next lemma will be the key element of the proof .
[ lem : rvplus ] let @xmath517 and @xmath518 be the @xmath519th largest order statistic of a standard exponential sample , then , for any intermediate sequence @xmath19 and @xmath520 , @xmath521 note that @xmath522 then , the result follows since @xmath523 and the convergence @xmath524 is locally uniform on @xmath525 . in order to secure dominated convergence arguments , we will use drees s improvement of potter s inequality ( see * ? ? ?
for every @xmath526 , there exists @xmath527 such that , for @xmath528 , @xmath529 to prove proposition [ sec : abel - vari - hill ] , we start from representation : @xmath530 by the pythagorean relation , @xmath531\right)+{\ensuremath{\mathbb{e}}}\left [ { \operatorname{var}}\left ( \widehat{\gamma}(k_n ) \mid y_{(k_n+1)}\right)\right ] \ , , \ ] ] so that @xmath532\right)}{\eta(n / k_n ) } + k_n { \ensuremath{\mathbb{e}}}\left [ \frac{{\operatorname{var}}\left ( \widehat{\gamma}(k_n ) \mid y_{(k_n+1)}\right)-\frac{\gamma^2}{k_n}}{\eta(n / k_n ) } \right ] \ , .
\end{aligned}\ ] ] the second summand can be further decomposed using .
@xmath533\right)}{\eta(n / k_n)}}_{\textsc{(i ) } } \\ & & + \underbrace{\eta ( \tfrac{n}{k_n}){\ensuremath{\mathbb{e}}}\left[{\operatorname{var}}\left [ \int_0^e \frac{\eta({\mathrm{e}}^{u+y_{(k_n+1 ) } } ) } { \eta ( n / k_n ) } { \mathrm{d}}u \mid y_{(k_n+1 ) } \right]\right]}_{(\textsc{ii})}\\ & & + \underbrace{2\gamma { \ensuremath{\mathbb{e}}}\left[{\operatorname{cov}}\left [ e,\int_0^e \frac{\eta({\mathrm{e}}^{u+y_{(k_n+1)}})}{\eta ( n / k_n)}{\mathrm{d}}u \mid y_{(k_n+1)}\right]\right]}_{(\textsc{iii } ) } .
\end{aligned}\ ] ] we check that ( i ) and ( ii ) tend to @xmath83 and then that ( iii ) converges towards a finite limit . fix @xmath526 and define @xmath534 .
+ let @xmath535 denote the event @xmath536 . for @xmath37 such that @xmath537 , as @xmath518 is sub - gamma with variance factor @xmath538 , @xmath539 we first check that ( ii ) tends to @xmath83 .
let @xmath37 be such that @xmath540 and @xmath541 denote the random variable @xmath542 .
note that , for @xmath543 , @xmath544 using jensen s inequality and fubini s theorem , @xmath545\right]}\\ & \leq & { \ensuremath{\mathbb{e}}}\left[{\ensuremath{\mathbb{e}}}\left [ e \int_0^e \left ( \frac{\eta({\mathrm{e}}^{u+y_{(k_n+1)}})}{\eta ( n / k_n)}\right)^2{\mathrm{d}}u \mid y_{(k_n+1 ) } \right]\right ] \\ & = & \int_0^\infty { \mathrm{e}}^{-v } v \int_0^v { \ensuremath{\mathbb{e}}}\left[\left ( \frac{\eta({\mathrm{e}}^{u+y_{(k_n+1)}})}{\eta ( n / k_n)}\right)^2 \right ] { \mathrm{d}}u \mathrm{d}v \\ & = & \int_0^\infty { \mathrm{e}}^{-v } v \int_0^v { \ensuremath{\mathbb{e}}}\left[\left ( \frac{\eta({\mathrm{e}}^{u+w_n}n / k_n)}{\eta ( n / k_n)}\right)^2 \right ] { \mathrm{d}}u \mathrm{d}v \end{aligned}\ ] ] we now apply potter s inequality ( [ potter : drees ] ) on the event @xmath546 with @xmath547 and @xmath548 : @xmath549\right ] } \\ & \leq & \int_0^\infty { \mathrm{e}}^{-v } v \int_0^v { \ensuremath{\mathbb{e}}}\left [ \mathbb{1}_{a_{n } } { \mathrm{e}}^{2\rho ( u+w_n ) } \left ( 1+\epsilon { \mathrm{e}}^{\delta ( u+|w_n| ) } \right)^2 + \mathbb{1}_{a_n^c } \frac{m^2}{\eta(n / k_n)^2 } \right ] { \mathrm{d}}u \mathrm{d}v \\ & \leq & \int_0^\infty { \mathrm{e}}^{-v } v \int_0^v { \ensuremath{\mathbb{e}}}\left [ { \mathrm{e}}^{2\rho w_n } 2 \left ( 1+\epsilon^2 { \mathrm{e}}^{2\delta ( u+|w_n| ) } \right ) \right ] { \mathrm{d}}u \mathrm{d}v+ \frac{2m^2}{\eta(n / k_n)^2}{\ensuremath{\mathbb{e}}}\mathbb{1}_{a_n^c } \ , . \end{aligned}\ ] ] the first summand has a finite limit thanks to lemma [ lem : rvplus ] .
the second summand converges to @xmath83 as @xmath550 tends to @xmath83 exponentially fast while @xmath551 tends to infinity algebraically fast .
bounds on ( i ) are easily obtained , using jensen s inequality and poincar inequality .
@xmath552\right)}{\eta(n / k_n ) } & = & \frac{k_n { \operatorname{var}}\left ( \int_0^\infty { \eta \left ( { \mathrm{e}}^{u+y_{(k_n+1 ) } } \right ) } { \mathrm{e}}^{-u } { \mathrm{d}}u\right)}{\eta(n / k_n ) } \\ & \le & 4 \eta(n / k_n ) { \mathbb{e}}\left [ \left ( \int_0^\infty \frac{\eta \left ( { \mathrm{e}}^{u+y_{(k_n+1 ) } } \right)}{\eta(n / k_n ) } { \mathrm{e}}^{-u } { \mathrm{d}}u \right)^2\right ] \\ & \leq & 4 \eta(n / k_n ) { \mathbb{e}}\left [ \int_0^\infty \left ( \frac{\eta \left ( { \mathrm{e}}^{u+y_{(k_n+1 ) } } \right)}{\eta(n / k_n)}\right)^2 { \mathrm{e}}^{-u } { \mathrm{d}}u \right ] \ , . \ ] ] using the line of arguments as for handling the limit of ( ii ) , we establish that ( i ) converges to @xmath83 .
we now check that ( iii ) converges towards a finite limit .
note that @xmath553\right ] } \\ & = & { \ensuremath{\mathbb{e}}}\left [ ( e-1)\int_0^e \frac{\eta({\mathrm{e}}^{u+y_{(k_n+1)}})}{\eta ( n / k_n)}{\mathrm{d}}u \right ] \ , .
\end{aligned}\ ] ] by lemma [ lem : rvplus ] , for almost every @xmath520 , @xmath554 and @xmath555 the first term is finite as the integral of a continuous function on a compact .
+ thus , @xmath556 the expected value of the last random variable is @xmath557 .
we check that , for sufficiently large @xmath37 , @xmath558}\\ & \leq & { \mathbb{e}}\left [ | e-1| \int_0^e { \mathrm{e}}^{\rho ( u+w_n ) } \left(1 + \epsilon { \mathrm{e}}^{\delta ( u+|w_n|)}\right ) + \mathbb{1}_{a_n^c } |e-1| \frac{m}{|\eta(n / k_n)| } { \mathrm{d}}u \right ] \\ & \leq & { \ensuremath{\mathbb{e}}}\left[{\mathrm{e}}^{\rho w_n } \left ( 2 + \frac{\epsilon}{\delta(1-\delta)^2 } { \mathrm{e}}^{\delta|w_n| } \right)\right ] + \frac{m}{|\eta(n / k_n)| } { \ensuremath{\mathbb{e}}}\mathbb{1}_{a_n^c } \\ & \leq & 4 { \mathrm{e}}^{\frac{\rho^2}{k_n } } + \frac{2 \epsilon}{\delta(1-\delta)^2 } { \mathrm{e}}^{\frac{(\delta-\rho)^2}{k_n } } + \frac{m}{|\eta(n / k_n)| } { \ensuremath{\mathbb{e}}}\mathbb{1}_{a_n^c } { \enspace .}\ ] ] we now way conclude by dominated convergence that @xmath559
the proof of proposition 4.3 from @xcite yields that , with probability larger than @xmath248 , for @xmath560 , @xmath561 we may choose @xmath562 and notice that @xmath563 .
this yields @xmath564
lower bounds on tail index estimation error @xcite are usually constructed by defining sequences of local models around a pure pareto distribution with shape parameter @xmath565 . when deriving lower bounds for the estimation error under constraints like @xmath54 is regularly varying , the elements of the local model for sample size @xmath37 may be defined by @xmath566 where @xmath199 is square integrable over @xmath567 $ ] , @xmath568 , @xmath569 @xcite .
the sequences @xmath570 and @xmath571 are chosen in such a way that @xmath572 satisfies the required constraint . if the local alternatives are pareto change point distributions as in @xcite and @xcite , @xmath573 , @xmath574 .
@xcite explores a richer collection of local alternatives in order to fit into the theory of weak convergence of local experiments . in order to explore adaptivity as in @xcite
, it is necessary to handle simultaneously a collection of sequences @xmath575 corresponding to different rates of decay of the von mises function .
the difficulty of estimation is connected with the difficulty of distinguishing alternatives with different tail indices that is , with the hardness of a multiple hypotheses testing problem . in order to lower bound the testing error
, @xcite chose to use fano s lemma ( * ? ? ? * see ) .
using fano s lemma requires bounding the kullback - leibler divergence between the different local alternatives which is not as easy as bounding the divergence between a pareto change point distribution and a pure pareto distribution .
the next lemma is from @xcite .
it can be used in the derivation of risk lower bounds instead of the classical fano lemma . just as fano s lemma
, it states a lower bound on the error in multiple hypothesis testing .
however , as it only requires computing the kullback - leibler divergence to the localisation center , in the present setting , it significantly alleviates computations and makes the proof more concise and more transparent .
[ thm - kc - lower - bound : appendix ] let @xmath12 , @xmath579 , and @xmath580 .
then , for any tail index estimator @xmath149 and any sample size @xmath37 such that @xmath581 , there exists a collection @xmath582 of probability distributions such that a. @xmath583 with @xmath584 , b. @xmath585 meets the von mises condition with von mises function @xmath586 satisfying @xmath587 where @xmath588 , c. @xmath589 + and @xmath590 \geq \frac{c_\rho}{4(1 + 2{\mathrm{e } } ) } \left(\frac{v\ln\ln n}{n}\right)^{|\rho|/(1 + 2|\rho| ) } \ , , \ ] ] with @xmath157 .
the center of localisation @xmath596 is the pure pareto distribution with shape parameter @xmath4 ( @xmath597 ) .
the local alternatives @xmath598 are pareto change point distributions .
each @xmath585 is defined by a breakpoint @xmath599 and an ultimate pareto index @xmath600 .
if @xmath601 denotes the distribution function of @xmath585 , @xmath602 karamata s representation of @xmath603 is @xmath604 with @xmath605 the kullback - leibler divergence between @xmath585 and @xmath596 is readily calculated , @xmath606 if @xmath584 , the next upper bound holds , @xmath607 the breakpoints and tail indices are chosen in such a way that all upper bounds are equal ( namely @xmath608 does not depend on @xmath64 ) , @xmath609 so that @xmath610 , for all @xmath611 .
now , let @xmath149 be any tail index estimator .
define region @xmath614 , as the set of samples such that @xmath600 minimises @xmath615 , for @xmath616 . then ,
if the event @xmath614 is not realised , @xmath617 by birg s lemma , @xmath618 in order to make the whole construction useful , it remains to choose the `` second - order parameters '' @xmath619 s ( the true second - order parameter of each @xmath585 is infinite ! ) .
we will need an upper bound on @xmath620 ( but we already have @xmath621 ) , as well as a lower bound on @xmath622 for @xmath623 that scales like @xmath624 .
following @xcite , we finally choose @xmath619 as @xmath625 for @xmath611 .
then , for @xmath626 , using that @xmath627 and @xmath628 , @xmath629 \\ & \geq & \frac{1}{2}\left(\frac{n}{v\ln m } \right)^{\rho_i/(1 + 2|\rho_i| ) } \left [ 1- \exp \left ( \frac{-(i - j)}{2(1 + 2|\rho_i|)(1 + 2|\rho_j|)}\right)\right ] \\ & \geq & \frac{c_\rho}{2 } \left(\frac{n}{v\ln m } \right)^{\rho_i/(1 + 2|\rho_i|)}\end{aligned}\ ] ] where @xmath630 may be chosen as @xmath631 . | this paper presents an adaptive version of the hill estimator based on lespki s model selection method .
this simple data - driven index selection method is shown to satisfy an oracle inequality and is checked to achieve the lower bound recently derived by @xcite . in order to establish the oracle inequality , we derive non - asymptotic variance bounds and concentration inequalities for hill estimators . these concentration inequalities are derived from talagrand s concentration inequality for smooth functions of independent exponentially distributed random variables combined with three tools of extreme value theory : the quantile transform , karamata s representation of slowly varying functions , and rnyi s characterisation for the order statistics of exponential samples .
the performance of this computationally and conceptually simple method is illustrated using monte - carlo simulations . |
the second - named author considered in @xcite so - called _ extremalities _ in the approximate integration .
let @xmath0 be the @xmath1-th degree legendre polynomial given by the rodrigues formula @xmath2 then @xmath0 has @xmath1 distinct roots @xmath3 .
the @xmath1-point gauss
legendre quadrature is the positive linear functional on @xmath4}$ ] given by @xmath5=\sum_{i=1}^nw_if(x_i)\ ] ] with the weights @xmath6 the @xmath7-point lobatto quadrature is the functional @xmath8=v_1f(-1)+v_{n+1}f(1)+\sum_{i=2}^n v_if(y_i)\,,\ ] ] where @xmath9 are ( distinct ) roots of @xmath10 and @xmath11 for these forms of quadratures as well as for another quadratures appearing in this paper see for instance @xcite .
recall that a continuous function @xmath12\to\r$ ] is @xmath1-_convex _ ( @xmath13 ) , if and only if @xmath14 is of the class @xmath15 and the derivative @xmath16 is convex ( ( * ? ? ?
* theorem 15.8.4 ) ) . for the needs of this paper
it could be regarded as a definition of @xmath1-convexity .
let @xmath17 be a positive linear functional defined ( at least ) on a linear subspace of @xmath4}$ ] generated by the cone of @xmath18-convex functions ( @xmath19\xge 0 $ ] for @xmath20 ) .
assume that @xmath17 is exact on polynomials of order @xmath21 , @xmath22=\int_{-1}^1 p(x){\mathrm{d}x}$ ] for any polynomial @xmath23 of order @xmath21 .
it was proved in ( * ? ? ?
* theorem 14 ) that the inequality @xmath24\xle\t[f]\xle\lob_{n+1}[f]\ ] ] holds for any @xmath18-convex function @xmath12\to\r$ ] .
then the functionals @xmath25 and @xmath26 restricted to the cone of @xmath18-convex functions are minimal and maximal , respectively , among all positive linear functionals defined ( at least ) on @xmath18-convex functions , which are exact on polynomials of order @xmath21 . in ( * ? ? ?
* theorem 15 ) there is a counterpart of the above result for @xmath27-convex functions with radau quadratures in the role of the minimal and maximal operators . studying the results of this kind
the following problem seems to be natural .
some quadrature operators are extremal in the sense of inequalities like . on the other hand ,
the set of all quadratures which are exact on polynomials of some given order is convex
. then it could be interesting to find its extreme points looking for the possible connections between extremalities in the approximate integration and the extreme points of convex sets .
in particular , are @xmath25 and @xmath26 extreme points of the above mentioned set
? if the answer is positive , are they the only extreme points , or there exist another ones ?
this is the starting point for our considerations .
we will observe that the extreme points in the set of all quadratures exact on polynomials of prescribed order could be determined with the aid of ( * ? ? ? * theorem 6.1 , p. 101 ) .
next we shall investigate the extreme points the set of all positive linear operators defined on @xmath28 $ ] with prescribed moments .
our research is far from being complete .
actually we are able to give a full description of the extreme points of the set of symmetric operators with four prescribed moments , @xmath29 .
let @xmath30 be a convex subset of a linear space . recall that @xmath31 is the extreme point of @xmath30 , if @xmath32 is not the `` interior '' point of any segment with endpoints in @xmath30 , @xmath33 for some @xmath34 and @xmath35 $ ] implies that @xmath36 .
the set of all extreme points of a set @xmath30 will be denoted by @xmath37 .
a _ quadrature _ on @xmath38 $ ] is the linear functional defined on @xmath4}$ ] by the formula @xmath39=\sum_{k=1}^{n({\mathcal{q}})}w^{{\mathcal{q}}}_kf(\xi^{{\mathcal{q}}}_k)\,,\ ] ] where @xmath40 , @xmath41 $ ] are the _ nodes _ and @xmath42 are the _ weights _ of @xmath43 ( for @xmath44 ) .
if all the weights of @xmath43 are positive , then @xmath43 is a _ positive quadrature _ , @xmath45\xge 0 $ ] for @xmath20 .
positive quadratures are often used in the approximate integration .
let @xmath46 , @xmath47 .
fix a vector @xmath48 .
let @xmath49 be the set of all positive quadratures @xmath43 with moments @xmath50=m_k$ ] , @xmath47 . in this section
we determine the extreme points of the ( convex ) set @xmath49 .
in particular , if @xmath51 is a vector of integral moments @xmath52 , @xmath47 , we will obtain a connection between the extremalities in the approximate integration and extreme points of a set of positive quadratures , which are exact on polynomials of a given order @xmath1 . every positive quadrature @xmath43
could be written in the form @xmath39=\int_{[-1,1]}f{\mathrm{d}\mu_{{\mathcal{q}}}}\,\quad\text{for } \mu_{{\mathcal{q}}}=\sum_{k=1}^{n({\mathcal{q}})}w^{{\mathcal{q}}}_k\delta_{\xi^{{\mathcal{q}}}_k},\ ] ] where @xmath53 stands for a dirac measure concentrated at @xmath32 . by the riesz
markov theorem ( @xcite ) the measure @xmath54 in the above representation is uniquely determined .
furthermore , if @xmath55 , then @xmath51 is the moment vector of the measure @xmath54 .
denote by @xmath56 the set of all discrete measures @xmath57 on @xmath58\bigr)$ ] with moments @xmath59}e_k{\mathrm{d}\mu}=m_k\,,\quad k=0,1,\dots , n\,.\ ] ] the set @xmath56 is convex .
a quadrature @xmath55 is an extreme point of @xmath49 if and only if @xmath60 .
a quadrature @xmath55 is an extreme point of @xmath49 if and only if the measure @xmath61 is the extreme point of @xmath56 . by virtue of (
* theorem 6.1 , p. 101 ) the extreme measures in @xmath56 are exactly the measures concentrated on at most @xmath62 points .
this finishes the proof . for @xmath52 , @xmath63
, we obtain immediately that the @xmath1-point gauss quadrature @xmath25 , as well as the @xmath7-point lobatto quadrature , are the extreme points of @xmath49
. nevertheless , there are infinitely many other extremal quadratures in this set . for instance , all gauss quadratures and with @xmath23 nodes ( @xmath64 ) , also all lobatto quadratures with the number of nodes @xmath65 , are the extreme points of @xmath49 .
to find extremal quadratures we needed to know the extreme points of a set of all discrete measures with finite spectrum . in this section
we consider all finite symmetric measures on @xmath66\bigr)}$ ] with prescribed moments @xmath67 .
let @xmath68\bigr)}$ ] be the set of all finite measures on @xmath66\bigr)}$ ] .
let @xmath69\bigr)}$ ] be a subset of @xmath68\bigr)}$ ] consisting of measures , which are symmetric @xmath70 , @xmath71\bigr)}$ ] if and only if @xmath72\bigr)}$ ] and @xmath73 , @xmath74\bigr)}$ ] .
let @xmath75\bigr)}$ ] , @xmath76 be the sets of probability measures on @xmath66\bigr)}$ ] and @xmath77 , respectively .
denote @xmath78\bigr)}={\mathcal{m}}^0{\bigl([-1,1]\bigr)}\cap{\mathcal{p}}(\r)\,.\ ] ] for a non - zero measure @xmath72\bigr)}$ ] define the measure @xmath79 by @xmath80\bigr)}}\,,\quad b\in{\mathcal{b}}{\bigl([-1,1]\bigr)}\,.\ ] ] let @xmath81 and @xmath82,a\bigr)}}$ ] be the set of all measures @xmath71\bigr)}$ ] satisfying @xmath83\bigr)}\,.\ ] ] set @xmath84,a\bigr)}}={\mathcal{m}}^0{{\bigl([-1,1],a\bigr)}}\cap{\mathcal{p}}{\bigl([-1,1]\bigr)}\,.\ ] ] clearly @xmath85 for any @xmath86,a\bigr)}}$ ] .
moreover , @xmath87,a\bigr)}}\iff { \widetilde{\mu}}\in{\mathcal{p}}^0{{\bigl([-1,1],a\bigr)}}\,,\ ] ] whenever @xmath57 is a non - zero measure . obviously , the set @xmath88,b\bigr)}}$ ] is the set consisting of all finite symmetric measures on @xmath66\bigr)}$ ] with prescribed moments @xmath67 .
we start with two lemmas .
the proof of the first of them is rather standard ad simple , so we omit it .
[ lem:1 ] let @xmath89\bigr)}$ ] and @xmath90 for @xmath81 . 1 .
if @xmath57 is concentrated on @xmath91 $ ] , then @xmath92 .
if @xmath57 is concentrated on @xmath93\cup[a,1]$ ] , then @xmath94 .
3 . if @xmath95 , @xmath57 is concentrated on the set @xmath96 , then @xmath97 .
4 . if @xmath57 is concentrated on @xmath91 $ ] and @xmath98 , then @xmath99 . 5 . if @xmath57 is concentrated on @xmath93\cup[a,1]$ ] and @xmath100 , then @xmath101 .
suppose that @xmath57 is concentrated on @xmath91 $ ] .
then @xmath97 if and only if @xmath57 is concentrated on @xmath96 .
suppose that @xmath57 is concentrated on @xmath93\cup[a,1]$ ] .
then @xmath97 if and only if @xmath57 is concentrated on @xmath96 .
let @xmath72\bigr)}$ ] , @xmath102\bigr)}$ ] .
then @xmath103 stands for the restriction of @xmath57 to the set @xmath104 , @xmath105 , @xmath74\bigr)}$ ] . similarly to , for any @xmath102\bigr)}$ ] with @xmath106 , we put @xmath107\bigr)}\,.\ ] ] [ lem:2 ] let @xmath81 and @xmath108,a\bigr)}}$ ] be a non - zero continuous measure , @xmath109 for all @xmath110 $ ]
. then 1 .
@xmath98 and @xmath100 ; 2 .
there exists a @xmath111 @xmath112 and @xmath113 .
the part ( i ) follows immediately by lemma [ lem:1 ] , because otherwise either @xmath101 , or @xmath99 . to prove ( ii ) assume , on the contrary , that for all @xmath114 , @xmath115\bigr)=0\quad\text{or}\quad\mu\bigl([\xi,1]\bigr)=0\,.\ ] ] we recursively define the sequence of sets @xmath116\subset[a,1]$ ] , @xmath13 starting with @xmath117 $ ] . using ( i ) and taking into account the symmetry of @xmath57 , we get @xmath118\bigr)>0 $ ] .
suppose that we have constructed the sets @xmath119 , @xmath120 @xmath121\bigr)$ ] , @xmath122 and @xmath123 , @xmath124 .
if @xmath125 , then , by , two cases are possible . if @xmath126\bigr)=0 $ ] , then @xmath127\bigr)=\mu\bigl([a,1]\bigr)$ ] and we take @xmath128 , @xmath129 . if @xmath127\bigr)=0 $ ] , then @xmath126\bigr)=\mu\bigl([a,1]\bigr)$ ] and we take @xmath130 , @xmath131
obviously , for @xmath132 $ ] we have @xmath133 and @xmath134\bigr)$ ] . by the above construction @xmath135\bigr)>0 $ ] and
there exists @xmath136 $ ] @xmath137 . because @xmath57 was continuous ,
we arrive at the contradiction .
this completes the proof .
below we prove some decomposition - type result .
[ th:3 ] let @xmath81 and @xmath86,a\bigr)}}$ ] be a continuous measure .
there exist the sets @xmath138\bigr)}$ ] @xmath139 , @xmath140 , @xmath141 and @xmath142,a\bigr)}}$ ] .
since @xmath57 is continuous and symmetric , lemma [ lem:2 ] implies @xmath143 consider the function @xmath144\to\r$ ] given by @xmath145}({\mathrm{d}u})\,,\quad a\xle x\xle 1\ ] ] and the measure @xmath146 , which is absolutely continuous @xmath57 .
in particular , @xmath147 is continuous .
denote by @xmath148 the distribution functions of the measures @xmath149 , respectively , and rewrite the function @xmath150 in the form @xmath151}({\mathrm{d}u})\biggr)\cdot\bigl(\mu\bigl([-x , x]\bigr)\bigr)^{-1}\\\nonumber & = \biggl(\int_{-x}^x u^2\mu({\mathrm{d}u})\biggr)\cdot\bigl(\mu\bigl([-x , x]\bigr)\bigr)^{-1}=\frac{\nu\bigl([-x , x]\bigr)}{\mu\bigl([-x , x]\bigr)}=\frac{f_{\nu}(x)-f_{\nu}(-x)}{f_{\mu}(x)-f_{\mu}(-x)}\ , .
\end{aligned}\ ] ] the distributions functions @xmath148 are continuous by continuity of the measures @xmath149 , respectively .
furthermore , @xmath152\bigr)\xge\mu\bigl([-a , a]\bigr)>0\ ] ] for all @xmath136 $ ] .
hence @xmath150 is continuous by . by lemma [ lem:1 ]
we infer that @xmath153 .
since @xmath86,a\bigr)}}$ ] , then @xmath154 .
continuity of @xmath150 implies that there exists @xmath155 @xmath156 then @xmath157 indeed , suppose that @xmath158 .
hence @xmath159 is a zero - measure and consequently @xmath160}({\mathrm{d}u})\biggr)\cdot\bigl(\mu\bigl([-b_1,b_1]\bigr)\bigr)^{-1}\\ & = \biggl(\int_{-1}^1 u^2\mu|_{[-a , a]}({\mathrm{d}u})\biggr)\cdot\bigl(\mu\bigl([-a , a]\bigr)\bigr)^{-1}=g(a)\ , , \end{aligned}\ ] ] which contradicts .
similarly , supposing that @xmath161 , we arrive at @xmath162 , which also contradicts .
now we define the function @xmath163\to\r$ ] by @xmath164\cup[x , b_1]}({\mathrm{d}u})\,,0\xle x\xle a\,.\ ] ] writing @xmath165\bigr)}{\mu\bigl([x , b_1]\bigr)}=\frac{f_{\nu}(b_1)-f_{\nu}(x)}{f_{\mu}(b_1)-f_{\mu}(x)}\ ] ] and using once more the continuity of @xmath148 , we obtain that @xmath166 is continuous .
we have @xmath167 and @xmath168\cup[a , b_1]}({\mathrm{d}u}).\ ] ] using lemma [ lem:1 ] we arrive at @xmath169 .
consequently , by continuity of @xmath166 we conclude that there exists @xmath170 @xmath171 define @xmath172\cup [ a_1,b_1]$ ] , @xmath173\setminus e_1 $ ] . by
we obtain @xmath174 taking into account the above equation and the moment condition @xmath175 , we have @xmath176 the following inequalities are true : @xmath177 indeed , suppose that @xmath178 .
then @xmath179 , which contradicts .
similarly , if @xmath180 , then @xmath181 , which also contradicts . by and
we infer that @xmath182 and @xmath183 . using and we arrive at @xmath142,a\bigr)}}$ ] .
the proof is now complete .
our next result offers some decomposition of a continuous measure .
[ th:4 ] let @xmath81 and @xmath86,a\bigr)}}$ ] be a continuous measure .
1 . there exist non - zero measures @xmath184,a\bigr)}}$ ] such that @xmath185 and @xmath186 , @xmath187 for any @xmath188 .
there exist measures @xmath189,a\bigr)}}$ ] and @xmath190 @xmath191 , @xmath192 and @xmath193 1 .
take @xmath138\bigr)}$ ] given by theorem [ th:3 ] and denote @xmath194 , @xmath195 .
since the measures @xmath196 are concentrated on disjoint sets and @xmath185 , we conclude that @xmath186 , @xmath187 for any @xmath188 .
2 . put @xmath197 , @xmath198 , where @xmath196 are defined in ( i ) . since @xmath186 and @xmath187 for any @xmath188 , then @xmath191 , @xmath192 . setting @xmath199
, we get .
this finishes the proof .
the corollary below follows immediately by theorem [ th:4 ] .
[ cor:5 ] let @xmath81 and @xmath86,a\bigr)}}$ ] .
if @xmath57 is a continuous measure , then @xmath57 is not the extreme point of @xmath88,a\bigr)}}$ ] .
let @xmath200 .
by @xmath201 we denote the set of all discrete symmetric probability measures @xmath57 on @xmath66\bigr)}$ ] with prescribed moments @xmath67 and admitting at most four mass points .
now we state for the symmetric probability measure @xmath57 a necessary condition to be the extreme point of @xmath88,b\bigr)}}$ ] .
[ th:6 ] let @xmath200 .
then @xmath202,b\bigr)}}\bigr ) \subset { \mathcal{k}}(b).$ ] let @xmath203,b\bigr)}}$ ] be the extreme point of @xmath88,b\bigr)}}$ ] .
then @xmath204 can be uniquely represented as the sum of a continuous measure and a discrete measure : @xmath205 where @xmath206 $ ] , @xmath207\bigr)}$ ] , @xmath208 is a continuous measure , while @xmath209 is a discrete measure .
observe that @xmath210 , because if @xmath211 , then @xmath204 was continuous and , by corollary [ cor:5 ] , @xmath204 was not the extreme point .
if @xmath212 , then @xmath204 is a discrete measure and the assertion follows by ( * ? ? ?
* theorem 6.1 , p. 101 ) .
suppose now that @xmath213 .
we claim that in this case @xmath204 is not the extreme point .
if we show it , the proof is finished .
it is not difficult to check that both @xmath208 and @xmath209 are symmetric @xmath70 . indeed ,
if @xmath214 for some @xmath110 $ ] , then @xmath215 .
therefore a discrete part of @xmath204 , @xmath216 , is symmetric , which implies that @xmath209 is symmetric .
hence also @xmath208 is symmetric . as a probability measure
, @xmath208 is non - zero .
then @xmath217,a\bigr)}}$ ] , where @xmath218 and @xmath81 .
now we apply theorem [ th:4 ] ( ii ) to the measure @xmath208 .
there exist the measures @xmath189,a\bigr)}}$ ] and @xmath219 @xmath220 and @xmath221 we have also @xmath222,c\bigr)$ ] , where @xmath223 and @xmath224 .
write @xmath225 using the properties of the measures @xmath226 we get @xmath227 by , , , @xmath204 can be written as @xmath228 with @xmath229 and @xmath230 . since @xmath220 , then @xmath231 .
using we arrive at @xmath232,b\bigr)}}$ ] , which implies that @xmath204 is not the extreme point of @xmath88,b\bigr)}}$ ] and completes the proof .
notice that every probability measure @xmath233 ( with @xmath200 ) can be written in the form @xmath234 for some @xmath235 and @xmath236 with @xmath237 .
observe that if @xmath238 then @xmath239 it is easy to see that if @xmath235 and @xmath236 with @xmath237 satisfying the above equation , then @xmath240 .
if additionally @xmath241 , then the numbers @xmath242 could be computed by @xmath243 [ rem:7 ] the set @xmath201 consists of all probability measures @xmath244 given by , where @xmath245 , @xmath236 with @xmath237 satisfying .
one could easily show the lemma .
[ lem:7c ] let @xmath200
. then @xmath246,b\bigr)}}\bigr).$ ] thus , we derive from theorem [ th:6 ] and lemma [ lem:7c ] the main result of this paper .
[ th:7 ] let @xmath200
. then @xmath202,b\bigr)}}\bigr)={\mathcal{k}}(b).$ ] of course the extreme points of @xmath88,b\bigr)}}$ ] are the measures admitting @xmath247 or @xmath248 mass points .
the only two - point extreme measure is @xmath249 all three - point extreme measures have the form @xmath250 where @xmath251 . in particular , for @xmath252 we get @xmath253
as an application of results obtained in the previous section concerning the extreme measures we shall give a theorem on integral representation of probability measures from the set @xmath88,b\bigr)}}$ ] .
the set of probability measures @xmath76 is metrizable . in metrizing of the weak convergence of probability measures on @xmath77
the lvy - prohorov distance ( see ( * ? ? ?
* chapter 11 , theorem 11.3.3 , p. 395 ) ) can be used .
the set @xmath75\bigr)}$ ] is a metrizable compact convex subset of @xmath76 .
it is not difficult to prove that @xmath88,b\bigr)}}$ ] is a closed and convex subset of @xmath75\bigr)}$ ] .
consequently , we have the following lemma . on @xmath88,b\bigr)}}$ ] consider the topology induced from @xmath76 and the mapping @xmath254\times[b,1]\to
{ \mathcal{k}}(b)$ ] given by @xmath255 taking into account remark [ rem:7 ] and the formulae , it is not difficult to prove the lemma .
[ th:8 ] let @xmath200 . for every probability measure @xmath203,b\bigr)}}$ ]
there exists a probability measure @xmath258 on @xmath259\times[b,1]\bigr)$ ] @xmath260\times[b,1]}\mu_{(x , y)}\gamma\bigl({\mathrm{d}(x , y)}\bigr)\biggr)({\mathrm{d}u})\ ] ] for any continuous function @xmath12\to\r$ ] .
we shall use choquet s representation theorem ( @xcite ) .
taking into account theorem [ th:7 ] we obtain that for every measure @xmath203,b\bigr)}}$ ] there exists a probability measure @xmath261 @xmath262 for any continuous linear functional @xmath263 on @xmath75\bigr)}$ ] .
let @xmath258 be the measure on @xmath259\times[b,1]\bigr)$ ] defined by @xmath264 , @xmath265 for @xmath266\times[b,1]\bigr)$ ] .
taking into account lemma [ lem:7b ] , for any continuous linear functional @xmath263 on @xmath75\bigr)}$ ] we have @xmath267\times[b,1]}l\circ t\bigl((x , y)\bigr)\gamma\bigl({\mathrm{d}(x , y)}\bigr)\,.\end{gathered}\ ] ] for every continuous function @xmath12\to\r$ ] the linear functional @xmath268 is continuous .
hence , for the probability measure @xmath203,b\bigr)}}$ ] we obtain by @xmath269\times[b,1]}\int_{-1}^{1 } f(u)\mu_{(x , y)}({\mathrm{d}u})\gamma\bigl({\mathrm{d}(x , y)}\bigr)\\ = \int_{-1}^{1}f(u)\biggl(\int_{[0,b]\times[b,1]}\mu_{(x , y)}\gamma\bigl({\mathrm{d}(x , y)}\bigr)\biggr)({\mathrm{d}u})\,.\end{gathered}\ ] ] the theorem is proved . | in the approximate integration some inequalities between the quadratures and the integrals approximated by them are called _
extremalities_. on the other hand , the set of all quadratures is convex .
we are trying to find possible connections between extremalities and extremal quadratures ( in the sense of extreme points of a convex set ) .
of course , the quadratures are the integrals discrete measures and , moreover , a quadrature is extremal if and only if the associated measure is extremal .
hence the natural problem arises to give some description of extremal measures with prescribed moments in the general ( not only discrete ) case . in this paper
we deal with symmetric measures with prescribed first four moments . the full description ( with no symmetry assumptions , and/or not only four moments are prescribed and so on ) is far to be done . |
we develop new entropy based statistics @xmath2 called _ slide statistics _ which can be computed from any sample data in a metric space . as an application
, we use these statistics to test whether financial returns are independent observations from a particular distribution .
for example , figure [ fig : fig1 ] shows plots of @xmath3 against @xmath14 for data @xmath15 consisting of n - tuples of consecutive returns regarded as a subset of the metric space @xmath16 with the usual metric .
as can be seen , the lower curve corresponding to the s@xmath0p @xmath1 is very different from the ones obtained for either the normal or laplace distributions . any potential model for
the returns of the s@xmath0p @xmath1 must be able simulate the @xmath3 curve in figure [ fig : fig1 ] which is a new requirement that is apparently difficult to meet .
the statistics @xmath17 contain information about the interplay between the dimension of a space and distributions on that space and provide a new way of describing continuous random variables and stochastic processes in general .
the @xmath17 will take some time to define but we will provide explicit formulas for slide statistics @xmath3 and @xmath4 that are easily computed for any sample in a metric space . against @xmath14 for the s@xmath0p @xmath1 for the @xmath18 trading days ending december 31 , 2014 .
the middle cluster of five graphs are the corresponding plots obtained from five different simulated sequences of returns with the laplace distribution .
the upper cluster of five graphs are the corresponding plots obtained from five different simulated sequences of normally distributed returns .
, width=384 ] our applications will focus on financial returns but many other important processes in science and mathematics also yield data in the form of a set of points in a metric space that must be quantified and interpreted .
fields like fractal analysis have developed in order to obtain dimensional information from a wide range of real world data using quantities like the hausdorff , information and correlation dimensions @xcite .
one of the problems with these traditional measures of dimension is that they tend to concentrate on the geometric properties of a set while ignoring the statistical origins of the data .
we will give examples where @xmath19 appears to converge to the reciprocal of the hausdorff dimension when @xmath20 is taken to be a a larger and larger random sample from a fractal .
it is not clear in general how to assign a dimension to a point process but our theory and simulations suggest a purely statistical way to identify a class of them for which it does make sense to assign a dimension .
specifically , we think the dimension of a point process @xmath21 should be defined to be @xmath22 provided @xmath23 and @xmath24 for @xmath25 in which case we say the process is tangible with dimension @xmath22 . in the case of a tangible process , the values of @xmath26
are entirely determined by the number @xmath22 but this relationship does not hold in general . a possible example of a tangible process is the random generation of points @xmath27 in @xmath8^d$ ] , where the @xmath28 are chosen independently and uniformly at random from @xmath8 $ ] , for which we provide evidence that @xmath29 and @xmath30 . in particular , @xmath3 and
@xmath4 are new spatial statistics for testing hypotheses concerning the distribution of points in @xmath8^d$ ] .
although we will sometimes use fractals as examples , these statistics can be applied to any random variable with values in a metric space . in particular
, the slide statistics appear to describe continuous random variables @xmath7 in an entirely new way that has nothing to do with mean and standard deviation since @xmath31 for @xmath32 .
for example , the first two slide statistics for any normal variable @xmath11 appear to converge to @xmath12 and @xmath13 which provide the basis for a new goodness of fit test for normality . for any exponential random variable @xmath33 , simulation gives that @xmath34 and @xmath35 and we conjecture from these and other simulations that @xmath36 for any continuous real - valued random variable @xmath7 .
if we think of @xmath37 as the _ dimension _ of the random variable @xmath7 , then this last conjecture says the uniform distribution on @xmath8 $ ] has the maximum dimension of any continuous real - valued random variable .
the second slide statistic @xmath4 is negative for the examples given so far but @xmath4 is positive for the cauchy distribution and is often found to be positive for the log daily returns of the standard and poor 500 index which is yet another demonstration of the non - normality of these returns .
the construction of @xmath38 requires several ideas working in concert and can be summarized as follows . in section [ s:2 ] , we introduce a variant of the differential entropy called the genial entropy which is the starting point for defining the slide statistics . unlike the differential entropy ,
the genial entropy is scale invariant and in section [ s : inequality ] we prove it is never negative which will give a new lower bound for the differential entropy . given a finite set of distinct points @xmath20 in a metric space , we find the distance from each point to its nearest neighbour and arrange these distances in descending order so @xmath39 .
let @xmath40 be the function on @xmath41 whose value on @xmath42 is @xmath43 .
let @xmath44 be the area under @xmath45 and let @xmath46 denote the genial entropy of the density @xmath47 which turns out to be @xmath48 at @xmath49 .
we then define @xmath50 to be the @xmath14th derivative from the right at @xmath48 of the function @xmath46 which is developed in section [ s : calculus ] . in section [ s : calculus ]
, we derive an explicit formula for @xmath19 and state a conjectured formula for @xmath51 .
the fact that @xmath3 and @xmath4 are easily evaluated using a computer makes them particularly suitable for practical applications .
the results obtained from the simulation of @xmath3 and @xmath4 in a variety of contexts are summarized in section [ s : slide ] and demonstrate the interplay between dimension and distribution that is captured by these statistics . as applications in section [ s : returns ] , we consider a variety of possible distributions as models for financial returns and use the slide statistics to illustrate how poorly these models fit the empirical data .
our starting point for the development of the slide statistics is the differential entropy @xmath52 of a density @xmath53 which is well known in statistics and information theory @xcite
. our focus will be on a variant of the differential entropy called the genial entropy or g - entropy which will be described in definition [ d:2.02 ] . in the simplest case of a probability density @xmath53 that has an inverse function such as @xmath54 on @xmath55 $ ]
, the genial entropy is just the sum of the differential entropies of @xmath53 and @xmath56 .
the next proposition shows how this sum can be written in a form that makes sense for densities that may not have inverses .
[ p:2.01 ] suppose @xmath53 is a continuous function on @xmath57 $ ] with the properties that the derivative of @xmath53 exists and is negative on @xmath58 , @xmath59 and @xmath60 . also assume the differential entropies of @xmath53 and @xmath56 both exist .
then the sum of their differential entropies is given by @xmath61 .
substitute @xmath62 into the integral @xmath63 to get @xmath64 which equals @xmath65 .
after integrating by parts this last integral becomes @xmath66 and the result follows .
the genial entropy will only be defined for densities of the following special form .
[ d:2.01 ] a corner density is a function @xmath67 where @xmath68 is a connected interval contained in @xmath69 , @xmath53 is monotone decreasing and @xmath70 .
here is our definition of genial entropy which is motivated by the conclusion of proposition [ p:2.01 ] .
[ d:2.02 ] let @xmath67 be a corner density .
the genial entropy or g - entropy of @xmath53 is defined by @xmath71 when this integral exists , with the usual convention that @xmath72 .
if @xmath53 satisfies the conditions of proposition [ p:2.01 ] , then @xmath73 so in particular @xmath74 and @xmath75 must have the same genial entropy which happens to be euler s constant as shown in table [ t:2.01 ] . in section [ s : inequality ] , we will prove the genial entropy is always nonnegative .
the genial entropy of some corner densities . .
[ cols="^,^,^",options="header " , ] the stable distributions are often considered @xcite as possible models for financial returns .
we now consider the family of stable distributions @xmath76 described in @xcite where @xmath77 $ ] is the stable parameter , @xmath78 $ ] is the skewness parameter , @xmath79 is the scale parameter and @xmath80 is the location parameter . by proposition [ p:5.01 ] , the values of @xmath17
are not affected by changes in @xmath81 and @xmath82 so we will simply work with the family @xmath83 which we write as @xmath84 .
figure [ fig : fig5 ] shows the points @xmath85 from table [ t:5.03 ] plotted together with @xmath86 samples from stable distributions with parameters @xmath87 and @xmath88 chosen uniformly at random from the intervals @xmath89 and @xmath90 respectively .
we see that the values in table [ t:5.03 ] lie outside the region corresponding to the stable distributions .
figure [ fig : fig6 ] show an expanded view of figure [ fig : fig5 ] showing that the slide statistics for the financial returns in table [ t:5.03 ] are well outside the region corresponding to samples from stable distributions .
this failure of the stable distributions to fit financial returns is consistent with the findings of @xcite .
obtained for @xmath86 samples from stable distributions with parameters @xmath87 and @xmath88 chosen uniformly at random from the intervals @xmath89 and @xmath90 respectively . the values in table [ t:5.03 ]
are shown plotted with squares .
, width=384 ] , width=384 ]
as we have seen , the statistics @xmath3 and @xmath4 can be used as the basis for goodness of fit test for financial returns .
much more research needs to be done to better understand the role of the slide statistics in characterizing financial data .
in particular , we would like to understand the relationship between the sign of @xmath4 and the behaviour of financial markets .
the simulations we have described suggest that @xmath3 and @xmath4 can be used to distinguish between probability distributions and are capapble of detecting dimensional information .
the values of @xmath91 we obtained through simulation point to a larger theory of these statistics which is currently being developed . at present
however , the @xmath91 are quite mysterious and much work will need to be done to understand all they are telling us about sets of points in metric spaces , random variables or point processes in general .
we gave examples of point processes on the cantor set and the sierpinski triangle for which @xmath5 converged to the dimension of the fractal .
we would like to understand when this occurs in general and also the relationship between @xmath5 and the usual definitions of dimension .
more generally , we would like to know when a point process is tangible in the sense of definition [ d:5.04 ] .
when a process is tangible it is possible to use the expression @xmath92{\frac{(-1)^{n+1}(n-1)!(n-1)\zeta(n)}{\rho_n(u)}}$ ] as a statistic for estimating the dimension .
we gave examples in which it converged to the dimension faster for @xmath93 than for @xmath94 and we would like to know what happens for larger values of @xmath14 .
in terms of calculations , we need to prove conjecture [ cj:4.01 ] concerning the calculation of @xmath51 and we need formulas for @xmath50 for @xmath95 . given the complexity of our conjecture for @xmath51 , the formulas for larger values of @xmath14 are likely to be very complicated .
the convergence of the @xmath50 for real - valued random variables can sometimes be improved by using the distances between consecutive points rather than the distances to nearest neighbours .
we would like to have a better understanding of this situation and also to know if there are any higher dimensional analogues . in defining the slide statistics
, we used the functions @xmath45 which can be thought of as a continuous deformation of @xmath40 at @xmath96 into the constant function @xmath97 at @xmath49 .
we can achieve the same effect using the functions @xmath98 and calculate the derivatives corresponding to the slide derivatives in section [ s : calculus ] .
it turns out that only the first of these statistics is interesting and while it is easier to calculate than @xmath3 it often does nt converge .
bonetti , m. , forsberg , l. , ozonoff , a. and pagano , m. , _ the distribution of interpoint distances .
mathematical modeling applications in homeland security _ , ht banks and c castillo chavez , eds . , 2003,87 - 106 . | we present a new approach to financial returns based on an infinite family of statistics called _ slide statistics _ that we introduce .
the evidence these statistics provide suggests that certain distributions such as the stable distributions are not good models for the financial returns from various securities or indexes like the s@xmath0p @xmath1 and the dow jones .
formally , we associate with any finite subset of a metric space an infinite sequence of scale invariant numbers @xmath2 derived from a variant of differential entropy called the genial entropy .
we give explicit formulas for @xmath3 and @xmath4 that are easily evaluated by a computer and make this theory particularly suitable for applications . as statistics for point processes , these numbers often appear to converge in simulations and we give examples where @xmath5 converges to the hausdorff dimension and we prove that @xmath6 . for a uniform random variable @xmath7 on @xmath8^n$ ] , the evidence from simulations suggests that @xmath9 and @xmath10 which yields new tests for spatial randomness .
the slide statistics describe continuous random variables in an entirely new way . for example , if @xmath11 is any normal variable then simulations suggest that @xmath12 and @xmath13 which provides new goodness of fit tests for normality . |
recent observations of gamma - ray bursts ( grbs ) suggest that long duration grbs and type ib / c supernova ( sn ) explosions are tightly connected .
for example , sn1998bw was observed in the positional error box of grb980425 @xcite . in this case
the grb / sn association was based on the spatial and temporal coincidence of both events .
the most remarkable example of long grb / sn link came in 2003 , when the spectra of both the grb030329 afterglow and of the sn2003dh were measured , since the burst happened closeby and it was quite bright .
the supernova spectrum , which includes many complex lines , gradually appeared from the decaying afterglow spectrum after a few tens of days from the burst .
the spectrum of sn2003dh after about a month from the explosion is quite similar with that of sn1998bw at the same stage .
both sn1998bw and sn2003dh are type ic supernovae @xcite whose progenitor had lost the hydrogen and the helium envelopes during the pre - supernova stage .
they are also categorized within a special class of supernova explosions , so - called hypernovae , whose explosion energy is about ten times higher , i.e. , @xmath6 , than that of ordinary supernova . indeed , our common view is that most long - lasting grbs are produced by core - collapse supernovae akin to sn1998bw .
grbs are not exclusively linked to hypernovae .
for instance , the long lasting burst ( @xmath7 s ) grb060218 ( or xrf060218 ) was associated with the type ic , which is not a hypernova .
the explosion energy falls within the regular range for type ic events @xcite .
@xcite argued that the progenitor star may not form a black hole but a neutron star , since the estimated mass of the progenitor during the main sequence is @xmath8 . on the other hand , there are recent examples of cases of long - lasting grbs ( and ) where no supernova signature was observed at all even if , considering the distance , the observational trace of a supernova should have been detected @xcite . though a variety of long duration grbs which have a strong connection with sns are observed , we still lack of a complete picture of the processes by which a sizable fraction of the energy involved in a type ib / c sn explosion is tapped in a relatively narrow channel , and produces a grb at a large distance from the original site of generation .
@xcite proposed that the death of massive stars can be an origin of grbs .
@xcite introduced the collapsar model to account for the progenitor system of long grbs . according to this model
, a non - spherical outflow could be formed from the deep inside of the progenitor where a black hole or a proto - neutron star is born as a result of the collapse of the iron core @xcite .
if the specific angular momentum of the iron core is sufficiently large , an accretion torus may develop around the central compact object .
the system formed by a central object and an accretion disk has the potential to launch a bipolar outflow .
the mechanisms proposed to extract energy out of such central engines are basically two ( e.g. , * ? ? ?
* ) : thermal or hydromagnetic .
thermal mechanisms rely on depositing a considerable amount of thermal energy in the vicinity of the rotation axis of the system , just above the poles of the central compact object , where a low density funnel has develop in the course of the evolution .
the accretion energy of the hot torus is converted into a copious flux of neutrinos @xmath9 and anti - neutrinos @xmath10 . from the @xmath11-annihilation a hot @xmath12-plasma results . in its turn , @xmath12-pairs annihilate yielding a _ fireball _ of high energy photons .
the conversion of the thermal energy of the fireball into kinetic energy partly determines the subsequent evolution of the plasma .
it accelerates to ultrarelativistic speeds ( reaching lorentz factors of @xmath13 @xcite ) while , at the same time , interacts with the progenitor system .
alternatively , mhd process may tap a fraction of the rotational energy of the bh or of the accretion disk to form an outflow @xcite .
a number of works have dealt with the hydrodynamic properties of the outflows generated in collapsar progenitors as they propagate through the progenitor system ( and in some cases beyond the surface of the progenitor star ) .
the problem is addressed by means of numerical relativistic hydrodynamic simulations with different degrees of complexity ( @xcite ) , which assume that a quasi - steady momentum flux has been produced at a certain distance from the region where the energy is released ( independent of which is the actual energy extraction mechanism -mhd or thermal- ) .
therefore , such numerical works assume the existence of a nozzle through which the injection of a supersonic jet is produced , and put their focus on the modification of the morphology and of the dynamics of collimated outflows as they travel through the progenitor star .
depending on the exact inflow conditions , a variety of different outflows result .
for instance , @xcite finds a whole spectrum of outflows ranging from collimated , relativistic jets to poorly collimated expanding winds .
thus , @xcite argue that such a variety of resulting outflows supports the idea that the same collapsar scenario can yield a number of different phenomena ( in agreement with the previous ideas of unification of high energy transients , e.g. , @xcite ) , such as , grbs , x - ray rich grbs , x - ray flashes , and normal supernovae . in the prompt grb phase we observe the emission of high energy photons from an ultrarelativistic outflow , which is generated at a distance @xmath14 cm , very far from the central engine .
due to the large optical thickness of the outflow at scales comparable to that of the progenitor , we have not observed any electromagnetic emission directly from the progenitor so far , except , perhaps , some precursor activity @xcite . since the direct detection of progenitors of grbs is nowadays impossible ( if they happen at cosmological distances ) , the observation of the association between grbs and sne probably provides the best clue to understand the progenitors of the grbs . though more than 100 grbs per year are identified , most of them occur far away .
thus , it is technically impossible to identify their accompanying supernovae .
another clue regarding the nature of the grb progenitors comes from the environment and the host galaxies in which the burst takes place .
both clues indicate that long - duration grbs are associated with the death of the most massive stars @xcite .
the typical hosts of long grbs are star - forming , low metallicity galaxies ( with an star formation rate @xmath15y@xmath3 ; @xcite ) but bluer than typical starburst galaxies , with little dust @xcite , and lower masses than current ellipticals , i.e. , they correspond to the typical environments of formation of massive stars .
@xcite also conclude , that the host galaxies of the grbs are significantly fainter and more irregular than the hosts of core - collapse supernovae .
a very interesting question is weather it is possible to get any information on the nature of a grb progenitor from the observation of the afterglow emission .
one possibility is to look at the angular distribution of integrated energy per unit of solid angle , as observed in the afterglow phase of the burst .
@xcite estimated theoretically such an angular distribution assuming , that the kinetic energy of the jet is converted to thermal energy in the cocoon , till the head of the jet reaches the progenitor surface .
the cocoon originates from jet material which crosses through the terminal strong shock of the collimated outflow and moves away from the center of the progenitor surrounding the beam of the jet .
a fat cocoon develops for light jets , i.e. , jets whose rest - mass density is much lower than that of the ambient gas into which the jet propagates .
when the jet breaks out the progenitor surface , the thermal energy is released to a low - density inter stellar medium ( ism ) .
@xcite concluded , that the energy distribution per solid angle ( @xmath16 ) of the jet displays a @xmath17 dependence with the viewing angle after its eruption through the progenitor surface .
recently , @xcite have used hydrodynamic simulations to test the theoretical prediction of @xcite , and found that their numerical models do not follow the inferred theoretical angular energy distribution . in this paper
we also try to verify the analytic relation for the angular dependence of the energy with the polar angle that was proposed by @xcite .
we explore a parameter space different from that of @xcite in order to compute the dependence of the angular energy distribution on the structure of the progenitor .
the progenitor models are built upon the pre - supernova models of @xcite . along the way
, we also characterize the hydrodynamic properties of relativistic jets propagating through different progenitor stars .
the paper is organized as follows .
we describe our physical model , the choice of stellar progenitors , and relevant numerical details in sect .
[ model ] . in the appendix
, we provide a study of a selected sample of models in order to justify our choice of numerical resolution and the effects it has on our conclussions .
the dynamics of the injected bipolar outflows , and the extrapolated angular energy distribution in the afterglow phase is considered in sec .
[ results ] .
finally , we discuss our results and write down the conclusions of this work in sect .
[ sec : conclusions ] .
we will investigate the dependence of the properties of relativistic jets , injected in a pre - supernova stellar model at a certain distance from the center , using relativistic hydrodynamic simulations . in sect .
[ sec : progenitor ] we show the different stellar progenitors used in this study .
we provide the technical details of the numerical simulations in sect . [
sec : numerics ] . finally , in sect .
[ sec : jet ] we specify the physical conditions used to inject relativistic jets in the pre - supernova progenitors described above . in the last years , some detailed calculations of stellar evolution of massive stars have been done including the effects of initial angular momentum , dynamo , metallicity , and mass loss rate @xcite . according to these studies
, the metallicity of the progenitor strongly affects the evolution of the angular momentum distribution at the pre - supernova stage , in such a way , that low metallicity is preferred to obtain a large angular momentum in the core of the progenitor . for the purposes of this work
, we employ some of the pre - supernova models computed by @xcite .
we stick to the same naming convention than the former authors , and consider several sets of models ( tab . [
tab : progenitor ] ) .
the first group corresponds to the he16-series of 16 models of @xcite , which include progenitor stars for which @xmath18 bare helium cores are evolved , that have solar metallicity , and different amounts of initial angular momentum , dynamo effects , mass - loss rates , etc .
the last three models of tab . [ tab : progenitor ] , 16oc , 16 tb , and 16tc form the second group of progenitors .
they correspond to stars with the same initial mass as those of the first group but with a smaller metallicity ( @xmath19 , @xmath20 ) .
the second set of low - metallicity models has been chosen among the many other possibilities available because their radius , at the pre - supernova stage , are the smallest among all other low - metallicity progenitors ( in all cases , their stellar radii are @xmath21 cm ) .
figure [ fig:1dradialmass ] shows the radial mass profiles of the models he16c , he16l , and he16n .
he16c is representative of progenitors whose pre - supernova mass is small @xmath22 ( due to the vigorous mass loss rate in the late phases of its evolution ; tab .
[ tab : progenitor ] ) .
other members of this group of low pre - supernova mass are he16b , he16j , and he16k .
we will refer to this group as type - l .
the model he16n belongs to the group of more massive progenitors ( @xmath23 ) , to which we will refer as type - h models .
finally , the model he16l ( also he16d ) falls in the middle of these two groups ( members of this group will be called type - m models ) .
its total mass is about @xmath24 .
figure [ fig:1dradialmass2 ] shows radial mass density profiles of the low metallicity models 16 tb , 16tc , and 16oc .
though the mass of the models is similar ( @xmath25 ) , the density profiles are slightly different . in total , 19 models are considered in this study ( tab .
[ tab : progenitor ] ) .
we neglect for the progenitors any deviation from spherical symmetry arising from the rotation of the models .
therefore , for the study we present here , each progenitor differs from the other mainly in its total mass , radius , and mass density profile at the pre - supernova stage . though a nonspherical structure is expected around the black hole due to the rapid rotation of the progenitor
, it is reasonable to assume spherical symmetry for the envelopes of the progenitor for radial distances @xmath26 cm , which is where we put the innermost radial boundary in the numerical simulations of this study .
thus , we only take from the models of @xcite the radial density and the radial velocity profiles which result by the end of the pre - supernova evolution .
we assume that the pressure of the progenitor is very low , or , equivalently , that the initial specific internal energy ( @xmath27 ) is set to be very low ( @xmath28 , where @xmath29 is the speed of light ) .
both the gravitational force produced by the central compact object , and the progenitor self gravity are ignored , since the timescale for the outflows to cross the progenitor and to break out from the surface of the progenitor , @xmath30s , is much shorter than free fall timescale for the stellar envelopes .
we extend the radial mass density profile to the outside of the progenitor up to the outer computational boundary located at @xmath31 cm .
for the models which barely loose mass during the latest stages of their evolution , the rest - mass density is assumed to be uniform @xmath32 and much smaller than that at the progenitor surface .
if the progenitor star has a non - null mass loss rate ( i.e. , if the parameter @xmath33 , according to the nomenclature of @xcite ) , we take a @xmath34 dependence in radial mass - density profiles from the surface of the progenitor ( see fig .
[ fig:1dradialmass ] ) .
we map the spherically symmetric progenitor models of sect .
[ sec : numerics ] into a two - dimensional grid in spherical coordinates ( @xmath35 ) .
we assume that our models are axial and equatorially symmetric and , therefore , specify reflection boundary conditions at the polar axis ( @xmath36 ) and at the equator ( @xmath37 ) .
the radial grid consists of @xmath38 points , uniformly spaced in @xmath39 , which extends from @xmath40 cm to @xmath41 cm .
the smallest radial grid spacing , besides @xmath42 , is @xmath43 cm , while the largest one , besides @xmath44 , is @xmath45 cm .
the resolution we have choosen here represents a trade - off between accuracy and feasibility of the numerical simulations , as we discuss in the appendix .
free outflow ( i.e. , zero gradient ) boundary conditions are set at @xmath46 and @xmath44 .
the polar grid has @xmath47 grid points uniformly spaced in the range @xmath48 , ( @xmath49 ) .
we use the same 2d special relativistic hydrodynamic code of @xcite to perform our simulations .
the code provides 3rd order accuracy in both space and time , by applying a ppm intra - cell interpolation and a tvd - runge kutta time integration . for the sake of simplicity
, we employ an ideal gas equation of state ( @xmath50 ) with uniform adiabatic index @xmath51 , where @xmath52 , @xmath53 and @xmath27 are the pressure and the rest - mass density , respectively .
we assume that a jet has been generated by the central engine , and that at a certain distance , quasi - steady injection conditions are settled through a well defined circular nozzle .
thus , we inject plasma , in the radial direction , through the innermost radial boundary at @xmath54 in a cone of half - opening angle @xmath55 .
the jet injection proceeds for a period @xmath56s .
we parametrize the outflowing plasma by assuming that it is hot ( we set @xmath57 ) and moderately relativistic ( the lorentz factor being @xmath58 ) .
we adopt the convention that the parameters of the outflow at the injection point are named with a subscript j. because of the conversion of thermal - to - kinetic energy , the injected flows have the potential to accelerate to bulk lorentz factors larger than 100 @xcite . during the first 3s
, the power of the injected outflow is @xmath59 , where @xmath60 is the area of the injection surface , @xmath61 is the specific enthalpy , and @xmath62 is the radial component of the 3-velocity . the density and pressure of the injected outflow are obtained by setting @xmath63 , @xmath64 , @xmath65 , @xmath66 , and @xmath42 .
we fix @xmath67 , which is higher than that adopted in previous studies @xcite .
the total injected energy is several times @xmath68erg .
since the main purpose of this study is to see the jet propagation and expansion of the cocoon into the interstellar medium after the shock breakout , we adopt this power to obtain a rapid propagation of the jet in the progenitor .
this fast propagation is necessary to be consistent with the fact that we neglect the self - gravity of the star .
if the jet crosses the progenitor much faster than the typical hydrodynamic timescale in the system , the progenitor remains roughly unchanged during the complete jet propagation through it and , therefore , we do not need to care about the progenitor evolution during such short timescales .
after the initial phase of constant kinetic power injection , both the kinetic power and the injection lorentz factor are linearly decreased according to the laws @xmath69 , and @xmath70 , respectively , for @xmath71 . in this period of decaying injection power ,
the specific energy is kept fixed to the same value as it had at @xmath72 , and the density and the pressure are obtained from the other parameters ( as in the constant injection power phase ) .
after @xmath73s , the flow injection ceases . with
the parametrization considered above , the rest - mass density ( @xmath74 ) during the constant power phase is 154g@xmath75 .
since the rest - mass density of the progenitor around the inner computational boundary is @xmath76g@xmath75 , the injected outflow is initially much lighter than medium in which it is injected .
thus it is expected that the jet propagation velocity across the progenitor is smaller than the speed of light , and one naturally expects to generate relatively thick cocoons surrounding the beam of the jet .
the dynamical evolution of our jet models can be split in two phases .
the first one happens during the period in which the jet drills its way through the progenitor star .
the second one shows up latter , when the jet breaks out of the stellar surface .
the dynamics of our models during this two phases if roughly similar to that outlined by some previous works ( e.g. , @xcite ) and , therefore , we limit ourselves here to provide a shallow description of the most salient features . figure [ fig : contour ] shows a snapshot of the evolution of the density of the model he16n at @xmath77s , when the head of the jet is still in the progenitor .
the left panel of the fig .
[ fig : contour1_2 ] shows the lorentz factor contour at the same time shown in fig .
[ fig : contour ] . the jet is well collimated both inside of the progenitor and as it travels through the ism .
the bow shock develops close to the head of the jet and rises the pressure and the temperature of the envelope region it sweeps up ( in agreement with the findings of @xcite ) .
it takes about 3.2s for the jet to cross the progenitor , hence , the average propagation velocity is @xmath78 .
the right panel of fig .
[ fig : contour1_2 ] shows the lorentz factor of the model he16c at @xmath79s , showing that the head of the jet in model he16c propagates faster than model he16n due to the lower density in model he16c .
s ) of the rest - mass density of model he16n .
the left and right panels are shown with different color scales in order to outline the most salient features of the cocoon and the cavity drilled by the jet ( left ) , and to show more clearly the structure of the beam ( right ) .
both , the vertical and the horizontal axis are scaled by @xmath80 cm .
a strong bow shock surrounds the jet , the cocoon , and the shocked progenitor gas .
a vigorous back flow from the head of the jet can be seen .
some vortices caused by kelvin - helmholtz instabilities are also indicated in the figure .
[ fig : contour ] ] s , and right he16c at t@xmath81s ) .
both , the vertical and the horizontal axis are scaled by @xmath80 cm .
the largest lorentz factors are reached close to the polar axis where the first reconfinement shock appears .
some weak reconfinement shocks can also be seen further downstream .
[ fig : contour1_2 ] ] although the outflow has a finite initial opening angle , the the beam of the jet is almost parallel to the polar axis .
the kinetic energy of the beam is dissipated when it crosses the reverse shock ( i.e , the mach disc ) at the head of the jet . after the beam plasma is decelerated at the mach disc and its pressure
is risen to a much higher value than in the beam , it expands and flows back in a thick cocoon .
the high pressure of the cocoon is the responsible for the beam collimation during the initial phase of propagation inside of the progenitor star .
also during this early stage of the evolution , a strong backflow can be seen flanking the beam of the jet .
some vortices develop between the jet and the backflow caused by the growth of kelvin - helmholtz modes .
an schematic view of this process can be seen in fig .
5a of @xcite .
the propagation of the jet inside the progenitor also drives a cavity limited by a shroud whose density and pressure is larger than in the cocoon .
the shroud is swept up by a reverse shock that results from the interaction between the cavity and the progenitor envelope . however , this reverse shock is not strong enough to rise the temperature above the threshold in which nuclear reactions can take place .
after the jet breaks out the progenitor surface , it proceeds to the ism , which is assumed to be rarefied for model he16n . in this phase ,
cocoon is almost freely released into the ism , because of the negligible pressure of the external medium ( fig .
[ fig : contour2 ] ) . in spite of the fact that the inertial confinement provided by the stellar progenitor is lost in the ism , by the time the jet reaches the surface of the star ,
the beam has accelerated to @xmath82 ( see sect . [
sec : acceleration ] ) and , thus , it has entered into a ballistic regime , where lateral expansion is strongly suppressed . thereby , the jet remains well collimated as it propagates through the ism , and the half - opening angle of the beam reaches only a few degrees .
these collimation properties have been confirmed by means of numerical models with better resolution in the @xmath83-direction ( see appendix ) .
s. both , the vertical and the horizontal axis are scaled by @xmath80 cm . by this time
, the head of the jet has already broken out of the progenitor surface , and the bow shock begins to expand into the ism .
[ fig : contour2 ] ] in this section we focus on the dynamics of the acceleration of the flow to large lorentz factors @xmath84 , as requested by the the standard fireball model ( e.g. , * ? ? ?
within such a theoretical model , an initial release of thermal energy is later converted into kinetic energy of the flow , as it expands into a dilute and cold environment .
on the basis of this model , a number of numerical results have assessed that a jet , injected through a pre - established nozzle , is able to drill its way through a collapsar can reach an ultrarelativistic regime under likely inflow conditions ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and this paper ) .
we focus here on the details of the dynamical phase in which a kinematically mildly relativistic jet ( @xmath85 ) speeds up to the ultrarelativistic regime @xmath84 converting its initial ( thermal ) energy @xmath86 into kinetic energy . to drive the discussion , we take as a prototype case that of model he16n .
the acceleration process of other models is very similar .
figures [ fig:1dprofiles]a-[fig:1dprofiles]c show one dimensional profiles of density , pressure , and lorentz factor along the polar axis , at different times .
these profiles are qualitatively similar to the ones shown in previous papers ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . during the injection the lorentz factor increases linearly ( fig .
[ fig:1dprofiles]c ) , whereas both the density and pressure decrease as @xmath87 and @xmath88 ( fig .
[ fig:1dprofiles]a , b ) , respectively .
this is not unexpected , since the fluid expands radially ( almost freely ) in that region , where there are no shocks .
note that this result differs a bit from models where the generation of the outflow is considered @xcite . if jet injection conditions are set through a nozzle at a certain distance to the center , the variability imprinted by the highly dynamical generation of the jet is erased .
clearly , this minimizes the number of internal shocks in the outflowing jet .
@xcite show in their fig . 2 that the outflow can accelerate to lorentz factors which are smaller than those attained in this work and in others where
jet injection conditions are assumed .
this is because of the modulation in the growth of the lorentz factor imposed by the development of kelvin - helmholtz modes in the course of the very early outflow evolution @xcite .
nevertheless , models which consistently include the outflow generation ( by thermal energy deposition ) , can still accelerate to ultrarelativistic lorentz factors at large distance from the source .
in such models the flow is kept hotter than in the present ones because of the occurrence of internal shocks already inside of the progenitor . however , since the outflow is optically thick , the thermal energy is not radiated away , but converted ( later ) into kinetic energy @xcite .
( d ) along the polar axis at times @xmath891.0 , 3.0 , 3.5 , 4.5 and 5.0s ( see legends ) , corresponding to model he16n .
[ fig:1dprofiles ] ] one can notice that the maximum lorentz factor is reached behind the first recollimation shock in the jet ( figs .
[ fig : contour1_2 ] and [ fig:1dprofiles]c ) coinciding with the location of the minimum density in the beam of the jet ( figs .
[ fig : contour ] and [ fig:1dprofiles]a ) . inside of the progenitor
star the jet reaches a maximum lorentz factor @xmath90 ( see dashed line in fig .
[ fig:1dprofiles]c corresponding to @xmath91s ) .
we note that the quick acceleration to large lorentz factor happening in the present models is a direct result of the fact that the flow is injected into a finite opening angle .
differently , @xcite , who injected the jet parallel to the polar axis of the progenitor , did not find such a fast acceleration . in the later case ,
the occurrence of recollimation shocks much closer to the injection nozzle prevents the development a freely expanding , unshocked jet extending so far away as in the models of this work .
indeed , such internal , recollimation shocks are the responsible of the confinement of the jet @xcite .
nonetheless , the head propagation speed is very similar in @xcite as it is found here and , hence , also the crossing time of the progenitor by the jet . in @xcite
an acceleration of the jet to @xmath84 at larger distances from the source happens ( by thermal - to - kinetic energy conversion ) , but it still takes place inside of the progenitor star .
this resembles more the acceleration process of outflows generated self - consistently from the inner engine as pointed out above .
the absolute maximum lorentz factor ( @xmath92 ) is attained by the time we begin to reduce the luminosity of the inflow ( at @xmath93s , see [ sec : numerics ] ) . by this time
, the head of the jet has already broken up the surface of the progenitor .
this numerical value is smaller than the maximum one which could be potentially reached , provided the conditions of the injected outflow , according to which @xmath94 , which is an indication of the maximum potentially reachable lorentz factor , along a flow line , if all thermal energy were converted to kinetic energy . after the energy deposition is switched off ( @xmath91s ) , the region previously occupied by the unshocked beam of the jet ( up to the first recollimation shock of the beam ; fig .
[ fig:1dprofiles]c ) begins to shrink and to incorporate mass from its lateral boundaries ( c.f .
, rest - mass density profiles between times 3.5s and 5.0s in fig .
[ fig:1dprofiles]a up to @xmath95 cm ) .
this fact reduces progressively both the beam lorentz factor ( fig .
[ fig:1dprofiles]c ) and the ability of the beam fluid to reach large asymptotic values of @xmath5 ( fig .
[ fig:1dprofiles]d ) . thereby ,
if the jet is injected with a finite half - opening angle , the ultrarelativistic part of the outflow ( i.e. , that where @xmath84 ) may persist only if the activity of the engine does not cease
. the evolution of our models has been followed until the head of the jet reaches the outermost radial computational boundary at @xmath96 cm . at such distances ,
the outflow is still optically thick and , therefore , radiation can not escape freely . the jet and the surrounding cocoon , past a transient phase , that happens after the jet breaks out of the stellar surface , enter into a quasi self - similar phase , where the properties of the outflow roughly scale with the distance of propagation and become almost time independent .
this fact allows us to extrapolate the properties of the jet from distances @xmath97 cm to the typical distances where the afterglow will take place ( namely , @xmath98 cm ) .
thus , we can roughly estimate the angular energy distribution per unit solid angle ( @xmath16 ) that can be potentially emitted at the afterglow phase . in order to derive @xmath16
we have to make several assumptions .
first , we assume that a fix fraction ( the same everywhere ) of the total energy ( internal plus kinetic ) will be converted into electromagnetic radiation .
basically , this assumption is equivalent to state that the angular profile of the observed non - thermal radiation is simply a scaled version of the total energy angular profile .
certainly , this is a rough approximation , since the non - thermal radiation from @xmath99-rays to radio frequencies will be produced by synchrotron ( and , perhaps , inverse compton ) processes of particles accelerated at shocks ( or , maybe , along the jet boundary layer ; e.g. , @xcite ) .
obviously , there are shocks of very different properties in the ultrarelativistic beam and in the cocoon and , thus , we may expect somewhat different conversion efficiencies of the outflow energy into radiation in the beam and in the cocoon .
finally , we assume that the angular energy distribution is _ frozen - in _ by the time when the head of the jet reaches the outer computational boundary .
as commented above , our models evolve almost self - similarly a bit after they break out of the stellar surface , and therefore , we expect only a minor time evolution of the angular profiles of @xmath16 .
we point out that the procedure we use to estimate @xmath16 differs from that of @xcite , who derived their @xmath16 profiles from the time integration of the energy flux trough a certain radius . under the hypothesis of self - similar evolution , this is equivalent to integrate , along the radial direction , the energy density of our models ( by the time they reach @xmath44 ) as follows , @xmath100 where the subscripts @xmath101 and @xmath102 are associated to the spherical grid coordinates @xmath103 , and @xmath104 , respectively , and @xmath105 and @xmath83 are the radial velocity in units of @xmath29 and the observer s viewing angle ( measured from the jet axis ) , respectively .
the expression ( [ eq : dedomega ] ) includes the radiation contributions coming from regions outside of the line of sight ( see @xcite ) .
the summation in the radial direction runs from the surface of the progenitor , located at @xmath106 or , equivalently , @xmath107 , to the outermost boundary .
the summation in the azimuthal angle runs from @xmath108 to @xmath109 ( note that due to the assumed axial symmetry , we can copy the computed data of the quadrant @xmath110 to the quadrant @xmath111 ) . in order to avoid accounting for subrelativistic regions , which will not contribute to the afterglow energetics
, we exclude the contributions of numerical cells where @xmath112 and @xmath113 in the expression ( [ eq : dedomega ] ) .
have the potential to accelerate , at most , to @xmath114 .
in actuality , the asymptotic lorentz factor of such parcels of fluid , will be much smaller , since they will decelerate as they incorporate mass from the external medium . ]
the absolute value of the observed @xmath115 along every radial direction forming an angle @xmath83 with the polar axis depends , among other things , on two parameters whose exact value is not well constrained , neither by observations nor by the present day theory .
these are ( i ) the efficiency of energy conversion to radiation , and ( ii ) the total energy injected .
therefore , we will show only the angular profiles of @xmath115 normalized to the maximum value @xmath116 found for each model .
figure [ fig : distri1 ] shows the normalized angular energy distributions corresponding to models he16c , he16l and he16n , which are prototypes of the types l , m and h , respectively . in the same figure we overplot
fits to the normalized @xmath115 profiles .
the fitting function is a smoothly broken power law ( sbp ) of the form , @xmath117^{1/n},\end{aligned}\ ] ] where @xmath118 is the value of the function @xmath119 at @xmath120 , @xmath121 is the angular location of the break point between the prebreak and postbreak power - laws , whose slopes are @xmath122 and @xmath123 , respectively , and @xmath124 is a numerical factor that controls the sharpness of the break .
note that the maximum value of @xmath119 occurs at @xmath125 } , \label{eq : theta_max}\ ] ] when @xmath126 . otherwise , if @xmath127 and @xmath128 , the function diverges as @xmath129 . by inspection of fig .
[ fig : distri1 ] , the angular energy distributions are remarkably well fitted by the function of eq .
( [ eq : fitting ] ) in the interval @xmath130 , i.e. , in the angular region occupied by the beam of the jet . at smaller latitudes ( @xmath131 )
the model data separates from the fitting function and presents systematically larger values than the latter .
indeed , the data in such an interval can be well fitted by a simple power law , with a slope in the range @xmath132 $ ] ( see the inset in fig . [
fig : distri1 ] ) .
the deviation from the spb function in this angular range is due to the contribution of the expanding , mildly relativistic cocoon .
this cocoon contribution shows up more clearly ( the energy distribution tends to flatten in the range @xmath133 ) if we lower the thresholds on the values of @xmath134 and @xmath62 used to compute the angular energy distribution of each model .
however , as we argued before , lowering these thresholds too much will catch up contributions from numerical cells whose asymptotic lorentz factor is too small to account for typical afterglow .
low metallicity models 16 tb , 16tc , and 16oc , are also well fit by the function of eq .
( [ eq : fitting ] ) in roughly the same interval as the solar metallicity models ( fig .
[ fig : distri2 ] ) .
the values of the fit parameters are comparable to those of type - h models , with which they share a very similar progenitor mass ( @xmath135 ) . however , in these models , the cocoon contribution , which can be fit by a simple power - law with a slope @xmath136 between @xmath137 ( fig .
[ fig : distri2 ] inset ) shows a faster decay of @xmath16 than type - h models for @xmath138 .
the reason for this difference is the much deeper density drop of low - metallicity models close to the star surface ( fig .
[ fig:1dradialmass2 ] ) compared with type - h models ( fig . [ fig:1dradialmass ] ) .
the density of low - metallicity models in the region @xmath139@xmath140 cm is @xmath141 times smaller than in type - h models .
hence , the beam of jets in such low - metallicity progenitors becomes much more ballistic than the corresponding beams of jets in the type - h group .
since more ballistic beams reduce the sideways expansion of their cocoons , this explains that the angular energy distribution in low - metallicity stars is more narrowly concentrated than in solar - metallicity , type - h progenitors . in order to show more clearly the existence of correlations between the properties of the progenitor star and the @xmath16 distribution , we show in fig .
[ fig : mass_al ] the dependence of the postbreak slope @xmath123 and on the stellar progenitor mass @xmath142 .
there exists a correlation between @xmath123 and @xmath142 , such that the slope of lighter progenitors is steeper than that of heavier ones .
there is roughly a linear dependence of @xmath123 on @xmath142 , which displays a relatively large dispersion .
the reason for the dispersion being that for very similar values of the total progenitor mass , the rest - mass density radial profiles can be appreciably different ( see , e.g. , figs . [ fig:1dradialmass ] and [ fig:1dradialmass2 ] ) .
this is particularly true in heavy progenitor models ( including type - h and low metallicity models ) . for the prebreak slope @xmath122
we find no obvious correlation with the progenitor mas , but in all the models considered here is very small ( @xmath143 ; tab . [ tab : progenitor ] ) . on the total mass of the progenitor .
we identify the models by the last letter in the model name , e.g. , the label a stands for model he16a .
the labels tb , tc , and oc stand for the models 16 tb , 16tc , and 16oc , respectively . [ fig : mass_al ] ] on the mass loss rate in units of the _ typical _ mass loss rate considered by @xcite .
the naming convention is the same as in fig .
[ fig : mass_al ] .
[ fig : massloss_al ] ] we have also investigated the dependence of the slope @xmath123 and @xmath122 on the mass loss rate @xmath144 assumed in models of @xcite ( see tab . [ tab : progenitor ] ) . figure [ fig : massloss_al ] shows that there exist a good correlation between @xmath123 and @xmath144 , while @xmath122 seems to be independent of @xmath144 .
the @xmath145 correlation tells us that , models with a larger mass loss rate posses a steeper slopes .
this is not surprising considering the previously found correlation between @xmath123 and @xmath142 , since the stellar progenitor mass is mostly determined by the amount of mass lost in the form of winds during the latest stages of its evolution .
we have not found any other good correlation between the fit parameters ( other than @xmath123 ) and the gross properties of the progenitors ( radius , average density , total angular momentum , rotation period , mass of the iron core , etc . ) .
we shall note that the angular distribution @xmath16 is not directly observable .
instead , the isotropic equivalent angular energy per solid angle @xmath146 can be detected .
figure [ fig : distri3 ] shows the equivalent isotropic angular energy distribution for prototype models belonging to the types l , m and h , normalized to the value of the distribution at @xmath147 . as can be seen from fig .
[ fig : distri3 ] ( see also tab . [
tab : progenitor ] ) , the values of @xmath148 are also negative for the @xmath146 , while the corresponding values @xmath122 are close to zerofor @xmath16 ( approximately , @xmath148@xmath149 ) .
this happens because of the small value of the solid angles close to the symmetry axis , which makes systematically larger the higher latitude values of @xmath146 than those of @xmath16 .
the values of @xmath150 are systematically smaller than the respective @xmath123 values ( roughly , @xmath150@xmath151 in our standard resolution runs ) . given the tight relations of the slopes in the @xmath16 and @xmath146 distributions , it is not surprising to find that there exists a good correlation between @xmath150 and @xmath142 ( fig .
[ fig : mass_al2 ] ) which follows the same qualitative trend as the correlation between @xmath123 and @xmath142 .
but for the equivalent isotropic energy distribution .
[ fig : mass_al2 ] ] to sum up our findings so far , we realize that the values of the parameters of the fit function ( eq.[eq : fitting ] ) are chiefly correlated with the mass of the progenitor . in type - l progenitors ,
the velocity of propagation of the jet is larger than in more massive models , which results in jets developing more massive and hotter cocoons in type - m and type - h models than in light progenitors .
this fact is evident by looking at fig .
[ fig : comparison ] , where the pressure in the cocoon of the model he16n is higher than that in the model he16c in the course of the propagation of the jet up to the progenitor surface .
the differences in the propagation velocity of the injected jets are set , to a large extend , by the differences in the average density between progenitors of different mass .
the less massive progenitors of our sample ( type - l models ) tend to have the smaller average densities ( see , e.g. , fig .
[ fig:1dradialmass ] , where model he16c displays a smaller density than model he16n at every radial point ) .
consistently , jets propagating in type - l progenitors are faster as can be observed in fig . [
fig : average_velo ] .
however , the time needed to reach the progenitor surface presents a dependence with the progenitor mass with a much larger scatter ( fig .
[ fig : time_cross ] ) .
particularly , models he16a , he16f , he16 g and he16h , all of which belong to the type - h group , display a progenitor crossing time comparable to that of the jets in the type - l group ( although its average propagation velocity is comparable to that of models of type - h ) . that s the reason why these models appear as _ outliers _ in the correlations between the slope @xmath152 and the progenitor mass ( figs . [ fig : mass_al ] and with the mass - loss rate ( fig . [ fig : massloss_al ] ) .
since more massive progenitor stars yield hotter cocoons , once the cocoon is erupted through the stellar surface , it undergoes a larger lateral expansion ( compare models he16c and he16n in the lower panels of fig .
[ fig:1dradialmass ] ) , i.e. , the energy carried by the jet spreads towards lower latitudes .
this explains why the energy per solid angle is more concentrated towards the axis when the jet crosses a low mass progenitor .
s ) and he16n ( right : at @xmath153s ) , when the head of the jet reaches progenitor surface in each model , i.e. , at shock break out . both , the vertical and the horizontal axis are scaled by @xmath80 cm .
the radii of the progenitors are @xmath154 cm and @xmath155 cm for the models he16c and he16n , respectively . at the same distance @xmath156 from the injection nozzle , the pressure in both the jet and in the cocoon is higher for model he16n than for model he16c .
this fact explains the larger lateral expansion after shock break out in model he16n than in model he16c .
[ fig : comparison ] ] if we assume that the detectability of an event , for an observer looking such event at a certain viewing angle ( @xmath83 ) , is proportional to @xmath157 ( see , e.g. , @xcite ) , it turns out that @xmath83 should be rather small ( @xmath158 ; fig .
[ fig : detecta ] ) to observe and event produced in collapsar progenitors ( note that the flanks of @xmath146 distribution are quite steep and , thus , it is very unlikely to detect events which are not directly pointing towards the observer ) .
jets produced in type - l progenitors exhibit narrower observability profiles than those injected in more massive starts ( c.f .
, compare the profiles of models he16c and he16n in fig .
[ fig : detecta ] ) .
therefore , it is more unlikely to detect off - axis events produced in light progenitors than in more massive ones .
alternatively , we may state the the lower degree of collimation of relativistic jets in high - mass stars results in a higher probability of observing an event from a high - mass progenitor than a low - mass progenitor .
we also find that the metallicity has little influence on the observability , because for similar progenitor masses , solar - metallicity models and low metallicity ones display almost identical observability angular profiles ( fig .
[ fig : detecta ] ) .
we also notice the very different observability profiles of jets produced in collapsars and jets produced in remnants of mergers of compact objects . in the latter case ,
a non - negligible observability is obtained in the viewing angle interval @xmath159 , because of the larger half - opening angles of the outflows originated in progenitors of short grbs ( see fig . 2 of @xcite ) .
another relevant difference is that collapsar - jets show a detectability with a fast rise close to @xmath160 , followed by a shallower decay beyond the most probable detection angle @xmath161 .
in contrast , most jets produced in merger remnants tend to decay very abruptly beyond the most likely observing angle , and have an observability rise much more moderate than that of collapsar - jets .
the reason for this difference is the much shallower decay of the isotropic equivalent energy at the flanks of the beam in collapsar - jets than in jets from merger remnants .
in this paper we explore the relation of the dynamical properties of ultrarelativistic jets generated in collapsars with the properties of the progenitor stars in which such jets propagate .
we have particularly focused on the correlations that exist between the angular profile of the energy per solid angle ( as seen in the afterglow phase ) and the properties of the progenitors . along the way
, we have pointed out which is the relevance of the fact that our numerical models are set up with a finite injection half - opening angle .
using a non - zero injection angle affects the way in which the conversion of internal - to - kinetic energy takes place .
if the flow is injected parallel to the polar axis , the development of reconfinement shocks happens closer to the injection nozzle than if the flow is injected radially within a cone of finite half - opening angle .
when the recollimation shock occurs far away from the nozzle , the unshocked beam flow accelerates along a larger distance in a rarefaction that precedes such shock .
thus , the beam reaches there larger lorentz factors than if the jet is injected parallel to the polar axis .
the dissipation in cross shocks acts by simply recycling part of the kinetic energy of the outflow into thermal energy .
the thermal energy is not lost , since the jet propagation is roughly adiabatic inside of the progenitor ( radiation losses are negligible there ) . instead
, this thermal energy can be further converted into kinetic energy at larger distances .
this process may happen several times before the outflow becomes transparent and radiation can freely escape .
this explains why , in spite of the differences in the beam dynamics , all models ( independent of the injection half - opening angle ; c.f .
, @xcite ) develop a roughly similar propagation speed and , by the time they reach the head of the jet , the gross properties of the outflow are similar .
we have estimated the angular distribution of energy per solid angle in the afterglow phase by extrapolation of the state of our models when they reach a distance of @xmath162 cm .
this extrapolation relies on the fact that the jets develop a rough self - similar behavior soon after they emerge from the progenitor surface .
our results show that the equivalent isotropic energy per solid angle @xmath163 is only partly consistent with that of @xcite .
however , the results in this paper do appear to be consistent with previous numerical simulations such as those of @xcite and @xcite .
@xcite obtain that the angular energy distribution displays a relatively flat core which is flanked by a region where @xmath164 .
our results show that the core of the distribution ( close to @xmath36 ) is not flat ( but decays as @xmath165 ) and that the energy per solid angle decays much faster than @xmath17 ( it does it as @xmath166 , with a value of @xmath167 depending on the mass of the progenitor ; see below ) .
we can fit the @xmath146-data with sbp functions up to @xmath168 . at smaller latitudes ,
a simple power - law with a slope close to @xmath169 fits better the data . in this region ( @xmath170 )
, the cocoon contribution is the dominant ( at smaller values of @xmath83 , the beam of the jet dominates the energetics ) , and we find there the best consistency with the model of @xcite . we have correlated the properties of the angular @xmath163 distribution of the jet with the fundamental parameters of the progenitor star in which the jet has propagated . we find that the shape of the distribution is mostly influenced by the mass of the progenitor .
when the mass of the progenitor is small ( @xmath171 ) , because of the occurrence of large mass losses due to winds in the latest stages of the star s evolution , then @xmath146 decays faster with @xmath83 than if the mass of the progenitor is large ( @xmath172 ) .
we find that the reason for this behavior is that the average density of the progenitors tends to grow ( approximately ) with the mass .
this means that the average jet propagation speed inside of the star is smaller , the larger is the mass . a smaller jet propagation speed results into thicker and hotter cocoons , since we fix the same mass , momentum and energy fluxes at the nozzle for all models .
also the beam of the jet in the more massive progenitors is wider .
this is the reason why low mass progenitors develop narrower @xmath146 profiles than high mass ones .
the difference in the collimation of the energy distributions resulting from low and high mass progenitors has a direct influence on the number of observed events .
we expect to see more events produced in heavy progenitors than in low mass ones .
one could question whether the correlation that we have found between the mass of the progenitor and the width of the @xmath163 is an artifact of our numerical set up .
we are fixing the luminosity of the jet to be the same independent of the progenitor mass .
however , progenitors of different mass may develop central engines which release different power .
a good proxy of the power of the central engine is the mass of the iron core of the progenitor .
one may expect that the collapse of more massive iron cores results into larger central compact objects .
if the power released by the central engine is dominated by the size of the central compact object , then models with more massive iron cores could release a larger power than models with low mass cores .
then , how do we justify our numerical assumption that the power injected in the jet is roughly independent of the progenitor s mass ? . in support of our point
we argue that , first , according to @xcite data , there is not a one to one correlation between the mass of the iron core and the mass of the progenitor and , second , the iron mass varies by less than a 30% in all the models considered here , while the total mass can be different by a factor of 3 .
thus , within the simplifications we do in our models , the assumption of a common luminosity independent of the mass of the progenitor is justified .
irrespective of these two arguments , if heavier progenitors would result into more luminous central engines , we shall point out that this trend will also result into wider jets and cocoons .
this is a result early pointed out by @xcite , were it was shown that , taking the same progenitor , but increasing the luminosity of the central engine by a factor of 10 , results into thicker cocoons than in cases in which the luminosity is more moderate .
the reason being that the release of a large power triggers large amplitude kelvin - helmholtz instabilities at the basis of the outflow , which transfer a large fraction of the momentum of the beam to a thicker shear layer between the beam and the cocoon . effectively , this process widens the cross sectional area of the beam , reducing its propagation speed and , consistently inflating larger cocoons .
we have also found , that comparing progenitors of similar mass , the metallicity of the star has a small impact on the extrapolated @xmath146 profiles .
the collimation of the jet is similar regardless of the stellar metallicity .
however , the cocoon is more narrowly collimated in low - metallicity stars , because of the large density drop close to their surfaces .
this reduced density makes the jets in low - metallicity stars much more ballistic , once they break out the stellar surface , than in solar - metallicity progenitors .
unfortunately , this difference might not be observable , since it happens in regions where the energy per solid angle is much smaller than at the jet core ( unless orphan afterglows could be detected ; see , e.g. , @xcite ) .
finally , we have found significant differences between the @xmath146 profiles of collapsar - jets and those of jets produced in merger remnants ( i.e. , between the angular energy distribution of jets associated to long and to short grbs ) .
collapsar - jets are more narrowly collimated than jets of merger remnants , and the decay of @xmath146 beyond the central flat core is much steeper in the latter than in the former .
this intrinsic difference manifest itself as a larger chance of detectability of jets from merger remnants at viewing angles up to @xmath173 , while , on the other hand , collapsar - jets could hardly be seen off - axis .
we would like to thank a. heger for his kindness to allow us to use his progenitor models for this study .
we would like to thank dr .
hanawa for his helpful comments .
we also thank the referee for his / her useful comments that helped to improve this paper .
this work is partly supported by the grants - in - aid of the japanese ministry of education , science , culture , and sport ( 19540236 , 20041002 , 20340040 a.m. ) , the grants of the spanish ministry of science and innovation csd2007 - 00050 and aya2007 - 67626-c03 - 01 . m. a. a. is a ramn y cajal fellow of the spanish ministry of education and science .
the calculations were carried out on a nec sx8 , at the cybermedia center , osaka university , on a nec sx8 , at yukawa institute for theoretical physics , kyoto university , on the space science simulator ( nec sx6 ) , at the japan aerospace exploration agency , and on a nex sx9 , at the center for computational astrophysics , national astronomical observatory of japan .
the main purpose of this appendix is to justify the choice of numerical resolution that we have used in the main body of this paper .
we have picked up three models , he16c , he16l and he16n ( representative of the models of the respective types - l , -m and h ) and performed a resolution study by progressively increasing the numerical resolution .
our standard models have a working resolution @xmath174 zones .
the standard computational grid is uniform in both @xmath39- and @xmath83-coordinates .
in addition to the standard resolution , two higher resolutions have been considered .
first , we compute models he16c - m , he16l - m , and he16n - m ( tab . [
tab : progenitor ] ) , which have an intermediate resolution of @xmath175 zones . in these models
the radial grid is uniform in @xmath39 and the smallest radial grid spacing is @xmath176 cm , i.e. , 3/4 times smaller than that of our standard resolution cases .
the polar grid possesses a uniform region close to the symmetry axis ( @xmath177 ) , where @xmath178 , followed by a uniformly spaced region in @xmath179 ( @xmath180 ) .
the reason to consider two different regions in the @xmath83-spacing is that all our jet models develop cocoons whose angular extension is @xmath181 , i.e. , all the dynamics develops in a wedge covered by a finer mesh in our computational grid .
thus , an increased resolution in the abovementioned wedge surrounding the polar axis yields an effective increase of the numerical resolution in the whole computational grid .
even higher resolution models ( he16c - h , he16l - h , and he16n - h ) have also been computed . in this case , the grid consists of @xmath182 zones uniformly spaced in @xmath39 , with a minimum radial grid spacing @xmath183 cm , i.e. , one half of the width of smallest radial spacing of the standard resolution case .
we take in this case @xmath47 zones , also split in two regions : a uniform grid in the interval @xmath184 ( @xmath185 ) , followed by a uniformly spaced region in @xmath179 ( @xmath180 ) . going to even larger resolutions in
the @xmath83-direction increases the total computational time up to prohibitive limits due to the increased number of time steps associated to the fulfillment of the courant condition .
we note that the jet dynamics is rather independent of the resolution , and also the gross morphological features are converged at the standard resolution , though , of course , finer details show up both in the cocoon and in the beam ( see fig . [
fig : pressure_reso ] ) .
therefore , when we take radial averages to compute the angular energy distribution of our models ( eq . [ eq : dedomega ] ) , the differences are relatively small ( see below ) , and our standard resolution models can be considered to be sufficiently resolved to account for such global energetic properties . the time - scales to cross the progenitor and/or to reach the outer computational boundary ( at @xmath186 cm ) are slightly different from each other .
since the jet propagates at slightly different speed depending on the resolution , the three snapshots correspond to slightly different evolutionary times .
both , the vertical and the horizontal axis are scaled by @xmath80 cm .
[ fig : pressure_reso ] ] figure [ fig : energy_distri_reso ] shows the isotropic equivalent angular energy distributions of models computed with three different resolutions . in all cases , a sbp function ( eq . [ eq : fitting ] ) fits properly the data up to @xmath187 . at larger latitude , the contribution of the expanding cocoon component dominates . since each distribution is normalized independently to its absolute maximum , the distributions corresponding to different resolutions do not overlap , but they show the same shape .
the fitting parameters for the higher resolution cases are also listed in tab .
[ tab : progenitor ] .
there , we can see that the same correlations found in models with the standard resolution are reproduced at higher resolutions , namely , the correlation between @xmath188 or @xmath150with the progenitor mass ( sect .
[ sec : dedomega ] ) .
however , at higher resolution , we could guess additional correlations which are not obvious in models run with the standard resolution .
particularly , the parameters @xmath121 and @xmath189 become smaller as the progenitor mass increases .
this trend is however , an artifact of the models chosen as prototypes of each mass type .
they are such , that the radius of the progenitor roughly grows with the mass .
as pointed out by @xcite , progenitors with larger radii provide a larger inertial confinement which prevents the lateral expansion of the jet .
hence , for the models chosen in this resolution study , a larger progenitor mass yields better collimated jets .
taking the whole sample of models , but run at higher resolution , the correlations between @xmath121 or @xmath189 would not exist . cccc|ccccc|ccccc model & mass & total & radius & @xmath122 & @xmath123 & @xmath190 & @xmath191 & @xmath124 & @xmath192 & @xmath193 & @xmath194 & @xmath195 & @xmath196 + & loss & mass /@xmath197 & /@xmath198 cm & /@xmath199 & & ergs & & & & & ergs & & + he16a & 0 & 15.70 & 3.86 & -1.22 & -3.27 & 8.65 & 2.37 & -0.686 & -1.01 & -4.24 & 2.45 & 2.35 & -0.696 + he16b & 1.0 & 5.10 & 4.91 & -1.72 & -3.68 & 6.02 & 2.12 & -0.632 & -1.02 & -4.45 & 2.14 & 2.02 & -0.694 + he16c & 1.0 & 5.15 & 4.75 & -2.99 & -3.58 & 6.61 & 2.03 & -0.706 & -1.03 & -4.38 & 2.40 & 1.95 & -0.765 + he16d & 0.3 & 9.53 & 4.42 & -1.22 & -3.05 & 8.84 & 2.21 & -0.746 & -1.01 & -3.99 & 2.72 & 2.18 & -0.768 + he16e & 0.1 & 12.86 & 5.71 & -2.60 & -2.73 & 9.64 & 1.84 & -0.896 & -1.03 & -3.64 & 3.64&1.80 & -0.942 + he16f & 0.03 & 14.80 & 3.52 & -1.33 & -3.15 & 9.28 & 2.41 & -0.726 & -1.01 & -4.09 & 2.63&2.37 & -0.746 + he16 g & 0.01 & 15.56 & 3.31 & -1.27 & -3.26 & 8.55 & 2.44 & -0.719 & -1.01 & -4.30 & 2.24&2.47 & -0.705 + he16h & 0 & 15.68 & 3.31 & -1.71 & -3.41 & 8.96 & 2.34 & -0.690 & -1.02 & -4.37 & 2.58&2.32 & -0.701 + he16i & 0 & 15.88 & 4.54 & -1.21 & -3.17 & 8.12 & 2.33 & -0.714 & -1.01 & -4.14 & 2.32&2.32 & -0.722 + he16j & 1.0 & 5.13 & 4.81 & -3.64 & -3.47 & 7.23 & 1.89 & -0.764 & -1.04 & -4.29 & 2.77&1.82 & -0.824 + he16k & 1.0 & 5.16 & 4.81 & -3.49 & -3.57 & 6.67 & 2.02 & -0.741 & -1.04 & -4.38 & 2.39&1.95 & -0.798 + he16l & 0.3 & 9.58 & 4.18 & -1.01 & -3.19 & 8.51 & 2.27 & -0.707 & -1.01 & -4.15 & 2.53&2.24 & -0.721 + he16 m & 0.1 & 13.04 & 6.29 & -2.29 & -2.80 & 8.89 & 1.93 & -0.851 & -1.03 & -3.69 & 3.24&1.88 & -0.903 + he16n & 0.03 & 14.95 & 6.17 & -2.91 & -2.73 & 9.92 & 1.79 & -0.914 & -1.03 & -3.64 & 3.86&1.75 & -0.965 + he16o & 0.01 & 15.62 & 5.96 & -1.88 & -2.74 & 9.11 & 1.99 & -0.855 & -1.02 & -3.67 & 3.16&1.94 & -0.891 + he16p & 0 & 15.88 & 4.63 & -1.40 & -2.99 & 9.78 & 2.18 & -0.762 & -1.01 & -3.93 & 3.05&2.15 & -0.781 + 16 tb & 0.1 & 15.29 & 4.45 & -1.42 & -3.14 & 8.38 & 2.19 & -0.733 & -1.01 & -4.11 & 2.55&2.18 & -0.741 + 16tc & 0.1 & 15.23 & 4.87 & -1.46 & -3.14 & 7.51 & 2.30 & -0.735 & -1.02 & -4.08 & 2.22&2.27 & -0.754 + 16oc & 0.1 & 14.26 & 4.42 & -1.62 & -2.95 & 9.05 & 2.09 & -0.798 & -1.02 & -3.94 & 2.85&2.09 & -0.798 + he16c - m & 1.0 & 5.15 & 4.75 & -0.72 & -5.77 & 6.71 & 2.70 & -0.486 & -1.01 & -9.81 & 0.310&3.43 & -0.241 + he16l - m & 0.3 & 9.58 & 4.18 & -1.16 & -4.45 & 11.8 & 2.40 & -0.572 & -1.01 & -5.04 & 2.97&2.29 & -0.569 + he16n - m & 0.03 & 14.95 & 6.17 & -1.27 & -3.33 & 17.9 & 1.65 & -0.685 & -1.01 & -4.41 & 3.64&2.06 & -0.645 + he16c - h & 1.0 & 5.15 & 4.75 & -1.99 & -6.18 & 4.40 & 2.64 & -0.377 & -1.00 & -7.76 & 1.08&2.98 & -0.397 + he16l - h & 0.3 & 9.58 & 4.18 & -1.04 & -3.90 & 11.0 & 2.22 & -0.600 & -1.01 & -6.36 & 2.04&2.74 & -0.446 + he16n - h & 0.03 & 14.95 & 6.17 & -1.18 & -3.22 & 12.4 & 1.96 & -0.704 & -1.01 & -4.62 & 5.74&1.79 & -0.596 + | collapsars are fast - spinning , massive stars , whose core collapse liberates an energy , that can be channeled in the form of ultrarelativistic jets .
these jets transport the energy from the collapsed core to large distances , where it is dissipated in the form of long - duration gamma - ray bursts . in this paper
we study the dynamics of ultrarelativistic jets produced in collapsars .
also we extrapolate our results to infer the angular energy distribution of the produced outflows in the afterglow phase .
our main focus is to look for global energetical properties which can be imprinted by the different structure of different progenitor stars .
thus , we employ a number of pre - supernova , stellar models ( with distinct masses and metallicities ) , and inject in all of them jets with fixed initial conditions .
we assume that at the injection nozzle , the jet is mildly relativistic ( lorentz factor @xmath0 ) , has a finite half - opening angle ( @xmath1 ) , and carries a power of @xmath2ergs@xmath3 . in all cases ,
well collimated jets propagate through the progenitor , blowing a high pressure and high temperature cocoon .
these jets arrive intact to the stellar surface and break out of it .
a large lorentz factor region @xmath4 develops well before the jet reaches the surface of the star , in the unshocked part of the beam , located between the injection nozzle and the first recollimation shock .
these high values of @xmath5 are possible because the finite opening angle of the jet allows for free expansion towards the radial direction .
we find a strong correlation between the angular energy distribution of the jet , after its eruption from the progenitor surface , and the mass of the progenitors .
the angular energy distribution of the jets from light progenitor models is steeper than that of the jets injected in more massive progenitor stars .
this trend is also imprinted in the angular distribution of isotropic equivalent energy . |
from the 1970s through the early 1990s , studies of galaxy clustering drove much of the progress in cosmology .
measurements of steadily improving dynamic range and precision demonstrated good agreement with the predictions of a cosmological model incorporating scale - invariant , gaussian primeval fluctuations modulated by the transfer function expected in a universe dominated by cold dark matter ( cdm ) with @xmath9 ( where @xmath1 is the matter density parameter and @xmath10 ) .
the advent of multi - fiber galaxy redshift surveys and improved photometric input catalogs has dramatically improved the precision of clustering measurements over the last decade , beginning with the las campanas redshift survey ( lcrs ; @xcite ) and continuing with the two - degree field galaxy redshift survey ( 2dfgrs ; @xcite ) and the sloan digital sky survey ( sdss ; @xcite ) . in parallel , numerical simulations and
numerically tested analytic approximations have turned the task of calculating non - linear dark matter clustering from specified initial conditions into an essentially solved problem .
the principal obstacle to drawing cosmological inferences from galaxy clustering measurements is now the uncertainty in the relation between the distribution of observable galaxies and the underlying distribution of dark matter , the problem known as galaxy bias .
much of the cosmological progress in the last decade has been driven by observations that circumvent this complication , such as cosmic microwave background ( cmb ) anisotropies , weak gravitational lensing , the ly@xmath11 forest , and the type ia supernova diagram .
these new observables favor an inflationary , low-@xmath1 , cdm - dominated model similar to that originally suggested by galaxy clustering , which in turn implies that the galaxies that dominate typical optically selected galaxy surveys must be approximately unbiased , in the sense that the rms galaxy count fluctuations are similar to the rms dark matter density fluctuations on large scales ( see , e.g. , @xcite ) .
however , observed galaxy clustering varies systematically with galaxy luminosity , color , and spectral or morphological type ( @xcite , and numerous references therein ) , and reproducing the observed galaxy correlation function in an inflationary cdm model requires that the bias of the correlation function vary with separation on scales below a few megaparsecs ( @xcite ) .
the advances in observational cosmology over the last few years have also raised the stakes for galaxy clustering studies .
we are no longer interested in , for example , distinguishing @xmath12 from @xmath13 ; instead , we want to constrain @xmath1 at the few percent level to increase the power of tests for the nature of dark energy . despite improvements in semi - analytic and numerical modeling of galaxy formation , it is not clear that these methods predict galaxy bias robustly enough for this kind of precision cosmology . faced with these challenges , most cosmological applications of the 2dfgrs and the sdss have focused on the linear or near - linear regime , where generic arguments suggest that the effects of galaxy bias should be relatively simple .
these `` perturbative '' analyses of large - scale structure play a significant role in the current web of cosmological constraints ( e.g. , @xcite ) , but they are restricted to large scales where even these enormous surveys have limited statistical precision . in this paper
we argue that recent developments in the theoretical description of galaxy bias allow a more aggressive approach to inferring cosmological constraints from galaxy clustering measurements .
we work in the framework of the halo occupation distribution ( hod ) , which characterizes galaxy bias in terms of the probability distribution @xmath14 that a dark matter halo of virial mass @xmath15 contains @xmath16 galaxies of a specified type , together with prescriptions for the spatial and velocity bias of galaxies within dark matter halos @xcite . here
the term `` halo '' refers to a bound dark matter structure of typical overdensity @xmath17 , in approximate dynamical equilibrium , which may be the individual halo of a single bright galaxy or the common halo of a galaxy group or cluster .- body simulations by a friends - of - friends algorithm with linking length @xmath18 , but the precise definition of halo does not matter provided that one is consistent throughout all calculations . ]
the flow diagram in figure [ fig : flowchart ] , adapted from @xcite , sketches the interplay between the `` cosmological model '' and the `` physics of galaxy formation '' in determining observable galaxy clustering , which we take to include both the traditional statistics measured from redshift surveys and the galaxy - matter correlations measured by galaxy - galaxy lensing ( e.g. , @xcite ) . on the left side ,
the cosmological model , which specifies the initial conditions and the energy and matter contents of the universe , determines the mass function , spatial correlations , and velocity correlations of the dark halo population .
the intervening box indicates that the only features of the cosmological model that really matter in this context are @xmath1 and the amplitude and shape of the linear matter power spectrum @xmath19 , here represented by @xmath2 ( the rms linear matter fluctuation in @xmath20 spheres ) , the inflationary spectral index @xmath21 , and the transfer function shape parameter @xmath3 , which itself depends on the values of @xmath1 , @xmath22 , and the baryon density @xcite .
other features of the cosmological model , such as the energy density and equation of state of the vacuum component , may have an important impact on other cosmological observables or on the _ history _ of matter clustering , but they have virtually no effect on the halo population at @xmath23 , if the shape of @xmath19 and the present day value of @xmath2 are held fixed ( see @xcite ) . on the right side of figure
[ fig : flowchart ] , the box titled `` galaxy formation physics '' represents the additional processes such as shock heating , radiative cooling , star formation , feedback , and mergers that are essential to producing galaxies and determining their masses , luminosities , diameters , colors , and morphologies .
these physical processes operate in the background provided by the evolving halo population , so together with the cosmological model they determine the hods of different classes of galaxies .
the halo population and the hod together determine galaxy clustering and galaxy - mass correlations .
one can , of course , use hydrodynamic simulations or semi - analytic models to predict galaxy clustering statistics directly , without computing the hod as an intermediate step ( e.g. , @xcite ) .
however , these predictions reflect the combination of the cosmological model and the galaxy formation theory , and unless one has complete confidence in the latter , one can not draw secure conclusions about the former .
we would therefore like to know how well cosmological parameters can be constrained _ without _ relying on a detailed theory of galaxy formation , by using the data themselves to determine the relation between galaxies and dark matter .
for this purpose , the hod formulation of bias has two key strengths , both emphasized by @xcite .
the first is the division of labor implied by figure [ fig : flowchart ] : galaxy formation physics influences the hod , but the properties of the halo population defined at @xmath17 are determined almost entirely by the much simpler physics of gravitational clustering , which can be modeled accurately using @xmath16-body simulations or numerically tested analytic approximations .
the second is completeness : given a cosmological model and a specified hod , one can calculate any galaxy clustering statistic on any scale , by populating @xmath16-body halos or by drawing on a steadily expanding array of analytic techniques .
the hods of different galaxy classes therefore encode , statistically , all the aspects of galaxy formation physics that are relevant to predictions of galaxy clustering .
( we discuss an important caveat to this statement below . )
hod modeling enables one to take full advantage of galaxy clustering data , with no restriction to large scales or particular statistics .
the goal of our empirical approach is to reverse the causal arrows in figure [ fig : flowchart ] , working backwards from the data to properties of the hod and the halo population , and from there to conclusions about galaxy formation physics and fundamental aspects of cosmology .
@xcite considered a fixed cosmological model and showed how changes to the hod affect many of the traditional statistical measures of galaxy clustering , such as correlation functions , the group multiplicity function , and redshift - space distortions .
they argued that the complementary information from different statistics and different spatial scales should permit an accurate empirical reconstruction of the hod for a specified cosmology , yielding physically informative tests of galaxy formation theories .
@xcite showed that @xmath1 , @xmath2 , and the power spectrum shape have non - degenerate effects on the _ halo _ population , and they argued that changes to the galaxy hod could not fully mask the effects of cosmological parameter changes . here
we complete this theoretical program by showing how galaxy clustering measurements can constrain simultaneous changes to the cosmological model and the galaxy hod .
our results add quantitative teeth to the qualitative arguments and speculations in @xcite , @xcite , and @xcite .
we define our cosmological parameter space by the values of @xmath1 , @xmath2 , and @xmath3 .
we concentrate mainly on @xmath1 and @xmath2 because @xmath3 ( or , more generally , the shape of the linear power spectrum ) can be well constrained by combining the large - scale galaxy power spectrum with other observables like cmb anisotropy and the ly@xmath11 forest .
we first compute predicted values of a number of galaxy clustering observables , assuming a central model with observationally motivated choices of cosmological parameters and the galaxy hod .
we then change one or more cosmological parameters , and we test how well this new cosmological model can reproduce the original clustering `` data '' given complete freedom to vary the hod within a very flexible parameterization . to gain insight , we first explore several interesting axes within the @xmath24 parameter space , varying parameters individually or in physically motivated combinations .
we arrive at definite numbers by assuming that each of our 30 observables , some of which represent the same clustering statistic measured at multiple scales , can be measured with 10% uncorrelated fractional uncertainty from a survey like the sdss .
this assumption seems roughly plausible , but the ultimate strength of cosmological conclusions will depend on the precision and dynamic range of the measurements . forecasting this precision for a survey like the sdss , including
the covariance of errors among different observables , is itself a major theoretical task , which we will not undertake here . for our theoretical investigation ,
we further restrict ourselves to observables for which we have reasonable analytic approximations , and we suspect that our quantitative conclusions will in the end prove overly pessimistic because we must omit some observables that contain significant additional information . as discussed by @xcite , the completeness of the hod as a description of bias rests on the assumption that the galaxy content of a halo of virial mass @xmath15 is statistically independent of the halo s larger scale environment .
@xcite show that this assumption accurately describes the galaxy population in weinberg et al.s ( @xcite ) hydrodynamic cosmological simulation .
it has fundamental theoretical roots in the excursion set model of @xcite , which predicts that the statistical features of a halo s assembly history depend only on its present mass ; indeed , all semi - analytic galaxy formation models that use statistically generated merger trees make this assumption implicitly .
since the halo mass function itself varies with environment , with high - mass halos absent in low - density regions , models that tie galaxy populations to halo masses still predict strong correlations between galaxy properties and the large - scale environment , in good agreement with observations ( e.g. , @xcite ) .
however , while @xcite show that @xmath16-body halos of mass @xmath25 have similar properties and formation histories in different environments , recent studies show a substantial correlation of halo formation redshift with large - scale overdensity for lower mass halos ( @xmath26 ) , contradicting the simplest form of the @xcite model ( @xcite ; see also @xcite ) . as discussed below in [ sec : environment ] , we expect the impact of such dependence on the clustering statistics of mass- or luminosity - thresholded galaxy samples to be small , but given the level of precision we eventually hope to reach , it will probably be necessary to allow for it when fitting the data .
strategies for incorporating environmental variations into hod modeling will require guidance from the next generation of hydrodynamic and semi - analytic models of galaxy clustering , and we reserve this task for future investigation .
ultimately , the assumption of an environment - independent hod , or of any particular parameterization of environmental dependence , must be tested empirically by checking that the adopted model can consistently explain the variations of galaxy clustering with large - scale environment .
many current efforts to constrain cosmological parameters with galaxy clustering data make ( explicitly or implicitly ) the much stronger assumption that galaxy bias can be adequately described by a linear model , @xmath27 , on the relevant scales . here
@xmath28 and @xmath29 are the galaxy and dark matter density contrast , respectively , and the linear bias factor @xmath30 may depend on galaxy type but is assumed to be independent of scale . in this case , @xmath31 , and the shape of the galaxy power spectrum provides powerful cosmological constraints in combination with cmb data even if @xmath30 is unknown @xcite .
large - scale redshift - space distortions measure @xmath32 @xcite , which in combination with clustering measurements constrains @xmath33 , where @xmath34 is the measured rms fluctuation of the _ galaxy _ density contrast at @xmath20 . the mass function or x - ray temperature function of galaxy clusters constrains a similar combination of @xmath2 and @xmath1 @xcite by a route that is independent of galaxy bias . at low redshift ,
the combination of galaxy - galaxy lensing with the galaxy - galaxy correlation function constrains a somewhat different combination of these parameters , @xmath35 , assuming a linear bias model with @xmath36 .
the triangle shape dependence of the reduced galaxy bispectrum can constrain @xmath30 directly @xcite , in this case assuming a quadratic bias model @xmath37 to relate the galaxy and mass density fields at second order @xcite .
general theoretical arguments suggest that linear bias should be a good approximation for the power spectrum on sufficiently large scales , provided that the efficiency of galaxy formation is determined by the local ( @xmath38 few mpc ) environment @xcite .
numerical experiments using such `` local '' bias models also show that the `` @xmath30 '' multiplying the power spectrum amplitude should be similar to the `` @xmath30 '' affecting redshift - space distortions @xcite .
thus , the combination of these perturbative galaxy clustering analyses can in principle yield separate constraints on @xmath3 ( measured directly from the power spectrum shape ) , @xmath1 , and @xmath2 , with the degeneracy of @xmath1 and @xmath2 broken either by the bispectrum analysis or by the different @xmath1 dependence of the galaxy - galaxy lensing constraint and the @xmath39 or cluster normalization constraints .
the applications of these techniques to 2dfgrs and sdss data show an impressive degree of internal consistency and good agreement with external cosmological constraints ( @xcite ) .
however , reliance on the linear or quadratic bias approximations restricts these analyses to large scales , and it is not clear that these approximations hold consistently at the level of accuracy desired for further improvements ( better than 10% , say ) , since scatter in the relation between galaxy and mass densities can have different effects on different statistics and at different scales @xcite .
the hod approach to galaxy clustering analysis substitutes a much more general model of galaxy bias , and it makes use of high - precision measurements from small and intermediate scales ( from sub - mpc to @xmath8 ) in addition to the lower precision measurements in the perturbative regime . furthermore , regardless of the strength of the cosmological constraints , the hod parameters derived from the data themselves provide detailed tests of galaxy formation models .
the program of hod - based interpretation of observed galaxy clustering is , in fact , well underway . @xcite and @xcite used hod - type bias models to interpret the correlation functions and pairwise velocity dispersions measured from the lcrs and the point - source catalog redshift survey ( pscz ; @xcite ) .
@xcite , @xcite , and @xcite used the group multiplicity function to constrain the galaxy occupations of high - mass halos .
@xcite , @xcite , and @xcite applied hod modeling to galaxy - galaxy lensing measurements from the sdss .
@xcite , analyzing a volume - limited sample of bright ( @xmath40 ) sdss galaxies , showed that hod models naturally explain the observed deviation from a power - law correlation function .
@xcite measured the luminosity and color dependence of the sdss galaxy correlation function and used it to infer the luminosity and color dependence of the galaxy hod , finding results in good qualitative agreement with theoretical predictions @xcite .
@xcite used a similar approach to investigate the halo occupations of early- and late - type galaxies in the 2dfgrs , while @xcite derived hods of red and blue 2dfgrs galaxies from the group catalog of @xcite , testing the consistency of their result by comparing predicted and observed correlation functions .
@xcite used the sdss measurements for @xmath40 galaxies in conjunction with cmb anisotropy data to infer simultaneous constraints on hod and cosmological parameters .
@xcite used hod modeling of zehavi et al.s ( @xcite ) clustering measurements to predict cluster mass - to - light ( @xmath41 ) ratios , and they inferred constraints on @xmath42 by comparing to published @xmath41 measurements .
hod models have been applied to the interpretation of high - redshift clustering by @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , and @xcite .
finally , @xcite have initiated a comprehensive program similar to the one described here , based on the closely related conditional luminosity function ( clf ) formalism ( see also @xcite ) .
we discuss the similarities and differences between the hod and clf approaches in
[ sec : discussion ] . in the next section
we define our class of cosmological models , list the analytic approximations we use for properties of the halo population , and describe our flexible parameterization of the galaxy hod . in [ sec : observables ] we describe the methods and approximations that we adopt for analytic calculation of galaxy clustering observables .
section [ sec : results ] presents our main results , showing the ability of complementary galaxy clustering measurements to constrain the galaxy hod and cosmological parameters simultaneously .
section [ sec : environment ] discusses the issue of environmental variations of the hod .
section [ sec : discussion ] summarizes our findings and discusses the overall prospects for application of this approach .
a reader familiar with the field who wants an overview of our main results can read the first paragraph of [ sec : cosmology ] , skim [ sec : hodpar ] to understand our hod parameterization , then skip to [ sec : results ] , paying particular attention to [ sec : omegam ] , [ sec : general ] , and [ sec : influence ] and figures [ fig : chi2_omega ] , [ fig : chgerr ] , and [ fig : matrix ] .
throughout this paper we adopt spatially flat @xmath43cdm cosmological models with gaussian initial density fluctuations .
the cosmological model is defined by the mass density parameter @xmath1 and the mass fluctuation power spectrum @xmath44 , where @xmath45 is the wavenumber , @xmath21 is the spectral index of the inflationary power spectrum , and @xmath46 is the transfer function with shape parameter @xmath3 .
we parameterize each cosmological model in this paper by @xmath1 , @xmath2 , @xmath21 , and @xmath3 , where @xmath2 is the rms fluctuation of the linear density field filtered with a top - hat filter of radius @xmath47 .
we adopt the parameterization of @xcite for the transfer function @xmath46 , which approximates the evolution of adiabatic primordial fluctuations .
this formulation suffices for our purpose of investigating how galaxy clustering data constrain hod and cosmological parameters , with a single parameter @xmath3 encoding the combined effects of the matter - radiation transition ( at the scale @xmath48 ) and the suppression of fluctuation growth in baryon or massive neutrino components .
observational analyses should use a more accurate transfer function , e.g. , from cmbfast @xcite , but this would not change the sensitivity to the shape and amplitude of @xmath19 . it is important to note that we treat @xmath1 and @xmath3 as independently variable quantities ; we do not automatically change the power spectrum shape when we change @xmath1 . in the context of @xmath43cdm , our approach implicitly assumes that a change in @xmath1 is compensated by a change in @xmath22 or some other parameter to keep the power spectrum at its empirically constrained shape .
our calculations of galaxy clustering statistics rely on analytic descriptions of the density profiles , mass function , spatial clustering , and velocity statistics of dark matter halos .
we draw on the extensive literature that presents numerically tested analytic approximations or numerically calibrated fitting formulae for these quantities .
the density profile of a dark matter halo of mass @xmath15 is assumed to have the navarro - frenk - white ( nfw ) form ( @xcite ) , @xmath49 where the characteristic radius @xmath50 is related to the virial radius @xmath51 of the halo through the concentration parameter @xmath52 , and @xmath53 we define the virial radius by the condition @xmath54 , independent of cosmology , where @xmath55 is the average mass density of the halo within @xmath51 in units of the mean matter density @xmath56 .
the concentration parameter @xmath57 is a function of halo mass .
here we adopt the relation given by @xcite after modifying it according to our constant-@xmath55 halo definition , @xmath58 where @xmath59 , @xmath60 , and @xmath61 is the nonlinear mass .
the mass function of halos can be expressed as @xmath62 where @xmath63 is the space density of halos with mass higher than @xmath15 , @xmath64 is the rms fluctuation of the mass overdensity at a mass scale @xmath15 , @xmath65 with @xmath66 is the threshold density contrast for collapse , and @xmath67 is a dimensionless function .
the mass @xmath61 in equation ( [ eqn : concentration ] ) is defined by the condition @xmath68 ( i.e. , @xmath69 ) .
the function @xmath67 can be approximately derived using the press - schechter ( @xcite ) or excursion set @xcite formalism , or it can be calibrated by fitting the results of @xmath16-body simulations ( e.g. , @xcite ) . here
we use the fitting formula of @xcite , @xmath70\nu \exp\biggl(-\frac{a\nu^2}{2}\biggr),\ ] ] where @xmath71 , @xmath72 , and @xmath73 .
this formula is consistent with the ellipsoidal collapse model ( @xcite ) , and it provides an excellent match to results from simulations ( see , e.g. , @xcite ) . at large scales , the halo two - point correlation function @xmath74 is biased relative to the matter two - point correlation function @xmath75 by a mass - dependent factor @xmath76 .
we adopt sheth et al.s @xcite formula for @xmath77 , based on the ellipsoidal collapse model , @xmath78 , \end{aligned}\ ] ] where @xmath72 , @xmath79 , @xmath80 , and @xmath81 and @xmath82 have the same meaning as in equation ( [ eqn : halomf ] ) .
our calculation of the galaxy 2-point correlation function , discussed in [ sec : spatial ] , also incorporates scale dependence of halo bias .
recent numerical work @xcite suggests that the bias factors given by equation ( [ eqn : halobias ] ) are systematically too high .
since we use the same approximations to calculate the `` observations '' of our central model and the `` predictions '' of others , we do not expect moderate errors in @xmath77 to alter our conclusions .
however , comparisons to real observational data should use the most accurate available bias factors .
the rms three - dimensional ( 3d ) space velocity of halos of mass @xmath15 can be approximated as @xmath83 where @xmath84,\ ] ] @xmath85 $ ] is the fourier transform of the top - hat filter on mass scale @xmath15 , and the smoothing radius @xmath86 is determined through @xmath87 @xcite .
the one - dimensional ( 1d ) rms velocity is simply the 3d one divided by @xmath88 .
the rms 1d velocity difference of halos of masses @xmath89 and @xmath90 at a separation @xmath91 is @xmath92 where @xmath93 represents the correlation term . in our calculations
( see [ sec : dynamical ] ) , we ignore the correlation term , which is an acceptable simplification @xcite .
we assume that the 1d velocity dispersion of dark matter _ within _ a halo of mass @xmath15 has the value @xmath94 expected for a singular isothermal sphere , independent of radius .
moderate departures from this assumption can be accommodated by the velocity bias parameter @xmath95 , described below .
the hod of a specified class of galaxies is defined by the probability distribution @xmath14 that a halo of mass @xmath15 hosts @xmath16 galaxies , together with prescriptions for the spatial and velocity distributions of galaxies within halos relative to those of the dark matter particles . for our purpose in this paper , we want a flexible parameterization of the hod that could describe the relevant features of any physically reasonable model .
we want the constraints on the hod to be imposed by the clustering measurements rather than by our parameterization , and we want to give each cosmological model its best shot at reproducing these measurements before rejecting it .
conceptually , it is useful to separate @xmath96 into a mean occupation function @xmath97 , defined by @xmath98 , and a distribution @xmath99 at fixed halo mass .
there have been numerous investigations of the predictions of semi - analytic models , hydrodynamic simulations , and high - resolution @xmath16-body simulations for @xmath96 ( e.g. , @xcite ; @xcite ; the last three are the most extensive and the most relevant to our considerations here ) .
the predicted hod depends on the defining characteristics of the galaxy sample , such as luminosity , color , and morphological type .
for the most sensitive cosmological investigations , it seems best to work with volume - limited samples defined by luminosity thresholds , which are the largest homogeneous samples ( i.e. , with the same distribution of galaxy types at all distances ) that one can construct from an apparent magnitude - limited survey . in this case
, @xmath97 should be a more or less monotonic function of halo mass , since more massive halos contain , on average , either more galaxies or brighter galaxies .
below some cutoff mass , @xmath97 should drop to zero , as low - mass halos do not contain enough cooled baryons to make a galaxy above the luminosity threshold . instead of a luminosity threshold
, one can impose a threshold in stellar mass , estimated for each galaxy from its luminosity and its color or spectral energy distribution ( e.g. , @xcite ) .
this approach should reduce the scatter between the galaxy observable and the host halo mass by reducing the impact of stellar population variations , thus sharpening the cutoff in @xmath97 and minimizing its environmental dependence .
however , if one starts from an apparent magnitude - limited survey , then the mass - thresholded samples one can create are smaller than the luminosity - thresholded samples because of the range of galaxy colors .
while we do not want to impose strong theoretical priors on our hod parameterization , we have decided to incorporate one of the most generic and useful results that comes from the galaxy formation papers cited above : the distinction between central and satellite galaxies . in hydrodynamic simulations and semi - analytic models ,
most halos contain a galaxy near the center of mass , moving close to the center - of - mass velocity , which is usually more massive and older than any other galaxies in the halo @xcite .
@xcite show that the central - satellite distinction naturally explains one of the quantitatively important properties of the hod , the sub - poisson width of @xmath99 at small mean occupation numbers . for satellite dark matter subhalos , the fluctuations about
the mean occupation are close to poisson , but the number of central galaxies is by definition either zero or one , thus obeying nearest - integer ( bernoulli ) statistics .
fluctuations in the full @xmath99 are substantially sub - poisson as long as the central galaxy makes a significant contribution to @xmath97 .
@xcite show that this argument carries over to galaxies in hydrodynamic simulations and semi - analytic models .
motivated by these results , we model the mean occupation function for galaxies above a luminosity threshold as @xmath100[1+s(m)],\ ] ] where the last set of square brackets represents the sum of one central galaxy and @xmath101 satellites and the first provides a smooth cutoff near @xmath102 .
the form of the cutoff profile corresponds to a log - normal distribution of central galaxy luminosity at fixed halo mass ( see @xcite ) .
we apply the same cutoff profile to satellite galaxy numbers , although in practice these are already well below unity close to @xmath102 . to model @xmath101 in a flexible way , we take the free parameters to be the values of @xmath103 at @xmath104 fixed values of @xmath105 and define the continuous function by a cubic spline that passes through these @xmath104 values . we set @xmath106 and choose values of @xmath103 at @xmath107 , 12 , 13 , 14 , and 15 for the spline fit .
halos more massive than @xmath108 are too rare to have much impact on the clustering statistics we consider here .
theoretical predictions of @xmath101 can be accurately described by a truncated power - law parameterized by a slope , amplitude , and low - mass cutoff ( @xcite ) , but our five - parameter model allows for curvature or inflections that depart from these theoretical predictions .
in addition to the @xmath101 values , the adjustable parameters of equation ( [ eqn : navg ] ) are @xmath109 and @xmath102 .
we always assume that the uncertainty in the galaxy mean density is negligible in comparison to the uncertainty in the clustering measurements .
therefore , once we have specified @xmath101 and @xmath109 , we choose the value of @xmath102 that yields the correct mean number density , rather than treating it as a free parameter .
( we could pick any of the parameters for this treatment , but the mean density is most sensitive to @xmath102 . ) for the clustering measures and analytic approximations used in this paper , the only property of @xmath99 that we need to specify is the second factorial moment @xmath110 , which determines the mean number of pairs per halo .
the number of central galaxies is zero or one with a relative probability determined by the mean @xmath111 . for satellite galaxies ,
we introduce a width parameter @xmath112 defined by @xmath113 the value @xmath114 corresponds to a poisson distribution of satellite numbers @xmath115 at fixed halo mass , consistent with theoretical predictions .
the parameter @xmath112 allows the satellite probability to be broader or narrower than poisson .
the full probability distribution becomes narrow at low @xmath97 in any case because of central galaxies .
the value of @xmath112 determines how quickly the distribution approaches a poisson - like width , thus influencing the number of galaxy pairs in halos with mean occupation numbers @xmath116 1@xmath117several . in the hydrodynamic simulation examined by @xcite ,
satellite galaxies have a radial profile similar to that of the dark matter . however , the physical processes that determine the satellite profile are complex ( see , e.g. , @xcite ) , so there is no reason to expect the profiles to be identical . here
we assume that satellite galaxies in a halo still follow an nfw profile , but we allow the galaxy concentration parameter @xmath118 to differ from the underlying dark matter concentration parameter @xmath57 .
we characterize the relative spatial distributions of satellite galaxies and dark matter by @xmath119 , where a positive ( negative ) @xmath120 means that galaxies are more ( less ) centrally concentrated than dark matter .
we assume that central galaxies always reside at the halo center of mass .
we also allow the velocity dispersion of satellites to differ from the velocity dispersion of dark matter by an average factor @xmath95 , which we assume to be independent of mass .
central galaxies are assumed to move at the halo center - of - mass velocity .
simulations suggest that @xmath121 , but the results are not entirely consistent on this point ( e.g. , @xcite ) , and in any event we want this dynamical question to be settled by observations rather than by theory . @xcite
discuss the influence of @xmath95 on redshift - space distortions in detail .
altogether we have 10 free parameters in our hod prescription seven in @xmath97 ( @xmath102 , @xmath109 , five spline points ) , one ( @xmath112 ) for the second moment of @xmath122 , one ( @xmath120 ) for the spatial bias within halos , and one ( @xmath95 ) for the velocity bias within halos .
one of these ( which in practice we take to be @xmath102 ) is determined by matching the mean galaxy space density once the values of other parameters are specified .
the remaining nine are varied to yield the best match to the observables described in
[ sec : observables ] .
the parameters @xmath112 , @xmath120 , and @xmath95 could in principle depend on halo mass , but the variation would have to be quite strong over a fairly narrow mass range to have an effect that could not be adequately described by an appropriate average value . the main theoretical preconception that is built into our parameterization is the existence of central galaxies .
the central location of these galaxies makes a small but not negligible difference to the two - point correlation function ( see @xcite ) , although this could probably be compensated by @xmath120 .
however , since @xmath123 in practice turns out to be small near the cutoff mass @xmath102 , the central galaxy parameterization necessarily leads to sub - poisson @xmath99 fluctuations at low @xmath97 .
since this aspect of galaxy formation physics is well grounded , we think it is reasonable to adopt it , and it allows us to represent a wider range of physically realistic hods with our finite parameter set .
however , even if we do not make the central - satellite distinction and instead use a functional form like that of @xcite to describe the width of @xmath99 as a function of halo mass , we reach similar conclusions about the ability to break degeneracies between cosmology and galaxy bias .
in essence , the observations drive us to sub - poisson fluctuations at low halo masses whether or not we impose this trend as a theoretical expectation ( see @xcite ) .
the observable quantities considered in this paper fall into three general categories .
first are quantities related to the real - space clustering of galaxies : the rms galaxy number density contrast , the galaxy two - point correlation function , the reduced galaxy bispectrum , and the group multiplicity function .
second are dynamical quantities that depend on galaxy peculiar velocities : the large - scale redshift - space distortion parameter @xmath39 , the galaxy pairwise velocity dispersion , and the estimated virial mass of galaxy groups as a function of group richness .
finally , we include the galaxy - matter two - point cross - correlation function , which can be measured by galaxy - galaxy lensing .
most of these quantities are continuous functions of scale or group multiplicity . to obtain a discrete set of observables , we sample these functions at a discrete set of separations or multiplicities , listed in [ sec : results ] .
we space our samplings widely so that we can treat the statistical errors in each measurement as approximately uncorrelated .
in addition to these observables , we impose the requirement that the hod model reproduce the correct value of the sample s mean space density , implying @xmath124 in effect , we treat @xmath125 as an observable with negligible statistical error . in this paper
we limit ourselves to observables for which we have some reasonable analytic approximation . in many cases ,
these approximations are not accurate enough for application to high - precision clustering measurements from large surveys like the sdss or 2dfgrs or at least their accuracy has not been extensively tested with numerical simulations .
however , we use the same approximations to calculate `` observed '' quantities for our central cosmological model and hod and `` predicted '' quantities for altered models . provided that our approximations realistically capture the variations of the observables with variations in the cosmological and hod parameters , they are adequate for our present purposes of forecasting the ability of clustering measurements to break the degeneracy between cosmology and galaxy bias .
observational applications of these techniques will require further testing and development of analytic approximations .
they could also benefit from analytic approximations for other statistics , or from numerical studies that provide accurate interpolation formulae in the relevant ranges of cosmological and hod parameters . at large scales ,
the galaxy two - point correlation function @xmath126 is biased relative to the matter correlation function by ( the square of ) a number - weighted average of the halo bias factor , @xmath127 we use equations ( [ eqn : halomf ] ) and ( [ eqn : fnu ] ) for @xmath128 and equation ( [ eqn : halobias ] ) for @xmath77 .
the fractional rms fluctuation @xmath129 of galaxy number counts in spheres of radius @xmath86 can be expressed as an integral over @xmath126 , so at large @xmath86 it is related to the rms mass fluctuation @xmath130 by @xmath131 at @xmath132 , the smaller of the two scales we use for @xmath129 , the approximation ( [ eqn : sig_g ] ) is good but not perfect ( see @xcite ) .
more generally , @xmath126 can be decomposed into one - halo and two - halo terms , @xmath133 and @xmath134 , which respectively represent pairs of galaxies residing in the same halo and in separate halos .
the one - halo term dominates at small separations and the two - halo term at large separations , with the transition occurring near the virial diameter of large halos . noting that the total number of galaxy pairs [ @xmath135 is simply the sum of the number of pairs from single halos [ @xmath136 and that from different halos [ @xmath137 , we have @xmath138+\xid(r)$ ] .
the one - halo term @xmath133 can be exactly computed in real space through @xcite @xmath139 where @xmath140 is the average number of pairs in a halo of mass @xmath15 and @xmath141 is the cumulative radial distribution of galaxy pairs , i.e. , the average fraction of galaxy pairs in a halo of mass @xmath15 ( virial radius @xmath51 ) that have separation less than @xmath91 .
the differential function @xmath142 is determined by the profile of the galaxy distribution within the halo .
we assume that there is always a galaxy located at the center of the halo , and others are regarded as satellite galaxies . with this assumption ,
@xmath142 is the pair - number weighted average of the central - satellite pair distribution @xmath143 and the satellite - satellite pair distribution @xmath144 ( see , e.g. , @xcite ) , @xmath145 the central - satellite galaxy pair distribution @xmath143 is just the normalized radial distribution of galaxies within halos , and the satellite - satellite galaxy pair distribution @xmath144 can be derived through the convolution of the galaxy distribution profile with itself ( @xcite ) . in the case of an nfw profile , as used in this paper , @xmath144 can be analytically expressed ( see appendix [ sec : appendixa ] ; also see @xcite ) . at large scales
, the two - halo term becomes @xmath146 , where @xmath147 is the galaxy number - weighted halo bias factor of equation ( [ eqn : bg ] ) . at smaller scales
, one must account for the convolution of galaxy profiles over the finite size of halos , scale dependence of the halo bias factor , and the fact that separate halos by definition do not interpenetrate . because convolutions become multiplications in fourier space , it is convenient to calculate the two - halo component of the galaxy power spectrum @xmath148 and fourier transform it to get @xmath134 @xcite .
our full method of calculating the two - halo term is described in appendix b of @xcite .
our adopted treatment of halo exclusion is referred to there as the `` spherical '' exclusion method , and we use the @xcite formula for @xmath77 instead of the modified formula proposed by @xcite .
we have used this method in earlier papers @xcite .
galaxies within the same dark matter halo constitute a galaxy group .
the cumulative multiplicity function of galaxy groups , which is the number density of galaxy groups that have @xmath16 or more members , is @xmath149 in this work we make the simplifying approximation that @xmath150 is a nearest - integer distribution , where @xmath16 is one of the two integers bracketing @xmath97 , and the relative probability is determined by having the right mean .
the space density of galaxy groups with multiplicity @xmath151 is approximately equal to the space density of halos with mass @xmath152 , so given a cosmological model that specifies @xmath63 one can effectively `` read off '' the high end of the mean occupation function @xcite .
a full observational implementation should take into account both the width of @xmath99 and the scatter between estimated and true multiplicities introduced by the group finding algorithm .
we consider galaxy groups with five or more members , so the approximation ( [ eqn : multi ] ) is fairly good . in redshift space ,
the peculiar velocities of galaxies distort the galaxy power spectrum and correlation function , producing anisotropy in which the line of sight is a preferred direction . by measuring this distortion on large scales , where it is caused by coherent flows into overdense regions and out from underdense regions , one can infer a combination @xmath153 of the density parameter and the large - scale galaxy bias factor @xcite .
we adopt @xmath39 as the first of our dynamically sensitive observables .
the task of inferring @xmath39 from clustering measurements in the context of hod models is discussed in detail by @xcite .
the next dynamical measurement we consider is the pairwise radial velocity dispersion of galaxies , defined as @xmath154 , where @xmath155 and @xmath156 ( @xmath157 ) are velocities and positions of the two galaxies , and the average is over all galaxy pairs with separations near @xmath158 .
the pairwise dispersion can be inferred by modeling the redshift - space correlation function ( e.g. , @xcite ) .
similar to the galaxy correlation function , on small scales , @xmath159 is dominated by a one - halo term , which can be computed through @xcite @xmath160 where @xmath161 should be understood as the pair - number weighted average dispersion of the relative radial velocities between pairs of galaxies in the same halo . if we assume that the velocity distribution of satellite galaxies is isotropic and isothermal and that the central galaxy is at rest with respect to the center - of - mass of the parent halo , then the pair - weighted average dispersion can be separated into central - satellite and satellite - satellite terms as in equation ( [ eqn : weightedpair ] ) , @xmath162 here @xmath163 is the 1d velocity dispersion of dark matter within the halo , @xmath95 is the velocity bias factor , and other quantities have the same meaning as in equation ( [ eqn : weightedpair ] ) .
note that there is a factor of 2 for satellite - satellite velocity dispersions , since both members of the pair are moving with respect to the halo center - of - mass . on large scales
, the two galaxies of each pair come from different halos , so the relative motions of halos also contribute to the galaxy pairwise dispersion .
we ignore the halo velocity correlation ( @xmath164 in eq .
[ [ eqn : halovdisp2 ] ] ) , which is an acceptable simplification ( @xcite ) . following the argument of @xcite but taking into account the separation of central and satellite galaxies , the galaxy pairwise radial velocity dispersion on large scales can be expressed as @xmath165 } \intdna \nonumber \\ & & \times \intdnb [ 1+\xi_{\rm hh}(r;m_1,m_2 ) ] \nonumber \\ & & \times [ \langle n(m_1)\rangle \langle n(m_2)\rangle \sigma_{h,{\rm 1d}}^2(r;m_1,m_2 ) \nonumber \\ & & \mbox { } + \langle n_{\rm sat}(m_1)\rangle \langle n(m_2)\rangle \alpha_v^2\sigma_m^2(m_1 ) \nonumber \\ & & \mbox { } + \langle n(m_1)\rangle \langle n_{\rm sat}(m_2)\rangle \alpha_v^2\sigma_m^2(m_2 ) ] , \end{aligned}\ ] ] where @xmath166 is the two - point correlation function between halos of mass @xmath89 and @xmath90 at separation @xmath91 .
the three terms inside the last set of square brackets represent contributions by the pairwise velocity dispersion of @xmath89 and @xmath90 halos , by the dispersion of satellite galaxies in @xmath89 halos , and by the dispersion of satellite galaxies in @xmath90 halos .
the halo - halo velocity dispersion is imprinted on all @xmath167 galaxy pairs .
however , the velocity dispersion of satellite galaxies in , say , @xmath89 halos only contributes to the dispersion of the @xmath168 pairs involving those satellites .
the halo - halo velocity dispersion can be computed through equation ( [ eqn : halovdisp2 ] ) , and on large scales we have @xmath169 and @xmath170 . with these substitutions ,
equation ( [ eqn : galvdispl ] ) reduces to @xmath171 \nonumber \\ & & + \frac{2\alpha_v^2}{1+\xigg(r ) } [ \langle \sigma_m^2\rangle_{n_s } + \xigg(r)\langle \sigma_m^2\rangle_{n_sb } ] , \end{aligned}\ ] ] where the number- and bias - weighted averages @xmath172 and @xmath173 are defined as @xmath174 @xmath175 and @xmath176 and @xmath177 are defined by replacing @xmath97 with @xmath123 in equations ( [ eqn : sig2n ] ) and ( [ eqn : sig2nb ] ) , respectively .
since we do not have a good approximation for @xmath159 at intermediate scales , we only consider pairwise dispersion measurements at small scales , where equation ( [ eqn : galvdisps ] ) applies , or on large scales , where equation ( [ eqn : galvdisps2h ] ) applies .
ultimately , the constraints obtainable from anisotropies in redshift space on small , intermediate , and large scales can be inferred using the methods of @xcite and @xcite .
we also consider dynamical measurements of the average virial masses of galaxy groups of fixed multiplicity , @xmath178 .
these can be inferred from galaxy positions and velocities using the kind of estimators described by @xcite .
these estimators are affected by velocity bias , with the estimated mass scaling as @xmath179 . for a given halo mass function and hod
, we therefore calculate this quantity as @xmath180 as in our multiplicity function calculation , we approximate @xmath99 by a nearest - integer distribution . with equation ( [ eqn : mvir ] ) , we implicitly assume that the mass estimators are applied only to satellite galaxies , or else that the reduced dispersion due to the central galaxy is properly taken into account . with large enough galaxy numbers
, more sophisticated mass estimators can circumvent the effects of velocity bias ( e.g. , @xcite ) .
in addition , average group masses can be estimated using gravitational lensing or x - ray properties instead of galaxy velocities .
we do not consider these kinds of @xmath95-free mass estimates in our analysis , but it is clear that they would add power to our overall set of constraints if they could be implemented robustly .
the galaxy - mass cross - correlation can be probed by galaxy - galaxy lensing ( e.g. , @xcite ) . in practice , what is usually inferred from galaxy - galaxy lensing at low redshift is the product of the mean matter density @xmath1 and the galaxy - mass two - point cross - correlation function @xmath181 , so we adopt values of @xmath182 as the observables in our analysis .
the galaxy - mass two - point cross - correlation function can also be decomposed into one - halo and two - halo terms . for simplicity ,
we only consider small - scale cross - correlation , where the one - halo term dominates .
similar to the one - halo term of the galaxy two - point correlation function , the one - halo term @xmath183 of the galaxy - mass two - point cross - correlation function is computed in real space by counting galaxy and mass particle pairs , @xmath184 where @xmath185 is the critical density .
the function @xmath186 is the cumulative radial distribution of galaxy and mass particle pairs ( not galaxy - galaxy pairs as in eq .
[ [ eqn : xigg1h ] ] ) , which is determined by the distribution profiles of galaxies and dark matter within the halo .
when the effect of the central galaxy is taken into account , we have ( similar to eq . [ [ eqn : weightedpair ] ] ) @xmath187 where @xmath188 ( @xmath189 ) is the cumulative distribution of pairs made of the central galaxy ( satellite galaxies ) and mass particles .
the function @xmath190 is simply the radial distribution of matter , which can be directly obtained from the nfw profile .
the function for the satellite - mass pairs , @xmath191 , can be derived by convolving satellite and dark matter profiles . under our assumption that the satellite distribution inside a halo also follows an nfw profile ,
@xmath191 is the convolution of two nfw profiles with different concentration parameters , which has an analytic expression given in appendix [ sec : appendixa ] .
more general discussions of galaxy - galaxy lensing in the halo occupation framework are given by @xcite , who focus on halo virial masses and satellite galaxy fractions , and @xcite , who focus on the derivable constraints on @xmath1 and @xmath2 .
at the final step of our analysis , we consider the large - scale behavior of the galaxy bispectrum , the fourier transform of the three - point correlation function .
although this is a measure of real - space clustering , like those in
[ sec : spatial ] , we treat it separately because its information content is rather different and because it is more difficult to measure observationally .
the specific observable that we consider is the reduced bispectrum @xmath192 where @xmath193 is the bispectrum , @xmath19 is the power spectrum , and the three wavevectors form a triangle ( @xmath194 ) .
in the weakly nonlinear regime , the local relation between the galaxy density contrast @xmath28 and the mass density contrast @xmath195 can be written as a taylor expansion @xmath196 .
the reduced galaxy bispectrum @xmath197 is related to the reduced matter bispectrum @xmath198 through @xmath199 which shows contributions from both gravitational clustering and galaxy bias ( @xcite ; see also @xcite ) . on large scales ,
@xmath200 and @xmath201 are simply the galaxy number - weighted halo bias factors @xmath202 and @xmath203 , respectively , where @xmath202 and @xmath203 are predicted by nonlinear perturbation theory @xcite .
the reduced matter bispectrum is calculated using the linear matter power spectrum @xmath19 and the second - order perturbative bispectrum @xmath204 ( @xcite ) , @xmath205 where @xmath206 , with @xmath207 .
the weak dependence on @xmath1 ( @xcite ) is neglected here .
the key diagnostic power of @xmath197 on large scales lies in the distinctive triangle - shape dependence of @xmath198 predicted by gravitational perturbation theory
. positive linear bias ( @xmath208 suppresses @xmath197 , and non - linear bias ( @xmath209 ) can restore the amplitude but not the shape dependence @xcite .
motivated by cmb anisotropy measurements ( e.g. , @xcite ) , the abundance of galaxy clusters ( e.g. , @xcite ) , high - redshift supernova observations ( e.g. , @xcite ) , and measurements of the 3d galaxy power spectrum and the ly@xmath11 forest power spectrum ( e.g. , @xcite ) , we assume a central cosmology with @xmath210 , @xmath211 , @xmath212 , and @xmath213 .
we model a galaxy population of mean space density @xmath214 , similar to that of galaxies with @xmath91-band absolute magnitude brighter than @xmath215 in the sdss .
we therefore use zehavi et al.s ( @xcite ) results for @xmath216 galaxies as a guide for our central model s hod parameters .
we assume a satellite occupation @xmath217 in equation ( [ eqn : navg ] ) , i.e. , the number of satellites is proportional to the halo mass well above the cutoff @xmath102 .
we set the central galaxy cutoff parameter @xmath109 to 0.2 , consistent with predictions from galaxy formation models ( @xcite ) . with these choices ,
the cutoff mass required to match @xmath214 is @xmath218 .
the probability distribution of satellite galaxies is taken to be a poisson distribution , i.e. , @xmath114 in equation ( [ eqn : nnm1sat ] ) .
we assume that satellite galaxies have the same radial profile and velocity dispersion as dark matter within halos , i.e. , @xmath219 and @xmath220 . for the central cosmology and the central hod
, we use the analytic formulae presented in [ sec : observables ] to calculate a set of observables , which we thereafter treat as observational measurements .
these observables are the rms galaxy overdensity @xmath221 on scales of @xmath222 and @xmath223 ; the group multiplicity function @xmath224 for @xmath225 10 , 20 , and 40 ; the galaxy two - point correlation function @xmath126 at @xmath226 0.3 , 0.5 , 1 , 2 , 5 , and 10 @xmath227 ; the @xmath39 parameter ; the galaxy pairwise velocity dispersion @xmath228 at @xmath229 0.5 , 3 , 5 , and 10 @xmath227 ; the average virial mass of galaxy groups @xmath178 for @xmath230 20 , 40 , and 80 ; the galaxy - mass cross - correlation function @xmath231 at @xmath232 and 0.5 @xmath227 ; and values of the reduced galaxy bispectrum on large scales , for wavevector triangles having two sides fixed at @xmath233 and subtending angles @xmath234 , @xmath235 , @xmath236 , @xmath237 , and @xmath238 .
we assume that each of the 30 `` measurements '' has a @xmath5 fractional uncertainty of 10% , and that the measurement errors are uncorrelated . on its own ,
the 10% error assumption is probably pessimistic ; even the largest scale observables can probably be measured at least this well with the full sdss , and some of the smaller scale quantities have already been measured more precisely ( we investigate the effect of reducing errors for selected observables in [ sec : general ] ) . however , the assumption that the measurement errors are uncorrelated is optimistic , even though we space the observables for any given statistic ( e.g. , multiple scales of the correlation function ) fairly widely . unless there are strong positive correlations among many observables , the constraints on the hod and cosmology should not degrade much with respect to our forecasts , but a full application to observations must include the covariances among observables , calculated from the data or from mock catalogs .
each time we change cosmological parameters , and thus the statistics of the halo population , we search for hod parameters that yield the best possible match to the observables of the central model .
specifically , we minimize @xmath239 where @xmath240 represents the @xmath241 observable with an observational error @xmath242 , the superscript `` @xmath243 '' indicates the predicted value for a given cosmology and hod model , and the superscript `` @xmath244 '' indicates the observed value calculated for the central cosmological model and hod parameters . to isolate the information content of different clustering statistics , we start by considering only the spatial clustering observables described in [ sec : spatial ] , then add new observables one at a time .
we use a gauss - newton scheme to carry out the @xmath0 minimization ( see appendix [ sec : appendixb ] for details ) .
we characterize the acceptability of models in terms of @xmath245 , appropriate for `` flat '' priors that treat all parameter values equally . before turning to cosmological parameter constraints
, we first ask how well the hod can be determined if we assume that the cosmological parameters are known perfectly from independent data .
observational constraints on the hod for an assumed cosmology have already been derived from the 2dfgrs , the sdss , and other data sets , but these studies have all used restricted forms of the hod and limited subsets of the galaxy clustering observables , such as the correlation function or the multiplicity function ( e.g. , @xcite ) . here
we adopt our much more flexible hod parameterization and examine the constraints that could be obtained with 10% measurements of all of the observables discussed above . like @xcite and @xcite
, we explore the hod parameter space using a monte carlo markov chain ( mcmc ) technique ( e.g. , @xcite ) .
we adopt the central cosmological model throughout , and we start the chain from the central hod parameters , which by definition yield @xmath246 . at any point of the chain , we generate a new set of hod parameters by taking a random walk in the hod parameter space with the step size for each parameter drawn from a gaussian distribution .
we accept the new hod model with a probability of 1 if @xmath247 and @xmath248 $ ] if @xmath249 , where @xmath250 and @xmath251 are values of @xmath0 for the new model and for the previous model , respectively .
flat priors are adopted for hod parameters @xmath101 , @xmath109 and @xmath112 in logarithmic space and @xmath120 and @xmath95 in linear space .
the constraints on @xmath97 and other hod parameters are illustrated in figure [ fig : mcmc_hodpar ] .
figure [ fig : mcmc_hodpar]@xmath252 plots the envelope of @xmath97 curves determined from hod models that have @xmath253 , thus showing the 1-@xmath254 uncertainty in @xmath97 . with our adopted set of observables , the shape of @xmath97 near the cutoff mass is poorly constrained .
the mean galaxy density imposes a strong constraint on the cutoff scale , but either hard or soft cutoffs can produce the same @xmath125 .
the halo bias factor is nearly independent of mass in the low - mass regime ( see , e.g. ,
@xcite ) , and these halos contain only a single galaxy , so variations in the cutoff profile do not affect either the two - halo or one - halo terms of @xmath126 .
observables that we have not considered , like the void probability function or the tully - fisher relation , might provide stronger constraints on the location and form of the @xmath97 cutoff . at higher halo masses ,
our observables provide much greater leverage , and @xmath97 is tightly constrained over the range @xmath255 to @xmath256 .
the constraints loosen at still higher halo masses , since these halos are too rare to have much impact on the correlation function , and we only consider group virial masses up to @xmath257 and the multiplicity function up to @xmath258 .
figures [ fig : mcmc_hodpar]@xmath30 [ fig : mcmc_hodpar]@xmath259 show the marginalized likelihoods of the parameters @xmath112 , @xmath260 , and @xmath95 , with dotted vertical lines marking the @xmath5 range . under our assumption of 10% observational errors
, the distribution width parameter @xmath112 can be constrained at the level of @xmath261% ( 1-@xmath254 ) . for the difference between galaxy and dark matter concentrations ,
the 1-@xmath254 constraint is @xmath262 dex , although with more small - scale points in @xmath126 this quantity could be pinned down more tightly .
the 1-@xmath254 uncertainty in the velocity bias @xmath95 is only a few percent , largely because the spatial clustering observables tightly constrain the true halo mass at fixed @xmath16 , and the average virial masses then determine the value of @xmath95 .
figure [ fig : mcmc_corr ] shows correlations between some pairs of hod parameters , inferred from the mcmc run .
if galaxies are more centrally concentrated than dark matter inside halos , then the preferred width of @xmath96 narrows to suppress one - halo pairs in halos with small virial radii ( fig .
[ fig : mcmc_corr]@xmath252 ) .
a wider @xmath14 correlates , weakly , with a lower velocity bias ( @xmath263 for @xmath264 ; fig . [ fig : mcmc_corr]@xmath30 ) , since more small - scale pairs involve satellite galaxies , which make larger contributions to the pairwise velocity dispersion .
there is a similar but stronger correlation with galaxy concentration ( fig .
[ fig : mcmc_corr]@xmath57 ) , since positive @xmath120 allows more small - scale pairs to come from massive halos with large velocity dispersions . finally , if the mean occupation at high halo masses increases ( higher @xmath265 in fig .
[ fig : mcmc_corr]@xmath259 ) , then the mean halo mass at a given high multiplicity decreases , so @xmath95 greater than unity is favored to keep the _ apparent _ virial mass @xmath178 near its original value . before considering general constraints in the @xmath266 parameter space in [ sec : general ] , we examine constraints along several axes in this space : a pure change in @xmath1 , a pure change in @xmath2 , `` cluster - normalized '' changes that preserve the value of @xmath267 at fixed @xmath21 and @xmath3 , and linked changes in @xmath1 , @xmath2 , and @xmath3 ( at fixed @xmath21 ) that preserve the amplitude and slope of the halo mass function at high masses . as discussed by @xcite , the halo population responds to these changes in relatively simple ways , so we gain physical insight into the origin of the more general cosmological constraints by studying the behavior along these axes .
these restricted constraints are also relevant to the case where other parameters are determined by independent data .
figure [ fig : mf_xim ] illustrates the variation of the halo mass function @xmath268 ( fig .
[ fig : mf_xim]@xmath252 ) and the non - linear matter correlation function @xmath269 ( fig .
[ fig : mf_xim]@xmath30 ) along these four axes .
open circles show the central model .
dotted lines show the effect of increasing @xmath1 from 0.3 to 0.5 , with other parameters fixed .
dashed lines show the effect of increasing @xmath2 from 0.9 to 1.0 .
dot - dashed lines show a cluster - normalized change , with @xmath1 increased to 0.5 and @xmath2 decreased to 0.7 to keep @xmath270 fixed .
solid lines show the same @xmath271 combination , but with a corresponding change in @xmath3 that keeps the slope of @xmath268 fixed at the cluster mass scale @xmath272 ( see @xcite ) . in all cases
we use the @xcite mass function formula ( our eq [ [ eqn : halomf ] ] ) and the @xcite prescription for computing the matter correlation function . note that peculiar velocities of halos on large scales are proportional to @xmath42 , similar to the scaling of velocity dispersions of cluster mass halos @xcite .
the constraints for a pure @xmath1 change are easiest to understand , so we examine this case first and in the greatest detail .
as shown in figure [ fig : mf_xim ] , a pure @xmath1 change simply shifts the mass function in proportion to @xmath1 . the effect on the matter correlation function
is very small , arising entirely from changes in the concentrations of halos .
once halo masses are scaled in proportion to @xmath1 , the spatial clustering of halos at a given scaled mass is virtually unchanged , although the velocities change in proportion to @xmath273 @xcite .
figure [ fig : chi2_omega ] shows @xmath245 for the best - fit hod model at each value of @xmath1 , as different observables are added in succession .
figures [ fig : navg1_omega ] and [ fig : hodpar_omega ] show the parameters of these best - fit hods .
if we include only the spatial clustering observables @xmath221 , @xmath224 , and @xmath126 , we obtain results virtually identical to those of the central model by simply shifting the mass scale of the mean occupation function in proportion to @xmath1 . for these observables
alone , we get @xmath274 at all @xmath1 ( _ horizontal dotted line in _ fig . [
fig : chi2_omega ] ) . figure [ fig : navg1_omega]@xmath275 plots the best - fit @xmath97 as a function of the scaled halo mass @xmath276 . because the constraints imposed by our observables on the low end of @xmath97 are weak , the cutoff profiles of the scaled mass functions vary , but above @xmath255 they are nearly indistinguishable .
thus , the spatial clustering observables effectively fix @xmath97 in terms of scaled halo masses , and only small further adjustments can be accommodated when other observables are introduced .
as shown in figure [ fig : hodpar_omega ] , the best - fit models have @xmath114 just like the central model hod , and they have @xmath277 , although there is a slight reduction in halo concentration at low @xmath1 to compensate for the higher concentrations of the halos themselves .
the @xmath1 degeneracy breaks as soon as we include dynamical measures that are sensitive to the absolute mass or velocity scale of halos .
the first such measure that we add is @xmath153 . for low @xmath1 , the cutoff of @xmath97 smooths out so that more massive galaxies come from lower mass halos with lower bias factors ( _ dashed lines in _ figs .
[ fig : navg1_omega]@xmath252 and [ fig : navg1_omega]@xmath30 ) .
the @xmath96 distribution broadens slightly ( @xmath278 in fig .
[ fig : hodpar_omega]@xmath252 ) to compensate for a drop in one - halo pairs , and concentrations drop ( @xmath279 in fig .
[ fig : hodpar_omega]@xmath30 ) so that these pairs move to larger separations .
however , within the constraints imposed by the multiplicity function , the correlation function , and the galaxy number density , there is virtually no room to change the galaxy bias factor in a way that compensates changes in @xmath1 . for @xmath1 in the range @xmath280 , the @xmath245 values in figure [ fig : chi2_omega ]
are almost equal to those expected for constant @xmath147 , given our adopted 10% uncertainty in @xmath39 ( and hence @xmath273 ) .
when we add galaxy pairwise velocity dispersions as observables , the velocity bias parameter @xmath95 adjusts to compensate for changes in @xmath1 ( fig .
[ fig : hodpar_omega]@xmath57 , _ long - dashed line _ ) .
however , while velocity bias can reduce velocity dispersions within halos , it can not affect the pairwise velocities of the halos themselves , which scale as @xmath273 .
for high @xmath1 , @xmath95 drops by more than the factor @xmath281 that would keep _ internal _ dispersions fixed , in an effort to compensate for the higher halo velocities .
this change would drive the pairwise dispersion too low on small scales , where the one - halo term dominates , so halo concentrations increase , allowing more of the small separation pairs to come from massive halos . in response to these higher concentrations , the width parameter @xmath112 drops , reducing the one - halo contribution to @xmath126 . despite the freedom introduced by velocity bias
, the combination of pairwise dispersions at small and large scales adds significant discriminatory power on the value of @xmath1 . for @xmath282 , for example , @xmath245 increases from @xmath283 ( @xmath39 alone ) to @xmath284 , with the two dynamical measurements producing most of the discrepancy .
when we add the @xmath178 observables , the velocity bias is forced close to the value @xmath285 that keeps these apparent virial masses fixed ( fig .
[ fig : hodpar_omega]@xmath57 , _ dot - dashed line , nearly obscured by solid line _ ) . since there is no longer room to improve the large - scale pairwise dispersion by `` overcompensating '' with @xmath95 , the concentration and width parameters return to the values preferred by the spatial clustering observables ( note the similarity of dotted and dot - dashed lines in figs .
[ fig : hodpar_omega]@xmath252 and [ fig : hodpar_omega]@xmath30 , and the difference between these lines and the long - dashed lines ) .
although @xmath95 can exactly compensate a pure @xmath1 shift for the @xmath178 observables , the match to the pairwise dispersion becomes worse as a result , and @xmath245 increases slightly .
the galaxy - galaxy lensing observable @xmath231 provides a diagnostic of host halo masses that is unaffected by velocity bias .
if @xmath97 stays fixed as a function of scaled mass , then a pure @xmath1 change simply scales @xmath231 in proportion to @xmath1 ( eq . [ eqn : xigm1h ] ) .
the effect of higher @xmath1 can be partly compensated by reducing halo concentrations ( fig .
[ fig : hodpar_omega]@xmath30 , _ solid line _ ) , which dilutes the contribution from satellite galaxy and matter particle pairs .
this decrease drives an increase in @xmath112 ( fig .
[ fig : hodpar_omega]@xmath252 ) to restore lost one - halo galaxy pairs at small separations ( note that @xmath286 itself is independent of @xmath112 ) . however , altering concentrations has limited power to compensate for a shift in the halo mass function , and @xmath231 measurements add substantially to @xmath245 .
measurements in the linear or near - linear regime , not considered here , would add still more discriminatory power . finally , large - scale bispectrum measurements have little sensitivity to a pure @xmath1 shift ( fig . [
fig : chi2_omega ] , _ long - dashed line _ ) .
this statistic is mainly a diagnostic for the galaxy bias factor @xmath147 . along this model sequence ,
the amplitude of dark matter clustering does not change , and the galaxy bias factor is therefore determined by the observed spatial clustering , independent of @xmath1 .
the bispectrum will become important when we consider sequences with changing @xmath2 .
overall , these galaxy clustering observables have great power to constrain any pure shift in @xmath1 .
a model with @xmath287 has a likelihood @xmath288 relative to the central model , so even a model with @xmath289 is disfavored by a factor of 10 .
including all observables , we find @xmath290 and 5.0 for @xmath291 and 0.35 , respectively , implying that @xmath292 can be firmly ruled out .
these constraints weaken significantly if we omit the weak - lensing observables with their direct sensitivity to mass scales .
still , even with @xmath39 and @xmath228 as the only dynamical observables , the @xmath293 and 0.4 models are rejected with @xmath294 .
we now consider models that differ only in @xmath2 , the amplitude of the linear matter power spectrum , with @xmath1 and the shape of @xmath19 held fixed .
the dashed lines in figure [ fig : mf_xim ] show the effect of raising @xmath2 from 0.9 to 1.0 .
the amplitude of the matter correlation function increases at all scales , and the space density of high - mass halos ( @xmath295 ) rises , while the space density of galaxy mass halos ( @xmath296 ) is nearly unchanged .
figure [ fig : chi2_sig8 ] shows @xmath245 for the best - fit hod models along the changing @xmath2 sequence , and figure [ fig : hodpar_sig8 ] shows the properties of these best - fit models .
figure [ fig : hodpar_sig8]@xmath30 shows the response of @xmath97 to @xmath2 when the spatial clustering observables @xmath297 , @xmath298 and @xmath299 are the only constraints . higher @xmath2 leads to lower @xmath97 at high masses , so that @xmath299 remains fixed despite an increased halo abundance .
this change reduces the galaxy bias factor @xmath147 , but on its own it is not enough to compensate for the increased matter clustering amplitude .
the @xmath97 cutoff therefore spreads to lower masses , putting more isolated galaxies in low bias halos while retaining the galaxy number density @xmath300 .
the @xmath96 width parameter rises to @xmath301 to restore small - scale pairs lost when the number of halos in the @xmath302 range is reduced ( fig .
[ fig : hodpar_sig8]@xmath57 , _ dotted line _ ) . for spatial clustering observables alone ,
these hod changes effectively compensate for changes in the halo population , and the @xmath245 curve is nearly flat in the vicinity of the central model ( fig .
[ fig : chi2_sig8 ] , _ dotted line _ ) . however , these observables pin down @xmath97 fairly precisely , leaving little room for adjustment when dynamically sensitive observables are added as constraints .
the @xmath245 from @xmath303 is close to that expected for our 10% error bar and the scaling @xmath304 implied by fixed galaxy spatial clustering .
the velocity bias parameter @xmath95 can partly compensate for changes in halo velocities and internal dispersions , but it can not fix the small- and large - scale values of @xmath228 simultaneously ( see discussion in [ sec : omegam ] ) , so @xmath305 rises when @xmath306 is added . adding @xmath178 effectively pins down @xmath95 to the value needed to keep the apparent virial masses fixed , making the overall fit of @xmath228 worse and further increasing @xmath305 .
the @xmath231 observable is unaffected by velocity bias , and the small adjustments still allowed in @xmath97 , @xmath120 , and @xmath112 do little to counter the changes in halo masses . finally , the galaxy bispectrum increases @xmath305 through its direct sensitivity to @xmath147 . with all observables considered together , @xmath305 exceeds 4.0 for @xmath307 , and it rises rapidly for larger changes .
the observed abundance of massive galaxy clusters effectively constrains the amplitude of the halo mass function at a scale @xmath308 .
the halo abundance at this scale is an increasing function of @xmath1 and @xmath2 , and the cluster normalization constraint is usually expressed in the form @xmath309 , with @xmath310 ( e.g. , @xcite ) . here , as in @xcite , we consider a sequence of cluster - normalized cosmological models that satisfy @xmath311 , with the shape of @xmath19 held fixed .
the dot - dashed lines in figure [ fig : mf_xim ] show the effect of simultaneously increasing @xmath1 to 0.5 and decreasing @xmath2 to 0.70 .
as expected , the halo mass function stays the same at @xmath308 , but relative to the fiducial model there are more low - mass halos and fewer halos with @xmath312 .
the matter correlation function is lower at all scales because of the lower value of @xmath2 .
figures [ fig : chi2_clnorm ] and [ fig : hodpar_clnorm ] show the @xmath305 and parameters of the best - fit hod models along the cluster - normalized sequence , in the same format as figures [ fig : chi2_sig8 ] and [ fig : hodpar_sig8 ] .
the multiplicity function fixes the high - mass end of @xmath97 in a way that is nearly independent of @xmath1 , although there are slight variations in slope that reflect the changing slopes of the halo mass function ( fig .
[ fig : hodpar_clnorm]@xmath30 ) .
the higher abundance of low - mass halos for high @xmath1 requires an increase in the cutoff mass to keep the galaxy number density constant .
the tightened mass range for single - galaxy halos leads to a larger one - halo contribution to @xmath126 , which is compensated by decreasing the width parameter @xmath112 and the concentration parameter @xmath120 .
although the matter correlation amplitude is reduced for high @xmath1 and low @xmath2 , the bias of halos at fixed mass is higher , and these two effects conspire to keep the large - scale amplitude of the galaxy correlation function nearly constant over a substantial range of @xmath1 . for spatial clustering observables alone
, @xmath305 is nearly flat over the range @xmath313 .
because cluster normalization keeps the mass and velocity scale of halos roughly fixed , the addition of dynamically sensitive observables makes relatively little difference to best - fit hod parameters , and @xmath305 increases more slowly than it does for pure @xmath1 or pure @xmath2 changes ( note the smaller vertical scale of figure [ fig : chi2_clnorm ] relative to figures [ fig : chi2_omega ] and [ fig : chi2_sig8 ] ) .
the detailed changes in @xmath112 , @xmath120 , and @xmath95 as new observables are added mostly reflect the changes in the cutoff of @xmath97 and corresponding changes in the fraction of single - occupancy halos .
the @xmath305 curves are asymmetric in @xmath1 because of the non - linear relation between @xmath1 and @xmath2 . for example
, cluster normalization requires a 13% reduction in @xmath2 to 0.78 at @xmath282 but a 22% increase to @xmath314 at @xmath293 , and the still larger @xmath2 values required at lower @xmath1 are easily ruled out .
the bispectrum constraint is especially useful in distinguishing cluster - normalized models because it responds to @xmath147 independent of @xmath1 . despite the compensating effects of @xmath2 and @xmath1 in cluster - normalized models , the full set of observables yields @xmath315 for @xmath282 and @xmath316 for @xmath317 .
as shown by @xcite , one can construct cluster - normalized model sequences that match the amplitude _ and slope _ of the halo mass function at cluster scales by changing the power spectrum shape parameter in concert with @xmath1 and @xmath2 .
the solid lines in figure [ fig : mf_xim ] show a model with @xmath318 , @xmath319 , and @xmath320 .
the halo mass function is nearly identical to the central model s for @xmath321 , although it is higher at low halo masses .
the _ shape _ of the matter correlation function is significantly different because of the different shape of @xmath19 .
figure [ fig : chi2_mfmatch ] shows @xmath305 as a function of @xmath1 along this model sequence .
the required values of @xmath2 and @xmath3 are shown in figure 13 of @xcite , and the values of @xmath3 are also marked on the top axis of our figure [ fig : chi2_mfmatch ] . as speculated by @xcite , matching the slope of the halo mass function by changing the shape of @xmath19 does more harm than good : @xmath305 values at a given @xmath1 are always higher than they are for the fixed @xmath19 , cluster - normalized models shown in figure [ fig : chi2_clnorm ] . with all observables included , models with @xmath293 and @xmath282 have @xmath322 and @xmath323 , respectively .
the four model sequences considered in figures [ fig : chi2_omega ] , [ fig : chi2_sig8 ] , [ fig : chi2_clnorm ] , and [ fig : chi2_mfmatch ] trace four different curves through the @xmath324 space of cosmological parameters .
we now examine the constraints and degeneracies in this parameter space in more general terms .
the computational expense of finding the best - fit , 10-parameter hod model for each cosmological model makes a comprehensive study of the full 3d space difficult .
however , the fact that changing the power spectrum shape to match the halo mass function makes models easier to distinguish suggests that the `` @xmath3 '' dimension is largely decoupled from the @xmath325 plane .
furthermore , the galaxy power spectrum shape can be measured directly on large scales ( e.g. , @xcite ) , and we have not incorporated these direct constraints into our observables .
to begin , therefore , we fix @xmath213 and map out @xmath245 contours in the @xmath325 plane , finding the best - fit hod parameters that minimize the values of @xmath245 at each point on the cosmological parameter grid .
we include all the observables discussed in
[ sec : observables ] and incorporated in our earlier constraints .
the solid contours in figure [ fig : contour]@xmath252 show @xmath326 and @xmath327 , corresponding to 68.3% and 95.4% confidence intervals for two parameters .
the most degenerate direction seems to be @xmath309 with @xmath328 , which deviates from the degeneracy axis of cluster normalization ( @xmath329 ) or large - scale redshift - space distortions ( @xmath330 ) .
this deviation reflects the steeper scaling favored by the galaxy - galaxy lensing constraint , which has @xmath331 even on these non - linear scales ( @xcite ) , and by pairwise velocity dispersions with @xmath95 as an adjustable parameter .
dotted contours in this panel show @xmath332 and 4 , from which one can read projected @xmath5 and @xmath333 constraints on the individual parameters @xmath2 and @xmath1 .
the remaining panels show the 1d likelihoods for these parameters and the combination @xmath334 , inferred by integrating the likelihood function @xmath288 over each bin of the relevant quantity . under our assumption of 10% fractional error in each observable ,
the 1-@xmath254 uncertainty in the matter density parameter @xmath1 is about 11% .
the amplitude @xmath2 of the matter power spectrum has a 1-@xmath254 constraint that is better than 10% .
the 1-@xmath254 uncertainty in the best constrained quantity , @xmath334 with @xmath335 , is at the 5% level .
if we force @xmath336 to 0.5 , the precision is almost equally good .
what constraints could we infer on @xmath3 from the galaxy clustering measurements considered in our observable set ? to address this question in approximate terms , we fix @xmath1 and @xmath2 to the central model values of 0.3 and 0.9 and vary @xmath3 alone , again finding best - fit hod parameters and minimum @xmath305 .
we obtain @xmath332 at @xmath337 , 0.219 and @xmath338 at @xmath339 , 0.240 . the 10% ( @xmath5 ) constraint on @xmath3
is similar to that found along our matched halo mass function sequence ( @xmath332 at @xmath340 , 0.223 and @xmath338 at @xmath341 , 0.250 ; figure [ fig : chi2_mfmatch ] ) , supporting our conjecture that the ( @xmath342 ) constraints are mostly orthogonal to the @xmath19 shape constraints . with our 30 observables and assumption of 10% , independent fractional errors , therefore , the full @xmath0 surface approximately follows the contours of figure [ fig : contour ] with an orthogonal gaussian of roughly 10% @xmath5 width in the @xmath3 direction . as discussed in the introduction , traditional approaches to testing cosmological models with large - scale structure focus on the perturbative regime , where it is reasonable to adopt linear theory and linear bias for describing the galaxy power spectrum and redshift - space distortions and second - order perturbation theory and a quadratic bias model for describing the galaxy bispectrum .
it is interesting to compare our hod - based constraints to those that could be derived from the same observables with the perturbative approach .
specifically , we use the @xmath297 , @xmath39 , and @xmath197 observables adopted previously , with the same 10% uncertainties .
we employ a two - parameter quadratic bias model , and we calculate the observables for a given cosmology and bias using @xmath343 , @xmath344 , and @xmath345 , where @xmath254 and @xmath198 are the rms fluctuations and reduced bispectrum for the dark matter .
we examine an ( @xmath342 ) grid with @xmath213 and find the values of @xmath30 and @xmath346 that minimize @xmath305 for each cosmological model .
the top left panel of figure [ fig : chgerr ] compares our hod - based constraints ( repeated from figure [ fig : contour ] ) to those obtained from the perturbative analysis .
the principal degeneracy axis for the latter constraints is @xmath347 , since @xmath39 constrains @xmath348 and @xmath297 constrains @xmath349 .
the bispectrum constraint on @xmath30 breaks this degeneracy and produces closed contours .
however , the constraints from the full hod - based analysis of all observables are much tighter than the constraints from the smaller set of observables that can be modeled by the perturbative approach alone .
of course , the perturbative cosmology constraints can be improved by including more observables , such as large - scale weak - lensing measurements or cmb anisotropies , but these observables will also improve the hod - based constraints .
thus , despite the need for a more complex , multi - parameter model of galaxy bias , the ability of the hod approach to model intermediate and small - scale clustering leads to substantial improvements in cosmological power . at the same time ,
an hod analysis of observed galaxy clustering yields valuable tests of theoretical models of galaxy formation .
the obtainable constraints on cosmological parameters depend , of course , on the precision of the clustering measurements .
our assumption of 10% fractional uncertainty on each observable is probably conservative , and current galaxy clustering data have already yielded higher precision for some of these observables , such as the two - point correlation function . to study the impact of improving measurement precision
, we change the assumed error bars of different sets of observables by turns and recalculate @xmath305 contours .
this investigation also highlights the relative cosmological sensitivity of the different observables .
each panel of figure [ fig : chgerr ] ( except for the top left panel discussed previously ) shows changes in the ( @xmath342 ) constraints caused by improving the measurement precision of one set of observables .
the two dotted contours are the 1-@xmath254 and 2-@xmath254 confidence levels ( @xmath350 6.17 ) obtained by assuming 10% fractional errors for all sets of observables ; they match the solid contours in figure [ fig : contour]@xmath252 .
the solid contours are obtained by assuming 5% errors for the set of observables indicated in the upper right corner of the panel , keeping 10% for other sets . with the galaxy bias factor constrained by other observables ,
a better measurement of galaxy density contrasts @xmath297 leads to a better constraint on @xmath2 , shrinking contours at both ends .
the tighter @xmath2 constraint in turn produces a slightly tighter @xmath1 constraint , but there is essentially no improvement in the @xmath334 constraint ( i.e. , the width of the contours in the narrow direction is unchanged ) . reducing the observational error bar on @xmath197 produces a similar improvement for a different reason
the more precise measurements of @xmath197 allow a better constraint on the galaxy bias factor @xmath147 , so they help to break the degeneracy between @xmath147 and the amplitude of the dark matter power spectrum .
improving the pairwise velocity dispersion measurements produces modest improvements in both the @xmath334 constraint and the individual constraints on @xmath2 and @xmath1 .
the former improvement comes mainly from the large - scale velocity dispersion , which depends on halo velocities and thus scales roughly with @xmath42 .
the individual parameter improvements come mainly from the interplay of small - scale @xmath306 with the apparent virial mass @xmath178 .
better @xmath306 measurements yield tighter constraints on the velocity bias parameter , and the small - scale @xmath306 and @xmath178 measurements then constrain the halo mass scale . improving the @xmath178 measurements on their own yields better @xmath2 and @xmath1 constraints but does not change the @xmath334 constraint . improving the @xmath126 measurements
leads to a significantly better determination of @xmath2 for two reasons : it tightens the constraints on @xmath14 and thus on the large - scale galaxy bias factor , and it improves the measurement of the large - scale galaxy clustering amplitude itself .
these two effects are analogous to those discussed earlier for @xmath197 and @xmath297 , respectively . although @xmath126 is not directly sensitive to @xmath1 , the tighter @xmath2 constraint produces a tighter @xmath1 constraint when combined with dynamical measurements . improving @xmath224 measurements produces surprisingly little improvement in the cosmological parameter constraints .
if we instead degrade the fractional errors on @xmath224 to 50% , the constraints ( shown by light solid contours ) get noticeably but not dramatically worse .
while it is always possible to reproduce @xmath224 exactly by choosing the appropriate @xmath97 , this choice pins down the mass associated with a given multiplicity , and it largely pins down the galaxy bias factor .
it appears , however , that the @xmath126 measurements already determine the high end of @xmath97 well enough for these purposes , so that moderate improvements in @xmath224 do not add much power . if we degrade the presumed measurement errors on @xmath126 to 25% , then the impact of improving the @xmath224 measurements is larger , but it is still not dramatic .
the apparent redundancy of the @xmath224 and @xmath126 constraints makes them a valuable consistency check on the underlying assumptions of the hod modeling .
for example , the bias factor of single - galaxy halos affects @xmath126 but not @xmath224 , so an hod model that makes the wrong assumption about this bias factor will lead to conflicting estimates of @xmath97 at high masses .
we have also examined the consequences of adopting priors on the hod parameters , which could be motivated by theoretical or observational considerations .
for example , numerical simulations give some idea of the plausible range of values for the velocity bias factor , and tully - fisher measurements can put limits on hod parameters describing the cutoff profile of the mean occupation function .
we characterize the priors by a 1-@xmath254 deviation allowed from a central value and minimize the sum of the @xmath0 function and a quadratic term for the prior ( see appendix [ sec : appendixb ] ) . with a 1-@xmath254 range of @xmath351 for the velocity bias factor @xmath95 or @xmath352 for the distribution width parameter @xmath112
, we find that there is almost no improvement in cosmological constraints .
the reason is that , even without these priors , the measurements have already excluded cosmological models that need extreme values of @xmath95 and @xmath112 . for example , models that need @xmath95 as high as 1.3 have @xmath353 and @xmath354 ( see figures [ fig : hodpar_omega ] and [ fig : hodpar_sig8 ] ) .
adopting a prior for the halo concentration parameter @xmath260 with 1-@xmath254 range of @xmath355 only slightly improves the cosmological constraints by excluding some small regions of extremely high or low values of @xmath2 .
priors on these hod parameters must therefore be quite strong before they would improve the constraints obtainable from the observable set considered here .
as we have shown in figure [ fig : mcmc_hodpar ] , for a fixed cosmology , the cutoff profile of the hod can not be well - constrained with the observables we adopt in this paper .
this result suggests that priors in the _ shape _ of the cutoff profile would not improve constraints on cosmological parameters . to verify this expectation
, we perform a test in which we fix the cutoff profile width @xmath109 to the central model value .
the resulting cosmological constraints are virtually identical to those in figure [ fig : contour ] , where @xmath109 is not fixed .
however , if the value of the absolute cutoff mass scale can also be pinned down , e.g. from the tully - fisher relation or other dynamical measurements in the single - galaxy regime , then cosmological constraints improve substantially .
we have investigated the idealized case in which @xmath102 and @xmath109 are fixed to the values of the central model .
we maintain the galaxy number density by adjusting the amplitude of the satellite galaxy occupation function .
the strong constraint on the halo mass scale from the value of @xmath102 produces a tight constraint on @xmath1 ( @xmath356 ) , which in turn leads to a much tighter constraint on @xmath2 ( @xmath357 ) . finally , we consider the constraints achievable with a brighter , lower density galaxy sample
. different samples of galaxies have different hods , but they should probe the same underlying cosmology .
all the investigations we have presented so far are based on the luminosity - threshold sample with a number density @xmath214 , roughly corresponding to the @xmath358 galaxies in the sdss .
we increase the mass scale of our central model hod to construct a galaxy sample with number density 10 times smaller ( @xmath359 ) , which approximates the sdss @xmath360 sample ( @xcite ) . with the same assumption of 10% fractional error on each of the 30 observables
, we find that the cosmological constraints from these two galaxy samples are quite similar , with the degeneracy direction @xmath309 slightly rotated to higher @xmath336 for the lower density sample .
thus , measurements for these two largely independent galaxy samples should allow an important consistency check on cosmological conclusions and , if the results are combined , a @xmath361 reduction in statistical errors .
perhaps the most important implication of figure [ fig : chgerr ] is that the cosmological constraints do not rely on one or two clustering statistics but instead emerge from the interlocking web of measurements .
figure [ fig : matrix ] demonstrates this point in a different way .
it shows the `` influence matrix , '' which we define in terms of the partial derivatives @xmath362 evaluated at the central hod and cosmological model .
each predicted observable @xmath363 is a function of cosmological and hod parameters , and each @xmath364 represents one of these parameters , so the influence matrix encodes the response of the observables to the individual model parameters . when evaluating the partial derivatives , we adjust @xmath102 so that the galaxy number density stays fixed , but we keep the remaining hod and cosmological parameters ( other than @xmath364 ) fixed .
the influence matrix is closely related to the fisher matrix , @xmath365 where @xmath366 is the vector of the predicted observables and @xmath367 is the covariance matrix of the errors . however , while the fisher matrix depends strongly on the assumed error covariance matrix @xmath367 , the influence matrix is independent of this assumption , and it reveals the dependence of observables on model parameters in a more transparent manner . figure [ fig : matrix ] shows the influence matrix calculated at the central model .
the area of each circle is proportional to the element of the matrix from the corresponding observable ( _ vertical axis _ ) and parameter ( _ horizontal axis _ ) .
the hod parameter @xmath102 , determined by matching the galaxy number density , is presented here as an observable , since its value could in principle be probed by dynamical or weak - lensing measurements in the single - galaxy regime .
if two columns have similar elements , then the corresponding parameters are largely degenerate , since an increase in one can be compensated by a decrease in the other .
the first three columns of figure [ fig : matrix ] show how clustering observables vary with cosmological parameters .
most elements of @xmath368 and @xmath369 are positive .
an increase in @xmath1 shifts the halo mass function to higher mass scales ( fig .
[ fig : mf_xim ] ) , and with the fixed satellite hod , this shift increases the group multiplicity function @xmath299 . to keep a fixed galaxy number density , @xmath102 increases by a larger factor than @xmath1 , and the shift of galaxies to halos with larger
@xmath370 leads to stronger galaxy bias for spatial clustering statistics @xmath297 and @xmath298 .
an increase in @xmath2 boosts the matter clustering directly . raising either parameter increases halo masses and velocities , and thus the values of dynamical observables .
the similar form of these vectors implies significant degeneracy between these two parameters , with an overall trend that roughly follows @xmath371 , but the degeneracy is not complete .
the power spectrum shape parameter @xmath3 is largely decoupled from @xmath1 and @xmath2 .
the much stronger sensitivity of @xmath102 to @xmath1 than to any other cosmological or hod parameter explains why a tight observational constraint on @xmath102 can produce such a tight constraint on @xmath1 ( [ sec : general ] ) .
the sensitivity of spatial clustering statistics to @xmath1 , which contrasts with the flat @xmath245 shown by the dotted line in figure [ fig : chi2_omega ] , arises because we here keep all mass scales in the hod fixed except for @xmath102 .
if we instead scale all masses @xmath372 , then the derivatives of @xmath298 , @xmath299 , and @xmath2 with respect to @xmath1 essentially vanish , as expected , while the derivatives of @xmath178 become one instead of zero .
the rightmost nine columns in figure [ fig : matrix ] show how clustering observables vary with the hod parameters , with cosmological parameters held fixed .
a degeneracy between a cosmological parameter and hod parameters exists to the extent that the `` influence vector '' @xmath373 can be approximated by a linear combination of the influence vectors @xmath374 , where @xmath375 is the cosmological parameter and @xmath376 are the hod parameters . as expected from our earlier results , a substantial degeneracy exists if we restrict @xmath366 to the spatial clustering observables @xmath298 , @xmath299 , and @xmath297 .
for example , an increase in @xmath2 can be largely compensated by decreases in the parameters @xmath377 and @xmath265 that describe the high end of @xmath97 , and changes in @xmath112 and @xmath120 can take up some amount of residual difference .
however , with the full set of observables , the structures of @xmath368 , @xmath369 , and @xmath378 are quite different from any of the influence vectors of individual hod parameters .
the visual structure of figure [ fig : matrix ] suggests that there is no linear combination of the hod influence vectors that accurately approximates the cosmological parameter influence vectors .
we verify this suggestion by solving for the combination coefficients that minimize the squared difference between the @xmath1 influence vector and the corresponding linear combination of the hod influence vectors .
( we do not include @xmath102 as an observable here , since it probably can not be measured as precisely as the other quantities . )
figure [ fig : lincomb]@xmath252 shows the hod influence vectors multiplied by these linear combination coefficients .
the @xmath377 and @xmath265 parameters dominate the best - fit linear combination , with @xmath95 compensating for the increase in @xmath178 that would otherwise occur .
figure [ fig : lincomb]@xmath57 shows that this best - fit linear combination ( _ open squares _ ) does not fully recover the @xmath1 influence vector ( _ filled circles _ ) , with pairwise dispersions and ( especially ) galaxy - mass correlations providing the largest discrepancy . if @xmath2 and @xmath3 are allowed to vary in addition to the hod parameters , then the best - fit linear combination ( _ open circles _ in fig .
[ fig : lincomb]@xmath57 ) is a better approximation to the @xmath1 influence vector , with clear improvement in @xmath306 and @xmath379 . figure [ fig : lincomb]@xmath30 shows that the contribution from the @xmath2 influence vector is much larger than that from any of the hod influence vectors .
nonetheless , the changes in @xmath377 and @xmath265 significantly improve the agreement with the multiplicity function and small - scale correlation function , relative to a pure @xmath2 change with no freedom in the hod .
as argued earlier , the shape parameter @xmath3 is largely decoupled from the strong cosmological degeneracy between @xmath1 and @xmath2 .
figure [ fig : matrix ] also shows that clustering observables are insensitive to some hod parameters , most notably @xmath380 and @xmath381 , which describe the low - mass end of the mean occupation function .
we reached similar conclusions from our analysis of @xmath97 constraints in [ sec : fixed ] and the effect of a @xmath109 prior in [ sec : general ] .
the insensitivity of clustering observables to these parameters means that it is difficult to constrain them with clustering data .
however , this insensitivity also means that poor knowledge of these parameters does not introduce uncertainty in the cosmological conclusions .
we have also investigated the influence matrix for a simpler hod parameterization in which the mean satellite occupation function is a truncated power - law ( @xcite ) .
the influence vector of the power - law normalization is similar to that of @xmath377 in figure [ fig : matrix ] , while the influence vector of the power - law slope is similar in overall form but with a different mass and length - scale dependence , reflecting its differential impact on high- and low - mass halos .
the influence of the low - mass cutoff is , again , weak .
the halo occupation function @xmath96 is , by definition , a distribution weighted by halo number , the probability that a randomly selected halo of mass @xmath15 contains @xmath16 galaxies .
our clustering calculations implicitly assume that this distribution and other parameters of the hod like @xmath120 and @xmath95 are independent of the larger scale environment .
a systematic dependence of a halo s galaxy occupation on the overdensity @xmath195 of its surroundings would alter the clustering predicted for a given halo population and ( number - averaged ) @xmath96 .
ignoring this dependence could therefore lead one to infer an incorrect hod or incorrect cosmological parameters when fitting observed galaxy clustering .
the assumption of an environment - independent hod is rooted partly in the @xcite excursion set derivation of the extended press - schechter formalism .
this derivation predicts that the statistical features of a halo s assembly history depend only on its present mass , a point emphasized by @xcite . in agreement with this prediction
, @xcite and @xcite found that @xmath16-body halos of fixed mass in different environments have similar properties and formation histories , although @xcite found a subtle trend for close pairs of halos to have higher formation redshifts .
the resolution of the simulations used in these studies effectively limited them to halo masses @xmath25 .
two recent studies @xcite , based on the higher resolution , millennium run @xmath16-body simulation @xcite , show that the correlation between halo formation time and large - scale environment becomes much stronger for halo masses @xmath382 . to the extent that galaxy properties depend on halo formation time and halo mass , this correlation will in turn produce an environmental dependence of @xmath96 .
there are several reasons for thinking that the quantitative impact of such a dependence is small , at least for galaxy samples defined by simple thresholds in luminosity or baryonic mass . empirically ,
@xcite show that the correlation of sdss galaxy properties with galaxy overdensity on a scale of @xmath20 can be entirely explained by the correlation with overdensity on the @xmath383 scale characteristic of large halos ; at fixed local overdensity , they detect no residual correlation with the larger scale environment . on the theoretical side , @xcite
show that the hod ( more precisely , the mean occupation function ) of galaxies above a baryonic mass threshold in @xcite s ( @xcite ) cosmological hydrodynamic simulation is independent of environment within the statistical uncertainties imposed by the @xmath384 simulation volume .
@xcite , analyzing the same simulation , show that explicitly eliminating any environmental dependence by randomly shuffling the galaxy populations among halos of similar mass changes the galaxy - galaxy and galaxy - matter correlation functions by @xmath385 on scales @xmath386 , again within the statistical uncertainties . in a similar experiment using semi - analytic galaxy populations in the millennium
run simulation , @xcite find that the correlation functions of luminosity - bin samples change by a few percent ; because of the large simulation volume , these shifts are measured with high precision .
since we hope in the long term to achieve few percent precision on cosmological parameter measurements , the millennium run results imply that environmental dependence of the hod must eventually be considered as part of the program outlined in this paper . allowing arbitrary dependence of @xmath96 on large - scale overdensity @xmath195
would probably make the problem intractable and degenerate , but given the small magnitude of the anticipated effects , a more restricted parameterization should suffice . for example , since the age dependence of halo clustering is found mainly in the mass range of individual galaxy halos , it may well be enough to allow @xmath102 to vary with @xmath195 , or to treat the effective bias parameter of low - mass halos as a parameter to be fitted instead of using the standard @xmath16-body result .
since much of the signal in most clustering statistics comes from multi - galaxy halos , addition of such parameters may not have much impact on the achievable constraints on cosmological parameters .
semi - analytic mock catalogs will be valuable in guiding the choice of parameterizations and in testing whether fitting methods that adopt a particular parameterization ( or that ignore environmental variations altogether ) yield incorrect estimates of hod or cosmological parameters . for cosmological applications , it will be desirable to work with galaxy samples defined by properties that are sensitive to halo mass and relatively insensitive to formation history , such as stellar mass or luminosity in a band dominated by old stellar populations . color - selected or morphologically selected samples will be more susceptible to environmental dependence of halo formation history at fixed mass .
conversely , samples with high luminosity or mass thresholds may be the least susceptible , since the correlation of formation time with environment is weak for high - mass halos .
while the breakdown of the @xcite prediction is an annoyance for the cosmological applications that are the focus of this paper , it could be an asset for efforts to understand the physical processes that determine galaxy morphology and spectral properties .
for example , if fitting the clustering of a selected class of galaxies _ requires _ environmental dependence of the hod , then the form and magnitude of that dependence will provide strong clues to the aspects of halo formation that determine whether galaxies are members of that class .
conversely , if the color and morphology dependence of observed clustering can be entirely explained without environmental variations of the hod , it will imply that aspects of halo formation that correlate with environment do not play a strong role in determining these properties of galaxies .
our analysis shows that hod modeling can substantially increase the cosmological power of galaxy clustering measurements , by breaking degeneracies between the clustering of dark matter and the bias of galaxies with respect to mass .
changing the shape or amplitude of the matter power spectrum or the value of @xmath1 alters the mass function , spatial clustering , and velocity statistics of the dark halo population in well - understood ways @xcite .
our experiments here , which verify the qualitative arguments of @xcite and @xcite , show that changes to the galaxy hod can not mask these changes in the underlying dark halo population . with our highly flexible parameterization of the hod , the set of observables considered here yields @xmath5 uncertainties of @xmath4 in @xmath2 , @xmath1 , and @xmath3 and @xmath7 uncertainty in the combination @xmath387 with @xmath388 .
we expect these forecasts to be conservative , as we have not included observables for which we did not have ready analytic approximations , and our assumption of 10% measurement errors is pessimistic in at least some cases .
the physical origin of these cosmological constraints is straightforward to understand for simple changes in @xmath1 , @xmath2 , or @xmath270 , as discussed in [ sec : omegam]-[sec : clnorm ] .
the general theme of these discussions is that , for a given cosmological model , the spatial clustering of galaxies largely determines the number of galaxies in halos of a given spatial abundance .
dynamically sensitive statistics then reveal the halo mass scale , which depends on @xmath2 and @xmath1 .
we allow an arbitrary bias @xmath95 between the galaxy and dark matter velocity dispersions within halos , but this freedom does not eliminate the constraining power of dynamical observables because the space velocities of the halos themselves do not change .
the main parameter degeneracy is approximately @xmath389 because fixing this combination roughly fixes the halo velocity scale and the abundance of halos at the mass scale of rich galaxy groups .
however , the changing shape of the mass function , the differing sensitivities of different velocity measures , and the different @xmath390 dependence of galaxy - galaxy lensing all serve to break this degeneracy . figures [ fig : chgerr ] and [ fig : matrix ] demonstrate that the cosmological constraints emerge from the full web of clustering observables and are not dominated by one or two statistics on their own . constraints on the galaxy hod will themselves provide valuable tests of galaxy formation models .
the cutoff regime of @xmath97 is difficult to pin down with the clustering statistics considered here , but for a known cosmological model the relation between average satellite number and halo mass is well determined , and the satellite distribution width @xmath112 , concentration @xmath118 , and velocity bias @xmath95 are measured to @xmath4 , 30% , and 3% , respectively ( fig .
[ fig : mcmc_hodpar ] ) .
all of these quantities depend in detail on the physics that governs the evolution of satellites in larger halos ( see , e.g. , @xcite ) , while the relative mass of halos that host central and satellite galaxies depends on the efficiency with which halos feed baryonic mass to their central objects ( see , e.g , discussions by @xcite ) .
these galaxy formation constraints will be especially powerful when derived as a function of luminosity , stellar mass , or other observables .
the two main assumptions built into our modeling are the central - satellite parameterization and environment independence of the hod .
the central - satellite distinction appears well rooted in galaxy formation physics , and it allows us to represent the range of plausible galaxy hods more completely with a moderate number of parameters .
however , we have confirmed with other tests that if we model galaxy bias with a flexible hod parameterization that does not impose a central - satellite distinction , but instead introduces a characteristic mass for the narrow - to - wide transition of @xmath96 as in @xcite , then we reach almost identical conclusions about the cosmological constraining power of the clustering observables considered in this paper .
recent numerical results imply an environmental dependence of halo formation times that opens the door to environmental variation of the hod , especially in the single - galaxy regime . as discussed in [ sec : environment ] , we expect the quantitative impact of such dependence to be small , but potentially significant at the high level of precision we ultimately hope to attain .
investigation of environmental dependence effects and methods of allowing for them in hod modeling are a high priority for future work . the cosmological modeling approach advocated here is closely related to the clf method introduced by @xcite and @xcite , who use clustering data and the global galaxy luminosity function to constrain the dependence of the luminosity function on halo mass . in principle , the clf and hod methods are equivalent
they are merely differential and integral forms of one another
. one can derive the clf from a series of hod fits to galaxy samples with different luminosity thresholds @xcite .
conversely , one can integrate the clf to infer @xmath97 for galaxies above a luminosity threshold @xcite .
the principal virtue of our hod - based approach is that by focusing on a single , well - defined class of galaxies , we can parameterize the hod in a way that seems likely to capture the predictions of any reasonable galaxy formation model .
this kind of comprehensive parameterization is more difficult to achieve for the full clf , and most analyses to date have assumed , for example , that the clf has a schechter form in halos of fixed mass .
nonetheless , it is valuable to pursue both hod and clf approaches and test for consistency of conclusions .
hod modeling complements rather than replaces the perturbative approach based on large - scale measures that can be modeled with linear or quadratic bias .
hod modeling is more complex , but it can take advantage of high - precision clustering measurements on small and intermediate scales .
hod modeling can also amplify the power of the perturbative approach , extending its reach further into the non - linear regime and checking its range of validity at a desired level of precision .
for example , @xcite show that an hod - based approach to redshift - space distortions can improve recovery of the perturbative parameter @xmath39 that controls large - scale flows .
@xcite show that the linear bias model provides an accurate description of galaxy - galaxy lensing for @xmath391 , and they show how to accurately model this phenomenon on smaller scales ( see also @xcite ) . j. yoo et al .
( 2007 , in preparation ) show that the scale - dependent bias factors derived by fitting the projected galaxy correlation function can extend recovery of the shape of the linear matter power spectrum into the mildly non - linear regime .
clf and hod analyses of the 2dfgrs and sdss redshift surveys have already produced a number of interesting results , even though they have considered only a fraction of the potential clustering observables .
these results confirm , at least qualitatively , many of the basic predictions of current galaxy formation models , including the general form of the mean occupation function , the dependence of this function on luminosity , the existence of a minimum mass - to - light ratio in the halos of @xmath392 galaxies where galaxy formation is most efficient , the large gap between the minimum halo mass for central and satellite galaxies above a luminosity threshold , the sub - poisson fluctuations of @xmath96 that are a consequence of this gap , and the strong preference of galaxies with older stellar populations for higher mass halos @xcite . in combination with cmb data , the cosmological constraints from hod modeling of the sdss projected galaxy correlation function are almost as tight as those from the large - scale galaxy power spectrum , and the two analyses are consistent within their statistical uncertainties @xcite .
for the most part , the cosmological inferences from clf / hod modeling of galaxy clustering agree with those from other methods , but matching the mass - to - light ratios of galaxy clusters simultaneously with other clustering data appears to require values of @xmath1 and/or @xmath2 that are substantially lower than the commonly adopted values of 0.3 and 0.9 @xcite .
if this conclusion is correct , then the evidence for it should become much stronger as more clustering observables are brought into play and the sdss data set itself moves to completion . we have focused in this paper on the cosmological parameter constraints that can be derived from galaxy clustering alone , using external data only to guide the choice of power spectrum shape and motivate the assumption of gaussian initial conditions . as with perturbative analyses of large - scale structure , the long - term interest lies in combining these constraints with those from the cmb , the ly@xmath11 forest , type ia supernovae , and other cosmological observables . the complementary sensitivities of these observables lead to much tighter parameter constraints .
more importantly , conflicts among them could point the way to physics beyond the simplest versions of @xmath43cdm , such as evolving dark energy , a gravitational wave contribution to cmb anisotropy , departures from scale invariance in the primordial power spectrum , non - zero space curvature , cosmologically significant neutrino masses , and so forth . by sharpening the constraints from large - scale structure in the new generation of galaxy redshift surveys ,
hod modeling can play a critical role in efforts to test the standard cosmological model and , perhaps , discover its breaking points .
we thank andreas berlind , jeremy tinker , and jaiyul yoo for valuable discussions on these topics .
we thank barth netterfield for suggesting a cubic spline parameterization of the mean occupation function , sandy faber for suggesting the influence matrix investigation , and andy gould for advice on the @xmath0 minimization method .
this work was supported by nsf grants ast 00 - 98584 and ast 04 - 07125 .
z. z. acknowledges the support of nasa through hubble fellowship grant hf-01181.01-a awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, for nasa , under contract nas 5 - 26555
. z. z. was also supported by a presidential fellowship from the graduate school of the ohio state university at an early stage of the project .
for the calculation of the one - halo term of the galaxy - galaxy or galaxy - mass two - point correlation function , we need to know the distribution of galaxy pairs or galaxy - mass particle pairs as a function of separation in each halo .
the pair distribution is basically the convolution of the density profiles of two populations . for
spherically symmetric density distributions @xmath393 and @xmath394 , the fraction of pairs with separation in the range of @xmath91 to @xmath395 is @xmath396 where @xmath397 is the fraction of pairs with separation less than @xmath91 , @xmath398 , @xmath399 , and @xmath11 is the angle between @xmath400 and @xmath401 .
if @xmath393 and @xmath394 follow the same nfw profile truncated at the virial radius , the analytic expression of @xmath402 can be found in @xcite ( see their eq.[1 ] and eq.[a25 ] ; note that @xmath403 in their equations is proportional to @xmath404 ) . in this paper we assume that , like the dark matter , the distribution of galaxies inside halos also follows the nfw profile .
therefore , we use the formula in @xcite when evaluating the one - halo term of galaxy - galaxy correlations .
however , we allow galaxies to follow a different nfw profile than the dark matter in halos of a given mass , which is our way to parameterize the spatial bias inside the halo in this paper .
so , for the purpose of calculating the one - halo term of the galaxy - mass two - point correlation function , we need the pair distribution function corresponding to two nfw profiles with different concentration parameters .
we evaluate the convolution in equation [ eqn : a0 ] for two nfw profiles with concentration parameters @xmath405 and @xmath406 , both truncated at the virial radius @xmath407 of the halo . after tedious algebra
, we derive the following analytic expression for the ( differential ) pair distribution function @xmath408 , where @xmath409 . for @xmath410 @xmath411s,\ ] ] where @xmath412 + \frac{c_1 s}{c_2(c_2+c_1+c_1 c_2 s)(1+c_1 s)},\ ] ] @xmath413 - \frac{c_1(1-s)}{c_2 ( c_2-c_1+c_1 c_2 s)(1+c_1 s)(1+c_1)},\ ] ] @xmath414 + \frac{c_1(1-s)}{c_2 ( c_2-c_1-c_1 c_2 s)(1+c_1-c_1 s)},\ ] ] @xmath415 and @xmath416 . for
the purpose of stableness , one should compute @xmath417 using the limit value @xmath418/2 $ ] as @xmath419 approaches @xmath420 and similarly @xmath421/2 $ ] as @xmath419 approaches @xmath422 . for @xmath423 , @xmath424 where @xmath425 + \frac{s-2}{(1+c_1)(1+c_2)(c_2+c_1+c_1 c_2 s)}.\ ] ] the distribution is normalized so that @xmath426 .
the normalization factor @xmath427 in equations ( [ eqn : a1 ] ) and ( [ eqn : a2 ] ) can be expressed as @xmath428 where we have the following fitting formula for @xmath429 , @xmath430^\mu \{1+b_0\sin [ \omega(\log c-\phi)]\},\ ] ] with @xmath431 , @xmath432 , @xmath433 , @xmath434 , @xmath435 , @xmath436 , @xmath437 , and @xmath438 .
the fitting formula has a fractional error less than 0.2% for @xmath439 .
it can be shown that the pair distribution derived here ( eqs.[[eqn : a1 ] ] and [ [ eqn : a2 ] ] ) reduces to equation ( a25 ) in @xcite for the case @xmath440 .
we also test the above formulae by generating random particle distributions following different nfw profiles and counting particle pairs directly .
the monte carlo result agrees with the analytic formulae perfectly .
we use the gauss - newton method to perform the @xmath0 minimization . here
we briefly describe the main procedure .
readers are referred to @xcite for more details .
assume that we have @xmath441 observables @xmath442 ( @xmath443 ) predicted by a model with @xmath104 parameters @xmath444 ( @xmath445 ) .
observations give @xmath446 .
the @xmath0 is defined as ( einstein s summation convention is used ) @xmath447b_{kl}[f_l({\bf a})-f_{l,{\rm obs}}],\ ] ] where @xmath448 is the inverse of the covariance matrix @xmath449 and has only diagonal components @xmath450 if observational errors are uncorrelated .
if the initial guess of the solution of @xmath451 is @xmath452 , through linearizing @xmath453 around @xmath452 and demanding that the first derivatives of @xmath0 are zero , we obtain an equation for the correction to the initial guess ( @xcite ) , @xmath454 where @xmath455 and @xmath456.\ ] ] the direction of correction for @xmath452 is given by @xmath457 .
we set @xmath458 as the new @xmath452 , where @xmath459 is a small number , and refine the estimation of the solution of @xmath451 .
the above process is repeated until @xmath0 reaches a minimum .
it can be proved that the covariance matrix of @xmath451 is just @xmath460 ( @xcite ) .
if some parameters have additional constraints , we introduce a cost function to realize the constraints and minimize the sum of @xmath0 and the cost function .
for example , if parameter @xmath461 can not exceed @xmath462 , the cost function can be defined as @xmath463 ^ 2 $ ] for @xmath464 and @xmath465 for @xmath466 , where @xmath254 is the tolerance .
another example is that a parameter has some priors .
if @xmath461 is assigned a prior of gaussian distribution with mean @xmath467 and standard deviation @xmath254 , the cost function can be defined as @xmath468 ^ 2 $ ] .
including a cost function leads to additional terms , @xmath469 , in @xmath470 ( eq . [ [ eqn : d_i ] ] ) .
abazajian , k. , et al .
2004 , , 128 , 502 abazajian , k. , et al .
2005 , , 625 , 613 bardeen , j. m. , bond , j. r. , kaiser , n. , & szalay , a. s. 1986 , , 304 , 15 bean , a. j. , ellis , r. s. , shanks , t. , efstathiou , g. , & peterson , b. a. 1983 , , 205 , 605 bell , e. f. & de jong , r. s. 2001 , , 550 , 212 benson , a. j. , cole , s. , frenk , c. s. , baugh , c. m. , & lacey , c. g. 2000 , , 311 , 793 benson , a. j. 2001 , , 325 , 1039 berlind , a. a. , narayanan , v. k. , & weinberg , d. h. 2001 , , 549 , 688 berlind , a. a. & weinberg , d. h. 2002 , , 575 , 587 berlind , a. a. , weinberg , d. h. , benson , a. j. , baugh , c. m. , cole , s. , et al .
2004 , , 593 , 1 berlind , a. a. , blanton , m. r. , hogg , d. w. , weinberg , d. h. , dav , r. , eisenstein , d. j. , & katz , n. 2005 , , 629 , 625 blanton , m. r. , eisenstein , d. j. , hogg , d. w. , & zehavi , i. 2006 , , 645 , 977 bond , j. r. , cole , s. , efstathiou , g. , & kaiser , n. 1991 , , 379 , 440 bullock , j. s. , kolatt , t. s. , sigad , y. , somerville , r. s. , klypin , a. a. , primack , j. r. , dekel , a. 2001 , , 321 , 559 bullock , j. s. , weschsler , r. h. , & somerville , r. s. 2002 , , 329 , 246 carlberg , r. g. , yee , h. k. c. , & ellingson , e. 1997 , , 478 , 462 cen , r. , & ostriker , j. p. 2000 , , 538 , 83 coil , a. l. , et al .
2006 , , 638 , 668 cooray , a , , & sheth , r. 2002 , , 372 , 1 cooray , a. 2006 , , 365 , 842 cole , s. , et al .
2005 , , 362 , 505 coles , p. 1993
, , 262 , 1065 colless , m. , et al .
2001 , , 328 , 1039 collister , a. a. , & lahav , o. 2005 , , 361 , 415 croton , d. j. , gao , l. , & white , s. d. m. 2006 , , in press , arxiv astrophysics e - prints , arxiv : astro - ph/0605636 croft , r. a. c. , weinberg , d. h. , bolte , m. , burles , s. , hernquist , l. , katz , n. , kirkman , d. , tytler , d. 2002 , , 581 , 20 davis , m. , & peebles , p. j. e. 1983 , , 267 , 465 dekel , a. , & lahav , o. 1999 , , 520 , 24 efstathiou , g. , bond , j. r. , & white , s. d. m. 1992 , , 258 , 1 fischer , p. , et al .
2000 , , 120 , 1198 fry , j. 1984 , , 279 , 499 fry , j. n. 1994 , , 73 , 215 fry , j. n. , & gaztaaga , e. 1993 , , 413 , 447 gao , l. , springel , v. , & white , s. d. m. 2005 , , 363 , l66 gilks , w. r. , richardson , s. , & spiegelhalter , d. j. 1996 , markov chain monte carlo in practice ( london : chapman and hall ) gould , a. 2003 , astro - ph/0310577 guzik , j. , & seljak , u. 2002 , , 335 , 311 hamilton , a. j. s. 1998 , in the evolving universe , ed .
d. hamilton ( dordrecht : kluwer ) , 185 harker , g. , cole , s. , helly , j. , frenk , c. , & jenkins , a. 2006 , , 367 , 1039 hawkins , e. , et al .
2003 , , 346 , 78 heisler , j. , tremaine , s. , & bahcall , j. n. 1985 , , 298 , 8 hoekstra , h. , yee , h. k. c. , & gladders , m. d. 2001 , , 558 , l11 hu , w. & sugiyama , n. 1996 , , 471 , 542 jenkins , a. , et al .
1998 , , 499 , 20 jenkins , a. , frenk , c. s. , white , s. d. m. , colberg , j. m. , cole , s. , evrard , a. e. , couchman , h. m. p. , & yoshida , n. 2001 , , 321 , 372 jing , y. p. 1998
, , 503 , l9 jing , y. p. , & brner , g. 1998 , , 503 , 37 jing , y. p. , b " orner , g. , & suto , y. 2002 , , 564 , 15 juszkiewicz , r. , weinberg , d. h. , amsterdamski , p. , chodorowski , m. , & bouchet , f. 1995 , , 442 , 39 kaiser , n. 1987 , , 227 , 1 kamionkowski , m. & buchalter , a. 1999 , , 514 , 7 kauffmann , g. , et al .
2003 , , 341 , 33 kauffmann , g. , colberg , j. m. , diaferio , a. , & white , s. d. m. 1999 , , 303 , 188 kauffmann , g. , nusser , a. , & steinmetz , m. 1997 , , 286 , 795 kravtsov , a. v. , berlind , a. a. , wechsler , r. h. , klypin , a. a. , gottloeber , s. , allgood , b. , & primack , j. r. 2004 , , 609 , 35 kochanek , c. s. , white , m. , huchra , j. , macri , l. , jarrett , t. h. , schneider , s. e. , & mader , j. 2003 , , 585 , 161 lahav , o. , et al .
2002 , , 333 , 961 lee , k. , giavalisco , m. , gnedin , o. y. , somerville , r. , ferguson , h. , dickinson , m. , & ouchi , m. 2006 , , 642 , 63 lemson , g. , & kauffmann , g. 1999 , , 302 , 111 lin , y .- t . , mohr , j. j. , & stanford , s. a. 2004 , , 610 , 745 ma , c. , & fry , j. n. 2000 , , 543 , 503 magliocchetti , m. , & porciani , c. 2003 , , 346 , 186 mandelbaum , r. , tasitsiomi , a. , seljak , u. , kravtsov , a. v. , & wechsler , r. h. 2005a , , 362 , 1451 mandelbaum , r. , seljak , u. , kauffmann , g. , hirata , c. m. , brinkmann , j. 2006 , , 368 , 715 mann , r. g. , peacock , j. a. , & heavens , a. f. 1998 , , 293 , 209 marinoni , c. , & hudson , m. j. 2002 , , 569 , 101 mo , h. j. , yang , x. , van den bosch , f. c. , & jing , y. p. 2004
, , 349 , 205 moustakas , l. a. , & somerville , r. s. 2002 , , 577 , 1 nagai , d. , & kravtsov , a. v. 2005 , , 618 , 557 narayanan , v. k. , berlind , a. a. , & weinberg , d. h. 2000 , , 528 , 1 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1995 , , 275 , 56 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1996 , , 462 , 563 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1997 , , 490 , 493 netterfield , c. b. , et al . 2002 , , 571 , 604 norberg , p. , et al .
2002 , , 332 , 827 ouchi , m , et al .
2005 , , 635 , l117 peacock , j. a. , & smith , r. e. 2000 , , 318 , 1144 pearce , f. r. , jenkins , a. , frenk , c. s. , white , s. d. m. , thomas , p. a. , couchman , h. m. p. , peacock , j. a. , & efstathiou , g. 2001 , , 326 , 649 pen , u. 1998 , , 504 , 601 percival , w. j. , et al . 2002 , , 337 , 1068 perlmutter , s. , et al . 1999 , , 517 , 565 pierpaoli , e. , scott , d. , & white , m. 2001 , , 325 , 77 pryke , c. , halverson , n. w. , leitch , e. m. , kovac , j. , carlstrom , j. e. , holzapfel , w. l. , & dragovan , m. 2002 , , 568 , 46 reed , d. , gardner , j. , quinn , t. , stadel , j. , fardal , m. , lake , g. , & governato , f. 2003 , , 346 , 565 riess , a. g. , et al .
1998 , , 116 , 1009 riess , a. g. , et al . 2004 , ,
607 , 665 rozo , e. , dodelson , s. , & frieman , j. 2004 , , 70 , 083008 saunders , w. , et al . 2000 , , 317 , 55 scherrer , r. j. , & weinberg , d. h. 1998 , , 504 , 607 scoccimarro , r. , sheth , r. k. , hui , l. , & jain , b. 2001 , , 546 , 20 scranton , r. 2003 , , 339 , 410 seljak , u. 2000 , , 318 , 203 seljak , u. & warren m. s. 2004 , , 355 , 129 seljak , u. , et al .
2005a , , 71 , 3511 seljak , u. , makarov , a. , mcdonald , p. , anderson , s. , et al .
2005b , , 71 , 3515 seljak , u. , & zaldarriaga , m. 1996 , , 469 , 437 shectman , s. a. , landy , s. d. , oemler , a. , tucker , d. l. , lin , h. , kirshner , r. p. , & schechter , p. l. 1996 , , 470 , 172 sheldon , e. s. , et al .
2004 , , 127 , 2544 sheth , r. k. & diaferio , a. 2001 , , 322 , 901 sheth , r. k. , mo , h. j. , & tormen , g. 2001 , , 323 , 1 sheth , r. k. & tormen , g. 1999 , , 308 , 119 sheth , r. k. , & tormen , g. 2004 , , 350 , 1385 smith r. e. , et al .
2003 , , 341 , 1311 spergel , d. n. , et al .
2003 , , 148 , 175 spergel , d. n. , et al .
2006 , , submitted , arxiv astrophysics e - prints , arxiv : astro - ph/0603449 springel , v. , et al .
2005 , , 435 , 629 springel , v. , yoshida , n. , & white , s. d. m. 2001 , new astronomy , 6 , 79 taylor , j. e. , & babul , a. 2004 , , 348 , 811 taylor , j. e. , & babul , a. 2005a , , 364 , 515 taylor , j. e. , & babul , a. 2005b , , 364 , 535 tegmark , m. , et al .
2004 , , 606 , 702 tegmark , m. , et al .
2004 , , 69 , 103501 tegmark , m , et al .
2006 , , in press , arxiv astrophysics e - prints , arxiv : astro - ph/0608632 tinker , j. l. , weinberg , d. h. , zheng , z. , & zehavi , i. 2005a , , 631 , 41 tinker , j. l. , weinberg , d. h. , & zheng , z. 2006 , , 368 , 85 tinker , j. l. 2006 , , submitted , arxiv astrophysics e - prints , arxiv : astro - ph/0604217 vale , a. , & ostriker , j. p. 2006
, , 371 , 1173 van den bosch , f. c. , mo , h. j. , & yang , x. h. 2003 , , 345 , 923 van den bosch , f. c. , yang , x. h. , & mo , h. j. 2003 , , 340 , 771 van den bosch , f. c. , norberg , p. , mo , h. j. , & yang , x. 2004 , , 352 , 1302 van den bosch , f. c. , yang , x. h. , mo , h. j. , & norberg , p. 2005
, , 356 , 1233 van den bosch , f. c. , et al .
2006 , , submitted , arxiv astrophysics e - prints , arxiv : astro - ph/0610686 verde , l. , et al .
2002 , , 335 , 432 wang , y. , yang , x. , mo , h. j. , van den bosch , f. c. , & chu , y. 2004 , , 353 , 287 weinberg , d. h. 1995 , in wide - field spectroscopy and the distant universe , ed .
s. j. maddox & a. aragn - salamanca ( singapore : world scientific ) , 129 weinberg , d. h. 2002 , a new era in cosmology , asp conference proceedings , vol .
t. shanks & n. metcalfe ( san francisco : astronomical society of the pacific ) , p.3 weinberg , d. h. , dav ' e , r. , katz , n. , & hernquist , l. 2004 , , 601 , 1 white s. d. m. , 1996 , in schaeffer r. , silk j. , spiro m. , zinn - justin j. , eds , cosmology and large scale structures , les houches session lx .
elsevier , amsterdam , p.77 s yan , r. , madgwick , d. s. , & white , m. 2003 , , 598 , 848 yang , x. h. , mo , h. j. , & van den bosch , f. c. 2003 , , 339 , 1057 yang , x. , mo , h. j. , jing , y. p. , van den bosch , f. c. , & chu , y. 2004 , , 350 , 1153 yang , x. , mo , h. j. , jing , y. p. , & van den bosch , f. c. 2005 , , 358 , 217 yoo , j. , tinker , j. l. , weinberg , d. h. , zheng , z. , katz , n. , & dav , r.2006 , , 652 , 26 york , d. g. , et al .
2000 , , 120 , 1579 yoshikawa , k. , taruya , a. , jing , y. p. , & suto , y. 2001 , , 558 , 520 zehavi , i. , weinberg , d. h. , zheng , z. , berlind , a. a. , frieman , j. a. , et al .
2004 , , 608 , 16 zehavi , i. , et al .
2005 , , 630 , 1 zentner , a. r. , berlind , a. a. , bullock , j. s. , kravtsov , a. v. , & wechsler , r. h. 2005 , , 624 , 505 zheng , z. , tinker , j. l. , weinberg , d. h. , & berlind , a. a. 2002 , , 575 , 000 | adopting the framework of the halo occupation distribution ( hod ) , we investigate the ability of galaxy clustering measurements to simultaneously constrain cosmological parameters and galaxy bias . starting with a fiducial cosmological model and galaxy hod , we calculate spatial clustering observables on a range of length and mass scales , dynamical clustering observables that depend on galaxy peculiar velocities , and the galaxy - matter cross - correlation measurable by weak lensing .
we then change one or more cosmological parameters and use @xmath0-minimization to find the galaxy hod that best reproduces the original clustering .
our parameterization of the hod incorporates a flexible relation between galaxy occupation numbers and halo mass and allows spatial and velocity bias of galaxies within dark matter halos . despite this flexibility , we find that changes to the hod can not mask substantial changes to the matter density @xmath1 , the matter clustering amplitude @xmath2 , or the shape parameter @xmath3 of the linear matter power spectrum cosmology and bias are not degenerate . with the conservative assumption of 10% fractional errors , the set of observables considered here can provide @xmath4 ( @xmath5 ) constraints on @xmath2 , @xmath1 , and @xmath3 , using galaxy clustering data _ alone_. the combination @xmath6 is constrained to @xmath7 . in combination with traditional methods that focus on large - scale structure in the `` perturbative '' regime , hod modeling can greatly amplify the cosmological power of galaxy redshift surveys by taking advantage of high - precision clustering measurements at small and intermediate scales ( from sub - mpc to @xmath8 ) . at the same time
, the inferred constraints on the galaxy hod provide valuable tests of galaxy formation theory . |
the parity - violating deep inelastic scattering ( pvdis ) experiment e08 - 011 was completed in december 2009 at the thomas jefferson national accelerator facility ( jlab ) .
the goal of this experiment @xcite was to measure with high precision the parity - violating asymmetry in deep inelastic scattering of a polarized 6 gev electron beam on an unpolarized liquid deuterium target .
this asymmetry is sensitive to the quark weak axial charge @xmath2 which corresponds to a helicity dependence in the quark coupling with the @xmath3 boson . for electron inclusive scattering from an unpolarized target ,
the electromagnetic interaction is parity conserving and is insensitive to the spin flip of the incoming electron beam .
only the weak interaction violates parity and causes a difference between the right- and the left - handed electron scattering cross - sections @xmath4 and @xmath5 .
the dominant contribution to the parity violation asymmetry , @xmath6 , arises from the interference between electromagnetic and weak interactions and is proportional to the four momentum transfer squared @xmath7 for @xmath8 .
the magnitude of the asymmetry is on the order of @xmath0 or @xmath9 parts per million ( ppm ) at @xmath10 ( gev/@xmath11)@xmath12 .
the goal of jlab e08 - 011 was to measure the pvdis asymmetries to statistical precisions of 3% and 4% at @xmath33 and @xmath34 ( gev/@xmath11)@xmath12 , respectively , and under the assumption that hadronic physics corrections are small , to extract the quark axial weak charge combination @xmath35 .
in addition , the systematic uncertainty goal was less than @xmath36 . for this experiment , the expected asymmetries were 91 and 160 ppm respectively at the two @xmath7 values @xcite . to achieve the required precision , an event rate capability of up to 600 khz was needed .
the main challenge of deep inelastic scattering experiments is the separation of scattered electrons from the pion background in the spectrometer and detector system .
the neutral pions would decay into @xmath37 pairs from which the electrons produced can not be rejected by detectors .
this pair production background was studied by reversing the spectrometer magnet settings and measure the @xmath38 yield , and the effect on the measured asymmetries was found to be negligible .
charged pions are produced primarily from nucleon resonance decays and could carry a parity violation asymmetry corresponding to the @xmath7 at which the resonances are produced , typically a fraction of the asymmetry of electrons with the same scattered momentum .
assuming that a fraction @xmath39 of the detected events are @xmath40 and @xmath41 are electrons , the measured asymmetry is @xmath42 where @xmath43 is the desired electron scattering asymmetry and @xmath44 is the asymmetry of the pion background . to extract @xmath43 to a high precision , one needs either to minimize the pion contamination @xmath39 to a negligible level , or to correct the measured asymmetry for the asymmetry of pions , which itself needs to be measured precisely . for the pvdis experiment ,
the goal was to control @xmath39 to the @xmath0 level provided that the pion asymmetries did not exceed those of electrons .
the experiment used a 100 @xmath45a electron beam with a polarization of approximately 90% and a 20-cm long liquid deuterium target .
the two high resolution spectrometers ( hrs ) @xcite were used to detect scattered events .
while the standard hrs detector package and data acquisition ( daq ) system routinely provide a @xmath46 pion rejection with approximately @xmath47 electron efficiency , they are based on full recording of the detector signals and are limited to event rates up to 4 khz @xcite .
this is not sufficient for the high rates expected for the experiment .
( the hrs daq will be referred to as `` standard daq '' hereafter . )
recent parity violation electron scattering experiments , such as happex @xcite , and prex @xcite at jlab , focused on elastic scattering from nuclear or nucleon targets that are typically not contaminated by inelastic backgrounds .
signals from the detectors can be integrated and a helicity dependence in the integrated signal can be used to extract the physics asymmetry .
an integrating daq was also used in the preceding pvdis measurement at slac @xcite in which approximately 2% of the integrated signal was attributed to pions .
the sample experiment @xcite at mit - bates focused also on elastic scattering but the inelastic contamination was more challenging to reject , and an air cherenkov counter was used to select only elastic scattering events . in the mainz pva4 experiment @xcite , particles were detected in a total absorption calorimeter and the integrated energy spectrum was recorded . charged pions and other background were separated from electrons in the offline analysis of the energy spectrum , and the pion rejection was on the order of 100:1 based on the characteristics of the calorimeter .
high performance particle identification can usually be realized in a counting - based daq where each event is evaluated individually . in the g0 experiment @xcite at jlab ,
a superconducting spectrometer with a @xmath48 azimuthal angle coverage was used to detect elastically scattered protons at the forward angle and elastic electrons at the backward angle . at the forward angle
, protons were identified using time - of - flight . at the backward angle
, pions were rejected from electrons using an aerogel cherenkov counter , and a pion rejection factor of @xmath49 or better was reported @xcite .
the deadtime correction of the counting system was on the order of a few percent @xcite .
while the pvdis experiment could fully utilize existing spectrometers and detectors at jlab , examination of all existing techniques for pv measurements made it clear that a custom electronics and daq were needed to keep the systematic uncertainties due to data collection to below 1% . in this paper
we describe a scaler - based , cost effective counting daq which limited the pion contamination of the data sample to a negligible level of @xmath50 .
basic information on the detector package and the daq setup will be presented first and followed by the analysis of electron detection efficiency , pion rejection and contamination , corrections due to counting deadtime , and the statistical quality of the asymmetry measurement .
the design goal of the daq is to record data up to 600 khz with hardware - based pid and well measured and understood deadtime effects .
the following detectors in the hrs @xcite were used to characterize scattered particles : two scintillator planes provided the main trigger , while a co@xmath51 gas cherenkov detector and a double - layer segmented lead - glass detector provided particle identification information .
the vertical drift chambers ( as the tracking detector ) were used during calibration runs but were turned off during production data taking because they were not expected to endure the high event rates . for the gas cherenkov and the lead - glass detector ,
a full recording of their output adc data was not feasible at the expected high rate .
instead their signals were passed through discriminators and logic units to form preliminary electron and pion triggers .
these preliminary triggers were then combined with the scintillator triggers to form the final electron and pion triggers , which were sent to scalers to record the event counts and used offline to form asymmetries @xmath52 , where @xmath53 is the integrated rate of the triggers normalized to the integrated beam charge for the right@xmath54 and left@xmath55 handed spin ( helicity ) states of the incident electron beam .
the scalers that counted triggers and the beam charge were integrated over the helicity period , which was flipped pseudo - randomly at 30 hz per the experimental technique used by the happex experiments @xcite . for the hrs
the two layers of the lead - glass detector are called `` preshower '' and `` shower '' detectors , respectively .
the preshower in the right hrs ( the spectrometer located to the right side of the beamline when viewed along the beam direction ) has @xmath56 blocks arranged in a @xmath57 array , with the longest dimension of the blocks aligned perpendicular to the particle trajectory .
for the two blocks in each row , only the ends facing outward are read out by photo - multiplier tubes ( pmts ) , while the other ends of the two blocks are facing each other and not read out .
therefore , the preshower detector has @xmath56 output channels .
all preshower blocks were individually wrapped to prevent light leak .
the shower detector in the right hrs had @xmath58 blocks arranged in a @xmath59 array with the longest dimension of the blocks aligned along the trajectory of scattered particles .
pmts were attached to each block of the right shower detector on one end only , giving normally @xmath58 output channels . however to minimize the electronics needed for this experiment ( see next paragraph ) , only 60 of the 75 shower blocks were used while signals from the 15 blocks on the edge were not utilized by the daq .
the reduction of the hrs acceptance due to not using these side blocks was negligible .
the preshower and the shower detectors in the left hrs are similar to the preshower detector on the right hrs except that for each detector there are @xmath60 blocks arranged in a @xmath61 array . because the lead - glass detectors in the left and right hrs are different , design of the lead - glass - based triggers of the daq is also different , as shown in fig .
[ fig : grouping ] . as a compromise between the amount of electronics needed and the rate in the front end logic modules ,
the lead - glass blocks in both the preshower and the shower detectors were divided into 6 ( 8) groups for the left ( right ) hrs , with each group consisting typically 8 blocks .
signals from the 8 blocks in each group were added using a custom - made analog summing unit called the `` sum8 module '' , then passed to discriminators .
the geometry and the position of each preshower group were carefully chosen to match those of the corresponding shower group to maximize electron detection efficiency . on the left hrs , adjacent groups in both preshower and shower had overlapping blocks , while for the right hrs only preshower groups were overlapping . to allow overlap between adjacent groups , signals from preshower blocks on the right hrs and from both preshower and shower blocks on the left hrs were split into two identical copies using passive splitters . a schematic diagram of the daq electronics for the right hrs is shown in fig .
[ fig : daqflowchart ] .
preliminary electron and pion triggers were formed by passing shower ( ss ) and preshower ( ps ) signals and their sums , called total shower ( ts ) signals , through discriminators with different thresholds . for electron triggers ,
logical ands of the ps discriminator and the ts discriminator outputs were used . for pions , low threshold discriminators on the ts signal alone
were sent to logical or modules to produce preliminary triggers .
additional background rejection was provided by the `` gate '' circuit , which combined signals from the gas cherenkov ( gc ) and the `` t1 '' signal @xcite from the scintillators ( sc ) . each valid coincidence between gc and t1
would produce a 150-ns wide electron gate signal that allowed an output to be formed by the logical and modules from the preliminary electron triggers .
each valid t1 signal without the gc signal would produce a 150-ns wide pion gate signal that allowed an output to be formed by the logical or modules from the preliminary pion triggers .
the outputs of the logical and and or modules are called group electron and pion triggers , respectively .
all six ( eight ) group electron or pion triggers were then ored together to form the global electron or pion trigger for the left ( right ) hrs .
all group and the final electron and pion triggers were counted using scalers . because pions do not produce large enough lead - glass signals to trigger the high threshold ts discriminators for the electron triggers
, pions do not introduce extra counting deadtime for the electron triggers .
however , the 150-ns width of the electron gate signal would cause pion contamination in the electron trigger .
this effect will be presented in section [ sec : pid ] . in order to monitor the counting deadtime of the daq
, two identical paths of electronics were constructed .
the only difference between the two paths is in the ps and the ts discriminator output widths , set at 30 ns and 100 ns for the `` narrow '' and the `` wide '' paths , respectively .
the scalers are rated for 250 mhz ( 4 ns deadtime ) and therefore do not add to the deadtime .
in addition , the output width of all logic modules was set to 15 ns , so the deadtime of the daq for each group is dominated by the deadtime of the discriminators .
detailed analysis of the daq deadtime will be presented in section [ sec : deadtime ] .
the sum8 modules used for summing all lead - glass signals also served as fan - out modules , providing exact copies of the input pmt signals .
these copies were sent to the standard hrs daq for calibration . during the experiment ,
data were collected at low rates using reduced beam currents with both daqs functioning , such that a direct comparison of the two daqs could be made .
vertical drift chambers were used during these low rate daq studies .
outputs from all discriminators , signals from the scintillator and the gas cherenkov , and all electron and pion group and global triggers were sent to fastbus tdcs ( fbtdc ) and were recorded in the standard daq .
data from these fbtdcs were used to align the amplitude spectrum and timing of all signals .
they also allowed the study of the cherenkov and the lead - glass detector performance for the new daq .
full sampling of partial analog signals was done using flash - adcs ( fadcs ) at low rates intermittently during the experiment . for one group on the left and one group on the right hrs ,
the preshower and the shower sum8 outputs , the intermediate logical signals of the daq , and the output electron and pion triggers were recorded .
these fadc data provided a study of pileup effects to confirm the deadtime simulation and to provide the input parameters for the simulation , specifically the rise and fall times of the signals and their widths .
during the experiment data were taken at two deep inelastic scattering ( dis ) kinematics at @xmath62 and @xmath34 ( gev/@xmath11)@xmath12 . these were the main production kinematics and will be referred to as dis#1 and dis#2 , respectively . due to limitation of the spectrometer magnets ,
dis#1 was taken only on the left hrs , while dis#2 was taken on both left and right hrss .
in addition , data were taken at five kinematics within or near the nucleon resonance region with their invariant mass @xmath63 between the @xmath64 resonance and just above @xmath65 gev .
these data were used for the purpose of radiative corrections and will be referred to as res i through v ( although kinematics v was located slightly above @xmath65 gev ) .
data for each of the resonance settings were taken only with one hrs because of the spectrometer magnet limitations as well as to optimize the beam time allocation .
the kinematic settings are shown in table [ tab : kine ] along with the observed electron rate @xmath66 and the pion to electron ratio @xmath67 in the hrs .
the highest electron rate occurred at res ii at approximately 600 khz .
.overview of kinematics settings during the experiment , including : the beam energy @xmath68 , the spectrometer central angle setting @xmath69 and central momentum setting @xmath70 , the observed electron rate @xmath66 and the @xmath71 ratio @xmath67 .
[ cols="^,^,^,^,^,^,^",options="header " , ] results shown in table [ tab : deadtime_all ] provide a direct correction to the measured asymmetry , and the uncertainties are small compared with other dominant systematic uncertainties such as the approximately 2% uncertainty from beam polarizations . in practice , the deadtime correction was applied to data on a run - by - run basis with the deadtime of each run calculated using the actual beam current during the run and the linear fitting results from fig .
[ fig : hats_dtoverall ] . simulated deadtime loss of the global electron trigger for the three dis spectrometer and kinematics combinations and the five resonance kinematics , for the narrow path ( open triangles , fit by dashed lines ) and the wide path ( open squares , fit by solid lines ) .
the error bars shown are due to statistical uncertainty of the simulation .
see table [ tab : deadtime_all ] for final uncertainty evaluation of the total deadtime loss . , title="fig:",scaledwidth=51.0% ] simulated deadtime loss of the global electron trigger for the three dis spectrometer and kinematics combinations and the five resonance kinematics , for the narrow path ( open triangles , fit by dashed lines ) and the wide path ( open squares , fit by solid lines ) .
the error bars shown are due to statistical uncertainty of the simulation .
see table [ tab : deadtime_all ] for final uncertainty evaluation of the total deadtime loss .
, title="fig:",scaledwidth=49.0% ] the physics asymmetries sought for in this experiment were expected to be in the order of @xmath9 ppm .
the measured asymmetries were about @xmath72 of the expected values due to beam polarization . to understand the systematics of the asymmetry measurement , a half - wave plate ( hwp )
was inserted in the beamline to flip the laser helicity in the polarized source during half of the data taking period .
the measured asymmetries flipped sign for each beam hwp change and the magnitude of the asymmetry remained consistent within statistical error bars . the asymmetries can be formed from event counts of each beam helicity pair , with 33-ms of helicity right and 33-ms of helicity left beam , normalized by the beam charge .
figure [ fig : asym ] shows the pull distribution of these pair - wise asymmetries with the `` pull '' defined as @xmath73 where @xmath74 is the asymmetry extracted from the @xmath75-th beam helicity pair with the hwp states already corrected and @xmath76 its statistical uncertainty with @xmath77 the event count from the right ( left ) helicity pulse of the pair , and @xmath78 is the asymmetry averaged over all beam pairs .
one can see that the asymmetry spectrum agrees to five orders of magnitude with the gaussian distribution , as expected from purely statistical fluctuations .
pull distribution [ eq.([eq : pull ] ) ] for the global electron narrow trigger for the three dis spectrometer and kinematics combinations , and the five resonance kinematics .
the dis#2 distribution has both the left and the right hrs data combined .
, title="fig:",scaledwidth=50.0% ] pull distribution [ eq.([eq : pull ] ) ] for the global electron narrow trigger for the three dis spectrometer and kinematics combinations , and the five resonance kinematics .
the dis#2 distribution has both the left and the right hrs data combined .
, title="fig:",scaledwidth=50.0% ]
a scaler - based counting daq with hardware - based particle identification was successfully implemented in the 6 gev pvdis experiment at jefferson lab to measure parity - violating asymmetries at the @xmath0 level at event rates of up to 600 khz .
asymmetries measured by the daq followed gaussian distributions as expected from purely statistical measurements .
particle identification performance of the daq was measured and corrections were applied to the data on a day - to - day basis .
the overall pion contamination in the electron sample was controlled to approximately @xmath1 or lower , with an electron efficiency above 91% during most of the data production period of the experiment .
the daq deadtime was evaluated from a full - scale timing simulation and contributed an uncertainty of no more than @xmath79 to the final asymmetry results .
systematic uncertainties from the pion contamination and the counting deadtime therefore were both negligible compared to the @xmath80 statistical uncertainty and other leading systematic uncertainties .
results presented here demonstrate that accurate asymmetry measurements can be performed with even higher event rates or backgrounds with this type of scaler - based daq .
this work was supported in part by the jeffress memorial trust under award no .
j-836 , the u.s .
national science foundation under award no . 0653347 , and the u.s .
department of energy under award no .
de - sc0003885 and de - ac02 - 06ch11357 . *
notice : * authored by jefferson science associates , llc under u.s .
doe contract no .
de - ac05 - 06or23177 .
government retains a non - exclusive , paid - up , irrevocable , world - wide license to publish or reproduce this manuscript for u.s .
government purposes .
publications on the e08 - 011 physics asymmetries are in preparation .
r. n. cahn and f. j. gilman , phys .
d * 17 * , 1313 ( 1978 ) .
k. nakamura _ et al .
_ [ particle data group ] , j. phys .
* g37 * , 075021 ( 2010 ) .
k. a. aniol _ et al . _
[ happex collaboration ] , phys .
c * 69 * , 065501 ( 2004 ) .
a. acha _
[ happex collaboration ] , phys .
lett . * 98 * , 032301 ( 2007 ) .
k. a. aniol _ et al . _
[ happex collaboration ] , phys .
lett . *
96 * , 022003 ( 2006 ) .
k. a. aniol _ et al . _
[ happex collaboration ] , phys .
b * 635 * , 275 ( 2006 ) .
z. ahmed _ et al .
_ [ happex collaboration ] , phys . rev .
lett . * 108 * , 102001 ( 2012 ) .
s. abrahamyan , z. ahmed , h. albataineh , k. aniol , d. s. armstrong , w. armstrong , t. averett and b. babineau _ et al .
_ , phys .
lett . * 108 * , 112502 ( 2012 ) .
et al . _ ,
phys . lett . *
b77 * ( 1978 ) 347 .
f. e. maas _ et al .
_ [ a4 collaboration ] , q**2 = 0.230-(gev / c)**2 , phys . rev .
lett . * 93 * , 022002 ( 2004 ) .
f. e. maas , k. aulenbacher , s. baunack , l. capozza , j. diefenbach , b. glaser , t. hammel and d. von harrach _ et al .
_ , * * 2 = 0.108 ( gev / c)**2 , phys . rev .
lett . * 94 * , 152001 ( 2005 ) .
s. baunack , k. aulenbacher , d. balaguer rios , l. capozza , j. diefenbach , b. glaser , d. von harrach and y. imai _ et al .
_ , phys .
lett . * 102 * , 151803 ( 2009 ) .
d. h. beck , phys .
d * 39 * , 3248 ( 1989 ) .
d. s. armstrong _ et al . _
[ g0 collaboration ] , phys .
lett . * 95 * , 092001 ( 2005 ) .
d. androic _ et al . _
[ g0 collaboration ] , phys . rev .
* 104 * , 012001 ( 2010 ) .
d. marchand , j. arvieux , g. batigne , l. bimbot , a. s. biselli , j. bouvier , h. breuer and r. clark _ et al .
instrum .
a * 586 * , 251 ( 2008 ) .
d. androic _ et al . _
[ g0 collaboration ] , nucl .
instrum .
a * 646 * , 59 ( 2011 ) . | an experiment that measured the parity - violating asymmetries in deep inelastic scattering was completed at the thomas jefferson national accelerator facility in experimental hall a. from these asymmetries , a combination of the quark weak axial charge could be extracted with a factor of five improvement in precision over world data . to achieve this , asymmetries at the @xmath0 level needed to be measured at event rates up to 600 khz and the high pion background typical to deep inelastic scattering experiments needed to be rejected efficiently . a specialized data acquisition ( daq ) system with intrinsic particle identification ( pid )
was successfully developed and used : the pion contamination in the electron samples was controlled at the order of @xmath1 or below with an electron efficiency of higher than 91% during most of the production period of the experiment , the systematic uncertainty in the measured asymmetry due to daq deadtime was below 0.5% , and the statistical quality of the asymmetry measurement agreed with the gaussian distribution to over five orders of magnitudes .
the daq system is presented here with an emphasis on its design scheme , the achieved pid performance , deadtime effect and the capability of measuring small asymmetries . , , , , , , , , jefferson lab ; hall a ; pvdis ; daq |
tropical algebraic geometry is the geometry of the tropical semiring @xmath0 .
its objects are polyhedral cell complexes which behave like complex algebraic varieties .
although this is a very new field in which many basic questions have not yet been addressed ( see @xcite for a nice introduction ) , tropical geometry has already been shown to have remarkable applications to enumerative geometry ( see @xcite ) , as well as connections to representation theory ( see @xcite , @xcite , @xcite ) . the classical theory of total positivity concerns matrices in which all minors are positive .
however , in the past decade this theory has been extended by lusztig ( see @xcite and @xcite ) , who introduced the totally positive variety @xmath11 in an arbitrary reductive group @xmath12 and the totally positive part @xmath13 of a real flag variety @xmath14 . in the process , lusztig discovered surprising connections between his theory of canonical bases for quantum groups and the theory of total positivity . in this paper
we introduce the totally positive part ( or positive part , for short ) of the tropicalization of an arbitrary affine variety over the ring of puiseux series , and then investigate what we get in the case of the grassmannian @xmath15 .
first we give a parameterization of the totally positive part of the grassmannian , largely based on work of postnikov @xcite , and then we compute its tropicalization , which we denote by @xmath1 .
we identify @xmath1 with a polyhedral subcomplex of the @xmath16-dimensional grbner fan of the ideal of plcker relations , and then show that this fan , modulo its @xmath17-dimensional lineality space , is combinatorially equivalent to an @xmath18-dimensional fan which we explicitly describe . as a special case , we show that @xmath2 is a fan which appeared in the work of stanley and pitman ( see @xcite ) , which parameterizes certain binary trees , and which is combinatorially equivalent to the ( type @xmath19 ) associahedron . we also show that @xmath4 and @xmath5 are fans which are closely related to the fans of the types @xmath6 and @xmath7 associahedra , which were first introduced in @xcite .
these results are strikingly reminiscent of the results of fomin and zelevinsky @xcite , and scott @xcite , who showed that the grassmannian has a natural cluster algebra structure which is of type @xmath19 for @xmath8 , type @xmath6 for @xmath9 , and type @xmath7 for @xmath10 .
( fomin and zelevinsky proved the @xmath8 case and stated the other results ; scott worked out the cluster algebra structure of all grassmannians in detail . ) finally , we suggest a general conjecture about the positive part of the tropicalization of a cluster algebra .
in this section we will define the tropicalization and positive part of the tropicalization of an arbitrary affine variety over the ring of puiseux series .
we will then describe the tropical varieties that will be of interest to us .
let @xmath20 and @xmath21 be the fields of puiseux series over @xmath22 and @xmath23 .
every puiseux series @xmath24 has a unique lowest term @xmath25 where @xmath26 and @xmath27 . setting @xmath28
, this defines the _ valuation map _ @xmath29 .
we define @xmath30 to be @xmath31 .
we will discuss the wisdom of this definition later ; for practically all purposes , the reader may think of @xmath32 as if it were @xmath22 and of @xmath30 as if it were @xmath33 .
let @xmath34 $ ] be an ideal .
we define the tropicalization of @xmath35 , denoted @xmath36 , to be the closure of the image under @xmath37 of @xmath38 , where @xmath35 is the variety of @xmath39 .
similarly , we define the positive part of @xmath36 , which we will denote as @xmath40 , to be the closure of the image under @xmath37 of @xmath41 .
note that @xmath42 and @xmath43 are slight abuses of notation ; they depend on the affine space in which @xmath44 is embedded and not solely on the variety @xmath44 . if @xmath45 \setminus \ { 0 \}$ ] , let the _ initial form _
@xmath46 $ ] be defined as follows : write @xmath47 for @xmath48 chosen as large as possible such that all powers of @xmath49 in @xmath50 are nonnegative
. then @xmath51 is the polynomial obtained from @xmath50 by plugging in @xmath52 . if @xmath53 , we set @xmath54 . if @xmath55 then @xmath56 is defined to be @xmath57 .
if @xmath34 $ ] then @xmath58 is the ideal generated by @xmath56 for all @xmath59 .
it was shown in @xcite that @xmath36 consists of the collection of @xmath60 for which @xmath58 contains no monomials .
the essence of this proof was the following : @xcite [ lift ] if @xmath61 and @xmath58 contains no monomial then @xmath62 is nonempty and any point @xmath63 of this variety can be lifted to a point @xmath64 with the leading term of @xmath65 equal to @xmath66 .
we now prove a similar criterion to characterize the points in @xmath40 .
[ poschar ] a point @xmath67 lies in @xmath40 if and only if @xmath58 does not contain any nonzero polynomials in @xmath68 $ ] . in order to prove this proposition , we will need the following result of @xcite , which relies heavily on a result of @xcite .
@xcite[einsedler ] an ideal @xmath39 of @xmath69 $ ] contains a nonzero element of @xmath70 $ ] if and only if @xmath71 for all @xmath72 .
we are now ready to prove proposition [ poschar ] .
define @xmath73 to be the image of @xmath74 under @xmath37 .
let @xmath75 denote the subset of @xmath76 consisting of those @xmath60 for which @xmath58 contains no polynomials with all positive terms . by definition
, @xmath40 is the closure of @xmath77 in @xmath76 .
we want to show that the closure of @xmath77 is @xmath75 .
it is obvious that @xmath77 lies in @xmath75 .
@xmath75 is closed , as the property that @xmath56 has only positive terms is open as @xmath60 varies .
thus , the closure of @xmath77 lies in @xmath75 .
conversely , suppose that @xmath78
. then , by proposition [ einsedler ] , for some @xmath72 , @xmath79 . for @xmath80
sufficently small , we have @xmath81 . therefore we can find a sequence @xmath82 , @xmath83 , approaching @xmath60 with @xmath84 .
as @xmath60 varies , @xmath58 takes on only finitely many values , and the subsets of @xmath76 on which @xmath58 takes a specific value form the relative interiors of the faces of a complete rational complex known as the grbner complex ( see @xcite ) .
these complexes are actually fans when @xmath39 is defined over @xmath23 ( @xcite ) .
therefore , we may perturb each @xmath85 , while preserving @xmath86 , in order to assume that the @xmath87 and we still have @xmath88 . then , by proposition [ lift ] , each @xmath89 , so @xmath60 is in the closure of @xmath77 as desired .
[ fancorollary ] @xmath36 and @xmath40 are closed subcomplexes of the grbner complex . in particular , they are polyhedral complexes . if @xmath39 is defined over @xmath23 , then @xmath36 and @xmath40 are closed subfans of the grbner fan
. one might wonder whether it would be better to modify the definition of @xmath30 to require that our power series lie in @xmath90 .
this definition , for example , is more similar to the appearance of the ring of formal powers series in @xcite .
one can show that in the case of the grassmanian , this difference is unimportant .
moreover , the definition used here has the advantage that it makes it easy to prove that the positive part of the tropicalization is a fan .
suppose @xmath91 and @xmath92 are varieties and we have a rational map @xmath93 taking @xmath94 . unfortunately , knowing @xmath95 for @xmath96 does not in general determine @xmath97 , so we do nt get a nice map @xmath98 . however , suppose that @xmath99 takes the positive points of @xmath35 surjectively onto the positive points of @xmath100 and suppose that @xmath101 is _ subtraction - free _ , that is , the formulas for the @xmath102 s are rational functions in the @xmath103 s whose numerators and denominators have positive coefficients .
define @xmath104 by replacing every @xmath105 in @xmath99 with a @xmath106 , every @xmath107 with a @xmath108 , every @xmath106 with a @xmath109 and every constant @xmath110 with @xmath111 .
[ blah ] suppose @xmath112 and @xmath113 are varieties .
let @xmath114 be a subtraction - free rational map taking @xmath35 to @xmath100 such that @xmath115 surjects onto @xmath116 .
then @xmath117 takes @xmath40 surjectively onto @xmath118 .
this follows immediately from the formulas @xmath119 and @xmath120 for @xmath121 and @xmath122 .
we now define the objects that we will study in this paper .
fix @xmath123 and @xmath17 , and let @xmath124 .
fix a polynomial ring @xmath125 in @xmath126 variables with coefficients in a commutative ring .
the _ plcker ideal _
@xmath127 is the homogeneous prime ideal in @xmath125 consisting of the algebraic relations ( called _ plcker relations _ ) among the @xmath128 minors of any @xmath129-matrix with entries in a commutative ring .
classically , the _ grassmannian _
@xmath15 is the projective variety in @xmath130 defined by the ideal @xmath127 of plcker relations .
we write @xmath131 for the variety in @xmath132 defined by the same equations .
similarly , we write @xmath133 for the real points of the grassmannian , @xmath134 for the real positive points , @xmath135 for those points of @xmath136 all of whose coordinates lie in @xmath30 and so on .
we write @xmath137 when we want to emphasize that we are using the field @xmath22 , and use @xmath15 when discussing results that hold with no essential modification for any field .
the _ totally positive grassmannian _ is the set @xmath134 .
an element of @xmath15 can be represented by a full rank @xmath129 matrix @xmath138 . if @xmath139 \choose k}$ ] we define the _ plcker coordinate _ @xmath140 to be the minor of @xmath138 corresponding to the columns of @xmath138 indexed by @xmath141 .
we identify the element of the grassmannian with the matrix @xmath138 and with its set of plcker coordinates ( which satisfy the plcker relations ) .
our primary object of study is the _ tropical positive grassmannian _
@xmath1 , which is a fan , by corollary [ fancorollary ] .
as in @xcite , this fan has an @xmath17-dimensional lineality space .
let @xmath142 denote the map from @xmath143 into @xmath144 which sends @xmath145 to the @xmath16-vector whose @xmath146-coordinate is @xmath147 .
we abuse notation by also using @xmath142 for the same map @xmath148 .
let @xmath149 denote the corresponding linear map which sends @xmath145 to the @xmath16-vector whose @xmath146-coordinate is @xmath150 .
the map @xmath149 is injective , and its image is the common lineality space of all cones in @xmath151 .
in this section we explain two equivalent ways to parameterize @xmath134 , as well as a way to parameterize @xmath152 .
the first method , due to postnikov @xcite , uses a certain directed graph @xmath153 with variables associated to each of its @xmath154 edges .
the second method is closely related to the first and uses the same graph , but this time variables are associated to each of its @xmath155 regions .
this has the advantage of giving a bijection between @xmath156 and @xmath134 .
finally , we use @xmath153 with variables labelling each of its @xmath157 inner regions in order to give a bijective parameterization of @xmath158 .
let @xmath153 be the directed graph which is obtained from a @xmath123 by @xmath159 grid , as shown in figure [ web49 ] .
it has @xmath123 incoming edges on the right and @xmath159 outgoing edges on the bottom , and the vertices attached to these edges are labelled clockwise from @xmath160 to @xmath17 .
we denote the set of @xmath154 edges by @xmath161 .
let us associate a formal variable @xmath162 with each edge @xmath163 , and if there is no ambiguity , we abbreviate the collection @xmath164 by @xmath121 .
if @xmath165 is a path on @xmath153 ( compatible with the directions of the edges ) , then we let @xmath166 denote @xmath167 . and
if @xmath125 is a set of paths on @xmath153 , then we let @xmath168 denote @xmath169 . as in @xcite , we define a @xmath129 matrix @xmath170 , whose entries @xmath171 are polynomials in the variables @xmath162 , by the following equation : @xmath172 where the sum is over all directed paths @xmath165 from vertex @xmath173 to vertex @xmath174 .
note that the @xmath128 submatrix of @xmath170 obtained by restricting to the first @xmath123 columns is the identity matrix .
in particular , @xmath175 is a full rank matrix and hence we can identify it with an element of @xmath15 .
also note that every element of the totally positive grassmannian @xmath134 has a unique matrix representative whose leftmost @xmath128 submatrix is the identity .
we shall see that as the @xmath164 vary over @xmath176 , the @xmath175 range over all of @xmath134 .
we now show it is possible to express the maximal minors ( plcker coordinates ) of @xmath175 as subtraction - free rational expressions in the @xmath162 , as shown in @xcite .
if @xmath177 \choose k}$ ] , then let @xmath178 denote the set @xmath179\backslash ( k\cap [ k ] ) \text { to } k\backslash ( [ k]\cap k ) \}.\ ] ] note that for @xmath180 $ ] , we consider the empty set to be a legitimate set of pairwise vertex - disjoint paths
. applied to @xmath153 , theorem 15.4 of @xcite implies the following .
[ plucker ] the plcker coordinates of @xmath175 are given by @xmath181 we give a brief proof of this result : the main idea is to use the well - known gessel - viennot trick @xcite .
first note that @xmath171 has a combinatorial interpretation : it is a generating function keeping track of paths from @xmath173 to @xmath174 .
thus , the determinant of a @xmath128 submatrix of @xmath175 corresponding to the column set @xmath141 also has a combinatorial interpretation : it is a generating function for all sets of paths from @xmath182\backslash ( k\cap [ k])$ ] to @xmath183\cap k)$ ] , with the sign of each term keeping track of the number of crossings in the corresponding path set .
what we need to show is that this is equal to the sum of the contributions from path sets which are pairwise vertex - disjoint . to see this ,
consider a path set which does have an intersection .
look at its lexicographically last intersection , and compare this path set to the one obtained from it by switching the two path tails starting at that point of intersection .
these two path sets get different signs , but have equal weights , and hence they cancel each other out .
proposition [ plucker ] allows us to define a map @xmath184 as follows .
let @xmath177 \choose k}$ ] , and define @xmath185 by @xmath186 clearly if we substitute positive values for each @xmath162 , then @xmath187 will be positive .
we now define @xmath188 by @xmath189\choose k}}.\ ] ] in other words , @xmath188 is the map which sends a collection of positive real numbers @xmath164 to the element of @xmath15 with plcker coordinates @xmath187 ( which is identified with the matrix @xmath175 ) . by theorem 19.1 of @xcite , the map @xmath188 is actually surjective : any point in @xmath134 can be represented as @xmath175 for some positive choices of @xmath164 . in summary , we have the following result ( which will also be a consequence of our theorem [ bij ] ) .
[ parameter ] the map @xmath190 is a surjection onto @xmath134 .
unfortunately , the method we have just described uses @xmath154 variables to parameterize a space of dimension @xmath155 .
we will now explain how to do a substitution of variables which will reduce the number of variables to @xmath155 .
we define an _ inner region _ of @xmath153 to be a bounded component of the complement of @xmath153 ( viewed as a subset of @xmath191 ) .
and we define an _ outer region _ of @xmath153 to be one of the extra inner regions we would obtain if we were to connect vertices @xmath173 and @xmath192 by a straight line , for @xmath173 from @xmath160 to @xmath193 .
a _ region _ is an inner or outer region .
note that there are @xmath155 regions , which we denote by @xmath194 , and there are @xmath157 inner regions .
let us label each region @xmath195 with a new variable @xmath196 , which we define to be the product of its counterclockwise edge variables divided by the product of its clockwise edge variables .
for example , the new variables @xmath197 shown in figure [ substitute ] would be defined by @xmath198 it is easy to check that for a path @xmath165 on @xmath153 , @xmath166 is equal to the product of the variables attached to all regions below @xmath165 . since @xmath168 and @xmath175 were defined in terms of the @xmath166 s , we can redefine these expressions in terms of the @xmath155 region variables .
proposition [ parameter ] still holds , but our map is now a map @xmath199 from @xmath156 onto @xmath134 , taking the region variables @xmath200 to the element of @xmath134 represented by @xmath170 .
since we are now parameterizing a space of dimension @xmath155 with @xmath155 variables , we should have a bijection .
we shall prove that this is so by constructing the inverse map .
[ bij ] the map @xmath201 , which maps @xmath202 to the grassmannian element represented by @xmath175 , is a bijection .
before we prove this theorem , we need a lemma about matrices and their minors .
we use a very slight generalization of a lemma which appeared in @xcite . for completeness
, we include the proof of this lemma .
first we must define some terminology .
let @xmath203 be a @xmath129 matrix .
let @xmath204 denote the minor of @xmath203 which uses row set @xmath39 and column set @xmath205 .
we say that @xmath204 is _ solid _ if @xmath39 and @xmath205 consist of several consecutive indices ; if furthermore @xmath206 contains @xmath160 , we say that @xmath204 is _
initial_. thus , an initial minor is a solid minor which includes either the first column or the first row .
@xcite[minors ] a matrix @xmath203 is uniquely determined by its initial minors provided that all these minors are nonzero .
let us show that each matrix entry @xmath207 of @xmath203 is uniquely determined by the initial minors .
if @xmath208 or @xmath209 , there is nothing to prove , since @xmath207 is an initial minor .
assume that @xmath210 .
let @xmath211 be the initial minor whose last row is @xmath173 and last column is @xmath174 , and let @xmath212 be the initial minor obtained from @xmath211 by deleting this row and column .
then @xmath213 , where @xmath214 is a polynomial in the matrix entries @xmath215 with @xmath216 and @xmath217 . using induction on @xmath218
, we can assume that each @xmath215 that occurs in @xmath214 is uniquely determined by the initial minors , so the same is true of @xmath219 .
we now define a _ reflected initial _ minor to be a solid minor @xmath204 such that @xmath39 contains @xmath123 or @xmath205 contains @xmath160 .
thus , a reflected initial minor is a solid minor which includes either the first column or the last row .
a trivial corollary of lemma [ minors ] is the following .
[ silly ] a matrix @xmath203 is uniquely determined by its reflected initial minors provided that all these minors are nonzero .
now we are ready to prove theorem [ bij ] .
to prove the theorem , we will construct an explicit inverse map @xmath220 .
the first step is to prove that @xmath221 .
let us index the regions in @xmath153 by ordered pairs @xmath222 as follows .
given a region , we choose @xmath173 to be the label of the horizontal wire which forms the upper boundary of the region , and choose @xmath174 to be the label of the vertical wire which forms the left boundary of the region .
now we define a map @xmath141 from the set of regions to @xmath223 \choose k}$ ] by @xmath224 if @xmath222 is not a region of @xmath153 , then we define @xmath225 .
let @xmath138 be a @xmath129 matrix whose initial @xmath226 minor is the identity .
we define @xmath227 by @xmath228 note that by convention , we define @xmath229 to be @xmath160 .
see figure [ inverse ] for the definition of @xmath230 in the case of @xmath231 .
note that for brevity , we have omitted the @xmath138 s from each term .
we claim that if @xmath232 , then @xmath233 . to prove this , we note that the variable in region @xmath222 can be expressed in terms of vertex - disjoint paths as follows .
first observe that if @xmath234 then there is a unique set of pairwise vertex - disjoint paths from @xmath182 \setminus ( [ k ] \cap k(i , j))$ ] to @xmath235 \cap k(i , j))$ ] . if one examines the terms in ( [ preimage ] ) and draws in the six sets of pairwise vertex - disjoint paths on @xmath153 ( say the three from the numerator in red and the three from the denominator in blue ) then it is clear that every region in @xmath153 lies underneath an equal number of red and blue paths except the region @xmath222 , which lies underneath only one red path .
thus , by definition of the maps @xmath236 , it follows that @xmath237 to complete the proof , it remains to show that @xmath230 is injective .
this will complete the proof because we know that @xmath238 , and @xmath230 injective then implies that @xmath239 .
choose an element of @xmath134 , which we identify with its unique matrix representative @xmath138 whose leftmost @xmath128 minor is the identity .
let @xmath240 denote the set of rational expressions which appear in the right - hand side of ( [ preimage ] ) for all regions @xmath222 in @xmath153 .
let @xmath241 denote the set of all individual plucker coordinates which appear in @xmath240 .
we prove that @xmath230 is injective in two steps .
first we show that the values of the expressions in @xmath240 uniquely determine the values of the plcker coordinates in @xmath241 .
next we show that the values of the plcker coordinates in @xmath241 uniquely determine the matrix @xmath138 .
the first step is clear by inspection .
we illustrate the proof in the case of @xmath231 . by the choice of @xmath138 , @xmath242 looking at the rational expressions in figure [ inverse ]
, we see that knowing the value @xmath243 determines @xmath244 ; the value @xmath244 together with the value @xmath245 determines @xmath246 ; and similarly for @xmath247 .
next , these values together with the value @xmath248 determines @xmath249 , and so on . for
the second step of the proof , let @xmath250 denote the @xmath251 matrix obtained from @xmath138 by removing the leftmost @xmath128 identity matrix .
note that the values of the plcker coordinates @xmath252 ( which are all elements of @xmath241 ) determine the values of all of the reflected initial minors of @xmath250 .
( each such plcker coordinate is equal to one of the reflected initial minors , up to sign . )
thus , by corollary [ silly ] , they uniquely determine the matrix @xmath250 and hence @xmath138 .
this completes the proof of theorem [ bij ] .
now let us parameterize @xmath253 .
we shall show that we can do this by using variables corresponding to only the @xmath157 inner regions of @xmath153 .
first recall that the @xmath17-dimensional torus acts on @xmath134 by scaling columns of a matrix representative for @xmath254 .
( although the torus has dimension @xmath17 , this is actually just an @xmath255-dimensional action as the scalars act trivially . )
namely , @xmath256 if @xmath254 then we let @xmath257 denote the torus orbit of @xmath138 under this action .
note that if @xmath258 , then @xmath259 .
we will now determine the corresponding torus action on @xmath156 such that the above bijection commutes with the actions .
if @xmath260 is an internal region then @xmath196 is a ratio of plcker coordinates with the same indices appearing on the top and bottom , so @xmath196 is not modified by the torus action .
a simple computation shows that the torus acts transitively on the values of the outer region variables .
thus , taking the quotient by @xmath261 on the right hand side of the equation corresponds to forgetting the outer variables on the left .
define a map @xmath262 by lifting a point @xmath263 to any arbitrarily chosen point @xmath264 and then mapping @xmath265 to @xmath266 .
we have just proven : [ secondbijection ] the map @xmath267 is a bijection .
in this section we will construct a lower - dimensional fan associated to the tropical positive grassmannian @xmath1 . by methods precisely analogous to those above
, we can prove an analogue of theorem [ secondbijection ] for the field of puiseux series .
the map @xmath268 is a bijection .
this theorem allows us to compute @xmath269 by applying the valuation map to the image of @xmath270 . by proposition [ blah ]
we can tropicalize the map @xmath270 , obtaining the following surjective map .
@xmath271 the map @xmath272 is the map we get by replacing multiplication with addition and addition with minimum in the definition of @xmath270 .
explicitly , it is defined as follows .
let @xmath139 \choose k}$ ] , and let inner region variables take on values @xmath200 in @xmath23 .
outer region variables are chosen arbitrarily .
if @xmath165 is a path on @xmath153 then we let @xmath273 denote the sum of all variables which label regions below @xmath165 . similarly ,
if @xmath125 is a set of paths , then let @xmath274 denote @xmath275 .
now define @xmath276 by @xmath277 the map @xmath272 is the map @xmath278 given by @xmath279 the fan @xmath280 is the complete fan in @xmath281 whose maximal cones are the domains of linearity of the piecewise linear map @xmath272 . because @xmath272 surjects onto @xmath269 , the fan @xmath280 reflects the combinatorial structure of the fan @xmath269 , which differs from @xmath1 only through modding out by the linearity space .
however , @xmath280 is much easier to work with , as it lives in @xmath157-dimensional space as opposed to @xmath16-dimensional space .
since the maps @xmath282 are piecewise linear functions , to each one we can associate a fan @xmath283 whose maximal cones are the domains of linearity for @xmath282 .
it is clear that the fan @xmath280 is the simultaneous refinement of all of the fans @xmath283 .
from now on , by abuse of notation , we will refer to @xmath269 as @xmath1 .
in this section we will describe the fan @xmath284 associated to @xmath2 .
we show that this fan is exactly the stanley - pitman fan @xmath285 , which appeared in the work of stanley and pitman in @xcite .
in particular , the face poset of @xmath284 , with a top element @xmath286 adjoined , is isomorphic to the face lattice of the normal fan of the associahedron , a polytope whose vertices correspond to triangulations of the convex @xmath17-gon .
( in the language of @xcite , this is the associahedron of type @xmath3 . )
let us first do the example of @xmath287 .
we use the web diagram @xmath288 , as shown in figure [ web25 ] .
the maps @xmath282 are given by : @xmath289 each map @xmath290 is piecewise linear and so gives rise to the complete fan @xmath283 .
for example , the map @xmath291 is linear on the region @xmath292 , where it is the function @xmath293 , and on the region @xmath294 , where it is the function @xmath295 .
thus , @xmath296 is simply the subdivision of the real plane into the regions @xmath297 and @xmath298 .
the three nontrivial fans that we get from the maps @xmath299 are shown in figure [ fan25 ] . in each picture , the maximal cones of each fan are separated by solid lines .
@xmath300 , which is the simultaneous refinement of the three nontrivial fans , is shown in figure [ bigfan25 ] . in @xcite
, it was shown that maximal cones of the fan @xmath301 correspond to trivalent trees on @xmath17 labelled leaves .
it turns out that maximal cones of the fan @xmath302 correspond to trivalent planar trees on @xmath17 labelled leaves , as is illustrated in figure [ bigfan25 ] .
we will now describe the fan that appeared in @xcite , but first , we must review some notions about trees .
plane binary tree _ is a rooted tree such that each vertex has either two children designated as left and right , or none at all ; and an _ internal vertex _ of a binary tree is a vertex which is not a leaf .
trivalent planar tree _ is an ( unrooted ) tree such that every vertex has degree three , and such that the leaves are labelled in a clockwise fashion .
it is known that both plane binary trees with @xmath193 leaves , and trivalent planar trees with @xmath17 labelled leaves , are counted by the catalan number @xmath303 .
there is a simple bijection between such trivalent planar trees and plane binary trees : if t a trivalent planar tree , then simply contract the edge whose leaf is labelled @xmath160 , and make this the root .
this bijection is illustrated in figure [ bigfan25 ] .
let us now define the stanley - pitman fan @xmath285 in @xmath304 .
( note that we use different indices than are used in @xcite ) .
the maximal cones of @xmath285 are indexed by plane binary trees with @xmath193 leaves , in the following manner .
let @xmath77 be a plane binary tree with @xmath193 leaves .
label the internal vertices of @xmath77 with the numbers @xmath305 in the order of the first time we drop down to them from a child when doing a depth - first search from left to right starting at the root .
( see figure [ bigfan25 ] for examples . )
let @xmath306 denote the coordinates in @xmath304 .
if the internal vertex @xmath173 of @xmath77 is the parent of vertex @xmath174 , and @xmath307 , then associate with the pair @xmath222 the inequality @xmath308 while if @xmath309 then associate with @xmath222 the inequality @xmath310 these @xmath311 inequalities define a simplicial cone @xmath312 in @xmath304 .
the result proved in @xcite is the following .
[ @xcite ] the @xmath313 cones @xmath312 , as @xmath77 ranges over all plane binary trees with @xmath193 leaves , form the chambers of a complete fan in @xmath304 . moreover , the face poset of @xmath285 , with a top element @xmath286 adjoined , is dual to the face lattice of the associahedron which parameterizes triangulations of the convex @xmath17-gon .
the key step in proving that our fan @xmath284 is equal to the stanley - pitman fan @xmath285 is the following lemma , also proved in @xcite .
let @xmath314 .
let @xmath315 consist of all plane binary trees with @xmath193 leaves and root @xmath173 . then @xmath316 .
the fan @xmath284 is equal to the fan @xmath285 .
first let us describe the fan @xmath284 as explicitly as possible . note that if we label the regions of @xmath317 with the variables @xmath306 from right to left , then all of the maps @xmath282 are of the form @xmath318 where @xmath319 .
since this map has the same domains of linearity as the map @xmath320 we can work with the maps @xmath321 instead .
let @xmath322 be the fan whose cones are the domains of linearity of @xmath321 .
then @xmath284 is the simultaneous refinement of all fans @xmath322 where @xmath319 .
now note that the previous lemma actually gives us an algorithm for determining which cone @xmath312 a generic point @xmath323 lies in .
namely , if we are given such a point , compute the partial sums of @xmath324 , for @xmath325 .
choose @xmath173 such that @xmath324 is the minimum of these sums .
( if @xmath208 , the sum is @xmath326 . )
then the root of the tree @xmath77 is @xmath173 .
the left subtree of @xmath77 consists of vertices @xmath327 , and the right subtree of @xmath77 consists of vertices @xmath328 .
we now compute @xmath329 and @xmath330 in order to compute the roots of these two subtrees and so forth .
now take a point @xmath331 in a cone @xmath332 of @xmath284 .
this means that the point is in a domain of linearity for all of the piecewise linear functions @xmath333 where @xmath319 , and we take @xmath334 to be @xmath326 . in other words , for each @xmath173 and @xmath174 , there is a unique @xmath123 such that @xmath335 . in particular , we can reconstruct the tree @xmath77 such that @xmath336 , and _ every _ point @xmath337 belongs to this same cone @xmath312 .
finally , we can show by induction that @xmath338 .
( we need to show that all of the functions @xmath321 are actually linear on @xmath312 . )
this shows that each cone @xmath332 in @xmath284 is actually equal to a cone @xmath312 in @xmath285 , and conversely .
in connection with their work on cluster algebras , fomin and zelevinsky @xcite recently introduced certain polytopes called _ generalized associahedra _ corresponding to each dynkin type , of which the usual associahedron is the type @xmath138 example .
when we computed @xmath339 , the fan associated with @xmath4 , we found that it was closely related to the normal fan of the type @xmath6 associahedron , in a way which we will now make precise .
( we defer the explanation of our computations to the end of this section . )
the @xmath99-vector of @xmath339 is @xmath340 .
the rays of @xmath339 are listed in table [ 16 ] , along with the inequalities defining the polytope that @xmath339 is normal to .
.rays and inequalities for @xmath339 [ cols= " < , < " , ]
cluster algebras are commutative algebras endowed with a certain combinatorial structure , introduced in @xcite and expected to be relevant in studying total positivity and homogeneous spaces , such as grassmannians
. we will not attempt to give a precise definition of a cluster algebra here , but will rather describe their key properties .
slightly varying definitions can be found in @xcite , @xcite and @xcite ; we follow @xcite but do not believe these small variations are important .
a cluster algebra is an algebra @xmath341 over a field @xmath123 , which in our examples can be thought of as @xmath23 .
additionally , a cluster algebra carries two subsets @xmath332 and @xmath342 , known as the coefficient variables and the cluster variables .
@xmath332 is finite , but @xmath343 may be finite or infinite .
if @xmath343 is finite , @xmath341 is known as a cluster algebra of finite type .
there is also a nonnegative integer @xmath260 associated to a cluster algebra and known as the rank of the algebra .
there is a pure @xmath344-dimensional simplicial complex called the _
cluster complex _ whose vertices are the elements of @xmath343 and whose maximal simplices are called _
clusters_. we will denote the cluster complex by @xmath345 .
if @xmath346 and @xmath347 is a cluster containing @xmath121 , there is always a unique cluster @xmath348 with @xmath349 .
let @xmath350 .
then there is a relation @xmath351 where @xmath14 is a binomial in the variables of @xmath352 . for any @xmath346 and any cluster @xmath211
, @xmath121 is a subtraction - free rational expression in the members of @xmath353 and is also a laurent polynomial in the members of @xmath353 .
conjecturally , this laurent polynomial has non - negative coefficients .
note that this conjecture does not follow from the preceding sentence : @xmath354 is a subtraction - free expression in @xmath121 and @xmath355 , and a polynomial in @xmath121 and @xmath355 , but it is not a polynomial with positive coefficients .
it was demonstrated in @xcite that the coordinate rings of grassmannians have natural cluster algebra structures .
usually these cluster algebras are of infinite type , making them hard to work with in practice , but in the cases of @xmath8 , @xmath356 for @xmath357 and their duals , we get cluster algebras of finite type . in the case of @xmath8 ,
the coefficient set @xmath332 is @xmath358 and the set of cluster variables @xmath343 is @xmath359 .
( note that these @xmath211 s are plcker coordinates and not simplices . )
label the vertices of an @xmath17-gon in clockwise order with the indices @xmath360 and associate to each member of @xmath361 the corresponding chord of the @xmath17-gon .
the clusters of @xmath8 correspond to the collections of chords which triangulate the @xmath17-gon .
thus , @xmath345 in this example is ( as an abstract simplicial complex ) isomorphic to the dual of the associahedron .
since we have shown that the fan of @xmath2 is combinatorially equivalent to the normal fan of the associahedron , it follows that @xmath2 is ( combinatorially ) the cone on @xmath345 . in the case of @xmath9 , the coefficient set
@xmath332 is equal to @xmath362 .
@xmath343 contains the other @xmath363 plcker coordinates , but it also contains two unexpected elements : @xmath364 and @xmath365 . by definition ,
all plcker coordinates are positive on the totally positive grassmannian , so by the results above on subtraction - free rational expressions , these new coordinates are positive on the totally positive grassmannian as well .
the new coordinates turn out to be laurent polynomials with positive coefficients in the region variables of section [ parameterize ] .
thus , we can tropicalize these laurent polynomials and associate a fan to each of them . when we refine @xmath339 by these fans ,
the refinement subdivides the two bipyramids and yields precisely the normal fan to the @xmath6 associahedron , which is again the cone over @xmath345 . in the case of @xmath10
, @xmath332 again consists of @xmath366 where indices are modulo @xmath367 .
@xmath343 contains all of the other plcker variables and the pullbacks to @xmath10 of the two new cluster variables of @xmath9 , along with the @xmath367 rational coordinate projections @xmath368 .
thus , @xmath343 contains @xmath369 plcker variables and @xmath363 other variables . as in the case of @xmath9 ,
the 14 new variables are laurent polynomials with positive coefficients in the region variables of section [ parameterize ] , so to each one we can associate a corresponding fan .
when we refine @xmath370 by these 14 new fans , we get a fan combinatorially equivalent to the fan normal to the @xmath7 associahedron .
we can describe what we have seen in each of these grassmannian examples in terms of the general language of cluster algebras as follows : embed @xmath371 in affine space by the variables @xmath372 .
then @xmath373 is a fan with lineality space of dimension @xmath374 . after taking the quotient by this lineality space
, we get a simplical fan abstractly isomorphic to the cone over @xmath345 .
this observation does not quite hold for an arbitrary cluster algebra of finite type .
for example , if we take the cluster algebra of @xmath375 and set all coefficient variables equal to @xmath160 , we get a different cluster algebra which is still of type @xmath376 .
however , when we compute the positive part of the corresponding tropical variety , we get a fan whose lineality space has dimension @xmath160 , not @xmath326 as the above would predict .
our fan is a cone over a hexagon cross a @xmath160-dimensional lineality space , which is a coarsening of the fan normal to the type @xmath376 associahedron .
based on this and other small examples , it seems that in order to see the entire cone over @xmath345 , one needs to use enough " coefficients .
let @xmath341 be a cluster algebra of finite type over @xmath23 and @xmath345 its associated cluster complex .
if the lineality space of @xmath373 has dimension @xmath374 then @xmath373 modulo its lineality space is a simplicial fan abstractly isomorphic to the cone over @xmath345 .
if the condition on the lineality space does not hold , the resulting fan is a coarsening of the cone over @xmath345 .
* remark : * the condition on the lineality space can be restated without mentioning tropicalizations . consider the torus @xmath377 acting on the affine space @xmath378 and let @xmath12 be the subgroup taking @xmath371 to itself .
we want to require that @xmath379 .
* remark : * in the notation of @xcite and @xcite , the condition on the dimension of the lineality space is equivalent to requiring that the matrix @xmath380 be of full rank .
we thank andrei zelevinsky for pointing this out to us .
note how surprising this conjecture is in light of how the two complexes are computed .
the fan described in the conjecture is computed as the refinement of a number of fans , indexed by the vertices of @xmath345 . that the rays of this fan , which arise as the intersections of many hypersurfaces , should again be in bijection with the vertices of @xmath345
is quite unexpected .
we expect an analogous statement to hold for infinite type cluster algebras .
we are grateful to bernd sturmfels for suggesting this problem to us , and for advice throughout our work . in addition , we thank komei fukuda and michael joswig for their help in performing the polyhedral computations and joshua scott , richard stanley , and andrei zelevinsky for their useful comments .
j. richter - gebert , b. sturmfels , t. theobald , first steps in tropical geometry , 29 pages , ` http://www.arxiv.org/math.ag/0306366 ` , to appear in `` idempotent mathematics and mathematical physics '' , proceedings vienna 2003 , ( editors g. litvinov and v. maslov ) . | tropical algebraic geometry is the geometry of the tropical semiring @xmath0 .
the theory of total positivity is a natural generalization of the study of matrices with all minors positive . in this paper
we introduce the totally positive part of the tropicalization of an arbitrary affine variety , an object which has the structure of a polyhedral fan .
we then investigate the case of the grassmannian , denoting the resulting fan @xmath1 .
we show that @xmath2 is the stanley - pitman fan , which is combinatorially the fan dual to the ( type @xmath3 ) associahedron , and that @xmath4 and @xmath5 are closely related to the fans dual to the types @xmath6 and @xmath7 associahedra .
these results are strikingly reminiscent of the results of fomin and zelevinsky , and scott , who showed that the grassmannian has a natural cluster algebra structure which is of types @xmath3 , @xmath6 , and @xmath7 for @xmath8 , @xmath9 , and @xmath10 .
we suggest a general conjecture about the positive part of the tropicalization of a cluster algebra . |
the discovery of superconductivity at elevated temperatures in the copper - oxide materials a quarter of a century ago sparked an intense effort to understand the mechanism of the superconductivity in these compounds . despite the wealth of information from the accumulation of hundreds of thousands of scientific papers , there is still no consensus on the pairing mechanism .
while superconductivity was originally observed in the single - layer la@xmath0ba@xmath1cuo@xmath2 system,@xcite single crystals of this material proved difficult to grow and attention quickly shifted to the related la@xmath0sr@xmath1cuo@xmath2 materials where large single crystals were available .
further investigations into systems with more than two copper - oxygen layers in the unit cell led to rapid increases in the critical temperature ( @xmath9 ) , with the current maximum @xmath16 k observed at ambient pressure in the hgba@xmath17ca@xmath17cu@xmath18o@xmath19 system.@xcite the physical properties of this class of correlated electron materials and the nature of the superconductivity have been described in numerous review articles.@xcite while most high - temperature superconductors display a doping - dependent dome - shaped superconducting phase boundary,@xcite an anomalous weakening of the superconductivity is observed at a doping of @xmath20 per copper atom in the related la@xmath0sr@xmath1cuo@xmath2 system,@xcite and more recently in the bi@xmath17sr@xmath17cacu@xmath17o@xmath19 and yba@xmath17cu@xmath18o@xmath21 materials.@xcite this so - called `` @xmath22 anomaly '' increases dramatically in la@xmath0ba@xmath1cuo@xmath2 for @xmath15 and results in the almost total destruction of the bulk three - dimensional ( 3d ) superconductivity.@xcite there is also a structural transition over much of the phase diagram from a low - temperature orthorhombic ( lto ) to a low - temperature tetragonal ( ltt ) symmetry@xcite that is not observed in pure la@xmath0sr@xmath1cuo@xmath2 crystals ; the lto@xmath23ltt transition is also accompanied by the formation of an antiferromagnetically ( afm ) ordered state.@xcite recently , large single crystals of la@xmath0ba@xmath1cuo@xmath2 have become available that have allowed a much more detailed investigation of the electronic@xcite and structural@xcite properties of the 1/8 phase . with the onset of the ltt phase both long - range charge - stripe order , followed by spin - stripe order at a slightly lower temperature , are observed.@xcite the question of what role stripes play in the formation of a superconducting state in the cuprate materials is the subject of some debate;@xcite however , in this case it appears that the formation of static charge - stripe order has the effect of electronically decoupling the copper - oxygen planes , significantly increasing the resistance along the _ c _ axis , and frustrating the formation of a bulk superconducting ground state.@xcite
the 1/8 phase exhibits a pseudogap@xcite in the lto phase,@xcite similar to what is observed in the underdoped cuprates.@xcite the remaining fermi arc ( or pocket ) in the nodal region becomes gapped in the ltt phase close to the spin - ordering temperature .
the gap in the nodal region is momentum - dependent with a _
d_-wave form , @xmath24 , where @xmath25 is the gap maximum and @xmath26 is a fermi surface angle , similar to the form of the superconducting gap observed in other cuprate superconductors.@xcite it has been proposed that this is a superconducting gap , but that the pairs lack the necessary phase coherence required to form a condensate.@xcite the optical properties of a material can provide important information about the nature of the lattice vibrations and the electronic transport , including the formation of charge and superconducting energy gaps . the role of charge - stripe and spin - stripe ordering in the 1/8 phase is of particular interest .
previous optical investigations of stripe correlations in an analogous phase of la@xmath27nd@xmath28sr@xmath29cuo@xmath2,@xcite and in la@xmath0sr@xmath1cuo@xmath2,@xcite have resulted in some disagreement over the association of certain infrared spectral features with charge order.@xcite in this work we determine the temperature dependence of the optical properties of single crystals of la@xmath0ba@xmath1cuo@xmath2 over a wide frequency range for three different dopings , @xmath3 , 0.125 and 0.145 with @xmath9 s of approximately 32 , 2.4 and 24 k , respectively , for light polarized in the conducting _ a - b _ planes , and along the poorly conducting _ c _ axis . the optical conductivity in the _ a - b _
planes for all three materials show a drude - like metallic behavior for the free - carriers at room temperature with a scattering rate that decreases with temperature ; for the superconducting @xmath3 and 0.145 samples there is a characteristic suppression of the low - frequency conductivity below @xmath9 , allowing the superfluid density @xmath10 to be determined from this so - called missing spectral weight.@xcite in contrast , in a refinement of earlier optical work,@xcite the 1/8 phase shows a dramatic narrowing of the free - carrier response below @xmath30 accompanied by the transfer of some spectral weight to both high and low frequencies .
this transfer of spectral weight is consistent with the formation of a momentum - dependent gap due to the formation of charge - stripe and spin - stripe order resulting in a nodal metal .
the loss of low - frequency spectral weight also mimics the missing spectral weight observed in the superconducting materials , and it is possible that the transfer of spectral weight is associated with the two - dimensional ( 2d ) fluctuating superconducting state prior to the onset of bulk 3d superconductivity.@xcite the two dopings for which superconductivity is observed fall on the universal scaling line , @xmath31 ( where @xmath32 is measured just above @xmath9),@xcite which is consistent with the observation that there is substantial dissipation in the normal state of these materials .
the optical properties along the poorly - conducting _ c _ axis are dominated by the infrared - active lattice modes superimposed on a weak electronic background . in the superconducting @xmath3 and 0.145
samples , sharp plasma edges in the low - frequency reflectance develop below @xmath9 due to josephson coupling between the planes . no josephson plasma edge , or any indication of metallic behavior ,
is observed for the 1/8 phase down to the lowest measured temperature , despite the dramatic reduction of the _ c_-axis resistivity below @xmath30 .
ba@xmath33cuo@xmath2 single crystals ( reproduced from ref . ) .
onset temperatures : @xmath9 for bulk superconductivity ( sc ) , @xmath34 for charge stripe order ( co ) , @xmath30 for spin stripe order ( so ) , and @xmath35 for the low - temperature structural phases ltt and ltlo .
the unit cell of la@xmath0ba@xmath33cuo@xmath2 is shown in the ( b ) high - temperature tetragonal ( htt ) , ( c ) low - temperature orthorhombic ( lto ) , and ( d ) low - temperature tetragonal ( ltt ) phases . ]
single crystals of la@xmath0ba@xmath1cuo@xmath2 were grown using the traveling - solvent floating - zone method for a series of three nominal ba concentrations , @xmath36 , 0.125 and 0.155 .
the @xmath36 and 0.155 compositions are superconducting with transitions at @xmath37 and 24 k ; the superconductivity in the @xmath5 sample is strongly suppressed with @xmath38 k ( all values for @xmath9 were determined by magnetic susceptibility ) . the @xmath9 for the nominal @xmath39 crystal studied here ( and in ref . )
is substantially lower than that of the piece ( from the same growth ) characterized in ref . .
interpreting the discrepancy in @xmath9 as a difference in ba concentration , we estimate that the present sample corresponds to @xmath40 , and we refer to it as @xmath7 throughout the rest of the paper .
( ref . ) . in the normal state
@xmath41 decreases linearly with decreasing temperature , while @xmath42 is observed to increase until just above @xmath9 , below which both decrease dramatically and a weak anomaly is observed in @xmath41 .
( b ) the _ ab_-plane and _ c_-axis resistivity for the @xmath5 material ( refs . ) .
weak anomalies are observed in @xmath41 and @xmath42 at @xmath34 ; @xmath41 continues to decrease until @xmath30 , below which it falls dramatically before once again adopting a linear temperature dependence .
along the _ c _ axis , @xmath42 continues to increase below @xmath34 until just below @xmath30 where it begins to decrease quickly until @xmath43 k , below which it decreases less rapidly .
note different scales for @xmath41 . ]
the phase diagram of la@xmath0ba@xmath1cuo@xmath2 shown in fig .
[ fig : phase](a ) is considerably more sophisticated than that of the related la@xmath0sr@xmath1cuo@xmath2 material.@xcite above room temperature , this material is in the @xmath44 tetragonal ( htt ) phase [ fig .
[ fig : phase](b ) ] .
with decreasing temperature it undergoes an orthorhombic distortion into a @xmath45 ( lto ) phase [ fig .
[ fig : phase](c ) ] which may be described as a tilting of the cuo@xmath46 octahedra towards the la ions leading to a roughly @xmath47 larger basal plane rotated by @xmath48 .
the htt@xmath23lto transition occurs at @xmath49 , 235 and 190 k for the @xmath3 , 0.125 and 0.145 materials , respectively.@xcite at dopings close to @xmath5 the tilt of the octahedra moves between the la ions and is accompanied by a weak in - plane distortion resulting in a low - temperature tetragonal ( ltt ) phase @xmath50 [ fig .
[ fig : phase](d ) ] ; this phase is associated with the development of charge - stripe and spin - stripe order.@xcite in the @xmath3 crystal close to @xmath51 the lto structure undergoes a weakening of the orthorhombic strain , leading to a `` low - temperature less orthorhombic '' ( ltlo ) @xmath52 phase at very low temperature.@xcite in the @xmath7 material the lto@xmath23ltt transition occurs at @xmath53 k , but charge order does not develop until @xmath54 k and the spin order is likely very weak and is not observed ; at low temperature the structure is thought to be a mixture of ltt@xmath55ltlo .
ba@xmath56cuo@xmath2 revealing a number of striations and cleavage steps ; the total area shown is roughly @xmath57 . ]
the _ ab_-plane and _ c_-axis resistivity of the @xmath3 material is shown close to @xmath37 k in fig .
[ fig : trans](a ) ( ref . ) . in the normal state @xmath41 displays a characteristic linear decrease with temperature , until just above @xmath9 where a weak anomaly is observed that is associated with the onset of the structural transition previously described.@xcite in contrast , in the normal state @xmath42 increases with decreasing temperature , leading to a resistivity anisotropy @xmath58 that continues to increase until just above @xmath9 .
ab_-plane and _ c _ axis resistivity of the @xmath5 material is shown in fig . [
fig : trans](b ) ( refs . ) . in the normal
state @xmath41 decreases with temperature in an approximately linear fashion until a small jump is observed at the charge - ordering temperature ( @xmath59 k ) , followed by a dramatic decrease at the spin - ordering temperature ( @xmath60 k ) ; this latter transition is extremely sensitive to the applied magnetic field.@xcite below @xmath30 a 2d fluctuating superconducting state is thought to form and the resistivity continues to decrease exponentially ; at the same time a strong diamagnetic response is observed.@xcite this material is conjectured to undergo a berezinskii - kosterlitz - thouless transition@xcite at @xmath61 k , below this temperature the resistivity is effectively zero.@xcite a true bulk ( 3d ) superconducting transition is finally achieved at @xmath6 k. in the normal state @xmath42 shows little change with decreasing temperature until @xmath34 where , similar to @xmath41 , a small jump is observed ; however , @xmath42 continues to increase indicating a dramatic increase in the resistivity anisotropy until slightly below @xmath30 , where @xmath42 begins to decrease quickly , reaching a negligible value below @xmath62 k.@xcite it is possible to cleave some of these materials , resulting in mirror - like _
ab_-plane surfaces .
however , in most cases cleaving yields a `` terraced '' surface consisting of numerous steps with exposed _ c_-axis faces , shown in fig . [ fig : cleave ] . in fig .
[ fig : faces](a ) the absolute reflectance of a freshly - cleaved _ ab_-plane surface for the @xmath3 material has been measured at room temperature at a near - normal angle of incidence over the infrared region using an _ in situ _
evaporation technique.@xcite in addition to the hagen - rubens @xmath63 behavior expected for a metallic system and the two normally infrared - active vibrations at about 120 and 360 @xmath12 , there are two strong antiresonances at @xmath64 and 575 @xmath12 associated with the longitudinal optic modes of the infrared - active _ c_-axis modes ( sec .
these features are absent from naturally - grown crystal faces , but are observed whenever a surface has been cut and polished , or as illustrated in the photomicrograph shown in fig .
[ fig : cleave ] , when a cleaved surface has numerous steps .
this phenomenon has been documented extensively in the literature and is attributed to the misorientation of the _ ab_-plane surface.@xcite clearly , it is desirable to minimize these extrinsic features while preserving the intrinsic behavior of the sample . pursuing this goal ,
the opposite face was polished using a series of successively finer diamond grits using kerosene as a suspension agent , with a final lap using a 0.1 @xmath65 m diamond paste .
the infrared reflectance shown in fig .
[ fig : faces](b ) is quite similar to the reflectance of the cleaved surface , and shows that the antiresonance features are still present .
it was subsequently determined from x - ray laue diffraction that this surface was misoriented by @xmath66 .
the sample was reoriented using the laue camera to better than @xmath67 and then repolished .
the resulting reflectance in fig .
[ fig : faces](c ) indicates that the antiresonance features have been minimized .
thus , while cleaved surfaces are often desirable , in those cases where strong antiresonances are observed , determining the correct orientation and careful polishing may remove most of these unwanted spectral features .
the reflectance of single crystal la@xmath0ba@xmath1cuo@xmath2 for @xmath3 , 0.125 and 0.145 has been measured at a near - normal angle of incidence for light polarized in the _ a - b _ planes , as well as along the _ c _ axis , over a wide frequency range from @xmath68 mev to over 3 ev ( @xmath69 to over 24,000 @xmath12 ) at a variety of temperatures using the previously described overcoating technique .
ab_-plane reflectance of the @xmath5 sample was measured from a cleaved surface , while for the other dopings the surface was oriented to better than @xmath67 and polished .
c_-axis was revealed by cutting the _ a - b _ face at @xmath70 , which was subsequently polished .
c_-axis direction was determined by using an infrared polarizer to check the modulation of the infrared signal , with the minima corresponding to the _
c_-axis direction .
ba@xmath71cuo@xmath2 at 295 k in the far - infrared region for light polarized in the _ a - b _ planes .
( a ) the reflectance from a cleaved surface .
there are two prominent antiresonances at @xmath64 and 570 @xmath12 which are associated with _
c_-axis longitudinal optic modes that become infrared - active due to a surface misorientation .
this effect can be seen in the _ c_-axis loss function ( dashed line ) .
( b ) the reflectance from the polished surface of the same sample with a measured surface misorientation of @xmath72 ; the antiresonances are still prominent .
( c ) the reflectance from the polished surface with a measured surface misorientation of @xmath73 ; the antiresonances are substantially weaker . ]
the reflectance is a complex quantity consisting of an amplitude and a phase , @xmath74 .
because only the amplitude @xmath75 is measured in this experiment it is often difficult to decipher how the reflectance is related to the real and imaginary optical properties .
thus , the complex optical properties have been calculated from a kramers - kronig analysis of the reflectance.@xcite the kramers - kronig transform requires that the reflectance be determined for all frequencies , thus extrapolations must be supplied in the @xmath76 limits . at low frequency , for light polarized in the _ a - b _ planes , above @xmath9 in the normal state the metallic hagen - rubens form for the reflectance
@xmath77 is employed .
below @xmath9 in the superconducting state @xmath78 ; however , it should be noted that when the reflectance is close to unity the analysis is not sensitive upon the choice of low - frequency extrapolation . for light polarized along the _ c _ axis , at low frequency
the hagen - rubens form is used in the normal state , while below @xmath9 we again use @xmath79 . for all polarizations
the reflectance is assumed to be constant above the highest measured frequency point up to @xmath80 @xmath12 , above which a free electron gas asymptotic reflectance extrapolation @xmath81 is employed.@xcite
the temperature dependence of the reflectance of la@xmath0ba@xmath1cuo@xmath2 for light polarized in the _ a - b _ planes is shown over a wide frequency range for @xmath3 , 0.125 and 0.145 in figs .
[ fig : reflec](a ) , ( b ) and ( c ) , respectively ; the insets show the temperature dependence of the reflectance in the far - infrared region .
ba@xmath1cuo@xmath2 for light polarized in the _ a - b _ planes over a wide frequency range for : ( a ) a polished sample with @xmath3 , ( b ) a cleaved sample for @xmath5 , and ( c ) a polished sample with @xmath7 .
insets : the temperature dependence of the reflectance in the far - infrared region . ] at room temperature the reflectance curves for all three dopings are similar in that the reflectance at low frequency is high and decreases with frequency until a plasma edge in the reflectance is encountered at about 1 ev .
the reflectance in the infrared region is observed to increase with electronic doping . in the three samples the temperature dependence of the reflectance all
show a similar increase in the low - frequency reflectance with decreasing temperature . in the superconducting state
the @xmath3 material displays an abrupt increase in the low - frequency reflectance , while the changes in the @xmath7 material below @xmath9 are not as pronounced . at low temperature
the reflectance of the @xmath5 sample is dramatically different ; below @xmath82 @xmath12 the reflectance increases down to the lowest measured temperature , but close to or below @xmath30 the reflectance in the @xmath83 @xmath12 region actually decreases with temperature,@xcite as shown in the inset of fig .
[ fig : reflec](b ) . while the reflectance contains a great deal of information about the electronic transport and structure in a material , the complicated behavior observed in the @xmath5 material makes it difficult to interpret .
the determination of the optical conductivity from a kramers - kronig analysis of the reflectance is more revealing .
the real part of the optical conductivity for light polarized in the _ a - b _ planes of la@xmath84ba@xmath71cuo@xmath2 is shown in the infrared region at temperatures above @xmath9 in fig .
[ fig : sigma1](a ) . at room temperature
the conductivity appears rather broad in the far - infrared region , but as the inset in this panel shows it is in fact decreasing rather rapidly before giving way above @xmath85 @xmath12 to a broad plateau in the mid - infrared ; there is a shoulder at about 4000 @xmath12 ( 0.5 ev ) above which the conductivity continues to decrease until about 1 ev , at which point it begins to slowly rise again .
superimposed on the room - temperature conductivity curve are several sharp resonances at @xmath86 , 358 and 682 @xmath12 ; these are the normally - active infrared vibrations which are observable due to the poor screening in this class of materials.@xcite the weak antiresonances observed in the reflectance in fig .
[ fig : faces](c ) at @xmath64 and 570 @xmath12 attributed to a surface misorientation manifest themselves as weak resonances or antiresonances in the conductivity . in the htt phase ,
the irreducible vibrational representation for the infrared active modes is @xmath87 where the singly - degenerate @xmath88 modes are active along the _ c _ axis and the doubly - degenerate @xmath89 modes are active in the _ a - b _ planes . below the structural phase transition in the lto phase the unit cell is larger and more complicated . the lowered symmetry results in the removal of degeneracy and more infrared modes @xmath90 where the @xmath91 modes are silent , and the singly - degenerate @xmath92 , @xmath93 and @xmath94 modes are active along the _ c _ , _ b _ and _ a _ axes , respectively .
the additional low - temperature structural transition to either a ltlo or ltt+ltlo phase represents a further reduction of symmetry ; for the ltt phase the irreducible vibrational representation becomes @xmath95 where the @xmath96 , @xmath92 and @xmath93 modes are silent ; as in the case of the htt phase the @xmath88 and @xmath89 modes are active along the _ c _ axis and the _ a - b _ planes , respectively . at room temperature la@xmath84ba@xmath71cuo@xmath2 appears to be in the htt phase ( demonstrated by the _
c_-axis properties in sec .
iiib ) ; of four possible in - plane modes , only three are clearly identified ( the remaining mode is likely too weak to be observed ) . at low temperature
the vibrations typically narrow and increase in frequency ; however , the low - frequency mode at @xmath86 @xmath12 in fig .
[ fig : sigma1](a ) appears to broaden anomalously at low temperature , suggesting it may be sensitive to the charge or spin order .
ba@xmath71cuo@xmath2 for light polarized in the _ a - b _ planes .
( a ) the infrared region for several temperatures above @xmath97 k. the antiresonances observed in the reflectance in fig .
[ fig : faces](c ) are indicated by asterisks , while the in - plane infrared - active lattice modes are indicated by the arrows .
inset : the drude - lorentz fit to the data at 295 k. the drude component ( dashed line ) and lorentz oscillators ( dash - dot lines ) combine ( long - dashed line ) to reproduce the data ( solid line ) quite well .
( b ) the optical conductivity just above and well below @xmath9 in the infrared region illustrating the transfer of spectral weight into the condensate .
inset : the drude - lorentz fit to the data at 35 k. ] as the temperature decreases the low - frequency spectral weight associated with the free - carrier component increases rapidly .
the spectral weight is defined here simply as the area under the conductivity curve over a given frequency interval : @xmath98 . for a metal , in the absence of other excitations , the area under the entire conductivity curve is the well - known _
f_-sum rule,@xcite @xmath99 , where @xmath100 is the classical plasma frequency .
the value for the dc resistivity for @xmath3 in fig .
[ fig : trans](a ) just above @xmath9 is in reasonably good agreement with the extrapolated value for @xmath101 , although it should be noted that in this temperature region the resistivity is changing rapidly .
.the parameters for the drude plasma frequency @xmath102 and scattering rate @xmath103 determined in the normal state at 100 k.@xmath104 the superconducting plasma frequency @xmath105 and effective penetration depth are determined for @xmath106 . [ cols="^,^,^,^,^,^ " , ] [ tab : props ] angle - resolved photoemission ( arpes ) studies of high - temperature superconductors in general show a single band crossing the fermi level,@xcite indicating that the intra - band excitations giving rise to the free - carrier component have a single origin . within this context , it is not unreasonable to try and model the optical conductivity using a simple drude - lorentz model for the complex dielectric function @xmath107 , @xmath108 where @xmath109 is the real part of the dielectric function at high frequency , @xmath110 and @xmath111 are the square of the plasma frequency and scattering rate for the delocalized ( drude ) carriers ; @xmath112 , @xmath113 and @xmath114 are the position , width , and strength of the @xmath115th vibration or excitation . the complex conductivity is @xmath116/4\pi$ ] , yielding the expression for the real part of the optical conductivity @xmath117 where the first term is the drude response and @xmath118 is the contribution from the bound ( lorentz ) excitations .
( when @xmath102 and @xmath103 are in units of @xmath12 , the conductivity has units of @xmath119@xmath12 . ) the drude - lorentz model has been fit to the normal state conductivity using a non - linear least - squares method at 295 and 35 k. in addition to the drude component , two lorentz oscillators have been included somewhat arbitrarily at @xmath85 and 3100 @xmath12 to reproduce the observed mid - infrared response.@xcite the fit to the data at 295 k is in good agreement with the experimental data and is shown over a broad spectral range in the inset of fig .
[ fig : sigma1](a ) .
the fit to the data in the normal state at 35 k is quite good and is shown in the inset in fig .
[ fig : sigma1](b ) .
the drude plasma frequency @xmath120 @xmath12 does not vary with temperature , while the scattering rate decreases dramatically with temperature , @xmath121 , 207 , 131 and 75 @xmath12 at 295 , 200 , 100 and 35 k , respectively ( table [ tab : props ] ) . for a superconductor with an isotropic energy gap @xmath122 ,
the gap in the optical conductivity is @xmath123 . in the superconducting state below @xmath9 , shown in fig .
[ fig : sigma1](b ) for @xmath3 , the dramatic decrease in the low - frequency optical conductivity below @xmath124 @xmath12 ( 25 mev ) is associated with the formation of a superconducting energy gap;@xcite this result is in reasonable agreement with the arpes estimate of @xmath125 mev for the maximum gap value in this material.@xcite the `` missing area '' is referred to as the spectral weight of the condensate @xmath126 and may be calculated from the ferrell - glover - tinkham sum rule@xcite @xmath127 here @xmath128 is the square of the superconducting plasma frequency , and the superfluid density is simply @xmath129 ; the cut - off frequency @xmath130 is chosen so that the integral converges smoothly .
the superconducting plasma frequency has also been determined from the real part of the dielectric function in the low frequency limit where @xmath131 .
yet another method of extracting @xmath105 from @xmath132 is to determine @xmath133^{1/2}$ ] well below @xmath9 in the @xmath134 limit.@xcite all three techniques yield @xmath135 @xmath12 , which corresponds to an effective penetration depth of @xmath136 .
the fact that less than two - thirds of the free - carriers in the normal state have condensed ( @xmath137 ) is consistent with the observation of strong dissipation in the normal state .
ba@xmath138cuo@xmath2 for light polarized in the _ a - b _ planes measured just above and well below the critical temperature .
the asterisks denote structure due to surface misorientation , while the arrows indicate the in - plane infrared - active lattice modes .
inset : the temperature dependence of the optical conductivity in the normal state .
] the temperature dependence of the real part of the optical conductivity for the @xmath7 material is shown in fig .
[ fig : sigma2 ] at a variety of temperatures above and below @xmath9 . in the normal state
the overall behavior is not unlike that of the @xmath3 material ; at low frequency a drude - like component narrows rapidly with decreasing temperature but gives way to a broad , incoherent component above @xmath139 mev .
the conductivity in the normal state can be fit reasonably well using the drude - lorentz model yielding @xmath140 @xmath12 and @xmath141 @xmath12 at room temperature .
the larger value for @xmath102 denotes the increase in the density of free carriers .
it is interesting to note that the normal - state scattering rate in the overdoped material is roughly twice that of the underdoped system ; this may be due to electronic correlations or it may be that the increasing cation disorder in the blocking layer results in a larger value of @xmath103 ( table [ tab : props ] ) .
upon entry into the superconducting state a decrease in the low - frequency conductivity is once again observed ; however , the onset at which this decrease occurs is rather indistinct and there is considerably more residual conductivity at low frequency than was observed in the underdoped material . using the techniques previously discussed , the superconducting plasma frequency is estimated to be @xmath142 @xmath12 , corresponding to a penetration depth of @xmath143 . as a result of the large normal state scattering rate ,
less than a fifth of the free - carriers have condensed ( @xmath144 ) .
the in - plane reflectance of la@xmath0ba@xmath1cuo@xmath2 for @xmath5 is shown in fig .
[ fig : reflec](b ) .
while some aspects of the reflectance have been previously discussed,@xcite there are several important points that bear repeating .
as the temperature is lowered , the reflectance is observed to increase in the far- and mid - infrared regions .
close to or below the spin - ordering temperature @xmath145 k the reflectance decreases in the @xmath83 @xmath12 interval ; however , below about 200 @xmath12 the reflectance continues to increase.@xcite the consequences of this behavior may be seen more clearly in the conductivity shown for temperatures above @xmath34 in fig .
[ fig : sigma3](a ) and mainly below @xmath34 in fig .
[ fig : sigma3](b ) .
ba@xmath56cuo@xmath2 for light polarized in the _ a - b _ planes at several temperatures above @xmath34 and @xmath30 .
inset : the infrared reflectance for the same temperatures .
( b ) the conductivity for several temperatures below @xmath34 and @xmath30 in the same region showing the transfer and loss of spectral weight occurring mainly below @xmath30 .
inset : the infrared reflectance for the same temperatures contrasting the different responses at low and high frequency .
the asterisks denote structure due to surface misorientation , while the arrows indicate the in - plane infrared - active lattice modes . ] at high temperature the optical conductivity can be modeled quite well using a drude - lorentz model ; as the temperature is lowered from room temperature to just above @xmath146 k , there is a characteristic narrowing of the drude - like component and spectral weight is transferred from high to low frequency .
the plasma frequency @xmath147 @xmath12 is relatively constant , while the scattering rate decreases from @xmath148 @xmath12 at room temperature to 81 @xmath12 at @xmath34 ( table [ tab : props ] ) . in fig .
[ fig : sigma3](b ) for @xmath149 the scattering rate continues to decrease without any obvious loss of spectral weight or change in @xmath102 .
below @xmath30 the drude - like component in the conductivity continues to narrow ; however , above @xmath150 @xmath12 there is now a noticeable increase in the conductivity , indicating a transfer of spectral weight from the low - frequency free - carrier component to localized or gapped excitations,@xcite the significance of which will be discussed in sec .
this is precisely the temperature at which the infrared reflectance , shown in the inset of fig .
[ fig : sigma3](b ) , begins to decrease above @xmath82 @xmath12 , while still increasing at low frequency .
below @xmath151 k little spectral weight is being transferred to high energy , instead the decrease in the low - frequency spectral weight is now associated almost entirely with the increasingly narrow free - carrier response , such as might be expected in a nodal metal.@xcite .
attempts to fit the free - carrier component with a drude - model yield a rapidly decreasing value for @xmath102 for @xmath152 , but at the same time @xmath103 is also decreasing dramatically.@xcite we note that in the drude model @xmath153 , so that the net effect is the rapidly decreasing resistivity observed in fig .
[ fig : trans](b ) . at @xmath154
k the free - carrier response is now so narrow it may no longer be accurately observed , giving way to a broad , incoherent background conductivity .
it is difficult to distinguish this behavior from the phenomenon of missing spectral weight in the superconducting state .
the drude model is a reasonable description of a non - interacting fermi liquid ; however , the cuprates are either moderately or strongly correlated electron systems.@xcite in this latter case , a more general description of the drude model is favored in which the scattering rate and the effective mass are allowed to adopt a frequency dependence,@xcite @xmath155\ ] ] and @xmath156 , \ ] ] where @xmath157 is the bare mass , @xmath158 and @xmath159 is a frequency - dependent electron - boson coupling constant .
in this instance we set @xmath160 and @xmath161 ( although the choice of @xmath109 has little effect on the scattering rate or the effective mass in the far - infrared region ) .
the infrared - active lattice modes observed in the conductivity have been fit using lorentzian oscillators superimposed on a linear background and then removed from the complex dielectric function .
the resulting temperature dependence of the in - plane scattering rate @xmath162 is shown in fig .
[ fig : tau ] for the three different dopings studied in this work .
a simple drude model would produce a frequency - independent scattering rate ; however , despite the reasonable agreement with the drude model , in figs .
[ fig : tau](a ) , ( b ) , and ( c ) , @xmath162 displays an approximately linear frequency dependence .
we note that in the @xmath134 limit the two models should yield the same scattering rate , and indeed @xmath163 . the temperature dependence @xmath103 ( shown on the plots and for each doping ) merits a brief discussion .
for the three dopings examined , @xmath3 , 0.125 , and 0.145 , the values of the drude scattering rate at 295 k increase steadily from @xmath164 @xmath12 to 439 @xmath12 and 500 @xmath12 , respectively , indicating either stronger electronic correlations or that the increase in barium concentration is accompanied by increasing out - of - plane cation disorder , which may in turn lead to an increase in elastic scattering in the copper - oxygen planes . while @xmath103 decreases reasonably quickly with decreasing temperature for @xmath3 ,
this trend is not as pronounced for @xmath7 .
as previously noted , in the 1/8 crystal @xmath103 decreases steadily with temperature , but for @xmath152 the reduction of the scattering rate is dramatic , an effect that is shown in more detail in fig .
[ fig : tau](d ) .
ba@xmath1cuo@xmath2 ( with the phonons removed ) in the far - infrared region for ( a ) @xmath3 , ( b ) the 1/8 phase , and ( c ) @xmath7 .
( d ) detailed temperature dependence of the scattering rate for the 1/8 phase below 60 k. the values for @xmath103 determined from the drude fits are indicated by the symbols at the origin .
the dashed line indicates @xmath165 . ]
the frequency - dependence of the scattering rate contains a wealth of information .
the dashed line in fig . [ fig : tau ] denotes @xmath166 ; the region below this line is associated with a landau - fermi liquid regime where the quasiparticles are well defined . in the materials examined , at high temperature , @xmath167 at low frequency , indicating that the conductivity in the _ a - b _ planes is strongly dissipative , and that the quasiparticles are not well defined . across the far - infrared region
@xmath162 decreases with temperature and in the @xmath3 material , just above @xmath9 at 35 k , and in the 1/8 phase at low temperature ( @xmath168 ) , @xmath169 , indicating well - defined quasiparticles and coherent transport ; this condition is never satisfied in the @xmath7 material . at low frequency , @xmath162 decreases with decreasing temperature ; however , in the 1/8 phase this trend is reversed for @xmath170 above about 150 @xmath12 , shown in fig .
[ fig : tau](d ) , where the scattering rate increases dramatically .
this behavior is consistent with the formation of a momentum - dependent gap .
it is interesting to note that , aside from some artifacts due to the imperfect nature of the phonon subtraction , for the @xmath3 material in fig .
[ fig : tau](a ) the frequency dependent scattering rate is essentially featureless except for a small kink at about 45 mev ( also observed in the @xmath7 material ) . in many high - temperature superconductors ,
@xmath162 often has sharp inflection points in the normal state;@xcite when an inversion of the scattering rate is performed@xcite these structures correspond to peaks in the electron - boson spectral function.@xcite it has been proposed that the peak location is proportional to the position of the magnetic scattering resonance@xcite in polarized neutron scattering measurements @xmath171 , and more generally that it scales with the superconducting transition temperature,@xcite @xmath172 . in the underdoped cuprates the presence of a pseudogap on the fermi surface
is expected to produce kinks in the scattering rate at roughly @xmath173 and @xmath174.@xcite in the @xmath3 ( @xmath97 k ) material , taking the average , @xmath175 mev , while @xmath176 mev ( ref . ) yielding @xmath177 and 46 mev ; the larger value is very close to where a kink is observed in @xmath162 in fig .
[ fig : tau](a ) . while it is tempting to associate this kink with a magnetic resonance , the absence of such a feature in the underdoped la@xmath0sr@xmath33cuo@xmath2 material@xcite suggests a different origin for this feature . in the more heavily doped @xmath7 ( @xmath178 k ) material , @xmath179
mev and @xmath180 mev suggesting features at @xmath181 and 32 mev ; however , no obvious structure in the scattering rate is observed in fig .
[ fig : tau](c ) at either of these energies .
it has been noted that the high - temperature superconductors obey the scaling relation @xmath182 , where @xmath183 is the dc conductivity in the normal state measured at @xmath184 , and the superfluid density @xmath10 is determined for @xmath106 ; this scaling relation is valid for both the metallic _ a - b _ planes as well as along the poorly conducting _ c _ axis , in both the electron and hole - doped cuprates.@xcite in this representation , the scaling relation is valid over five orders of magnitude and as such is usually shown as a log - log plot .
however , when only the _ a - b _ plane results are considered , the results span less than two orders of magnitude and it is convenient to rewrite this relation as @xmath185 where @xmath186 .
the scaling relation now has a similar appearance to the well known uemura relation,@xcite @xmath187 , and may be shown using a linear scale . in fig .
[ fig : scaling ] the values for @xmath9 are plotted against @xmath188 determined in the _ a - b _
planes for a variety of single - layer and double - layer cuprates .
the values for la@xmath0ba@xmath33cuo@xmath2 for @xmath3 ( @xmath4 k , @xmath189 @xmath119@xmath12 just above @xmath9 , and @xmath190 @xmath191 ) and @xmath7 ( @xmath192 k , @xmath193 @xmath119@xmath12 just above @xmath9 , and @xmath194 @xmath191 ) are highlighted within the circles and fall on the general scaling line .
these points also lie quite close to the related la@xmath0sr@xmath33cuo@xmath2 materials with comparable doping .
this result is consistent with the observation that there is strong dissipation in the normal state.@xcite versus the product of the dc resistivity measured just above @xmath9 and the spectral weight of the condensate , @xmath188 , in the copper - oxygen planes for a variety of electron and hole - doped cuprates , compared with the _ a - b _ plane results for la@xmath0ba@xmath1cuo@xmath2 for @xmath3 and 0.145 ( highlighted within the circles ) .
the dashed line corresponds to the general result for the cuprates @xmath195 . ]
the temperature dependence of the reflectance of la@xmath0ba@xmath1cuo@xmath2 for light polarized along the poorly - conducting _ c _ axis is shown in figs . [ fig : creflec](a ) , ( b ) and ( c ) for the @xmath3 , 0.125 and 0.145 dopings , respectively .
the reflectance in this polarization is dramatically different than what is observed in the metallic _ a - b _ planes ( fig .
[ fig : reflec ] ) ; as a result of the near total absence of a free - carrier contribution to the dielectric function , the reflectance is dominated by the normally infrared - active _ c _ axis lattice modes . in the normal state as
@xmath134 the reflectance displays a slight upturn , indicative of a weakly - conducting state ; this is seen most clearly in the @xmath7 material .
ba@xmath1cuo@xmath2 for light polarized along the _ c _ axis for ( a ) @xmath3 , ( b ) @xmath5 , and ( c ) @xmath7 .
the wave number ( photon energy ) scale is linear up to 60 @xmath12 ; above this it is logarithmic . ] below @xmath9 a sharp plasma edge is observed in the reflectance at low frequency in figs . [
fig : creflec](a ) and ( c ) for the @xmath3 and 0.145 dopings , respectively .
the onset of the plasma edge is extremely rapid in both materials .
the 1/8 phase material shows no hint of a plasma edge in the reflectance ; however , this is not surprising as the bulk @xmath9 in this material is too low to be accessed in our current experimental configuration .
c_-axis plasma edge in the reflectance is observed in many of the cuprate superconductors,@xcite and is attributed to the josephson coupling of the copper - oxygen planes resulting in the formation of a supercurrent and the onset of a bulk 3d superconducting state.@xcite it is interesting to note that the josephson plasma edge is considerably broader in the more heavily - doped material ; this observation is in agreement with the result for heavily - doped la@xmath0sr@xmath1cuo@xmath2 where it is suggested that the smearing of the plasma edge is due to spatial variations of the superconducting condensate.@xcite cccc + mode & @xmath112 & @xmath113 & @xmath114 + @xmath88 & 234.6 ( 0.2 ) & 29 ( 0.6 ) & 1085 ( 10 ) + @xmath88 & 343.6 ( 0.1 ) & 9.2 ( 0.5 ) & 61 ( 4 ) + @xmath88 & 381.0 ( 0.6 ) & 34 ( 3 ) & 71 ( 10 ) + @xmath88 & 495.9 ( 0.1 ) & 25 ( 0.6 ) & 312 ( 5 ) + & & & + + @xmath92 & 132.7 ( 0.2 ) & 5.7 ( 0.9 ) & 55 ( 6 ) + @xmath92 & 235.4 ( 0.2 ) & 22 ( 0.6 ) & 1110 ( 10 ) + @xmath92 & 314.2 ( 0.1 ) & 7.1 ( 0.2 ) & 85 ( 2 ) + @xmath92 & 348.0 ( 0.1 ) & 4.6 ( 0.3 ) & 113 ( 6 ) + @xmath92 & 380.4 ( 0.9 ) & 39 ( 6 ) & 109 ( 13 ) + @xmath92 & 495.8 ( 0.2 ) & 19 ( 0.6 ) & 338 ( 5 ) + [ tab : caxis ] the behavior along the _ c _ axis is easier to understand when the conductivity is considered , shown in figs .
[ fig : csigma](a ) , ( b ) and ( c ) for @xmath3 , 0.125 and 0.145 , respectively . in each case
the conductivity is dominated by a very strong infrared lattice mode at @xmath196 @xmath12 with another prominent mode at @xmath197 @xmath12 ; the conductivity is shown on a logarithmic scale to allow the emergence of weaker modes , as well as the behavior of the electronic background , to be examined .
the results of oscillator fits to the conductivity for the @xmath3 material at 295 and 35 k are listed in table [ tab : caxis ] . in the htt phase a total of four @xmath88 modes
are expected along the _ c _ axis ; despite @xmath198 k in the @xmath3 material , it is clear from the conductivity in fig .
[ fig : csigma](a ) that only four modes at @xmath196 , 344 , 381 and 496 @xmath12 are present at room temperature , indicating that the material is indeed in the htt phase . for
@xmath199 two new modes at @xmath200 and 314 @xmath12 quickly emerge , in agreement with the six @xmath92 modes expected in the lto phase .
it is unclear what effect , if any , the ltlo phase would have on the nature of the _ c _ axis vibrations ; however , at the lowest measured temperature there is a slight asymmetry in the strong 235 @xmath12 mode , suggesting a possible further reduction of symmetry .
the vibrational properties in the @xmath5 and 0.145 materials are similar . at room temperature
both of these materials are in the htt phase and four modes are observed ; two new modes emerge below the respective values for @xmath201 . in the 1/8 phase , at @xmath154 k this material
is expected to be in an ltt phase and a total of seven @xmath88 modes are expected to be active along the _ c _ axis ; the shoulder in the strong 235 @xmath12 mode suggests that all seven @xmath88 modes may be observed .
the picture in the more heavily doped @xmath7 material is less clear .
while six modes are observed at low temperature as expected for the ltt phase , the modes in this material have broadened and it is no longer obvious if there is any asymmetry in the 235 @xmath12 mode at low temperature .
this may be a reflection of the somewhat ambiguous nature of the low - temperature structure ( ltt@xmath55ltlo ) for this doping.@xcite ba@xmath1cuo@xmath2 for light polarized along the poorly - conducting _ c _ axis for ( a ) @xmath3 , ( b ) @xmath5 , and ( c ) @xmath7 . ] the conductivity along the _ c _ axis is orders of magnitude lower than what is observed in the metallic _ a - b _ planes . at room temperature in the @xmath3 material @xmath202
@xmath119@xmath12 , increasing to @xmath62 @xmath119@xmath12 in the 1/8 phase [ both values are comparable with the _ c _ axis resistivity shown in figs .
[ fig : trans](a ) and ( b ) ] , and @xmath203 @xmath119@xmath12 in the @xmath7 material . the normal - state transport along the _ c _ axis is attributed to thermally - activated hopping , which is consistent with the observation of an electronic background that decreases ( weakly ) with temperature .
in addition , some of this decrease in the optical conductivity may also be due to a narrowing of vibrational features . despite the small value for the background conductivity , a suppression of the low - frequency conductivity due to the formation of a condensate
is observed below @xmath9 in the @xmath3 and 0.145 materials .
the decrease in the low - frequency conductivity in the 1/8 phase below @xmath34 is likely due to the decoupling of the planes due to the formation of charge - stripe order;@xcite however , the dramatic decrease in the _
c_-axis resistivity below @xmath30 [ fig .
[ fig : trans](b ) ] is not reflected in the optical properties as no commensurate increase in the conductivity is observed down to the lowest measured frequency .
this suggests that the decrease in the low - frequency spectral weight at low temperature is associated with an increase in the optical conductivity at very low energies ( possibly microwave or radio frequencies ) , and is likely associated with the onset below @xmath30 of superconducting correlations in the planes.@xcite in the @xmath3 and 0.145 compounds the energy scale for the loss of spectral weight is roughly 80 @xmath12 ( 10 mev ) and 140 @xmath12 ( 18 mev ) , respectively .
it is interesting to note that these energy scales correspond to @xmath25 , rather than @xmath204 observed in the _ a - b _
planes.@xcite this type of behavior has been previously observed in the cuprates;@xcite however , it remains unexplained .
, in la@xmath0ba@xmath1cuo@xmath2 for @xmath3 ( ) and 0.145 ( @xmath205 ) and optimally - doped la@xmath206sr@xmath207cuo@xmath2 ( @xmath208 ) ( ref . ) .
the dashed line is a guide to the eye . ]
while it is possible to determine @xmath105 from the ferrell - glover - tinkham sum rule , the small values of the conductivity and the proximity of strong lattice vibrations can lead to significant uncertainties . under these circumstances a more reliable approach is to examine the low - frequency response of the real part of the dielectric function , @xmath209 . for @xmath106 this method yields @xmath210 @xmath12 for @xmath3 , and @xmath211 @xmath12 for @xmath7 , which corresponds to effective penetration depths of approximately 13.4 and 9.1 @xmath65 m , respectively .
it is also possible to track the temperature dependence of the superfluid density using this method .
the temperature dependence of the superfluid density for the @xmath3 and 0.145 samples , expressed here as the ratio value of the square of the penetration depths @xmath212 , is plotted against the reduced temperature @xmath213 and compared with optimally - doped la@xmath206sr@xmath207cuo@xmath2 , shown in fig .
[ fig : lambda ] . in a comprehensive study of the evolution of the _ c_-axis superfluid density in la@xmath0sr@xmath1cuo@xmath2 for a wide variety of dopings,@xcite
it may be noted that the superfluid density develops most rapidly for the optimally - doped material .
it is therefore surprising that in underdoped la@xmath84ba@xmath71cuo@xmath2 the superfluid density develops extremely quickly just below @xmath9 before slowing below @xmath214 ; in this regard it is reminiscent of the behavior of a system with an isotropic _ s_-wave energy gap,@xcite while the more heavily doped @xmath7 sample develops relatively slowly and is similar to what is observed in the copper - oxygen planes in a _
d_-wave system.@xcite we attribute this difference in behavior to the presence of a pseudogap in the underdoped materials where the antinodal region is gapped well above @xmath9 .
if the _ c_-axis transport is dominated by the antinodal zone boundary region of the fermi surface,@xcite then when the fermi pocket is gapped below @xmath9 the _ c_-axis response will be dominated by the gap maximum @xmath25 well away from the nodal region , and thus will not be sensitive to the _
d_-wave nature of the system , instead resembling a more isotropic system.@xcite the formation of charge order in the @xmath3 material below @xmath43 k may frustrate the josephson coupling between the planes,@xcite leading to the observed slowing of the formation of the condensate below roughly @xmath214 . on the other hand , the pseudogap is generally not present in materials close to optimal doping ; below @xmath9 the entire fermi surface is gapped .
this suggests that in a optimally - doped material the _
c_-axis response will be more sensitive to the
_ d_-wave nature of the energy - gap and therefore evolves more slowly , similar to the response observed in the copper - oxygen planes.@xcite vs @xmath215 along the _ c _ axis for a variety of electron and hole - doped cuprates compared with la@xmath0ba@xmath1cuo@xmath2 for @xmath3 and 0.145 ( highlighted within the circles ) .
the superscripts next to the symbols for la@xmath0sr@xmath1cuo@xmath2 and yba@xmath17cu@xmath18o@xmath216 refer to different chemical dopings.@xcite the dashed line corresponds to the general result for the cuprates @xmath217 . ]
it was previously noted that the scaling relation @xmath218 is valid not only in the _ a - b _ planes , but also along the _ c _ axis.@xcite while it is possible to plot the _
c _ axis data in the form of @xmath9 versus @xmath188 , the highly anisotropic nature of these materials results in conductivities ( resistivities ) along the _ c _ axis that are orders of magnitude smaller ( larger ) than those observed in the _ a - b _ planes , resulting in a significant uncertainty in @xmath183 and the corresponding value for @xmath219 . in addition , the scaling behavior of @xmath10 emerges naturally if the superconductivity along the _ c _ axis is considered to originate from josephson coupling.@xcite the _ c _ axis values for the spectral weight of the condensate @xmath220 are plotted as a function of @xmath215 in fig .
[ fig : cscale ] for a variety of single - layer and double - layer cuprates .
the dashed line is the best fit to the data , @xmath182 .
the _ c_-axis values for la@xmath0ba@xmath33cuo@xmath2 for @xmath3 ( @xmath4 k , @xmath221 @xmath119@xmath12 just above @xmath9 , and @xmath222 @xmath191 ) and 0.145 ( @xmath192 k , @xmath223 @xmath119@xmath12 just above @xmath9 , and @xmath224 @xmath191 ) are highlighted in the circles .
these values fall on the general scaling line and once again are quite close to the related la@xmath0sr@xmath33cuo@xmath2 materials with comparable doping .
the behavior of the optical conductivity at low temperature in the @xmath15 phase is fundamentally different than the other two dopings , where a response that is consistent with the opening of a superconducting _ d_-wave energy gap is observed .
the transfer of spectral weight from low to high frequency that develops close to or below @xmath30 is extremely unusual for a hole - doped cuprate and has the characteristics of an anisotropic transport gap on the fermi surface due to the formation of charge or spin order .
a partial gapping of the fermi surface has been observed in other 2d materials with incommensurate charge - density waves ( cdws).@xcite unlike one - dimensional systems , where the modulation wave vector can perfectly span the fermi surface and the cdw almost always leads to the complete gapping of the fermi surface , in 2d perfect nesting is not possible@xcite and as a result the cdw gaps only a portion of the fermi surface .
fortunately , arpes measurements have been performed on this composition,@xcite so that we can analyze nesting of the fermi surface by the measured modulation wave vectors.@xcite the charge - order wave vector is approximately equal to @xmath225 in the antinodal region .
there is certainly a gap on that portion of the fermi surface , but it is already there as a pseudogap before the charge - stripe order appears.@xcite therefore , the temperature dependence of the optical spectral weight transfer suggests a connection with the spin - stripe order for which the wave vector would nest parts of the fermi arc .
the arpes measurements indicate that the gapping of the fermi arc below @xmath30 is indistinguishable from the @xmath226-wave - like gap found in superconducting samples.@xcite while the conventional density - wave picture is a possible candidate to explain the observed redistribution of spectral weight at low temperature , another possibility is that this shift is associated with superconducting correlations in this material.@xcite we are not able to access the bulk superconducting state in the @xmath15 material ; however , the previously discussed scaling relation@xcite indicates that because of the dramatically reduced value for the bulk superconductivity in this material ( @xmath6 k ) the josephson plasma frequency along the _ c _ axis is estimated to be @xmath227 @xmath12 .
the josephson plasma edge in the reflectance typically occurs in a region close to the renormalized superconducting plasma frequency @xmath228 . given the experimentally - determined value of @xmath229 at 50 @xmath12 , this yields a value of @xmath230 @xmath12 , below our ability to measure using current optical techniques .
the strongly reduced value for the _ c_-axis jopsephson plasma edge for @xmath15 is consistent with the disappearance of this feature in la@xmath231nd@xmath28sr@xmath207cuo@xmath2 , which occurs when the increasing nd concentration causes the structure to change from lto to ltt , resulting in stripe order.@xcite to explain the abrupt change in the josephson coupling in the latter case , despite the much weaker reduction in @xmath9 , himeda _
et al._@xcite proposed a pair - density - wave ( pdw ) superconducting state involving local @xmath226-wave symmetry but modulated by a sinusoidal envelope function ; the envelope function has the same period as the spin modulation , but its extrema are aligned with the maxima of the charge modulation .
the rotation of the stripe ( and pdw ) orientation by @xmath70 from layer to layer leads to a frustration of the interlayer josephson coupling .
this concept was rediscovered@xcite after the observation of quasi - two - dimensional superconductivity by transport measurements@xcite on la@xmath0ba@xmath1cuo@xmath2 with @xmath15 .
the pdw proposal also has implications for the in - plane conductivity .
given a sinusoidally - modulated pair wave function , the superfluid density can be written as @xmath232,\ ] ] where @xmath233 [ or @xmath234 is the ordering wave vector ( in reciprocal lattice units based on the htt phase ) , and @xmath26 is the phase of the modulation with respect to the crystal lattice .
et al._@xcite showed that inhomogeneity of the superfluid density will cause some of the conductivity that would normally be at @xmath235 to shift to finite @xmath236 .
orenstein@xcite went further , and showed that sinusoidal modulation of @xmath237 shifts conductivity to a peak at the renormalized superconducting plasma frequency @xmath238 . from table
[ tab : props ] we extrapolate an in - plane value of @xmath239 @xmath12 for @xmath15 . using the experimentally - determined value of @xmath240 at 2 ev for @xmath109 [ there is little or no temperature dependence of @xmath241 in this frequency region ] , then we estimate the in - plane @xmath242 @xmath12 . in considering fig .
[ fig : sigma3](b ) , we already noted that some spectral weight is shifted from low frequency to above 300 @xmath12 . to emphasize this behavior , fig .
[ fig : ratio ] shows the ratio of @xmath243 to that at @xmath244 k. the enhancement of @xmath245 at @xmath246 mev is clearly present at 40 k and has doubled by 6 k ; however , there is no enhancement at 50 k , which is already below @xmath34 .
the development of the 80-mev feature on cooling correlates with the drop in in - plane resistivity . to compare with orenstein
s analysis,@xcite we have to allow for the fact that we have superconducting correlations but not long - range order for 40 k@xmath247 k. as a consequence , there should be a finite scattering rate associated with the `` superconducting '' features .
the small but finite width of the @xmath235 ( `` drude '' ) peak for 16 k@xmath248 k is consistent with this expectation .
the development of the finite - frequency peak is qualitatively compatible with orenstein s prediction for a modulated superconductor , although it appears at a frequency somewhat lower than @xmath249 .
the shifted spectral weight appears to be pushed above @xmath204 , where arpes studies@xcite indicate @xmath250 mev in the antinodal region .
there are some experimental results that complicate this picture .
the arpes studies@xcite indicate a @xmath226-wave - like gap along the near - nodal arc for @xmath251 k , whereas the pdw state is predicted to have a gapless near - nodal arc.@xcite the @xmath226-wave - like gap suggests the presence of a more uniform superconducting component that develops below @xmath252 k. such a uniform component might be necessary to get bulk superconductivity at low temperature , as the correlation length of the pdw order is limited by that of the stripe order , which is restricted to several hundred .@xcite of course , if a uniform component is present , then one returns to the challenge of explaining why the interlayer josephson coupling is frustrated .
a different complication is that a substantial narrowing of the zero - frequency peak is already apparent at 50 k , where we have no enhancement at @xmath253 mev . at that temperature
, we have charge stripe order but not spin order , and the spin order may be necessary to pin the phase of @xmath237 with respect to the lattice .
perhaps at 50 k , in the absence of pinning , the pdw correlations start to show up at low frequency but not at higher frequency . in any case
, the high - frequency peak remains at 6 k , below the point where 2d superconducting order develops,@xcite so that the modulated @xmath254 should still be present there . .
inset : the ratio over a wide frequency range . ]
the temperature dependence of the optical properties of la@xmath0ba@xmath1cuo@xmath2 for light polarized in the _ a - b _ planes and along the poorly - conducting _ c _ axis has been determined for @xmath3 , 0.125 and 0.145 .
the @xmath3 and 0.145 samples are superconducting , with @xmath97 and @xmath255 k , respectively .
the in - plane optical conductivity can be characterized as a drude - like response that narrows with decreasing temperature , superimposed on a flat , incoherent background ; below @xmath9 missing low - frequency spectral weight is the hallmark of the formation of a superconducting condensate .
interestingly , in both cases @xmath256 , indicating that only a fraction of the normal - state carriers condense ; this is consistent with the observation of strong dissipation in the normal state .
the optical properties along the poorly - conducting _ c _ axis in the @xmath3 and 0.145 materials are insulating in character , dominated by the normally infrared - active phonons with only a weak electronic background ; new lattice modes appear at the htt@xmath23lto transition .
below @xmath9 a sharp plasma edge is observed in the _
c_-axis reflectance ; this feature is attributed to the formation of a supercurrent along the _ c _ axis due to josephson coupling between the copper - oxygen planes . despite the dramatic reduction in the _
c_-axis resistivity below @xmath30 in the 1/8 material , no josephson plasma edge or hint of metallic behavior is observed . in the superconducting materials , the spectral weight of the condensate for the _ a - b _ plane and the _ c _
axis fall on the universal scaling line @xmath182 .
the behavior of the anomalous @xmath15 phase in which the superconductivity is dramatically weakened ( @xmath6 k ) stands out in stark contrast .
the in - plane optical properties are similar to the other two dopings at high temperature ; however , close to or below @xmath30 there is an unusual suppression of the reflectance over much of the infrared region and a transfer of spectral weight from low frequency to energies above @xmath257 mev , indicating the formation of a momentum - dependent energy gap .
below @xmath30 the free - carrier response continues to narrow dramatically with decreasing temperature , to the extent that at the lowest measured temperature the spectral weight associated with the free carriers can no longer be observed .
this response mimics the missing spectral weight associated with the formation of a condensate ; however , we are currently unable to distinguish a momentum - dependent cdw gap from the similar behavior expected from a pdw state .
we would like to thank a. akrap , a. auerbach , d. n. basov , d. a. crandles , s. v. dordevic , j. hwang , s. a. kivelson , m. reedyk , t. timusk , and n. l. wang for useful discussions , and m. g. rechner for careful reading of this manuscript .
research supported by the u.s .
department of energy , office of basic energy sciences , division of materials sciences and engineering under contract no .
de - ac02 - 98ch10886 .
146ifxundefined [ 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 * * , ( ) link:\doibase 10.1038/365323a0 [ * * , ( ) ] , ed . , @noop _ _ ( , , ) @noop * * , ( ) link:\doibase 10.1088/0034 - 4885/62/1/002 [ * * , ( ) ] link:\doibase 10.1126/science.288.5465.468 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.75.473 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.77.721 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.78.17 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.83.471 [ * * , ( ) ] link:\doibase 10.1103/physrevb.40.2254 [ * * , ( ) ] link:\doibase 10.1016/0921 - 4534(91)90700 - 9 [ * * , ( ) ] link:\doibase 10.1103/physrevb.51.12911 [ * * , ( )
] link:\doibase 10.1038/nphys909 [ * * , ( ) ] link:\doibase 10.1103/physrevb.38.4596 [ * * , ( ) ] link:\doibase 10.1016/0921 - 4534(92)90029-c [ * * , ( ) ] link:\doibase 10.1103/physrevb.57.7491 [ * * , ( ) ] link:\doibase 10.1103/physrevb.73.180505 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.62.2751 [ * * , ( ) ] link:\doibase 10.1143/jpsj.60.1448 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.71.1903 [ * * , ( ) ] link:\doibase 10.1016/0921 - 4534(93)90108 - 3 [ * * , ( ) ] link:\doibase 10.1016/0921 - 4534(91)91811-h [ * * , ( ) ] \doibase 10.1023/a:1012670430673 [ * * , ( ) ] \doibase 10.1023/a:1022966211887 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.96.257002 [ * * , ( ) ] link:\doibase 10.1103/physrevb.76.014508 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.99.067001 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.174529 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.212506 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.180513 [ * * , ( ) ] link:\doibase 10.1007/s10948 - 008 - 0422 - 0 [ * * , ( ) ]
link:\doibase 10.1103/physrevlett.102.157001 [ * * , ( ) ] @noop ( ) , link:\doibase 10.1103/physrevb.83.060506 [ * * , ( ) ] link:\doibase 10.1103/physrevb.70.104517 [ * * , ( ) ] link:\doibase 10.1038/nature02574 [ * * , ( ) ] link:\doibase 10.1038/nphys178 [ * * , ( ) ] link:\doibase 10.1038/nature04704 [ * * , ( ) ] link:\doibase 10.1103/physrevb.76.184121 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.224410 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.092507 [ * * , ( ) ] link:\doibase 10.1063/1.2889162 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.214507 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.104.057004 [ * * , ( ) ] link:\doibase 10.1103/physrevb.83.104506 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.195101 [ * * , ( ) ] link:\doibase 10.1103/physrevb.40.7391 [ * * , ( ) ] link:\doibase 10.1038/375561a0 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.75.1201 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.83.3916 [ * * , ( ) ] link:\doibase 10.1080/00018730903122242 [ * * , ( ) ] link:\doibase 10.1088/0953 - 8984/23/38/385701 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.99.127003 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1126/science.1134742 [ * * , ( ) ] link:\doibase 10.1038/nphys1159 [ * * , ( ) ] link:\doibase 10.1038/nphys334 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.98.267004 [ * * , ( ) ] link:\doibase 10.1088/0034 - 4885/71/6/062501 [ * * , ( ) ] link:\doibase 10.1209/epl / i1999 - 00449 - 1 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.86.500 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.88.147003 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.90.037002 [ * * , ( ) ] link:\doibase 10.1103/physrevb.72.205101 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.91.129702 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.91.129701 [ * * , ( ) ] @noop _ _ ( , , ) link:\doibase 10.1103/physrevb.82.100505 [ * * , ( ) ] link:\doibase 10.1038/nature02673 [ * * , ( ) ] link:\doibase 10.1103/physrevb.72.134517 [ * * , ( ) ] link:\doibase 10.1103/physrevb.71.184515 [ * * , ( ) ] link:\doibase 10.1103/physrevb.80.180509 [ * * , ( ) ] link:\doibase 10.1016/0038 - 1098(90)90504 - 5 [ * * , ( ) ] @noop ( ) , @noop * * , ( ) , link:\doibase 10.1088/0022 - 3719/6/7/010 [ * * , ( ) ] link:\doibase 10.1364/ao.32.002976 [ * * , ( ) ] link:\doibase 10.1364/ao.46.007884 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.69.2705 [ * * , ( ) ] link:\doibase 10.1103/physrevb.49.13065 [ * * , ( ) ] link:\doibase 10.1016/s0921 - 4534(99)00397 - 4 [ * * , ( ) ] link:\doibase 10.1103/physrevb.60.13196 [ * * , ( ) ] link:\doibase 10.1103/physrevb.71.094508 [ * * , ( ) ] link:\doibase 10.1103/physrevb.75.045128 [ * * , ( ) ] @noop _ _ ( , , ) @noop _ _ ( , , ) pp .
link:\doibase 10.1103/physrevlett.84.5391 [ * * , ( ) ] in @noop _ _ , ( , , ) pp .
@noop link:\doibase 10.1103/physrev.109.1398 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.2.331 [ * * , ( ) ] link:\doibase 10.1103/physrevb.54.1264 [ * * , ( ) ] link:\doibase 10.1103/physrevb.67.184516 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.87.017001 [ * * , ( ) ] link:\doibase 10.1103/physrevb.67.174520 [ * * , ( ) ] link:\doibase 10.1103/physrevb.72.054529 [ * * , ( ) ] link:\doibase 10.1038/nphys1343 [ * * , ( ) ] link:\doibase 10.1103/physrevb.15.2952 [ * * , ( ) ] link:\doibase 10.1088/0953 - 8984/8/48/023 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.134522 [ * * , ( ) ] link:\doibase 10.1103/physrevb.67.214508 [ * * , ( ) ] link:\doibase 10.1103/physrevb.71.104529 [ * * , ( ) ] link:\doibase 10.1103/physrevb.75.144508 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevlett.100.137005 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.86.1610 [ * * , ( ) ] link:\doibase 10.1126/science.1067877 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.027003 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.100.177005 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.98.077004 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.104525 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.62.2317 [ * * , ( ) ] link:\doibase 10.1038/430512a [ * * , ( ) ] link:\doibase 10.1103/physrevb.73.180504 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.054512 [ * * , ( ) ] link:\doibase 10.1038/nphys1975 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.69.1455 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.71.1645 [ * * , ( ) ] link:\doibase 10.1016/0921 - 4534(95)00579-x [ * * , ( ) ] link:\doibase 10.1103/physrevb.53.14558 [ * * , ( ) ] link:\doibase 10.1103/physrevb.55.6051 [ * * , ( ) ] link:\doibase 10.1103/physrevb.60.r15051 [ * * , ( ) ] link:\doibase 10.1103/physrevb.61.r9269 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.91.167401 [ * * , ( ) ] link:\doibase 10.1103/physrevb.71.054503 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.72.2263 [ * * , ( ) ] link:\doibase 10.1103/physrevb.50.3511 [ * * , ( ) ] link:\doibase 10.1016/0921 - 4534(95)00359 - 2 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.77.4090 [ * * , ( ) ] link:\doibase 10.1103/physrevb.43.5370 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.73.2484 [ * * , ( ) ] link:\doibase 10.1126/science.261.5119.337 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.77.4632 [ * * , ( ) ] link:\doibase 10.1126/science.285.5431.1241 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.81.1298 [ * * , ( ) ] link:\doibase 10.1116/1.577016 [ * * , ( ) ] link:\doibase 10.1103/physrevb.42.8890 [ * * , ( ) ] link:\doibase 10.1103/physrevb.58.13565 [ * * , ( ) ] link:\doibase 10.1103/physrevb.64.161103 [ * * , ( ) ] @noop _ _ ( , , ) link:\doibase 10.1103/physrevlett.88.117001 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/11/11/115004 [ * * , ( ) ] link:\doibase 10.1103/physrevb.61.r14924 [ * * , ( ) ] link:\doibase 10.1016/s0921 - 4534(03)00705 - 6 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.174502 [ * * , ( ) ] | the optical properties of single crystals of the high - temperature superconductor la@xmath0ba@xmath1cuo@xmath2 have been measured over a wide frequency and temperature range for light polarized in the _ a - b _ planes and along the _ c _ axis .
three different ba concentrations have been examined , @xmath3 with a critical temperature @xmath4 k , @xmath5 where the superconductivity is dramatically weakened with @xmath6 k , and @xmath7 with @xmath8 k. the in - plane behavior of the optical conductivity for these materials at high temperature is described by a drude - like response with a scattering rate that decreases with temperature . below @xmath9 in the @xmath3 and 0.145 materials
there is a clear signature of the formation of a superconducting state in the optical properties allowing the superfluid density ( @xmath10 ) and the penetration depth to be determined . in the anomalous 1/8 phase , some spectral weight shifts from lower to higher frequency ( @xmath11 @xmath12 ) on cooling below the spin - ordering temperature @xmath13 k , associated with the onset of spin - stripe order ; we discuss alternative interpretations in terms of a conventional density - wave gap versus the response to pair - density - wave superconductivity .
the two dopings for which a superconducting response is observed both fall on the universal scaling line @xmath14 , which is consistent with the observation of strong dissipation within the _ a - b _ planes .
the optical properties for light polarized along the _ c _ axis reveal an insulating character dominated by lattice vibrations , superimposed on a weak electronic background .
no josephson plasma edge is observed in the low - frequency reflectance along the _ c _ axis for @xmath15 ; however , sharp plasma edges are observed for @xmath3 and 0.145 below @xmath9 . |
x - ray observations have shown that hot gas in groups and clusters of galaxies has been heated by some sources in addition to gravity .
this was shown by the fact that the luminosity and temperature of a group or cluster follow a scaling relationship ( the @xmath3@xmath4 relation ) , @xmath5 , @xcite , which is at odds with that expected for groups and clusters formed by gravitational structure formation , with @xmath6 @xcite .
more recently , it was shown that the entropies of the hot gas in groups and clusters , especially groups , are higher than those predicted by models of gravitational structure formation @xcite .
supernova - driven galactic winds have been considered the heating source @xcite .
however , the energy from supernovae alone seems to be insufficient to heat the hot gas to the observed level @xcite .
therefore , active galactic nuclei ( agns ) are now recognised as another promising candidate of the heating source @xcite .
agns are often found at the centres of groups and clusters @xcite .
the energy ejected by an agn can cancel radiative cooing of the hot gas and may prevent development of a cooling flow in the central region of a group or a cluster . for most of the agns at @xmath7 , however , the power is not enough to heat the hot gas on a group or cluster - scale , and would be insufficient to account for the entropy excess found in groups and clusters .
recently , however , shock waves associated with extremely powerful agn outbursts have been found in some clusters @xcite .
they seem to be powerful enough to heat the gas on a group or cluster - scale .
such strong outbursts would also affect the surrounding environment in a form other than heating .
for example , @xcite showed that the shock wave excited by an outburst accelerates particles and that the emission from the accelerated particles could be responsible for radio mini - halos observed in clusters .
@xcite discussed the regulation of galaxy formation by outbursts . in this letter
, we consider another effect of strong agn outbursts on the environment .
we focus on the interaction between the shock wave produced by an outburst and the galaxies surrounding the agn .
we consider an agn and galaxies in a small group of galaxies with the mass of @xmath8 rather than a cluster of galaxies . in clusters ,
another interaction between the hot gas and the galaxies , called ram - pressure stripping , is effective ( see section [ sec : discuss ] ) .
thus , the influence of agn outbursts would be obscured .
it should be noted that such strong outbursts are rare phenomena at low redshift .
@xcite estimated that strong outbursts ( @xmath9 erg ) are likely to occur only @xmath10% of clusters at low redshift .
for groups , the fraction may even be smaller because such outbursts have not been observed .
however , at high redshift , they would be more common ( * ? ? ?
* see section [ sec : discuss ] ) . in this letter ,
the cosmological parameters are @xmath11 , @xmath12 , and @xmath13 , where @xmath14 .
using one - dimensional hydrodynamic simulations , we estimate the impact of the shock wave created by an agn outburst on galaxies .
we assume that an agn is located at the centre of a group of galaxies with the virial mass of @xmath15 and that the group is spherically symmetric . for the mass distribution of the group
, we adopt the so - called nfw profile @xcite , although later studies indicated that the central cusp would be steeper ( e.g. * ? ? ?
the mass profile is written as @xmath16\:,\ ] ] where @xmath17 is the characteristic radius of the group .
the normalisation can be given by @xmath18 , where @xmath19 is the virial radius of the group .
the virial radius is given by @xmath20^{1/3}\:,\ ] ] where @xmath21 is a spherical over - density of the virialized dark halo within @xmath22 at redshift @xmath23 , in units of the critical density of the universe at @xmath23 , or @xmath24 .
for @xmath21 , we use the fitting formula of @xcite for a flat universe with a non - zero cosmological constant , @xmath25 , where @xmath26 and @xmath27 is the cosmological density parameter at redshift @xmath23 .
the concentration parameter of the group , @xmath28 , is given by @xmath29 @xcite .
initially , hot gas or intragroup medium ( igm ) is in pressure equilibrium with the gravitational potential formed by the group .
we assume that the initial igm density and temperature profiles follow the ` universal profile ' derived by @xcite .
they can respectively be written as @xmath30 @xmath31 where @xmath32\:,\ ] ] @xmath33 the parameters @xmath34 and @xmath35 can be derived from the condition that the igm and dark matter profiles are the same in the outermost region of the group @xcite .
we solve the following equations : @xmath36 @xmath37 @xmath38 = -n_\rmn{e}^2 \lambda(t_\rmn{igm } ) \nonumber\\ -\rho_\rmn{igm } v_\rmn{igm } g m(r)/r^2 \;,\end{aligned}\ ] ] where @xmath39 is the gravitational constant , and @xmath40 and @xmath41 are the pressure and velocity of the igm , respectively
. the total energy is defined as @xmath42 , where @xmath43 .
although we include the cooling function @xmath44 , the evolution of the shock is faster than the radiative cooling .
the electron density is defined as @xmath45 , where @xmath46 is the proton mass .
we ignore the self - gravity of igm . the cold interstellar medium ( ism ) in the disc of a galaxy
would be stripped when a shock wave excited in the igm passes the galaxy .
since the shock wave passes the galaxy in a short time ( less than the dynamical time or the rotation time of the galaxy ) , momentum transfer causes stripping to occur .
the integrated momentum from the igm per unit area at radius @xmath47 is @xmath48 thus , the criterion of stripping for a galaxy at radius @xmath47 is @xmath49 where @xmath50 is the column density of the cold ism in the galactic disc , and @xmath51 is the escape velocity of the galaxy . strictly speaking ,
this relation is valid only when the galaxy is face - on ; if not , the stripping would be less efficient . in this relation
, we assume that the galaxy is not moving relative to the group .
the agn explodes at @xmath52 .
the escape velocity is given by @xmath53 , where @xmath54 is the rotation velocity of the galaxy .
( the ism may stay in the halo of the galaxy with this velocity . )
@xcite indicated that for a given @xmath54 the total disc surface density of a galaxy , @xmath55 , is proportional to the hubble constant at redshift @xmath23 : @xmath56^{1/2}\ ] ] following @xcite , we assume that the ism column density is proportional to the disc surface density ( @xmath57 ) . thus , the former is given by @xmath58 moreover , we assume that the disc radius of a galaxy has a relation of @xmath59^{-1}\:\ ] ] @xcite .
we consider the influence on galaxies in groups exerted by an extremely strong agn outburst that has not been observed in low - redshift groups .
we assume that the mass of a galaxy group is @xmath60 .
we set @xmath61 and @xmath62 kpc .
we fix the rotation and escape velocities of the galaxy at @xmath63 and @xmath64 , respectively .
these parameters are those for the galaxy ( e.g. * ? ? ?
since we do not know much about the sources of the non - gravitational heating in groups and clusters ( section [ sec : intro ] ) , we assume that the igm has not been non - gravitationally heated at @xmath52 for the sake of simplicity .
in other words , the igm is non - gravitationally heated for the first time by the agn outburst we consider below .
thus , we assume that the mass fraction of the igm in the group is the same as the baryon fraction of the universe and is 0.15 .
if the igm has been non - gravitationally heated , the mass fraction and density of the igm would be lower and the influence of the shock wave on galaxies would be smaller .
first , we assume that the agn at the group centre ejects an energy of @xmath65 erg , which is the one estimated for the cluster ms 0735.6@xmath667421 @xcite .
the energy is kinematically given to the igm for @xmath6720 kpc at @xmath52 ; the details of the energy input do not affect the results .
we use 1000 unequally spaced meshes in the radial coordinate to cover a region with a radius of 600 kpc .
the inner boundary is set at @xmath68 kpc .
[ fig : z0](a ) shows the evolution of the igm density profile for the group at @xmath69 .
the parameters of the group are @xmath70 kpc and @xmath71 .
the outburst is strong enough to blow away most of the igm .
the shock reaches the virial radius at @xmath72 yr .
[ fig : z0](b ) shows the evolution of @xmath73 , where @xmath74 ( see equation [ eq : criterion ] ) .
[ fig : z0](b ) indicates that the cold ism in the galactic disc is stripped for @xmath75 kpc because @xmath76 at @xmath77 .
the profile @xmath78 does not change after the shock passes the outermost region of the group .
we refer to this final profile as @xmath79 [ @xmath80 . of course , the ism of galaxies with @xmath81 is more easily stripped for a given @xmath50 .
[ fig : z2 ] shows the evolutions at @xmath82 .
the parameters of the group are @xmath83 kpc and @xmath84 .
the shock reaches the virial radius at @xmath85 yr . the profile @xmath86 in fig .
[ fig : z2](b ) indicates that the ism in the galactic disc is stripped for @xmath87 kpc .
we also considered the case when the energy ejected by the agn is smaller .
[ fig : z2_01 ] shows the evolutions at @xmath82 when @xmath88 erg .
the shock reaches the virial radius at @xmath89 yr .
since @xmath90 in the entire group , the ism in the galactic disc is not stripped ( fig .
[ fig : z2_01]b ) . .
( a ) the density profiles at @xmath52 , 5 , 20 , and 80 myr .
the curve at @xmath91 myr is omitted at small @xmath47 because it is affected by the inner boundary condition .
( b ) the profiles of @xmath73 at @xmath92 , 20 , 80 myr , and @xmath77.,width=279 ] but for @xmath82.,width=279 ] but for @xmath82 and @xmath93 erg.,width=279 ]
we found that in the inner region of a group , the cold ism in the disc of a galaxy is stripped by the shock wave if @xmath94 .
we compare the effect with that of the usual ram - pressure stripping through the motion of a galaxy in the igm .
the typical velocity of galaxies in a group is @xmath95 for the group with @xmath96 , the velocity is @xmath97 at @xmath69 and @xmath98 at @xmath82 .
the condition of usual ram - pressure stripping owing to the motion of a galaxy is determined by the long - term ( larger than the dynamical time of the galaxy ) balance between the ram - pressure from the igm and the gravity of the galaxy .
thus , the ram - pressure stripping is effective when @xmath99^{-1 } \left[\frac{\sigma_{\rm ism}(z ) } { 8\times 10^{20 } m_{\rm p}\;\rm cm^ { -2}}\right ] \label{eq : grav3}\:\end{aligned}\ ] ] @xcite . from equations ( [ eq :
sig_ism ] ) and ( [ eq : r_gal ] ) , one finds that the right hand of the relation is proportional to @xmath100 for a given @xmath54 .
thus , from equations ( [ eq : vgal ] ) and ( [ eq : strip ] ) , the ram - pressure stripping is effective for @xmath101 at @xmath69 and @xmath102 at @xmath82 .
[ fig : z0](a ) and [ fig : z2](a ) indicate that the effect of the ram - pressure stripping can be ignored .
this means that the stripping by a shock wave overwhelms the usual ram - pressure stripping in groups .
however , in groups and clusters with larger @xmath15 , the ram - pressure stripping should be more effective because the typical velocity of galaxies , @xmath103 , is larger .
even if @xmath104 , the shock wave would affect the evolution of a galaxy .
it is often assumed that the cold ism in the disc of a galaxy is supplied from the warm gas in the halo .
if most of the gas in the halo is stripped , the ism in the disc is gradually consumed through star formation without supply .
this is followed by gradual decrease of star formation activity of the galaxy , which is often called ` strangulation ' and is considered to be associated with the evolution of galaxies in galaxy groups @xcite . from x - ray observations , @xcite actually detected the halo gas in a massive galaxy ngc 5746 .
however , the rotation velocity of the galaxy is very large ( @xmath105 ) and the result could not apply to the galaxy we consider here ( @xmath106 ) .
therefore , we use the results of the observations of ngc 5170 ( @xmath107 ) for which the halo gas was not detected @xcite . for our cosmological parameters ,
the upper limit of the mass of the halo gas is @xmath108 @xcite . assuming that the halo gas resides in a cylinder of a base radius of 38 kpc @xcite , the column density in the direction of the height is @xmath109 .
therefore , we assume that the column density of the halo gas is @xmath110 and @xmath111 as equation ( [ eq : sig_ism ] ) .
since @xmath112 for our model galaxy , the halo gas would be stripped even when @xmath113 or smaller . for @xmath114 , fig .
[ fig : z0](b ) and fig .
[ fig : z2](b ) show that @xmath115 up to the virial radius of the group ( 560 kpc for @xmath69 and 230 kpc for @xmath82 ) .
thus , most galaxies in the group would be affected by the outburst of the central agn .
even if @xmath116 , the influence of the agn outburst can not be ignored for @xmath117 ( fig .
[ fig : z2_01]b ) .
after the cold ism ( or the halo gas ) is stripped by the outburst , the colour of the galaxies would become red due to the lack of star formation .
some of the galaxies would be observed as passive spiral galaxies ( e.g. * ? ? ?
. some would also fall into the group centre through dynamical friction , and eventually merge with the central galaxy in which the agn responsible for the outburst resides .
since the fallen galaxies no longer have ism , they do not supply gas to the supermassive black hole in the agn .
thus , the outburst would serve as a kind of feedback mechanism for the growth and activity of the black hole . at low redshift , radio and x - ray observations
indicate that the stripping by agn outbursts may not be common . after the outburst , the x - ray luminosities of the groups we considered reduce to @xmath118% of the initial values and become @xmath119 for @xmath120 . here
, we do not integrate the luminosities up to @xmath19 , because it is difficult to detect x - ray in the outermost regions of groups @xcite .
the luminosities of the igm after the outburst are smaller than those of groups from which x - ray emission from the igm has been detected ( @xmath121 ; @xcite ) .
if the stripping by agn outbursts were common , the lack of cold ism in galaxies should be confirmed more in groups lacking for x - ray emission .
however , radio ( hi ) and x - ray observations of nearby groups showed the opposite trend @xcite , although the cause is not known . on the other hand , we suppose that the stripping by agn outbursts occurred more often at high redshift ( say @xmath122 ) .
this is because strong outbursts were more common at that time .
for example , radio observations showed that there are a number of agns with the jet power of @xmath123 at @xmath124 @xcite .
considering typical duration of an agn activity ( @xmath125@xmath126 yr ; @xcite ) , the total energy injected through an activity is @xmath127 erg .
moreover , it has been indicated that luminous agns tend to be found at higher redshifts ( @xmath122 ) in comparison with less luminous agns @xcite .
furthermore , optical observations often found clustering of red galaxies around radio galaxies at @xmath128 ( e.g. * ? ? ?
* ; * ? ? ?
the activities of the radio galaxies might have affected the star formation of the surrounding galaxies as we predict . from a theoretical point of view
, @xcite indicated that strong agn outbursts at @xmath128 blew out the igm from groups ( ancestors of present - day groups or clusters ) , and they could be responsible for the low metal abundance observed in groups ( @xmath129 ) at @xmath7 @xcite .
although these studies do not directly prove the stripping by agn outbursts , they are at least consistent with our model . in the future , galaxies lack of cold gas would be observed inside the cocoon or the shock produced by an outburst .
we have investigated the influence of an agn outburst at the centre of a galaxy group on the galaxies surrounding the agn . if the energy of the agn is extremely large ( @xmath130 ) as the one recently observed in clusters , the shock wave excited by the outburst strips the cold ism in the disc of the surrounding galaxies locating at @xmath131200 kpc from the agn .
the effect of the stripping can be much stronger than that of usual ram - pressure stripping owing to the motion of the galaxies in a group . however , the stripping of the cold ism in the disc is not effective if @xmath132 .
we also showed that even if the cold ism in the disc of the galaxies is not stripped , the warm gas in the halo would be stripped by the shock wave .
this could be effective for the whole galaxies in the group up to the virial radius even when @xmath133 .
the star formation activities of those galaxies would decrease gradually because the ism consumed by the star formation is not compensated from the gas in the halo ( strangulation ) .
our results indicate that even one strong agn outburst significantly affects the evolution of galaxies in groups .
after the stripping , the galaxies should become red because of the decrease of the star formation rate .
the success of this model depends on the assumption of an extremely strong outburst . since such outbursts
have not been observed in galaxy groups at low - redshift , the stripping by them would not have strong influence on the evolution of galaxies at low - redshift .
however , we suppose that it would be at high redshift where extremely strong agn activities have often been observed .
the author wishes to thank the referee for useful comments .
y. f. was supported in part by grants - in - aid from the ministry of education , science , sports , and culture of japan ( 17740162 ) . | we study the influence of a strong agn outburst on the surrounding galaxies .
the agn is assumed to reside in a group of galaxies , and an outburst excites a shock wave in the hot gas in the group .
we calculate the impact of the shock wave on the galaxies .
we find that if the energy of the outburst is extremely large ( @xmath0 erg ) as the one recently observed in clusters , the impact is strong enough to strip the cold interstellar medium in the disc of the galaxies in the inner region of the group .
moreover , even in the outer region of the group , the warm gas in the halo of the galaxies would be stripped , even if the energy of the outburst is @xmath1 erg .
these would decrease star formation activity of the galaxies .
if these galaxies fall into the group centre through dynamical friction and their interstellar medium is the fuel of the supermassive black hole in the agn , the outburst would serve as feedback . while this mechanism works only when @xmath2 is extremely large , such outbursts
have not been observed in groups at low redshift ; it would work at high redshift rather than at low redshift .
[ firstpage ] galaxies : active galaxies : clusters : general galaxies : interactions galaxies : intergalactic medium . |
coronal mass ejections ( cmes ) are a major source of potentially destructive space weather conditions ( e.g. , geomagnetic storms , solar energetic particles ) . due to our increasing dependence on advanced technology , which is vulnerable to severe space weather conditions
, there is a high national priority to establish a reliable space weather forecasting capability .
however , available cme observations that may provide a basis for forecasts are very limited . in particular , erupting magnetic fields can not be directly observed in the solar corona , which is of critical importance given that interplanetary magnetic field ( imf ) is a major driver of the geomagnetic storms . for these reasons , there is a great need to be able to predict cme magnetic fields based on photospheric observations , both for scientific understanding and for forecasting purposes . in the past two decades
, many cme forecasting models have been developed , which can be divided mainly into three different categories .
the first category is empirical forecasting models , which use near - sun cme observations to estimate the arrival time of cmes at 1 au through empirical relations built through a large number of observations ( e.g. , @xcite ) .
recently , the data - mining techniques have been utilized to establish empirical models ( riley et al .
. the second category is called kinematic models , which solve a reduced form of fluid equations without dynamics .
successful examples include the 3-d hakamada - akasofu - fry version 2 ( hafv.2 ) model , in which remote - sensing data is used to derive the shock speed and direction @xcite , and the cone model , which fits cme observations with three free parameters : angular width , speed , and central cme position @xcite .
the cone model has been widely used by the research community to predict the cme / cme - driven shock velocity ( e.g. , @xcite ) . in order to provide more accurate forecasts as well as the cme plasma parameters ( i.e. , density , temperature , velocity ) in addition to the arrival time , the third category of forecasting models couple kinematic models with heliospheric magnetohydrodynamics ( mhd ) models ( inner boundary outside magneto sonic point ) . in this case , the kinematic models are used to prescribe the initial conditions of the mhd models . by combing the cone model with enlil @xcite , the average error in the cme analysis tool ( cat ; @xcite)-wang - sheeley - arge ( wsa ; @xcite)-enlil operational forecasting model at space weather prediction center ( swpc ) is @xmath07.5 hours , which represents the state - of - the - art cme forecasting model in operation .
more recent mhd models introduce the flux rope beyond the alfvn surface allowing for prediction of the magnetic field at 1 au . by utilizing a spheromak cme model in an mhd inner heliosphere solar wind model
, @xcite demonstrated that the model is capable of predicting magnetic profile at the earth .
however , since the model starts outside the alfvn surface , the cme model is less constrained by observations . to provide magnetic field forecasting ,
several models have been developed which self - consistently simulate the cme from the corona to 1 au by combining realistic mhd corona models ( e.g. , @xcite ) and magnetically driven eruptions ( e.g. , @xcite ) .
this approach represents the most sophisticated cme propagation model so far . and
many case studies have provided high fidelity simulations both at the sun and at 1 au ( e.g. , @xcite ) . as the first step to transfer a research model to operational forecasting
, it is important to characterize the behavior of the magnetic - driven eruptions in the ambient solar wind solutions . using wsa - enlil modeling system ,
@xcite suggest that the cme propagation is non - chaotic and relates to a finite set of inputs in the mhd simulations .
therefore , it is possible to control the cme behavior through varying the initial cme parameters to match observations . in our study
, we follow this path by combing the alfvn wave solar model ( awsom ; @xcite ) with the gibson - low ( gl ) analytical flux rope model , and we present a method to determine the initial flux rope parameters from available observations so that the cme speed can be well reproduced near the sun . with the cme structure
self - consistently propagating into the heliosphere , the model has the capability to follow the erupting magnetic field from the sun to 1 au and provide a forecast of cme arrival time and all plasma parameters ( i.e. , velocity , density , temperature , and magnetic field ) . the paper is organized as follows : in section 2 , we briefly describe the awsom model used to construct the background solar wind and the gl flux rope model for the cme initiation . the methodology of determining the gl flux rope parameters will be described in detail in section 3 , followed by the discussion and summary in section 4 .
here , we describe the eruptive event generator using gibson - low configuration ( eeggl ) , an automated tool for finding parameters of the gl flux rope to reproduce observed cme events . in figure 2 ,
an example of the gl flux rope is shown embedded into the global solar wind solution of carrington rotation ( cr ) 2107 .
figure 4 shows a chart demonstrating how cme events can be simulated in the awsom .
first , the synoptic magnetogram is used to specify the inner boundary condition of the magnetic field for awsom , which is then employed to generate a steady state solar wind solution . at the same time , the input magnetogram and observed cme speed ( from soho / lasco and/or stereo / cor observations ) are used by eeggl to determine the gl flux rope parameters . with the derived parameters ,
a gl flux rope is inserted into the steady state solar wind to initiate the cme event . in this section ,
we briefly describe the awsom and gl analytical flux rope model .
the mhd solar wind model used in this study is the recently developed awsom @xcite , which is a data - driven model with a domain extending from the upper chromosphere to the corona and heliosphere .
a steady - state solar wind solution is obtained with local time stepping and second - order shock - capturing scheme @xcite . the inner boundary condition for the magnetic field
can be specified by different magnetic maps ( e.g. , from gong , soho / mdi , or sdo / hmi ) . in this study ,
the magnetograms from gong are used .
the inner boundary conditions for electron and proton temperatures @xmath1 and @xmath2 and number density @xmath3 are assumed at @xmath4 50,000 k and @xmath5 2@xmath610@xmath7 m@xmath8 , respectively . at the base of the atmosphere ,
the temperature is held fixed at 50,000 k while the density falls off exponentially until it reaches a level where the radiative losses are sufficiently low that the temperature increase monotonically with height . above this height , the temperature increases rapidly forming the transition region .
this procedure allows chromospheric evaporation to self - consistently populate the corona with an appropriately high plasma density .
the inner boundary density and temperature do not otherwise have a significant influence on the global solution @xcite .
the initial conditions for the solar wind plasma are specified by the parker solution @xcite , while the initial magnetic field is based on the potential field source surface ( pfss ) model solved with the finite difference iterative potential solver ( fdips , @xcite ) . in the awsom ,
alfvn waves are represented by energy densities of counter - propagating waves .
the dissipation is based on the interaction between the counter - propagating waves with self - consistent wave reflection .
alfvn waves are specified at the inner boundary with a poynting flux that scales with the surface magnetic field strength .
the solar wind is heated by alfvn wave dissipation and accelerated by thermal and alfvn wave pressure .
electron heat conduction ( both collisional and collisionless ) and radiative cooling are also included in the model .
these energy transport terms are important for self - consistently creating the solar transition region . in order to produce physically correct solar wind and cme structures , such as shocks ,
the electron and proton temperatures are treated separately @xcite .
thus , while the electrons and protons are assumed to have the same bulk velocity , heat conduction is applied only to the electrons , owing to their much higher thermal velocity @xcite . by using a physically consistent treatment of wave reflection , dissipation , and heat partitioning between electrons and protons ,
the awsom has demonstrated the capability to reproduce the solar corona environment with only 3 free parameters that determine poynting flux ( @xmath9 ) , wave dissipation length ( @xmath10 ) , and stochastic heating parameter ( @xmath11 ) @xcite . in figure 1 ,
an example of a steady state solar wind solution is shown .
panel ( a ) shows the steady state solar wind speed of a meridional slice at @xmath12 .
panel ( b ) shows the 3d field configuration near the sun with large - scale helmet streamer belt and open / active region field lines marked by white and green , respectively .
carrtington coordinates are used , in which the z axis corresponds with the rotation axis . also , in figure 2(b)-(d ) , a steady state solar wind solution of velocity , proton , and electron temperatures near the sun are shown .
the electron and proton temperatures are very close near the sun due to collisions and they diverge further away from the sun .
in this study , we initiate cmes using the analytical gibson - low ( gl ; @xcite ) flux rope model implemented in the eruptive event ( ee ) of space weather modeling framework ( swmf ; @xcite ) .
this flux rope model has been successfully used in numerous modeling studies of cmes ( e.g. , @xcite ) .
the gl flux rope are obtained by finding an analytical solution to the magnetohydrostatic equation @xmath13 and the solenoidal condition @xmath14 .
this solution is derived by applying a mathematical stretching transformation @xmath15 to an axisymmetric solution describing a twisted toroidal flux rope contained within a sphere of radius @xmath16 centered relative to the heliospheric coordinate system at @xmath17 ( black contour in figure 2(a ) ) .
the full 3d field of @xmath18 can be expressed by a scalar function @xmath19 and a free parameter @xmath20 that determines the magnetic field strength @xcite .
the transformed flux rope appears as a tear - drop shape of twisted magnetic flux ( red contour in figure 2(a ) ) . at the same time , lorentz forces are introduced , which leads to a density - depleted cavity in the upper portion and a dense core approximating a filament at the lower portion of the flux rope .
this flux rope structure helps to reproduce the commonly observed 3-part density structure of the cme @xcite . in sum
, the gl flux rope is mainly controlled by five parameters : the stretching parameter @xmath21 determines the shape of the flux rope , the distance of torus center from the center of the sun @xmath22 determines the initial position of the flux rope before it is stretched , the radius of the flux rope torus @xmath23 determines the size of the flux rope , the flux rope field strength parameter @xmath20 determines the magnetic strength of the flux rope , and a helicity parameter to determine the positive ( dextral)/negative ( sinistral ) helicity of the flux rope .
the relative magnitudes of b@xmath24 and b@xmath25 at any point in the flux rope model are of fixed functional forms such that the magnetic field makes a transition from pure toroidal field at the core to pure poloidal field at the outer surface .
the gl flux rope and contained plasma are then superposed onto the steady state solar corona solution : i.e. @xmath26 , @xmath27 , @xmath28 .
the combined background - flux rope system is in a state of force imbalance ( due to the insufficient background plasma pressure to offset the magnetic pressure of the flux rope ) , and thus erupts immediately when the numerical model is advanced forward in time . in figure 2(a ) , the gl flux rope in 2d is shown with magnetic field lines and plasma density .
we can see that the main flux rope structure is embedded in a coronal streamer with high plasma density while a low density cavity fills the outer part of the rope and dense filament material fills the bottom . in figure 2(b)-(d ) , the 3d gl flux rope is shown with velocity , proton temperature , and electron temperature on the central meridional plane of the flux rope . in figure 3 , we show three gl flux ropes with different size ( @xmath16 ) and magnetic field strength ( @xmath20 ) parameters .
the left panel shows the initial configuration of the gl flux ropes .
the blue and red isosurfaces represent a density ratio of 0.3 and 2.5 between the solutions before and after the flux rope insertion .
the middle panel shows the resulting cme evolution at 10(20 ) min .
the background color shows the density ratio between the solution at 10(20 ) min and the background solar wind .
the red , white , and green field lines represent cme flux rope field lines , large - scale helmet streamers , and field lines from surrounding active regions and open field .
the right panel shows the synthesized soho / lasco white light images .
the color scale shows the white light total brightness divided by that of the pre - event background solar wind .
the first ( @xmath29 , @xmath30 ) and second ( @xmath29 , @xmath31 ) cases have the same flux rope size but different magnetic field strength . comparing figure 3(a ) and 3(d )
, we can see that with a higher magnetic field strength parameter , more plasma is added at the bottom of the flux rope ( red isosurface ) .
the slight size difference is due to the background field .
since the gl flux rope solution is added to the background solar wind solution , the field near the gl flux rope boundary may be affected with weak magnetic field strength parameter .
after initiation , the higher flux rope magnetic field leads to a higher cme speed and stronger density pile - up in front of the flux rope .
the second and third ( @xmath32 , @xmath31 ) cases have the same magnetic field strength parameter but with different flux rope size . in this case
, we can see the flux rope is considerably smaller at the beginning . with this smaller flux rope ,
the resulting cme speed is reduced and the morphology of cme in the synthesized white light image is quite different with narrower cme width angle .
there are several advantages of using this force imbalanced flux rope for this study .
first , the flux rope fits into an active region with free energy for eruption , which does not require time - consuming energy build - up process in the simulation .
second , the gl flux rope comes with dense plasma over the polarity inversion line ( pil ) and a low density cavity above it , which mimics cme observations .
third , the eruption speed can be controlled by the gl flux rope parameters .
moreover , previous studies have shown that the gl flux rope is capable of producing magnetic cloud signature at 1 au in the simulations ( e.g. , @xcite ) .
eeggl is designed to determine the gl flux rope parameters , including flux rope location , orientation , and five parameters to control the characteristics of the flux rope in order to model specific cme events .
eeggl operates under the assumption that the cmes originate from closed corona flux forming loop structures anchored to the photosphere in bipolar active regions . from eeggl s graphical interface ( shown in figure 5 )
, the user selects the active region from which the cme originates .
eeggl then calculates the weighted center for both positive and negative polarities of the source region . at the same time , eeggl determines the pil locations of the source region .
the flux rope location then can be determined by the intersection between the pil and the line connecting positive and negative weighted centers .
also , the orientation of the flux rope is set to be parallel to the pils . in the cases that the pils are complex curves ,
a straight line is fitted using the pil points near the flux rope center to represent the pils . in case of non - bipolar regions , the pil that is responsible for the eruption
will be used to determine the orientation of the flux rope .
one example of non - bipolar source region is the ar0 in figure 5 .
the pil of ar0 has an up - side - down
l " shape . when the eruption site is determined from the observation , the gl flux rope orientation and the source region size are calculated by the erupting part of the pil . the flux rope helicity ( dextral / sinistral )
is determined according to the hemispheric helicity rule that the dextral / sinistral filament dominates the northern / southern hemisphere .
a recent study by @xcite found that @xmath33 of 151 ars observed with hmi obey this helicity rule .
several examples are shown in figure 5 , in which the gong carrington magnetogram of cr 2107 is shown180 . the latitude grid is evenly spaced in sine latitude . ] .
the active region markers ( ar0 ar6 ) represent the cme source regions we used for the parameter study .
note that , to reduce the complexity of the photospheric magnetic field configuration , we smooth the magnetogram by a 5@xmath65 pixel window . of the four gl flux rope parameters , eeggl calculates the flux rope size @xmath23 and flux rope field strength @xmath34 and fixes the other two parameters ( @xmath22=1.8 , @xmath21=0.6 ) .
the reason for fixing these two parameters is that they can not be easily related to or obtained from normal observations . on the other hand , by varying these parameters , two main cme characteristics ( cme width and speed ) can be adjusted to match observations .
the fixed values were selected a priori for optimal performance , in particular how it relates to the amount of disconnected flux and the cme mass .
the mass falls in the range of typical cmes , of the order of 10@xmath35 grams @xcite and increases with field strength .
this is the best we can do , since we can not regularly determine the filament mass of the cme . in order to find a useful empirical relationship between the gl flux rope parameters and the resulting cme speeds
, we examined the observational data in @xcite where we found an approximately linear relationship between cme speed and reconnected flux .
the data from @xcite is replotted in figure 6 . from this result , we searched for an empirical relationship between cme speed and poloidal magnetic flux of the gl flux rope
. the poloidal flux can be calculated by integrating @xmath36 over a central surface across magnetic axis of gl flux rope .
the expression of @xmath36 can be found in the appendix of @xcite . with the two gl flux rope parameters fixed ( @xmath21 and @xmath37 ) ,
the gl poloidal flux is only determined by the size ( @xmath16 ) and magnetic strength ( @xmath20 ) parameters : @xmath38 , where @xmath39 is a constant .
we initiate cmes with different flux rope sizes and magnetic field strengths and run each simulation 30 minutes after the flux rope insertion .
we then derive the cme speeds in the simulation taken as the average speeds of the cme at the outer - most front between 20 and 30 minutes .
the outermost front of the cme is determined by finding the cme propagation plane and extracting the line profiles along the cme propagation path ( see @xcite for an example ) .
the typical plasma parameters we use to identify the outermost front are proton temperature , velocity , or density profiles .
the 20 - 30 min window is chosen because during that period of time the cme is reaching a nearly constant speed both in the simulation and in the observation .
therefore , the speed obtained is more reliable and consistent . according to the method we used to determine the cme speed , the major source of uncertainty comes from the grid size at the outermost front in the simulation that can affect the determination of the exact outermost point on the profile ( e.g. , figure 5 and 6 in @xcite ) .
therefore , we estimate the uncertainty in velocity by @xmath40 , where @xmath41 and @xmath42 are the cme front grid size at 20 min and 30 min , respectively .
@xmath43 is 600 s in this case . to investigate the relationship between the poloidal flux of the gl flux rope and the resulting cme speed ,
we initiate 8 cmes from ar0 ( marked in figure 5 ) with different flux rope size and strength parameters .
the cme parameters can be found in table 1 .
the poloidal fluxes vary from @xmath44 to @xmath45 mx , which lead to cme speeds varying from 758 km s@xmath46 to 3150 km s@xmath46 . in figure 6 , the relationship between the input poloidal fluxes and the resulting cme speeds in the simulation is shown , in which we find a very good linear relationship between the two variables .
the red dashed line shows a linear fit result .
we next investigate how the same flux rope behaves in different active regions .
we test all the 7 major active regions in cr2107 magnetogram ( marked in figure 5 ) and the results are shown in table 2 and figure 7(b)/(d ) .
the flux rope parameters ( @xmath23=0.8 , @xmath34=1.6 ) are fixed for all the 7 runs .
the average magnetic field @xmath47 ( measured both around the flux rope center and along the pil of the source region ) is used to quantify the source region . with the same flux rope , we find that the cme speed is inversely related to the average @xmath47 of the active region through a nonlinear relationship that is not as well defined as the cme speed - poloidal flux relationship ( figure 6 ) .
the spread of points suggest a dependence of cme speed on more complex features of the active region that are not captured by the average field strength . with these two empirical relationships obtained from the parameter study ,
it is possible to derive an equation to calculate the gl flux rope parameters based on the observed cme speed and average magnetic field of the source region : @xmath48 where @xmath34 is the parameter that determines the gl flux rope field strength , @xmath49 is the observed cme speed near the sun , @xmath50 is the area average radial field strength of the source region ( @xmath51 ) , @xmath23 is the size of the gl flux rope , @xmath52 , @xmath53 and @xmath54 are the fitting constants . the observational cme velocity can be obtained through stereocat cme analysis tool at ccmc .
this tool provides the measured cme speed through stereoscopic reconstruction using both soho / lasco and stereo / cor observations . for determining the gl flux rope size , we investigate two different algorithms : ( 1 ) relate the flux rope size to the length of the pil ; ( 2 ) relate the flux rope size to the size of the source region
. since the complexity of the pils , the first algorithm is not robust and may lead to extremely large flux ropes .
therefore , we use the second algorithm to calculate the gl flux rope size in this study . in figure 5 , we show all the 7 source region boundaries in cyan contours calculated by the algorithm . in case of
very large or very small complex non - bipolar source regions , the gl flux rope size may be over- or under - estimated .
we set up a lower and upper limit of gl flux rope size ( @xmath55 $ ] ) .
note that , since the cme speed in the simulation is determined by gl poloidal flux , the uncertainty of flux rope size will not influence the final cme speed . in figure 7
, we show the fitting results based on the current tests . in figure 7(a)-(b ) , the average @xmath56 is calculated around the center of the source region ( @xmath57 in latitude and longitude ) .
the fitted constants are : @xmath58 , @xmath59 , and @xmath60 . in figure 7(c)-(d ) , the average @xmath56 is calculated along the pil is chosen by @xmath61 pixel around the pil . ] .
the fitted constants are : @xmath62 , @xmath63 , and @xmath64 .
the fitting curves based on eq .
( 1 ) are shown in red dashed lines .
the fitting results are very close for both methods of determining the average @xmath56 .
although the current empirical relationship is obtained from limited number ( 14 ) of runs , there is no sudden change in the curves shown in figure 6 and 7 that needs to be resolved .
therefore , the empirical relationship is expected to be valid for the parameter space of moderate to fast cmes .
we note that the fitted empirical relationship may not work for slow cmes in a weak source region , which may lead to negative calculated flux rope magnetic strengths .
for example , if the source region has an average @xmath47 of 10 gauss , then the input cme speed can not be less than 740 km s@xmath46
. the eeggl will give an error message when negative magnetic strength are calculated . to further validate this empirical relationship obtained from cr 2107 active regions , we choose 4 different crs and initiate 5 cmes using eeggl calculated flux rope parameters .
the input cme speeds into the eeggl are shown in the third column of table 3 . with each cme simulation runs to 30 min
, we calculate the actual cme speed in the simulation taken as the average cme speed at the outer - most front between 20 and 30 minutes , the same definition as cr 2107 runs .
then the simulated and input cme speeds can be compared .
the validation result is shown in table 3 , which confirms that the empirical relationship works well for other crs and different source region configurations .
the error ( defined by the relative difference between the input and simulated cme speeds ) varies from @xmath65 to @xmath66 .
in summary , we present the first data - driven mhd cme model in which the driving flux rope is initiated in the corona and the parameters of the model are constrained by observations near the sun .
we present a new tool , eeggl , which automatically calculates the parameters to model cme events using synoptic magnetogram data along with the observed cme speed .
this cme model and eeggl are now available at community coordinated modeling center ( ccmc ) for runs on demand , where it provides a new forecast capability for the icme magnetic field at 1 au that is critically important for space weather . in a companion paper by @xcite
, we simulate a cme event on 2011 march 7 from the sun to 1 au using awsom+eeggl and present detailed analysis of the results to show the new model can reproduce many of the observed features near the sun and in the heliosphere .
the availability of modeling magnetic eruptions from the low corona to 1 au also will greatly facilitate research on the coronal and heliospheric responses to cmes , as well as provide valuable information for improving flux rope driven numerical models including the empirical relationships applied in eeggl . for simplicity ,
the gl flux ropes used in this study have a fixed positive helicity . in future studies
, we will set the flux rope helicity based on the vector magnetic field observations and investigate how the helicity of the flux rope affects cme propagation .
active region helicity is available as a data product of the space - weather hmi active region patches ( sharps ; @xcite ) derived from sdo / hmi can now provide the helicity information calculated from the vector magnetic field for all major active regions on the sun .
therefore , it is possible to obtain the helicity information derived from sharps and use it in eeggl to specify the direction of the toroidal field in the gl flux rope .
moreover , the hek output of the sdo automatic filament detection @xcite will be utilized to improve the flux rope size calculation . with increasing number of events
studied , the empirical relationship will be improved consistently .
also , more factors ( e.g. , background density , large - scale magnetic structure ) may be considered that affect the cme propagation from the sun to 1 au . with more events studied , we could introduce new parameters to account for these effects and further revise the empirical relationship to determine the gl flux rope parameters . to achieve forecasting ,
the model must run faster than real time .
the computational expense using the current awsom model to simulate a cme event from the sun to 1 au is about 40,000 cpu hours on the nasa pleiades supercomputer .
this means 1000 - 2000 cpus are needed to run the model faster than the real time . with increasing computational power nowadays , the computational requirements may be much more readily available .
moreover , a new model in development , awsom - r ( threaded field line model ; sokolov et al . 2016 ) may run 10 - 100 times faster than the current awsom so that far fewer cpus will be needed for real time simulations .
a preliminary test at the ccmc shows that with 120 cpus on ccmc cluster hilo , the awsom - r steady state simulation takes 17 hours , and the cme evolution of 3 days takes 16 hours .
we are grateful to the anonymous referee for constructive comments that helped improve the paper .
m.jin is supported by nasa / ucar jack eddy postdoctoral fellowship and nasa s sdo / aia contract ( nng04ea00c ) to lmsal .
the work performed at the university of michigan was partially supported by national science foundation grants ags-1322543 , ags1408789 and phy-1513379 , nasa grant nnx13ag25 g , the european union s horizon 2020 research and innovation program under grant agreement no 637302 progress .
we would also like to acknowledge high - performance computing support from : ( 1 ) yellowstone ( ark:/85065/d7wd3xhc ) provided by ncar s computational and information systems laboratory , sponsored by the national science foundation , and ( 2 ) pleiades operated by nasa s advanced supercomputing division .
this work utilizes data obtained by the global oscillation network group ( gong ) program , managed by the national solar observatory , which is operated by aura , inc . under a cooperative agreement with the national science foundation . with magnetic field lines .
the black grid show the simulation cells .
( b ) 3d field configuration of the steady state solution .
the white field lines represent the large - scale helmet streamer belt .
the active region and open fields are marked in green . ] and
@xmath67 . ( a ) initial configuration of the gl flux rope .
the blue and red isosurfaces represent a density ratio of 0.3 and 2.5 between the solutions after and before the gl flux rope insertion .
( b ) the resulting cme evolution at @xmath68 min .
the background shows the density ratio between the solution at 20 min and the background solar wind .
the red , white , and green field lines represent gl flux rope field lines , large - scale helmet streamers , and field lines from surrounding active regions and open field .
( c ) the synthesized soho / lasco white light image .
the color scale shows the white light total brightness divided by that of the pre - event background solar wind .
( d)-(f ) : gl flux rope with @xmath69 and @xmath70 .
( g)-(i ) : gl flux rope with @xmath71 and @xmath70 . ] | we present a first - principles - based coronal mass ejection ( cme ) model suitable for both scientific and operational purposes by combining a global magnetohydrodynamics ( mhd ) solar wind model with a flux rope - driven cme model .
realistic cme events are simulated self - consistently with high fidelity and forecasting capability by constraining initial flux rope parameters with observational data from gong , soho / lasco , and stereo / cor .
we automate this process so that minimum manual intervention is required in specifying the cme initial state . with the newly developed data - driven eruptive event generator gibson - low ( eeggl ) , we present a method to derive gibson - low ( gl ) flux rope parameters through a handful of observational quantities so that the modeled cmes can propagate with the desired cme speeds near the sun . a test result with cmes launched with different carrington rotation magnetograms are shown .
our study shows a promising result for using the first - principles - based mhd global model as a forecasting tool , which is capable of predicting the cme direction of propagation , arrival time , and icme magnetic field at 1 au ( see companion paper by @xcite ) . |
dwarf novae a subclass of cataclysmic variable stars
are quite well studied interacting binary systems composed of late - type red dwarf secondary and white dwarf primary stars ( warner 1995 , hellier 2001 ) .
matter transferred from the red dwarf forms an accretion disc around the white dwarf .
although in the last decade significant progress has been made in explaining the behaviour of dwarf novae light curves , some physical processes ongoing in these systems are still not fully understood ( see for example smak 2000 , schreiber and lasota 2007 ) . in particular , the thermal - tidal instability model of osaki ( 1996 , 2005 ) describing the phenomenon of superoutbursts and superhumps may be tested by examination of su uma - type dwarf novae light curves .
additionally , objects near and inside the so called period gap are very important from an evolutionary point of view .
those systems give us an unprecedented opportunity to study the evolution of dwarf novae .
v419 lyr is a poorly studied cataclysmic variable discovered by kurochkin ( 1990 ) and originally classified as a z cam - type dwarf nova .
later , nogami et al . ( 1998 ) caught this object in outburst and found superhumps in its light curve .
detection of superhumps together with characteristic properties of the outburst allowed them to classify v419 lyr as a su uma - type dwarf nova , but short coverage of the eruption did not allow accurate determination of the superhump period .
nevertheless , there was a strong suggestion that v419 lyr has one of the longest orbital periods known among su uma variables .
this object has been monitored at various photometric bands by the variable star network ( vsnet ) ( see for example kato et al .
the observations from that program enabled a tentative determination of the supercycle period to be about @xmath4 days ( katysheva and pavlenko 2003 ) .
moreover , morales - rueda and marsh ( 2002 ) obtained a spectrum of v419 lyr during outburst showing a relatively broad absorption feature around 430 - 440 nm . in this work we present an analysis of photometric data collected during the 2006 july superoutburst of v419 lyr .
the data are much richer than previous studies and provide us with an opportunity to determine parameters describing this system more precisely .
the curious variable experiment ( curve ) team ( see for example olech et al .
2004 , 2006 ) , alerted by the vsnet mailing list , found v419 lyr in a very bright state on 2006 july 17/18 .
subsequently the object was monitored on 13 consecutive nights ( with a gap on july 24/25 ) until its return to quiescence on august 2/3 .
the observations were performed using a 0.6-m cassegrain telescope equipped with a tektronix tk512cb back - illuminated ccd camera .
the image scale was @xmath5pixel providing a @xmath6 field of view ( udalski and pych , 1992 ) observations were made unfiltered for two reasons .
first , due to lack of an autoguiding system , we wished to keep exposures short in order to minimize guiding errors .
second , because our main goal was analysis of the temporal behaviour of the light curve , the use of filters might cause the object to be too faint to observe in quiescence .
exposure times were 90 seconds during the bright state and 100 - 150 seconds at minimum light .
all curve team data reductions were performed using a standard procedure based on iraf package and the profile photometry has been derived using the daophotii package ( stetson 1987 ) . during preliminary analysis of the data we found the aavso archive containing several ccd observations of the same superoutburst of v419 lyr made by observers from england ( d.b . ) and the united states ( r.k . ) .
therefore we decided to combine our data to obtain better results .
d.b . used a 0.35-m meade schmidt - cassegrain telescope with a starlight xpress sxv - h9 ccd camera .
data were taken unfiltered .
exposures were 40 , 50 or 60 sec depending on conditions .
the aip4win package ( berry and burnell 2000 ) was used to dark - subtract and flat - field all images before measuring them using aperture photometry .
the magnitude of v419 lyr was determined by differential photometry with respect to an ensemble of two nearby comparison stars .
r.k . used 0.25-m meade lx-200 schmidt - cassegrain telescope equipped with an apogee ap-47 ccd camera and clear filter characterized by ir block at 700 nm .
mpo canopus software ( warner 2007 ) was used for differential photometry of the variable with an average of four comparison stars .
table 1 presents a journal of our ccd observations of v419 lyr . in total
, we observed the star for more than 48 hours on 14 nights and collected 3179 exposures .
figure 1 presents the photometric behaviour of v419 lyr in 2006 july and august .
dots and open circles correspond to our ccd observations and visual estimates of aavso observers , respectively .
the shape of the light curve corresponds to the standard picture of a superoutburst .
first , the star rose rapidly from its quiescent level to a peak magnitude of around 14.5 mag . due to the lack of observations , the exact time of the rise
is unknown .
v419 lyr was seen by aavso observers in a bright state on july 16 .
two nights later , during our first run , the star was still at the same brightness and showing fully developed superhumps with a peak - to - peak amplitude of 0.3 mag - a feature characteristic of the beginning of a superoutburst .
thus we conclude that the july 2006 superoutburst started around july 15 . after reaching peak magnitude ,
v419 lyr entered the plateau stage lasting about 11 days with an average decline rate of about 0.1 mag d@xmath7 .
the plateau stage ended rapidly on july 27 and during the next two days we observed the final decline stage with a change of brightness of about 1 mag d@xmath7 . on july 30 the star was again in quiescence showing a mean brightness of 18.1 mag .
thus the entire superoutburst lasted about 15 days .
the linear fit to the plateau stage in figure 1 is helpful in showing there was no rebrightening as has been observed in some su uma systems ( kato et al .
figure 2 shows the light curves of v419 lyr during thirteen individual nights .
filled circles correspond to the ostrowik observatory data , while open circles and squares represent d.b . and r.k .
measurements respectively .
the magnitudes have been transformed to a common @xmath8 system and detrended for purposes of fourier power spectrum analysis .
the superhumps are clearly visible and have an initial amplitude of 0.3 mag
. later the superhump amplitude gradually decreases and on august 2/3 is practically indistinguishable from noise .
as we noted earlier , all light curves of v419 lyr in superoutburst were detrended by removing a fit based on a first or second order polynomial .
then we analyzed them using anova statistics ( schwarzenberg - czerny 1996 ) .
the resulting periodogram is shown in figure 3 .
it shows a very clear and dominant peak at frequency @xmath9 c / d , which we interpret as due to superhumps and corresponds to the period @xmath10 days .
the spectrum has almost no 1-day aliases due to good coverage and the use of data from three sites - two from europe and one from the united states .
a small peak is also observed around 22 c / d , the first harmonic of the main frequency .
we then prewhitened the light curve of v419 lyr during the superoutburst with the main period and its first harmonic .
the power spectrum of the resulting light curve shows no other periodicities . to check the stability of the superhump period and to better determine its value
, we constructed an @xmath11 diagram . because the maxima in our case were almost always clearly visible and easier to measure than minima
, we decided to use the timings of the former .
finally , we were able to determine 27 times of maxima which are listed in table 2 together with associated errors , cycle numbers @xmath12 , and @xmath11 values .
.times of maxima observed in the light curve of v419 lyr during its 2006 superoutburst [ cols=">,^,^,>",options="header " , ] a least - squares linear fit to the data taken during the plateau phase of superoutburst gives the following ephemeris for the maxima : @xmath13 the above equation indicates that the mean superhump period was @xmath14 days , which agrees within errors with the determination based on anova statistics .
combining these two measurements gives us our final estimate of the mean superhump period of v419 lyr during its 2006 july superoutburst which is @xmath15 days ( @xmath2 min ) .
the @xmath11 values computed according to the ephemeris ( 1 ) are listed in table 2 and also shown in figure 4 .
it is clear that v419 lyr , during its 2006 superoutburst , showed clear changes of superhump period . in the cycle range @xmath16 the period was quickly decreasing .
a second - order polynomial fit to @xmath17 dependence in this range corresponds to the solid line in the bottom panel of figure 4 and is expressed by the following ephemeris : @xmath18 this equation indicates that the period derivative has the large value @xmath3 .
this is the second largest negative value detected in su uma stars .
figure 5 , taken from kato et al .
( 2003b ) and olech et al .
( 2003 ) , shows the position of v419 lyr on the @xmath19 diagram .
only kk tel had a faster period decrease during its 2002 june superoutburst ( kato et al .
it is interesting that both v419 lyr and kk tel are long superhump period dwarf novae with orbital periods very close or even within the period gap .
it is worth commenting on the behaviour of the superhump period after cycle number @xmath20 . at that moment ,
the star was still in the plateau phase of the superoutburst , three days before entering the final decline , and superhumps were still clearly visible in the light curve .
the @xmath11 values for cycle numbers from 70 to 111 can be roughly fitted with a straight line , which indicates that the period decrease had stopped and its value stabilized at @xmath21 days . the noisy maxima with cycle numbers above 120
are shifted by @xmath22 cycle with respect to the superhump maxima from earlier nights .
this indicates that they may be connected with late superhumps or even with the orbital wave .
it is interesting that our @xmath11 diagram could be also interpreted in diffrent way than `` common superhump followed by transition to late superhump '' scenario .
fitting a 5th order polynomial to the moments of maxima appears to work with the data about as well as a quadratic followed by a linear trend as described previously .
we are not arguing in favour of this interpretation , simply saying that there might be other interpretations of the data .
the upper panel of figure 4 shows the evolution of the amplitude of superhumps through the entire superoutburst . in a typical su uma star ,
fully developed superhumps have an amplitude of about 0.3 mag and a characteristic tooth shape .
interestingly , these properties seem to be completely independent of the inclination of the orbit of the binary .
as the outburst progresses , the amplitude gradually decreases and the profile of the humps changes . a few days after maximum , so called secondary humps or interpulses become visible . in the beginning they are small but with time they may become as high as the main maxima - most probably evolving towards late superhumps .
v419 lyr is an interesting case because it seems not to follow this scenario . on the nights of july 19/20 and 20/21 it clearly showed double maxima .
however , these quickly disappeared . during subsequent nights
there was hardly a trace of secondary humps .
very weak humps could be seen only on july 23/24 and 25/26 and quite strong ones on july 27/28 .
we were curious about the reason for such behaviour .
the only property which strongly differs in v419 lyr from typical su uma stars is its long superhump / orbital period placing it within the period gap .
we therefore reviewed the literature to investigate the occurrence of secondary humps among long period systems ( i.e. with superhump period @xmath23 days ) .
the summary of our review is given in table 3 .
l l c c c c c star & @xmath24 [ days ] & clear & poorly & invisible & ref + hs vir & 0.08077 & - & + & - & 1,2 + v359 cen & 0.08092 & - & + & - & 3,4 + v660 her & 0.081 & - & - & + & 5 + v503 cyg & 0.08101 & + & - & - & 6 + br lup & 0.08220 & + & - & - & 7,8 + v877 ara & 0.08411 & + & - & - & 9 + ab nor & 0.08438 & - & + & - & 10 + v369 peg & 0.08484 & - & - & + & 11 + hv aur & 0.08559 & + & - & - & 12 + ef peg & 0.08705 & - & + & - & 13,14 + ty psa & 0.08765 & - & + & - & 15,16 + bf ara & 0.08797 & - & - & + & 17 + kk tel & 0.08803 & - & - & + & 18 + dv uma & 0.08869 & - & + & + & 19,20 + v419 lyr & 0.0901 & - & + & - & this study,21 + uv gem & 0.0902 & - & + & - & 22 + v344 lyr & 0.09145 & - & - & + & 23 + yz cnc & 0.09204 & - & + & - & 24,25 + gx cas & 0.09297 & - & - & + & 26 + v725 aql & 0.09909 & - & + & - & 27 + mn dra & 0.1055 & - & - & + & 28,29 + + + + + + + + + among 21 reviewed stars only four show
clear secondary humps .
the rest of them show no interpulses at all or only a weak trace of them .
the orbital period of v419 lyr is unknown .
however it is possible to estimate its value using the relation in stolz and schoembs ( 1984 ) connecting the period excess @xmath25 defined as @xmath26 with the orbital period of the binary .
this empirical relation is as follows : using the definition of @xmath25 and knowing @xmath24 for v419 lyr , we were able to estimate the orbital period as @xmath28 days .
this is slightly longer than two hours which indicates that v419 lyr is a dwarf nova in the period gap
. many characteristics of v419 lyr are typical of su uma stars .
it goes into superoutburst every year or so , the eruption lasts about two weeks and has an amplitude of @xmath0 mag .
superhumps appear shortly after the beginning of the superoutburst and have a maximum amplitude of 0.3 mag , which decreases to 0.1 mag at the end of the outburst .
in addition to its long orbital period , v419 lyr is unusual in two other properties .
its superhump period derivative has one of the largest negative values known and it shows only a weak trace of secondary humps in the final stages of the superoutburst .
* acknowledgments . *
we acknowledge generous allocation of the warsaw observatory 0.6-m telescope time .
this work used the online service of the vsnet and aavso .
we would like to thank prof .
jzef smak for fruitful discussions .
antipin , s. v. , & pavlenko , e. p. 2002 , _ astron .
_ , 391 , 565 barwig , h. , kudritzki , r. p. , vogt , n. , & hunger , k. 1982 , _ astron .
_ , 114 , l11 berry r. , burnell j. 2000 , the handbook of astronomical image processing , willmann - bell feline , w. j. , dhillon , v. s. , marsh , t. r. , & brinkworth , c. s. 2004 , _ mnras _ , 355 , 1 harvey , d. , skillman , d. r. , patterson , j. , & ringwald , f. a. 1995 , _ pasp _ , 107 , 551 hellier , c. 2001 , cataclysmic variable stars , springer howell , s. b. , schmidt , r. , deyoung , j. a. , fried , r. , schmeer , p. , & gritz , l. 1993 , _ pasp _ , 105 , 579 kato , t. 1993 , _ pasj _ , 45 , l67 kato , t. , nogami , d. , masuda , s. , & baba , h. 1998 , _ pasp _ , 110 , 1400 kato , t. , sekine , y. , & hirata , r. 2001 , _ pasj _ , 53 , 1191 kato , t. , & uemura , m. 2001 , informational bulletin on variable stars , 5158 , 1 kato , t. 2001 , informational bulletin on variable stars , 5104 , 1 kato , t. , et al .
2002 , _ astron .
_ , 395 , 541 kato , t. 2002 , _ pasj _ , 54 , 87 kato , t. , nogami , d. , moilanen , m. , yamaoka , h. 2003a , _ pasj _ , 55 , 989 kato , t. , et al . 2003b , _ mnras _ , 339 , 861 kato , t. , bolt , g. , nelson , p. , monard , b. , stubbings , r. , pearce , a. , yamaoka , h. , richards , t. 2003c , _ mnras _ , 341 , 901 kato , t. , uemura , m. , ishioka , r. , nogami , d. , kunjaya , c , baba , h. , yamaoka , h. 2004a , _ pasj _ , 56 , 1 kato , t. , et al .
2004b , _ mnras _ , 347 , 861 katysheva , n.a . , pavlenko , e.p .
2003 , _ astophysics _ , 46 , 114 kuroshkin , n.e . 1990 , _
peremennye zvezdy _
, 17 , 186 mennickent , r. e. , & sterken , c. 1998 , _ pasp _ , 110 , 1032 nogami , d. , kato , t. , masuda , s. , & hirata , r. 1995 , informational bulletin on variable stars , 4163 , 1 nogami , d. , kato , t. , & masuda , s. 1998 , _ pasj _ , 50 , 411 nogami , d. , kato , t. , baba , h. , novk , r. , lockley , j. j. , somers , m. 2001 , _ mnras _ , 322 , 79 nogami , d. , et al .
2003 , _ astron .
_ , 404 , 1067 odonoghue , d. 1987 , _ astrophys . space science _ , 136 , 247 olech , a. , schwarzenberg - czerny , a. , kedzierski , p. , zloczewski , k. , mularczyk , k. , wisniewski , m. 2003 , _ acta astron .
_ , 53 , 175 olech , a. , zloczewski , k. , mularczyk , k. , kedzierski , p. , wisniewski , m. , stachowski , g. 2004 , _ acta asron .
_ , 54 , 57 olech , a. , zloczewski , k. , cook , l. m. , mularczyk , k. , kedzierski , p. , wisniewski , m. 2005 , _ acta astron . _ , 55 , 237 olech , a. , mularczyk , k. , kedzierski , p. , zloczewski , k. , wisniewski , m. , szaruga , k. 2006 , _ astron
. astrophys .
_ , 452 , 933 osaki , y. 1996 , _ pasp _ , 108 , 39 osaki , y. 2005 , _ proc .
of the japan academy _
, seires b , 81 , 291 patterson , j. 1979 , _ astron . j. _
, 84 , 804 schreiber , m.r , lasota , j .-
2007 , arxiv:0706.3888 schwarzenberg - czerny , a. 1996 , _ apj letters _ , 460 , l107 smak , j. 2000 , _ new astronomy _ , 44 , 171 stetson , p.b .
1987 , _ pasp _ , 99 , 191 stolz , b. , schoembs , r. 1984 , _ astron .
_ , 132 , 187 udalski a. , pych w. 1992 , _ acta astron . _ , 42 , 285 uemura , m. , kato , t. , pavlenko , e. , baklanov , a. , & pietz , j. 2001 , _ pasj _ , 53 , 539 warner , b. , odonoghue , d. , & wargau , w. 1989 , _ mnras _ , 238 , 73 warner b. 1995 , cataclysmic variable stars , cambridge university press .
warner b. 2007 , http://www.minorplanetobserver.com/mposoftware/mpocanopus.htm woudt , p. a. , & warner , b. 2001 , _ mnras _ , 328 , 159 | we report extensive photometry of the dwarf nova v419 lyr throughout its 2006 july superoutburst till quiescence . the superoutburst with amplitude of @xmath0 magnitude lasted at least 15 days and was characterized by the presence of clear superhumps with a mean period of @xmath1 days ( @xmath2 min ) . according to the stolz - schoembs relation
, this indicates that the orbital period of the binary should be around 0.086 days i.e. within the period gap . during the superoutburst the superhump period was decreasing with
the rate of @xmath3 , which is one of the highest values ever observed in su uma systems . at the end of the plateau phase ,
the superhump period stabilized at a value of 0.08983(8 ) days .
the superhump amplitude decreased from 0.3 mag at the beginning of the superoutburst to 0.1 mag at its end . in the case of v419 lyr
we have not observed clear secondary humps , which seems to be typical for long period systems .
* key words : * stars : individual : v419 lyr binaries : close novae , cataclysmic variables |
because of the limited spatial resolutions of the currently available instruments , the geometry and kinematics of broad - line region ( blr ) in active galactic nucleus ( agn ) is still an important and unresolved problem .
this poor understanding results in an uncertainty for the virial coefficient @xmath2 that depends not only on the kinematics , structure and orientation of the blr , but also on the virial width estimator used .
@xmath2 is equal to 3 for a spherical isotropic velocity distribution , while @xmath2 can be several times larger if the blr can be described as a circular rotation disk .
the poor determination of @xmath2 finally causes an uncertainty in the estimation of the mass of central supermassive black hole ( smbh ) of an agn in the virial assumption ( e.g. , onken et al .
2004 ; kaspi et al . 2005 ; marconi et al . 2008
; woo et al .
2010 ) .
recent spectral profile modelings indicate that the blr is much more complex than a single virial component .
in addition to the traditional broad component ( hereafter bc , @xmath3 ) , a very broad component ( hereafter vbc , @xmath4 ) is usually required to reproduce the observed line profiles of both h@xmath0 and h@xmath1 in some agns ( e.g. , sulentic et al . 2000 ; netzer & trakhtenbrot 2007 ; hu et al .
2008 ; marziani et al . 2009 ; zhu et al . 2009
; zamfir et al . 2010 ) .
the physical interpretation of the two components is , however , still under debate at present .
previous studies have frequently suggested that the vbc likely arises in optically thin clouds that are closer to the central engine than the bc since 1980 s ( e.g. , shields et al . 1995 ; sulentic et al .
2000 ; zamfir et al .
2010 ) . by examining a large sample of quasars selected from sdss , hu et al .
( 2008 ) proposed that the bc ( the intermediate - line region in hu et al .
2008 ) can be explained by an inflow because it is systematically redshifted with respect to the quasar rest - frame .
the comprehensive photoionization models calculated by korista & goad ( 2004 ) indicate that the vbc can arise from inner blr ( i.e. , not a distinct emission - line region ) where the balmer line emission rate ( and also responsivity ) is low .
sulentic et al .
( 2000 ) detected a violent variability in both continuum and h@xmath0 emission line in radio - quite luminous quasar pg1416 - 129 ( hereafter 2000 spectrum for short ) . by comparing it with the spectrum ( hereafter 1990 spectrum for short ) taken by boroson & green ( 1992 , bg92 )
, the 2000 spectrum shows a dramatic disappearance for the bc of the h@xmath0 emission line .
however , the vbc can be still clearly identified in the 2000 spectrum . basing upon these observational facts , sulentic et al .
( 2000 ) argued that a large fraction of the vbc is likely produced in optically thin clouds that are closer to the central source . in this letter , we re - observed the luminous quasar pg1416 - 129 ( r.a.=14@xmath5:19@xmath6:[email protected] , dec=13:10@xmath8:44@xmath9 , j2000 , @xmath10 ) at 10 years after the demise of its h@xmath0 bc to examine the variability of the h@xmath0 line profile .
in addition to h@xmath0 , the h@xmath1 emission line is also covered in our new spectra . the observations and data reductions
are described in 2 . 3
presents the results and discussion .
the @xmath11 cold dark matter ( @xmath11cdm ) cosmology with parameters @xmath12 , @xmath13 , and @xmath14 ( spergel et al .
2007 ) is adopted throughout the letter .
two new optical spectra were obtained by the national astronomical observatories , chinese academy of sciences ( naoc ) 2.16 m telescope in xinglong observatory at january 25 , 2010 , and march 29 , 2011 .
the spectroscopic observations were carried out with the omr spectrograph equipped with a back - illuminated spec @xmath15 ccd as the detector .
the used grating is 600@xmath16 , and the slit oriented in the south - north direction corresponds a width of @xmath17 .
this setup finally translates into a spectral resolution of @xmath18 as measured from the sky emission lines and comparison arcs .
this spectral resolution is comparable with the previous observations in bg92 .
the blazed wavelength was fixed at 6500 in both observations in order to allow the observed spectra cover the wavelength range from 4600 to 6900 in the observer frame ( i.e. , cover both h@xmath0 and h@xmath1 emission regions ) .
the observations were carried out as close to median as possible . in each of the observations ,
the object was observed successively twice with the exposure time of each frame of 3600s .
the two frames were combined prior to the extraction to enhance the signal - to - noise ratio and to eliminate the contamination of cosmic - rays easily .
the wavelength calibration was carried out by the helium - neon - argon comparison arc taken between the two successive frames , i.e. , at the position being nearly identical to that of the object . the observed flux
was calibrated by the kitt peak national observatory ( kpno ) standard stars hd117880 and feige56 in the 2010 and 2011 observations , respectively ( massey et al .
1988 ) .
the 2-dimensional spectra were reduced by the standard procedures through the iraf package , including bias subtraction , flat - field correction and cosmical - rays removal before the extraction of the 1-dimensional spectra . each extracted spectrum
was then calibrated in wavelength and flux by the corresponding comparison arc and standard . in the observer frame ,
the redshifted h@xmath1 emission profile is strongly distorted by the a - band telluric absorption feature at @xmath197600 - 7630 due to @xmath20 molecules . in order to properly model the h@xmath1 profile ,
the two telluric features around @xmath216800 and @xmath217600 were removed from each observed spectrum by the corresponding standard .
the galactic extinction was corrected for both spectra by the color excess @xmath22 taken from the nasa / iapc extragalactic database ( ned ) , assuming the @xmath23 extinction law of our galaxy ( cardelli et al .
the spectra were then transformed to the rest frame , along with the correction of the relativity effect on the flux , according to the narrow peaks of h@xmath0 .
both reduced spectra ( hereafter 2010 and 2011 spectra ) are displayed in figure 1 by the bottom two curves .
the top two curves show the 1990 and 2000 spectra for a comparison . for each of the four spectra , we modeled the continuum by a broken - powerlaw superposed by a broadened template using a @xmath24 minimization .
the template provided in bg92 is broadened to have a fwhm that is identical to that of total h@xmath0 line .
the minimization was carried out in the wavelength range free of strong emission lines ( e.g. , h@xmath1 , h@xmath0 , [ ] @xmath194959 , 5007 , @xmath214686 ) of agns .
the continuum fittings are illustrated in figure 1 by the superposed red lines for the four spectra .
the isolated emission - lines were then profiled by the specfit task ( kriss 1994 ) in the iraf package .
we modeled each emission line by a sum of a set of several gaussian components .
as an illustration , the modelings in the 2011 spectrum are shown in the bottom - left panel and bottom - right panel in figure 2 for the h@xmath0 and h@xmath1 regions , respectively .
the two middle panels show the same as the bottom ones but for the 2010 spectrum .
in addition to the new spectra , the 1990 and 2000 spectra were also modeled by the same method in order to examine the variability of h@xmath25 and h@xmath26 .
the deblendings of the h@xmath0 region are illustrated in the two top panels in figure 2 .
the measured emission - line properties and continuum fluxes at 5100 are tabulated in table 1 .
all the fluxes are reported after scaling the spectra to the total [ ] @xmath215007 flux of the 1990 spectrum , i.e. , @xmath27}\lambda5007)=(3.10\pm0.23)\times10^{-14}\ \mathrm{ergs\ cm^{-2}\ s^{-1}}$ ] .
all the uncertainties presented in the table only include the errors resulted from the spectral modelings .
the red - shelf of the h@xmath0 line profile of the object is most unlikely contaminated by the broad @xmath28 emission lines ( veron et al .
2002 ) . at first
, the broad @xmath29 line emission is not detectable in both new spectra .
secondly , a corresponding red - shelf can be obviously identified in the h@xmath1 emission line because of the clear inflection of the profile .
is strongly redshifted with respect to the narrow core ( e.g. , marziani et al .
2009 ) at a velocity shift @xmath31 in the 2010 spectrum , which is highly consistent with that of h@xmath32 , @xmath33 .
the consistence occurs not only in the velocity shift , but also in the line width of @xmath34 .
such consistence can be also identified in the 2011 spectrum , although the h@xmath30 bulk velocity is clearly less than that of h@xmath32 ( note the large uncertainty for the h@xmath0 line ) .
we calculate the virial mass of the central smbh ( @xmath35 ) from the four spectra taken in different epochs since we believe that the blackhole mass increases negligibly during the 20 years . in each epoch , @xmath35 is estimated from h@xmath1 and h@xmath0 emission lines by adopting the calibrations that are based on line width and line luminosity ( e.g. , peterson et al .
2004 ; kaspi et al . 2005 ; green & ho 2005 ; bentz et al .
2006 ; vestergaard & peterson 2006 ) . the luminosities are obtained after the correction of the local extinction .
the extinction is inferred from the measured narrow - line flux ratio h@xmath1/h@xmath0 by assuming the standard case b recombination and a galactic extinction curve with @xmath23 . the line width measured from the total broad - line profile
is usually adopted as a good virial broadening parameter because of the produced close virial relationship ( e.g. , peterson & wandel 2000 ; onken & peterson 2002 ; kollatschnny 2003 ; ) . by re - analyzing the reverberation - mapping data , peterson et al .
( 2004 ) claimed that the best line - width measurement is the line dispersion calculated from the variable part of the spectrum . in order to measure the fwhms of the entire broad - line profiles ,
the modeled narrow emission - lines are subtracted from each of the observed spectra before the measurements .
in addition to the entire line profiles , @xmath35 is also estimated by using the fwhms from the bcs .
note that the @xmath35 estimation is seriously complicated by the significant variation in the line width and velocity shift between 2000 and 2010/11 when the bcs are considered only .
the vbcs are usually excluded from the @xmath35 estimation because of the large relative velocity with respect to the systematic velocity .
@xmath35 is calculated from the line widths and luminosities .
the labels v and g are used to denote the calibrations in green & ho ( 2005 ) and vestergaard & peterson ( 2006 ) , respectively .
the calibrations finally yield consistent values of @xmath35 . for h@xmath0 line ,
the average estimates are ( @xmath36 , @xmath37)@xmath38 = ( @xmath39 , @xmath40)@xmath41 , and ( @xmath36 , @xmath37)@xmath42 = ( @xmath43 , @xmath44)@xmath41 , where the formal errors are estimated from the observations at the different epochs . for h@xmath1 line , we have average estimates of ( @xmath45)@xmath38= @xmath46 and ( @xmath45)@xmath42=@xmath47 from the two new observations .
note that these values are consistent with each other because an uncertainty of a factor of 4 - 5 is produced by the current available scaling relationships ( peterson 2008 ) .
one can see clearly that the object shows negligible line profile variation within the time scale of one year according to our new observations .
the light - travel time cross a spherically symmetric blr can be evaluated as @xmath48days , where @xmath49 is the h@xmath0 line luminosity in unit of @xmath50 , @xmath51 is the fill factor of the emission gas in unit of @xmath52 , and @xmath53 is the number density of the gas in unit of @xmath54 . the local - extinction corrected h@xmath0 luminosities are @xmath55 for both bc and vbc components .
the luminosities yield a light - travel time @xmath56 day for both emission - line regions , which agrees with the observed fact that the line profile is non - variable within one year .
the nature of the region emitting h@xmath32 is still under debate .
many authors believe that the h@xmath32 components detected in many agns come from a separate inner optically thin emitting region according to the analysis of line ratios and the line - continuum variation studies ( e.g. , ferland et al .
1990 ; marziani & sulentic 1993 ; corbin 1995 , 1997 ; corbin & smith 2000 ; sulentic et al .
2000 ; zhu et al .
the spectroscopic monitors previously claimed that the variations in line cores are stronger than in wings for both high - ionization and low - ionization emission lines ( e.g. , peterson et al . 1993 ; kassebaum et al .
1997 ; corbin & smith 2000 ) , although the opposite cases are found in ly@xmath1 and @xmath211549 lines for high - z quasars ( e.g. , obrien et al . 1989 ) . in an optically thin cloud ,
the emissivity of recombination lines is mainly determined by the volume and the density of the cloud rather than the incident ionizing continuum , which results in a weak response to the continuum level for the line intensity . by contrast , modern photoionization model of agn
s blr suggests that the h@xmath32 component can be attributed to the inner part of the traditional blr , because the calculated responsivity of balmer emission of optically thick clouds increases steeply with the distance from the central engine ( korista & goad 2004 ) .
snedden & gaskell ( 2007 ) compared their photoionization models with the broad emission - line profiles of both hst and optical spectra .
their analysis points out that the optically thick model provides better consistence with the observations for the very broad component , while the optically thin scenario poses many problems .
hu et al . (
2008 ) recently proposed another scenario that the classical blr produces the h@xmath32 component , meanwhile the h@xmath57 component is produced by an inflow of material .
comparing our new spectra with the previous ones allows us to identify a strong responsivity to the continuum for the h@xmath32 component . as the continuum level at 5100 decreases from @xmath58 to @xmath59 , the measured flux of h@xmath32
is reduced by a factor of 2 - 3 .
the corresponding equivalent widths ( ews ) maintain a constant value @xmath60 .
all the ews are measured against the continuum levels at 5100 . as an additional investigation
, the upper panel in figure 3 shows the smoothed line profiles for h@xmath1 ( the red lines ) and h@xmath0 ( the blue lines ) .
the profiles are produced by a subtraction of the modeled narrow - line profiles , and are smoothed by a box of 10 pixels .
we present the average h@xmath1/h@xmath0 ratios as a function of radial velocity in the lower panel .
the average values and the corresponding errors are calculated after binning the spectra over 1000@xmath61 .
the results from the 2010 and 2011 spectrum are plotted by the red and blue symbols , respectively . in both observations ,
the calculated balmer decrements have a plateau with @xmath62 from -5000@xmath61 to + 5000@xmath61 , and decrease rapidly in the wings .
the models in snedden & gaskell ( 2007 ) indicate that the decrements at the wings can not be reproduced by the optically thin photoionization model in which the predicted balmer decrement is rather uniformly around the value of case b recombination .
snedden & gaskell ( 2007 ) claimed that the decreased line ratios at wings in their sample are best explained by the optically thick model .
the object shows peculiar variation for the h@xmath57 component .
when compared with the 1990 spectrum , the component is decreased by a factor of 10 in flux in the 2000 spectrum .
however , the flux returns to the 1990 level in the 2010/2011 spectra in which the continuum level is comparable with that of the 2000 spectrum .
in addition to the flux change , the 2010/2011 spectra show a large variation in the bulk blueshift velocity and line width for the bc when compared with the previous observations .
the velocity and line width in the 2010 ( 2011 ) spectrum are larger than those of the 1990 ( 2000 ) spectrum by nearly an order of magnitude .
putting the variations in flux and in velocity together suggests a possibility that the h@xmath57 line might be emitted from an episodic outflow from the central engine .
the outflow is at first radially accelerated at the launch point , which results in the increasing bulk blueshift and line width .
the outflow then reaches a terminal velocity at larger distance where the radiative force and gravitation are negligible , and has a reduced optical depth .
the dynamical timescale of blr is @xmath63 , where @xmath64 is the distance of the blr from the central smbh and @xmath65 is the typical velocity .
estimating the distance as @xmath66 ld ( kaspi et al .
2000 ) yields a dynamical timescale @xmath67years , which does not pose fundamental challenge for the above scenario .
the second possibility is that the object is a `` double - peaked '' emission - line qso whose broad balmer line profiles are produced by an accretion disk ( e.g. , chen et al .
1989 ; gaskell 2010 ) .
in contrast with the object , the typical `` double - peaked '' emitters show line profiles with two peaks connected by a depressed plateau at the line center , and show rapid line profile variation with a timescale typical of a few months .
finally , the current data can not exclude a possibility that the two balmer components come from two separated smbhs orbiting with their blrs within a common nlr ( e.g. , sdssj153636.22 + 044127.0 , boroson & lauer 2009 ; lauer & boroson 2009 ) .
unusual broad line profile ( e.g. , two displaced peaks or large shifted broad line ) is usually used as an indicator for binary smbhs candidates ( see recent review in popovic 2011 and the references therein ) .
the binary blr model indicates a significant variation of line profile for different blr orbiting phases . in pg1416 - 129 , assuming a @xmath68 for each of the binary smbhs , the eq .
( 2 ) in popovic ( 2011 ) results in a distance between the two black holes of @xmath69 pc when the orbital period of @xmath70years is used .
alternatively , the significant variability could be produced from the merging of the blrs when the two black holes are in the `` sim - detached '' configuration .
long term spectroscopic monitor , especially in x - ray , is required to test which hypothesis is correct for the object .
the spectra of luminous radio - quite quasar pg1416 - 129 were re - observed in 2010 and 2011 to study the variation of its broad - line emission .
the two spectra show negligible line variation within a timescale of one year .
by re - analyzing the previous spectra , we identify a strong response to the continuum level for the very broad component .
the wide wavelength coverage of the new spectra indicate a flat balmer decrements at the line wings . with these results
, we argue that a separate inner optically thin emission line region might not be necessary in the object .
we would like to thank the anonymous referee for his / her very useful comments and suggestions .
the authors thank profs .
todd a. boroson and richard f. green for providing the optical template and spectrum taken in 1990 .
we are grateful to prof .
w. sulentic for providing the digital spectrum observed in 2000 .
special thanks go to the staff at xinglong observatory as a part of national astronomical observatories , china academy of sciences for their instrumental and observational help .
the study is supported by the open project program of the key laboratory of optical astronomy , naoc , cas .
this search is supported by the nfs of china ( grant 10803008 ) , and by the national basic research program of china ( grant 2009cb824800 ) .
bentz , m. c. , peterson , b. m. , pogge , r. w. , et al .
2006 , , 644 , 133 boroson , t. a. , & green , r. f. 1992 , , 80 , 109 ( bg92 ) boroson , t. a. , & lauer , t. r. 2009 , nature , 458 , 53 cardelli , j. a. , clayton , g. c. , & mathis , j. s. 1989 , , 345 , 245 chen , k. halpern , j. p. & filippenko , a. v. 1989 , , 339 , 742 corbin , m. r. 1995 , , 447 , 496 corbin , m. r. 1997 , , 485 , 517 corbin , m. r. , & smith , p. s. 2000 , , 532 , 136 gaskell , c. m. 2010 , nature , 463 , 1 greene , j. e. , & ho , l. c. 2005 , , 630 , 122 ferland , g. j. , korista , k. t. , & peterson , b. m. 1990 , , 363 , 21 hu , c. , wang , j. m. , ho , l. c. , et al .
2008 , , 683 , 115 kaspi , s. , maoz , d. , netzer , h. , et al .
2005 , , 629 , 61 kassebaum , t. m. , peterson , b. m. , wanders , i. , et al .
1997 , , 475 , 106 kollatschny , w. 2003 , , 407 , 461 korista , k. t. , & goad , m. r. 2004 , , 606 , 749 kriss , g. 1994 , adass , 3 , 437 lauer , t. r. , & boroson , t. a. 2009 , , 703 , 930 marconi , a. , axon , d. j. , maiolino , r. , et al .
2008 , , 678 , 693 marziani , p. , & sulentic , j. w. 1993 , , 409 , 612 marziani , p. , sulentic , j. w. , stirpe , g. m. , zamfir , s. , & calvani , m. 2009 , , 495 , 83 massey , p. , strobel , k. , barnes , j. v. , et al .
1988 , , 328 , 315 netzer , h. , & trakhtenbrot , b. 2007 , , 654 , 754 obrien , p. t. , zheng , w. , & wilson , r. 1989 , , 240 , 741 onken , c. a. , ferrarese , l. , merritt , d. , et al .
2004 , , 615 , 645 onken , c. a. , & peterson , b. m. 2002 , , 572 , 746 peterson , b. m. 2008 , newar , 52 , 240 peterson , b. m. , ali , b. , horne , k. , et al .
1993 , , 402 , 469 peterson , b. m. , ferrarese , l. , gilbert , k. m. , et al .
2004 , , 613 , 682 peterson , b. m. , & wandel , a. 2000 , , 540 , 13 popovic , l. c. 2011 , astro - ph / arxiv1109.0710 snedden , s. a. , & gaskell , c. m. 2007 , , 669 , 126 shields , j. c. , ferland , g. j. , & peterson , b. m. 1995 , 441 , 507 spergel , d. n. 2007 , , 170 , 337 sulentic , j. w. , marziani , p. , zwitter , t. , et al .
2000 , , 545 , 15 veron , p. , goncalves , a. c. , & veron - cetty , m .-
2002 , , 384 , 862 vestergaard , m. , & perterson , b. m. 2006 , , 641 , 689 woo , j. , treu , t. , barth , a. j. , et al .
2010 , , 716 , 269 zamfir , s. , sulentic , j. w. , marziani , p. , et al .
2010 , , 403 , 1759 zhu , l. , zhang , s. n. , tang , s. 2009 , , 700 , 1173 lccc & + + & flux & fwhm & @xmath71 + line identification & @xmath72 & @xmath61 & @xmath61 + h@xmath73 & @xmath74 & @xmath75 & @xmath76 + h@xmath32 & @xmath77 & @xmath78 & @xmath79 + h@xmath80 & @xmath81 & & + & + + & flux & fwhm & @xmath71 + line identification & @xmath72 & @xmath61 & @xmath61 + h@xmath73 & @xmath82 & @xmath83 & @xmath84 + h@xmath32 & @xmath85 & @xmath86 & @xmath87 + h@xmath80 & @xmath88 & & + & + + & flux & fwhm & @xmath71 + line identification & @xmath72 & @xmath61 & @xmath61 + h@xmath73 & @xmath89 & @xmath90 & @xmath91 + h@xmath32 & @xmath92 & @xmath93 & @xmath94 + h@xmath80 & @xmath95 & & + h@xmath96 & @xmath97 & @xmath98 & @xmath99 + h@xmath100 & @xmath101 & @xmath102 & @xmath103 + h@xmath104 & @xmath105 & & + & + + & flux & fwhm & @xmath71 + line identification & @xmath72 & @xmath61 & @xmath61 + h@xmath73 & @xmath106 & @xmath107 & @xmath108 + h@xmath32 & @xmath109 & @xmath110 & @xmath111 + h@xmath80 & @xmath112 & & + h@xmath96 & @xmath113 & @xmath114 & @xmath115 + h@xmath100 & @xmath116 & @xmath117 & @xmath118 + h@xmath104 & @xmath119 & & + | we report new spectroscopic observations performed in 2010 and 2011 for luminous radio - quite quasar pg1416 - 129 .
our new spectra with high quality cover both h@xmath0 and h@xmath1 regions , and show negligible line profile variation within a timescale of one year .
the two spectra allow us to study the variability of the balmer line profile by comparing the spectra with the previous ones taken at 10 and 20 years ago . by decomposing the broad balmer emission lines into two gaussian profiles , our spectral analysis suggests a strong response to the continuum level for the very broad component , and significant variations in both bulk blueshift velocity / fwhm and flux for the broad component .
the new observations additionally indicate flat balmer decrements ( i.e. , too strong h@xmath0 emission ) at the line wings , which is hard to be reproduced by recent optically thin models . with these observations
we argue that a separate inner optically thin emission - line region might not be necessary in the object to reproduce the observed line profiles . |
being located two orders of magnitude closer than the nearest large galaxy , the galactic centre ( gc ) can be studied with much higher spatial resolution and sensitivity than is possible for other galaxies .
because of this advantage , we can identify unique objects like the radio - arc consisting of linear parallel filaments ( yusef - zadeh et al .
1984 ) , or the 2.6@xmath6 m@xmath7 black hole most probably associated with the compact radio source sgr a * ( ghez et al . 1998 ) .
these studies have also established a concentration of gas , mostly in the form of molecular clouds characterised by densities @xmath810@xmath9 molecules @xmath10 , velocity dispersions @xmath810 km s@xmath11 , temperatures @xmath870 k and apparent magnetic fields @xmath81 mg in the region .
most of these quantities are one to two orders of magnitude larger than those found in the disk of our galaxy .
the review by morris & serabyn ( 1996 ) , describes this research .
melia & falcke ( 2001 ) provide the recent advances made in the research on the black hole and the central region ( a few pc ) of the galaxy .
the three - dimensional source distribution in the gc region is complex , and despite the studies mentioned above , identification of the objects and the physical processes operating in the region are difficult to understand , some of which we discuss below . from the high - resolution ( @xmath8 arcsec ) observations with the vla at radio frequencies ( pedlar et al .
1989 ) , the following sources within the central 15@xmath0 of the galaxy have been identified or reconfirmed .
( i ) at the dynamical centre of the galaxy is the compact nonthermal radio source known as sgr a*. ( ii ) around sgr a * is the hii region sgr a west , whose morphology resembles a face - on spiral galaxy .
( iii ) near sgr a west is sgr a east , which is believed to be a supernova remnant ( snr ) .
( iv ) a 7@xmath0 halo , which has been proposed to be a mixture of thermal and non - thermal emission ( pedlar et al .
1989 ) .
sgr a * ( see melia & falcke 2001 for a review ) was not detected at frequencies below 960 mhz and observations at 408 mhz ( davies et al .
1976 ) and at 330 mhz ( pedlar et al . 1989 ) provided upper limits ( @xmath120.1 jy at 330 mhz ) on its flux density .
sgr a * probably has a low frequency turnover below 1 ghz , but the nature of the turnover has never been examined in detail ( melia & falcke 2001 ) .
recently , nord et al . (
2004 ) claim to have detected sgr a * at 330 mhz .
however , their detection at 330 mhz remains provisional ( roy & rao 2004 ) .
magnetic fields can be strong enough to have a significant influence on the dynamics and evolution of the sources in the central region of our galaxy .
magnetic pressure can contribute significantly to the overall pressure balance of the interstellar medium ( ism ) and can even influence the scale height of the gas .
it can affect the formation and motion of clouds and perhaps mediate in the star formation process ( beck et al . 1996 ) .
therefore , it is important to measure the magnetic field geometry and strength in the central part of the galaxy .
however , other than the central 200 pc region , no systematic study has been made to estimate the magnetic fields in the inner 5 kpc region of the galaxy ( davidson 1996 ) .
the estimates within the central 200 pc region are mainly based on the observations of the non - thermal filaments or ntfs ( gray et al .
1995 ) . however , if the ntfs are manifestations of a peculiar local environment ( shore & larosa 1999 ) , inferences drawn from these observations can be misleading . here
we describe ( i ) the 620 mhz gmrt observations of the gc , which was aimed at studying the low frequency spectrum and the location of sgr a * with respect to sgr a west hii region , ( ii ) atca and vla observations of 65 background sources located in the central @xmath16@xmath2@xmath36@xmath2 , @xmath12@xmath2@xmath42@xmath2 region of the galaxy to measure faraday rotation measure ( rm ) , which could yield a model dependent magnetic field in the gc region .
the 620 mhz observations of the gc were carried out with the gmrt with a nominal bandwidth of 16 mhz .
full - synthesis ( @xmath87 hours ) observations were performed .
absolute flux density calibration was performed using 3c48 and 3c286 data .
the increase of @xmath13 from the calibrator field to the target source affects the source visibility amplitudes in the default observing mode ( i.e. the automatic level control ( alc ) in the system is turned on ) , and this was tackled by ( i ) multiplication of the target source data with ratio of total power ( when alcs were off ) between source and calibrator , or ( ii ) observations with alc off and using unnormalised cross - correlation data ( see appendix a ) .
the data were processed within astronomical image processing system ( aips ) using standard programs .
images of the fields were formed by fourier inversion and cleaning ( imagr ) .
the initial images were improved by phase only self - calibration . to improve the deconvolution of the extended emission
, we made the final image using multi - resolution clean ( wakker & schwarz 1988 ) .
the gmrt map at 620 mhz ( fig
. 1 ) has a dynamic range of about 150 , and is limited by systematics .
a total of 37 sources were observed with the atca using their 4.8 ghz and 8.5 ghz bands .
the atca observations were made using a 6 km array configuration .
each target source was typically observed for a total of 4050 minutes . since we used an e - w array configuration , to get a satisfactory _ uv_-coverage each source
was observed @xmath1410 times almost equally spaced in hour angle .
calibration and editing of the atca data were performed using miriad .
the calibrated visibilities were imaged using standard aips .
we used the vla in its bna configuration to observe 27 relatively weak sources in the sample .
the default continuum mode , with a single frequency channel of bandwidth 50 mhz in each if band was used .
observations were centred at frequencies of 4.63 , 4.88 , 8.33 and 8.68 ghz .
each source was observed for typically 5 minutes at 2 different hour angles .
all the data were calibrated and processed using the aips package .
the central 15@xmath0 region of the galaxy at 620 mhz is shown in fig . 1 .
the compact source sgr a * , along with other well known sources like sgr a east and the 7@xmath0 halo are clearly seen . to compare the smaller - scale features near sgr a west with higher frequency maps
, we have overlayed in fig . 2 , the 620 mhz map in gray scale on the 4.8 ghz vla map in contours convolved to a common resolution .
there is almost one to one correspondence between the brighter emission features at 4.8 ghz comprising the sgr a west region and a drop in the total intensity at 620 mhz ( indicated by a white region seen along sgr a west ) , indicating that the thermal emission is optically thick near 620 mhz .
while estimating the flux density of sgr a * at 620 mhz , we reduced the confusion due to the extended emission by imaging only those visibilities having _ uv _ distance @xmath157 k@xmath16 .
the flux density estimated from the image plane is about 0.5 @xmath170.1 jy .
we note that even after applying a lower _ uv _ cutoff , there is significant background confusion within a beam of size 7.@xmath18 .
this confusion causes an uncertainty of about 0.1 jy in the estimated flux density .
although the high radio frequency spectrum of sgr a * is well established , the spectrum below 1.4 ghz is not well estimated . at 1.4 ghz ,
the flux density of sgr a * is about 0.5 jy and its time averaged spectral index , @xmath19 defined as s@xmath20 , between 1.4 and 8.5 ghz is + 0.17 ( melia & falcke 2001 ) .
davies et al .
( 1976 ) found the flux density of sgr a * at 960 mhz to be a factor of two less than at 1.6 ghz and suggested that it has a low frequency turnover around 1 ghz .
this appears to be confirmed from their upper limit of 50 mjy at 408 mhz and the 100 mjy upper limit set by pedlar et al .
( 1989 ) at 330 mhz . however , the present 620 mhz observations along with its tentative 330 mhz detection ( nord et al .
2004 ) raise questions about earlier measurements . based on the average flux density of sgr a * at 1.4 ghz ( zhao et al .
2001 ) and its known spectrum between 1.4 and 8.5 ghz , we expect a flux density of 0.44 jy at 620 mhz , which is consistent with our measurements within the error - bars .
therefore , at frequencies of a few ghz to 620 mhz , the observed spectral index of sgr a * is nearly flat . though the advection - dominated accretion flow ( adaf ) model of emission from sgr a * fails to explain the observed low frequency emission , adaf along with self - absorbed synchrotron emission either from a relativistic jet , or from a small relativistic fraction of electrons embedded in the accretion flow ( yuan et al .
2003 ) could explain the observed spectrum of sgr a * from x - ray ranges to low radio frequencies of about 620 mhz . at 620 mhz
, sgr a west shows evidence for free - free self - absorption . in roy & rao ( 2004 )
, we have shown the optical depth of sgr a west hii region near sgr a * to be [email protected] .
if sgr a * was located behind sgr a west , then its flux density would have been attenuated by a factor of 10 due to thermal absorption by sgr a west .
however , the spectral index of sgr a * between 620 mhz and 1.4 ghz is roughly the same as between 1.4 ghz and 8.5 ghz showing no effect of free - free absorption by sgr a west .
this indicates that sgr a * is located in front of sgr a west .
it is possible to invoke alternate scenarios like holes in sgr a west along the line of sight to sgr a * , but these appear to be unlikely ( see roy & rao 2004 for details ) .
based on the atca data , 24 sources were found to have at least one polarised component .
the fwhm sizes of the beams in the atca 4.8 ghz images are @xmath146@xmath21 , and the typical rms noise is 0.23 mjy / beam in stokes i and about 0.15 mjy / beam in stokes q and u. of the sources observed using the vla , 21 have at least one polarised component .
the vla images have a beam of fwhm size @xmath82@xmath22 and the rms noise is typically 75 @xmath23jy / beam . a map of one of the sources is presented in fig . 3 . the estimated rm towards 45 sources are plotted in fig .
5 . in the plot , the positive rms have been indicated by cross ( @xmath24 ) and the negative rms by a circle ( @xmath25 ) , where the size of the symbols increases linearly with the @xmath26rm@xmath26 .
the region appears to be dominated by positive rm .
the median value of the rms is estimated to be 467@xmath1787 rad m@xmath27 . to detect small - scale fluctuations in the rm ,
we have plotted the rm structure function in fig .
as can be seen from the plots , the rm structure function tends to zero as the separation of the polarised components tends to zero . since intrinsic rms of different sources
are independent of their angular separations , our result supports an interstellar origin ( plasma turbulence ) of the rm towards the gc .
the structure function tends to saturate for components separated by more than 0.@xmath23 . from the plot ,
the coherence scale ( where the structure function attains half of its maximum value ) of the faraday screen is estimated to be about 0.@xmath22 ( 30 pc at a distance of 8.5 kpc ) .
our observations show a large - scale positive bias in rm towards the gc , with a mean rm of 525@xmath17129 rad m@xmath27 towards the sources with positive galactic longitude , and 389@xmath17103 rad m@xmath27 towards the sources with negative galactic longitude .
the absence of any reversal of sign of the rm ( 4@xmath28 result ) across the rotation axis of the galaxy ( @xmath29=0@xmath2 ) , indicates that the magnetic field across the rotation axis of our galaxy is not circularly symmetric .
two models , the axisymmetric and the bisymmetric spiral model of the magnetic fields are generally invoked to explain the field geometry in galaxies ( beck et al . 1996 ) .
it is thought that current loop generated by rotation of the matter around the centre ( dynamo model ) gives rise to the axisymmetric spiral configuration of the magnetic field
. however , if the magnetic fields at the present time are the result of concentration of a primordial magnetic field due to collapse of matter while the galaxy was formed , then it gives rise to the bisymmetric spiral configuration of the magnetic field in galaxies . in our galaxy ,
the magnetic field configuration has been suspected to be bisymmetric ( simard - normandin & kronberg 1980 ) , in which case , the field does not undergo a reversal of sign across the centre .
our observations support the bisymmetric spiral configuration in the inner parts of our galaxy . in order to estimate the line of sight ( _
los _ ) component of the magnetic fields from the rm , we need a model of the electron density in the region .
based on the very long baseline array ( vlba ) observations of the scattering diameter of extragalactic sources ( bower et al .
2001 ) , we consider an electron density of 0.4 @xmath30 in the region , which is twice the value predicted by the taylor & cordes ( 1993 ) model for the inner galaxy .
if the central 2 kpc is considered responsible for the faraday rotation , then the _ los _ averaged magnetic field is estimated to be 0.7 @xmath23 g ( roy , rao & subrahmanyan 2004 ) . to compute the random component of the magnetic field
, we note that the typical scale size of the faraday screen as estimated in @xmath313.3.1 is @xmath830 pc .
hence , there will be about 70 cells of irregularities within the central 2 kpc of the galaxy .
assuming that the rm contributions from the individual cells along the _ los _ are random about the mean , the cell contributions would have an rms rm value of 83 rad m@xmath32 .
if the rms electron density within such a cell is 0.4 @xmath30 , the estimated magnetic field is about 9 @xmath23 g , which is more than an order of magnitude higher than its _ los _ average value .
this small field raises doubts about an all - pervasive mg field in the central few hundred pc of the galaxy ( yusef - zadeh & morris 1987 ) .
the large - scale picture suggested by our observations is that the magnetic field even a few degrees away from the gc is oriented nearly perpendicular to the _ los _ pointing towards us on both sides of the gc .
gmrt is run by the national centre for radio astrophysics of the tata institute of fundamental research .
the australia telescope is funded by the commonwealth of australia for operation as a national facility managed by csiro .
the national radio astronomy observatory is a facility of the national science foundation operated under cooperative agreement by associated universities , inc .
beck , r. , brandenburg , a. , moss , d. , shukurov , a. , sokoloff , d. 1996 , ara&a , 34 , 155 + bower , g. c. , backer , d. c. , sramek , r. a. 2001 , apj , 558 , 127 + davidson , j. a. 1996 , in asp conf .
97 : polarimetry of the interstellar medium , 504 + davies , r. d. , walsh , d. , booth , r. s. 1976 , mnras , 177 , 319 + ghez , a. m. , klein , b. l. , morris , m. , becklin , e. e. 1998 , apj , 509 , 678 + gray , a. d. , nicholls , j. , ekers , r. d. , cram , l. e. 1995 , apj , 448 , 164 + + melia , f. , falcke , h. 2001 , ara&a , 39 , 309 + morris , m. , serabyn , e. 1996 , ara&a , 34 , 645 + nord , m. e. , lazio , t. j. w. , kassim , n. e. , goss , w. m. , duric , n. 2004 , apjl , 601 , l51 + pedlar , a. , anantharamaiah , k. r. , ekers , r. d. , et al .
1989 , apj , 342 , 769 + roy , s. , rao , a. p. 2004
, mnras , 349 , l25 + roy s. , rao a.p . , subrahmanyan r. 2004 , submitted to mnras + shore , s. n. , larosa , t. n. 1999 , apj , 521 , 587 + simard - normandin , m. , kronberg , p. p. 1980 , apj , 242 , 74 + taylor , j. h. , cordes , j. m. 1993 , apj , 411 , 674 + wakker , b. p. , schwarz , u. j. 1988 , a&a , 200 , 312 + yuan , f. , quataert , e. , narayan , r. 2003 , apj , 598 , 301 + yusef - zadeh , f. 1989 , in iau symp . 136 : the center of the galaxy , eds m. morris , dordrecht : kluwere , 243 + yusef - zadeh , f. , morris , m. , 1987 , aj , 94 , 1178 + yusef - zadeh , f. , morris , m. , chance , d. 1984 , nature , 310 , 557 + zhao , j. , bower , g. c. , goss , w. m. 2001 , apjl , 547 , l29
the automatic level control ( alc ) is a negative feedback device that keeps the output power of the antennas at the optimal levels .
when @xmath13 changes , the alc changes the effective amplitude gain of the amplifiers in order to counteract the changes in output power , leading to changes in the visibility amplitudes .
therefore , switching off the alcs avoids the problem .
however , with alcs off , there can be sufficient output power to saturate the electronics system .
therefore , system gains need to be adjusted in a way to avoid this problem .
this is usually achieved by changing the amplitude gains of all the antennas using attenuators , so that the output power of all the antennas are equal to their default values . since
, @xmath13 is maximum when the sky temperature is maximum , this needs to be performed while observing the strongest source .
keeping visibility amplitudes free from above problem also requires data , where the cross - correlation coefficients between different antenna pairs are not normalised by the total power of the antennas . as shown in the appendix - b , it also avoids any possible change of bandshape due to change of sky temperature .
[ [ appendix - b - variation - of - spectral - bandshape - in - normalised - cross - correlated - data - due - to - change - in - sky - temperature ] ] appendix b : variation of spectral bandshape in normalised cross correlated data due to change in sky temperature ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ the shape of the instrumental bandpass ( bandpass function ) is typically determined by observing a suitably strong calibrator during observations .
it is known that @xmath13 is the sum of the electronics temperature ( @xmath33 ) and the antenna temperature ( @xmath34 ) , and @xmath34 is directly proportional to the antenna efficiency ( @xmath35 ) .
therefore , if @xmath35 is a function of frequency ( @xmath36 ) , then both @xmath34 and @xmath13 are functions of @xmath36 .
as normalised visibilities are directly proportional to @xmath35 and inversely proportional to @xmath13 , it implies that the bandpass function is proportional to @xmath37 .
since the bandpass calibration is performed with a calibrator typically away from the galactic plane , the @xmath38 may be small and neglected in the denominator .
however , @xmath38 is substantial while observing any source in the galactic plane at low radio frequencies ( @xmath391.4 ghz ) .
as the relationship of the bandpass function with @xmath38 is non - linear , it changes the overall shape of the bandpass function .
for example , consider the antenna efficiency ( @xmath35 ) to fall from 0.6 at the centre of the observing band to 0.4 at the edge of the observing band .
assuming @xmath33=70 k , and the sky temperature ( @xmath40 ) ( near galactic plane ) to be 200 k , the normalised bandpass function at the centre of the band is equal to @xmath41 , and is @xmath42 at the edge of the band .
however , while observing a bandpass calibrator away from galactic plane , the @xmath34 is negligible , ( @xmath33 is assumed to be independent of @xmath36 ) and the bandpass function is proportional to @xmath35 . however , if this bandshape is used to correct the visibilities of the target source , the true visibilities at the edge of the observing band will be overestimated by about 25% compared to the centre of the band . | we have observed the central 45@xmath0 region of the galaxy at 620 mhz band of the giant metrewave radio telescope ( gmrt ) in radio continuum , and measured the polarisation properties of 64 small diameter background extragalactic sources seen through the @xmath16@xmath2@xmath36@xmath2 , @xmath12@xmath2@xmath42@xmath2 region with the australia telescope compact array ( atca ) and the very large array ( vla ) .
our 620 mhz observations show that sgr a * is located behind the hii region sgr a west . using the atca and the vla observations
, we measured the faraday rotation measure ( rm ) of the polarised sources .
the measured rms are mostly positive , and show no reversal of sign across the rotation axis of the galaxy .
this rules out any circularly symmetric model of magnetic field in the region .
we estimate the magnetic field strength in the region to be @xmath5 g , which raises doubts against an all pervasive mg field in the central few hundred pc of the galaxy . galaxy : centre galaxy : nucleus ism : magnetic fields hii regions |
the weak decays of heavy baryons provide testing ground for the standard model .
they reveal some important features of the physics of heavy quarks . from the study of the heavy quark physics , some important parameters of the standard model , for instance
, the cabbibo - kobayashi - maskawa ( ckm ) matrix element @xmath4 can be extracted by comparing experiments with theoretical calculations from the decay mode @xmath5 .
the main difficulties in the standard model calculations , however , are due to the poor understanding of the nonperturbative aspects of the strong interactions ( qcd ) . for the heavy hadrons containing a single heavy quark , an effective theory of qcd based on the heavy quark symmetry in the heavy quark limit @xcite , the so - called heavy quark effective theory ( hqet ) , has been proposed @xcite .
the classification of the weak decay form factors of heavy baryons has been simplified greatly in hqet @xcite . at the leading order of heavy quark expansion , only one universal form factor , the isgur - wise function ,
is required to describe the @xmath0 semileptonic decay . to the order of @xmath6 @xcite , one more universal function and one mass parameter
are introduced @xcite .
however , the heavy quark symmetry itself has no power to give information about the details of the universal form factors and the mass parameter . for a complete analysis to the heavy baryons , we still need to employ some other nonperturbative methods .
interesting results about the heavy baryon weak decay form factors have been obtained by various nonperturbative methods .
they are qcd sum rules @xcite , large @xmath3 limit @xcite , lattice simulation @xcite , dispersion relation and analyticity @xcite , and quark models @xcite . in this paper
, we apply the results of qcd sum rules and large @xmath3 limit to analyze in detail the weak decays of @xmath7 to the order of @xmath1 and @xmath2 .
the analysis is useful to experiments in the near future . in sec.2 ,
@xmath0 semileptonic decay form factors are discussed .
while there is no large @xmath3 calculation for the universal form factor appeared in @xmath6 corrections , we argue that it is zero in the large @xmath3 limit . in sec.3 , the numerical results for the decay rates , distributions and various angular asymmetry parameters are calculated . in sec.4 ,
several non - leptonic decay modes of @xmath8 are discussed .
we summarize the results in sec.5 .
by qcd sum rules , @xmath19 , @xmath20 and @xmath21 have been obtained @xcite .
qcd sum rule is regarded as a nonperturbative method rooted in qcd itself @xcite . in a linear approximation ,
the leading isgur - wise function is fit as @xmath23 on the other hand , the subleading isgur - wise function is negligibly small , @xmath24 and the parameter @xmath21 is determined to be @xmath25 it is interesting to compare above qcd sum rule results with that of large @xmath3 .
large @xmath3 limit is one of the most important and model - independent method of nonperturbative qcd in spite of the realistic @xmath26 @xcite .
hqet in this limit , for the heavy baryon case , is often believed to be the heavy quark skyrme model @xcite .
the leading isgur - wise function is predicted as @xcite @xmath27~. \label{eq : nc}\ ] ] the slope of this isgur - wise function is steeper than that of the sum rule . in the large @xmath3 limit , the parameter @xmath21 equals to the proton mass @xcite . in the following analysis , we take it as @xmath28 gev @xcite .
this result is in agreement with that obtained by qcd sum rules .
however , there is no skyrme model calculation for the subleading isgur - wise function .
we will assume that the subleading isgur - wise function is negligible in the skyrme model analysis .
in the following , we argue that this assumption is true in the large @xmath3 limit .
the subleading isgur - wise function @xmath29 is defined by @xmath30\end{aligned}\ ] ] where @xmath31 denotes the heavy quark field defined in the hqet with velocity @xmath13 , and @xmath32 is some gamma matrix .
@xmath29 just measures the amplitude of the brown muck transfer through a strong interaction , described by above matrix element , from the heavy quark which has a velocity change from @xmath13 to @xmath14 due to the weak decay . in the hartree - fock picture of large @xmath3 hqet
, heavy baryon has @xmath33 light quarks .
any @xmath34 transition in fact is suppressed when @xmath3 is large , because that involves changing the momenta of all the light quarks inside the baryon . in the limit @xmath35 ,
we expect @xmath36 . furthermore , it is well - known that @xmath37 due to luke theorem @xcite
. therefore , we get that @xmath29 vanishes in the large @xmath3 limit .
although the above argument makes our assumption for @xmath29 reasonable , it should be noted that there is still a subtle point which distinguishes the skyrme model from the large @xmath3 limit .
the point is that for heavy baryon weak decay form factors , the skyrme model result is not exactly identical to that of large @xmath3 . consider the leading isgur - wise function , our large @xmath3 argument for @xmath29 also applies to @xmath38 , that is @xmath39 . because @xmath40 , we expect that the leading isgur - wise function is @xmath41-function like , @xmath42 in the large @xmath3 limit .
however , this result in principle agrees with that of skyrme model @xcite @xmath43 $ ] , if @xmath3 is taken to be @xmath44 . the so called skyrme model result eq.([eq : nc ] )
can be obtained by taking @xmath26 .
therefore , @xmath45 can be understood as the result of large @xmath3 limit for the skyrme model .
with the knowledge of the form factors from qcd sum rule and large @xmath3 limit , we can calculate the rates , distributions and various asymmetry parameters for the @xmath0 semileptonic decay .
the standard expressions for these ovservables are given in ref .
@xcite in terms of helicity amplitudes .
the process @xmath46 is considered as a two - successive decay @xmath47 , @xmath48 .
let @xmath49 be the polarization vector of @xmath50 , where @xmath51 denotes the helicity state .
longitudinal state corresponds to @xmath52 , whereas transverse state , @xmath53 .
the helicity amplitudes are defined by @xmath54 where @xmath55 stands for the vector(axial vector ) current , and @xmath56 in the subscript is the helicity of the daughter baryon @xmath15 .
they can be expressed by the form factors , @xmath57 where the upper(lower ) sign is for the vector(axial vector ) current . with the notation of the total helicity amplitude @xmath58 , and the parity relation @xmath59
, the differential decay rate can be expressed as @xmath60 , \label{eq : a}\end{aligned}\ ] ] where @xmath61 is the angle between @xmath62 and @xmath63 measured in the @xmath50 rest frame .
the @xmath64 distribution of the decay rate is obtained by the integration over @xmath65 , @xmath66\nonumber\\ & = & \frac{d\gamma_{t+}}{dy}+\frac{d\gamma_{t-}}{dy } + \frac{d\gamma_{l+}}{dy}+\frac{d\gamma_{l-}}{dy}~ , \label{eq : dgdy}\end{aligned}\ ] ] where @xmath67 , @xmath68 are defined as the transerve and longitudinal contribution to the decay rate with @xmath69 final baryon helicity , respectively .
+ the numerical results are obtained by inputting the form factors discussed in last section .
we have taken @xmath70 , @xmath71 , @xmath72 and @xmath73 .
the partial decay distributions are plotted as a function of @xmath64 in fig.1 and fig.2 for both qcd sum rule and large @xmath3 predictions .
it is easy to see the dominance of @xmath74 and @xmath75 over other plus helicity components . as discussed in ref .
@xcite , this is due to the left - handed v - a current . from fig.2 , we can see that the discrepancy of different model gets larger as @xmath64 goes larger . as a result of the fact that the slope of the isgur - wise function of large @xmath3 is steeper than that of qcd sum rule , the decay distributions on @xmath64 predicted by large @xmath3 are smaller than that of qcd sum rule explicitly when @xmath76 .
it is experimentally useful to calculate the lepton energy distribution .
we obtain , are different from that given in ref .
] @xmath77 \nonumber\\ & \equiv&\frac{d\gamma_{t_+}}{de_l } + \frac{d\gamma_{t_-}}{de_l } + \frac{d\gamma_{l_+}}{de_l } + \frac{d\gamma_{l_-}}{de_l } \label{eq : b}\end{aligned}\ ] ] where @xmath78 the lepton energy spectrums of the decay rates are given in fig.3 and fig.4 for the qcd sum rule and large @xmath3 isgur - wise function . as in the case of @xmath64 distribution ,
the helicity minus components dominate the plus ones . and the decay distributions on @xmath79 by large @xmath3 are always smaller than that by qcd sum rule .
the decay rates are obtained from eq.([eq : dgdy ] ) by integrating over @xmath64 , or from eq.([eq : b ] ) by integrating over @xmath79 .
the numerical results for the partial decay rates into given helicity states are listed in table 1 , where the quark model results @xcite are also listed for comparison .
the total decay rate is obtained by summing them up , @xmath80 for qcd sum rule linear fitting , and @xmath81 for large @xmath3 approximation .
the quark model results given in ref .
@xcite are @xmath82 and @xmath83 for @xmath84 .
the qcd sum rule predicts larger decay branching ratio than large @xmath3 model , as we have expected .
we also see that both qcd sum rule and large @xmath3 predicts larger results for the decay than quark model of ref .
up to the leading order , we have @xmath85 for qcd sum rule and @xmath86 for large @xmath3 limit .
it means that @xmath1 and @xmath2 corrections yield about @xmath87 enhancement for the total decay rate . + now let s turn to the various asymmetry paremeters .
the polarization effects in the process @xmath0 are revealed in various angular distributions .
first , from eq.([eq : a ] ) , the polar angle distribution is @xmath88 where @xmath89 and @xmath90 are asymmetry parameters which can be expressed as @xmath91 there are other asymmetry parameters if the successive hadronic cascade decay @xmath92 where @xmath93 and @xmath94 are some hadrons , are considered .
two more angles are involved , @xmath95 and @xmath20 .
@xmath95 is the angle beween @xmath15 s momentum in @xmath8 rest frame and @xmath93 s momentum in @xmath15 rest frame assuming @xmath96 and @xmath97 .
@xmath20 is the relative azimuthal angle between the decay planes defined by @xmath98 and @xmath99 .
@xmath95 and @xmath20 distributions of the decay are @xcite @xmath100 where @xmath101 is the asymmetry parameter in the @xmath15 hadronic decay . in this case , the related asymmetry parametes , @xmath102 and @xmath103 are given by @xmath104 @xmath105 when the @xmath8 polarization is further considered , additional asymmetry parameters can be introduced .
the new decay angles related are @xmath106 and @xmath107 .
@xmath106 is the angle between @xmath8 polarization and @xmath15 momentum , and @xmath107 is the azimuthal angle between the plane of @xmath8 polarization , @xmath15 momentum and that of @xmath93 , @xmath94 s momenta .
the decay distributions are @xcite @xmath108 where @xmath109 is the degree of polarization of @xmath8 .
the asymmetry parameters @xmath110 and @xmath111 are @xmath112 @xmath113 these asymmetry parameters are functions of @xmath64 . on averaging over @xmath64 , the numerators and denominators are integrated separately with proper weight , @xmath114 . +
our numerical results on the mean values of the asymmetry parameters are listed in table 2 where both qcd sum rule and large @xmath3 results are given .
note that these results include @xmath1 and @xmath2 corrections .
the quark model results @xcite are also listed for comparison .
note that all the results include @xmath1 and @xmath2 corrections .
in this section , we will consider the two - body nonleptonic decay modes @xmath115 and @xmath116 .
all these decays involve external @xmath117-emission diagrams which can be analyzed by using the factorization approximation @xcite .
@xmath115 also get contributions from internal @xmath117-emission which , however , is non - factorizable and is difficult to calculate reliably
. we will simply neglect this contribution in the analysis .
penguin diagrams do not contribute to these decays .
there are contributions from w - exchange diagrams in both @xmath115 and @xmath116 channels . because such diagrams are suppressed by the possibility of the two valence quark lines meeting in the region of @xmath118 , they are neglected .
this is also justified by a detailed quark model analysis for b - baryon nonleptonic decays [ 20 ] . in short ,
the decays will be analyzed by using factorization assumption which is expected to be good except for the @xmath119 channels . in a recent study @xcite , the non - factorizable effect in decay
@xmath120 has been estimated .
the total non - factorizable contribution is about 30% of the factorizable one .
although it is indeed sizable , the factorizable effect is still dominant .
after factorization , the amplitude of the process can be expressed by the product of two matrix elements to which the form factors given in sec.ii can be applied just as in the case of the semileptonic decay . with the definition of the @xmath0 matrix element eq.([eq : form ] ) , the widths of the decays into pseudoscalar meson(p ) and vector meson(v ) can be easily calculated : @xmath121[(1-r)^2-t_p^2]}\nonumber\\ % & & \big\{[(1+r)^2-t_p^2]|a|^2+[(1-r)^2-t_p^2]|b|^2\big\ } \frac{r^2\kappa_p}{(1-\kappa_p^2)^2}\big[|a|^2+\kappa_p^2|b|^2\big]~,\ ] ] where @xmath122~,\nonumber\\ b&=&(1+r)\big[g_1-\frac{1-r}{2}\big(g_2+\frac{g_3}{r}\big)\big]~,\nonumber\\ \kappa_{p(v)}&=&\sqrt{\frac{(1-r)^2-t_{p(v)}^2}{(1+r)^2-t_{p(v)}^2}}~,\end{aligned}\ ] ] and @xmath123~,\ ] ] where @xmath124 here @xmath125 denotes @xmath126 for @xmath127 and @xmath128 for @xmath129 , and @xmath130 , @xmath131 where @xmath132 is the mass of the pseudoscalar ( vector ) meson . and @xmath133 .
@xmath134 is the qcd coefficient which is taken as a free parameter in the discussion of nonleptonic decays @xcite .
the above expressions are spin averaged results .
if we take into account the spin effects , the spin up - down asymmetry of @xmath8 is a good parameter to analyze @xmath0 nonleptonic decay . in this case , the decay rates are @xcite @xmath135 where @xmath136 and @xmath137 where @xmath138}~.\ ] ] here @xmath139 is the spin vector of @xmath140 and @xmath141 is the unit vector of momentum of @xmath15 .
@xmath142 is the spin up - down asymmetry parameter of @xmath8 .
the numerical results from qcd sum rule and large @xmath3 limit are given in table 3 . because @xmath64 is near to 1.45 , in the qcd sum rule case
, we used the nonlinear fitting of the isgur - wise function @xmath143}$ ] @xcite .
the value of @xmath134 is taken to be 0.98 which is obtained from the decay @xmath144 @xcite .
various decay constants are taken as follows ; @xmath145 , @xmath146 and @xmath147 . the quark model results of ref .
@xcite are also listed for comparison .
we have analyzed the @xmath0 semileptonic decay in the framework of hqet to the order of @xmath1 and @xmath2 .
the predictions for the isgur - wise functions and mass parameters from qcd sum rule and large @xmath3 method are used . in the large @xmath3 limit
, we argued that the subleading isgur - wise function vanishes .
the decay rates , distributions and varuous asymmetry parameters are calculated numerically . some of the @xmath8 nonleptonic decays are also calculated .
the numerical results can be checked by the experiments in the near future .
+ we would like to thank jungil lee for helpful discussion .
this work was supported in part by the kosef through src program , and in part by the korean ministry of education , project no .
bsri-97 - 2418 . after finishing this work
, we noticed a paper @xcite which calculated the nonleptonic @xmath8 decays in the large @xmath3 and heavy quark limits .
e. jenkins , a.v .
manohar and m.b .
wise , nucl . phys .
* b 396 * ( 1993 ) 27 ; + c.g .
callan and i. klebanov , nucl .
* b 262 * ( 1985 ) 222 ; + m. rho , d.o .
riska and n.n .
scoccola , phys . lett .
* b 251 * ( 1990 ) 597 ; z. phys . * a 341 * ( 1992 ) 343 ; + y. oh , d .-
min , m. rho and n.n .
scoccola , nucl . phys . * a 534 * ( 1991 ) 493 .
c. chow and m.b .
wise , phys .
d 50 * ( 1994 ) 2135 ; + c. liu , phys .
* b 389 * ( 1996 ) 347 .
krner and m. krmer , phys . lett .
* b 275 * ( 1992 ) 495 ; + p.bialas , j.g .
krner , m. krmer and k. zalewski , z. phys .
* c 57 * ( 1993 ) 115 .
b. knig , j.g .
krner , m. krmer and p. kroll , phys . rev . *
d 56 * ( 1997 ) 4282 .
m. bauer , b. stech and m. wirbel , z. phys .
* c 34 * ( 1987 ) 103 .
for a recent review , see m. neubert and b. stech , hd - thep-97 - 23 , hep - ph/9705292 .
cheng , phys . rev . *
d 56 * ( 1997 ) 2799 .
m. a. ivanov , j. g. krner , v. e. lyubovitskij , and a. g. rusetsky , hep - ph/9709372 ; hep - ph/9710523 .
a.k . giri , l. maharana and r. mohanta , hep - ph/9710285 . fig 1 .
@xmath64 distribution of the decay rates for ( a ) qcd sum rule and ( b ) large @xmath3 limit .
@xmath148 and @xmath149 are abbreviated as @xmath67 and @xmath68 , respectively .
+ .3 cm fig 2 .
comparison of the two models in the helicity componential @xmath64 distribution of the decay rates : ( a ) @xmath150 ( b ) @xmath151 ( c ) @xmath152 ( d ) @xmath153 ( e ) @xmath154 .
+ .3 cm fig 3 .
lepton energy distribution of the decay rates for ( a ) qcd sum rule and ( b ) large @xmath3 limit .
+ .3 cm fig 4 .
comparison of the two models in the helicity componential lepton energy distribution of the decay rates : ( a ) @xmath155 ( b ) @xmath156 ( c ) @xmath157 ( d ) @xmath158 ( e ) @xmath159 .
+ table 1 . the partial decay rates(in @xmath160 ) .
+ .3 cm table 2 .
asymmetry parameters .
+ .3 cm table 3 .
numerical results for @xmath8 two - body nonleptonic decays .
the qcd coefficient @xmath134 is taken to be 0.98@xcite . | the @xmath0 semileptonic decay is analyzed in the framework of heavy quark effective theory to the order of @xmath1 and @xmath2 .
the qcd sum rule and large @xmath3 predictions to the decay form factors are applied .
it argues that the subleading baryonic isgur - wise function in the large @xmath3 limit vanishes .
the decay rates , distributions and asymmetry parameters are calculated numerically .
some of the nonleptonic decay modes are discussed in the end . |
in cold dark matter ( cdm ) cosmology , galaxies form hierarchically and present - day galaxies are the products of successive mergers @xcite .
recent observations show that almost all galaxies harbor supermassive black holes ( smbhs ) at their centers @xcite and the black hole ( bh ) masses tightly correlate with the properties of their host galaxies such as the mass of stellar bulge @xcite , the bulge luminosity @xcite , and the nuclear stellar velocity dispersion @xcite .
the correlations between bh mass and galaxy properties imply that the growth of smbhs and the formation and evolution of galaxies are closely linked .
the correlation is likely induced by galaxy major mergers ( merging of galaxies with comparable mass ) in which both rapid star formation and gas accretion onto smbhs are triggered and the feedback from the central active galactic nuclei ( agns ) regulates the growth of both smbhs and galaxy bulges .
the coevolution scenario can successfully explain not only the correlations between smbhs and their host galaxies but also many of the observed evolutions of galaxies and agns @xcite .
during galaxy mergers part of the gas originally in galactic plane is driven to galaxy center , triggering starburst and the accretion of central smbh @xcite . if both galaxies in the merging system harbor smbhs at their centers , a pair of agns could form which is an intriguing object for both observations and theories .
because of dynamical friction the two smbhs quickly sink to the common center of the two merging galaxies , forming a bound , compact ( @xmath2 pc ) supermassive black hole binary ( smbhb , * ? ? ?
* ) , which would be difficult to resolve in any but the closest galaxies using current telescopes . as the separation of the binary shrinks , dynamical friction becomes less and less efficient and
three - body interaction between smbhb and the stars passing by becomes more and more important . at separations
@xmath3 pc smbhb becomes hard in the sense that the binary loses energy and angular momentum mainly through three - body interaction .
the stars passing by the binary , if not tidally disrupted or swallowed by the bhs , would be expelled by the `` slingshot effect '' ( * ? ? ?
* q96 thereafter ) with velocities comparable to the host galaxy s escape velocity .
if the refilling of the reservoir of stars on orbits that interact with the smbhb is inefficient , the evolution of smbhb slows down and the evolution timescale may even exceed the hubble time , with the smbhb separation stalling well outside the radii needed for the final gravitational wave radiation and coalescence stage ( * ? ? ?
* y02 hereafter ) .
this seems inconsistent with that fact that smbhbs have not yet been observed in nearby galaxies .
therefore , many stellar - dynamical mechanisms are proposed to boost the hardening rate of smbhbs in gas - poor systems , including non - equilibrium stellar distribution @xcite , brownian motion of smbhb @xcite , and non - spherical ( axisymmetric , triaxial , or irregular ) stellar distribution ( y02 ; * ? ? ?
* ; * ? ? ?
however , the conclusions in the above scenarios depend on processes which are not well understood , so it is still unclear whether smbhbs could coalesce within a hubble time using purely stellar - dynamical processes . in gas - rich systems it has been argued that the interaction between binary bhs and the gaseous environment could efficiently drive the binary to the gravitational radiation dominant domain within a timescale of @xmath4 @xcite .
although many physical processes have been proposed in the literature to boost the hardening rates of smbhbs , the abundance of binary bhs in galaxy centers has not yet been strongly constrained observationally .
because of their compactness , bound smbhbs are very difficult to resolve directly with current telescopes , and only unbound smbhbs in a couple of gas - rich ( wet ) galaxy mergers have been identified @xcite .
so far all the evidence for bound and hard smbhbs are indirect and model dependent .
the prototype evidence for smbhbs in agns is the helical morphology of radio jets in many radio galaxies which may be due to the precession of jet orientation in a smbhb system @xcite .
it has been suggested that the periodic outbursts observed in some agns @xcite are due to the orbital motion of the jet - emitting bh in smbhb @xcite , the precession of spin axis of the rotating primary bh @xcite , the interaction between smbhb and a standard accretion disk @xcite or an advection dominated accretion flow ( adaf ) @xcite .
x - shaped features in a subclass of radio galaxies have been attributed to the interaction and alignment of smbhbs and standard accretion disks ( * ? ? ?
* alternatives see @xcite ) .
there is also some evidence for the coalescence of smbhbs in agns .
@xcite suggested that the double - double radio galaxies and the restarting jet formation in some radio galaxies are due to the removal and refilling of inner accretion disk because of the interaction with smbhbs .
although there is much circumstantial evidence for both active and coalesced smbhbs in agns , it is still unclear what fraction of smbhbs would coalesce during or before the agn phase and how many could survive to the later dormant or inactive phase . because present observations can not give constraints on the physical processes which have been proposed to boost the hardening rates of smbhbs , @xcite suggested measuring the acceleration of jet precession as a way of determing the hardening rates of smbhbs in agns and to place strong constraints on models of smbhb evolution .
in inactive galaxies or gas - poor mergers the observational evidence for smbhbs is very rare .
one possible way to identify uncoalesced smbhbs in inactive galaxies is to detect hypervelocity binary stars ejected by smbhbs with the three - body sling - shot effect @xcite .
however , this method can not be applied to distant galaxies .
therefore , in this paper we suggest a way to determine statistically whether smbhbs in nearby galaxies have coalesced .
one inevitable impact of a non - rorating smbh on its stellar environment is that stars passing by the bh as close as the tidal radius @xmath5 would be tidally disrupted @xcite , where @xmath6 and @xmath7 are the radius and mass of the stars and @xmath8 is the bh mass .
part of the stellar debris will be spewed to highly eccentric bound orbits and later fall back onto the bh , giving rise to an outburst decaying within months to years @xcite .
these tidal flares are definitive evidences of smbhs in inactive galaxies @xcite . since the frequency of stellar disruption depends on the surrounding stellar distribution (
* ; * ? ? ?
* thereafter mt99 ) , the stellar disruption rate can be used as a probe of the inner structure of galactic nucleus .
for nearby early - type galaxies recent theoretical calculation with single bh without taking into account general relativistic ( gr ) effects gives an averaged stellar disruption rate of @xmath9 per galaxy if the stellar distribution is spherically symmetric @xcite or possibly @xmath10 orders of magnitude higher if the stellar distribution is non - spherical ( axisymmetric or triaxial , * ? ? ?
however , if smbhbs reside in galactic centers the stellar distribution would be dramatically changed due to the slingshot effect , and the stellar disruption rate would be very different .
@xcite studied the interaction between smbhb and dense galactic cusp ( @xmath11 pc ) for a bound , but non - hard , binary at the dynamical friction stage and found a high disruption rate of @xmath12 . @xcite
considered the stellar disruption by single bhs immediately after smbhb coalescence .
they found that the tidal disruption rate is significantly suppressed due to the low density core forming during the hard stage of smbhbs .
since the two stages investigated by @xcite and @xcite are two short phases in the evolution of smbhbs and binary bhs with mass ratios @xmath13 spend most of their lifetime on the hard stage ( y02 ) , it would be very interesting to calculate the tidal disruption rate of stellar objects for hard smbhbs .
therefore , in this paper we calculate tidal disruption rates for hard and stalling smbhbs .
we suggest that the comparison of expected tidal disruption rate by single smbhs or hard smbhbs and the future survey of x - ray or uv flares at nearby galactic centers could tell us whether most of the smbhs in nearby galaxies are single or binary and whether the smbhbs would coalesce rapidly , as expected by some models .
the stellar disruption rate in a galaxy harboring a smbhb depends on : ( 1 ) the rate that stars are fed from large - scale to the vicinity of smbhb and ( 2 ) the probability that a star moving in the vicinity of the smbhb passes close enough to one of the bhs to be tidally disrupted .
although there has been much work on the former part , intended to study the evolution timescale of smbhb ( q96 ; y02 ; * ? ? ?
* ) , the latter is largely unaddressed in the literature .
because of the chaotic nature of three - body interaction , numerical scattering experiments are needed to give the probability of tidal disruption .
previous numerical experiments on star - smbhb interactions mainly focus on energy and angular momentum exchanges ( q96 ; * ? ? ?
* ) and the limited particle numbers are insufficient to tackle the rare events of very close encounters such as tidal disruptions .
therefore , we begin with intensive numerical scattering experiments ( normally @xmath1 particles in each run ) to calculate the tidal disruption cross sections for hard smbhbs . then we apply our results to a sample of nearby galaxies and estimate the tidal disruption rates of stellar objects both by single and binary bhs in different types of galaxies .
this paper is organized as follows . in [ bg ]
we briefly review some basic properties of hard smbhbs and show how to calculate stellar disruption rate .
we describe our numerical scattering experiments in [ experi ] and present the results in [ results ] . in [ apply ] we calculate stelar disruption rates with realistic galaxy models and discuss the differences between the tidal disruption rates for single and binary bhs .
we also estimate the density of tidal disruption events in local universe and make predictions for future surveys of uv / x - ray flares in [ prediction ] . in [ discussion ]
we discuss the observational signatures of smbhbs in inactive galaxies and the implications for the evolution of binary bhs .
finally we give our conclusions in
[ conclusion ] .
a smbhb with masses @xmath14 and @xmath15 ( @xmath16 ) becomes hard when the semimajor axis @xmath17 decreases to @xmath18 ( q96 ) where @xmath19 is the hardening radius , @xmath20 is the one - dimensional stellar velocity dispersion of background stars , and @xmath21 is the reduced mass with @xmath22 .
observations of nearby galaxies show that bh mass tightly correlates with bulge velocity dispersion @xcite . if the total bh mass @xmath23 of smbhb also follows the empirical @xmath24 relation : @xmath25 @xcite , equation ( [ ah ] ) reduces to @xmath26 where @xmath27 and @xmath28 is the mass ratio of the binary . in this paper
we only consider @xmath29 because the dynamical friction timescale for smbhbs with @xmath30 is very long ( y02 ) .
a hard smbhb loses energy and becomes more bound through three - body interaction @xcite . as a result stars intruding a sphere of radius @xmath31 about the mass center of the binary ( if not tidally disrupted or swallowed by the bhs )
will be eventually expelled with an average energy gain @xmath32 where @xmath33 is a dimensionless factor about @xmath34 ( q96 ; y02 ) . in this paper ,
unless noted otherwise , we assume that galaxies are spherical . in a spherical potential
there is just one family of orbits ( centropilic loops ) and any such orbit can be labelled by its four integrals of motion , namely the specific angular momentum @xmath35 and specific binding energy @xmath36 .
we define the _ loss cone _ to consist of all orbits with pericenters less than @xmath17 , which results in a conical region with the boundary given by @xmath37 $ ] . in a non - spherical galaxy angular momentum
is no longer conserved and there can be more than one orbit family .
nevertheless , instead of a `` loss cone '' one can still define a `` loss zone '' inside which stars will be delivered to the smbhb .
mechanisms such as two - body relaxation tend to refill the loss cone so the decay of smbhb continues .
the stars refilled into loss cone are originally far from the smbhb so that @xmath38 and @xmath39 if the characteristic change of @xmath40 during one orbital period of a star is much smaller than @xmath41 , the loss cone refilling behaves like a diffusion process in which the loss cone filling rate is not sensitive to @xmath42 or @xmath17 ( diffusive regime , * ? ? ?
* ) . otherwise , the stars entering the loss cone have a uniform distribution with respect to @xmath43 so that the loss cone filling rate is proportional @xmath42 or @xmath17 ( pinhole regime ) . in inactive galaxies gas dynamics is probably not important , so a hard binary loses energy mainly through interacting with loss cone stars .
the evolution timescale @xmath44 at this hard stage depends on the loss cone filling rate @xmath45 , the number of stars fed into a sphere of radius @xmath17 about the mass center of the binary per unit time . from equation ( [ deltae ] ) and
@xmath46 we have @xmath47\label{th}\end{aligned}\ ] ] ( see also y02 ) , which implies that about an amount of @xmath48 of stars should be consumed to reduce @xmath17 by a single @xmath49-fold , in agreement with recent simulations @xcite .
a fraction of the stars fed into the loss cone of smbhb would be scattered to the vicinity of a bh and get tidally disrupted . the ratio of the tidal radius of the primary bh and the separation of the binary is about @xmath50 ( from eq .
[ [ rt ] ] and [ [ ahpc ] ] ) , which is small for typical hard smbhbs this implies that most stars in the loss cone would be expelled instead of being tidally disrupted .
to further quantify the probability of tidal disruption , we introduce @xmath51 ( @xmath52 ) , the cross section for stars to be scattered to a distance less than @xmath53 from the @xmath54th bh . the probability that a loss cone star is disrupted by the @xmath54th bh estimates @xmath55 , where @xmath56 is the tidal radius of the @xmath54th bh .
then the stellar disruption rate is related to the loss cone filling rate and the tidal disruption cross section as @xmath57 for a single bh the geometrical cross section can be written out analytically as @xmath58 which increases with the gravitational potential @xmath59 and decreases with stars kinetic energy @xmath60 at infinity .
for bound stars equation ( [ singlecs ] ) is still valid as long as the stars fall to the bh on near - radial orbits , and in the bound case @xmath61 is the velocity at apocenter .
when gravitational focusing is important ( @xmath62 ) the cross section scales as @xmath53 . in smbhb systems
the presence of a companion bh tends to increase the term @xmath59 by deepening the gravitational potential and also increase the term @xmath60 by inducing orbital motion which raises the relative velocity between each binary member and ambient stars .
therefore in the rest frame of the primary member , the equivalent increase in @xmath63 is @xmath64 and the equivalent increase in @xmath60 is about @xmath65 .
so the cross section of the primary bh is approximately @xmath66 where @xmath67 is a correction factor of order unity .
we will see below that for different @xmath68s the best fit is obtained when @xmath69 . for the secondary , the roles of @xmath14 and @xmath15 exchange , so the cross section can be obtained by replacing @xmath68 in equation ( [ appcs ] ) with @xmath70 .
quinlan has separately explained the behavior of the cross sections at @xmath71 and @xmath72 from another way of understanding ( q96 ) , while our explanation and empirical formula are general for any @xmath73 .
because of the effect of gr equation ( [ appcs ] ) is no longer valid when @xmath74 , where @xmath75 is the schwarzschild radius . since the ratio of @xmath76 and @xmath77 is @xmath78 gr effect is important when @xmath79 ( @xmath80 for solar type stars ) . in practice , one can use the pseudo - newtonian potential @xmath81 @xcite to simulate the gr effect so equation ( [ appcs ] ) becomes @xmath82 ( notice that @xmath83 ) , equivalent to multiplying the non - relativistic cross section by a correction factor @xmath84 .
equation ( [ grappcs ] ) diverges when @xmath85 , but since a star plunging into the sphere of marginally bound radius @xmath86 about a bh eventually falls into to the event horizon , the cross section at @xmath87 $ ] should be constant , equal to the cross section at @xmath88 .
both equation ( [ appcs ] ) and ( [ grappcs ] ) are evaluated at the first time that a star passes the binary . in some cases
a star could be scatted onto temporally bound orbit and encounters with the binary many times before expelled or disrupted ( @xcite ; q96 ) . during such multi - encounters the probability of tidal disruption is expected to be enhanced .
since it is very difficult to derive analytical cross sections for these complicated encounters , scattering experiments are needed .
stars on near - radial orbits have @xmath89 , then equations ( [ rtah ] ) , ( [ appcs ] ) , and ( [ grappcs ] ) suggest that the probability of tidally disruption scales as @xmath90 , which is of order @xmath91 .
therefore , to get statistically meaningful tidal disruption cross sections a large number of particles should be used in the scattering experiments .
in this section we describe our numerical scattering experiments of the restricted three - body problem .
the method adopted here is similar to those in q96 and @xcite but the number of test particles in our experiments is much larger , typically @xmath1 , to give statistically meaningful cross sections at @xmath92 . in the scattering experiments , instead of using @xmath36 , the specific binding energy of a particle in the combined potential of stars and bhs
, we use a more convenient parameter @xmath93 , the specific binding energy about the bh binary , where @xmath53 is the radius about the mass center and @xmath94 is the velocity of particle .
we denote the initial specific binding energy with @xmath95 and initial velocity with @xmath61 . in our problem
the relevant energy range is @xmath96 ( from eq .
[ [ ah ] ] and eq .
[ [ energy ] ] ) . in each scattering experiment ,
the origin is chosen at the mass center of the binary . in the case
@xmath97 ( unbound case ) , initially particles come form infinity @xmath98 with asymptotic velocity @xmath99 , where @xmath100 is the impact parameter , the minimum separation between particle and mass center if the particle feels no gravitational field .
the fiducial value for @xmath61 is @xmath101 , to reproduce fig . 3 in q96 . given @xmath61 , @xmath102
is uniformly sampled in the range @xmath103 $ ] , where @xmath104 .
particles with @xmath105 hardly reach @xmath106 therefore contribute little to the tidal disruption cross section ( see [ jtheta ] ) . in the other case @xmath107
( bound case ) , initially we put particles at @xmath108 with an isotropic velocity distribution ( @xmath109 ) .
@xmath110 satisfies @xmath111 because in realistic smbhb systems most stars enter the loss cone on near - radial orbits .
another six parameters ( four if the binary is circular ) are set to fix the initial conditions of the binary : ( 1 ) the mass ratio @xmath68 ; ( 2 ) the eccentricity of binary orbit @xmath49 ; ( 3 ) the inclination of orbital plane ; ( 4 ) the argument of pericenter ; ( 5 ) the longitude of ascending node ; ( 6 ) the initial binary phase . in each set of experiments with fixed @xmath68 and @xmath49 , the cosine of the orbital inclination angle
is evenly sampled in @xmath112 $ ] and the other three angular parameters are uniformly distributed in @xmath113 $ ] , resulting in an isotropically filling of the loss cone . given the above initial conditions , the motion of a massless particle is governed by the following coupled first - order differential equations : @xmath114 where @xmath115 is the position of the @xmath54th bh . in each scattering experiment
we first move a particle from its initial position to @xmath116 along a keplerian orbit about a point mass @xmath23 at the origin . at @xmath117
the quadrupole force from the binary is eight orders of magnitude smaller than @xmath118 .
then we integrate the particle s orbit with the subroutine * dopri8 * @xcite , an explicit runge - kutta method of order ( 7)8 .
we set the threshold of fractional error per step in @xmath119 and @xmath120 to be @xmath121 .
raising this threshold up to @xmath122 does not significantly change our results .
the integration stops if one of the following conditions is satisfied ( 1 ) the particle leaves the sphere of radius @xmath117 with negative binding energy ; ( 2 ) the physical integration timescale exceeds @xmath123 yr ; ( 3 ) the integration timestep reaches @xmath124 . a small fraction ( @xmath125 percent ) of the particles
are scattered to wide , bound orbits and survive many revolutions so conditions ( 2 ) and ( 3 ) are adopted to save computational time .
we have tested our code by reproducing fig .
3 in q96 and found very good agreement .
we have also used the pseudo - newtonian potential ( eq.[[grp ] ] ) in our numerical experiments to investigate the effect of gr . in this case equation ( [ accel ] ) becomes : @xmath126 and we stop the integration once a particle reaches @xmath127 schwarzschild radius about either of the bhs . in the gr experiment
the ratio of @xmath128 and @xmath19 , @xmath129 should be set in priori .
for the illustrative purpose we always set @xmath130 .
we will discuss the effect of changing @xmath131 on tidal disruption cross sections in
[ grcs ] . after each scattering experiment
we record the minimum separation between the particle and each bh . at the end of all experiments we count the number @xmath132 of particles whose minimum separations from the @xmath54th bh are less than @xmath53 .
then the normalized multi - encounter cross section ( particle are allowed to encounter the binary as many times as they could until they are expelled ) is calculated with @xmath133 .
the poissonian error in the counts is @xmath134 , so the error for the normalized cross section is @xmath135 and the corresponding fractional error in statistics is @xmath136 .
following q96 , we also record the minimum separation between a particle and each bh during their first encounter and derive the single - encounter cross sections . in [ nongrcs ]
we compare the single - encounter cross sections and those from the analytical approximations ( eq . [ [ appcs ] ] and [ [ grappcs ] ] ) .
first we consider the unbound case .
we set @xmath137 ( after fig . 3 in q96 ) ,
@xmath138 , and use the newtonian potential .
the normalized mullti- and single - encounter cross sections from @xmath139 particles are presented as solid lines in the top panel of figure [ cs001 ] .
the cross sections are plotted as a function of @xmath73 so that they can be easily scaled .
the perturbations at the lower ends are due to statistical fluctuation . in the top panel of figure [ cs001 ] we also plot the empirical single - encounter cross sections ( see eq . [ [ appcs ] ] ) as dashed lines .
we estimate the fractional error of the empirical formula with @xmath140 $ ] , where @xmath141 is from scattering experiments and @xmath142 from equation ( [ appcs ] ) . @xmath143 and
@xmath144 for the primary and secondary bhs are shown in the middle and bottom panels of figure [ cs001 ] .
for both bhs we found that at @xmath145 @xmath146 is almost always below @xmath144 , indicating that the empirical cross sections agrees very well with the numerical results . at @xmath147 , although @xmath146 is larger than @xmath148 , the difference does not exceed @xmath149 .
+ for @xmath150 and @xmath151 , the numerical ( solid lines ) and empirical approximations ( dashed lines ) cross sections are presented in the top left and top right panels of figure [ fitnongr ] .
@xmath148 ( solid lines ) and @xmath146 ( dotted lines ) for the single - encounters are presented in the middle and bottom panels .
although the approximations become worse when @xmath68 is large , the difference between approximation error and statistic fluctuation is always below @xmath149 .
+ figure [ cs001 ] and figure [ fitnongr ] show that at @xmath152 both the multi- and single - encounter cross sections scales linearly as @xmath73 , implying that gravitational focusing is important for very close encounters .
the cross sections at @xmath153 could be obtained by using this scaling .
it is also clear that the normalized multi - encounter cross sections are greater than the single - encounter ones as is predicted in [ cstheory ] .
however , we find that in our experiments although a particle could encounter with the binary many times , no particle more than once enters the sphere of tidal radius ( for @xmath130 ) about either bh .
this is because the probability for a particle to enter the sphere of tidal radius about the @xmath54th bh @xmath154 times roughly scales as @xmath155^n$ ] or @xmath156 , which becomes extremely low for @xmath157 and @xmath158 .
for the similar reason , the event that a particle successively approaches the tidal radii of the primary and secondary bhs is also extremely rare . to investigate the effect of binding energy
, we carried out two test experiments for @xmath159 and @xmath160 ( in unit of @xmath161 ) , each with @xmath162 particles and @xmath138 . in the unbound case
the asymptotic velocity of the intruding particles is @xmath163 , slightly greater than the limit @xmath164 ( from the fitting formula eq .
17 in q96 ) above which the binary becomes soft . in the bound case ,
initially the particles are at @xmath165 with an isotropic velocity of @xmath163 .
we do not consider @xmath166 because such initial condition results in unrealistically bound particles with @xmath167 .
+ results from these test experiments are presented in figure [ bound ] as solid lines .
cross sections for the fiducial value @xmath168 ( from [ nongrcs ] ) are also plotted for reference .
although the physical cross sections increase prominently with @xmath95 which is expected by equation [ appcs ] , figure [ bound ] shows that in both bound and unbound cases the normalized cross sections for both multi- and single - encounters seem not varying significantly with @xmath95 .
this is because when @xmath169 , the velocity of the particles passing @xmath170 from the binary is not sensitive to @xmath95 but to the depth of the gravitational potential at @xmath53 .
therefore , different particles passing by the same bh binary feel the similar strength of gravitational focusing and have similar interaction cross sections with the bh components of binary . since the normalized cross sections do not vary significantly with @xmath95 as long as @xmath169 , it is reasonable to apply the fiducial cross sections to various binaries with a wide range of semimajor axis .
for example , according to equation ( [ ah ] ) , varying @xmath95 from @xmath171 to @xmath172 corresponds to increasing @xmath17 from @xmath173 to @xmath174 . in both experiments for @xmath175 and @xmath176
initially the particle follow a uniform distribution in @xmath43 and the loss cone filling rate is isotropic , i.e. , the cross sections obtained in [ nongrcs ] and [ boundstars ] have been equally averaged over the whole loss cone and over all directions .
such averaging may be oversimplified in realistic smbhb systems . to investigate the dependence of the cross sections on the amplitude of initial angular momentum , in figure [ dfjdr ] we plot the differential tidal disruption cross sections @xmath177 with respect to @xmath43 .
the triangles and squares , respectively , refer to the primary and secondary bhs .
@xmath177 is obtained as follows . for each bin of
@xmath178 we count the number @xmath179 of particles whose initial angular momenta fall in this bin and minimum separations from the @xmath54th bh are less than @xmath56 . here
@xmath56 is from equation ( [ rtah ] ) with @xmath130 and we only use the data for multi - encounters . then @xmath180 is calculated with @xmath181 so that the integration over @xmath43 results in the normalized cross section of tidal disruption .
the poissonian errors in the counts have been indicated by the error bars in figure [ dfjdr ] .
note that the differential cross section [ or @xmath182 alone ] derived in this way is proportional to the probability of tidal disrupting a particle with initial angular momentum @xmath40 . generally speaking ,
both differential cross sections for the primary and secondary bhs are flat at @xmath183 .
the differential cross sections start to cut off above @xmath42 and particles with initial angular momentum greater than @xmath184 contribute little to the cross sections . when @xmath185 the differential cross section for primary bh rises steeply with decreasing @xmath43 at @xmath186 but even in this case the contribution to the total cross section from particles with larger @xmath40 is still significant .
+ the cross sections also depend on the direction of the initial angular momentum vector @xmath187 .
we define @xmath188 as the relative angle between @xmath187 and binary s orbital angular momentum .
figure [ dfthetadr ] shows the differential tidal disruption cross sections with respect to @xmath189 .
the differential cross sections are obtained following the same described above . when @xmath185 the particles on corotating orbits ( @xmath190 ) are more likely to be disrupted , while for secondary bhs the dependence of the cross sections on @xmath188 is weak .
unless the stellar distribution is extremely flattened or the star cluster around smbhb is significantly rotating the total cross section would not differ much from those presented in figures [ cs001 ] and [ fitnongr ] . + to see clearly the effect of the eccentricity of binary s orbit , we set an extreme eccentricity @xmath191 and carried out two test experiments for @xmath138 and @xmath192 , each with @xmath124 particles .
the results ( solid lines ) are presented in figure [ quilane ] .
cross sections for @xmath193 ( dotted lines , from [ nongrcs ] ) are also plotted for reference . increasing eccentricity
only slightly increase the normalized multi - encounter cross sections for both bh components .
to study the effect of gr we set the gravitational potential to be pseudo - newtonian with equation [ [ grp ] ] and @xmath130 and then repeated the scattering experiments . for each @xmath68
the result from @xmath194 particles is shown in figure [ fit ] as solid lines .
dotted lines are results from the non - relativistic experiments . below the marginally bound radius @xmath86 both
the cross sections for multi- and single - encounters become constant , representing the relativistic effect .
+ the dashed lines are empirical cross section in the gr case . for the single - encounters we have used equation ( [ grappcs ] ) and set the cross sections constant between @xmath195 and @xmath86 .
the resulting cross sections agree well with the numerical ones .
the approximate cross sections to the multi - encounters are obtained by multiplying the non - relativistic cross sections by the correction factor @xmath196 and keeping the cross sections constant below @xmath86 .
these resulting cross sections well agree with the numerical ones as long as @xmath197 , i.e. , in the gravitational focusing dominated regime . due to
gr effect the cross sections at @xmath195 are four times greater than the non - relativistic ones .
for other values of @xmath131 the location of schwarzschild radius moves along the @xmath73 axis . instead of repeating the whole simulation we use non - gr cross sections to calculate the approximate relativistic cross sections , following the procedure described above .
in this section we estimate the tidal disruption rates ( eq .
[ nd ] ) of stellar objects for a sample of nearby galaxies .
the largest uncertainty in our calculation comes from the loss cone refilling rate @xmath45 ; for simplicity , and to ease comparison with earlier work , we assume that two - body relaxation is the only process contributing to @xmath45 .
our galaxy sample consists of @xmath198 nearby elliptical galaxies which are listed in table [ table1 ] . for each galaxy
the bh mass is obtained with @xcite s @xmath199 relation , where the bulge mass @xmath200 is from the scaling relation between stellar mass and luminosity @xcite .
the assumed mass to light ratios and the resulting bh masses are given by columns 2 and 3 in table [ table1 ] .
lcccrrrrrrrrr a2052 & 7.41 & 9.17 & -5.89 & 1.76 & 13.55 & -4.88 & -9.75 & -12.54 & -10.33 & -11.87 & -11.11 & -11.41 + ngc 1023 & 4.88 & 7.84 & -4.13 & 0.25 & 9.98 & -2.65 & -6.47 & -9.71 & -7.03 & -8.54 & -7.59 & -7.61 + ngc 1172 & 5.39 & 8.16 & -4.21 & 1.09 & 10.50 & -2.85 & -7.38 & -10.20 & -7.97 & -9.49 & -8.76 & -9.06 + ngc 1316 & 7.60 & 9.25 & -4.64 & 0.89 & 12.46 & -3.71 & -7.72 & -10.47 & -8.25 & -9.80 & -9.04 & -8.94 + ngc 1399 & 6.33 & 8.67 & -5.28 & 0.87 & 12.73 & -4.57 & -8.73 & -11.60 & -9.27 & -10.83 & -10.10 & -10.10 + ngc 1400 & 5.69 & 8.33 & -4.79 & 0.97 & 11.49 & -3.67 & -8.05 & -11.33 & -8.59 & -10.17 & -9.57 & -9.57 + ngc 1426 & 5.04 & 7.95 & -4.44 & 0.95 & 10.38 & -2.94 & -7.45 & -10.29 & -7.97 & -9.51 & -8.78 & -8.78 + ngc 1600 & 7.44 & 9.19 & -5.80 & 1.33 & 13.81 & -5.13 & -9.57 & -12.49 & -10.13 & -11.70 & -10.97 & -10.91 + ngc 1700 & 6.25 & 8.63 & -4.37 & 1.18 & 10.84 & -2.71 & -7.22 & -10.06 & -7.74 & -9.24 & -8.55 & -8.55 + ngc 221 & 2.71 & 5.96 & -3.68 & -3.34 & 8.52 & -3.16 & -4.96 & -6.48 & -4.84 & -6.07 & -5.37 & -5.37 + ngc 224 & 4.62 & 7.67 & -4.54 & -0.47 & 10.91 & -3.75 & -7.00 & -9.81 & -7.49 & -8.98 & -8.04 & -8.10 + ngc 2636 & 3.93 & 7.15 & -5.05 & 0.88 & 9.95 & -3.30 & -7.99 & -11.30 & -8.53 & -10.10 & -9.49 & -9.49 + ngc 2832 & 7.76 & 9.32 & -5.53 & 1.57 & 13.57 & -4.75 & -9.41 & -12.69 & -9.94 & -11.44 & -10.95 & -10.95 + ngc 2841 & 4.66 & 7.69 & -4.59 & 0.74 & 10.21 & -3.02 & -7.36 & -10.68 & -7.92 & -9.46 & -8.92 & -8.92 + ngc 3115 & 5.39 & 8.16 & -4.03 & 0.56 & 10.19 & -2.53 & -6.56 & -9.59 & -7.12 & -8.62 & -7.87 & -7.80 + ngc 3377 & 4.53 & 7.60 & -3.85 & -0.40 & 10.04 & -2.94 & -6.28 & -9.07 & -6.77 & -8.26 & -7.37 & -7.36 + ngc 3379 & 5.21 & 8.05 & -4.93 & 0.63 & 11.48 & -3.93 & -8.07 & -10.89 & -8.62 & -10.13 & -9.31 & -9.31 + ngc 3599 & 4.55 & 7.61 & -4.61 & 0.92 & 10.16 & -3.05 & -7.63 & -10.41 & -8.18 & -9.70 & -9.04 & -9.26 + ngc 3605 & 4.13 & 7.31 & -4.78 & 0.44 & 10.00 & -3.19 & -7.36 & -10.26 & -7.94 & -9.42 & -8.79 & -8.87 + ngc 3608 & 5.48 & 8.21 & -4.79 & 0.94 & 11.12 & -3.41 & -7.80 & -11.07 & -8.35 & -9.91 & -9.31 & -9.31 + ngc 4168 & 6.39 & 8.70 & -5.56 & 1.20 & 12.77 & -4.58 & -9.09 & -11.93 & -9.61 & -11.11 & -10.45 & -10.39 + ngc 4239 & 3.50 & 6.78 & -5.24 & -0.02 & 10.02 & -3.74 & -7.67 & -10.47 & -8.19 & -9.70 & -8.90 & -8.90 + ngc 4365 & 6.71 & 8.86 & -5.24 & 0.96 & 12.82 & -4.46 & -8.65 & -11.50 & -9.21 & -10.69 & -10.06 & -10.14 + ngc 4387 & 3.95 & 7.17 & -4.83 & 0.06 & 10.02 & -3.36 & -7.26 & -10.09 & -7.78 & -9.25 & -8.40 & -8.49 + ngc 4434 & 4.01 & 7.22 & -4.48 & -0.37 & 9.98 & -3.27 & -6.74 & -9.42 & -7.25 & -8.74 & -7.86 & -7.85 + ngc 4458 & 4.01 & 7.22 & -4.31 & -0.39 & 9.99 & -3.28 & -6.73 & -9.43 & -7.25 & -8.73 & -7.86 & -7.84 + ngc 4464 & 3.54 & 6.82 & -3.99 & -2.38 & 9.78 & -3.56 & -5.73 & -7.62 & -5.79 & -7.13 & -6.40 & -6.40 + ngc 4467 & 2.92 & 6.20 & -4.52 & -2.89 & 9.59 & -3.99 & -6.01 & -7.73 & -5.96 & -7.27 & -6.55 & -6.55 + ngc 4472 & 7.29 & 9.12 & -5.30 & 0.83 & 13.29 & -4.67 & -8.66 & -11.43 & -9.20 & -10.72 & -9.95 & -9.89 + ngc 4478 & 4.49 & 7.58 & -4.78 & 0.80 & 10.27 & -3.20 & -7.66 & -10.55 & -8.20 & -9.76 & -9.03 & -8.97 + ngc 4486 & 7.05 & 9.02 & -5.57 & 0.83 & 13.40 & -4.89 & -7.27 & -10.08 & -7.76 & -9.25 & -8.34 & -8.40 + ngc 4486b & 3.18 & 6.48 & -5.11 & -0.89 & 10.00 & -4.03 & -8.91 & -11.65 & -9.45 & -10.98 & -10.24 & -10.14 + ngc 4551 & 4.10 & 7.28 & -4.70 & 0.01 & 10.00 & -3.23 & -7.03 & -10.27 & -7.56 & -9.09 & -8.18 & -8.19 + ngc 4552 & 5.66 & 8.32 & -4.81 & 0.82 & 11.79 & -3.98 & -8.19 & -11.03 & -8.77 & -10.32 & -9.58 & -9.88 + ngc 4564 & 4.72 & 7.73 & -4.24 & 0.23 & 10.00 & -2.77 & -6.61 & -9.86 & -7.16 & -8.66 & -7.79 & -7.88 + ngc 4570 & 4.80 & 7.79 & -4.11 & -0.20 & 9.97 & -2.69 & -6.16 & -8.93 & -6.67 & -8.14 & -7.25 & -7.23 + ngc 4621 & 5.88 & 8.43 & -4.11 & 0.82 & 10.53 & -2.60 & -6.78 & -9.65 & -7.35 & -8.83 & -8.14 & -8.14 + ngc 4636 & 6.28 & 8.65 & -5.52 & 0.80 & 12.95 & -4.81 & -8.93 & -11.85 & -9.49 & -11.00 & -10.14 & -10.14 + ngc 4649 & 6.80 & 8.90 & -5.37 & 0.82 & 13.09 & -4.70 & -8.75 & -11.76 & -9.30 & -10.79 & -10.04 & -9.97 + ngc 4697 & 5.64 & 8.30 & -4.69 & 0.66 & 11.05 & -3.25 & -7.34 & -10.61 & -7.90 & -9.45 & -8.59 & -8.55 + ngc 4742 & 4.18 & 7.35 & -3.50 & -1.98 & 9.80 &
-3.05 & -5.35 & -7.34 & -5.48 & -6.85 & -6.11 & -6.10 + ngc 4874 & 8.57 & 9.64 & -6.07 & 1.61 & 14.50 & -5.37 & -9.92 & -12.73 & -10.51 & -12.06 & -11.29 & -11.59 + ngc 4889 & 8.33 & 9.54 & -5.77 & 1.60 & 14.13 & -5.09 & -9.67 & -12.45 & -10.24 & -11.78 & -11.07 & -11.30 + ngc 524 & 6.13 & 8.57 & -5.10 & 1.00 & 12.13 & -4.06 & -8.38 & -11.73 & -8.94 & -10.45 & -9.84 & -9.97 + ngc 5813 & 6.44 & 8.73 & -5.14 & 1.10 & 12.51 & -4.29 & -8.67 & -11.95 & -9.21 & -10.79 & -10.09 & -10.21 + ngc 5845 & 4.66 & 7.69 & -4.41 & 0.78 & 10.00 & -2.81 & -7.20 & -10.48 & -7.77 & -9.31 & -8.73 & -8.73 + ngc 596 & 5.52 & 8.24 & -4.61 & 0.95 & 10.80 & -3.07 & -7.46 & -10.73 & -8.02 & -9.57 & -8.97 & -8.97 + ngc 6166 & 8.46 & 9.60 & -6.24 & 1.69 & 14.68 & -5.59 & -10.26 & -13.55 & -10.81 & -12.39 & -11.79 & -11.79 + ngc 720 & 6.23 & 8.62 & -5.56 & 0.96 & 13.00 & -4.89 & -9.17 & -11.93 & -9.72 & -11.23 & -10.57 & -10.79 + ngc 7332 & 4.70 & 7.72 & -3.73 & -0.56 & 10.01 & -2.80 & -5.96 & -8.73 & -6.44 & -7.90 & -6.98 & -7.02 + ngc 7768 & 7.73 & 9.31 & -5.42 & 1.62 & 13.36 & -4.56 & -9.27 & -12.19 & -9.82 & -11.36 & -10.76 & -10.76 + for each galaxy we built simple spherical models using the same nuker law parameters as those in @xcite and separately calculated the loss cone filling rates due to two - body interaction in single and binary bh cases ( see mt99 and y02 for detailed description of the calculation ) . for single bhs we did two sets of calculations to investigate the effect gr on loss cone filling rate @xmath201 .
in the non - gr case we calculated @xmath41 with @xmath77 from equation ( [ rt ] ) while in the gr case we substituted @xmath77 with @xmath86 if @xmath202 . in both cases the resulting loss cone filling rates are similar , within a factor of @xmath203 even when @xmath204 .
this is because the loss cones of the most massive bhs are diffusive so that @xmath201 is insensitive to the size of loss cone . in table
[ table1 ] we only give @xmath201 for the non - gr case .
the mean of @xmath205 is @xmath206 for all the galaxies or @xmath207 for galaxies with @xmath208 , which is within a factor of @xmath209 of the rates obtained by @xcite , the differences being completely accounted for by our different assumed bh masses
. for binary bhs gr effect is not important to loss cone filling rate because we are interested in the case @xmath210 . for each galaxy
we have calculated the loss cone filling rates for binary bhs with a set of @xmath17 and derived the evolution timescale @xmath211 with @xmath212 ( y02 ) , where @xmath213 is the evolution timescale due to gravitational wave radiation @xcite .
then the binary s stalling radius @xmath214 is determined by locating the peak of the @xmath211 curve .
if @xmath215 yr , which is the case for most of our model galaxies , we increase @xmath214 until ( 1 )
@xmath216 yr or ( 2 ) @xmath217 arcsec , whichever comes first .
the latter criteria is motivated by that smbhbs in nearby galaxies , if wider than @xmath192 arcsec , would have been observed .
notice that @xmath214 , @xmath218 , and @xmath219 given by either criteria ( 1 ) or ( 2 ) are independent of @xmath68 .
the results for @xmath150 , the typical mass ratio in the literature for the present - day smbhbs , are given in table [ table1 ] , while the results for @xmath138 and @xmath220 are essentially identical .
although the loss cone of smbhb in @xmath221 space could be more than four orders of magnitude wider than that of a single bh ( eq.s [ [ jlc ] ] and [ [ rtah ] ] ) , the loss cone filling rate for binary bh increases only by a factor of @xmath222 , which implies that the loss cone of smbhb is also in the diffusive regime .
first , we ignore gr effect and only consider tidal disruption of solar type stars . in this case
the stellar disruption rates for single bhs have been given by column 4 in table [ table1 ] .
for binaries the rates are calculated according to equation ( [ nd ] ) with the loss cone filling rates listed in table [ table1 ] and the non - relativistic cross sections from [ nongrcs ] .
in figure [ case1 ] we present the stelar disruption rates for both single ( triangles ) and binary ( dots and circles ) bhs .
it is clear that the total disruption rates for smbhb ( @xmath223 ) is significantly lower than that for single bh ( @xmath224 ) .
the contrast between @xmath225 and @xmath224 is more than one order of magnitude and increases with @xmath68 .
the significant reduction of tidal disruption rates in binary bh systems originates in the depletion of loss - cone stars and the suppression of loss cone filling rate by the smbhb .
+ mechanisms other than two - body relaxation would increase the loss cone filling rates in both single and binary smbh systems ( mt99 ; y02 ; @xcite ) .
the enhancement in loss cone filling rate is uncertain because it depends on processes which are not well understood .
however , equation ( [ th ] ) implies that the lifetime of hard smbhb decreases with loss cone filling rate .
so for long - life smbhb with @xmath226 yr the loss cone filling rate should not significantly exceed @xmath227 .
correspondingly the stellar disruption rate in smbhb system is unlikely to be much greater than @xmath228\,.\label{ndcri}\end{aligned}\ ] ] these upper limits of stellar disruption rates for binary bhs have been presented in figure [ case1 ] as dashed lines . in @xmath229 out of @xmath198 cases @xmath225 is below @xmath230 , reflecting that two - body relaxation is inefficient in refilling the loss cone . for galaxies with @xmath231
@xmath230 is always lower than @xmath224 , so long - life smbhbs in these galaxies would manifest themselves by prominently suppressing stellar disruption rates no matter what mechanism is responsible for the loss cone replenishing . at @xmath232 @xmath230 could be higher than @xmath224 if @xmath233 or @xmath185 , implying that in these galaxies smbhbs are not so distinct in the efficiency of stellar disruption .
in the relativistic case there is a critical mass @xmath234 above which a bh would swallow the whole star without tidal disruption @xcite .
for solar type stars recent relativistic simulation @xcite gave that @xmath235 if the bh is non - spinning and @xmath236 if the bh is maximally spinning .
due to the ambiguity of bh spin it is not clear whether a star approaching a bh more massive than @xmath237 would necessarily be disrupted and produce a flare .
however , it is expected that stellar disruptions by these massive bhs are rare because only those stars approaching the spinning bh along a corotating orbit close to the equatorial plane could get disrupted . for less massive bhs ( @xmath238 ) although equation ( [ rt ] ) suggests that tidal disruption is inevitable ( @xmath239 ) , part of the stellar debris with low angular momentum may directly plunge into the event horizon without producing a flare @xcite .
the proportion of such debris is sensitive to the spin of the star just before tidal disruption and is uncertain so far , but the proportion should be low when @xmath240 .
because of all these uncertainties , we adopted a rough assumption that flares are produced only in the condition @xmath241 .
figure [ flare ] shows the stellar disruption rates when gr effect is taken into account .
the tidal disruption rates for binary bhs are calculated with the loss cone filling rates from table [ table1 ] and the relativistic tidal disruption cross sections presented in [ grcs ] .
the resulting relativistic tidal disruption rates are about @xmath242 times greater than the non - relativistic ones if @xmath243 , or @xmath244 times greater if @xmath245 .
the dashed lines show the thresholds @xmath230 above which smbhbs would coalesce within @xmath123 yr .
they are derived according to equation ( [ ndcri ] ) but with relativistic cross sections . to study the frequency of tidal flares
we have set the stellar disruption rates zero when @xmath246 , thus the sudden cutoff at @xmath247 and @xmath248 is artificial .
+ at @xmath249 the stellar disruption rates are similar to those presented in figure [ case1 ] and again we find that the stellar disruption rates for smbhbs are considerably lower than those for single bhs . while at @xmath233 the only sources producing tidal flares are the secondary bhs , though the rates are low .
the results of previous sections are based on the assumption that each galaxy is composed of stars with solar mass and radius .
since none of our sample galaxies exhibits recent nuclear star formation , taking into account a mass spectrum of main - sequence stars would result in numerous low mass stellar objects which have smaller mean tidal disruption cross sections and are less efficient to relax via two - body interaction .
but these effects would not qualitatively change the tidal disruption rates since : ( a ) the tidal radius is not sensitive to the stellar mass because @xmath250 @xcite so that @xmath251 ; ( b ) the increment in stellar number density to reproduce the mass distribution inferred from observations compensates for the decrease in the efficiency of two - body relaxation .
however , tidal disruption of off - main - sequence giant stars becomes important when bhs are more massive than @xmath237 because in this case the tidal radius is greater than the schwarzschild radius .
following mt99 we assume that giant stars with time - averaged radius @xmath252 contribute @xmath253 to the total stellar population .
these values are consistent with the approximate stellar evolution model given in @xcite . for single bhs
the disruption rates of giants in the diffusive loss cone limit are obtained with @xmath254 ( mt99 ) , where the loss cone filling rates @xmath201 in relativistic case are from table [ table1 ] . for binary bhs
the disruption rates are calculated in the same way as in [ grdr ] except that the loss cone filling rates and tidal radii for giant stars are used .
since we are interested the frequency of tidal flares we have set the stellar disruption rates zero when @xmath246 .
+ figure [ giant ] shows the stellar disruption rates and the thresholds @xmath230 when both solar type and giant stars are taken into account . at @xmath231
the stellar disruption rates are dominated by disruption of solar type stars therefore similar to those in figure [ flare ] . at @xmath255 disruption of giant stars
take over in both single and binary bh systems , resulting in low but no longer zero flaring rates .
the suppression of stellar disruption rate by smbhb is still obvious if the loss cone refilling is dominated by two - body relaxation , but other mechanisms could potentially enhance the stellar disruption rates for massive smbhbs ( @xmath255 ) to a level indistinguishable from those for single bhs .
to estimate the density of tidal flares in local universe we adopted the mass functions of smbh for early ( e+s0 ) and late ( sabcd ) type galaxies from @xcite , which are converted from the distribution of stellar velocity dispersion using the empirical @xmath256 relation @xcite .
the mass functions are presented in figure [ mf ] and in the following we assume that the total bh mass ( @xmath23 ) in smbhb systems follows the same distribution . at @xmath257 the mass distribution of bhs is unclear because the demographics of the bhs in dwarf galaxies is not well established . the total bh mass density according to figure [ mf ]
is @xmath258 ( @xmath259 throughout this paper ) , consistent with the value given by @xcite .
+ the flaring rates per unit volume are calculated by integrating the bh mass functions weighted by the flaring rates .
the flaring rates for single and binary bhs are from figures [ flare ] ( we have smoothed the flaring rates ) and the fiducial field of integration is @xmath260 $ ] . in single bh case
we do a second set of calculations with @xmath261 $ ] to account for the flaring rates from the intermediate - mass bhs ( imbhs , @xmath262 ) at the centers of dwarf galaxies , while for binary bhs we skip such calculations because most dwarf galaxies have not experienced mergers @xcite .
the flaring rates at @xmath263 are from the linear extrapolation of the logarithmic flaring rate at @xmath264 .
we do not consider flares produced by giants because we will see below that these flares differ from those produced by normal stars .
our result is not sensitive to this treatment because the mass function of smbhs cuts off steeply above @xmath237 where flares from giants become dominant .
llllll early & @xmath265 & @xmath265 & 0.01 & @xmath266 & @xmath267 + early & @xmath265 & @xmath265 & 0.1 & @xmath268 & @xmath269 + early & @xmath265 & @xmath265 & 1 & @xmath270 & @xmath270 + late & @xmath271 & @xmath272 & 0.01 & @xmath273 & @xmath274 + late & @xmath271 & @xmath272 & 0.1 & @xmath275 & @xmath276 + late & @xmath271 & @xmath272 & 1 & @xmath277 & @xmath277 + the flaring rates per unit volume for different types of galaxies and for both single and binary bhs are presented in table [ table2 ] .
note that taking into account asymmetric stellar distributions and relaxation mechanisms other than two - body interaction would significantly increase the loss cone filling rates in both single and binary bh systems , therefore dramatically increase the flaring rates .
according to table [ table2 ] , it is obvious that if smbhbs are ubiquitous in the centers of both early and late type galaxies the flaring rate in local universe would be more than one order of magnitude lower than that in single bh case .
it is also clear that dwarf galaxies , if they follow the same mass distribution as that of massive galaxies and harbor imbhs at their centers , would contribute the majority to the flares in late type galaxies but have little effect on the flaring rate in early type galaxies . + if a fraction , @xmath278 , of smbhs in the investigated galaxies are in binary and a fraction @xmath279 are single , the expected tidal disruption rate is @xmath280 @xmath281 as a function of @xmath278 is presented in figure [ constrain ] .
it shows that @xmath281 is not sensitive to @xmath68 unless @xmath282 . in [ discussion ] we will use these @xmath283 diagrams to constrain @xmath284 for different types of galaxies .
the hierarchical structure formation model in cdm cosmology predicts that smbhbs are continuously formed across the merging history of galaxies . constraining the abundance of smbhbs among inactive galaxies is essential to test the theoretical models of the dynamic evolution of binary bhs . in this paper
we have studied the effect of hard binary bhs on stellar disruption rates , trying to find distinct observational signatures of smbhbs in galaxy centers .
we focus our attention on inactive galaxies which contain the final products of bh mergers so we do not consider the effect of gas on smbhb evolution .
we have carried out numerical scattering experiments designed for hard smbhbs to investigate the probability of stellar disruption . since @xmath285 , in each set of experiments usually a large number of test particles ( @xmath139 ) were used to give statistical meaningful results . for solar type stars
the test particle approximation should be valid outside @xmath77 as long as @xmath286 .
in this situation the self - binding energy of a star @xmath287 is smaller by a factor @xmath288 than its kinetic energy at @xmath77 so the loss of orbital energy due to tidal dissipation is not important .
the test particle approximation may be problematic for giant stars because they are much less concentrated and may suffer from mass stripping during their encounters with bhs .
this uncertainty could be incorporated into the uncertainty in @xmath289 . in single bh case
the stellar disruption rate is not sensitive to @xmath289 if the loss cone is in the diffusive regime but would be proportional to @xmath289 if the loss cone is in the pinhole regime ( mt99 ) . for binary bhs the loss cone filling rate does not depend on @xmath289 but the tidal disruption cross section scales with @xmath289 when gravitational focusing is important . therefore unless the effective @xmath289 deviates significantly from our choice
@xmath252 our results for giant stars would not be qualitatively different .
we find that in binary bh systems the multi - encounter cross sections are greater than the single - encounter ones .
this is because binary bhs tend to scatter particles onto temporary bound orbits which enhances the number of star - binary encounters .
if a star is only partly disrupted during each close passage from the bhs , multi - flares would be produced occur due to the multi - encounters , which provides a distinct observational signature of smbhbs .
however , we have seen in [ nongrcs ] that the probability of multi - flaring may be extremely low because @xmath290 . even if @xmath291 equation ( [ rsrt ] ) implies that @xmath292 so gravitational wave radiation would quickly drive the bhs to coalesce , making the multi - flares unlikely .
a secondary flare could also be produced if the secondary member of smbhb captures sufficient debris from the star disrupted by the primary bh , or vise versa .
the probability of these events depends on the opening angle of all the spewed debris about the stellar disrupting bh , which needs to be determined by hydrodynamic simulations .
we have shown that the normalized cross sections are not sensitive to the star s initial binding energy or to the eccentricity of the bh binary , but do depend on the amplitude and orientation of the star s initial angular momentum .
the cross sections presented in figures [ cs001 ] and [ fit ] , which have been equally averaged over the whole loss cone and over all directions , can be directly used if the loss cone of smbhb is isotropic and is in the pinhole regime . in the diffusive loss cone limit the stars falling to the binary bhs satisfy @xmath293 . according to figure [ dfjdr ] the differential tidal disruption cross sections at @xmath293 are lower than those inside the loss cone , so the mean tidal disruption cross section in the diffusive limit is smaller than those presented in figures [ cs001 ] and [ fit ] .
we find that the tidal disruption rates in binary bh systems are more than one order of magnitude lower than those in single bh systems .
the suppression of stellar disruption rate by smbhb originates in the low loss cone filling rate in binary bh systems , which results from the stellar cusp destruction due to the hardening of smbhbs .
even efficient loss cone filling mechanisms are taken into account the contrast between the stellar disruption rates in single and binary bh cases are still prominent in less massive galaxies with @xmath231 , because the survival of smbhbs requires that the loss cone filling rate should not exceed a critical value which is proportional to the total bh mass of smbhb .
our result is qualitatively different from that of @xcite , who suggested that a smbhb will enhance the stellar disruption rates .
the difference is due to that we are studying a much later evolution stage at which the dense galactic cusps considered by @xcite have been destroyed during the hardening of smbhbs .
according to the calculation in y02 binary bhs reach the hard radius in about a dynamic friction timescale of the merging galaxy .
so the situation considered by @xcite applies for a short period relative to the lifetime of a galaxy .
@xcite suggested that because loss cone refilling takes time the stellar disruption rate immediately after smbhb coalescence is significantly below the final steady disruption rate . but
this effect is prominent only in massive galaxies with @xmath255 , where the relaxation timescale is long .
our result implies that in less massive galaxies the flaring rate could give stringent constraint to the abundance of smbhb . in larger galaxies with @xmath255 tidal flares
are dominated by disruption of giants .
however , their spectral and variable properties may be different from those of solar type stars .
a tidal flare is expected to be produced at a scale about @xmath77 , initially radiating at the eddington limit @xmath294 with a thermal spectrum of effective temperature @xmath295^{1/4}=2.4\times10 ^ 5(r_*/r_\odot)^{-1/2}(m_*/m_{\odot})^{1/6}m_8^{1/12}~{\rm k}$ ] and decaying on a timescale @xmath296 , where @xmath297 characterizes the span of the binding energy of the debris @xcite . for solar type
stars the spectrum of tidal flare peaks at uv or soft x - ray and the luminosity decays in about @xmath298 year . while for giant stars with @xmath299 and @xmath300 the spectrum peaks at optical or uv band and the flare dims on a timescale of about @xmath301 years
therefore if a uv / x - ray outburst decaying within months to one year is detected in a galaxy with total bh mass much greater than @xmath237 ( determined by other observations ) , it is likely that a less massive secondary smbh resides in the galactic center .
such event is another signature of smbhb , though figure [ giant ] suggests that the frequency of such event is low .
so far six candidate tidal disruption events have been observed in nearby inactive galaxies . among them , one was observed at @xmath302 by the uv telescope _ galex _ @xcite and
the rest were discovered in galaxies at @xmath303 during the one year rosat all sky survey ( _ rass _ ) @xcite .
the hubble types for the host galaxies of the five _ rass _ flares are not well determined : ngc 5905 is of a sb galaxy and the other four galaxies look like ellipticals @xcite . from the observations of _ rass _ @xcite inferred a total flaring rate of @xmath304 for all types of galaxies .
from the _ rass _ results we can also give a rough estimate of the flaring rates for different types of galaxies : @xmath305 for early type galaxies and @xmath306 for late types .
we plot @xmath307 as long - dashed horizontal lines in figure [ constrain ] and the corresponding uncertainties with shaded areas . because of the poor statistics in @xmath307 and the ill - determination of hubble types of the host galaxies , the results are highly uncertain .
we did the calculation based on the assumptions that galaxies are spherical and two - body scattering relaxation dominates the loss - cone refilling .
however , taking into account more realistic non - spherical stellar distributions and other loss - cone refilling mechanisms besides two - body interaction will probably significantly increase the loss - cone refilling rates in both single and binary smbh systems . mt99 and y02 showed that taking into account axisymmetric stellar distribution will increase the loss cone refilling rates by a factor of @xmath209 for both single smbhs and hard smbhbs .
the loss cone refilling rates in triaxial galaxies are slightly higher than those in axisymmetric ones if the triaxiality is less than @xmath192 but can be two orders of magnitude higher if the triaxiality is more extreme ( y02 ; * ? ? ?
resonant relaxation would increase the loss cone refilling rates by a factor of @xmath308 relative to those in the case of two - body relaxation , depending on the eccentricities of stellar orbits @xcite . to calculate the tidal disruption rates by taking into account the above physical processes
is out of the scope of this paper . to illustrate the effects of increased loss - cone refilling rates and tidal disruption rates on the estimated fraction of binary bhs , @xmath284
, we simply multiply our calculated stellar disruption rates @xmath281 by @xmath209 , @xmath222 , and @xmath309 and plot the results in figure [ constrain ] . in figure
[ constrain ] , the intersections of long - dashed and solid lines give the lower limit to @xmath278 , while the intersections of long - dashed and dot - dashed lines roughly give the upper limits .
when all types of galaxies are considered , binary fraction @xmath278 is between @xmath310 and @xmath311 . for early type galaxies
, the solid lines do not intersect the long - dashed line , implying that other loss - cone filling mechanisms in addition to two - body relaxation should also be important . for late type galaxies ,
figure [ constrain ] suggests a binary fraction @xmath312 and that extreme triaxiality may not be common , otherwise the current observation would imply an extremely high fraction of smbhbs .
we would like to note that the constraints on @xmath278 are very uncertain due to the large uncertainty in the current @xmath307 .
detection of more candidate tidal disruption events is needed to give better constraints on @xmath278 and on the dominant loss - cone filling mechanisms in different types of galaxies .
the non - detection of x - ray or uv flares in dwarf galaxies seems inconsistent with the calculated high flaring rates .
the discrepancy may be due to ( 1 ) a lack of sufficient sensitivity in current surveys , ( 2 ) the continuous and steady accretion of stellar debris onto imbhs during successive tidal disruption events @xcite , or ( 3 ) a lack of imbhs in the centers of dwarf galaxies .
future x - ray and uv surveys with improved sensitivity and angular resolution would help constraining the fraction of imbhs in galaxy centers and give interesting implications to the formation and the merging history of imbhs @xcite .
to investigate whether smbhbs are ubiquitous in nearby inactive galaxies or coalesce rapidly in galaxy mergers , we have studied the tidal disruption rates of stellar objects in both single and binary smbh systems .
we have calculated the interaction cross sections between hard smbhbs and intruding stars by carrying out intensive numerical scattering experiments with typically @xmath194 particles , taking into account the initial binding energy and angular momentum of particle , the eccentricity of the orbit of smbhb , and including general relativistic effects .
we have also derived empirical formulae for the relativistic cross sections , which can be applied to smbhbs with a wide range of semimajor axis and todal bh mass .
we have calculated the rate of loss cone refilling due to two - body relaxation for a sample of @xmath198 nearby galaxies , assuming that each galaxy is spherical . the steady loss cone filling rates in binary bh systems
would be significantly suppressed due to the three - body interaction between smbhbs and stars passing by .
we have calculated the tidal disruption rates respectively for single and binary smbhs by combining the loss cone filling rates and the tidal disruption cross sections .
we find that the tidal disruption rate in smbhb systems is more than one order of magnitude lower than that in single smbh systems . for galaxies with bhs
more massive than @xmath237 , a uv / x - ray flare at galactic center decaying within one year provides strong evidence of a secondary bh , although the probability of such events is low .
finally we have calculated the flaring rates in local universe using the bh mass function given in the literature .
the comparison of the calculated flaring rates and the preliminary results from current x - ray surveys could not yet tell whether smbhbs are ubiquitous in local universe .
future uv / x - ray surveys with improved sensitivity and duration are needed .
we are grateful to dr .
xue - bing wu , dr .
stefanie komossa , and dr .
vladimir karas for many constructive comments .
we thank the referee for helpful comments and suggestions .
many thanks are due to bing - xiao xu , ran wang , lei qian and zhao - yu li for fruitful discussions . during this work we have used
the sgi altix 330 system at the department of astronomy , peking university ( pku ) and ccse - i hp cluster of pku .
this work is supported by the national natural science foundation of china ( no . 10573001 ) and by the national 973 program ( no .
2007cb815405 ) .
jm thanks the royal society for financial support . | supermassive black hole binaries ( smbhbs ) are expected by the hierarchical galaxy formation model in @xmath0cdm cosmology .
there is some evidence in the literature for smbhbs in agns , but there are few observational constraints on the evolution of smbhbs in inactive galaxies and gas - poor mergers . on the theoretical front , it is unclear how long is needed for a smbhb in a typical galaxy to coalesce . in this paper
we investigate the tidal interaction between stars and binary bhs and calculate the tidal disruption rates of stellar objects by the bh components of binary .
we derive the interaction cross sections between smbhbs and stars from intensive numerical scattering experiments with particle number @xmath1 and calculate the tidal disruption rates by both single and binary bhs for a sample of realistic galaxy models , taking into account the general relativistic effect and the loss cone refilling because of two - body interaction .
we estimate the frequency of tidal flares for different types of galaxies using the bh mass function in the literature . we find that because of the three - body slingshot effect , the tidal disruption rate in smbhb system is more than one order of magnitude smaller than that in single smbh system .
the difference is more significant in less massive galaxies and does not depend on detailed stellar dynamical processes .
our calculations suggest that comparisons of the calculated tidal disruption rates for both single and binary bhs and the surveys of x - ray or uv flares at galactic centers could tell us whether most smbhs in nearby galaxies are single and whether the smbhbs formed in gas - poor galaxy mergers coalesce rapidly . |
the first of the above issues has been addressed very often in the past and we will restrict ourselves here to a brief summary . in the lf framework ,
non - trivial vacuum structure can reside only in zero - modes ( @xmath0 modes ) . since these are high - energy modes ( actually infinite energy in the continuum ) one often does not include them as explicit degrees of freedom but assumes they have been integrated out , leaving behind an effective lf - hamiltonian .
the important points here are the following .
if the zero - mode sector involves spontaneous symmetry breaking , this manifests itself as explicit symmetry breaking for the effective hamiltonian . in general
, these effective lf hamiltonians thus have a much richer operator structure than the canonical hamiltonian .
therefore , compared to a conventional hamiltonian framework , the question of the vacuum has been shifted from the states to the operators and it should thus be clear that the issues of renormalization and the vacuum are deeply entangled in the lf framework .
despite widespread confusion on this subject , vector current conservation ( vcc ) is actually not a problem in the lf framework .
many researchers avoid the subject of current conservation because the divergence of the vector current q_j^(q ) = q^-j^+ + q^+j^- - q__involves @xmath1 , which is quadratic in the constrained fermion spinor component j^-= ^_(-)_(- ) , and thus @xmath1 contains quartic interactions making it at least as difficult to renormalize as the hamiltonian .
we know already from renormalizing @xmath2 that the canonical relation between @xmath3 and @xmath4 is in general not preserved in composite operators ( such as @xmath5 ) .
it is therefore clear that a canonical definition for @xmath1 will in general violate current conservation since it does not take into account integrating out zero - modes and other high - energy degrees of freedom .
however , in lf gauge @xmath6 , @xmath1 does not enter the hamiltonian and therefore one can address its definition separately from the construction of the hamiltonian .
in fact , it is very easy to find a pragmatic definition which obviously guarantees manifest current conservation , namely j^-(q^+ ) & = & - ( 1 + 1 ) [ eq : j - def ] + j^-(q^+,q _ ) & = & - \ { - q__(q^+,q _ ) } ( 3 + 1)in 1 + 1 and 3 + 1 dimensions respectively . the corresponding expressions in coordinate space @xmath7 can be obtained by fourier transform . in summary ,
most importantly , vcc is no problem in lf quantization and is manifest at the operator level , provided @xmath1 is _ defined _ using eq .
( [ eq : j - def ] ) . since vcc can easily be made manifest ( by using the above construction ! ) , there is no point in _ testing _
its validity and it can not be used as a non - perturbative renormalization condition either . for a non - interacting theory , eq .
( [ eq : j - def ] ) reduces to the canonical definition of @xmath1 , but in general this is not the case when @xmath2 contains interactions or even non - canonical terms . note that ( as so often on the lf ) @xmath8 appears in the denominator of eq .
( [ eq : j - def ] ) . therefore , as usual
, one should be very careful while taking the @xmath9 limit and while drawing any conclusions about this limit .
an example of this kind are the pair creation terms in @xmath1 .
naively they do not contribute for @xmath9 , since the @xmath10 pair emanating from @xmath1 necessarily carries positive @xmath8 .
however , since @xmath1 often has very singular matrix elements for @xmath8 , such seemingly vanishing terms nevertheless survive the @xmath9 limit , which often leads to confusion . for an early example of this kind see ref .
@xcite . a more recent discussion can be found in ref .
a parity transformation , @xmath11 , @xmath12 leaves the quantization hyperplane ( @xmath13 ) in equal time ( et ) quantization invariant and therefore the parity operator is a kinematic operator in such a framework .
it is thus very easy to ensure that parity is a manifest symmetry in et quantization by tracking parity at each step in a calculation .
the situation is completely different in the lf framework , where the same parity transformation exchanges lf-`time ' ( @xmath14 ) and space ( @xmath15 ) directions , i.e. x^+ x^- and therefore the quantization hyperplane @xmath16 is _ not _ invariant .
hence , the parity operator is a dynamical operator on the lf and , except for a free field theory , it is probably impossible to write down a simple expression for it in terms of quark and gluon field operators .
thus , parity invariance is not a manifest symmetry in this framework .
note that the situation is the other way round for the boost operator ( kinematic and manifest on the lf , dynamical and non - manifest in et ) .
it thus depends on the physics application one is interested in and the symmetries that one considers the most important ones for that particular physics problem , which framework is preferable .
for most applications of lf quantization , the lack of manifest parity actually does not constitute a problem in fact , one can view it as an opportunity rather than a problem .
the important point here is the following : due to the lack of manifest covariance in a hamiltonian formulation , lf hamiltonians in general contain more parameters than the corresponding lagrangian .
parity invariance may be very sensitive to some of these parameters . in order to illustrate this important point ,
let us consider the example of a 1 + 1 dimensional yukawa model @xcite @xmath17 this model actually has a lot in common with the kind of interactions that appear when one formulates qcd ( with fermions ) on a transverse lattice @xcite .
the main difference between scalar and dirac fields in the lf formulation is that not all components of the dirac field are dynamical : multiplying the dirac equation @xmath18 by @xmath19 yields a constraint equation ( i.e. an `` equation of motion '' without a time derivative ) @xmath20 where @xmath21 for the quantization procedure , it is convenient to eliminate @xmath3 , using _ ( - ) = ( m_f+g_5 ) _ ( + ) from the classical lagrangian before imposing quantization conditions , yielding @xmath22 the rest of the quantization procedure very much resembles the procedure for self - interacting scalar fields .
the above canonical hamiltonian contains a kinetic term for the fermions , a fermion boson vertex and a fermion 2-boson vertex .
while the couplings of these three terms in the canonical hamiltonian depend only on two independent parameters ( @xmath23 and @xmath24 ) , it turns out that these terms are renormalized independently from each other once zero - mode and other high - energy degrees of freedom are integrated out .
more explicitly this means that one should make an ansatz for the renormalized lf hamiltonian density of the form @xmath25 where the @xmath26 do not necessarily satisfy the canonical relation @xmath27 .
however , this does not mean that the @xmath26 are completely independent from each other .
in fact , eq.([eq : pren ] ) will describe the yukawa model only for specific combinations of @xmath26 .
it is only that we do not know the relation between the @xmath26 .
thus the bad new is that the number of parameters in the lf hamiltonian has increased by one ( compared to the lagrangian ) .
the good news is that a wrong combination of @xmath26 will in general give rise to a parity violating theory : formally this can be seen in the weak coupling limit , where the correct relation ( @xmath27 ) follows from a covariant lagrangian .
any deviation from this relation can be described on the level of the lagrangian ( for free massive fields , equivalence between lf and covariant formulation is not an issue ) by addition of a term of the form @xmath28 , which is obviously parity violating , since parity transformations result in @xmath29 for lorentz vectors @xmath30 ; i.e. @xmath31 .
this also affects physical observables , as can be seen by considering boson fermion scattering in the weak coupling limit of the yukawa model . at the tree level
, there is an instantaneous contact interaction , which is proportional to @xmath32 .
the ( unphysical ) singularity at @xmath33 is canceled by a term with fermion intermediate states , which contributes ( near the pole ) with an amplitude @xmath34 .
obviously , the singularity cancels iff @xmath35 .
since the singularity involves the lf component @xmath8 , this singular piece obviously changes under parity .
this result is consistent with the fact that there is no zero - mode induced renormalization of @xmath36 and thus @xmath35 at the tree level .
this simple example clearly demonstrates that a ` false ' combination of @xmath37 and @xmath36 leads to violations of parity for a physical observable , which is why imposing parity invariance as a renormalization condition may help reduce the dimensionality of coupling constant space .
of course there are an infinite number of parity sensitive observables , but not all of them are easy to evaluate non - perturbatively in the hamiltonian lf framework .
furthermore , we will see below that some relations among observables , which seem to be sensitive to parity violations , are actually ` protected ' by manifest symmetries , such as charge conjugation , or by vcc . from the brief discussion of the ( to - be - canceled ) singularity above it seems that the most sensitive observable to look for violations of parity in qcd would be compton scattering cross sections between quarks and gluons ( or the corresponding fermions and bosons in other field theories ) because there one could tune the external momenta such that the potential singularity enters with maximum strength .
however , this is not a very good choice : first of all non - perturbative scattering amplitudes are somewhat complicated to construct on the lf .
secondly , quarks and gluons are confined particles , which makes @xmath38 compton scattering an unphysical process . a much better choice are any matrix elements between bound states .
bound states are non - perturbative and all possible momentum transfers occur in their time evolution
. therefore , any parity violating sub - amplitude would contribute at some point and would therefore affect physical observables .
secondly , since one of the primary goal of lfqcd is to explore the non - perturbative spectra and structure of hadrons , matrix elements in bound states are the kind of observables for which the whole framework has been tailored .
one conceivable set of matrix elements are those of the vector current operator @xmath39 .
for example , consider the vacuum to meson matrix elements 0 | j^+ |n , p&= & p^+ f_n^(+ ) + 0 | j^- |n , p&= & p^- f_n^(- ) .
[ eq : fpm ] by boost invariance , the couplings defined in eq .
( [ eq : fpm ] ) must be independent of the momenta .
obviously , parity invariance requires @xmath40 , which makes it a useless relation for the purpose of parity tests .
similar statements can be made about elastic formfactors but we will omit the details here .
the basic upshot is that the same relations between the matrix elements of @xmath39 that arise from parity invariance can often also be derived using only vcc , i.e. such relations are in general ` protected ' by vcc one may also consider non - conserved currents , such as the axial vector current .
however , there one would have to face the issue of defining the ` minus ' component before one can test any parity relations .
parity constraints are probably very helpful in this case when constructing the minus components , but then those relations can no longer be used to help constrain the coefficients in the lf - hamiltonian .
another class of potentially useful operators consists of the scalar and pseudoscalar densities @xmath5 and @xmath41 . obviously ,
if parity is conserved then at most one of the two couplings f_s&= & 0 | || n + f_p&= & 0 | |_5| ncan be nonzero at the same time , since a state @xmath42 can not be both scalar and pseudoscalar .
what restricts the usefulness of this criterion is the fact that the same ` selection rule ' also follows from charge conjugation invariance ( @xmath43 and @xmath41 have opposite charge parity ! ) which is a manifest symmetry on the lf .
therefore , only for theories with two or more flavors , where one can consider operators such as @xmath44 and @xmath45 , does one obtain parity constraints that are not protected by charge conjugation invariance .
but even then one may have to face the issue of how to define these operators .
parity may of course be used in this process , but then it has again been ` used up ' and one can no longer use these selection rules to test the hamiltonian . in summary , many parity relations are probably very useful to determine the lf representation of the operators involved , but may not be very useful to determine the hamiltonian .
we have seen above that relations between @xmath46 and @xmath1 are often protected by vector current conservation and are therefore not very useful as parity tests .
a much more useful parity test can actually be obtained by considering the ` plus'-component only : let us now consider a vector form factor , @xmath47 where @xmath48 , between states of opposite parity . when writing the r.h.s . in terms of one invariant form factor
, use was made of both vector current conservation and parity invariance .
a term proportional to @xmath49 would also satisfy current conservation , but has the wrong parity .
a term proportional to @xmath50 has the right parity , but is not conserved and a term proportional to @xmath51 is both not conserved and violates parity . other vectors do not exist for this example . the lorentz structure in eq .
( [ eq : form ] ) has nontrivial implications even if we consider only the `` plus '' component , yielding @xmath52 1.cm ( 10,12)(0,1 . )
that this equation implies nontrivial constraints can be seen as follows : as a function of the longitudinal momentum transfer fraction @xmath53 , the invariant momentum transfer reads ( @xmath54 and @xmath55 are the invariant masses of the in and outgoing meson ) @xmath56 this quadratic equation has in general , for a given value of @xmath57 , two solutions for @xmath58 which physically correspond to hitting the meson from the left and right respectively .
the important point is that it is not manifestly true that these two values of @xmath58 in eq .
( [ eq : formplus ] ) will give the same value for the form factor , which makes this an excellent parity test . in ref .
@xcite , the coupling as well as the physical masses of both the fermion and the lightest boson where kept fixed , while the `` vertex mass '' was tuned ( note that this required re - adjusting the bare kinetic masses ) . figure [ fig : parity ] shows a typical example , where the calculation of the form factor was repeated for three values of the dlcq parameter k ( 24 , 32 and 40 ) in order to make sure that numerical approximations did not introduce parity violating artifacts . for a given physical mass and boson - fermion coupling ,
there exists a `` magic value '' of the vertex mass and only for this value one finds that the parity condition ( [ eq : formplus ] ) is satisfied over the whole range of @xmath57 considered .
this provides a strong self - consistency check , since there is only one free parameter , but the parity condition is not just one condition but a condition for every single value of @xmath57 ( i.e. an infinite number of conditions ) . in other words , keeping the vertex mass independent from the kinetic mass is not only necessary , but also seems sufficient in order to properly renormalize yukawa@xmath59 . in the above calculation
, it was sufficient to work with a vertex mass that was just a constant .
however , depending on the interactions and the cutoffs employed , it ma be necessary to introduce counter - term functions@xcite .
but even in cases where counter - term functions need to be introduced , parity constraints may be very helpful in determining the coefficient - functions for those more complex effective lf hamiltonians non - perturbatively .
99 m. burkardt , _ advances nucl .
phys . _ * 23 * , 1 ( 1996 ) .
w. a.bardeen et al .
, , 1037 ( 1980 ) .
s. dalley and b. van de sande , , 1088 ( 1999 ) ; , 065008 ( 1999 ) .
m.burkardt , , 762 ( 1989 ) ; , 613 ( 1992 ) .
s.j . brodsky and d.s .
hwang , , 239 ( 1998 ) ; h .-
choi and c.r .
ji , , 071901 ( 1998 ) .
m. burkardt , , 2913 ( 1996 ) .
m. burkardt and h. el - khozondar , , 054504 ( 1999 ) .
perry , lectures given at nato advanced study institute on confinement , duality and nonperturbative aspects of qcd , cambridge , u.k . , 1997 , hep - th/9710175 . | issues that are specific for formulating fermions in light - cone quantization are discussed .
special emphasis is put on the use of parity invariance in the non - perturbative renormalization of light - cone hamiltonians .
# 1#2#3#4#1 * # 2 * , # 3 ( # 4 ) light - front ( lf ) quantization is the most physical approach to calculating parton distributions on the basis of qcd @xcite . before one can formulate qcd with quarks ,
it is necessary that one understands how to describe fermions in this framework .
this in turn requires that one addresses the following issues how is spontaneous symmetry breaking ( chiral symmetry ! ) manifested in the lf framework , where the vacuum appears to be trivial ?
is it possible to preserve current conservation and parity invariance in this framework ?
how does one formulate fermions on the transverse lattice , which seems to be a very promising approaches to pure glue lfqcd @xcite ? |
this work was sponsored by the oak ridge national laboratory , managed by ut - battelle , llc , for the u.s.department of energy under contract de - ac05 - 00or22725 . | a novel approach that enables the study of parallel transport in magnetized plasmas is presented .
the method applies to general magnetic fields with local or nonlocal parallel closures .
temperature flattening in magnetic islands is accurately computed . for a wave number @xmath0 , the fattening time scales as @xmath1 where @xmath2 is the parallel diffusivity , and @xmath3 ( @xmath4 ) for non - local ( local ) transport .
the fractal structure of the devil staircase temperature radial profile in weakly chaotic fields is resolved . in fully chaotic fields ,
the temperature exhibits self - similar evolution of the form @xmath5 $ ] , where @xmath6 is a radial coordinate . in the local case
, @xmath7 is gaussian and the scaling is sub - diffusive , @xmath8 . in the non - local case , @xmath7 decays algebraically , @xmath9 , and the scaling is diffusive , @xmath10 .
the study of transport in magnetized plasmas is a problem of fundamental interest in controlled fusion , space plasmas , and astrophysics research .
three issues make this problem particularly challenging : ( i ) the _ extreme anisotropy _ between the parallel ( i.e. , along the magnetic field ) , @xmath11 , and the perpendicular , @xmath12 , conductivities ( @xmath13 may exceed @xmath14 in fusion plasmas ) ; ( ii ) magnetic _ field lines chaos _ which in general complicates ( and may preclude ) the construction of magnetic field line coordinates ; and ( iii ) _ nonlocal parallel transport _ in the limit of small collisionality . as a result of these challenges ,
standard finite difference and finite elements numerical methods suffer from a number of ailments .
chief among them are the pollution of perpendicular dynamics due to truncation errors in the discrete representation of the parallel heat flux , and the lack of a discrete maximum principle ( to enforce temperature positivity ) . despite the severity of these issues
, recent studies have succeeded in partially addressing some of them , and important progress has been made in the study of parallel transport .
reference @xcite discussed a finite element numerical implementation of nonlocal heat transport with applications including temperature flattening across magnetic islands and tokamak disruptions .
the use of high - order discretizations has been shown to mitigate numerical pollution of the perpendicular dynamics in finite - difference @xcite and finite - element methods @xcite .
a maximum principle has been shown to be enforceable by the use of limiters at the discrete level in finite differences @xcite , and finite elements @xcite , albeit with the effect of rendering both spatial discretizations formally first - order accurate . in ref .
@xcite a second order finite - difference iterative krylov method was used to find the steady state solution of the heat transport equation in a weakly chaotic magnetic field . motivated by the strong anisotropy typically encountered in magnetized plasmas ( @xmath15 ) , we study parallel heat transport in the extreme anisotropic regime ( @xmath16 ) .
in addition to the previously mentioned numerical difficulties , this regime presents potentially insurmountable issues regarding the algorithmic inversion of the discretized transport equation .
in particular , the potentially degenerate null space of the parallel transport operator might render the use of state - of - the - art scalable iterative inversion methods ( e.g. multigrid methods ) impractical . to overcome these numerical challenges , we present a novel lagrangian green s function approach .
the proposed method bypasses the need to discretize and invert the transport operators on a grid and allows the integration of the parallel transport equation without perpendicular pollution , while preserving the positivity of the temperature field .
the method is applicable to local and non - local transport in integrable , weakly chaotic , and fully chaotic magnetic fields . beyond the integration method ,
this letter presents novel physics results .
the difference between local and non - local parallel temperature mixing and flattening inside magnetic islands is studied .
of particular interest is the dependence of the temperature relaxation time on the wave number of the temperature perturbation .
understanding this problem is of significant interest to assess the impact of magnetic islands on magnetic confinement .
parallel transport is also studied in the case of weakly chaotic and fully chaotic @xmath17-d magnetic fields . in the case of weakly chaotic fields ,
the fractal hamiltonian structure of the magnetic islands implies that , as the ratio @xmath18 increases , the radial temperature profile approaches a devil - staircase .
going beyond previous studies @xcite , we unveil the fractal structure of the devil - staircase in the previously unaccessible @xmath19 regime .
these results open the possibility of a deeper understanding of the role of cantori which have been observed to act as partial transport barriers in numerical studies @xcite and experiments @xcite .
another problem of considerable interest in fusion and astrophysical plasmas is the understanding of electron heat transport in fully chaotic fields .
the case of local parallel transport has been extensively studied since the pioneering work in refs .
@xcite .
however , most studies have restricted their attention to local transport . here , in addition of the study of the local transport , we present novel results on the self - similar , non - gaussian spatio - temporal scaling of radial heat transport in the non - local case .
our starting point is the heat transport equation in a constant density plasma [ eq_1 ] _ t t = - , where @xmath20 is the heat flux . for local transport , in the limit @xmath21 , [ eq_2 ] * q*= - _ , where @xmath22 .
substituting ( [ eq_2 ] ) into ( [ eq_1 ] ) , we get [ eq_3 ] _ t t = -_s q _ , q_=-__s t , where @xmath23 is the derivative along the field line , and we have assumed the tokamak ordering @xmath24 . in the case of non - local transport , following ref .
@xcite , we consider [ eq_5 ] q_= _ 0^ dz , with @xmath25 . the case @xmath26 reduces to the collisionless heat flux characterized by the free - streaming of electrons along magnetic field lines @xcite , and the case @xmath27 reduces to the local diffusive case
the model in eq .
( [ eq_5 ] ) allows the interpolation between these two regimes .
in particular , eq .
( [ eq_5 ] ) allows the incorporation of parallel transport processes with underlying non - gaussian ( levy @xmath28-stable ) stochastic processes @xcite .
temperature mixing and homogenization in a magnetic island .
panel ( a ) shows the initial condition @xmath29 , and panel @xmath30 the final relaxed state , @xmath31 , where @xmath32 is the magnetic potential , for both local and non - local transport .
panels ( c ) and ( d ) show the solution at @xmath33 in the local and non - local cases respectively . ] substituting eq .
( [ eq_5 ] ) into eq .
( [ eq_1 ] ) , and assuming @xmath11 constant , the transport equation can be written in a compact form as [ eq_6 ] _ t t = _ ^_|s| t , where the operator @xmath34 denotes the symmetric fractional derivative of order @xmath28 , defined as @xmath35=-|k|^\alpha \hat{t}$ ] . here
, @xmath36 denotes the fourier transform @xcite . as expected , in the limit @xmath4 , eq .
( [ eq_6 ] ) reduces to the diffusion equation in eq .
( [ eq_3 ] ) .
the proposed method is based on the green s function solution of eq .
( [ eq_6 ] ) along the lagrangian trajectories of the magnetic field . the unique magnetic field line trajectory ( parametrized by the arc length ) that goes through a point @xmath37 is given by the solution of [ eq_9 ] = , ( s=0)=*r*_p .
thus , given an initial condition in the whole domain , @xmath38 , the initial condition along the field line is @xmath39 and the temperature at @xmath40 at time @xmath41 is given by [ eq_8 ] t(*r*_p ,
t ) = _ s_1^s_2 t_0 g_(s , t ) ds , where @xmath42 is the green s function of eq .
( [ eq_6 ] ) . for unbounded field lines , @xmath43 and the green s function
is given by [ greens ] g_(s , t)= _
-^ e^- t | k |^-
i k s d k . for @xmath4 , eq . ( [ greens ] ) gives the gaussian distribution [ eq_10 ] g_2(s , t ) =( _ t)^-1/2 , and in the non - local free streaming case , @xmath3 , it gives the cauchy distribution [ eq_11 ] g_1(s , t ) = .
for general @xmath28 , @xmath44 $ ] where @xmath45 is the symmetric @xmath28-stable levy distribution , see for example ref .
@xcite . in the case of
closed ( periodic ) field lines , the integration domain in eq .
( [ eq_8 ] ) is a finite interval @xmath46 with @xmath47 , and one has to use the periodic green s function , @xmath48 , obtained by mapping the unbounded green s function , @xmath42 , into the periodic domain .
poincare plot of the weakly chaotic magnetic field used in solution of the parallel heat transport equation shown in fig .
[ fig_3 ] . ] radial temperature profile of the time - asymptotic solution of the parallel heat transport equation for the weakly chaotic magnetic field in fig .
[ fig_2 ] .
the zooms in the successive panels unveil the fractal structure of the devil - staircase profile . ] at this point it is important to indicate a fundamental difference between the work in refs @xcite and the method proposed here . in refs .
@xcite , the flux is calculated throughout the computational domain by integrating along the field lines . however , in these references , once the flux is computed , the lagrangian approach is abandoned and the flux is mapped to gaussian quadrature points for the finite - element standard integration of the temperature evolution equation on a grid . on the other hand ,
the method proposed here is fully lagrangian and completely bypasses the use of finite differences or finite elements integration schemes , circumventing the numerical limitations discussed before .
another clear advantage of the use of the green s function is that the solution at an arbitrary time @xmath49 is obtained directly from the integral in eq .
( [ eq_8 ] ) .
this is to be contrasted with the standard explicit and implicit time discretization schemes that require the knowledge of the solution at previous times @xmath50 before the solution at @xmath49 can be constructed . a similar decoupling occurs in space . according to eq .
( [ eq_8 ] ) , the evaluation of @xmath51 at @xmath37 is fully decoupled from the solution at @xmath52 , yielding naturally to a massively parallel approach . in
what follows , we use the lagrangian green s function method to study local and non - local parallel transport in magnetic fields of increasing levels of complexity .
we start with the study of transport in magnetic islands , which we model with a @xmath53-d magnetic field with vector potential @xmath54 , where @xmath55 and @xmath56 are radial and angular coordinates respectively .
of particular interest is the difference in the temperature relaxation properties in the presence of local and nonlocal transport .
figure 1 shows the solution of the parallel transport of an initial condition of the form @xmath57 with @xmath58 .
as expected , for large enough times , the temperature relaxes to a unique solution of the form @xmath31 in both the local and non - local cases .
however , the dynamic process leading to the final relaxed state is different .
as the figure shows , temperature mixing in the non - local case is less homogeneous and slower than in the local case . in general , from eq .
( [ greens ] ) it follows that the homogenization time of a temperature perturbation with mode number @xmath0 scales as @xmath59 with @xmath4 ( @xmath3 ) in the case of local ( non - local ) transport . for the study of transport in @xmath17-d magnetic fields ,
we assume cylindrical geometry periodic in @xmath60 with period @xmath61 where @xmath62 is a constant .
the magnetic field consists of a perturbed screw - pinch of the form = ( r b /)/[1+(r/)^2 ] _
+ b_0 _ z + * b*_1(r,,z ) , with @xmath63 , @xmath64 , and @xmath65 constants .
the magnetic potential of the perturbation , @xmath66 , consists of a superposition of modes [ eq_20 ] a_z(r,,z)=_m , n a_mn(r ) ( m - n z / r + _ mn ) , with [ eq_21 ] a_mn= a(r ) ( ) ^m . for each @xmath67 , the values of @xmath68and @xmath69 are chosen so that the safety factor satisfies @xmath70 and @xmath71 .
the prefactor @xmath72 is included to guarantee the regularity of the radial eigenfunction near the origin , @xmath73 .
the function , @xmath74/2 $ ] , is introduced to guarantee the vanishing of the perturbation for @xmath75 and the existence of well - defined flux surfaces at the plasma boundary .
self - similar spatio - temporal evolution of the radial temperature profile for local , @xmath4 , transport in a fully chaotic magnetic field . in this case
the scaling function is gaussian and the scaling exponent is @xmath8 .
] self - similar spatio - temporal evolution of the radial temperature profile for non - local , @xmath3 , transport in a fully chaotic magnetic field . in this case
the scaling function is strongly - non - gaussian and , as the dashed - line fit shows , exhibits algebraic decay of the form @xmath76 .
the scaling exponent is @xmath10 . ] in the study of transport in weakly chaotic fields , only two modes were included .
as the poincare plot in figure [ fig_2 ] shows , the magnetic field in this case exhibits a rich fractal - like structure resulting from the existence of higher - order resonances .
figure [ fig_3 ] shows the time asymptotic , final radial temperature solution along the @xmath77 horizontal line in fig .
[ fig_2 ] , corresponding to the initial condition @xmath78 .
as expected , the same state is reached in the presence of local and non - local transport .
however , as mentioned before , the relaxation time is longer in the non - local case .
the fractal structure of the magnetic field gives rise to a devil - staircase temperature profile in which horizontal sections ( resulting from high - order resonances ) are mixed with vertical sections ( resulting from kam invariant circles and cantori ) . to study transport in a fully chaotic magnetic field we considered a set of @xmath79 strongly overlapping modes . in this case , the poincare plot ( not shown ) is fully hyperbolic and does not exhibit any structure .
the initial condition consists of a narrow cylindrical shell " of the form @xmath80 ^ 2 $ ] , with @xmath81 and @xmath82 .
figures [ fig_4 ] and [ fig_5 ] show the time evolution of the radial profile of the temperature averaged in @xmath56 and @xmath60 in the local and the non - local free streaming cases . in both cases
, the temperature exhibits an asymptotic self - similar evolution of the form t _ , z ( , t ) = ( t ) ^-/2 l ( ) , where the similarity variable is defined as @xmath83 with @xmath84 the scaling exponent . from here
, it follows that the second moment scales as @xmath85 . as fig .
[ fig_4 ] shows , in the local transport case the scaling function is approximately gaussian , @xmath86 .
consistent with ref .
@xcite , in this case @xmath87 .
this subdiffusive scaling results from the combination of the diffusive transport along the field line and the quasilinear diffusion of the field line itself due to the magnetic field line chaos . on the other hand , in the nonlocal case
, the scaling function is strongly non - gaussian and exhibits an algebraic decay of the form @xmath88 .
interestingly , in the non - local case @xmath89 .
this diffusive scaling results from the coupling of the quasilinear diffusion of the field line due to chaos and the free - streaming transport along the field line .
summarizing , this letter presented a lagrangian green s function method for the accurate and efficient computation of purely parallel ( @xmath21 ) local and non - local transport in arbitrary magnetic fields with constant @xmath11 .
because of the parallel nature of the lagrangian calculation , the formulation naturally leads to a massively parallel implementation , suitable for today s supercomputers .
the method was applied to compute temperature mixing in magnetic islands , for which the ratio of the non - local free - streaming , @xmath90 , and diffusive , @xmath91 , transport relaxation times scales as @xmath92 . radial transport in @xmath17-d magnetic fields in cylindrical geometry was also studied . in the case of weakly chaotic fields ,
the fractal structure of the devil staircase of the radial temperature profile was resolved . in the case of fully chaotic fields ,
the radial transport exhibited self - similar spatio - temporal behavior .
contrary to the well - known diffusive case , the non - local case exhibits a non - gaussian , algebraic decaying temperature profile with scaling exponent @xmath10 .
current work includes the implementation of the lagrangian green s function method as integral part of a more general multiscale framework that incorporates finite @xmath12 , non - constant @xmath11 , and heat sources . |
let @xmath2 be the fock space representation of @xmath0 introduced by hayashi @xcite and further studied by misra and miwa @xcite , stern @xcite , and kashiwara , miwa and stern @xcite .
it has a standard basis @xmath3 indexed by the set @xmath4 of all integer partitions . in @xcite two canonical bases @xmath5 and @xmath6 of @xmath2 have been constructed .
the subset of @xmath7 consisting of the @xmath8 s for which @xmath9 is @xmath10-regular coincides with kashiwara s lower global basis ( or lusztig s canonical basis ) of the irreducible sub - representation of @xmath2 generated by the highest weight vector @xmath11 . the main motivation for introducing the bases @xmath7 and @xmath12 was their conjectural relation with the decomposition matrices of the @xmath13-schur algebras @xmath14 defined by dipper and james @xcite in connection with the modular representation theory of the finite groups @xmath15 in non - describing characteristic .
conjecture 5.2 of @xcite was proved by varagnolo and vasserot @xcite , who established that the coefficients of the expansion of @xmath16 on the basis @xmath17 are equal to the kazhdan - lusztig polynomials appearing in lusztig s character formula for @xmath18 , where @xmath19 is a complex primitive @xmath10th root of 1 .
let @xmath20 be a prime number coprime to @xmath13 and such that the multiplicative order of @xmath13 in @xmath21 is equal to @xmath10 .
to @xmath22 one associates a `` large @xmath10-core partition '' @xmath23 ( see below , definition [ ckcore ] ) . set @xmath24 , and let @xmath25 be the unipotent block of @xmath26 containing the specht - type modules @xmath27 labelled by partitions @xmath9 with @xmath10-core @xmath28 and @xmath10-weight @xmath29 ( see @xcite ) .
assume that @xmath30 , and let @xmath31 denote the radical series of @xmath27 . in @xcite
the graded composition multiplicities @xmath32 \ , v^i\ ] ] of all @xmath27 in @xmath25 were computed explicitely in terms of the littlewood - richardson coefficients , using a morita equivalence between @xmath25 and the principal block of @xmath33 established in @xcite .
this morita equivalence is similar to a morita equivalence for blocks of symmetric groups conjectured long ago by rouquier ( see @xcite ) and recently proved by chuang and kessar @xcite .
it was also conjectured in @xcite that @xmath34 where @xmath35 denotes the @xmath36-decomposition number of @xcite , that is , the coefficient of @xmath37 in the expansion of @xmath38 .
the main result of this paper is a proof of this conjecture .
simultaneously , we also determine a similar expression for the coefficient @xmath39 of @xmath40 in the expansion of @xmath16 . as a consequence
we obtain a closed and combinatorial expression for two families of parabolic affine kazhdan - lusztig polynomials .
it is remarkable that in fact all these polynomials are just monomials .
another consequence is that we obtain a representation - theoretical interpretation of these particular @xmath36-decomposition numbers : the exponent of @xmath36 in @xmath35 indicates to which layer of the radical filtration of @xmath27 the copies of the simple module @xmath41 belong . in terms of the fock space
, the block @xmath25 corresponds to a distinguished weight space .
more precisely , there is a natural bijection between the set of @xmath10-core partitions and the set of extremal weights @xmath42 of @xmath2 . here
@xmath43 is the highest weight of @xmath2 and @xmath44 belongs to the affine weyl group @xmath45 of @xmath46 .
let @xmath47 be the extremal weight corresponding to @xmath28 in this bijection .
then our results give some closed formulas in terms of the littlewood - richardson coefficients for the @xmath17-expansions of all the @xmath38 s and @xmath16 s of weight @xmath48 , where @xmath49 is the imaginary root .
let @xmath50 denote the set of weights of @xmath2 .
we have the decomposition @xmath51 , where @xmath52 is the @xmath53-orbit of @xmath54 . for @xmath55 ,
let @xmath56 ( @xmath57 ) denote the transition matrix from @xmath17 to @xmath7 ( from @xmath17 to @xmath12 ) in the weight space @xmath58 .
having computed the matrices @xmath59 and @xmath60 , it is easy to determine @xmath56 and @xmath57 for many other weights @xmath61 . indeed , suppose that @xmath55 and @xmath62 is a simple root of @xmath46 such that @xmath63 .
let @xmath64 be the simple reflection of @xmath53 associated with @xmath62 . then @xmath65 and @xmath66 .
it follows that , for each @xmath29 , the orbit @xmath67 can be splitted into a finite number of classes on which the matrices @xmath56 and @xmath57 remain constant .
the class of @xmath68 for which we have found closed formulas is always infinite , hence for each @xmath29 our computations give the transition matrices for an infinite number of weights of @xmath67 .
moreover , in the case of @xmath69 this class is the only infinite one and therefore our formulas calculate in this case the matrices for all but a finite number of weights in each @xmath67 .
these facts , which are easily deduced from the theory of crystal bases , are the fock space counterpart of scopes results @xcite about morita equivalence for blocks of the symmetric groups , and of the analogues of these results for blocks of hecke algebras and unipotent blocks of @xmath70 @xcite .
the paper is structured as follows . in section [ sect2 ]
we review the correspondence between @xmath10-core partitions and the @xmath53-orbit of the fundamental weight @xmath43 of @xmath46 .
section [ sect3 ] and [ sect4 ] recall the main facts about the fock representations of @xmath46 and @xmath0 , and introduce the canonical bases . in section [ sect5 ]
we consider the space @xmath71 of symmetric functions in @xmath10 independent sets of variables @xmath72 with coefficients in @xmath73 .
the standard basis of @xmath71 is given by the products of schur functions @xmath74 .
we introduce two new bases @xmath75 and @xmath76 which are canonical with respect to a certain bar - involution sending @xmath36 to @xmath77 and two crystal lattices @xmath78 and @xmath79 , and we calculate their expansions on the standard basis in terms of the littlewood - richardson coefficients .
heuristically , we regard @xmath71 ( or rather its specialization @xmath80 at @xmath81 ) as the carrier space of the `` canonical commutation relations '' representation of the homogeneous heisenberg subalgebra of @xmath82 .
in fact , the sum of the homogeneous components of @xmath80 of degree less than @xmath20 can be identified to the sum of the complexified grothendieck groups @xmath83 where @xmath7 means the principal block . in this identification
the vectors @xmath84 are mapped to the classes of the unipotent ordinary irreducible modules , and the vectors @xmath85 to the classes of the simple modules . in section [ sect6 ]
we state our main result ( theorem [ th1 ] ) , namely that the canonical bases of certain weight spaces of @xmath2 coincide with the canonical bases of the corresponding homogeneous components of @xmath71 under a natural vector space isomorphism . at @xmath81
this vector space isomorphism is essentially the natural intertwining operator between the principal and the homogeneous realization of the basic representation of @xmath46 ( see @xcite ) .
this is the counterpart in this setting of the morita equivalence of @xcite , and it gives immediately the above - mentioned formulas for @xmath35 and @xmath86 when @xmath9 and @xmath87 have @xmath10-core @xmath28 and @xmath10-weight @xmath88 ( corollary [ formule ] ) .
section [ sect7 ] is devoted to the proof of theorem [ th1 ] .
finally , section [ sect8 ] develops in the context of the fock space the combinatorics underlying scopes isometries between blocks of symmetric groups .
[ [ abacus ] ] a partition is a finite non - increasing sequence of positive integers .
we shall denote by @xmath4 the set of all partitions . by convention
@xmath4 contains the empty partition @xmath89 . to a partition @xmath90 one
associates the infinite decreasing sequence of @xmath91-numbers @xmath92 where @xmath93 is assumed to be @xmath94 for @xmath95 .
fix an integer @xmath96 , and form an infinite abacus with @xmath10 runners labelled @xmath97 from left to right .
the positions on the @xmath98th runner are labelled by the integers having residue @xmath98 modulo @xmath10 . by placing a bead on each @xmath91-number of @xmath99 one
gets the abacus representation of @xmath99 . as is well known , sliding all the beads in the @xmath10-abacus representation of @xmath99 as high as they will go produces the @xmath10-core @xmath100 of @xmath99 .
this is illustrated in figure [ fig1 ] for @xmath101 and @xmath102 . in that case @xmath103 .
note that in the abacus representation of a partition , the number of occupied positive positions is always equal to the number of vacant non - positive positions . in particular , we see that the @xmath10-core partitions are in one - to - one correspondence with the @xmath10-tuples of integers @xmath104 such that @xmath105 .
thus , as shown in figure [ fig1 ] , to the @xmath106-core @xmath107 is associated the @xmath106-tuple @xmath108 .
the next definition introduces certain `` large @xmath10-core partitions '' .
they were first considered by rouquier as labels of certain good blocks of the symmetric groups .
[ ckcore]let @xmath22 .
the partition @xmath23 associated with @xmath29 is the @xmath10-core corresponding to the @xmath10-tuple @xmath109 for @xmath110 and @xmath111 , we have @xmath112 for @xmath110 and @xmath113 , we have @xmath114 [ [ section ] ] let @xmath115 be the affine lie algebra of type @xmath116 @xcite
. denote by @xmath117 the fundamental weights of @xmath118 , by @xmath119 its simple roots , and by @xmath49 the imaginary root .
they are related by @xmath120 if @xmath121 , and @xmath122 if @xmath123 . the free @xmath124-modules @xmath125 and
@xmath126 are called respectively the weight lattice and the root lattice of @xmath118 .
let @xmath127 be the coefficient of @xmath128 in @xmath62 .
the @xmath129 matrix @xmath130 is called the cartan matrix of @xmath118 .
one defines a non - degenerate symmetric bilinear form on @xmath131 by @xmath132 the weyl group @xmath53 of @xmath118 is the subgroup of @xmath133 generated by the simple reflections @xmath64 defined by @xmath134 this is a coxeter group of type @xmath116 , and it is isomorphic to the semi - direct product of the symmetric group @xmath135 by the root lattice @xmath136 of the finite - dimensional lie algebra @xmath137 .
[ [ x ] ] it will be convenient to realize @xmath131 and @xmath138 as sublattices of a lattice @xmath139 of rank @xmath140 .
namely we set @xmath141 and we identify the simple roots and fundamental weights to the following elements of @xmath139 : @xmath142 @xmath143 note that the root lattice @xmath144 of @xmath145 gets identified in this realization to @xmath146 it is easy to check that the above bilinear form is the restriction to @xmath131 of the bilinear form on @xmath139 given by @xmath147 this implies that the action of @xmath53 on @xmath131 can be lifted to @xmath139 by setting @xmath148 [ [ n - cores ] ] the stabilizer @xmath149 of @xmath150 is the subgroup of @xmath53 generated by @xmath151 .
therefore , it is isomorphic to the symmetric group @xmath135
. hence , the orbit @xmath152 is in a natural one - to - one correspondence with the root lattice @xmath144 of @xmath145 .
namely , for @xmath153 write @xmath154 by the formulas of [ x ] for the action of @xmath53 on @xmath139 , we see that the integer coordinates @xmath155 satisfy @xmath105 , and that the projection @xmath156 induces a bijection from @xmath157 to @xmath144 . comparing with [ abacus ] ,
we obtain that the orbit @xmath157 can be identified with the set @xmath158 of @xmath10-core partitions . in other words , there is a natural action of the affine weyl group @xmath53 on @xmath158 , which can be easily described using the abacus representation of the elements of @xmath158 .
namely , since @xmath149 acts on @xmath144 by permuting the vectors @xmath159 , we see that for @xmath160 , the action of @xmath64 on an @xmath10-core partition is obtained by switching the beads of the @xmath98th and @xmath161th runners of its abacus representation .
this is illustrated in figure [ fig2 ] .
on the other hand , we have @xmath162 so that the action of @xmath163 on an @xmath10-core partition is obtained by switching the beads of the first and last runners of its abacus representation , adding one more bead to the first runner and removing one bead from the last runner .
[ [ section-1 ] ] let @xmath164 denote the irreducible @xmath118-module with highest weight @xmath150 ( @xcite , 9 ) .
it is the simplest infinite - dimensional @xmath118-module , and as such , it is often called the basic representation of @xmath118 . the easiest way to construct
it is to embed it into a larger representation of @xmath118 called the fock space representation @xcite .
the fock space @xmath165 has a distinguished @xmath166-basis @xmath167 , on which the chevalley generators @xmath168 of @xmath118 act by @xmath169 where the first sum is over all partitions @xmath87 obtained by removing from the young diagram of @xmath99 a node with @xmath10-residue @xmath98 , and the second sum is over all @xmath170 s obtained from @xmath99 by adding an @xmath98-node . denoting by @xmath89 the empty partition
, we see immediately that @xmath11 is a highest weight vector of @xmath165 with weight @xmath150 .
the cyclic submodule @xmath171 is isomorphic to @xmath164 .
these infinite - dimensional @xmath166-vector spaces are graded by putting @xmath172 if @xmath99 is a partition of @xmath173 .
the dimension of the degree @xmath173 component of @xmath164 is known to be equal to the number of @xmath10-regular partitions of @xmath173 . for a @xmath118-module @xmath174 and a weight @xmath175 ,
we denote by @xmath176 the @xmath177-weight space of @xmath174 .
let @xmath178 denote the set of weights of @xmath174 , that is , the set of @xmath179 such that @xmath176 is non - zero .
it is known ( see @xcite , 12.6 ) that @xmath180 recall from [ n - cores ] that there is a natural bijection , say @xmath181 , from the orbit @xmath157 to the set @xmath158 of @xmath10-cores .
then the weight - space @xmath182 associated with the weight @xmath183 consists precisely of the span of the vectors @xmath37 where @xmath9 has @xmath10-core @xmath184 and @xmath10-weight @xmath29 .
[ [ integrable ] ] the fock space is an integrable @xmath118-module , that is , the operators @xmath185 and @xmath186 are locally nilpotent on @xmath165 .
this implies that @xmath187 for all @xmath188 and @xmath153 .
in fact , it is well known that @xmath189 where @xmath190 is the number of integer partitions of @xmath191 .
define the following linear operators on @xmath165 : @xmath192 ( see @xcite 3.8 ) .
they satisfy @xmath193 and @xmath194 , hence @xmath195 moreover , @xmath196 for all @xmath188 .
it is easy to see that the divided powers @xmath197 and @xmath198 preserve the @xmath124-lattice @xmath199 spanned by the basis @xmath200 .
it follows that @xmath201 also preserves @xmath199 for all @xmath98 .
[ [ scalar ] ] @xmath165 is endowed with a canonical scalar product defined by @xmath202 this form is compatible with the actions of @xmath118 and of the @xmath201 , in the sense that @xmath203 hence we see that @xmath201 is an orthogonal transformation of @xmath165 .
[ [ bosons ] ] for @xmath204 , define an endomorphism @xmath205 of @xmath165 by @xmath206 where the sum is over all @xmath207 such that @xmath208 is a horizontal @xmath10-ribbon strip of weight @xmath191 , and @xmath209 is the spin of the strip ( see @xcite ) . it is known that the @xmath205 pairwise commute , and also commute with the action of @xmath210 $ ] on @xmath165 .
moreover @xmath165 becomes a cyclic module under the action of @xmath211 $ ] , that is , @xmath212 .
[ [ section-2 ] ] let @xmath214 be the quantum affine algebra of type @xmath116 over the field @xmath215 .
let @xmath2 be the @xmath36-deformation of the fock space introduced by hayashi @xcite and further studied by misra and miwa @xcite . as a vector space @xmath216 , and
@xmath217 is a @xmath218-basis of @xmath2 .
the chevalley generators @xmath219 of @xmath213 act on @xmath2 via some simple @xmath36-analogues of the formulas ( [ robinson ] ) , as follows .
let @xmath9 and @xmath207 be two young diagrams such that @xmath207 is obtained from @xmath9 by adding an @xmath98-node @xmath181 .
such a node is called a removable @xmath98-node of @xmath207 , or an indent @xmath98-node of @xmath9 .
let @xmath220 ( @xmath221 ) be the number of indent @xmath98-nodes of @xmath9 ( of removable @xmath98-nodes of @xmath9 ) situated to the right of @xmath181 ( @xmath181 not included ) .
set @xmath222 .
then @xmath223 where the sum is over all partitions @xmath207 such that @xmath208 is an @xmath98-node .
similarly @xmath224 where the sum is over all partitions @xmath9 such that @xmath208 is an @xmath98-node , and @xmath225 is defined as @xmath226 but replacing right by left .
[ [ q - bosons ] ] the cyclic submodule @xmath227 is isomorphic to the basic representation @xmath228 of @xmath213 . as in [ bosons ] , define operators @xmath229 acting on @xmath165 by @xmath230 the sum being over all @xmath207 such that @xmath208 is a horizontal @xmath10-ribbon strip of weight @xmath191 .
they pairwise commute and they also commute with the action of the subalgebra @xmath231 of @xmath213 obtained by omitting the degree generator of the cartan part . the fock space @xmath2 becomes a cyclic module under the action of @xmath232 $ ] , generated by @xmath11 .
[ [ sect4.3 ] ] by kashiwara s theory of crystal bases @xcite , @xmath228 has two canonical bases dual to each other called the lower global base and the upper global base . in @xcite ( see also @xcite )
the canonical bases @xmath233 and @xmath234 of @xmath2 were introduced .
we shall review their definition .
let @xmath235 be the bar involution of @xmath2 defined in @xcite .
it is the unique semi - linear map satisfying @xmath236 and @xmath237 let @xmath238 ( @xmath239 ) be the free @xmath240$]-module ( @xmath241$]-module ) with basis @xmath242 .
the bases @xmath7 and @xmath12 are characterized by the properties @xmath243 for @xmath244 . the subset of @xmath7 consisting of the @xmath8 for which @xmath9 is @xmath10-regular coincides with the lower global base of @xmath245 .
[ [ section-3 ] ] define polynomials @xmath246 and @xmath86 by @xmath247 and @xmath248 the polynomials @xmath246 and @xmath86 are known to be parabolic kazhdan - lusztig polynomials and they belong to @xmath249 $ ] ( see @xcite ) .
[ [ sect5.1 ] ] let @xmath250 denote the @xmath218-algebra of symmetric functions in a countable set @xmath139 of indeterminates ( see @xcite ) .
it is known that @xmath250 is a polynomial algebra in the power - sums symmetric functions @xmath251 other systems of generators are the elementary symmetric functions @xmath252 and the complete symmetric functions @xmath253 .
the schur functions @xmath254 form a linear basis of @xmath250 , and we denote by @xmath255 the scalar product for which this basis is orthonormal
. we may identify @xmath250 to the fock space representation @xmath2 of @xmath0 by identifying the basis @xmath256 with the standard basis @xmath17 of @xmath2 . for @xmath257 ,
denote by @xmath258 the operator of multiplication by @xmath259 and by @xmath260 the adjoint operator , that is , @xmath261 it is known that @xmath262 ( see @xcite , i.5 ex.3 ) .
it follows that the operators @xmath258 , @xmath260 @xmath263 generate the enveloping algebra of the heisenberg lie algebra @xmath264 spanned by the operators @xmath265 , @xmath266 @xmath267 and the identity . thus , @xmath268 can be regarded as the `` canonical commutation relations '' representation of @xmath264 .
[ [ sect5.2 ] ] we introduce the generating function @xmath269 and we put @xmath270 . then , if we denote by @xmath271 a second countable set of indeterminates , and write @xmath272 , we have the cauchy identity ( see @xcite i.4.3 ) @xmath273 more generally , if @xmath274 , @xmath275 are two homogeneous bases of @xmath250 with @xmath276 , then @xmath277 and @xmath278 are adjoint to each other with respect to @xmath255 if and only if @xmath279 ( see @xcite i.4.6 ) .
[ [ sect5.3 ] ] let @xmath280 be @xmath10 countable sets of indeterminates .
we denote by @xmath281 the algebra over @xmath218 of functions symmetric in each set @xmath282 separately .
this is the polynomial algebra in the variables @xmath283 .
a linear basis of @xmath71 is given by the products @xmath284 the space @xmath71 carries a scalar product defined by @xmath285 for @xmath286 .
clearly , the basis @xmath287 is orthonormal for this scalar product . as in [ sect5.1 ] , we regard @xmath71 as the `` canonical commutation relations '' representation of the heisenberg algebra @xmath288 generated by the operators @xmath289 in the rest of this section we are going to introduce two canonical bases of @xmath71 .
[ [ section-4 ] ] given @xmath290 $ ] and @xmath291 , we define symmetric functions of the formal set of variables @xmath292 as follows . for @xmath293
we let @xmath294 denote the image of @xmath259 by the algebra homomorphism from @xmath250 to @xmath71 which maps @xmath295 to @xmath296 ( for the reader familiar with the language of @xmath9-rings @xcite , we consider @xmath297 as elements of a @xmath9-ring , @xmath36 being invertible of rank 1 . ) for example , for @xmath298 we define @xmath299 then we have ( see @xcite i.5.9 ) @xmath300 @xmath301 where the @xmath302 are the littlewood - richardson coefficients , that is , @xmath303 similarly , for @xmath304 , we set @xmath305 and we have ( see @xcite i.3.10 ) @xmath306 where @xmath307 denotes the partition conjugate to @xmath91 . [ [ section-5 ] ] for @xmath308 we write @xmath309
. consider two bases of @xmath71 @xmath310 and @xmath311 consisting of homogeneous elements of degree @xmath312 .
then [ sect5.2 ] may clearly be generalized , as follows .
[ lem3 ] the bases @xmath277 and @xmath278 are adjoint to each other with respect to @xmath255 if and only if @xmath313 [ [ section-6 ] ] for @xmath314 , we define the following elements of @xmath71 : @xmath315 it is easy to see that these are two bases of @xmath71 . by [ sect5.3 ] , their expansions on the orthonormal basis @xmath316 are readily computed in terms of the littlewood - richardson coefficients .
in particular , [ lem4 ] for @xmath317 we have @xmath318 where @xmath319 run through @xmath4 subject to the conditions @xmath320 and @xmath321 here we have used the convention that an empty sum is equal to 0 . thus @xmath322 and @xmath323 , so that @xmath324 is in fact equal to @xmath325 . similarly , @xmath326 and @xmath327 reduces to @xmath328 . using ( [ eq3 ] )
we obtain @xmath329 therefore , @xmath330 and the result follows from the identity @xmath331 and from the fact that @xmath332 is non zero only if @xmath333 . [ [ section-7 ] ] for @xmath334 we set @xmath335 .
let @xmath336 denote any total ordering of @xmath4 such that @xmath337 implies @xmath338 .
we also denote by @xmath336 the corresponding reverse lexicographic ordering of @xmath339 , that is , @xmath340 if and only if there exists @xmath191 such that @xmath341 and @xmath342 for @xmath343 .
the functions @xmath344 and @xmath345 enjoy the following properties .
[ prop1 ] ( i ) for @xmath317 there holds @xmath346 .
\(ii ) for @xmath317 there holds @xmath347 .
\(iii ) the matrices @xmath348_{\ul,\um\in\p^n } , \qquad \left[\ <
\vp_{\ul}(v ) \,,\,s_{\um } \ > \right]_{\ul,\um\in\p^n } \ ] ] are lower unitriangular if their rows and columns are arranged according to the total ordering @xmath349 .
\(i ) follows from lemma [ lem4 ] and the symmetry property @xmath350 of the littlewood - richardson coefficients . to prove ( ii ) we use lemma [ lem3 ] .
we have to show that @xmath351 the left - hand side is equal to @xmath352 here we have used ( [ cauchy ] ) and the basic identities @xmath353 valid for any formal sets of variables @xmath139 and @xmath271 . hence ( ii ) is proved .
suppose that @xmath354 .
then , by lemma [ lem4 ] we see that there exists @xmath355 such that @xmath356 hence either @xmath357 or @xmath358 . in the second case @xmath340 , and in the first one we have @xmath359 , therefore @xmath360 it follows that either @xmath361 or @xmath362 . in the second case
, we obtain that @xmath340 , and in the first one we have @xmath363 .
thus , by induction we see that @xmath354 .
implies that @xmath364 .
moreover , clearly @xmath365 .
next , by ( ii ) we have @xmath366 hence the expansion of @xmath84 on @xmath367 involves only multi - partitions @xmath368 , and by solving this unitriangular system we obtain that @xmath369 implies that @xmath370 and @xmath371 .
[ [ section-8 ] ] define @xmath372_{\ul,\um\in\p^n},\qquad a = s\overline{s^{-1}}=\left[a_{\ul\,\um}(v)\right]_{\ul,\um\in\p^n}.\ ] ] we introduce a semi - linear involution @xmath373 on @xmath71 by requiring that @xmath374 the identity @xmath375 shows that @xmath376 on the other hand , let @xmath78 be the @xmath240$]-lattice in @xmath71 spanned by the vectors @xmath84 .
it follows easily from lemma [ lem4 ] that @xmath377 by proposition [ prop1 ] ( iii ) @xmath378 is unitriangular , hence @xmath1 is also unitriangular .
so by a classical argument ( see @xcite , 7.10 ) , the basis @xmath379 is uniquely determined by ( [ equa1 ] ) and ( [ equa2 ] ) .
this is the canonical basis of @xmath71 associated with the involution @xmath373 and the lattice @xmath78 .
[ [ section-9 ] ] define @xmath380 obviously , @xmath381 is another basis of @xmath71 .
let @xmath79 be the @xmath241$]-lattice in @xmath71 spanned by the vectors @xmath84 .
one can check easily that @xmath382 on the other hand , for @xmath383 we have the formal identities
@xmath384a_0 + \sum_{1\le j\le k-1 } [ k - j ] ( a_j - va_{j-1})\,,\ ] ] where the coefficients @xmath385:=(v^j - v^{-j})/(v - v^{-1})$ ] are bar - invariant .
therefore the expansion of @xmath386 on the basis @xmath387 will have only bar - invariant coefficients .
hence @xmath388 and @xmath381 is the canonical basis of @xmath71 associated with @xmath373 and the lattice @xmath79 .
[ [ section-10 ] ] the coefficients of the expansion of @xmath386 on @xmath316 are given by the next lemma , which is proved in the same way as lemma [ lem4 ] .
[ lem6 ] for @xmath389 we have @xmath390 where the sum is over all families of partitions @xmath391 subject to the conditions @xmath392 and @xmath393 let @xmath394 denote the semi - linear involution of @xmath71 defined by @xmath395 .
the next proposition is easily checked by expansion on @xmath316 . [ prop7 ] the adjoint basis @xmath396 of the canonical basis @xmath379 is related to the basis @xmath381 by @xmath397
[ [ section-11 ] ] we fix @xmath398 . let @xmath399 be the large @xmath10-core associated with @xmath29 ( see definition [ ckcore ] ) . let @xmath400 be the set of partitions with @xmath10-core @xmath28 and let @xmath401 be the subset of partitions with @xmath10-weight @xmath88 .
note that @xmath28 is equal to its conjugate partition @xmath402 .
it follows that the map @xmath403 induces a bijection of @xmath401 .
to @xmath404 we associate its @xmath10-quotient @xmath405 and its @xmath10-sign @xmath406 ( see @xcite ) .
the @xmath10-quotient of a partition is well - defined only up to circular permutations .
we remove this ambiguity by requiring that the partition @xmath407 have @xmath10-quotient @xmath408 .
it is well - known that @xmath409 is a bijection from @xmath400 onto @xmath339 . comparing lemma [ lem4 ] with @xcite
we can see that for @xmath410 @xmath411 where @xmath412 has been defined in section [ sect1 ] .
this motivates the next definition .
[ [ section-12 ] ] let @xmath413 denote the subspace of @xmath2 spanned by all @xmath37 with @xmath414 .
we define a linear map @xmath415 from @xmath413 to @xmath71 by setting @xmath416 this is an isomorphism of vector spaces , which has already been considered in @xcite for @xmath123 and in @xcite in general .
our main result is [ th1 ] for @xmath417 one has @xmath418 theorem [ th1 ] will be proved in section [ sect7 ] .
let @xmath102 , @xmath111 , @xmath419 and @xmath420 .
then @xmath421 therefore @xmath422 on the other hand , @xmath423 , @xmath424 and @xmath425 theorem [ th1 ] and lemma [ lem4 ] readily imply [ formule ] conjecture 5.5 of @xcite is true and we have for @xmath426 @xmath427 where @xmath319 run through @xmath4 subject to the conditions latexmath:[\[|\a^i| = \sum_{0\le j \le i-1 } |\l^j|-|\m^j|\,,\qquad @xmath321 moroever , we also have @xmath429 where the sum is over all families of partitions @xmath391 subject to the conditions @xmath392 and @xmath393 this can be regarded as a combinatorial description of the parabolic kazhdan - lusztig polynomials @xmath35 and @xmath86 in this case . in particular , we note that these polynomials are just monomials when @xmath9 and @xmath207 belong to @xmath401 .
when @xmath123 and @xmath207 is a strict partition , @xmath430 is a decomposition number for the hecke algebra over a field of characteristic @xmath94 at @xmath431 @xcite . in this case , the decomposition numbers for partitions with a large core have been first calculated by james and mathas @xcite . finally , we deduce from proposition [ prop1 ] ( i ) and proposition [ prop7 ] the following interesting symmetries for @xmath426 there holds @xmath432
[ [ section-13 ] ] we want to construct certain elements of @xmath0 whose action on @xmath433 corresponds via @xmath415 to the multiplication by @xmath434 in @xmath71 for @xmath435 .
let @xmath436 be the residue of @xmath437 modulo @xmath10 .
thus @xmath438 if @xmath439 . for convenience we allow the indices @xmath98 of the chevalley generators @xmath440 to belong to @xmath124 by setting @xmath441 . [ lemck ]
let @xmath460 with @xmath461 .
if @xmath462 ( the young diagram of @xmath207 is contained in that of @xmath9 ) then there exists @xmath463 such that @xmath464 for @xmath465 and @xmath466
. moreover @xmath467 is the young diagram of the hook partition @xmath468 .
thus , if @xmath469 and if we consider an @xmath10-ribbon tiling of @xmath467 , we see that all ribbons contributing to @xmath470 have the same shape @xmath468 and the same spin @xmath471 .
let us say that such ribbons have color @xmath98 .
in fact , for @xmath472 any ribbon of colour @xmath98 is situated to the left of any ribbon of colour @xmath473 , as shown in figure [ fig3 ] with @xmath110 . moreover
, there is a unique @xmath10-ribbon tiling of @xmath467 , and since the ribbons with different colors do not mix , for any ribbon tableau @xmath474 of shape @xmath467 the corresponding @xmath10-tuple of tableaux @xmath475 under the stanton - white correspondence @xcite is obtained in a trivial way , as illustrated in figure [ fig4 ] . _
proof of proposition [ prop2 ] _ since @xmath476 , all the @xmath40 occuring in the expansion of @xmath477 have weight @xmath478 , hence the corresponding partitions @xmath207 are obtained from @xmath9 by adding @xmath191 @xmath10-ribbons . since by definition @xmath479
, all these ribbons must be such that one of their ends is a removable cell of @xmath207 with content equal to @xmath480 mod @xmath10 . with our convention for the definition of the @xmath10-quotient ,
this means that all these ribbons have either colour @xmath473 or @xmath481 .
moreover , the ribbons of colour @xmath481 must form a vertical ribbon strip and those of colour @xmath473 must form a horizontal ribbon strip .
conversely , let @xmath207 be a partition obtained from @xmath9 by adding a horizontal ribbon strip of @xmath482 ribbons of colour @xmath473 and a vertical ribbon strip of @xmath483 ribbons of colour @xmath481 .
then there is a unique way of obtaining @xmath207 from @xmath9 by adding in that order : | we give some closed formulas for certain vectors of the canonical bases of the fock space representation of @xmath0 . as a result
, a combinatorial description of certain parabolic kazhdan - lusztig polynomials for affine type @xmath1 is obtained . 0.6 cm |
rare @xmath0 decays induced by flavor changing neutral currents ( fcnc ) play important roles in particle physics , where they are always regarded as ideal places for probing signals of new physics .
the gim suppression of fcnc amplitude is absent in many new physics scenarios beyond the standard model ( sm ) , which could give large enhancement of fcnc contributions over the sm predictions .
however , due to our poor knowledge of non - perturbative qcd , predictions for many interesting exclusive decays are polluted by large hadronic uncertainties .
therefore , it would be of great interest to explore rare @xmath0 decays , which are induced with few hadronic parameters as well as only by fcnc currents .
two body radiative @xmath0 decays involve simple hadronic dynamics with only one hadron in the final states , so they suffer much less pollution than non - leptonic decays . in studying
the radiative decays such as @xmath12 , @xmath13 , the isospin breaking effects between the charged @xmath14 and neutral @xmath15 in these modes are mainly from the annihilation type diagrams @xcite .
many of pure annihilation type radiative decays , such as @xmath16 and @xmath17 , have been analyzed in the qcd factorization approach @xcite and in the perturbative qcd approach @xcite .
we find the branching ratio of @xmath18 is at the order of @xmath19 in the sm .
the decay rate is too small to be observed at presently running @xmath0 factories , babar and belle .
any measurements of the decays at babar and belle would be direct evidences of new physics . in this work
, we explore the decay @xmath16 and similarly the decay @xmath20 in the non - universal @xmath7 model @xcite , which could be naturally derived in certain string constructions @xcite , e6 models @xcite and so on . generally speaking , within the such model
a flavor mixing can be induced at the tree level in the up - type and/or down - type quark sector after diagonalizing their mass matrices . in some new physics model , fcncs due to @xmath7 exchange
can be induced by mixing among the sm quarks and the exotic quarks , which have been predicted to have different @xmath7 quantum numbers . here
we will consider the model in which the interaction between the @xmath7 boson and fermions are flavor non - universal for left handed couplings and flavor diagonal for right handed couplings .
the effects of the @xmath7 on other processes of the interest have been investigated in a number of papers such as @xcite , especially in @xmath0 physics @xcite .
the recent review about @xmath7 in detail is referred to ref .
@xcite . to keep completeness ,
we first calculate these decays in the naive factorization approach .
then we add the vertex corrections to the four quark operators , which have been performed in the so called qcd factorization approach @xcite in the sm , utilizing the light - cone wave functions of the light vector mesons .
a similar work within the r - parity violating susy can be also found in ref .
. however , in this work we will revisit these processes with the updated parameters in the non - universal @xmath7 model .
in the sm , the common starting point is the effective weak hamiltonian which mediates flavor - changing neutral current transitions of the type @xmath21 : @xmath22.\end{aligned}\ ] ] the explicit forms of the operators @xmath23 and the corresponding wilson coefficients @xmath24 at the scale of @xmath25 can be found in ref .
@xmath26 , @xmath27 are the cabibbo - kabayashi - maskawa ( ckm ) matrix elements . according to the effective hamiltonian ( [ hamiton ] ) , we can draw the lowest order diagram of this channel , as shown in fig .
1 , which is dominated by the photon radiated from light quark in @xmath28 meson . when the photon is radiated from heavy @xmath29 quark , the energetic up or down quark will suppress the @xmath30 production by a power of @xmath31 , therefore , here we neglect its contribution .
this point have been clearly discussed in @xmath12 decays @xcite .
[ fig:1 ] for convenience , we denote that @xmath32 and @xmath33 are transverse polarization vectors of the final vector meson and photon , respectively .
the photon energy and momentum are defined as @xmath34 and @xmath35 , and the momentum of @xmath36 meson is @xmath37 . in the rest frame of @xmath36 meson
, we take the photon and the vector meson moving along the @xmath38 and @xmath39 directions , respectively . within the effective hamiltonian and naive factorization hypothesis , we can write down the amplitudes as follows : @xmath40 \\
\times \sqrt{4\pi\alpha_e}f_{\rho}m_{\rho } f_v \biggl\ { -\epsilon_{\mu\nu\rho\sigma } \eta^{*\mu}_{\perp}\epsilon^{*\nu}_{\perp } v^{\rho}q^{\sigma } + i\left[(\eta_{\perp}^{*}\cdot\epsilon^{*}_{\perp } ) ( v\cdot q)-(\eta^{*}_{\perp}\cdot q ) ( v\cdot\epsilon^{*}_{\perp})\right]\biggr\},\end{gathered}\ ] ] @xmath41
\\ \times \sqrt{4\pi\alpha_e}f_{\rho}m_{\rho } f_v \biggl\ { -\epsilon_{\mu\nu\rho\sigma } \eta^{*\mu}_{\perp}\epsilon^{*\nu}_{\perp } v^{\rho}q^{\sigma } + i\left[(\eta_{\perp}^{*}\cdot\epsilon^{*}_{\perp } ) ( v\cdot q)-(\eta^{*}_{\perp}\cdot q ) ( v\cdot\epsilon^{*}_{\perp})\right]\biggr\},\end{gathered}\ ] ] @xmath42 \\ \times \sqrt{4\pi\alpha_e}f_{\rho}m_{\rho } f_v \biggl\ { -\epsilon_{\mu\nu\rho\sigma } \eta^{*\mu}_{\perp}\epsilon^{*\nu}_{\perp } v^{\rho}q^{\sigma } + i\left[(\eta_{\perp}^{*}\cdot\epsilon^{*}_{\perp } ) ( v\cdot q)-(\eta^{*}_{\perp}\cdot q ) ( v\cdot\epsilon^{*}_{\perp})\right]\biggr\},\end{gathered}\ ] ] where @xmath43 is defined as the combination of the wilson coefficients , @xmath44 for an odd ( even ) value of @xmath45 .
the form factor @xmath46 has been defined in refs .
@xcite @xmath47.\end{gathered}\ ] ] in order to calculate the form factor @xmath46 , we need two - particle light - cone projector for an initial @xmath0 meson : @xmath48 \biggr\}_{\alpha\beta},\end{aligned}\ ] ] where @xmath49 and @xmath50 are the leading twist light - cone distribution functions @xcite .
thus , we obtain the standard result : @xmath51 where @xmath52 is the charge of the @xmath53 quark in units of the proton s charge . because we have little knowledge about the distribution of the heavy meson , the integral in above formulae is often parameterized as : @xmath54 consequently , we write down the helicity amplitudes for these channels as : @xmath55,\\ { \cal m}_{b_s\to \omega \gamma}^{+ + } & = & i\frac{g_f}{2}\sqrt{4\pi\alpha_e}f_{v } f_{\phi}m_{\phi}m_{b}\left[v_{ub } v_{us}^*a2-v_{tb } v_{ts}^ * \left(2a_3 + 2 a_5 + \frac{1}{2}c_7+\frac{1}{2 } a_9 \right)\right],\\ { \cal m}_{b_d\to \phi \gamma}^{+ + } & = & i\frac{g_f}{\sqrt{2}}\sqrt{4\pi\alpha_e}f_{v } f_{\phi}m_{\phi}m_{b}\left[-v_{tb } v_{td}^ * \left(a_3+a_5-\frac{1}{2 } a_7 -\frac{1}{2}a_9 \right)\right],\\ { \cal m}_{b_s\to \rho
\gamma}^{--}&=&{\cal m}_{b_s\to \omega \gamma}^{--}={\cal m}_{b_d\to \phi \gamma}^{--}=0.\end{aligned}\ ] ] .summary of input parameters [ cols="^",options="header " , ] because there are both weak and strong phases in the decay modes @xmath56 and @xmath57 , we can get the cp asymmetries of these two channels as follows , @xmath58 by the definition of cp asymmetry @xmath59 for the decay mode @xmath60 , there is only weak phase from @xmath61 , so that the cp asymmetry in this decay disappears within the sm .
now we consider the effects due to an extra @xmath62 gauge boson @xmath7 .
usually , the flavor mixing can be induced at the tree level in up - type and/or down - type quark sector after diagonalizing their mass matrices . in some new physics model , fcncs due to @xmath7 exchange
can be induced by mixing among the sm quarks and the exotic quarks , which is predicted and can make different @xmath7 quantum numbers after the mixing . here
we will consider the model in which the interaction between the @xmath7 boson and fermions are flavor non - universal for left handed couplings and flavor diagonal for right handed couplings .
for simplicity , we neglected the mixing between the @xmath63 and @xmath7 and the evolution effect from the high scale @xmath64 to the @xmath65 scale .
we start to set up the relevant interactions with the new @xmath7 gauge particle . following the convention in ref .
@xcite , we write the couplings of the @xmath7-boson to fermions as @xmath66\psi_i , \label{eq : jzprime}\end{aligned}\ ] ] where @xmath45 is the family index and @xmath67 labels the fermions and @xmath68 . according to some string construction or gut models such as @xmath69 , it is possible to have family non - universal @xmath70 couplings .
that is , even though @xmath71 are diagonal , the couplings are not family universal . after rotating to the physical basis ,
fcncs generally appear at tree level in both left handed and right handed sectors .
explicitly , @xmath72 moreover , these couplings may contain cp - violating phases beyond that of the sm .
the effective hamiltonians describing the transition mediated by the @xmath11 boson have the form as : @xmath73,\;\nonumber\\ { \cal h}_{eff}^{z'}(b\to dq\bar q)= - \frac{4 g_f}{\sqrt 2 } v_{tb}v_{td}^ * \left [ \left ( \frac{g ' m_z } { g_1 m_{z ' } } \right ) ^2 \frac{b_{db}^l}{v_{tb } v_{td}^ * } ( b_{qq}^l o_9 + b_{qq}^r o_7)\right ] , \;\end{aligned}\ ] ] where @xmath74 and @xmath75 denote the left ( right ) handed effective @xmath11 couplings of the quarks @xmath45 and @xmath76 at the weak scale .
the diagonal elements are real due to the hermiticity of the effective hamiltonian but the off diagonal elements may contain effective weak phases . with the definition @xmath77 we can parameterize these coefficients as @xmath78 where @xmath79 , ( @xmath80 is the weak phase associated with @xmath81 ) . in the following discussion
, we adopt @xmath82 for convenience ( note that a possible negative sign can be accounted for by shifting @xmath80 by @xmath83 ) , which has been stated in detail in ref . @xcite . in order to see the effect of @xmath11 boson
, we have to know the values of the @xmath84 and @xmath85 or equivalently @xmath81 , @xmath86 and @xmath87 .
generally , one expects @xmath88 , if both the @xmath89 gauge groups have the same origin from some grand unified theories , and @xmath90 for a tev scale neutral @xmath11 boson , which yields @xmath91 .
however , in ref .
@xcite assuming a small mixing between @xmath92 bosons the value of @xmath93 is taken as @xmath94 . it has been shown in @xcite that the mass difference of @xmath95 mixing can be explained if @xmath96 . similarly , the cp asymmetry anomaly in @xmath97 can be resolved if @xmath98 .
so , we assume that @xmath99 assuming only left handed couplings are present , the bound on fcnc @xmath11 coupling ( @xmath100 from @xmath101 mass difference has been obtained in @xcite as @xmath102 using @xmath91 , one can obtain a more stringent bound on @xmath103 .
from these two relations one can obtain @xmath104 .
thus , it is expected that @xmath105 with the ckm matrix elements considered . however ,
in our analysis here we vary their values within the range @xmath106 , since the major purpose of this work is searching for new physics signal rather than obtaining acute numerical results . ,
variation of the cp averaged branching ratio ( in units of @xmath107 ) with @xmath108 ( in units of @xmath109 ) and the new weak phase @xmath80 ( left panel ) and the variation of direct cp asymmetry ( in % ) with the new weak phase @xmath80 ( right panel ) where the solid , dot - dashed and dashed lines correspond to @xmath110 and @xmath111.,title="fig:",width=359,height=226 ] , variation of the cp averaged branching ratio ( in units of @xmath107 ) with @xmath108 ( in units of @xmath109 ) and the new weak phase @xmath80 ( left panel ) and the variation of direct cp asymmetry ( in % ) with the new weak phase @xmath80 ( right panel ) where the solid , dot - dashed and dashed lines correspond to @xmath110 and @xmath111.,title="fig:",width=226,height=226 ] , variation of the cp averaged branching ratio ( in units of @xmath112 ) with @xmath108 ( in units of @xmath109 ) and the new weak phase @xmath80 ( left panel ) and the variation of direct cp asymmetry ( in % ) with the new weak phase @xmath80 ( right panel ) where the solid , dot - dashed and dashed lines correspond to @xmath110 and @xmath111.,title="fig:",width=302,height=226 ] , variation of the cp averaged branching ratio ( in units of @xmath112 ) with @xmath108 ( in units of @xmath109 ) and the new weak phase @xmath80 ( left panel ) and the variation of direct cp asymmetry ( in % ) with the new weak phase @xmath80 ( right panel ) where the solid , dot - dashed and dashed lines correspond to @xmath110 and @xmath111.,title="fig:",width=226,height=226 ] , variation of the cp averaged branching ratio ( in units of @xmath112 ) with @xmath108 ( in units of @xmath109 ) and the new weak phase @xmath80 ( left panel ) and the variation of direct cp asymmetry ( in % ) with the new weak phase @xmath80 ( right panel ) where the solid , dot - dashed and dashed lines correspond to @xmath110 and @xmath111.,title="fig:",width=264,height=226 ] , variation of the cp averaged branching ratio ( in units of @xmath112 ) with @xmath108 ( in units of @xmath109 ) and the new weak phase @xmath80 ( left panel ) and the variation of direct cp asymmetry ( in % ) with the new weak phase @xmath80 ( right panel ) where the solid , dot - dashed and dashed lines correspond to @xmath110 and @xmath111.,title="fig:",width=226,height=226 ] it is noted that the other wilson coefficients may also receive contributions from the @xmath70 boson through renormalization group ( rg ) evolution . with our assumption that no significant rg running effect between @xmath113 and @xmath65 scales , the rg evolution of the modified wilson coefficients is exactly the same as the ones in the sm @xcite . using the values of these coefficients at @xmath114 scale we can analogously obtain the new contribution to the transition amplitude as done in the case of @xmath115 boson .
the @xmath116 induced by the @xmath117are given as : @xmath118 and the contributions of new physics can be formulated as @xmath119,\\ \delta{\cal m}_{b_s\to \omega \gamma}^{+ + } & = & i\frac{g_f}{2}\sqrt{4\pi\alpha_e}f_{v } f_{\phi}m_{\phi}m_{b}\left[v_{ub } v_{us}^*\delta a_2^{\prime}-v_{tb } v_{ts}^ * \left(2 \delta a_3^{\prime}+2\delta a_5^{\prime } + \frac{1}{2}\delta a_7^{\prime}+\frac{1}{2}\delta a_9^{\prime } \right)\right],\\ \delta{\cal m}_{b_d\to \phi \gamma}^{+ + } & = & i\frac{g_f}{\sqrt{2}}\sqrt{4\pi\alpha_e}f_{v } f_{\phi}m_{\phi}m_{b}\left[-v_{tb } v_{td}^ * \left(\delta a_3^{\prime}+\delta a_5^{\prime}-\frac{1}{2}\delta a_7^{\prime } -\frac{1}{2}\delta a_9^{\prime } \right)\right].\end{aligned}\ ] ] noted that @xmath116s involve the new weak phase , which may change the cp asymmetries remarkably .
now using @xmath120 and taking the decay mode @xmath121 as an example , we also list the correction to @xmath122 from @xmath7 boson in the last column of table 2 . from the table , we note that the corresponding wilson coefficient of electro - weak penguin is enhanced remarkably with suitable parameter @xmath108 , which may affect the branching ratio and other observed values . in fig .
[ x1 ] , fig .
[ x2 ] and fig .
[ x3 ] , we show the variation of the cp averaged branching ratios of @xmath16 , @xmath121 and @xmath123 with @xmath108 and the new weak phase @xmath124 ( left panel ) and the corresponding direct cp violation with @xmath124 ( right panel ) , respectively . as anticipated , if the @xmath125 , the branching ratios can be enhanced remarkably , which can reach to @xmath8 for @xmath126 and @xmath5 for the @xmath127 .
all results are enhanced two orders of magnitude over the predictions of the sm . from the figures
, we can also argue that there have been significant enhancements in the branching ratios for large @xmath108 , or in other words for a lighter @xmath11 boson . in the experimental side , these results may be inaccessible at the belle and babar presently
. however , it is large enough for lhc - b and/or super b - factories .
moreover , we find that these decay modes may have large cp asymmetries when @xmath128 and suitable weak phase @xmath80 .
it implies that the contributions from new physics and from the sm can be comparable , and the interference between them leads to large asymmetries .
furthermore , future observations of these modes could in turn help us to constrain the mass of @xmath11 boson within the model .
in this work , we have studied pure annihilation type radiative processes @xmath129 , @xmath130 and @xmath131 within the qcd factorization . after adding the vertex corrections to the naive factorization approach
, we find that the non - factorizable contributions can enhance the branching ratio of @xmath132 decays , however , the branching ratio of @xmath133 is almost unchanged , but for @xmath134 the branching ratio is even lowered , because the corrections to ew penguin operators are much smaller than those to qcd penguin operators .
the smallness of these decays within the sm makes them sensitive probes of flavor physics beyond the sm . to explore new physics potential ,
we have estimated the contribution of the non - universal @xmath7 model to the decays . if @xmath125 , the branching ratios can be enhanced remarkably , and reach to @xmath8 for @xmath126 and @xmath5 for the @xmath127 .
moreover , we have also predicted large cp asymmetries in suitable parameter spaces .
these results can be tested at the lhc - b and/or super b - factories in future .
the observations of these modes could in turn help us to constrain the mass of @xmath11 within the model .
the work of c.s.k . was supported in part by basic science research program through the nrf of korea funded by moest ( 2009 - 0088395 ) and in part by kosef through the joint research program ( f01 - 2009- 000 - 10031 - 0 ) .
the work of y.li was supported by the brain korea 21 project and by the national science foundation under contract nos.10805037 and 10625525 .
y.li thanks prof y
.- d yang for valuable discussion .
s. w. bosch and g. buchalla , nucl .
b * 621 * , 459 ( 2002 ) [ arxiv : hep - ph/0106081 ] .
p. ball , g. w. jones and r. zwicky , phys .
d * 75 * , 054004 ( 2007 ) [ arxiv : hep - ph/0612081 ]
. w. wang , r. h. li and c. d. lu , arxiv:0711.0432 [ hep - ph ] .
x. q. li , g. r. lu , r. m. wang and y. d. yang , eur .
j. c * 36 * , 97 ( 2004 ) [ arxiv : hep - ph/0305283 ] .
g. r. lu , r. m. wang and y. d. yang , eur .
j. c * 34 * , 291 ( 2004 ) [ arxiv : hep - ph/0308256 ] .
y. li and c. d. lu , phys . rev .
d * 74 * , 097502 ( 2006 ) [ arxiv : hep - ph/0605220 ] .
p. langacker and m. plumacher , phys .
d * 62 * , 013006 ( 2000 ) [ arxiv : hep - ph/0001204 ] .
g. buchalla , g. burdman , c. t. hill and d. kominis , phys .
d * 53 * , 5185 ( 1996 ) [ arxiv : hep - ph/9510376 ] .
e. nardi , phys .
d * 48 * , 1240 ( 1993 ) [ arxiv : hep - ph/9209223 ] .
j. erler , p. langacker , s. munir and e. r. pena , jhep * 0908 * , 017 ( 2009 ) [ arxiv:0906.2435 [ hep - ph ] ] . s. l. chen and n. okada , phys .
b * 669 * , 34 ( 2008 ) [ arxiv:0808.0331 [ hep - ph ] ] .
v. barger , c. w. chiang , p. langacker and h. s. lee , phys .
b * 598 * , 218 ( 2004 ) [ arxiv : hep - ph/0406126 ] ; v. barger , l. everett , j. jiang , p. langacker , t. liu and c. wagner , phys .
d * 80 * , 055008 ( 2009 ) [ arxiv:0902.4507 [ hep - ph ] ] ; v. barger , l. l. everett , j. jiang , p. langacker , t. liu and c. e. m. wagner , arxiv:0906.3745 [ hep - ph ] ; v. barger , c. w. chiang , p. langacker and h. s. lee , phys .
b * 580 * , 186 ( 2004 ) [ arxiv : hep - ph/0310073 ] .
k. cheung , c. w. chiang , n. g. deshpande and j. jiang , phys .
b * 652 * , 285 ( 2007 ) [ arxiv : hep - ph/0604223 ] . c. w. chiang , n. g. deshpande and j. jiang , jhep * 0608 * , 075 ( 2006 ) [ arxiv : hep - ph/0606122 ] . c. h. chen and h. hatanaka , phys .
d * 73 * , 075003 ( 2006 ) [ arxiv : hep - ph/0602140 ] .
q. chang , x. q. li and y. d. yang , jhep * 0905 * , 056 ( 2009 ) [ arxiv:0903.0275 [ hep - ph ] ] ; q. chang , x. q. li and y. d. yang , arxiv:0907.4408 [ hep - ph ] . c. h. chen , arxiv:0911.3479 [ hep - ph ] . c. w. chiang , r. h. li and c. d. lu , arxiv:0911.2399 [ hep - ph ] .
r. mohanta and a. k. giri , phys .
d * 79 * , 057902 ( 2009 ) [ arxiv:0812.1842 [ hep - ph ] ] .
p. langacker , arxiv:0801.1345 [ hep - ph ] .
m. beneke , g. buchalla , m. neubert and c. t. sachrajda , phys .
lett . * 83 * , 1914 ( 1999 ) [ arxiv : hep - ph/9905312 ] ; m. beneke , g. buchalla , m. neubert and c. t. sachrajda , nucl .
b * 591 * , 313 ( 2000 ) [ arxiv : hep - ph/0006124 ] . for a review , see g. buchalla , a.j .
buras , m.e .
lautenbacher , rev .
phys . * 68 * , 1125 ( 1996 ) .
e. lunghi , d. pirjol and d. wyler , nucl .
b * 649 * , 349 ( 2003 ) [ arxiv : hep - ph/0210091 ] .
s. descotes - genon and c. t. sachrajda , nucl .
b * 650 * , 356 ( 2003 ) [ arxiv : hep - ph/0209216 ] .
p. ball and e. kou , jhep * 0304 * , 029 ( 2003 ) [ arxiv : hep - ph/0301135 ] .
a. g. grozin and m. neubert , phys .
d * 55 * , 272 ( 1997 ) [ arxiv : hep - ph/9607366 ] .
p. ball , v. m. braun , y. koike and k. tanaka , nucl .
b * 529 * , 323 ( 1998 ) [ arxiv : hep - ph/9802299 ] . | we study charmless pure annihilation type radiative @xmath0 decays within the qcd factorization approach . after adding the vertex corrections to the naive factorization approach
, we find that the branching ratios of @xmath1 , @xmath2 and @xmath3 within the standard model are at the order of @xmath4 , @xmath5 and @xmath6 , respectively .
the smallness of these decays in the standard model makes them sensitive probes of flavor physics beyond the standard model . to explore their physics potential ,
we have estimated the contribution of @xmath7 boson in the decays . within the allowed parameter space ,
the branching ratios of these decay modes can be enhanced remarkably in the non - universal @xmath7 model : the branching ratios can reach to @xmath8 for @xmath9 and @xmath5 for the @xmath10 , which are large enough for lhc - b and/or super b - factories to detect those channels in near future .
moreover , we also predict large cp asymmetries in suitable parameter space .
the observation of these modes could in turn help us to constrain the @xmath11 mass within the model . |
the problem of compactifying the generalized jacobian of a singular curve has been studied since igusa s work @xcite around 1950 .
he constructed a compactification of the jacobian of a nodal and irreducible curve @xmath0 as the limit of the jacobians of smooth curves approaching @xmath0 .
igusa also showed that his compactification does not depend on the considered family of smooth curves .
an intrinsic characterization of the boundary points of the igusa s compactification as the torsion free , rank 1 sheaves which are not line bundles is due to mumford and mayer .
the complete construction for a family of integral curves over a noetherian hensel local ring with residue field separably closed was carried out by dsouza @xcite .
one year later , altman and kleiman @xcite gave the construction for a general family of integral curves . when the curve @xmath0 is reducible and nodal , oda and seshadri @xcite produced a family of compactified jacobians @xmath5 parameterized by an element @xmath6 of a real vector space .
seshadri dealt in @xcite with the general case of a reduced curve considering sheaves of higher rank as well . in 1994 , caporaso showed @xcite how to compactify the relative jacobian over the moduli of stable curves and described the boundary points of the compactified jacobian of a stable curve @xmath0 as invertible sheaves on certain deligne - mumford semistable curves that have @xmath0 as a stable model . recently
, pandharipande @xcite has given another construction with the boundary points now representing torsion free , rank 1 sheaves and he showed that the caporaso s compactification was equivalent to his .
on the other hand , esteves @xcite constructed a compactification of the relative jacobian of a family of geometrically reduced and connected curves and compared it with seshadri s construction @xcite using theta funtions and alexeev @xcite gave a description of the jacobian of certain singular curves in terms of the orientations on complete subgraphs of the dual graph of the curve .
simpson s work @xcite on the moduli of pure coherent sheaves on projective spaces allows us to define in a natural way the jacobian of any projective curve @xmath0 as the space @xmath7 of equivalence classes of stable invertible sheaves with degree @xmath1 .
this is precisely the definition we adopt and we also denote by @xmath8 the space of equivalence classes of semistable , pure dimension 1 , rank 1 sheaves with degree @xmath1 . in some recent papers about the moduli spaces of stable vector bundles on elliptic fibrations , for instance @xcite , the jacobian in the sense of simpson of spectral curves appears .
beauville @xcite uses it as well in counting the number of rational curves on k3 surfaces .
this suggests the necessity to determine the structure of these simpson jacobians . the aim of this work is to describe the simpson jacobian of a generalized tree - like curve , that is , a projective , reduced and connected curve such that the intersection points of its irreducible components are disconnecting ordinary double points .
let @xmath2 be the fixed polarization on @xmath0 .
let @xmath9 denote the irreducible components of @xmath0 and let @xmath10 be the intersection points of @xmath11 .
the first result we use is a lemma of teixidor @xcite that allows to order @xmath4 and to find subcurves @xmath12 of @xmath0 that are also generalized tree - like curves and intersect its complement in @xmath0 at just one point @xmath13 .
then , we write @xmath14 for @xmath15 , where @xmath16 is the degree of @xmath2 , @xmath17 is the degree of the induced polarization on @xmath12 and @xmath18 is the residue class of @xmath1 minus the arithmetic genus of @xmath0 modulo @xmath16 .
the theorem that gives the description of the schemes @xmath7 and @xmath8 is the following : * theorem [ t ] . *
_ let @xmath19 , @xmath20 , be a generalized tree - like curve with a polarization @xmath2 .
if @xmath21 is not integer for @xmath22 , then @xmath7 is equal to @xmath23 and @xmath24 where @xmath25 are integer numbers inductively constructed . _
if @xmath21 is integer for some @xmath22 , then @xmath7 and @xmath26 are empty and @xmath27 where @xmath3 are integer numbers recursively constructed with an algorithm .
as we will see , these integer numbers , @xmath3 and @xmath25 , depend only on the degree of @xmath2 and on the fixed ordering of the irreducible components of @xmath0 .
the first we prove is that a pure dimension 1 sheaf of rank 1 and degree @xmath1 on @xmath0 that is locally free at @xmath13 , @xmath28 , is stable if and only if @xmath29 for @xmath15 and @xmath30 is obtained by gluing torsion free rank 1 sheaves on @xmath4 whose degrees are precisely the integers @xmath25 ( see lemmas [ lem5 ] and [ lem6 ] if @xmath30 is a line bundle , otherwise the proof is the same ) . by the contrary , if @xmath30 is not locally free at some intersection point @xmath13 , the sheaf can not be stable .
finally , we prove that strictly semistable torsion free rank 1 sheaves exist only when @xmath31 for some @xmath15 and in this case we construct a jordan - hlder filtration for it to conclude that its @xmath32-equivalence class belongs to @xmath33 .
it is important to note that , although @xmath34 for any integers @xmath1 and @xmath35 , because @xmath4 is an integral curve , to obtain a point in @xmath8 it is necessary to consider sheaves on @xmath4 of degrees exactly the integer numbers @xmath3 theorem [ t ] allows also to describe a wall - crossing structure for these simpson jacobians , in other words , to analyze the variation of the moduli spaces @xmath7 as the polarization changes .
the variation of the polarization is studied by considering the variation of the numbers @xmath21 , ( @xmath15 ) .
the space of possible values of such numbers contains a finite number of walls , the hyperplanes @xmath36 where @xmath37 is an integer number @xmath38 .
the complement of the walls has a finite number of connected components each of them called a chamber .
therefore , theorem [ t ] means that in the interior of a given chamber the jacobian @xmath7 is not empty and independent of the choice of numbers @xmath21 .
when the numbers @xmath21 lie in a wall , the theorem shows that the simpson jacobian of @xmath0 becomes empty . finally ,
if we choose two families of numbers @xmath21 lying in the interior of two different chambers , then the structure of the corresponding simpson jacobians is the same but the degrees @xmath25 of the induced line bundles on every component @xmath4 change according to . if @xmath39 is a real number , we use [ @xmath39 ] to denote the greatest integer less than or equal to @xmath39 .
let @xmath0 be a projective , reduced and connected curve over an algebraically closed field @xmath41 .
let @xmath9 denote the irreducible components of @xmath0 .
let @xmath42 be an ample invertible sheaf on @xmath0 , let @xmath2 be the associated polarization and let @xmath16 be the degree of @xmath2 .
let @xmath30 be a coherent sheaf on @xmath0 .
we say that @xmath30 is pure of dimension one or torsion free if for all nonzero subsheaves @xmath43 the dimension of @xmath44 is 1 .
the rank and the degree ( with respect to @xmath2 ) of @xmath30 are the rational numbers @xmath45 and @xmath46 determined by the hilbert polynomial @xmath47 the slope of @xmath30 is defined by @xmath48 the sheaf @xmath30 is stable ( resp .
semistable ) with respect to @xmath2 if @xmath30 is pure of dimension one and for any proper subsheaf @xmath43 one has @xmath49 in @xcite simpson defined the rank as the integer number @xmath50 and the slope as the quotient @xmath51 stability and semistability considered in terms of simpson s slope and in terms of @xmath52 are equivalent
. we adopt these definitions of rank and degree of @xmath30 because they coincide with the classical ones when the curve is integral .
* notation : * we shall use the following notation . for every proper subcurve @xmath53 of @xmath0
, we will denote by @xmath54 the restriction of @xmath30 to @xmath53 modulo torsion , that is , @xmath55 , @xmath56 will be the surjective morphism @xmath57 and @xmath58 .
we shall denote by @xmath59 the degree of the induced polarization @xmath60 on @xmath53 .
if @xmath61 then we shall write @xmath62 .
we first recall some general properties we will use later .
[ lem1 ] let @xmath30 be a pure dimension one sheaf on @xmath0 supported on a subcurve @xmath53 of @xmath0 .
then @xmath30 is stable ( resp .
semistable ) with respect to @xmath60 if and only if @xmath30 is stable ( resp .
semistable ) with respect to @xmath2 .
it follows from the equality @xmath63 where @xmath64 is the inclusion map .
[ lem2 ] a torsion free rank 1 sheaf @xmath30 on @xmath0 is stable ( resp .
semistable ) if and only if @xmath65 ( resp .
@xmath66 ) for every proper subcurve @xmath53 of @xmath0 . given a subsheaf @xmath67 of @xmath30 such that @xmath68 , let us consider the complementary subcurve @xmath69 of @xmath53 in @xmath0 , i.e. the closure of @xmath70 .
since @xmath71 is torsion free , we have @xmath72 with @xmath73 so that @xmath74 and the result follows .
let @xmath75 denote the intersection points of @xmath76 .
it is known ( see @xcite ) that for every pure dimension one sheaf @xmath30 on @xmath0 there is an exact sequence @xmath77 where @xmath78 is a torsion sheaf whose support is precisely the set of those points @xmath13 , @xmath79 , where @xmath30 is locally free .
from this exact sequence , we get @xmath80 in particular , if @xmath30 is a torsion free sheaf of rank 1 with respect to @xmath2 , then for every proper subcurve @xmath53 of @xmath0 @xmath54 is torsion free of rank 1 with respect to @xmath60 .
the following lemma , also due to seshadri , describes the stalk of a torsion free sheaf on @xmath0 at the points @xmath13 .
[ lem3 ] let @xmath30 be a pure dimension one sheaf on @xmath0 .
if @xmath13 is an ordinary double point lying in two irreducible components @xmath81 and @xmath82 , then @xmath83 where @xmath84 are the integer numbers determined by : @xmath85 see @xcite , huitime partie , prop .
3 . according to the general theory , for every semistable sheaf @xmath30 with respect to @xmath2 there is a jordan - hlder filtration @xmath86 with stable quotients @xmath87 and @xmath88 for @xmath89 .
this filtration need not be unique , but the graded object @xmath90 does not depend on the choice of the jordan - hlder filtration .
two semistable sheaves @xmath30 and @xmath91 on @xmath0 are said to be @xmath32-equivalent if @xmath92 .
note that two stable sheaves are @xmath32-equivalent only if they are isomorphic . in the relative case , given a scheme @xmath32 of finite type over @xmath41 , a projective morphism of schemes @xmath93 whose geometric fibers are curves and a relative polarization @xmath2
, we define the relative rank and degree of a coherent sheaf @xmath30 on @xmath0 , flat over @xmath32 , as its rank and degree on fibers , and we say that @xmath30 is relatively pure of dimension one ( resp .
stable , resp .
semistable ) if it is flat over @xmath32 and if its restriction to every geometric fiber of @xmath94 is pure of dimension one ( resp .
stable , resp .
semistable ) .
let @xmath95 ( resp .
@xmath96 ) be the functor which to any @xmath32-scheme @xmath78 associates the set of equivalence classes of stable invertible ( resp .
relatively torsion free rank 1 ) sheaves on @xmath97 with relative degree @xmath1 .
two such sheaves @xmath30 and @xmath91 are said to be equivalent if @xmath98 , where @xmath99 is a line bundle on @xmath78 and @xmath100 is the natural projection .
similarly , we define the functor @xmath101 ( resp .
@xmath102 ) of semistable invertible ( resp .
relatively torsion free rank 1 ) sheaves . as a particular case of the simpson s work @xcite
, there exists a projective scheme @xmath103 which coarsely represents the functor @xmath102 .
rational points of @xmath104 correspond to @xmath32-equivalence classes of semistable torsion free sheaves of rank 1 and degree @xmath1 on a fiber @xmath105 ( @xmath106 ) .
moreover , @xmath101 is coarsely represented by a subscheme @xmath107 and there are open subschemes @xmath108 and @xmath109 which represent the other two funtors .
the simpson jacobian of @xmath0 is @xmath110 .
we denote @xmath111 and @xmath112 .
when @xmath0 is an integral curve every torsion free rank 1 sheaf is stable , and then @xmath113 is equal to the picard scheme @xmath114 and @xmath115 coincides with the altman - kleiman s compactification @xcite .
a generalized tree - like curve is a projective , reduced and connected curve @xmath19 over @xmath41 such that the intersection points , @xmath116 , of its irreducible components are disconnecting ordinary double points .
henceforth we shall assume that @xmath0 is a generalized tree - like curve , and then @xmath117 .
[ lem4 ] it is possible to order the irreducible components @xmath76 of the curve @xmath0 , so that for every @xmath22 all but one of the connected components of @xmath118 consists entirely of irreducible components with subindices smaller than @xmath119 .
then there are irreducible components , say @xmath120 , with all subindices smaller than @xmath119 , such that @xmath121 is connected and intersects its complement @xmath122 in @xmath0 in just one point @xmath13 .
teixidor proves this lemma in ( @xcite , lem .
1 ) when @xmath0 is a tree - like curve , that is , its irreducible components are smooth , but her proof is valid for our curve .
let us suppose from now on that an ordering of the components of @xmath0 as in lemma [ lem4 ] , has been fixed .
* notation : * if @xmath123 denote the arithmetic genus of @xmath0 , that is , the dimension of @xmath124 , for any torsion free rank 1 sheaf @xmath30 en @xmath0 of degree @xmath1 , let @xmath18 be the residue class of @xmath125 modulo @xmath16 so that @xmath126 for every proper subcurve @xmath53 of @xmath0 , we shall write @xmath127 lemma [ lem4 ] allows us to define inductively integer numbers @xmath25 as follows : @xmath128 + 1-d_{i_1}^x-\hdots - d_{i_k}^x,\ \text { for } i=1,\hdots , n-1 \notag\\ & d_n^x = d - d_1^x-\hdots - d_{n-1}^x.\end{aligned}\ ] ] we are now going to modify the above numbers to obtain new numbers @xmath3 associated with @xmath0 .
this is accomplished by a recurrent algorithm . in order to describe it we start by saying that a connected subcurve @xmath129 , @xmath130 , of @xmath0 ordered according to lemma [ lem4 ]
is _ final _ either when the numbers @xmath131 are not integers for @xmath132 or @xmath53 is irreducible .
* algorithm : * if the curve @xmath0 is final , @xmath139 for all @xmath119 .
otherwise , let @xmath119 be the first index for which @xmath140 .
we consider the two connected components , @xmath141 and @xmath142 , of @xmath143 and we reorder them according with lemma [ lem4 ] .
this induces a new ordering @xmath144 ( resp .
@xmath145 ) of the points @xmath146 ( @xmath147 ) in @xmath148 ( resp .
@xmath149 ) . then
, \b ) if @xmath148 is not final , we take the first index @xmath150 of @xmath148 for which @xmath151 , consider the connected components , @xmath152 and @xmath153 , of @xmath154 and reorder them according with lemma [ lem4 ] .
if @xmath152 and @xmath153 are final , the process finishes for @xmath148 . otherwise , we iterate the above argument for those components that are not final and so on .
the process finishes for @xmath148 when all subcurves that we find are final .
\c ) if @xmath149 is not final , we take the first index @xmath155 of @xmath149 such that @xmath156 , consider the connected components , @xmath157 and @xmath158 , of @xmath159 and reorder them according with lemma [ lem4 ] .
if @xmath157 and @xmath158 are final , the process finishes for @xmath149 . otherwise , we repeat the above argument for those components that are not final and so on .
the process finishes for @xmath149 when all subcurves that we obtain are final .
[ lem5 ] let @xmath164 be a line bundle on @xmath0 of degree @xmath1 .
if @xmath164 is stable , then @xmath21 is not an integer for every @xmath22 and @xmath164 is obtained by gluing invertible sheaves @xmath165 on @xmath4 of degrees @xmath25 , @xmath166 .
let us consider the subsheaves @xmath167 of @xmath164 , @xmath15 , @xmath12 being the subcurves of @xmath0 given by lemma [ lem4 ] .
by the stability of @xmath164 , we get @xmath168 considering the subsheaves @xmath169 of @xmath164 , @xmath15 , yields @xmath170 since @xmath171 and @xmath12 , @xmath172 meet only at @xmath13 , we have @xmath173 , @xmath174 and @xmath175 . then , and give @xmath176 we have @xmath177 so that @xmath178 is an integer .
then , if @xmath140 for some @xmath22 , becomes a contradiction .
thus @xmath29 for all @xmath179 and there is only one possibility for @xmath180 , namely @xmath181 + 1,\ \ \
\text { for } i=1,\hdots , n-1.\ ] ] from @xmath182 and the exact sequence @xmath183 we deduce that @xmath184 for all @xmath119 and the proof is complete .
[ lem6 ] let @xmath164 be an invertible sheaf of degree @xmath185 obtained by gluing line bundles @xmath165 on @xmath4 of degrees @xmath25 , @xmath166 .
suppose that @xmath21 is not integer for every @xmath22 and let @xmath186 be a proper subcurve of @xmath0 .
then , \1 .
if @xmath53 is a connected subcurve of @xmath0 , let us denote by @xmath189 , @xmath190 , ( resp .
@xmath191 ) the component of @xmath53 ( resp .
@xmath69 ) which contains the point @xmath192 ( it is possible to have @xmath193 for @xmath194 ) .
let us see that @xmath198 . since the inclusion @xmath199 is clear we have only to prove that @xmath200 contains no component of @xmath53 and that @xmath201 if @xmath202 .
if @xmath203 for some @xmath204 , then @xmath205 and @xmath206 . hence , @xmath207 where we denote by @xmath208 ( resp .
@xmath209 ) the connected component of @xmath210 that contains @xmath211 ( resp .
@xmath212 ) .
analogously , @xmath191 , @xmath189 are contained in @xmath213 and @xmath214 where @xmath215 ( resp .
@xmath216 ) is the connected component of @xmath217 containing @xmath189 ( resp .
@xmath191 ) .
we deduce that @xmath218 , @xmath219 and @xmath220 .
thus , @xmath221 and @xmath222 , which is absurd . therefore
, none of the components @xmath211 of @xmath53 is in @xmath223 and , since @xmath196 are connected , no other component of @xmath53 is either .
moreover , a similar argument shows that @xmath196 and @xmath224 have no common components for @xmath202 .
let now @xmath225 be the number of irreducible components in @xmath196 . then , @xmath226 is equal to @xmath227 . on the other hand , the number of intersection points in @xmath228 is @xmath229 , where @xmath155 is the number of intersection points of @xmath196 , @xmath190 .
since @xmath229 is the number @xmath230 of intersection points in @xmath69 , we get @xmath231 and @xmath198
. then , @xmath232 , @xmath233 and @xmath234 assume now that @xmath240 .
then , @xmath241 is the connected component of @xmath242 which contains @xmath243 .
since @xmath53 is connected , @xmath241 contains @xmath53 and then @xmath241 contains @xmath189 and @xmath191 for @xmath244 .
this implies that @xmath245 for @xmath244 .
if we had @xmath246 for some @xmath244 , then @xmath53 would be contained in @xmath247 as before and @xmath248 , which is absurd .
thus , @xmath195 for @xmath244 .
in this situation , one has that @xmath249 , where @xmath250 and @xmath53 intersects @xmath152 in @xmath251 points .
we prove now the remaining statements of the theorem .
if @xmath164 is a strictly semistable line bundle on @xmath0 of degree @xmath1 then , @xmath257 is equal to one of the two extremal values of the inequality . in particular
, @xmath21 is an integer for some @xmath22 .
let us construct a jordan - hlder filtration for @xmath164 in both cases . since case a ) and case b ) are the same but with the roles of @xmath12 and @xmath172 intertwined , we give the construction in the case a ) .
since the residue class of @xmath274 modulo @xmath275 is @xmath276 , the numbers @xmath277 are not integers for @xmath272
. then , from lemma [ lem6 ] , we have only to prove that @xmath268 is in @xmath278 , where @xmath279 are the integer numbers defined as @xmath25 but with the new ordering of @xmath148 and @xmath150 runs through the irreducible components of @xmath148 .
this is equivalent to proving that @xmath280 + 1\ \ \
\text { for } r=\sigma(i_0),\hdots,\sigma(i_{k-1}).\ ] ] actually , since @xmath164 is semistable and @xmath271 is a proper subcurve of @xmath0 , arguing as in lemma [ lem6 ] , we obtain @xmath281 where @xmath39 is the number of intersection points of @xmath271 and its complement in @xmath0 .
we have that @xmath282 and @xmath283 is not equal to the extremal values of because @xmath284 .
moreover , if it were @xmath280 + 2,\ ] ] since @xmath285 , @xmath286 and @xmath287 , @xmath288 being the complement of @xmath271 in @xmath148 , then @xmath289\ ] ] which contradicts the semistability of @xmath164 .
thus , @xmath280 + 1\ ] ] and the proof of the claim 1 is complete . on the other hand ,
the irreducible components of @xmath152 are ordered according the instructions in lemma [ lem4 ] and the subcurves @xmath290 , where @xmath155 runs through the irreducible components of @xmath152 and @xmath291 , are equal to either @xmath105 or @xmath292 . since the residue class of @xmath295 modulo @xmath296 is @xmath297 and @xmath298 , by the hypothesis , the numbers @xmath299 are not integers for @xmath300 and , by the choice of @xmath119 , they are nt for @xmath301 either .
then , by lemma [ lem6 ] , it is enough to prove that @xmath302 + 1\ \ \
\text { for } s\leq n-1.\ ] ] since @xmath164 is semistable , we have that @xmath303 + 1.\ ] ] moreover , if @xmath304 then , @xmath305 and if @xmath306 then , @xmath307 . we obtain the desired result in both cases and the proof of the claim 2 is complete .
we return now to the proof of the theorem .
if @xmath273 and @xmath293 are not integers for @xmath272 and @xmath294 , then @xmath308 is a jordan - hlder filtration for @xmath164 and the @xmath32-equivalence class of @xmath164 belongs to @xmath309 . on the other hand , if @xmath273 is integer for some @xmath272 , the sheaf @xmath268 is strictly semistable and we have to repeat the above procedure with @xmath268 in the place of @xmath164 and the curve @xmath148 in the place of @xmath0 .
similarly , if @xmath293 is integer for some @xmath294 , the sheaf @xmath269 is strictly semistable .
then , we have to repeat the above procedure for @xmath269 . by iterating this procedure
, we get a jordan - hlder filtration for @xmath268 : @xmath310 and another for @xmath269 : @xmath311 therefore , a filtration for @xmath164 is given by @xmath312 thus , the @xmath32-equivalence class of @xmath164 belongs to @xmath313 , where @xmath3 are the integer numbers constructed with the algorithm .
finally , let us consider a torsion free sheaf @xmath30 on @xmath0 of rank 1 and degree @xmath1 which is not locally free .
when @xmath30 is locally free at the intersection points @xmath13 for all @xmath15 , calculations and results are analogous to the former ones .
if @xmath30 is not locally free at @xmath13 for some @xmath15 , then there is a natural morphism @xmath314 where @xmath148 , @xmath152 are the connected components of @xmath315 , that is clearly an isomorphism outside @xmath13 . but
this is an isomorphism at @xmath13 as well because by lemma [ lem3 ] , @xmath316 and this is precisely the stalk of @xmath317 at @xmath13 .
we conclude that if @xmath30 is strictly semistable , then @xmath318 y @xmath319 are integers , @xmath320 and @xmath321 are given by @xmath322 and @xmath323 and @xmath324 are semistable as well .
then , the construction of a jordan - hlder filtration for @xmath30 can be done as above and thus the @xmath32-equivalence class of @xmath30 belongs to @xmath33 .
* example 1 .
* let @xmath19 , @xmath160 , be a generalized tree - like curve with a polarization @xmath2 whose degree @xmath16 is a prime number .
suppose that the irreducible components of @xmath0 are ordered according with lemma [ lem4 ] .
then , since @xmath17 is not divisible by @xmath16 , @xmath14 is an integer if and only if @xmath325 .
therefore , by theorem [ t ] , we have that if @xmath326 , @xmath327 and @xmath328 whereas for @xmath325 , @xmath7 and @xmath26 are empty and @xmath329 moreover , in this case the number @xmath3 is given by @xmath330 actually , if @xmath164 is a strictly semistable line bundle of degree @xmath1 on @xmath0 , the first index @xmath119 such that @xmath21 is integer is @xmath331 and the connected components of @xmath242 are @xmath332 and @xmath333 .
suppose , as in the proof of the theorem , that @xmath334 then , @xmath268 and @xmath335 are semistable and @xmath336 . on the other hand , since @xmath293 is integer for @xmath337
, we have to apply the procedure to the curve @xmath152 . here , the first index @xmath155 such that @xmath293 is integer is @xmath338 and the connected components of @xmath339 are @xmath340 and @xmath341 . proceeding exactly as in case of @xmath0
, we obtain a sheaf belonging to @xmath342 and another supported on @xmath341 . since @xmath293 is integer for @xmath343 , we apply the procedure to the curve @xmath341 , and so on .
the iteration of this procedure will only finish when we obtain a sheaf supported on @xmath344 that belongs to @xmath345 .
thus , @xmath346 for @xmath166 .
* example 2 .
* we are now going to recover examples 1 and 2 of @xcite .
there , the irreducible components @xmath4 are taken to be smooth and @xmath347 .
then , the residue class of @xmath125 modulo @xmath16 is @xmath325 and @xmath348 .
it follows that @xmath21 is integer for @xmath166 and , arguing as in our example 1 , we have that @xmath349 and @xmath350 are empty and @xmath351 as asserted in @xcite .
the connected components of @xmath357 are @xmath358 and @xmath359 .
suppose @xmath164 is a strictly semistable line bundle of degree @xmath1 on @xmath0 and that @xmath360 ( the other case is similar ) .
then @xmath268 and @xmath361 are semistable . moreover , @xmath362 so that @xmath363 we now have to fix a new ordering for @xmath148 according with lemma [ lem4 ] .
we can take , for instance , @xmath364 with @xmath365 , @xmath366 and @xmath367 .
then , @xmath368 and @xmath369 . therefore , since @xmath370 and @xmath371 are not integers
, we conclude that @xmath372 + 1=d^x_1,\notag\\ d_3&=d_{\sigma(2)}^y=-\chi(\mathcal{o}_{y_{\sigma(2)}})+h_{y_{\sigma(2)}}t+ [ k_{y_{\sigma(2)}}]+1-d^y_{\sigma(1)}=\notag\\ & = -\chi(\mathcal{o}_y)+h_yt+k_y+1-d^x_1-d^x_2 - 1=d^x_3 - 1,\notag\\ d_2&=d_{\sigma(3)}^y = d_y - d_{\sigma(1)}^y - d_{\sigma(2)}^y = d^x_2.\notag\end{aligned}\ ] ] | the compactified jacobian of any projective curve @xmath0 is defined as the simpson moduli space of torsion free rank one degree @xmath1 sheaves that are semistable with respect to a fixed polarization @xmath2 on @xmath0 . in this paper
we give explicitly the structure of this compactified simpson jacobian in the case where @xmath0 is a generalized tree - like curve , i.e. , a projective , reduced and connected curve such that the intersection points of its irreducible components are disconnecting ordinary double points .
we prove that it is isomorphic to the product of the compactified jacobians of a certain degree @xmath3 of its components @xmath4 , where the degrees @xmath3 depend on @xmath1 , @xmath2 and on the particular structure of the curve .
we find also necessary and sufficient conditions for the existence of stable points which allow us to study the variation of these simpson jacobians as the polarization @xmath2 changes . |
in hard x - ray ( hxr ) emission of many flares quasi - periodic variations were observed with time - intervals between pulses , @xmath3s [ see ; see also review of and references therein ] . in our previous paper ( , paper i )
we attempted to investigate relationship between hxr loop - top ( lt ) sources and the quasi - periodic variations .
main difficulty was that sequences of pulses are usually short , so that it is difficult to carry out comprehensive analysis of their quasi - periodicity . therefore we used time - interval , @xmath4 , between the strongest pulses as a simple estimate of the quasi - period @xmath5 . in the present paper
we have selected four flares which have longer sequences of hxr pulses , so that it was possible to carry out detailed analysis of their quasi - periodicity ( section [ qper ] ) .
section [ 16jan ] contains detailed analysis of 16 january 1994 flare . in section [ det ]
we estimate values of electron density and magnetic field strength inside several hxr loop - top sources .
section [ sum ] contains discussion and conclusions .
we used hxr observations recorded by _ yohkoh _ hard x - ray telescope , hxt , @xcite [ light - curves and images ] and _ compton gamma ray observatory _ burst and transient source experiment , batse , @xcite [ light - curves ] .
we have selected four flares which have longer sequences of hxr pulses , so that detailed analysis of their quasi - periodicity was possible .
main difficulty in the analysis was the fact that during impulsive phase the pulses occur simultaneously with quick increase of total hxr intensity ( see figure [ 93may14pgm]a ) .
therefore we have applied a method of normalization which is commonly used by radioastronomers ( see ) .
a normalized time series , @xmath6 , is : @xmath7 where @xmath8 is the measured hxr flux and @xmath9 is a running average of @xmath8 .
the red line in figure [ 93may14pgm]a shows @xmath9 calculated with averaging time @xmath10s . the normalized time series , @xmath6 is shown in figure [ 93may14pgm]b . quasi - periodicity of the pulses is clearly shown in this figure and it is confirmed by power spectrum seen in figure [ 93may14pgm]c ( fourier transform of @xmath6 ) . the power spectrum has been calculated for time - interval 22:01 - 22:05:20 ut ( impulsive phase , ip ) .
s. ( b ) normalized light - curve , @xmath6 [ see equation ( 1 ) ] .
vertical dashed lines show the time - interval which was used to calculate power spectrum .
( c ) power spectrum calculated for the normalized light - curve , @xmath6 .
@xmath5 is the period corresponding to the peak in power spectrum.,scaledwidth=100.0% ] we see in figures [ 93may14pgm]a and [ 93may14pgm]b that before the impulsive phase , between 21:58:40 and 22:00:20 ut , three increasing pulses occurred .
mean time - interval between the pulses is about 31s which is close to the quasi - period , @xmath5 , seen in the power spectrum ( 28s ) .
this suggests that these pulses also belong to the quasi - periodic sequence seen during impulsive phase .
figure [ 94jan16lc ] shows a hxr light - curve of 16 january 1994 flare .
like in figure [ 93may14pgm]a , during impulsive phase we see a sequence of increasing pulses .
the red line connects tops of the pulses to show their quick non - linear increase and saturation . before the impulsive phase ( bip )
there is a sequence of weak pulses .
normalization and power - spectrum analysis is shown in figure [ 94jan16pgm ] . the power spectrum in figure
[ 94jan16pgm]c has been calculated for time interval a , i.e. bip plus ip .
figure [ 94jan16pgm]d shows power spectrum calculated for the decay phase ( time - interval b ) , where the pulses were weak and their profiles were disturbed , but the quasi - period had been retained .
figure [ 98aug18lc ] shows hxr light - curve of 18 august 1998 flare .
its characteristic feature is that quasi - period during increase of hxr emission is shorter than during decrease of the emission .
therefore analysis of the light - curve has been done separately for time - intervals a and b ( see figures [ 98aug18pgma ] and [ 98aug18pgmb ] ) . in our model of oscillating magnetic traps changes of quasi - period are explained as being due to changes of length of oscillating traps ( see section 3 ) .
figure [ 93mar12lc ] shows hxr light - curves of 12 march 1993 flare .
general behavior is similar to that seen in figure 2
: there are low - amplitude oscillations before impulsive phase and quick increase of hxr emission after the onset of impulsive phase . specific feature are short - period oscillations after hxr maximum . therefore analysis of the light - curve has been done separately for time - intervals a and b ( see figures [ 93mar12pgma ] and [ 93mar12pgmb ] ) .
change of the quasi - period is clearly seen .
we have chosen the flare of 16 january 1994 for further analysis because loop - top and footpoint sources can be easily recognized in its hxr images ( figure [ mozaic ] ) .
we wanted to investigate what happened in the loop - top source during transition from low - amplitude oscillations to quick increase of hxr emission seen in figures [ 94jan16lc ] and [ 94jan16pgm ] . toward this end
we have reconstructed _
yohkoh_/hxr images in energy channels l ( 14 - 23 kev ) , m1 ( 23 - 33 kev ) , and m2 ( 33 - 53 kev ) for the time of before impulsive phase ( bip ) oscillations ( upper row in figure [ mozaic ] ) , about the beginning of impulsive phase ( middle row ) and for maximum of ip ( lower row ) .
we see that before impulsive phase hxr footpoints are strong which means that accelerated electrons easily escape from the lt source .
but at the beginning of impulsive phase the footpoints are weaker than the loop - top source ( f @xmath11 lt ) in 14 - 23 and 23 - 33 kev emission which indicates that most of accelerated electrons deposit their energy within the lt source .
next we have determined mean temperature , @xmath12 , emission measure , @xmath13 , and mean electron number density , @xmath14 , in the hxr lt source , from its _ yohkoh _ soft - x - ray ( sxr ) images , using filter - ratio method ( from be119 and al12 images ) .
this analysis has been done in the following way : we have integrated sxr fluxes from area @xmath15 _ yohkoh _ pixels [ i.e. 4.9 @xmath16 4.9 ( arcsec)@xmath17 at the center of hxr lt source .
next we determined diameter , @xmath18 , of the hxr source according to isocontour @xmath19 , where @xmath20 is the maximum intensity within the source .
we assumed that extension of the hxr source along the line of sight is also @xmath18 and calculated the volume of the emitting plasma as @xmath21 ( here 1.15 is a correction factor which takes into account that actual size of the hxr source is somewhat larger than that determined from isocontour @xmath19 ) .
mean electron number density was calculated as @xmath22 and obtained time - variation of the temperature , @xmath23 , and density , @xmath24 , are shown in figures [ temp ] and [ nel ] .
figure [ temp ] shows that increase of energy release in the lt source began already about 23:10:40 ut .
sharp peaks of temperature about 23:12:30 and 23:16 ut ( figure [ temp ] ) are correlated with peaks in hxrs ( figure [ 94jan16pgm]b ) which confirms that significant part of the energy of accelerated electrons is deposited within the loop - top source ( the temperature peaks are somewhat , @xmath2516 s , delayed relative to the hxr peaks which is due to accumulation of plasma heating by accelerated electrons ) .
significant random fluctuations of @xmath24 in figure [ nel ] are mostly due to random errors in estimates of the source volume from individual hxr images .
systematic increase of density with time seen in figure [ nel ] is due to chromospheric evaporation flow .
x - ray images for a flare of 30 october 1992 are shown in figure [ 92oct30img ] .
we see similar behavior like in figure [ mozaic ] : footpoint sources were stronger than lt source before impulsive phase and the lt source was stronger during impulsive phase . in paper i we used @xmath26 @xmath1 as typical value of the electron number density in hxr lt sources during impulsive phase .
this value has been obtained by from _ rhessi _ soft - x - ray images of the sources . in paper
i we used this value of @xmath14 to estimate magnetic field strength in oscillating magnetic traps : @xmath27 where @xmath28 is the wave speed estimated from analysis of hxr oscillations and @xmath29 , @xmath30 is the mass of proton . obtained values of @xmath31 are given in table [ tab1 ] . for eight flares of those investigated in paper
i it was possible to determine mean electron density , @xmath14 , for hxr lt source from sxr images using the method described in section [ 16jan ] .
the values are given in table [ tab1 ] .
we see that most values are significantly higher than that assumed in paper i ( @xmath32 @xmath1 ) .
this indicates that the value of @xmath14 used in paper i , and therefore also magnetic field strength , @xmath31 , were underestimated .
we have calculated corrected values , @xmath33 , of the magnetic field strength using equation ( 2 ) with the new values of @xmath14 ( see table [ tab1 ] ) .
let us note that values of electron number density , @xmath14 , derived from _
yohkoh_/ sxt images are reliable , since : 1 . they have been confirmed by independent method ( see ) , 2 .
determination of @xmath14 from _ yohkoh_/al12 images does not depend significantly on temperature estimates , since instrumental response function for the al12 images is nearly constant in wide range of temperatures ( 7 - 40 mk see curve _ f _ in figure 9 of ) .
.values of mean electron density and magnetic field strength [ cols="^,^,^,^,^",options="header " , ] magnetic field strength estimated in paper i ( assuming @xmath34 @xmath1 ) mean electron number density in hxr lt source at hxr maximum , estimated in the present paper magnetic field strength calculated with the new value of @xmath14
in paper i we have proposed a model of oscillating magnetic traps to explain quasi - periodic oscillations seen in hxrs . in this model
we have assumed that : the cusp - like structure is now a standard model of solar flare magnetic field see . at the top of this structure ( near p in figure [ scheme ] ) magnetic
reconnection occurs .
reconnected magnetic field and plasma flow into the confined volume bpc with high velocity .
hence , it seems to be obvious that , this should excite magnetosonic oscillations within the bpc volume .
magnetic fields pb and pc reconnect at p ( figure [ scheme]a ) .
this generates a sequence of magnetic traps which move downward , the traps overtake each other , they collide and undergo compression ( figure [ scheme]b ) . during the compression particles
are accelerated within the traps , magnetic pressure , gas pressure , and the pressure of accelerated particles increase , so that the compression is stopped , the traps can expand and undergo magnetosonic oscillations . during the compression parameters of the traps undergo strong changes . at the beginning of compression
the trap ratio , @xmath35 , is high ( @xmath36 is the magnetic field strength at magnetic mirrors and @xmath37 is the strength at the middle of the trap ) .
the trap ratio decreases during compression and it reaches the lowest value , @xmath38 , at the end of compression . then the electrons reach highest energies and they can most easily escape from the trap . plasma density determines the fraction of accelerated electrons which lose their energy within the trap , i.e. before they escape from the loop - top source .
this fraction depends on the energy , since coulomb losses depend on energy ( see changes of the footpoint / loop - top intensity ratio along rows and along columns in figure [ mozaic ] ) .
distribution of the oscillations within the bpc volume is determined by the time - profile , @xmath39 , of the velocity of reconnection flow which excites the oscillations .
its central part ( @xmath40 ) determines a `` main trap '' which contains many electrons and which is responsible for strong hxr pulses .
wings of the velocity profile ( @xmath41 ) are responsible for excitation of oscillations of many other ( `` secondary '' ) traps whose oscillations are shifted in phase .
superposition of many hxr pulses coming from these secondary traps gives a `` quasi - smooth '' emission ( seen as the emission below the pulses in figures [ 93may14pgm]-[93mar12pgmb ] ) .
we see that this quasi - smooth emission is stronger than pulses in the investigated flares which indicates that most of the volume bpc was filled with the `` secondary '' traps ( their volume was greater than the volume of main trap ) . in many flares hxr
pulses are not clearly seen during quick increase of hxr emission , i.e. during beginning of impulsive phase ( see example in figure [ 93mar12pgmb ] ) . according to our model of oscillating magnetic traps , in such cases
the time - profile , @xmath39 of reconnection flow is quasi - rectangular , i.e. its maximum is flat .
therefore , it excites many oscillating traps of similar power , which have different length ( see figure [ scheme]b ) and their oscillations are shifted in phase .
this gives strong quasi - smooth hxr emission .
clear quasi - periodic pulses are seen only near maximum of hxr emission ( see figure [ 93mar12pgmb ] ) . 1 .
the oscillations have been excited by a short pulse of reconnection flow ( otherwise the oscillations would be more chaotic ) .
2 . the oscillations are self - maintained , i.e. some feedback mechanism operates which causes that the oscillations do not decay , but they increase during impulsive phase
. investigated hxr lt sources from _ rhessi _ observations and they have found that the energy contained in non - thermal electrons is usually higher than thermal energy .
this means that also the pressure of non - thermal electrons is higher than gas pressure , @xmath42 .
this suggests that the pressure , @xmath43 , of non - thermal electrons is an important factor in the feedback mechanism which maintains the oscillations of magnetic traps . during compression of a magnetic trap
the pressure @xmath43 steeply increases and this causes increase of the amplitude of next expansion of the trap .
this , in turn , causes increase of restoring force ( i.e. the tension of bent magnetic field lines ) and therefore the pulse of the pressure @xmath43 will be stronger during the next compression . hence , there is a feedback between the intensity of pulses of the pressure @xmath43 and the amplitude of the magnetic trap oscillation and this feedback is responsible for quick increase of the oscillations during impulsive phase
. it may be interesting to note that this mechanism of maintaining the oscillations is analogous to the mechanism of instability which is responsible for stellar pulsations ( `` non - adiabatic pulsations '' ) : in both cases the point is that additional energy is accumulated during compression of the gas ( it is the energy of helium ionization in the case of cepheids and it is energy of non - thermal electrons in our case ) . 1 .
it has been confirmed that quasi - periodic oscillations ( qpo ) occur in hxr emission of solar flares .
2 . we have found that low - amplitude qpo occur before impulsive phase of some flares .
3 . we have found that quasi - period of the oscillations can change in some flares .
we interpret this as being due to changes of the length of oscillating magnetic traps .
4 . during impulsive phase
most of the energy of accelerated ( non - thermal ) electrons is deposited within the hxr loop - top sources ( section [ 16jan ] ) . [ for weak hxr flares this is seen at lower energies ( 14 - 23 kev ) , but not at 23 - 33 kev . ]
we argue that the basic properties of the hxr oscillations can be explained in terms of a simple model of oscillating magnetic traps ( see paper i ) .
this model allows us also to explain large number of electrons which are accelerated during impulsive phase : observations show that the amplitude of the oscillations quickly increases and therefore the traps are filled with increasing amount of plasma coming from chromospheric evaporation .
this is main source of electrons which undergo acceleration .
we suggest that a feedback between the pressure of accelerated electrons and the amplitude of the following expansion of magnetic trap is the mechanism which causes the quick increase of the amplitude of oscillations . 7 . we have also determined improved values of electron number density and magnetic field strength for hxr loop - top sources of several flares which were investigated in paper i. obtained values fall within the limits of @xmath0 @xmath1 , @xmath2 gauss .
1 . acceleration of electrons occurs in a large volume ( bpc in figure [ scheme ] ) .
2 . during development of impulsive phase plasma
is delivered into bpc volume by chromospheric evaporation flow whose density is higher than the density of surrounding corona .
this explains why number of accelerated electrons is very high .
the _ yohkoh _ satellite is a project of the institute of space and astronautical science of japan .
compton gamma ray observatory _ is a project of nasa .
we would like to thank anonymous referee for valuable remarks which helped us to improve this paper .
this work was supported by polish ministry of science and high education grant no .
nn2031937 33 .
fishman , g.j .
, meegan , c.a . ,
wilson , r.b . ,
paciesas , w.s . , & pendleton , g.n . : 1992 , in the compton observatory science workshop ( nasa cp 3137 ) , ed .
shrader , n. gehrels , & b. dennis ( grenbelt : gsfc ) , 26 . | in our recent paper ( solar physics * 261 * , 233 ) we investigated quasi - periodic oscillations of hard x - rays during impulsive phase of solar flares .
we have come to conclusion that they are caused by magnetosonic oscillations of magnetic traps within the volume of hard - x - ray ( hxr ) loop - top sources . in the present paper
we investigate four flares which show clear quasi - periodic sequences of hxr pulses .
we also describe our phenomenological model of oscillating magnetic traps to show that it can explain observed properties of hxr oscillations .
main results are the following : 1 . we have found that low - amplitude quasi - periodic oscillations occur before impulsive phase of some flares .
2 .
we have found that quasi - period of the oscillations can change in some flares .
we interpret this as being due to changes of the length of oscillating magnetic traps .
3 . during impulsive phase a significant part of the energy of accelerated ( non - thermal ) electrons
is deposited within the hxr loop - top source .
4 .
our analysis suggests that quick development of impulsive phase is due to feedback between pulses of the pressure of accelerated electrons and the amplitude of magnetic - trap oscillation .
5 .
we have also determined electron number density and magnetic filed strength for hxr loop - top sources of several flares .
the values fall within the limits of @xmath0 @xmath1 , @xmath2 gauss . |
when considering quantum field theories on non - commutative spaces , e.g. by employing the so - called groenewold - moyal star product , one inevitably has to deal with the infamous uv / ir mixing problem ( see @xcite for reviews of the topic ) : new kinds of non - local ir divergences prevent the model from being renormalizable .
in fact , so far only some modified scalar field theories on euclidean non - commutative spaces have been found to be renormalizable by adding new types of terms in the action .
the first successful approach was introduced by grosse and wulkenhaar @xcite , and proofs of renormalizability have been achieved mainly by utilizing multiscale analysis @xcite , or formally in the matrix base @xcite . recently , and quite independent of former developments , gurau _ et al . _
@xcite introduced a term of the type @xmath1 into the lagrangian which modifies the theory in the infrared region and , in this way , renders it renormalizable .
this was in fact proven to all orders by the authors using multiscale analysis .
a thorough study of the divergence structure of this model ( referred to as @xmath2-model ) , including explicit renormalization at one - loop level @xcite has been carried out as well as a computation of the beta functions @xcite .
the task of constructing a renormalizable non - commutative gauge theory is even more involved , and although there have been several interesting ideas using additional constraints @xcite , or generalizing existing scalar models @xcite , proofs of renormalizability are still missing .
motivated by the inherent translation invariance and simplicity of the scalar @xmath2-model , a possible generalization to gauge theories was introduced in @xcite and further discussed in @xcite .
the main idea was to include a gauge invariant non - local term in the action @xmath3 where the field strength tensor @xmath4 associated to the gauge field @xmath5 and the covariant derivative @xmath6 are defined by @xmath7 } , \nonumber\\ & d_{{\mu}}\cdot = \partial_{\mu}\cdot -{{\rm i}g}{\left [ a_{\mu}\stackrel{\star}{,}\cdot\right ] } , \qquad \text{and}\quad { \varepsilon}{\widetilde{d}}_{{\mu}}={\varepsilon}{\theta}_{{\mu}{\nu}}d_{{\nu}}\,.\end{aligned}\ ] ] this model is formulated on euclidean @xmath8 with the moyal - deformed product @xmath9 } \equiv x_{\mu } \star x_{\nu } -x_{\nu } \star x_{\mu } = { { \rm i}}{\varepsilon}{\theta}_{\mu \nu}\,,\end{aligned}\ ] ] of regular commuting coordinates @xmath10 .
the real parameter @xmath11 has mass dimension -2 , rendering the constant antisymmetric matrix @xmath12 dimensionless . in eqn .
( [ eq : action_nloc ] ) , the non - local second term implements an ir damping mechanism similar to the one of the corresponding scalar model , but involves an infinite number of vertices in the action @xcite .
therefore , two alternatives of localizing that term by introducing additional fields were presented in @xcite .
initially the intent was to apply the technique of algebraic renormalization , however it turned out that this scheme may not be applicable to non - commutative spaces , as has been discussed recently in ref .
the reason for these obstacles is that arbitrary powers of the gauge invariant dimensionless operators @xmath13 and @xmath14 may appear as counter terms in the effective @xmath15-loop action .
in addition , these models suffer from a high degree of complexity due to a large number of ( ghosts and ) fields , being introduced in order to implement the demanded symmetries .
finally , due to potentially divergent contributions by auxiliary fields , it is not clear if the models are at all renormalizable .
the aim of the current work is to put forward an alternative action which implements the same - model . ] damping mechanism in the gauge field propagator in such a way , that the problems of the former models do not appear . in section
[ sec : def_newmodel ] , we present the action of our new model and its ( symmetry- ) properties .
we then discuss one - loop calculations including the respective renormalized propagator in section [ sec : one - loop ] before explaining why we expect no ir problems in higher loop graphs in section [ sec : higher - loops ] . in the following
we will use the abbreviations @xmath16 for vectors @xmath17 and @xmath18 for matrices @xmath19 . for the deformation
, we furthermore consider the simplest block - diagonal form @xmath20 for the dimensionless matrix describing non - commutativity .
bearing in mind the problems of the previous approaches @xcite to the generalization of the scalar @xmath2 model to gauge theories , we now attempt to make a different ansatz . as a starting point
we have to take into account that our action should meet the following requirements : * the tree level action should provide a counter term for the quadratic ir divergence in the external momentum @xmath21 @xmath22 which appears at one - loop level in all versions of non - commutative gauge theories on @xmath23-deformed spaces .
* all relevant propagators should be infrared finite and feature damping factors similar to the scalar @xmath24 model @xcite . in this way , ir damping at higher loop orders is implemented without breaking translation invariance . *
any auxiliary fields and related additional ghosts should be decoupled from the gauge sector , i.e. no according interactions should appear so that the physical content is not altered compared to traditional implementations of nc yang - mills theory see e.g. @xcite .
* the model should be as simple as possible .
the present idea is to discard the original approach of adding the non - local term of eqn .
( [ eq : action_nloc ] ) , and instead to implement the damping entirely within a so - called ` soft - breaking ' term a method well known from the gribov - zwanziger action @xcite in qcd .
let us discuss the most important steps leading to our new model .
the starting point is the action proposed in ref .
@xcite , which has been introduced in order to localize the operator @xmath25 appearing in eqn .
( [ eq : action_nloc ] ) .
that action consists of the usual yang - mills term @xmath26 plus the terms of below .
the one proportional to @xmath27 represents the gribov - zwanziger soft breaking , and the second term implements the coupling between auxiliary fields and the gauge sector .
our new proposal is derived in two steps : [ eq : brss_constr_evol ] @xmath28\label{eq : constr_step0}\,,\\ \phantom{\text{step 1}}\downarrow\text{step 1}\notag\\ { \int\!\ ! { \rm d}^4x}\left[\frac{\gamma^3}{2}\left(b_{{\mu}{\nu}}+{\bar{b}}_{{\mu}{\nu}}\right){\frac{1}{{\widetilde{\square}}}}f_{{\mu}{\nu}}-{\bar{b}}_{{\mu}{\nu}}d^2 b_{{\mu}{\nu}}\right]\label{eq : constr_step1}\,,\\ \phantom{\text{step 2}}\downarrow\text{step 2}\notag\\ { \int\!\ ! { \rm d}^4x}\left[\frac{\gamma^2}{2}\left(b_{{\mu}{\nu}}+{\bar{b}}_{{\mu}{\nu}}\right){\frac{1}{{\widetilde{\square}}}}\left(f_{{\mu}{\nu}}+{\sigma}\frac{{\theta}_{{\mu}{\nu}}}{2}\tilde{f}\right)-{\bar{b}}_{{\mu}{\nu}}b_{{\mu}{\nu}}\right]\,,\label{eq : constr_step2}\,\end{gathered}\ ] ] with several new definitions to be explained subsequently . to understand the first step we note that the divergences in the @xmath29 , @xmath30 , @xmath31 , and @xmath32 propagators ( see ref .
@xcite ) are mainly caused by the appearance of the operator @xmath33 sandwiched between @xmath34 and @xmath35 .
on the other hand this term is crucial to the construction of the correct damping factor for the gauge boson propagator @xmath36 .
a detailed analysis @xcite of the interplay between terms in the action and the resulting propagators leads to the insight that it is possible to move one part ( @xmath37 ) of the problematic operator into the soft breaking term .
aiming to eventually avoid couplings between auxiliary fields and the gauge sector , @xmath37 is furthermore replaced by is not a gauge invariant operator , but as we will discuss subsequently , brst invariance of the action will be guaranteed by additional sources . ]
thereby , the desired damping is maintained while the ir divergences in the propagators of the auxiliary fields are eliminated .
note also that the dimensionful @xmath11 has been absorbed into the new parameters @xmath39 and @xmath40 , which have mass dimensions @xmath41 and @xmath42 .
in step 2 , we note that the regularizing effects are solely implemented in the bi - linear part of the action , therefore opening the option to reduce the field strength tensor @xmath4 in the soft breaking term to its bi - linear part @xmath43 noting furthermore , that the @xmath44 operator in the @xmath45 sector is not required any more for the implementation of the damping mechanism , we may entirely omit these covariant derivatives .
consequently , the mass dimensions @xmath46 of the fields @xmath35 and @xmath34 change from 1 to 2 .
any interaction ( represented by @xmath15-point functions with @xmath47 ) of @xmath5 with auxiliary fields and related ghosts is now eliminated .
this represents the advantage that no additional feynman diagrams appear since the auxiliary fields only enter the bi - linear part of the action . finally , in order to implement a suitable term to absorb the @xmath48-contracted one - loop divergence ( see eqn .
( [ eq : generic - ir - div ] ) and @xcite ) , we further modify the soft breaking part by the insertion of the term @xmath49 , resulting in .
the new parameters @xmath40 and @xmath39 are intended to receive corrections in the renormalization process , as will be discussed in section [ sec : renorm ] .
obviously the motivation for the new breaking term is of purely technical nature since it simply allows to achieve the requirements we have imposed to the new model at the beginning of this section .
physically , as has been discussed extensively in refs .
@xcite , a soft breaking term in the action ( which should be understood as an additional gauge fixing as in the gribov - zwanziger case ) enables us to modify the behavior at low energy ( i.e. to damp ir divergences ) while not spoiling the symmetries of the theory in the uv .
this should be understood as an additional gauge fixing as in the gribov - zwanziger case .
however , the freedom in the choice of this soft breaking is not yet entirely understood . clearly , due to the uv / ir mixing problem , a non - commutative gauge field model only has the chance to be renormalizable with this or a similar additional ir modification .
this is inherently different from the well - known yang - mills theory on commutative space and of course requires further investigation in the future . in view of these considerations ,
the following brst invariant action formulated in euclidean @xmath8 is put forward : @xmath50=\nonumber\\ * & = { \int\!\ ! { \rm d}^4x}\bigg[\!\!\left(\bar{j}_{{\mu}{\nu}{\alpha}{\beta}}b_{{\mu}{\nu}}+j_{{\mu}{\nu}{\alpha}{\beta}}\bar{b}_{{\mu}{\nu}}\right)\!{\frac{1}{{\widetilde{\square}}}}\!\left(f_{{\alpha}{\beta}}+{\sigma}\frac{{\theta}_{{\alpha}{\beta}}}{2}\tilde{f}\right)\
! - \bar{q}_{{\mu}{\nu}{\alpha}{\beta}}\psi_{{\mu}{\nu}}{\frac{1}{{\widetilde{\square}}}}\!\left(f_{{\alpha}{\beta}}+{\sigma}\frac{{\theta}_{{\alpha}{\beta}}}{2}\tilde{f}\right)\nonumber\\ * & \qquad \qquad -\left(\bar{q}_{{\mu}{\nu}{\alpha}{\beta}}b_{{\mu}{\nu}}+q_{{\mu}{\nu}{\alpha}{\beta}}\bar{b}_{{\mu}{\nu}}\right){\frac{1}{{\widetilde{\square}}}}\mathop{s}\left(f_{{\alpha}{\beta}}+{\sigma}\frac{{\theta}_{{\alpha}{\beta}}}{2}\tilde{f}\right)\bigg]\ , , \nonumber\\ { s}_{\text{ext}}&={\int\!\ ! { \rm d}^4x}\left({\omega}^a_{\mu}sa_{\mu}+{\omega}^c sc\right)\,,\end{aligned}\ ] ] where @xmath51 , and all products are implicitly assumed to be deformed ( i.e. star products ) .
this will also apply to the rest of our paper .
the non - commutative generalization of a @xmath52 gauge field is denoted by @xmath5 , @xmath53 and @xmath54 are the ( anti-)ghosts and the multiplier field @xmath55 implements the landau gauge fixing @xmath56 . @xmath57 and @xmath58 are external sources introduced for the non - linear brst transformations @xmath59 and @xmath60 , which will be defined shortly .
furthermore , the complex field @xmath35 and its conjugate @xmath34 as well as associated ghosts @xmath61 and @xmath62 have been introduced into the bilinear part of the action in order to implement the ir damping and brst invariance ( compatible to the ` soft breaking ' technique ) .
these new unphysical fields do not interact with the gauge field @xmath5 .
notice , that the landau gauge fixing can surely be replaced by any other choice , however the freedom in the choice of the soft breaking is not yet entirely understood . the additional sources @xmath63 are needed in order to ensure brst invariance of @xmath64 in the ultraviolet . at the opposite end of the spectrum ( in the infrared ) these sources take their ` physical values ' @xmath65
see refs .
@xcite for details on this technique .
the action is invariant under the brst transformations @xmath66 the auxiliary fields form brst doublets reflecting their unphysical nature .
dimensions and ghost numbers of the fields involved are given in table [ tab : field_prop ] .
l c c c@ c@ c@ c@ c@ c@ c@ c@ c@ c@ c@ c@ ' '' '' field & @xmath5 & @xmath54 & @xmath53 & @xmath35 & @xmath34 & @xmath61 & @xmath62 & @xmath67 & @xmath68 & @xmath69 & @xmath70 & @xmath71 & @xmath72 & @xmath55 + @xmath73 & 0 & 1 & -1 & 0 & 0 & 1 & -1 & 0 & 0 & -1 & -1 & -1 & -2 & 0 + mass dim .
& 1 & 0 & 2 & 2 & 2 & 2 & 2 & 2 & 2 & 2 & 2 & 3 & 4 & 2 + statistics & b & f & f & b & b & f & f & b & b & f & f & f & b & b + the slavnov - taylor identity describing the brst symmetry content of the model is given by @xmath74 from which one derives the linearized slavnov - taylor operator @xmath75 furthermore we have the gauge fixing condition @xmath76 the ghost equation @xmath77 and the antighost equation @xmath78 finally , we also have the symmetry @xmath79 : @xmath80=0\,,\end{aligned}\ ] ] whose trace is connected to the reality of the action . ] , and is hence denoted ` reality charge ' @xmath81 @xcite : @xmath82 obviously , @xmath81 also generates a symmetry of the action .
having defined the operators @xmath83 , @xmath84 and @xmath81 we may derive the following graded commutators : @xmath85}=0\ , , & & { \left[{\mathcal{q}},{\mathcal{q}}\right]}=0\ , , & & { \left[{\mathcal{b}_{s}},{\mathcal{q}}\right]}=0\,,\end{aligned}\ ] ] which means that these symmetry operators form a closed algebra .
in fact , each of the relations in eqn .
( [ eq : algebra ] ) gives rise to a constraint to possible counter terms in the effective action .
however , first we will present the situation at one - loop level .
the simplest way to calculate the new gauge field propagator is to integrate over the auxiliary fields @xmath86 in the path integral after taking the physical values of @xmath87 given in eqn .
( [ jq - phys_new ] ) : @xmath88\bigg)\bigg\}\nonumber\\ & = \int\mathcal{d}({\bar{b}}ba)\,\exp\bigg\{-\bigg({s}_{\text{inv}}+{s}_{\text{gf}}+{s}_{\text{ext}}+{\int\!\ ! { \rm d}^4x}\bigg[\tfrac{\gamma^4}{4{\widetilde{\square}}}\!\left(f_{{\mu}{\nu}}+{\sigma}\tfrac{{\theta}_{{\mu}{\nu}}}{2}\tilde{f}\right){\tfrac{1}{{\widetilde{\square}}}}\!\left(f_{{\mu}{\nu}}+{\sigma}\tfrac{{\theta}_{{\mu}{\nu}}}{2}\tilde{f}\right)\nonumber\\ & \hspace{3.1 cm } -\left({\bar{b}}_{{\mu}{\nu}}-\tfrac{\gamma^2}{2{\widetilde{\square}}}\!\left(f_{{\mu}{\nu}}+{\sigma}\tfrac{{\theta}_{{\mu}{\nu}}}{2}\tilde{f}\right)\right)\left(b_{{\mu}{\nu}}-\tfrac{\gamma^2}{2{\widetilde{\square}}}\!\left(f_{{\mu}{\nu}}+{\sigma}\tfrac{{\theta}_{{\mu}{\nu}}}{2}\tilde{f}\right)\right)\bigg]\bigg)\bigg\}\nonumber\\ & = \int\mathcal{d}a\exp\bigg\{-\bigg({s}_{\text{inv}}+{s}_{\text{gf}}+{s}_{\text{ext}}+{\int\!\ ! { \rm d}^4x}\tfrac{\gamma^4}{4{\widetilde{\square}}}\!\left(f_{{\mu}{\nu}}+{\sigma}\tfrac{{\theta}_{{\mu}{\nu}}}{2}\tilde{f}\right){\tfrac{1}{{\widetilde{\square}}}}\!\left(f_{{\mu}{\nu}}+{\sigma}\tfrac{{\theta}_{{\mu}{\nu}}}{2}\tilde{f}\right)\bigg)\bigg\}\,.\end{aligned}\ ] ] this leads to the action @xmath89+s\left({\bar{c}}{\partial}_{\mu}a_{\mu}\right)\right)\,,\end{aligned}\ ] ] which , using the abbreviation in eqn .
( [ eq : def - eth ] ) follows @xmath90 , which is , however , not inserted at this point in order to keep the results more general . ]
@xmath91 , and with the definition of @xmath92 , reduces to @xmath93+s\left({\bar{c}}{\partial}_{\mu}a_{\mu}\right)\right).\end{aligned}\ ] ] the gauge field propagator hence takes the form @xmath94}\frac{{\tilde{k}}_{\mu}{\tilde{k}}_{\nu}}{{\tilde{k}}^2}\right)\nonumber\\ & = \left[k^2+\frac{{\gamma}^4}{{\tilde{k}}^2}\right]^{-1}\left[{\delta}_{{\mu}{\nu}}-\frac{k_{\mu}k_{\nu}}{k^2}-\frac{{\bar{\sigma}}^4}{\left(k^2+\left({\bar{\sigma}}^4+{\gamma}^4\right){\frac{1}{{\tilde{k}}^2}}\right)}\frac{{\tilde{k}}_{\mu}{\tilde{k}}_{\nu}}{({\tilde{k}}^2)^2}\right]\ , , \label{eq : prop_aa}\end{aligned}\ ] ] where we have introduced the abbreviation @xmath95 and considered the case where @xmath12 has the simple block diagonal form given in so that @xmath96 .
two limits are of special interest : the ir limit @xmath97 and the uv limit @xmath98 .
a simple analysis reveals that @xmath99 , & \text{for } { \tilde{k}}^2\to0\,,\\[0.8em ] { \frac{1}{k^2}}\left({\delta}_{{\mu}{\nu}}-\frac{k_{\mu}k_{\nu}}{k^2}\right ) , & \text{for } k^2\to\infty\,.\\ \end{cases } \label{eq : prop_aa_limits}\ ] ] from eqn .
( [ eq : prop_aa_limits ] ) one can nicely see the appearance of a term of the same type as in the ir limit .
this , by construction , admits the absorption of the problematic divergent terms appearing in the one loop results @xcite .
another advantageous property of the gauge propagator is that the uv limit ( from which due to uv / ir mixing all divergences originate ) , admits to neglect the term proportional to @xmath39 which reduces the number of terms in feynman integrals considerably .
the ghost propagator takes the simple form @xmath100 which , as usual in a covariant gauge , is quadratically ir divergent .
alternatively , one could add a damping factor to the gauge fixing term and the ghost sector as has been done in ref .
however , such dampings appear in vertex expressions with an inverse power relative to the respective propagators and , thus , cancel each other .
moreover , these factors contribute to uv divergences at higher loop orders , and are omitted , hence . due to the missing coupling between the @xmath5 and the remaining fields ( @xmath86 ) no other propagator will contribute to physical results . however , for the sake of completeness , we give the respective expressions : [ eq : otherprops ] @xmath101}\frac{{\tilde{k}}_{\rho}{\tilde{k}}_{\sigma}}{{\tilde{k}}^2}\right]\nonumber\\ & = g^{{\bar{b}}a}_{{\mu}{\nu},{\rho}}(k)\,,\label{eq : prop_ba}\\ g^{bb}_{{\mu}{\nu},{\rho}{\sigma}}(k)&=-{\gamma}^4\frac{\left(k_{\mu}k_{\rho}{\delta}_{{\nu}{\sigma}}+k_{\nu}k_{\sigma}{\delta}_{{\mu}{\rho}}-k_{\mu}k_{\sigma}{\delta}_{{\nu}{\rho}}-k_{\nu}k_{\rho}{\delta}_{{\mu}{\sigma}}\right)}{2k^2{\tilde{k}}^2\left({\tilde{k}}^2+\frac{{\gamma}^4}{{\tilde{k}}^2}\right)}\nonumber\\ * & \quad+\frac{{\gamma}^4}{4{\tilde{k}}^2\left[k^2\left({\tilde{k}}^2+\frac{{\gamma}^4}{{\tilde{k}}^2}\right)+{\bar{\sigma}}^4\right]}\bigg[{\sigma}{\theta}_{{\mu}{\nu}}\left(k_{\rho}{\tilde{k}}_{\sigma}-k_{\sigma}{\tilde{k}}_{\rho}\right)+{\sigma}{\theta}_{{\rho}{\sigma}}\left(k_{\mu}{\tilde{k}}_{\nu}-k_{\nu}{\tilde{k}}_{\mu}\right)\nonumber\\ & \quad\qquad\;-{\sigma}^2{\tilde{k}}^2{\theta}_{{\mu}{\nu}}{\theta}_{{\rho}{\sigma}}-{\bar{\sigma}}^4\frac{\left(k_{\mu}{\tilde{k}}_{\nu}{\tilde{k}}_{\rho}k_{\sigma}+k_{\rho}{\tilde{k}}_{\sigma}{\tilde{k}}_{\mu}k_{\nu}-k_{\mu}{\tilde{k}}_{\nu}{\tilde{k}}_{\sigma}k_{\rho}+k_{\sigma}{\tilde{k}}_{\rho}{\tilde{k}}_{\mu}k_{\nu}\right)}{k^2{\tilde{k}}^2\left({\tilde{k}}^2+\frac{{\gamma}^4}{{\tilde{k}}^2}\right)}\bigg]\nonumber\\ & = g^{{\bar{b}}{\bar{b}}}_{{\mu}{\nu},{\rho}{\sigma}}(k),\label{eq : prop_bb}\\ g^{b{\bar{b}}}_{{\mu}{\nu},{\rho}{\sigma}}(k)&=-{\frac{1}{2}}\left({\delta}_{{\mu}{\rho}}{\delta}_{{\nu}{\sigma}}-{\delta}_{{\mu}{\sigma}}{\delta}_{{\nu}{\rho}}\right)+g^{bb}_{{\mu}{\nu},{\rho}{\sigma}}(k)\,,\label{eq : prop_bbb}\\ g^{{\bar{\psi}}\psi}_{{\mu}{\nu}{\rho}{\sigma}}(k)&=-{\frac{1}{2}}\left({\delta}_{{\mu}{\rho}}{\delta}_{{\nu}{\sigma}}-{\delta}_{{\mu}{\sigma}}{\delta}_{{\nu}{\rho}}\right)\,.\label{eq : prop_psi}\end{aligned}\ ] ] notice , that all four propagators tend towards a constant in the infrared as well as in the ultraviolet .
the model features three vertices which equal those of the ` nave ' implementation of qed on non - commutative spaces ( see for example ref .
@xcite ) : @xmath102{p-2_vertex_aaa.eps}}&=\widetilde{v}^{3a}_{\rho{\sigma}\tau}(k_1 , k_2 , k_3)=2{{\rm i}g}(2\pi)^4{\delta}^4(k_1+k_2+k_3)\sin\left(\tfrac{{\varepsilon}}{2 } k_1{\tilde{k}}_2\right)\times\nonumber\\[-12pt ] & \phantom{=\widetilde{v}^{3a}_{\rho{\sigma}\tau}(k_1 , k_2 , k_3)\,}\quad\times\left[(k_3-k_2)_\rho { \delta}_{{\sigma}\tau}+(k_1-k_3)_{\sigma}{\delta}_{\rho\tau}+(k_2-k_1)_\tau { \delta}_{\rho{\sigma}}\right],\label{eq : vert_3a}\\ \hspace{-3ex}\raisebox{-23pt}{\includegraphics[scale=0.8,trim=0 2pt 0 5pt , clip = true]{p-2_vertex_4a.eps}}\hspace{-9pt}&=\widetilde{v}^{4a}_{\rho{\sigma}\tau{\epsilon}}(k_1 , k_2 , k_3 , k_4)=-4g^2(2\pi)^4{\delta}^4(k_1+k_2+k_3+k_4)\times\nonumber\\*[-16pt ] & \phantom{=\widetilde{v}^{4a}_{\rho{\sigma}\tau{\epsilon}}(k_1 , k_2 , k_3 , k_4)=}\;\times\big[({\delta}_{\rho\tau}{\delta}_{{\sigma}{\epsilon}}-{\delta}_{\rho{\epsilon}}{\delta}_{{\sigma}\tau})\sin\left(\tfrac{{\varepsilon}}{2}k_1{\tilde{k}}_2\right)\sin\left(\tfrac{{\varepsilon}}{2}k_3{\tilde{k}}_4\right)\nonumber\\ * & \phantom{=\widetilde{v}^{4a}_{\rho{\sigma}\tau{\epsilon}}(k_1 , k_2 , k_3 , k_4)=}\;\phantom{\big[}+({\delta}_{\rho{\sigma}}{\delta}_{\tau{\epsilon}}-{\delta}_{\rho{\epsilon}}{\delta}_{{\sigma}\tau})\sin\left(\tfrac{{\varepsilon}}{2}k_1{\tilde{k}}_3\right)\sin\left(\tfrac{{\varepsilon}}{2}k_2{\tilde{k}}_4\right)\nonumber\\ * & \phantom{=\widetilde{v}^{4a}_{\rho{\sigma}\tau{\epsilon}}(k_1 , k_2 , k_3 , k_4)=}\;\phantom{\big[}+({\delta}_{\rho{\sigma}}{\delta}_{\tau{\epsilon}}-{\delta}_{\rho\tau}{\delta}_{{\sigma}{\epsilon}})\sin\left(\tfrac{{\varepsilon}}{2}k_2{\tilde{k}}_3\right)\sin\left(\tfrac{{\varepsilon}}{2}k_1{\tilde{k}}_4\right)\big],\label{eq : vert_4a}\\ \hspace{-3ex}\raisebox{-23pt}{\includegraphics[scale=0.8,trim=0 2pt 0 5pt , clip = true]{p-2_vertex_cac.eps}}\hspace{0.1ex}&=\widetilde{v}^{{\bar{c}}ac}_\mu(q_1 , k_2 , q_3)=-2{{\rm i}g}(2\pi)^4{\delta}^4(q_1+k_2+q_3)q_{3\mu}\sin\left(\tfrac{{\varepsilon}}{2}q_1{\tilde{q}}_3\right)\label{eq : vert_cabc}\,.\end{aligned}\ ] ] concerning the superficial degree of uv divergence @xmath103 one can set up the following relations for the number of loops @xmath104 , external lines @xmath105 , internal lines @xmath106 and vertices @xmath107 for fields @xmath108 : @xmath109 this system of equations can be resolved by eliminating the @xmath106 and @xmath107 , finally leading to @xmath110 which , again , shows remarkable agreement with the respective relations for the ` nave ' implementation of non - commutative @xmath0 gauge theory .
the feynman rules , , , and give rise to the three graphs in fig .
[ fig:1loop_vacpol_all ] contributing to the vacuum polarization @xmath111 . in the light of renormalization one is interested mainly in the divergence structure of these expressions in the limit of small external momenta @xmath112 .
therefore , we apply an expansion of the following form to the integrands @xmath113 , @xmath114_{p\to0 } \nonumber\\ & \hspace{2.8cm}+\frac{p_\rho p_{\sigma}}{2 } \left[\partial_{p_\rho}\partial_{p_{\sigma}}\mathcal{i}_{\mu\nu}(p , k)\right]_{p\to0 } + \mathcal{o } \left(p^3\right)\bigg ] .\end{aligned}\ ] ] the phase factors are not expanded in order not to lose the damping effect of the highly oscillating functions at high momenta @xmath21 .
notice , that since all potential divergences ( uv as well as ir coming from uv / ir mixing ) arise from the region of _ large _ internal momentum @xmath21 , one may use the uv approximation for the gauge field propagator eqn .
( [ eq : prop_aa_limits ] ) for all one - loop calculations . conducting the expansions and integrations for the graphs depicted in fig .
[ fig:1loop_vacpol_all ] one finally arrives at @xmath115 where @xmath116 denotes an ultraviolet cutoff , and ` finite terms ' collects contributions being finite in the limits @xmath117 and @xmath118 , respectively .
as expected this result exhibits a quadratic ir divergence showing the tensor structure of eqn .
( [ eq : generic - ir - div ] ) . in principal , the calculation of
the vertex corrections proceeds along the lines of the previous section . however , one has to guarantee that the limits @xmath119 , @xmath120 commute for the integrand and its derivatives . a typical divergent expression , as it appears in the calculation of the graphs depicted in fig .
[ fig:1loop_3a_all ] , reads @xmath121 where @xmath122 denotes a regular smooth function . in order to evaluate this integral
it is convenient to utilize the identity @xmath123 the sum of the graphs depicted in fig .
[ fig:1loop_3a_all ] results in a linear ir divergence of the form @xmath124 as well as a logarithmic uv divergence @xmath125\nonumber\\ & = -\frac{17\,g^2}{6(4\pi)^2}\ln ( \lambda ) \widetilde{v}^{3a,\text{tree}}_{{\mu}{\nu}{\rho}}(p_1,p_2,p_3)\,,\end{aligned}\ ] ] where due to momentum conservation @xmath126 .
we note that our results and resemble the ones given in the literature @xcite .
the feynman graphs contributing to the one - loop 4-a vertex corrections are depicted in fig .
[ fig:1loop_4a_all ] .
they are all ir finite .
gauge invariance and the renormalization of the coupling constant from the 3-a vertex , as discussed in section [ sec : betafun ] , lead to the following logarithmic uv divergence in the 4-a vertex correction ( cf .
@xcite ) : @xmath127 where due to momentum conservation @xmath128 . for a renormalizable theory , the form of the action is invariant under quantum corrections (
provided the subtraction scheme respects all the symmetries ) and the parameters are fixed by renormalization conditions on the vertex functions . recall that the tree - level gauge field propagator has the form @xmath129 where we have introduced the abbreviations @xmath130 so far
, we have worked in landau gauge with @xmath131 in order to simplify loop - calculations ( cf .
( [ eq : prop_aa ] ) ) .
however , in the following a more general gauge fixing with arbitrary gauge parameter @xmath132 will be advantageous , as we explicitly need the inverse of the propagator , which for obvious reasons diverges for @xmath133 due to the elimination of the lagrange multiplier field @xmath55 : @xmath134 of course , one could do these computations also in the landau gauge , but one would have to take @xmath55 and its ( mixed ) propagator into account . instead , in order to keep things simple , we will take the limit @xmath133 only in the final step . in section [ sec : vac - pol ]
we have computed the one - loop corrections to the tree - level two - point vertex function .
it is given by eqn .
( [ eq : res_vac_pol_divergence ] ) , i.e. @xmath135 where @xmath116 was an ultraviolet cutoff . hence
, with these definitions for @xmath136 we find @xmath137 introducing the wave - function renormalization @xmath138 and the renormalized parameters @xmath139 and @xmath140 according to @xmath141 the one - loop two - point vertex function is cast into the same form as its tree - level counter part , i.e. @xmath142 finally , we may also write @xmath140 in terms of the renormalized @xmath143 : @xmath144 for the sake of completeness , we note that the renormalized propagator in landau gauge ( @xmath133 ) becomes @xmath145 in the following , we will provide renormalization conditions for the two - point vertex function for the gauge boson @xmath146 where the vertex function has been split into a transversal and longitudinal part .
we have used the identifications @xmath147 which finally allow to formulate the following renormalization conditions : @xmath148 regarding the renormalization arising from the three - point functions the ir divergent result corresponds to the counter term @xmath149 such a term can readily be introduced into the soft - breaking part of the action @xmath150 in eqn .
( [ eq : renormalizable_action ] ) .
but in order to do so , we have to restore brst invariance in the uv regime .
again , this can be achieved by introducing sources @xmath151 and @xmath152 , which form a brst doublet , @xmath153 consequently , we insert the following terms into @xmath150 in eqn .
( [ eq : renormalizable_action ] ) : @xmath154\,.\end{aligned}\ ] ] this term is brst invariant , _ per se_. in the ir , the sources take on their physical values @xmath155 ( cf . eqn .
( [ jq - phys_new ] ) and refs .
@xcite ) , and the counter term in eqn .
( [ 3a - correction ] ) leads to a renormalized @xmath156 , which is another parameter of dimension 1 . adding the term to the action
does not alter the divergence structure of the theory , due to the damping behavior of the propagators in the ir , and the fact that the additional contribution to the 3a - vertex is subleading in the uv , i.e. it does not modify the uv power counting .
in contrast to the ir singular terms , the uv divergences of and can be absorbed into the coupling constant , which is discussed subsequently in section [ sec : betafun ] . as usual , the @xmath157-function is given by the logarithmic derivative of the coupling @xmath158 with respect to the cut - off for fixed @xmath159 : @xmath160 the renormalized coupling is obtained from the relation @xmath161 where @xmath162 denotes the multiplicative correction to the three - photon vertex , and @xmath138 is the wave function renormalization given in .
explicitly , we have [ eq : betafun_explicit_corr ] @xmath163 note that the appearance of the wave function renormalization @xmath138 in eqn .
( [ gg ] ) is reasoned by the vertex correction being computed with the unrenormalized fields @xmath5 , which need to be replaced by their renormalized counter parts @xmath164 . since this topic
is treated rather novercally in the literature , let us briefly sketch how the @xmath157-function is obtained in practise .
bearing in mind that , according to eqn .
( [ eq : betafun_def ] ) , the renormalized coupling is being held constant , we differentiate ( after expansion for small @xmath158 ) with respect to @xmath165 and multiply the result by @xmath165 . with the general definition @xmath166 ,
this yields @xmath167 the solution of is then obtained by the ansatz @xmath168 and we finally obtain @xmath169 inserting the explicit definitions , the @xmath157-function reads @xmath170 the negative sign of the @xmath157-function indicates asymptotic freedom and the absence of a landau ghost .
this qualitatively compares to the results in @xcite , where also a negative @xmath157-function is obtained for the case of @xmath23-deformed qed .
in the light of renormalizability at higher loop orders , it is important to investigate the ir behavior of such integrands with insertions of the one - loop corrections which were discussed in the previous section . the aim is to identify possible poles give rise to ir divergences . upon insertion of these functions into higher - loop graphs
the former external momenta are being integrated over and , hence , also touch the value @xmath171 .
] at @xmath172 .
hence , we consider a chain of @xmath15 non - planar insertions , which may be part of a higher loop graph .
non - planar insertions , concatenated by gauge field propagators . ] considering , for example , the ir behavior of the chain depicted in fig .
[ fig : nloop_aa ] we find by employing eqns . and : @xmath173^{n+1}}{\tilde{p}}_{\mu}{\tilde{p}}_{\nu}\,,\end{aligned}\ ] ] i.e. the expression is well - behaved in the limit of small @xmath174 .
therefore , no ir problems are expected in higher loop graphs with insertions of the above type .
in fact , represents a nice illustration of the damping mechanism implemented by the soft breaking term in the action and parametrized by @xmath39 . in the limit @xmath175 ( implying @xmath176 ) the divergent behavior of the ` nave ' model without soft breaking is recovered .
after suffering a setback in our enterprise to construct a renormalizable non - commutative gauge field model @xcite we have thoroughly analyzed the obstacles raised by the introduction of a parameter of non - commutativity with negative mass dimension @xcite .
benefiting from the insight gained from this analysis we have worked out a non - commutative @xmath0 gauge model which provides the necessary terms to absorb all divergences appearing at one - loop level .
the main concepts of the construction are : * the ir damping of the @xmath24 model is implemented by a soft breaking term .
this allows us to use ordinary derivatives instead of inverse covariant derivatives which need to be localized by auxiliary fields and ghosts in order to be interpreted in a reasonable way .
thereby , the gauge field @xmath177 is completely decoupled ( in terms of interactions ) from the new auxiliary fields , and respective quantum corrections reduce ( diagrammatically ) to the ones known from the nave non - commutative ym and qed theories @xcite . *
the tree level action features terms @xmath178 , with powers @xmath179 being designed to absorb all types of divergences appearing at one - loop level . from transversality with respect to external momenta , dimensional considerations , and power counting , one can derive that these , in fact , are the only ir divergent tensor structures appearing in the corrections .
explicit one - loop computations showed the expected results .
the vacuum polarization of the gauge boson behaves for small external momenta as @xmath180 a divergence which has been shown to be absorbable into the parameter of the @xmath181 term in the action .
the correction of the @xmath182 vertex reads @xmath183 corresponding to the renormalization of the ( massive ) parameter of the term @xmath184 .
these one - loop results lead to a negative @xmath157-function ( [ beta ] ) , indicating asymptotic freedom and the absence of a landau ghost .
this remarkable fact is due to the ` non - abelian ' nature of the non - commutative @xmath0 theory .
although , at the moment , we have not been able to construct a rigorous proof , there are strong indications that the properties mentioned above indeed lead to renormalizability of this model .
we may conjecture that this can be confirmed in the framework of multiscale analysis ( for a review see ref .
this , however , will require extensive work to establish the foundations @xcite for the application of this scheme to gauge theories .
the authors are indebted to h. grosse , f. heindl , e. kronberger and h. steinacker for valuable discussions . the work of d. n. blaschke , a. rofner , r. i. p. sedmik and m. wohlgenannt was supported by the ` fonds zur frderung der wissenschaftlichen forschung ' ( fwf ) under contracts p20507-n16 and p21610-n16 .
d. n. blaschke , e. kronberger , a. rofner , m. schweda , r. i. p. sedmik and m. wohlgenannt , _ on the problem of renormalizability in non - commutative gauge field models a critical review _ , _ fortschr .
phys . _ * 58 * ( 2010 ) 364 , http://www.arxiv.org/abs/0908.0467[`[arxiv:0908.0467 ] ` ] .
d. n. blaschke , a. rofner and r. i. p. sedmik , _ one - loop calculations and detailed analysis of the localized non - commutative @xmath24 @xmath52 gauge model _ , _ sigma _ * 6 * ( 2010 ) 037 , http://www.arxiv.org/abs/0908.1743[`[arxiv:0908.1743 ] ` ] .
v. rivasseau , _ non - commutative renormalization , _ in _ quantum spaces poincar seminar 2007 _ , b. duplantier and v. rivasseau eds . ,
birkhuser verlag , http://www.arxiv.org/abs/0705.0705[`[arxiv:0705.0705 ] ` ] .
h. grosse and r. wulkenhaar , _ renormalisation of @xmath185 theory on noncommutative @xmath186 in the matrix base _ , _ jhep _ * 12 * ( 2003 ) 019 , http://www.arxiv.org/abs/hep-th/0307017[`[arxiv:hep-th/0307017 ] ` ] .
v. rivasseau , f. vignes - tourneret and r. wulkenhaar , _ renormalization of noncommutative @xmath185-theory by multi - scale analysis _ ,
* 262 * ( 2006 ) 565594 , http://www.arxiv.org/abs/hep-th/0501036[`[arxiv:hep-th/0501036 ] ` ] .
r. gurau , j. magnen , v. rivasseau and f. vignes - tourneret , _ renormalization of non - commutative @xmath187 field theory in x space _ , _ commun .
* 267 * ( 2006 ) 515542 , http://www.arxiv.org/abs/hep-th/0512271[`[arxiv:hep-th/0512271 ] ` ] .
h. grosse and r. wulkenhaar , _ renormalisation of @xmath185 theory on noncommutative @xmath188 in the matrix base _ , _ commun .
* 256 * ( 2005 ) 305374 , http://www.arxiv.org/abs/hep-th/0401128[`[arxiv:hep-th/0401128 ] ` ] .
r. gurau , j. magnen , v. rivasseau and a. tanasa , _ a translation - invariant renormalizable non - commutative scalar model _ , _ commun . math .
phys . _ * 287 * ( 2009 ) 275290 , http://www.arxiv.org/abs/0802.0791[`[arxiv:0802.0791 ] ` ] .
d. n. blaschke , f. gieres , e. kronberger , t. reis , m. schweda and r. i. p. sedmik , _ quantum corrections for translation - invariant renormalizable non - commutative @xmath189 theory _ , _ jhep _ * 11 * ( 2008 ) 074 , http://www.arxiv.org/abs/0807.3270[`[arxiv:0807.3270 ] ` ] .
j. b. geloun and a. tanasa , _ one - loop @xmath157 functions of a translation - invariant renormalizable noncommutative scalar model _ , _ lett . math .
* 86 * ( 2008 ) 1932 , http://www.arxiv.org/abs/0806.3886[`[arxiv:0806.3886 ] ` ] .
d. n. blaschke , h. grosse and m. schweda , _ non - commutative u(1 ) gauge theory on @xmath188 with oscillator term and brst symmetry _ , _ europhys .
* 79 * ( 2007 ) 61002 , http://www.arxiv.org/abs/0705.4205[`[arxiv:0705.4205 ] ` ] .
d. n. blaschke , f. gieres , e. kronberger , m. schweda and m. wohlgenannt , _ translation - invariant models for non - commutative gauge fields _ , _ j. phys .
_ * a41 * ( 2008 ) 252002 , http://www.arxiv.org/abs/0804.1914[`[arxiv:0804.1914 ] ` ] .
d. n. blaschke , a. rofner , m. schweda and r. i. p. sedmik , _ one - loop calculations for a translation invariant non - commutative gauge model _ , _ eur .
_ * c62 * ( 2009 ) 433 , http://www.arxiv.org/abs/0901.1681[`[arxiv:0901.1681 ] ` ] . l. c. q. vilar , o. s. ventura , d. g. tedesco and v. e. r. lemes , _ on the renormalizability of noncommutative u(1 ) gauge theory an algebraic approach _ , _ j. phys .
_ * a43 * ( 2010 ) 135401 , http://www.arxiv.org/abs/0902.2956[`[arxiv:0902.2956 ] ` ] .
d. n. blaschke , a. rofner , m. schweda and r. i. p. sedmik , _ improved localization of a renormalizable non - commutative translation invariant u(1 ) gauge model _ , _ epl _ * 86 * ( 2009 ) 51002 , http://www.arxiv.org/abs/0903.4811[`[arxiv:0903.4811 ] ` ] .
f. r. ruiz , _ gauge - fixing independence of ir divergences in non - commutative u(1 ) , perturbative tachyonic instabilities and supersymmetry _ , _ phys .
* b502 * ( 2001 ) 274278 , http://www.arxiv.org/abs/hep-th/0012171[`[arxiv:hep-th/0012171 ] ` ] .
d. dudal , j. gracey , s. p. sorella , n. vandersickel and h. verschelde , _ a refinement of the gribov - zwanziger approach in the landau gauge : infrared propagators in harmony with the lattice results _ , _ phys . rev . _ * d78 * ( 2008 ) 065047 , http://www.arxiv.org/abs/0806.4348[`[arxiv:0806.4348 ] ` ] . l. baulieu and s. p. sorella , _ soft breaking of brst invariance for introducing non - perturbative infrared effects in a local and renormalizable way _ , _ phys .
lett . _ * b671 * ( 2009 ) 481 , http://www.arxiv.org/abs/0808.1356[`[arxiv:0808.1356 ] ` ] .
c. p. martin and f. ruiz ruiz , _ paramagnetic dominance , the sign of the beta function and uv / ir mixing in non - commutative u(1 ) _ , _ nucl .
phys . _ * b597 * ( 2001 ) 197227 , http://www.arxiv.org/abs/hep-th/0007131[`[arxiv:hep-th/0007131 ] ` ] .
m. attems , d. n. blaschke , m. ortner , m. schweda , s. stricker and m. weiretmayr , _ gauge independence of ir singularities in non - commutative qft - and interpolating gauges _ , _ jhep _ * 07 * ( 2005 ) 071 , http://www.arxiv.org/abs/hep-th/0506117[`[arxiv:hep-th/0506117 ] ` ] .
c. p. martin and d. sanchez - ruiz , _ the one - loop uv divergent structure of @xmath52 yang - mills theory on noncommutative @xmath188 _ , _ phys .
* 83 * ( 1999 ) 476479 , http://www.arxiv.org/abs/hep-th/9903077[`[arxiv:hep-th/9903077 ] ` ] . | based on our recent findings regarding ( non-)renormalizability of non - commutative @xmath0 gauge theories @xcite we present the construction of a new type of model . by introducing a soft breaking term in such a way that only the bilinear part of the action is modified , no interaction between the gauge sector and auxiliary fields occurs .
demanding in addition that the latter form brst doublet structures , this leads to a minimally altered non - commutative @xmath0 gauge model featuring an ir damping behavior .
moreover , the new breaking term is shown to provide the necessary structure in order to absorb the inevitable quadratic ir divergences appearing at one - loop level in theories of this kind . in the present paper
we compute feynman rules , symmetries and results for the vacuum polarization together with the one - loop renormalization of the gauge boson propagator and the three - point functions .
institute for theoretical physics , vienna university of technology + wiedner hauptstrasse 8 - 10 , a-1040 vienna ( austria ) + faculty of physics , university of vienna + boltzmanngasse 5 , a-1090 vienna ( austria ) + |
zinc oxide is a prominent example of an intrinsically n - type semiconductor .
its abundant availability , its low - cost production , and its wide electronic band gap of approximately 3.4 ev make zno attractive for transparent and conducting oxide layers in consumer - electronics devices . however , a break - through of zno - based transparent electronics has not been achieved so far due to the lack of sufficiently conductive and long - term stable zno layers .
the problem of p - type doping of zno has by now been a scientific challenge for more than two decades . besides the doping with single elements like n , p , as , or sb , substituting o and li ,
na , k , cu , ag or au , substituting zn@xcite co - doping was discussed as well.@xcite apparently , the most promising candidate so far is still nitrogen .
however , it was shown theoretically@xcite that the substitutional defect of n on o sites ( n@xmath2 ) can not be responsible for the p - type conduction which has been observed in some experiments,@xcite since it creates a defect level which is located too high above the valence band ( vb ) edge with respect to thermal excitation energies .
recent theoretical and experimental studies consider more complicated defect complexes which involve zinc vacancies ( v@xmath4 ) or co - doping , like the n@xmath2-v@xmath4 or n@xmath2-v@xmath4-h defect complexes , as potential sources of p - type conduction.@xcite for polycrystalline zno it has been proposed that conduction channels of p - type can be found along grain boundaries.@xcite here it is an interplay of grain boundaries , their depletion regions , and point defects that results in the desired low lying acceptor levels .
it has been reported by hierro et al.@xcite that the use of ternary zno - based alloys like zn@xmath0mg@xmath1o ( denoted by zmo in the following ) facilitates the p - type conductivity due to lower net electron concentrations than in pure zno .
these authors relate the reduced net electron concentration to two deep lying mg - related defects which act as electron traps .
zmo retains the wurtzite crystal structure of zno for mg substituting zn for concentrations up to @xmath5 , and also the lattice parameters change only slightly . hence combining zmo with n - type zno layers
leads to negligible lattice misfits.@xcite up to now it is not yet clear which atomic or extended defects provide the key to promote p - type conductivity in zno and related materials like zmo .
an understanding of this defect physics would allow a targeted development of experimental routes to synthesize materials having these particular acceptor states .
in this paper we present a density functional theory ( dft ) study determining the electronic defect levels of the defects n@xmath2 , v@xmath4 , and n@xmath2-v@xmath4 in zno and zmo .
we discuss our results in the context of the accumulated experimental and theoretical knowledge , and thereby we shed light on the question of why some experimental routes recently taken have lead to p - conduction but others not .
are the defect states that originate from n@xmath2 indeed too deep@xcite in the band gap ? what changes when n@xmath2 forms a defect complex with a zn vacancy , and how do mg atoms nearby affect its defect level position ? for the determination of the electronic defect levels we use a self - interaction corrected ( sic ) local - density approximation ( lda ) approach@xcite that is capable to give far better results than both commonly used approximations , the lda and the generalized gradient approximation ( gga ) .
the latter two have limited predictive power for wide band gap semiconductors due to their inherent significant underestimation of the band gap due to their self - interaction error .
lda overcomes this limitation and yields electronic band structures with band gaps that are in good agreement with experiment@xcite and in particular for zno a correct separation of the zn-3d and o-2p bands.@xcite by means of this approach , the defect levels in various amorphous oxide semiconductors based on in , sn , and zn have been predicted reliably.@xcite the results presented in this paper may solve the controversial problem that individual n@xmath2 defects are indeed creating levels lying too deep inside the band gap even though nitrogen doping has been observed to lead to p - type zno and zmo , probably due to an interplay with other defects .
moreover , we relate our present results to the p - type conduction effects which have been observed for grain boundaries in zno , and we relate all this to one common origin .
the emerging theoretical picture will presumably be useful for the experimental optimization of p - type conduction in zno and zmo .
mg@xmath6o@xmath7 with randomly distributed mg atoms ( left ) and with clustered mg atoms ( right ) .
large gray and blue spheres represent zn and mg atoms , respectively .
small red spheres represent oxygen atoms .
[ fig : znmgo : cells],title="fig:",width=151 ] mg@xmath6o@xmath7 with randomly distributed mg atoms ( left ) and with clustered mg atoms ( right ) .
large gray and blue spheres represent zn and mg atoms , respectively .
small red spheres represent oxygen atoms .
[ fig : znmgo : cells],title="fig:",width=151 ]
for the study of defect complexes we have chosen a supercell model of the wurtzite crystal structure of zno containing 72 atoms and consisting of @xmath8 primitive unit cells . since zn@xmath0mg@xmath1o retains the wurtzite structure for mg concentrations @xmath9%@xcite the corresponding supercells were generated by randomly substituting zn atoms by mg atoms .
for the study of the bulk crystal properties like lattice parameters and band gap as function of the mg concentration we compared several realizations of random mg distributions within the 72-atoms supercell , and we find a negligible dependency of these properties on the mg distribution . for comparison
, we have also examined results from a 36-atom supercell in the study of bulk properties .
for the subsequent study of defects we have chosen the specific mg concentration of @xmath1011% ( i.e. replacing 4 of 36 zn atoms by mg atoms in the 72-atom supercell ) which is representative for typical experimental concentrations.@xcite furthermore , the defect analysis concentrates on specific realizations of the mg distribution which differ in the number of mg atoms that cluster in the direct neighborhood of the @xmath11-point defect under study .
two such realizations are displayed in fig .
[ fig : znmgo : cells ] .
the structure optimization of all the constructed supercell models was carried out using the projector augmented - wave ( paw ) method@xcite as implemented in the vasp code.@xcite the lda was used for the exchange - correlation functional , and paw potentials describing the zn-(3d,4s,4p ) , mg-3s , and o-(2s,2p ) as valence electrons .
all vasp calculations were carried out with a plane - wave cutoff energy of 400 ev , a @xmath12 monkhorst - pack k - mesh , and a gaussian broadening of 0.1 ev .
for all bulk structures the volume optimization was carried out by fitting the universal equation of state@xcite to a dataset of minimal total energies calculated at different cell volumes .
optimization of the lattice parameter ratio c / a and internal relaxation of the atomic positions were achieved by minimizing the elastic stress and the forces acting on the atoms , respectively .
the supercells containing defects were internally relaxed at fixed volume and c / a ratio .
the electronic - structure calculations on the basis of the sic - lda were performed using the computational mixed - basis pseudopotential ( mbpp ) method@xcite with the same calculation setup as reported in previous papers.@xcite we have taken the lda for exchange - correlation as parametrized by perdew and zunger.@xcite for zn , mg , and o optimally smooth norm conserving pseudopotentials@xcite were constructed , and a mixed basis of plane waves and non - overlapping localized orbitals were used . due to the localized orbitals a plane wave cutoff energy of 20 ry ( 1ry = 13.606 ev ) is sufficient for obtaining well converged results . for the k - point sampling of the brillouin - zone
integrals a monkhorst - pack mesh of @xmath12 and a gaussian broadening of 0.2 ev were used .
the dos of the supercells were evaluated with the same mesh and a gaussian broadening of 0.1 ev .
the self - interaction of the lda is corrected by an incorporation of the sic in the pseudopotentials.@xcite the sic procedure uses weight factors @xmath13@xmath14 accounting for the occupations of the individual s , p and d valence orbitals .
we corrected the zn 3d semicore - orbitals by 100% , i.e. @xmath13@xmath15=(0 , 0 , 1 ) .
the localized o-2s semi - core orbitals are also corrected by 100% while for the spatially more extended o-2p valence - band orbitals we have taken 80% which implies the weight factors @xmath13@xmath16=(1 , 0.8 , 0 ) .
we did not apply a sic to the almost unoccupied mg-3s , 3p and zn-4s , 4p conduction - band orbitals of the almost fully ionic compounds zn@xmath0mg@xmath1o . for the adjustment of the atomic sic to the crystal field in the solids , a scaling factor @xmath17 is applied .
the functional dependencies of the lattice parameters @xmath18 and @xmath19 on the mg content are displayed in fig .
[ fig : a_c ] .
our results underestimate the experimental values by about one percent , which is a typical systematic deviation for lda results .
the increase of @xmath18 and the decrease of @xmath19 with increasing mg content are quantitatively correct and just slightly shifted with respect to previous experimental@xcite and theoretical@xcite findings .
this good agreement on the structural parameters is the prerequisite for the electronic structure calculations which are the focus of this work .
note that the deviations from vegard s rule in some of the values of the experimentally observed lattice constants is probably related to the quality of the thin films and maybe due to the presence of compensated defects or residual strain as discussed , e.g. , by ryoken _
_ et al.__@xcite , width=321 ] mg@xmath1o as function of the mg content @xmath20 .
the data points calculated with sic - lda are connected by a black dashed line to guide the eyes .
experimental data from several groups are shown for comparison.@xcite [ fig : bandgap ] , width=321 ] theoretical and experimental values of the band gap of zn@xmath0mg@xmath1o as function of the mg content @xmath20 are displayed and compared in fig .
[ fig : bandgap ] . the results obtained by the sic - lda approach
do not only show the rising trend correctly , but also the calculated values agree quantitatively well to the experimental data .
the dependency of the band gap on the specific distribution of mg atoms within the supercell is small , e.g. for the sample structures with @xmath21 we find an energy scatter of only @xmath22mev .
the wide scatter of experimental results in fig.[fig : bandgap ] is probably due to the varying quality of the thin films but also due to different experimental approaches of determining the band gaps.@xcite above the thermodynamical solubility limit of @xmath23 a non - equilibrium solid solution is formed .
depending on the route of synthesis , this results in compositional non - uniformity which then leads to the generation of non - equilibrium compensated defects@xcite or even structural inhomogeneity@xcite and residual elastic stress .
naturally , all of this is reflected in the measured physical properties of the films , as discussed in the respective articles . for our investigation the calculation of the band gap as function of the mg content serves as a validation that the sic - lda describes the electronic structure of the zn - mg - o system rather well .
the total density of states ( dos ) of a representative zn@xmath24mg@xmath6o@xmath7 sample calculated with the sic - lda approach is shown in figure [ fig : dos : all ] .
the upper part of the vb is mainly formed by the o 2p valence orbitals while the lower part is dominated by the zn 3d semicore orbitals .
experimentally , the average d - band energy of zno was determined to be at -7.8 ev relative to the vb edge.@xcite this is almost unchanged when mg is introduced .
the sic - lda approach reproduces this experimental observation in contrast to lda , gga and several commonly used hybrid functionals which place the zn 3d levels too high in energy and therefore overestimate their hybridization with the o 2p levels .
the latter can lead to wrong positions of defect levels in the band gap.@xcite thus these results validate our approach to analyze individual electronic levels of point defects by using the sic - lda approach .
mg@xmath6o@xmath7 without ( bulk ) and with n@xmath2 , v@xmath4 , and n@xmath2-v@xmath4 defects .
the inset magnifies the region close to the vb maximum .
small colored arrows on the energy axis mark the highest occupied levels .
the curves are aligned with respect to the valence band and the energy of the bulk vb edge is set to zero .
[ fig : dos : all],width=321 ] from the total density of states ( dos ) in fig .
[ fig : dos : all ] it can be seen that the defects n@xmath2 , v@xmath4 and n@xmath2-v@xmath3 create energy levels close the vb edge .
however , a closer inspection which is provided in the inset shows that the defect levels that originate from n@xmath2 are separated from the vb with a peak at about 0.25 ev . in the left panel of fig .
[ fig : defects_mg ] the position of the n@xmath25 defect level is shown as function of the number of nearest mg neighbors ( nn ) . here , nn-1 mg indicates a zn@xmath24mg@xmath6o@xmath26n ( or zn@xmath27mg@xmath6o@xmath26n ) supercell in which one mg atom is adjacent to the n@xmath16 ( or n@xmath16-v@xmath15 ) defect , respectively , while in the nn-2 mg and nn-3 mg supercells , two and three mg atoms are nearest neighbors of the defect .
an increasing number of nn mg atoms leads to an up - shift of the defect level .
therefore , concerning p - type conductivity , mg doping makes n@xmath16 defects less attractive since the level becomes more narrow and moves deeper into the band gap . on the other hand , the energy level position of the n@xmath28-v@xmath15 defect complex remains almost unchanged for different numbers of nn mg atoms .
the substitution of zn by mg in zno does not help to make the acceptor levels of individual n@xmath2 defects shallow , which would facilitate p - type conductivity at room temperature , but rather provokes the contrary .
moreover , it has been reported@xcite that the presence of mg atoms in the neighborhood of a n@xmath2 defect increases the defect formation energy considerably and therefore reduces the respective defect concentration in the material .
our calculations are in line with the findings of ref . , see discussion in the appendix .
zinc vacancy defects alone create very shallow defect levels ( see inset of fig .
[ fig : dos : all ] ) , which is a well known result from previous studies.@xcite experimentally , defect levels at either 130mev@xcite , 160 mev@xcite , or 190 mev@xcite are reported , depending on the specific experiment . according to fig .
[ fig : defects_mg ] , the levels of the n@xmath2-v@xmath4 defect complex lie higher above the vb edge than that of the individual v@xmath4 defect and fit to the experimentally determined values rather well . we thus propose to assign the experimentally found acceptor levels to a n@xmath2-v@xmath4 defect complex .
this complex is very promising since it provides shallow defect levels that are not separated in energy from the vb edge , and thus continuous low energy excitation are possible .
the position of the defect level varies only slightly with the number of mg atoms near the defect complex ( see right panel of fig .
[ fig : defects_mg ] ) .
mg@xmath6o@xmath7 doped with n@xmath2 ( left ) and n@xmath2 + v@xmath4 ( right ) , calculated with the sic - lda , as function of the number of nearest - neighbor mg atoms . for comparison ,
results for defects in zno are shown as dashed lines and the dos of perfect zno is included as a dotted line .
[ fig : defects_mg],width=321 ] -defect level ( left ) , the unoccupied part of the ( n@xmath16-v@xmath15)-defect level ( middle ) and the respective vb edge ( right ) , shown for the n@xmath16 atom and its most relevant neighbors .
the relative positions of these atoms , which are the ones most involved in the defect complex , are sketched on the left in fig .
[ fig : positions ] . for the oxygen and metal ions
the summed contributions are shown .
[ fig : defects ] , width=321 ] and n@xmath2 + v@xmath4 .
metal atoms on positions 46 can be either zn or mg .
position 7 corresponds to either a zn atom or a zn vacancy .
( b ) dependency of the n atom position ( height h@xmath29 ) above the atomic plane defined by the metal atoms 46 on the number of mg atoms involved .
[ fig : positions],width=283 ] in order to analyze the spatial distribution of the defect levels we have evaluated the atom - projected local densities of states ( ldos ) for the most relevant atoms involved in the defect ( complex ) . figure [ fig : defects ] displays their contributions to the defect levels while the relative locations of these atoms are sketched in fig .
[ fig : positions]a . a contribution of an individual neighbor atom to a defect level
is evaluated by integrating the ldos of the respective atom in the relevant energy interval and subsequent division by the sum of the ldos of all atoms in the supercell , integrated for the same energy range .
the energy interval [ -0.1 , 0.5]ev was chosen for the analysis of the n@xmath16 defect , while the analysis of the n@xmath16-v@xmath15 defect complex is carried out separately for the intervals [ -0.2 , 0.0]ev ( vb edge ) and [ 0.0 , 0.4]ev ( unoccupied states above the vb edge ) , c.f .
[ fig : defects_mg ] . note that the contributions for the oxygen ions ( red line ) and metal ions ( green line ) shown in fig .
[ fig : defects ] are the sums for the 3 o - atoms ( numbered 13 in fig .
[ fig : positions]a ) and the 3 or 4 metal atoms ( numbered 47 in fig .
[ fig : positions]a ) , respectively . the metal atoms at positions 46 can be either zn or mg , while position 7 corresponds to either a zn atom in the case of the n@xmath16 defect or to a zn vacancy , in the case of the n@xmath30-v@xmath15 defect complex . the n@xmath16 defect level is highly localized on the central n - atom , by more than @xmath31 , as can be seen from the left panel in fig .
[ fig : defects ] . about @xmath32
is shared by the four nn metal atoms , while the oxygen atoms of the second neighbor shell contribute with @xmath33 or less each .
the localization on the n - atom increases to @xmath34 with an increasing number of nn mg - atoms , while the total contribution of the nn metal atoms remains rather constant .
this increase in localization is connected to an increasing spatial separation of the n atom from the plane defined by the three metal atoms 46 with increasing number of mg atoms therein , see graph in fig .
[ fig : positions]b . the analysis of the spatial distribution of the n@xmath16-v@xmath15 defect ( c.f .
middle and right panel in fig .
[ fig : defects ] ) is divided into two parts , namely ( i ) the unoccupied states above the vb edge and ( ii ) the vb edge which is modified with respect to the bulk material by the presence of the defect . the unoccupied states ( middle panel in fig .
[ fig : defects ] ) are reminiscent of the n@xmath16 defect and the n atom contributes more than 40% in pure zno which increases up to 55% with increasing number of nn mg .
the hybridization with the v@xmath15 defect state results in a @xmath32 contribution of the three oxygen atoms next to the vacancy site .
the three remaining nn metal atoms ( numbered 46 in fig . [ fig : positions]a ) only contribute @xmath3523% ( zn ) and @xmath351% ( mg ) , each .
for the vb edge ( right panel in fig .
[ fig : defects ] ) , reminiscent of the v@xmath15 defect , the 3 oxygen atoms are more important ( @xmath3527% ) , but the n - atom still contributes on average with @xmath3520% .
a change in the number of nn mg atoms has only little effect on this spatial distribution with a small shift of weight from the nn metal atoms to n. for p - type conduction a fermi level close the vb edge and shallow acceptor levels in sufficiently high concentration are needed .
since the thermal excitation energy at room temperature is only about 25 mev , p - type conduction with a significant hole concentration can at best be achieved by acceptor defect levels which are distributed continuously right above the vb edge .
our sic - lda analysis of the most relevant defects in n - doped zmo shows that v@xmath4 and n@xmath2-v@xmath4 indeed provide such defect levels .
our results support the hypothesis that co - doping of zno with mg and n can be a possible route to obtain p - type conductivity .
the ( n@xmath2-v@xmath4 ) defect complex yields shallow acceptor levels which can be activated by the experimentally reported lowering of the net electron concentration in zmo.@xcite this carrier compensation effect increases linearly with the mg content and can be explained by deep - lying mg - related defects which act as charge traps.@xcite zinc vacancies are low lying energy defects which exist also in pure zno.@xcite however , pure zno usually is an intrinsically n - type semiconductor since donor defects dominate .
only grain boundaries in polycrystalline zno may provoke p - type conducting films@xcite under oxygen rich conditions .
in general the addition of oxygen is thought to reduce the number of oxygen vacancies which are supposed to be the key donor defects .
we believe that an understanding of the effect of oxygen addition has to be extended to take the grain boundaries into account . under oxygen
rich conditions the percentage of grain boundaries with undercoordinated oxygen atoms increases . in previous work@xcite
we have shown that such grain boundaries contribute shallow levels above the vb edge .
these are an equivalent to the defect levels of v@xmath4 in bulk - like zno or zmo discussed here , since the oxygen atoms at the grain boundaries are undercoordinated due to local zn deficiencies .
furthermore , nitrogen at grain boundaries can form shallow defect levels which are very similar to those shown in fig .
[ fig : defects_mg ] ( see fig . 11 in ref . ) .
we argue therefore that all these shallow levels are activated due to the depletion zones near grain boundaries.@xcite zhang et al .
interpret their experimental results along this line,@xcite and wang et al .
even suppose that oxygen excess causes the formation of a quasi two dimensional hole gas which could explain the very high mobilities seen in some experiments ( see ref . and references therein ) .
putting all from above together , this rather general defect - complex mechanism for promoting p - type conductivity in zno , zmo , and related materials emerges : defect complexes which are connected to the zinc vacancies in the bulk interior or at grain boundaries provide shallow defect levels which can serve as acceptor defects .
the activation of these defects by lowering the net electron concentration is provided either by co - doping ( e.g. n and mg@xcite ) or by depletion zones around grain boundaries .
we have studied the electronic densities of states of crystalline zno and zn@xmath0mg@xmath1o with @xmath36 .
we have shown that the sic - lda allows to calculate the electronic band gap as function of the mg content and leads to quantitatively reasonable results with respect to experimental data .
we have determined the electronic defect levels which are generated by the introduction of n@xmath2 substitutional defects .
they are at 0.25 ev above the vb or higher .
an increasing number of neighboring mg atoms to the n@xmath2 site shifts the defect level even deeper into the band gap . on the other hand ,
isolated v@xmath4 defects generate shallow levels attached to the vb edge . in combination with n@xmath2 , the defect complex n@xmath2-v@xmath4 produces low lying defect levels which emerge continuously above the vb and no detached gap states ( like in the case of the isolated n@xmath2 defect ) appear .
it is likely that these levels which are peaked at 125140 mev are connected to experimentally determined acceptor levels which are reported at 130 mev@xcite , 160 mev@xcite or 190 mev.@xcite neighboring mg atoms to the n@xmath2-v@xmath4 defect complex do not change the situation significantly .
apparently the role of mg is mainly a lowering of the net electron concentration which lowers the fermi level and therefore contributes the n@xmath2-v@xmath4 defect complexes as acceptors .
our presented results are consistent with those of an earlier study on zno grain boundaries@xcite , in which we found that at grain boundaries with local zn deficiencies ( which is equivalent to undercoordinated oxygen atoms ) , n@xmath2 defects at the grain boundary produce shallow defect levels without gap to the vb edge and are thus promising acceptor levels .
the following conclusion emerges from this work : one condition for p - type conductivity in zno , zmo , or related materials is to have enough p - type defects like for example n@xmath2 which interact with v@xmath4 point defects , or local zn deficiencies in extended defects like grain boundaries , in order to form shallow gapless defect levels above the vb edge .
moreover , the net electron concentration has to be low enough in order to make these shallow acceptor states partially unoccupied .
this second condition can be achieved by mg@xmath4/n@xmath2 co - doping in the bulk interior or by n@xmath2 doping at grain boundaries to create depletion zones .
financial support for this work was provided by the european commission through contract no . nmp3-la-2010 - 246334 ( orama ) .
increase of the defect formation energy of the n@xmath2 defect ( left ) and the n@xmath2-v@xmath4 complex ( right ) with increasing number of next neighbor mg atoms .
the energy differences @xmath37 are calculated with respect to the values obtained for bulk zno .
results from ref . are shown as a dotted line for comparison in the left panel .
the sum of the formation energies of isolated n@xmath2 and v@xmath4 ( without next neighbor mg atoms ) is shown as a dashed line in the right panel . ]
the defect formation energies of the zinc vacancy and the n@xmath2 substitutional defect in zno have recently been discussed in great detail ( see e.g. refs . [ ] and refs . therein ) . for a discussion of defect formation energies in zno obtained from sic - lda
we refer the reader to refs .
[ ] . here we want to focus on the effect of the mg atoms on the defect formation energies ( e@xmath38 ) of n@xmath2 and the n@xmath2-v@xmath4 complex .
e@xmath38 is found to depend strongly on the number of mg atoms being nearest neighbor to the defect site , which can be explained by a stronger zn - n bond compared to the mg - n bond .
figure [ fig : eform ] shows the increase in e@xmath38 with respect to bulk zno with an increasing number of next neighbor mg atoms .
our results are in line with the findings of gai _ _ et al__@xcite who have performed dft simulations using cubic zinc - blende zmo alloys .
note that the formation of a n@xmath2-v@xmath4 complex is energetically favorable by @xmath39 compared to a pair of isolated n@xmath2 and v@xmath4 point defects without mg atoms in direct vicinity of the n atom . in order to check the influence of the finite supercell size on the defect levels obtained from our dos calculations ,
we have compared calculations for three different supercell sizes , namely n=36 , 72 and 108 atoms . as shown in fig .
[ fig : supercellsize ] , there is only very little change in level position and peak shape , when comparing the results for 72 and 108 atom supercells . on the other hand ,
the 36-atom supercells do not yet yield fully converged results , although the general conclusions could qualitatively also be drawn from these rather small cells .
we therefore decided to use the 72-atom supercells for the results presented in the main part of the paper .
b. wang , j. min , y. zhao , w . sang and c. wang , appl .
. lett . * 94 * , 192101 ( 2009 ) .
a. hierro , g. tabares , j. m. ulloa , e. muoz , a. nakamura , t. hayashi and j. temmyo , appl .
. lett . * 94 * , 232101 ( 2009 ) .
w. krner and c. elssser , phys .
b * 83 * , 205315 ( 2011 ) .
h. ryoken , n. ohashi , i. sakaguchi , y. adachi , s. hishita and h. haneda , j. cryst .
growth * 287 * , 134 ( 2006 ) .
r. ghosh and d. basak , j. appl . phys .
* 101 * , 023507 ( 2007 ) .
a. kaushal and d. kaur solar energy materials & solar cells * 93 * , 193 ( 2009 ) .
h. nishinaka1 , y. kamada , n. kameyama , and s. fujita , phys .
( b ) * 247 * , 1460 ( 2010 ) .
b. laumer , f. schuster , m. stutzmann , a. bergmaier , g. dollinger , and m. eickhoff j. appl . phys . * 113 * , 233512 ( 2013 ) . c. franz , m. giar , m.heinemann , m. czerner , and c. heiliger ,
mrs proceedings * 1494 * , 57 ( 2013 ) .
e. gr , g. tabares , a. arehart , j. m. chauveau , a. hierro , and s. a. ringel , j. appl . phys . * 112 * , 123709 ( 2012 ) . c. w. teng , j. f. muth , u. ozgur , m. j. bergmann , h. o. everitt , a. k. sharma , c. jin , and j. narayan , appl .
. lett . * 76 * , 979 ( 2000 ) . w. i. park , g .- c .
yi , and h. m. yang , appl .
. lett . * 79 * , 2022 ( 2001 ) . _
semiconductors : physics of group 4 elements and 3 - 5 compounds _ , edited by k. h. hellwege and o. madelung , landbolt - brnstein , new series , group 3 , vol .
17 , pt . a ( springer , berlin 1982 ) ; _ semiconductors : intrinsic properties of group 4 elements and 3 - 5 - 7 , and 1 - 7 compounds _ , edited by o. madelung , w. von der osten , and u. rssler ,
landolt - brnstein , new series , group 3 , vol 22 , pt .
a ( springer , berlin 1987 ) .
y. q. gai , b. yao , z. p. wei , y. f. li , y. m. lu , d. z. shen , j. y. zhang , d. x. zhao , x. w. fan , jingbo li , and jian - bai xia , appl .
. lett . * 92 * , 062110 ( 2008 ) .
p. muret , d. tainoff , c. morhain and j. m cauveau , appl .
. lett . * 101 * , 122104 ( 2012 ) .
y. f. li , b. yao , y. m. lu , z. p. wei , y. q. gai , c. j. zheng , z. z. zhang , b. h. li , d. z. shen , x. w. fan , and z. k. tang , appl .
. lett . * 91 * , 232115 ( 2007 ) . | possibilities of turning intrinsically n - type oxide semiconductors like zno and zn@xmath0mg@xmath1o into p - type materials are investigated .
motivated by recent experiments on zn@xmath0mg@xmath1o doped with nitrogen we analyze the electronic defect levels of point defects n@xmath2 , v@xmath3 , and n@xmath2-v@xmath4 pairs in zno and zn@xmath0mg@xmath1o by means of self - interaction - corrected density functional theory calculations .
we show how the interplay of defects can lead to shallow acceptor defect levels , although the levels of individual point defects n@xmath2 are too deep in the band gap for being responsible for p - type conduction .
we relate our results to p - type conduction paths at grain boundaries seen in polycrystalline zno and develop an understanding of a p - type mechanism which is common to zno , zn@xmath0mg@xmath1o , and related materials . |
two - dimensional ( 2d ) materials present some fascinating challenges to condensed - matter theory , with even the most simple 2d systems harboring surprises .
one of the most famous problems involves the precise description of melting in 2d solids made up of hard , disk - like particles with short - range repulsive interactions.@xcite specifically , does the fluid undergo a weak first - order transition to the solid , or is there an intermediate hexatic phase linked by two continuous phase transitions ? related avenues of research concern the existence of exotic phases in systems made up of more complex particles , such as ( non)periodic solids of hard - disk dimers,@xcite pentamers and hexamers,@xcite tetratic phases of hard squares@xcite and hard rectangles,@xcite and orientationally ordered solids of hard pentagons and heptagons.@xcite the effects of additional interactions on the phase behavior and dynamics of 2d systems are also of interest , as evidenced by recent studies on dipolar potentials in the context of magnetic colloids.@xcite such models provide an ideal testing ground for condensed - matter theories , and in some cases challenge our most fundamental understanding of the properties of matter . despite their simplicity
, 2d models can provide reliable descriptions of some real , and rather complex , experimental situations .
for example , in a number of recent studies , 2d models have been employed to help interpret and understand the clustering and crystallization of proteins at interfaces .
the conformations and interactions of proteins are central to biological activity , and ideally one would like to investigate these properties _ in vivo_. unfortunately , structural information is most commonly obtained from x - ray diffraction studies on crystals .
there is a class of proteins , however , that can be studied under conditions resembling those _ in vivo_. membrane proteins constitute a large class of molecules found within the lipid bilayers that constitute cell walls .
they fulfill a variety of roles , such as controlling the selective transport of ions and molecules across cell membranes , or providing binding sites for other molecules on to the membrane .
the structures of membrane proteins can be studied by deposition on to a surface , alongside lipids , to form either low surface - coverages or 2d crystals ; the hydrophobic lipids help to mimic the interior of the membrane .
electron microscopy or atomic - force microscopy can then be used to image directly the clustering and packing of proteins at the solid - air interface.@xcite in many cases , the ordering of proteins can be rationalized on the basis of their gross shapes ( the way in which those shapes would ` tile the plane ' ) and the presence of specific binding interactions between domains on different molecules .
for example , the surface structure of bacteriorhodopsin ( a transmembrane protein ) is comprised of a close - packed array of trimers , each made up of monomers that resemble @xmath0 sectors of a circle .
monte carlo ( mc ) simulations of hard sectors with an additional attractive square - well potential to mimic specific binding interactions
yield insight on the self - assembly and subsequent crystallization processes.@xcite in another application , the ordering in 2d crystals of annexin v another ` triangular ' membrane protein was reproduced in simulations of a hard - disk model decorated with an appropriate orientation - dependent potential to mimic the locations of the specific binding sites on the protein . experimentally observed honeycomb and triangular structures were captured by the molecular model .
these examples show that the basic physics of large - scale structural order in 2d protein crystals can be studied with simple models , and without resorting to atomistically detailed and hence very expensive computer simulations
. there are a large number of proteins which are either inherently triangular in shape , or otherwise form trimeric structures .
for example , rotavirus inner capsid protein v6 forms trimers resembling equililateral triangles , which pack in 2d crystals ( space group @xmath1).@xcite specific fragments of prion proteins found _ in vivo _ form trimeric units that crystallize in to a 2d structure ( space group @xmath2).@xcite finally , we highlight an example in which a membrane fusion protein ( from the semiliki forest virus ) is seen to form pentagons of trimers , with the center of the pentagon raised slightly out of the plane.@xcite some semblance of local five - fold coordination can also been seen in teta a roughly triangular transporter protein at moderate surface coverages.@xcite motivated by the diversity of 2d crystal structures exhibited by trimeric protein units , and also by the observation of five - fold coordination,@xcite we have investigated the structure and phase behavior of model trimeric molecules made up of hard disks . in order to mimic specific binding interactions , such as those that might give rise to local five - fold coordination , we focus on an equililateral triangle of three hard disks at contact , in which one disk can interact with the corresponding disks on other molecules _ via _ a short - range attractive square - well potential .
as we will show below , this raises the possibility of generating orientational order within simple close - packed structures , and also offers the opportunity of forming clusters at low surface coverages . using mc simulations ,
we map out the phase diagram of the model system , and characterize the structures of the low - density clustered fluid and high - density solids which are formed at low temperature .
the remainder of the article is organized as follows . in section [ sec : model ]
we describe the molecular model , and summarize the simulation methods .
the results are presented in section [ sec : results ] , and section [ sec : discussion ] concludes the paper .
the molecular model consists of three hard disks , each of diameter @xmath3 , fused at mutual contact to form an equilateral triangle .
two of the disks on each molecule are purely repulsive , and interact with all other disks in the system through the potential @xmath4 where @xmath5 is the separation between the centers of two disks .
the third disk on each molecule carries a central attractive interaction site ; these ` attractive ' disks interact with each other _ via _ the potential @xmath6 where @xmath7 controls the range of the attraction .
this potential crudely mimics an effective attraction between vertices of the molecular triangles , which might arise through specific interactions ( e.g. hydrogen bonding , disulfide bridges , effective solvophobic interactions ) .
the parameter @xmath8 will clearly have a crucial role to play in the thermodynamics of the system .
if @xmath9 then one should anticipate a conventional phase diagram containing a vapor - liquid transition , and a fluid - solid transition .
the orientation of a trimer can be defined by a vector @xmath10 joining the geometrical center of the trimer with the center of the attractive disk .
it is unlikely that there would be any periodic orientational ordering of @xmath10 in the solid phase ; if two trimers can interact favorably irrespective of the mutual orientation , then on entropy grounds the orientations will be disordered . in the opposite extreme , @xmath11
, the molecules will feel the orientation dependence of the net trimer - trimer potential , and ultimately we might expect the vapor - liquid transition to disappear from the equilibrium phase diagram . indeed , in a pure square - well hard - sphere fluid , condensation becomes metastable with respect to freezing when @xmath12.@xcite in the present case , an interaction range @xmath13 guarantees that attractive sites must face each other directly in order to interact ; when @xmath14 it is possible for an attractive disk to be within interaction range of a trimer even if it approaches from ` behind ' . with these comments in mind
, we have chosen to study a system with @xmath15 .
the ratio of @xmath16 to the ( angle - averaged ) diameter of the trimer is smaller than that in a pure square - well hard - sphere system with the same value of @xmath8 , and assuming some sort of correspondence between two and three - dimensional systems , we do not anticipate there being a vapor - liquid transition in the equilibrium phase diagram . on the other hand , because the trimers have to attain quite specific mutual orientations in order to interact favorably ( since @xmath13 ) , we should expect to see some sort of non - trivial structure in fluid and solid phases at low temperatures .
systems of @xmath17 trimers were studied using mc simulations either in the isothermal - isobaric ( @xmath18 ) ensemble or the canonical ( @xmath19 ) ensemble.@xcite the simulation cell was rectangular with dimensions @xmath20 and @xmath21 , and area @xmath22 .
each mc cycle consisted of one translational trial move and one orientational trial move for each of @xmath23 randomly selected molecules . to help equilibrate dense phases ,
every fifth mc cycle included @xmath23 trial moves in which a randomly selected trimer was rotated by @xmath24 . in @xmath18 simulations of solid phases ,
@xmath20 and @xmath21 were varied independently ; in @xmath18 simulations of fluid phases , the simulation cell was constrained to be square . for most thermodynamic state points
typical equilibration runs consisted of @xmath25 mc cycles , but some points ( close to phase transitions ) required @xmath26 mc cycles .
production runs were typically @xmath25 mc cycles .
we define the following dimensionless units in terms of the square - well depth , @xmath27 , and the hard - disk diameter , @xmath3 : number density @xmath28 ; temperature @xmath29 ; pressure @xmath30 .
the phase diagram of the model trimers in the density - temperature ( @xmath31-@xmath32 ) plane is sketched in fig .
[ fig : phasediag ] . before detailing the determination of the phase boundaries , the characteristics of the different phases will be described .
there are four distinct regions in the phase diagram . at low density and high temperature ,
a normal fluid phase is in evidence ( fluid i ) .
a typical simulation configuration is shown in fig .
[ fig : snapshot](a ) .
there is neither translational nor long - range orientational order in the system .
phase diagram of the model trimer system in the density - temperature ( @xmath31-@xmath32 ) plane : ( solid points and solid lines ) approximate fluid - solid phase boundaries , assumed to be first order ; ( open points and dashed lines ) boundaries between high - temperature unclustered states and low - temperature clustered states , as evidenced by maxima in the heat capacity along isobars ; ( dot - dashed line ) close - packed density , @xmath33 . ] at high density and high temperature , the stable solid phase ( solid i ) possesses an orientationally disordered structure ( in the sense that @xmath34 is disordered ) with the trimers close - packed to form alternating rows displaced by @xmath35 .
figure [ fig : snapshot](b ) shows both the lack of orientational order , and the registry between alternating rows .
notice the black bonds showing how the disks are connected within the trimers ; we call this an ` @xmath36 ' structure to denote the alternating alignment of the rows .
the close - packed rows resemble those formed by vp6,@xcite although the registry between the rows is different . at the end of this section
, we will briefly discuss the possibility of solids with other close - packed structures . at low temperature and low density
we find a highly associated fluid ( fluid ii ) , in which the attractive disks aggregate to form distinct clusters .
a typical configuration is shown in fig .
[ fig : snapshot](c ) , which exhibits a broad distribution of cluster sizes . to identify clusters
, we employ the obvious criterion that two trimers with attractive disks within interaction range belong to the same cluster . with this definition in mind ,
[ fig : snapshot](c ) shows that , in general , the attractive disks within the clusters form close - packed motifs , rather than loose arrangements of disks on the circumference of a ring . for clusters of three trimers
there is no distinction , whereas for four or more trimers the close - packed arrangement is more favorable ; in a ring , each disk would have two nearest neighbors , whereas close - packed motifs can accommodate more than two direct contacts . in fig .
[ fig : cluster ] we show the probability distribution function of clusters containing @xmath37 molecules , at different pressures along an isotherm with @xmath38 . as the pressure and density increase , the distributions show peaks at progressively higher values of @xmath37 .
at the highest fluid - density shown
@xmath39 , fig .
[ fig : cluster](e ) the most probable cluster size is @xmath40 .
we had hoped that these clusters would adopt a pentagonal structure , but instead the attractive disks form ` olympic rings ' motifs , such as those shown in fig .
[ fig : snapshot](c ) .
the maximum disk - disk separation in a perfect pentagon of disks is @xmath41 , which is longer than the range of the potential studied in this work .
hence , to minimize the energy , the cluster will contract to form a close - packed structure .
perhaps pentagonal clusters would be formed in a system with @xmath42 ? ( the upper limit means that there can be no other disks between two interacting attractive disks . )
we did some test runs in the fluid phase with @xmath43 , but no pentagonal clusters were observed .
if anything , fewer distinct clusters were in evidence as compared to @xmath15 , presumably because it is less crucial that the trimers attain a specific mutual orientation in order to interact .
cluster distributions for systems along the isotherm @xmath38 : ( a ) @xmath44 , @xmath45 ; ( b ) @xmath46 , @xmath47 ; ( c ) @xmath48 , @xmath49 ; ( d ) @xmath50 , @xmath51 ; ( e ) @xmath52 , @xmath39 ; ( f ) @xmath53 , @xmath54 . in ( a)-(e )
the system is fluid , whilst in ( f ) the system is solid ( ii ) . ] upon compression of the low - temperature fluid we often encountered metastable structures , such as that shown in fig .
[ fig : snapshot](d ) .
this clearly shows a predominance of @xmath55 clusters , with the attractive disks close packed to form a parallelogram motif , but the clusters are not yet fully packed in to a solid structure .
this process is completed upon further compression , to form a @xmath56 periodic solid ( solid ii ) , a defective example of which is shown in fig .
[ fig : snapshot](e ) . in simulations of the high - density solid ii phase , the initial configuration consisted of the appropriate @xmath36 structure , but
with @xmath10 for each molecule chosen randomly from the three molecular arms ; the orientational structure shown in fig .
[ fig : snapshot](e ) develops spontaneously .
the cluster distribution for such a solid at temperature @xmath38 and density @xmath54 is shown in fig . [
fig : cluster](f ) .
the primary peak is at @xmath55 , but the presence of defects such as those shown in fig .
[ fig : snapshot](e ) gives rise to smaller ` clusters ' of attractive disks .
the fluid - solid phase boundaries were located by monitoring the equation of state @xmath57 along selected isotherms in @xmath18 simulations . for each isotherm ,
two sets of simulations were performed : a compression branch , starting from a low - density fluid configuration ; and an expansion branch , starting from the perfect solid structure corresponding to that found in the compression branch at high pressure .
portions of two representative examples ( @xmath38 and @xmath58 ) are shown in fig .
[ fig : eos ] . of course
, the fluid equations of state extend to much lower densities , but these exhibit entirely conventional behavior and hence are not shown ; in particular , there is no sign of a ` van der waals ' loop which would indicate a vapor - liquid phase transition .
the main features of interest are the apparent discontinuities in the density at what are assumed to be first - order phase transitions ( we will not open up the can of worms associated with the precise nature of two - dimensional melting and freezing ) . in fig .
[ fig : eos ] we indicate distinct fluid and solid branches in the equations of state , a number of putative metastable states ( as discussed above ) , and approximate tie - lines connecting the fluid and solid coexistence densities , obtained as follows .
the fluid branch was fitted with a virial expansion containing terms up to @xmath59 , i.e. , @xmath60 , while the solid branch was found to be fitted rather well by a simple van der waals equation@xcite of the form @xmath61 , which contains a free - volume term arising from repulsive interactions , and a mean - field term arising from the attractions .
the coexistence densities were then estimated by extrapolating the fitted branches of the equation of state to a pressure half way between those in the highest - density stable fluid and the lowest - density stable solid ; the metastable states were identified as those that did not fit on to either branch and/or for which the simulation configuration was clearly neither pure solid nor pure fluid , e.g. fig .
[ fig : snapshot](d ) .
obviously this approach provides only very rough locations for the phase boundaries shown in fig .
[ fig : phasediag ] , but some general trends are nonetheless apparent . at very low temperatures ,
the coexistence densities decrease as the system is cooled , and the transition appears to be getting weaker . at high temperatures ( @xmath62 )
the fluid coexistence density ( @xmath63 ) is very similar to the density at which the pure hard - disk fluid undergoes its transition , either to a hexatic or a solid ( disk density @xmath64,@xcite ` trimer ' density @xmath65 ) .
the apparent trimer solid coexistence density ( @xmath66 ) is significantly larger than the melting density of hard disks ( disk density @xmath67,@xcite ` trimer ' density @xmath68 ) .
equations of state along isotherms with @xmath38 ( solid symbols , solid lines ) , and @xmath58 ( open symbols , dashed lines ) : ( circles ) state points from @xmath18 simulations , with @xmath36 solid phases ; ( squares ) state points from @xmath18 simulations , with @xmath69 solid phases ( @xmath58 only ) ; ( crosses ) putative metastable state points ; ( triangles ) approximate coexistence densities ; ( lines ) fits to the fluid and solid branches ( see text ) .
the statistical errors in the @xmath18 simulation points are smaller than the symbols . ] the final piece of the equilibrium phase diagram concerns the crossover from high - temperature orientationally disordered states to low - temperature states that possess structural motifs arising from the clustering of the attractive disks . to delineate the boundary between these two regimes ,
we calculated the heat capacity appropriate to the statistical mechanical ensemble being sampled . in general
we used @xmath18 simulations to measure @xmath70 where @xmath71 is the enthalpy ( minus the kinetic contribution ) as a function of temperature along an isobar .
since clustering must be accompanied by a drop in the configurational energy , and enthalpy , a peak in @xmath72 would seem to be an obvious signal of a crossover from unclustered to clustered states . in simulations we evaluated the usual fluctuation formula , @xmath73/k_{b}t^{2}$ ] , and , as a check ,
differentiated an @xmath74 $ ] pad approximant fitted to the enthalpy as a function of @xmath75 ; @xmath76 these two approaches yielded consistent results , and the peak in @xmath72 was easy to locate accurately .
in general the peak height is less pronounced at high densities , mainly due to the fact that even in the high - temperature phase there must be some attractive disks within interaction range due to the confinement .
thus , the most difficult situation obtains at close packing of the trimers , @xmath77 . in this case
we studied a perfect close - packed @xmath36 solid , and carried out @xmath19 mc simulations with @xmath78 rotations only .
we show results for the configurational energy , @xmath79 , and the excess constant - area heat capacity , @xmath80 , in fig .
[ fig : heatcap ] . a [ 5,5 ] pad
fit provides a reliable description of the energy , and the corresponding results for @xmath81 are consistent with those obtained _ via _ the fluctuation formula .
configurational energy @xmath79 ( left ) and excess heat capacity @xmath81 ( right ) as functions of reduced temperature @xmath32 at the close - packed density @xmath82 : ( circles ) simulation results ; ( lines ) results derived from a pad [ 5,5 ] fit ( see text ) .
] in fig .
[ fig : phasediag ] we show the positions of the maxima in @xmath72 and @xmath81 at @xmath83 along with separate cubic fits to the points in the fluid and solid regions of the phase diagram .
it appears that the two branches would meet up somewhere in the fluid - solid coexistence region .
we stress that the boundaries indicated do not represent thermodynamic phase transitions ; rather , they separate different regimes of trimer association . finally , we briefly consider the possibility of the trimer system adopting other solid structures , such as the @xmath56 @xmath69 structure shown in fig .
[ fig : snapshot](f ) , in which the close - packed ( horizontal ) rows are matched with the neighboring rows . in this case , the low - temperature , orientationally ordered solid exhibits rhombic cluster - motifs containing only four attractive disks . out of those four disks
, two are interacting with two other disks , and two are interacting with three other disks .
hence , the minimum configurational energy for an @xmath69 solid is @xmath84 per trimer . in the @xmath36 structure , there are six attractive disks per parallelogram motif , of which two have two neighbors , two have three neighbors , and two have four neighbors , giving a minimum energy of @xmath85 per trimer . hence , on energetic grounds , we should expect the @xmath36 structure to be thermodynamically favored . even at high temperature
, the @xmath69 structure appears to be less stable with respect to the @xmath36 structure . as an example , in fig .
[ fig : eos ] , we show an @xmath69-solid branch of the equation of state at @xmath58 , alongside the @xmath36-solid branch . for a given pressure ,
the @xmath36 solid has the higher density which makes this state at least mechanically stable with respect to @xmath69 .
indeed , we only ever observed the fluid spontaneously freezing in to an @xmath36 structure .
although we have not performed free - energy calculations , it would be very surprising if an entropic effect could compensate for the relative energetic and mechanical stability of the @xmath36 phase with respect to the @xmath69 phase . another possible close - packed structure
is illustrated in fig .
[ fig : hexagonal](a ) , without any indication of the attractive disks .
this structure resembles that adopted by 2d crystals of teta,@xcite although we never saw this packing structure emerge from our simulations .
as far as our model is concerned , the absence of this structure at low temperature is easy to understand . in figs .
[ fig : hexagonal](b ) and [ fig : hexagonal](c ) we illustrate mirror images of the most obvious periodic arrangement of the attractive disks ( space group @xmath2 ) .
the energy per trimer is only @xmath86 , and so this is not competitive with the @xmath36 structure that is seen to emerge spontaneously in our simulations .
free - energy calculations would be of interest , particularly at high temperatures where entropy is everything !
( color online ) illustrations of an alternative close - packed structure : ( a ) without an assignment of attractive disks ; ( b ) and ( c ) mirror images of a possible structural motif for a periodic arrangement of attractive disks .
the attractive disks are colored dark gray ( red online ) , the repulsive disks are colored light gray , and all disks are drawn with diameter @xmath87 . ]
in this article we have described the structure and phase behavior of a generic model of trimeric molecules , largely motivated by recent experimental 2d microscopy studies of clustering and crystallization in triangular proteins and protein trimers .
the molecular model consists of a triangle of hard disks , with one of the disks participating in attractive square - well interactions with similar disks on other trimers .
the range of the square - well potential , @xmath16 , was @xmath88 times the disk diameter .
this system crudely mimics the general shape and specific interactions of a wide range of proteins .
the model system exhibits fluid and solid phases which , at low temperatures , possess interesting structural motifs arising from the clustering of the ` attractive ' disks . in the fluid ,
a distribution of clusters is in evidence , including tetramers , pentamers , and hexamers ( of trimers ) . in the pentamers and hexamers , the attractive disks close - pack to form ` olympic rings ' and parallelogram shapes , respectively .
we had hoped to find more open pentagonal clusters of trimers , such as those reported in ref . .
to investigate the formation of such clusters further , it might be interesting to study a system of hard isosceles triangles with the unique angle equal to @xmath89 , and a short - range attraction operating between the corresponding vertices .
in the low - temperature solid , the basic structural motif consists of clusters of six molecules , with the attractive disks close - packed to form a parallelogram .
a metastable solid possessing a motif made up of four molecules was also identified .
the fundamental difference between the two situations is the registry between neighboring close - packed rows of trimers ( @xmath36 versus @xmath69 ) . even at high temperatures ,
the orientationally disordered @xmath36 solid is at least mechanically stable with respect to the @xmath69 solid .
we identified a third structure based on hexagonal close packing , but this structure is not competitive either , at least in terms of energy .
it would be worth performing free - energy calculations to study these issues further .
finally , it is worth commenting that a diverse range of 2d structures can be generated from very simple molecular models . fully atomistic calculations of 2d protein structures are expensive , and , it could be argued , yield little insight on the fundamental physics behind clustering and crystallization . as has been shown in a variety of cases , including the present study , the process of developing and studying simple models of complex systems can yield some surprising results . | motivated by the diversity and complexity of two - dimensional crystals formed by triangular proteins and protein trimers , we have investigated the structures and phase behavior of hard - disk trimers . in order to mimic specific binding interactions ,
each trimer possesses on ` attractive ' disk which can interact with similar disks on other trimers _ via _ an attractive square - well potential . at low density and low temperature ,
the fluid phase mainly consists of tetramers , pentamers , or hexamers .
hexamers provide the structural motif for a high - density , low - temperature periodic solid phase , but we also identify a metastable periodic structure based on a tetramer motif . at high density there is a transition between orientationally ordered and disordered solid phases .
the connections between simulated structures and those of 2d protein crystals as seen in electron microscopy are briefly discussed . |
the rapidly increasing relaxation timescales , the presence of non - exponential relaxation , as well as the violation of stokes - einstein relations between viscosity and diffusivity are some features observed close to the glass transition in glassy systems @xcite .
the appearance of these features suggests that relaxation dynamics is heterogeneous , i.e. that it is faster in some regions and slower in others @xcite .
direct microscopic evidence for this behavior has been found both in simulations @xcite and in experiments @xcite .
the understanding of dynamical heterogeneity is believed to be crucial to explain anomalous behavior of materials near the glass transition , and even possibly to explain the very presence of the glass transition itself @xcite . despite many efforts trying to address the origin of dynamical heterogeneity
, this question still remains open @xcite . in recent years
, different tools have been used to probe dynamical heterogeneity .
one such tool is the dynamical susceptibility , @xmath0 , which depends both on the strength of the local fluctuations and the spatial extent of their correlations .
the peak value of @xmath0 has been observed to grow while approaching the glass transition .
however , this tool by itself can not tell us much about the origin of dynamical heterogeneity and it may be desirable to supplement it with other ways of probing the fluctuations in the system . in this work , we test the predictions of a theoretical framework that aims to describe the slow part of the fluctuations in the relaxation dynamics @xcite .
this framework is based on the hypothesis that in glassy systems the long time dynamics is invariant under reparametrizations of the time variable @xcite , but this invariance is broken , giving rise to goldstone modes that manifest themselves with the emerge of heterogeneous dynamics @xcite .
the goldstone modes correspond to fluctuations in the time reparametrization @xmath1 , i.e. @xmath2 , \label{hyp0}\ ] ] where @xmath3 is a global two time correlation function .
some indirect evidence in favor of the presence of this kind of fluctuation in structural glasses has been presented in @xcite . in this work ,
we present results for a more direct test , based on decomposing fluctuations into a _
transverse part _ satisfying eq .
( 1 ) and a _ longitudinal part _ containing all other fluctuations .
this procedure allows one to separately quantify the strength and spatial correlations of both kinds of fluctuations , as a function of temperature and timescales , for a variety of glass - forming models both below and above the mode coupling critical temperature @xmath4 .
the same kind of test can also be applied to particle tracking experimental data from colloidal and granular systems @xcite , thus allowing to investigate a possible unified explanation of dynamical heterogeneity in diverse systems .
a summary of the results of an early version of our analysis was published in ref .
@xcite .
this manuscript is organized in the following way . in sec .
[ tri ] we discuss the hypothesis and define the quantities we use to test it . in sec .
[ systems ] we present the details of the numerical simulations used to test the hypothesis and present the results of our analysis . finally , we discuss our conclusions in sec .
[ conclusions ] .
we start by discussing in general terms the expected effects of time reparametrization invariance on the fluctuations .
let s consider the global two time correlation function @xmath5\ } \right\rangle , \label{global - correlation}\ ] ] where @xmath6 is the number of particles , @xmath7 is the position of particle @xmath8 at time @xmath9 , and @xmath10 is the wave - vector that corresponds to the main peak of the static structure factor . here
@xmath11 denotes an average over thermal fluctuations , which in our case is approximated by an average over independent molecular dynamics ( md ) runs . in equilibrium , and more generally in time translationally invariant ( tti ) systems , this correlation function depends only on the difference between the two times @xmath12 . in the case of aging ,
the system is no longer tti and @xmath3 depends nontrivially on both times @xmath9 and @xmath13 . to analyze the fluctuations in the dynamics we define the local correlation as @xcite @xmath14\},\ ] ] where the average over particles in eq .
( [ global - correlation ] ) is now restricted to a region @xmath15 around point @xmath16 , which contains @xmath17 particles at time @xmath18 .
@xmath19 probes the mobility of the particles in a region near @xmath16 between times @xmath13 and @xmath9 : it is close to zero for regions where the particle configuration has changed significantly , and much closer to unity for regions where it has changed little or not at all . in this work
we will consider a slightly more restrictive version of eq .
( [ hyp0 ] ) , namely @xcite @xmath20 , \label{hyp}\ ] ] which corresponds to the case when the global correlation has the form @xcite @xmath21 .
\label{hyp - global}\ ] ] in principle , the functions @xmath22 in eq .
( [ hyp ] ) and eq .
( [ hyp - global ] ) could be different .
for example , it has been claimed @xcite that stretched exponential global relaxation could be the result of combining local exponential relaxations with different relaxation times , in which case we would have @xmath23 in eq .
( [ hyp - global ] ) and @xmath24 in eq .
( [ hyp ] ) @xcite . in this work , for simplicity
, we impose the condition @xmath25 . both this restriction and the restriction imposed in eq .
( [ hyp - global ] ) could in principle make the results appear to be slightly less consistent with the hypothesis than they would be otherwise .
the forms of the functions @xmath22 and @xmath26 can be obtained by fitting the global correlation @xmath3 , as we do in the next section .
we will assume , for the moment , that @xmath22 and @xmath26 are known .
as mentioned before , we refer to the fluctuations described by eq .
( [ hyp ] ) as _ transverse fluctuations _ and the rest of the fluctuations as _ longitudinal fluctuations _
@xcite . to visualize these ideas in more detail
, we define @xmath27 with @xmath28 @xcite , where @xmath29
. the fluctuating quantity @xmath30 is composed of a transverse contribution , @xmath31 and a longitudinal contribution , @xmath32 , i.e. @xmath33 this means that in the absence of longitudinal fluctuations , eq . ( [ hyp ] ) would be exact . in order to quantify both kinds of fluctuations ,
we define @xmath34 \nonumber \\ & & = \frac{1}{\sqrt{3 } } \left ( \phi_{21,{\vec{r}}}{}^{l } + \phi_{32,{\vec{r}}}{}^{l}-\phi_{31,{\vec{r}}}{}^{l } \right ) , \label{sigma}\end{aligned}\ ] ] @xmath35 \nonumber \\ & = & \frac{1}{\sqrt{2}}\left [ ( \phi_{21,{\vec{r}}}{}^{t}+\phi_{21,{\vec{r}}}{}^{l})- ( \phi_{32,{\vec{r}}}{}^{t}+\phi_{32,{\vec{r}}}{}^{l } ) \right ] \label{pi1}\end{aligned}\ ] ] and @xmath36 \nonumber \\ & = & \frac{1}{\sqrt{6 } } \left [ ( \phi_{21,{\vec{r}}}{}^{t}+\phi_{21,{\vec{r}}}{}^{l})+(\phi_{32,{\vec{r}}}{}^{t}+\phi_{32,{\vec{r}}}{}^{l } ) + 2(\phi_{31,{\vec{r}}}{}^{t}+\phi_{31,{\vec{r}}}{}^{l } ) \right ] , \label{pi2}\end{aligned}\ ] ] with @xmath37 .
as shown above , @xmath38 contains only longitudinal fluctuations , but @xmath39 and @xmath40 contain both transverse and longitudinal components . if eq .
( [ hyp ] ) was an exact identity , the local two - time function would verify the following relation @xmath41 then , in the case where no longitudinal fluctuations are present , the vector @xmath42 would be restricted to be fluctuating in the plane @xmath43 .
we expect that as the temperature becomes lower , the timescales become longer , and the system becomes more glassy , transverse fluctuations should become progressively more dominant , according to the hypothesis ( eq . ( [ hyp0 ] ) ) and
therefore the probability distribution @xmath44 should become anisotropic , extending mostly along the @xmath43 plane and not away from it .
an analogous set of quantities can be defined starting from the global correlation @xmath3 .
we can write the analogs of eqs .
( [ transver - eq ] ) and ( [ hyp3 ] ) , i.e. @xmath45 @xmath46=\phi_{ab}=\phi_{ab}{}^t+\phi_{ab}{}^l=\phi(t_a)-\phi(t_b)+\phi_{ab}{}^l . \label{hyp3-global}\ ] ] however , in this case , the meanings of the symbols are different . @xmath47 and @xmath48 do not fluctuate , because they are computed from the global correlation in the thermodynamic limit , which is self averaging . if eq .
( [ hyp - global ] ) was exact , then @xmath49 .
however , in practice eq .
( [ hyp - global ] ) is only approximate : in sec .
[ systems ] we fit the lhs by an expression with the form given in the rhs .
thus @xmath47 represents the part of @xmath50 that can be represented as a difference @xmath51 according to the fits , and @xmath48 represents the part of @xmath50 that the fit does not capture , or in other words a fitting error .
additionally , the global variables @xmath52 , @xmath53 and @xmath54 can be defined , by analogy to eqs .
( [ sigma ] ) , ( [ pi1 ] ) and ( [ pi2 ] ) , in the following way : @xmath55 @xmath56 and @xmath57 with @xmath37 . if eq .
( [ hyp - global ] ) was an exact identity , i.e. if the fit of @xmath3 by @xmath58 $ ] had exactly zero residuals , then @xmath59 and the global two - time function would verify the relation @xmath60+g^{-1}[c(t_3,t_2)]-g^{-1}[c(t_3,t_1)]=0 . \label{phil=0-global}\ ] ] a more extensive analysis of the fluctuations can be performed by separating longitudinal and transverse components in eqs .
( [ pi1 ] ) and ( [ pi2 ] ) .
first , we recall that @xmath52 is a purely longitudinal quantity ( see eq .
( [ sigma ] ) ) , therefore @xmath61 and @xmath62 .
now , we make the following two assumptions .
one is that the transverse and longitudinal fluctuations are not correlated to each other .
the other is that all slow fluctuations are transverse , or in other words , that longitudinal fluctuations are short range correlated in time , or at least they are correlated over times that are shorter than the shortest time interval between the configurations that are being considered .
this leads to the conditions @xmath63 and @xmath64 here @xmath65 . by using eq .
( [ transver - eq ] ) , it can be shown that the transverse components of eqs .
( [ pi1 ] ) and ( [ pi2 ] ) are given by @xmath66 and @xmath67 regarding the longitudinal components , by using eq .
( [ def1 ] ) it can be shown that @xmath68 , \label{long - sigma}\ ] ] @xmath69 \label{long - pi1}\ ] ] and @xmath70 .
\label{long - pi2}\ ] ] from the last three equations it can be easily seen that @xmath71 by using eq .
( [ def ] ) we find that , all together , we can compute the variance of the transverse fluctuations by combining the fluctuations of eqs . ( [ sigma ] ) , ( [ pi1 ] ) and ( [ pi2 ] ) in the following way @xmath72
further , we can estimate correlation volumes ( in units of the coarse graining volume @xmath73 ) by using the formula @xmath74,\ ] ] where @xmath75 , @xmath76 is a local coarse grained variable , @xmath77 is the spatial average of @xmath76 , and @xmath78 is the volume of the system .
therefore , by using these equations together with eq .
( [ var - tran ] ) we can estimate the correlation volume of transverse and longitudinal fluctuations , respectively given by @xmath79 and @xmath80 if the time reparametrization hypothesis is correct , we expect that the variance as well as the correlation volume of the transverse fluctuations will grow together with those corresponding to the dynamical heterogeneities .
we also expect that the variance and correlation volume of the longitudinal fluctuations should be insensitive to changes in the variance and correlation volume of the dynamical heterogeneities .
we performed classical molecular dynamics simulations in systems of @xmath6 particles that were initially equilibrated at high temperature @xmath81 ( where @xmath4 is the mode coupling critical temperature @xcite ) , then instantaneously quenched to a final temperature @xmath82 and allowed to evolve for times several orders of magnitude longer than their typical vibrational times @xcite .
we determined @xmath4 by fitting @xmath83 , where @xmath84 is the equilibrium @xmath85-relaxation time defined by the conditions @xmath86 and @xmath87 , as shown in ref .
we generated eight datasets by simulating four atomistic glass - forming models @xcite .
two of the systems are 80:20 mixtures of a and b particles , interacting via either lennard - jones ( lj ) potentials @xcite or via purely repulsive weeks - chandler - andersen ( wca ) truncation of the lj potentials @xcite .
the interactions in the particle systems have the same length parameters , @xmath88 ( @xmath89 ) , and energy parameters , @xmath90 as in the standard kob - andersen mixture , namely @xcite : @xmath91 , @xmath92 , @xmath93 , @xmath94 , @xmath95 and @xmath96 .
the other two systems are models of short ( 10-monomer ) polymers , in which all particles interact with each other via either lj potentials or via wca potentials , with length parameter @xmath91 and energy parameter @xmath94 .
additionally , in our polymer models , nearest - neighbor monomers along a chain are connected by a fene anharmonic spring potential .
for both particle and polymer systems , the lj potential is truncated at the cut - off distance @xmath97 and the wca potential is truncated at @xmath98 .
we choose the unit of length as @xmath99 , the unit of energy as @xmath100 and the unit of time as @xmath101 . for the particle systems ,
the simulations were performed in an nvt ensemble , with the temperature being fixed by rescaling the velocities every 50 timesteps . for the polymer systems ,
the simulations were performed in an npt ensemble , with both the pressure and the temperature being controlled by the nose - hoover method .
the details of the simulations are summarized in table [ tabla ] .
c c c c r c l c + & & & & & + & & & & & & & + & & & & & & + a & 8000 & 10 & lj & 0.6 & 0.833&@xmath102 & 100 + [ 0.5ex]b & 8000 & 10 & wca & 0.4&0.503 & @xmath103 & 800 + [ 0.5ex]c & 8000 & 1 & lj & 0.4&0.435&@xmath104 & 250 + [ 0.5ex]d & 1000 & 1 & wca & 0.236 & 0.263&@xmath104 & 5000 + [ 0.5ex]e & 1000 & 1 & wca & 0.29 & 0.263&@xmath105 & 9000 + [ 0.5ex]f & 1000 & 1 & wca - eq & 0.29&0.263 & @xmath105 & 9000 + [ 0.5ex]g & 1000 & 1 & wca & 0.4&0.263&@xmath106 & 4999 + [ 0.5ex]h & 1000 & 1 & wca - eq & 0.4&0.263&@xmath107 & 4999 + [ 1ex ] [ tabla ] we begin by identifying the functions @xmath22 and @xmath26 ( eq . ( [ hyp - global ] ) ) that best describe the global correlations @xmath3 computed from our datasets by using eq .
( [ global - correlation ] ) .
we find that , for all datasets , the two time correlation @xmath3 can be well fitted by using the form @xmath108 $ ] .
however , as shown in the inset of fig .
[ global ] , the relaxation for different systems presents different behaviors , which leads to different forms of @xmath26 .
the best fits of @xmath3 that we obtained correspond to the following forms : for aging polymer systems @xmath109 , for aging particle systems @xmath110 , and for equilibrium @xmath111 .
we can verify our proposed eq .
( [ hyp - global ] ) for the different cases by using the form of the obtained functions @xmath22 and @xmath26 . for equilibrium
we trivially recovered , as expected , the case of tti . in the aging cases
we can verify that eq .
( [ hyp - global ] ) can be rewritten in the form @xmath112 $ ] @xcite , which is found in many aging systems @xcite .
once the fitting procedure is performed for a given dataset , we can use the known values of the parameters @xmath113 , @xmath114 and @xmath115 to compute @xmath116=\theta_0\{-\ln[q_{ea}^{-1}c(t_a , t_b)]\}^{1/\beta}\ ] ] and , by using eqs .
( [ sigma - global][pi2-global ] ) , to compute @xmath52 , @xmath53 , and @xmath54 .
[ global ] shows the results of plotting the global values of @xmath52 against the global values of @xmath117 for all times , @xmath118 , and all systems .
as discussed before , since the fits are not perfect , the results do not fall exactly on the line @xmath59 , but the collapse and the fits are good enough to allow us to test the hypothesis .
pairs for all datasets and all possible times @xmath119 .
_ inset : _ relaxation time @xmath120 as a function of the waiting time @xmath121 for all the models and temperatures considered .
@xmath120 is defined by the condition @xmath122 . ] as mentioned in the previous section , if eq .
( [ hyp0 ] ) is satisfied , we expect the local quantities to satisfy @xmath43 . in fig .
[ local - contours ] we show the results of plotting for all systems the 2d projection of the joint probability density of the coarse - grained local correlations , @xmath123 , with @xmath124 and @xmath125 .
the values of the global quantities are subtracted from the local quantities to avoid trivial effects due to differences in the global values . by doing this
we are able to better compare the contours independently of the choice of @xmath126 and @xmath127 .
we do , however , keep the value of @xmath128 approximately the same for all systems , in this case @xmath129 .
the three contours shown for each dataset enclose respectively @xmath130 , @xmath131 and @xmath132 of the total probability .
we coarse grain over moderately large regions , containing on average 125 particles , in order to detect collective fluctuations and , since time reparametrization symmetry is a long time asymptotic effect , we choose the times as late as possible . as the time reparametrization hypothesis predicts , for datasets with @xmath133 ( a - d ) , the purely longitudinal fluctuations @xmath134 are clearly smaller than the @xmath135 fluctuations , which contain both transverse and longitudinal contributions .
this behavior is more noticeable for the @xmath136 contour , which encloses the most likely fluctuations , than for the @xmath137 and @xmath132 contours .
for moderately higher temperature , @xmath138 ( datasets e and f ) , the anisotropy @xmath139 is still present , but less pronounced . in the case of the highest temperature , @xmath140 ( datasets g and h ) ,
the anisotropy is either very slight , or absent . in the case of the systems that are equilibrated ,
f and h , we find that the shapes of their contours are similar but slightly more anisotropic then the ones obtained for the same temperature in the aging regime , e and g , respectively .
the effect of temperature in the anisotropy of the contours can be observed in more detail in fig .
[ fig : comparison](a ) where the @xmath130 contour of the probability density for the systems of particles with wca interactions is shown for three temperatures ( datasets d , e , g ) .
this can be directly connected to the fact that , as the temperature is increased , the separation of timescales is less pronounced , the finite time corrections to the time reparametrization symmetry become larger , and the effects of local time variable fluctuations become weaker .
the same trends can be observed in figs .
[ fig : comparison](b ) and [ fig : comparison](c ) for the same datasets as in fig .
[ fig : comparison](a ) but for different values of the global correlation @xmath128 . , computed using coarse graining boxes containing 125 particles on average . by subtracting the global quantities to the local quantities we avoid trivial effects due to differences in the global values .
each panel from a to h contains results from the corresponding dataset , for @xmath141 , with the times chosen as late as possible within each dataset .
each set of three concentric contours is chosen so that they enclose @xmath142 and @xmath132 of the total probability . ] of the probability for the aging wca systems at three different temperatures ( d , e and g ) .
the three different panels correspond to different values of @xmath128 : ( a ) @xmath141 , which is the same value as in fig .
[ local - contours ] , ( b ) @xmath143 , and ( c ) @xmath144 . ]
we can further analyze the effects of choosing different conditions from the ones chosen in fig .
[ local - contours ] , for instance , by comparing the results shown in fig .
[ local - contours ] with results obtained for smaller coarse graining regions or for shorter times in the aging regime .
exactly this kind of comparison is shown in fig .
[ diff ] , where the @xmath130 probability contours for dataset b are shown for three conditions .
the contour labeled b is the one shown already in fig .
[ local - contours ] .
the contour labeled b corresponds to the same time , but with coarse graining regions containing on average 23 particles instead of 125 .
this leads to less averaging and stronger fluctuations , but also , since fluctuations correlated over shorter distances are no longer preferentially suppressed , the shape of the contour is no longer dominated by collective modes , and thus contour b extends more in the direction of @xmath145 than contour b. the contour labeled b corresponds to the same coarse graining size of contour b , but with much shorter times .
this leads to stronger finite time effects , analogous to the ones found at slightly higher temperatures , and as expected the contour is less anisotropic , and indeed , it resembles the contours corresponding to @xmath146 .
probability contours for dataset b with @xmath141 .
contour b corresponds to the @xmath130 of the probability density shown in fig .
[ local - contours ] .
contour b corresponds to the same times as contour b , but with a much smaller coarse graining size .
contour label b corresponds to the same coarse graining size as b , but with much shorter times . ]
we now move to a more quantitative analysis of the _ strength _ and _ spatial correlations _ of the fluctuations by making use of the results derived in sec .
[ tri ] . in fig .
[ fig : evolution - tw](a ) we show the ratio between the variances of local transverse fluctuations and longitudinal fluctuations , @xmath147 ( variance anisotropy ratio " ) ( see eq .
( [ var - tran ] ) ) , as a function of the initial time @xmath148 . , for @xmath149 .
( b ) ratio between the correlation volumes of transverse and longitudinal fluctuations as a function of @xmath148 , for the value @xmath149 . ]
similarly , in fig .
[ fig : evolution - tw](b ) we plot the ratio between the correlation volumes of transverse and longitudinal fluctuations , @xmath150 ( correlation volume anisotropy ratio " ) ( see eqs .
( [ vol - trans ] ) and ( [ vol - long ] ) ) , also as a function of @xmath148 .
we find that for aging systems both ratios grow as @xmath148 increases , as one could expect from the fact that at later times the reparametrizations symmetry breaking terms in the action should become progressively weaker @xcite .
for the equilibrium datasets , the dynamics is time translation invariant ( tti ) , and we observe , as expected , that both anisotropies are independent of @xmath148 . in fig .
[ fig : evolution ] , we show the same two ratios as functions of the strength of the dynamical heterogeneities , measured by the dynamical susceptibility @xmath151 $ ] . we observe that both anisotropy ratios grow when @xmath152 increases , i.e. as the dynamical heterogeneity becomes more pronounced .
although the same qualitative behavior is observed for all datasets , the curves are different for different systems and temperatures .
, for the value @xmath149 .
( b ) ratio between the correlation volumes of transverse and longitudinal fluctuations as a function of @xmath153 , for @xmath149 . ] the results presented in fig . [
fig : evolution ] correspond to the value @xmath154 , but similar results can be obtained for different values of @xmath128 , as one can expect from fig .
[ fig : comparison ] . results for three different values of @xmath128 are shown in fig . [
fig : diff - c13 ] , where the variance anisotropy ratio for datasets b and d is plotted as a function of @xmath152 for the values @xmath155 @xmath156 , @xmath157 and @xmath158 .
this figure shows that the variance anisotropy ratio grows with the strength of the dynamical heterogeneity , for all three fixed values of @xmath128 .
, for systems b and d for the values @xmath141 , @xmath157 , and @xmath158 . ]
according to the hypothesis , dynamical heterogeneity originates in the goldstone modes associated to fluctuations in the time reparametrization ( see eq .
( [ hyp ] ) ) .
therefore , the hypothesis implies that the correlation volume of the dynamical heterogeneity should be similar to the correlation volumes of the transverse components of the variables @xmath53 and @xmath54 , and the longitudinal variable @xmath52 should be less correlated in space .
our results , shown in fig . [ fig : corr - vol ] , show that this is indeed the case .
the correlation volume corresponding to the transverse fluctuations , @xmath159 , closely tracks the one corresponding to the dynamical heterogeneities , @xmath160 , and they both grow together as the temperature is reduced or the timescale is increased . by contrast , for the longitudinal fluctuations @xmath52 we find that their correlation volume @xmath161 is small and essentially constant ; it does not correlate with the correlation volume of the dynamical heterogeneity , nor with the temperature or the time scale .
in fact , despite the large error bars and the presence of some outlier points , the figure shows a partial data collapse between different systems , both for the case of transverse and for the case of longitudinal correlation volumes . in the case of longitudinal fluctuations
, this may be a trivial effect due to the correlation volumes being smaller than the volume of the coarse graining regions used to define the variables . in the case of the transverse fluctuations ,
the correlation volumes go well beyond the volume of the coarse graining regions , and the partial collapse in the results might be evidence of some sort of universality , but more work will be needed in order to decide this question one way or another . , the correlation volume for the fluctuations of @xmath162 , calculated for the value @xmath149 . ]
in this paper we tested the hypothesis that dynamical heterogeneity arises from goldstone modes related to a broken continuous symmetry under time reparametrizations .
in other words , we tested whether dynamical heterogeneity is associated with the presence of spatially correlated fluctuations in the time variables . to verify this , we studied probability distributions that allowed us to distinguish between time reparametrization fluctuations ( transverse fluctuations ) and other fluctuations ( longitudinal fluctuations ) .
we also tested for possible correlations of both the strength and the correlation volume of the fluctuations in the time variable with the dynamical susceptibility @xmath0 , which is normally used to probe dynamical heterogeneity .
altogether , we found that at the lowest temperatures , for the longest timescales and for the largest coarse graining lengths , the transverse fluctuations became stronger than the longitudinal fluctuations , which is consistent with the hypothesis .
we also found that the correlation volumes of the time reparametrization fluctuations were proportional to the correlation volume of the dynamical heterogeneity , while the correlation volumes of the longitudinal fluctuations were small and independent of the correlation volumes of the dynamical heterogeneity .
these observations apply to all the systems examined , regardless of the details of the interaction ( purely repulsive in the case of wca vs. repulsive and attractive in the case of lj ) , the kinds of objects ( binary systems of particles vs. systems of short polymers ) , or the ensembles used in the simulations ( nvt vs. npt ) .
all of this was despite the fact that , to simplify the analysis , we imposed some extra conditions on the form of the correlations , which may have made the agreement with the hypothesis appear less good than it would have been otherwise . with regards to universality , the evidence we found was mixed .
on the one hand , there were clear differences in the details of the results for different systems , for example for the anisotropy ratios .
on the other hand , all the trends we observed were the same across systems , and the results for the correlation volumes did show some hints of universality , although the relatively large noise in this measurement did not allow for definite conclusions to be drawn . in any case
, the commonality in the results is strong enough to suggest that other systems may display similar qualitative behaviors .
thus , we expect that it would be very instructive to apply the same kind of test to data from other slowly relaxing systems , such as particle tracking data from glassy colloids @xcite and from granular systems close to jamming @xcite .
finally , considering the success of the tests presented here , it becomes natural to ask if it is possible to extract from the data the actual fluctuating reparametrization @xmath163 , and to study its properties directly .
in fact , ref .
@xcite shows that some progress can be made in that direction .
h. e. c. thanks e. flenner , m. kennett , g. szamel , and e. weeks for discussions and c. chamon and l. cugliandolo for suggestions and for many discussions .
this work was supported in part by doe under grant de - fg02 - 06er46300 , by nsf under grants phy99 - 07949 and phy05 - 51164 , and by ohio university .
numerical simulations were carried out at the ohio supercomputing center .
h. e. c. acknowledges the hospitality of the aspen center for physics and the kavli institute for theoretical physics , where parts of this work were performed .
h. e. castillo , c. chamon , l. f. cugliandolo , j. l. iguain , and m. p. kennett , phys .
b * 68 * , 134442 ( 2003 ) . c. chamon , m. p. kennett , h. e. castillo , and l. f. cugliandolo , phys . rev . lett . *
89 * , 217201 ( 2002 ) .
h. e. castillo , c. chamon , l. f. cugliandolo , and m. p. kennett , phys .
lett . * 88 * , 237201 ( 2002 ) . c. chamon , p. charbonneau , l. f. cugliandolo , d. r. reichman , and m. sellitto , j. chem
. phys . * 121 * , 10120 ( 2004 ) . h. e. castillo and a. parsaeian , nature physics * 3 * , 26 ( 2007 ) .
a. parsaeian and h. e. castillo , phys .
e * 78 * , 060105(r ) ( 2008 ) .
a. parsaeian and h. e. castillo , phys .
lett . * 102 * , 055704 ( 2009 ) .
a. parsaeian and h. e. castillo , arxiv:0811.3190 ( 2008 ) .
k. e. avila , h. e. castillo , and a. parsaeian , phys .
* 107 * , 265702 ( 2011 ) .
j. p. bouchaud , l. f. cugliandolo , j. kurchan , and m. mzard .
young , _ spin glasses and random fields ( world scientific , 1998 ) .
_ our notation differs from the one used in @xcite . here
, we use @xmath164 $ ] instead of @xmath165 $ ] .
this corresponds to the changes of variables @xmath166 $ ] and @xmath167 .
also , in this work , our definition of @xmath50 corresponds to the logarithm of the definition of @xmath50 used in @xcite . | we test a hypothesis for the origin of dynamical heterogeneity in slowly relaxing systems , namely that it emerges from soft ( goldstone ) modes associated with a broken continuous symmetry under time reparametrizations .
we do this by constructing coarse grained observables and decomposing the fluctuations of these observables into transverse components , which are associated with the postulated time - fluctuation soft modes , and a longitudinal component , which represents the rest of the fluctuations .
our test is performed on data obtained in simulations of four models of structural glasses .
as the hypothesis predicts , we find that the time reparametrization fluctuations become increasingly dominant as temperature is lowered and timescales are increased .
more specifically , the ratio between the strengths of the transverse fluctuations and the longitudinal fluctuations grows as a function of the dynamical susceptibility , @xmath0 , which represents the strength of the dynamical heterogeneity ; and the correlation volumes for the transverse fluctuations are approximately proportional to those for the dynamical heterogeneity , while the correlation volumes for the longitudinal fluctuations remain small and approximately constant . |
recent analyses of hypernuclear nonmesonic decay reveal the importance of short - range interaction of baryonic weak transition @xcite .
it is , in fact , realized that the momentum transfer is large due to the @xmath10 mass difference , and therefore the short range interaction becomes important .
we proposed to treat the short range part of @xmath7 interaction using the valence quark picture of the baryon and the effective four - quark weak hamiltonian @xcite .
we found that the direct quark ( dq ) process gives significantly large contribution and shows qualitatively different features from the meson exchange contribution , especially in its isospin structure .
the kaon exchange contribution also largely enhance the @xmath11 ratio due to its spin structure @xcite . there remains , however , another problem , i.e. , experimental values of the proton asymmetry from polarized hypernucleus are incompatible with the theoretical values , which is closely related to the @xmath8 transition amplitudes @xcite .
the @xmath8 amplitude is relatively small and therefore masked by the @xmath9 amplitude , but it is important to clarify the breaking of the @xmath12 rule for the nonmesonic weak decay of hypernuclei .
in fact , the quark model treatments of @xmath7 transition predict large breaking of the @xmath13 rule in @xmath8 transition channel @xcite . for the case of double-@xmath6 hypernuclei ,
the nonmesonic decay is induced not only by the nucleon but also by the other @xmath6 hyperon : @xmath14 .
the two - body part of the nonmesonic decay rate of double-@xmath6 hypernuclei is classified as @xmath15 where @xmath16 stands for the @xmath17 transition which is available when the virtual @xmath18 exists @xcite .
the @xmath19 represents the @xmath0 transition , which is a characteristic transition for the multi-@xmath6 system .
experimentally , the @xmath0 transition should be distinguishable from the @xmath7 decay mode because the hyperon with high momentum is emitted once it occurs in double-@xmath6 hypernucleus .
thus the hyperon - induced decay mode provides us with valuable information on the @xmath20 baryon - baryon interaction @xcite . in sec.[sec : trr ] , the hyperon - induced decay rate , @xmath19 , is expressed in terms of two - body amplitudes corresponding to the @xmath0 transition . in sec.[sec : pot ] , we construct the meson exchange and dq potential for the @xmath0 transition . in sec.[sec : res ] , we show the result of the @xmath0 transition rate in @xmath1he .
finally , sec.[sec : con ] is devoted for the conclusion of this work .
for the nonmesonic weak decay of double-@xmath6 hypernuclei , the hyperon induced decay mode , @xmath0 , becomes possible besides the usual @xmath7 transition .
now we concentrate on the novel @xmath0 decay .
the @xmath5 transition rate is given by @xmath21 where @xmath22 is the @xmath5 transition amplitude , @xmath23 is the total spin of initial hypernucleus , and @xmath24 and @xmath25 are momenta of emitted particles , _
i.e. _ , hyperon and nucleon .
the summation indicates a sum over all quantum numbers of the initial and final particle systems .
the @xmath26-function for the energy conservation is given by @xmath27 where the hyperon binding energy is taken as @xmath28mev @xcite .
after the decomposition of angular momentum , the explicit form of @xmath22 is @xmath29 where @xmath30 is the transition potential and @xmath31 is the wave function of the @xmath32 or yn two - body system in the configuration space .
the @xmath33 , @xmath34 , and @xmath35 indicate the angular momentum , spin , and total spin for two - body system , respectively . in order to extract the @xmath32 state from the double-@xmath6 hypernucleus , we assume , for simplicity , that the @xmath32 state with @xmath8 and @xmath36 is decoupled from the residual nuclei : @xmath37 where the double-@xmath6 state is always in @xmath38 state in the ground state . the important point of the study
the @xmath0 process is that the initial @xmath32 pair is in the @xmath38 state and thus the @xmath0 transition occurs only in the @xmath8 channel .
the @xmath8 transition amplitude is not well studied in the @xmath7 decay , because the @xmath9 amplitude is much larger .
in contrast , the @xmath0 decay directly probes the @xmath8 transition amplitude and therefore plays a complementary role .
the initial double-@xmath6 state is written in terms of the shell model wave function of @xmath39 state : @xmath40^{j=0 } = \left [ \phi_{0s}^{b_r}(\vec{r } ) \times \phi_{0s}^{b_r}(\vec{r } ) \right]^{j=0 } \label{eq : wf}\ ] ] where @xmath41 is assumed to be a gaussian with size parameter @xmath42 .
the last expression of eq.([eq : wf ] ) is the wave function in terms of the relative and center - of - mass coordinate with @xmath43 and @xmath44 .
the @xmath45 and @xmath46 are @xmath47 and @xmath48 , respectively .
the wave function for the final @xmath49 state is assumed as the plane wave with @xmath50 and @xmath51 .
the @xmath42 parameter is taken as @xmath52fm which fit on the r.m.s .
radius of @xmath6 in @xmath1he @xcite .
in addition , we introduce phenomenological correlation functions . for the initial @xmath32 wave function ,
we multiply to the uncorrelated wave function @xmath53 with @xmath54fm , @xmath55@xmath56 , @xmath57fm , @xmath58 , according to ref.@xcite . on the other hand , for the final @xmath59 state
, we multiply @xmath60 where @xmath61 is a spherical bessel function with @xmath62 @xmath63 .
in order to describe the weak @xmath64 transition , we adopt the quark - meson hybrid model , which is focused on the quark degrees of freedom in a short - range region . at long and medium distances ,
the transition is induced mainly by the one pion exchange ( ope ) and one kaon exchange ( oke ) processes .
the general form of @xmath64 transition potential in the meson exchange picture is given by @xmath65 { \frac{1}{{\vec q}^2 + { \tilde m}_{i}^2 } } \left ( { \frac { \lambda_{i}^2 -{\tilde m}_{i}^2 } { \lambda_{i}^2 + { \vec q}^2 } } \right)^2
g_w [ \bar{u}_n ( a+b \gamma_5 ) u_\lambda ] \label{eq : ope}\ ] ] where the coupling constants @xmath66 , @xmath67 , @xmath68 and @xmath69 , shown in table [ tab : swcc ] , are chosen properly for each transition . a monopole form factor with cutoff parameters , @xmath70 = 800 mev and @xmath71 mev ,
are employed for each vertex . as the energy transfer is significantly large ,
we introduce the effective meson mass : @xmath72 the kaon exchange potential can be constructed similarly .
both the strong and weak coupling constants are evaluated under the assumption of the flavor su(3 ) symmetry .
note that , for oke , it involves the strangeness transfer and thus the strong and weak vertices should be exchanged .
.the strong and weak coupling constants for the @xmath2 and @xmath3 exchange potentials .
the strong coupling constants are taken from nijmegen strong couplings @xcite .
the weak coupling constants are in unit of @xmath73 .
[ cols="^ , < , < , > , < , < , > , < , < , > " , ] [ tab : he6ll ] we compare our results with the previous calculations carried out by itonaga et al .
@xcite and parreo et al .
we all agree the bare pion - exchange amplitudes in the @xmath74 transitions @xcite . in table
[ tab : he6ll ] , ref .
@xcite includes effects of two - pion exchange in the scalar ( @xmath75 ) channel as well as the vector ( @xmath76 ) channel .
@xcite contains all the pseudoscalar mesons and the vector mesons exchanged between @xmath32 .
ref.@xcite took two different approaches for the final state interaction . the full calculation ( f )
is based on the scattering @xmath77 matrix in the final states , while an effective treatment ( e ) evaluates the transition amplitudes by multiplying a correlation function to the relative wave function .
they pointed out that the full calculation enhances @xmath78 while it reduces @xmath79 .
the effective treatment does not include the conversion from @xmath80 to @xmath81 . in the present work , we make an effective treatment of the final state interactions .
therefore , our results have to be compared to the ones in ref.@xcite(e ) .
our calculation is similar to their effective treatment and thus the @xmath78 enhancement due to the final state interaction is not shown .
the total decay rates in our results are reduced by a factor about two due to the larger @xmath42 parameter for the @xmath6 wave function .
we note that if the same @xmath42 value is used , our results agree with the ( e ) results in ref.@xcite .
we also point out that the ratio @xmath82 is fixed to @xmath83 by the @xmath12 rule in the meson exchange calculations .
in the double @xmath6 hypernuclear system , the novel decay mode , @xmath0 , is allowed .
the hyperon - induced decay rates are much smaller than the decay rate of free @xmath6 ( @xmath84 ) .
this is because the @xmath0 transition occurs only in the @xmath8 channel . although the @xmath0 transition is not easy to detect experimentally , it is important to extract the information about the @xmath8 transition .
our result shows that the dq process gives significant contribution to the @xmath0 transition .
it strongly suppresses the pc amplitude for the @xmath85 transition and the pv amplitude for the @xmath86 transition .
furthermore the dq largely breaks the empirical @xmath12 rule . unlike the @xmath7 transition
, we have an opportunity to check directly the breaking of @xmath12 dominance for the @xmath86 transition .
the authors acknowledge profs .
a. parreo , k. itonaga , h. nemura , and y. akaishi for variable discussions .
one of authors , k. s. , thanks to jsps research fellowship for financial support .
this work is supported in part by grants - in - aid for scientific research ( no.11640261 ) from japan society for the promotion of science ( jsps ) .
t. inoue , s. takeuchi , and m. oka , nucl . phys . *
a577 * ( 1994 ) 281c ; + _ ibid _ * a597 * ( 1996 ) 563 a. parre@xmath87o , a. ramos , and c. bennhold , phys . rev . *
c56 * ( 1997 ) 339 ; + a. parre@xmath87o and a. ramos , phys . rev .
* c65 * ( 2002 ) 015204 .
t. inoue , m. oka , t. motoba , and k. itonaga , nucl .
* a633 * ( 1998 ) 312 .
k. sasaki , t. inoue , and m. oka , nucl .
* a669 * , 331 ( 2000 ) ; + erratum , _ ibid _ * a678 * ( 2000 ) 455 .
d. jido , e. oset , and j. e. palomar , nucl . phys .
* a694 * ( 2001 ) 525 .
k. itonaga , t. ueda , and t. motoba , phys .
* c65 * ( 2002 ) 034617 .
k. sasaki , t. inoue , and m. oka , nucl .
a707 * ( 2002 ) 477 .
h. nabetani , t. ogaito , t. sato , and t. kishimoto , phys . rev .
* c60 * ( 1999 ) 017001 .
k. maltman and m. shmatikov , phys .
b331 * ( 1994 ) 1 .
k. itonaga , t. ueda , and t. motoba , nucl . phys .
* a691 * ( 2001 ) 197c , and private communications .
a. parreo , a. ramos , and c. bennhold , phys . rev . *
c65 * ( 2002 ) 015205 .
h. takahashi et al . , phys .
* 87 * ( 2001 ) 212502
. y. akaishi , and h. nemura , private communications th .
a. rijken , v. g. j. stoks , and y. yamamoto , phys . rev .
* c59 * ( 1999 ) 21 .
v. g. j. stoks and th .
a. rijken , phys .
* c59 * ( 1999 ) 3009 .
k. itonaga , and a. parreo , private communications | we calculate the @xmath0 transition rate of @xmath1he by the hybrid picture , the @xmath2 and @xmath3 exchanges plus the direct quark processes . it is found that the hyperon - induced decay is weaker than the nucleon - induced decay , but the former may reveal the short - range mechanism of the weak transition and also give a clear signal of the strong @xmath4 transition .
the @xmath5 transition in double-@xmath6 hypernucleus is complement to the @xmath7 transition as it occurs only in the @xmath8 channel , while the @xmath9 transition is dominant in the @xmath7 case . , , and hypernuclei , nonmesonic decay , direct quark process 21.80.+a , 12.39.-x |
the higgs boson is the last remaining ingredient of a complete standard model .
it s persistent elusiveness is perhaps not surprising . within the framework of the standard model
, there are no symmetries which can be invoked to make a fundamental scalar light .
the existence of a light scalar degree of freedom which remains fundamental above the weak - scale would argue for supersymmetry since supersymmetry provides the only explicitly known solution to the naturalness problem which accompanies fundamental scalars @xcite .
of course , the higgs boson may not be fundamental at all , and the only testament to its existence may be the eventual unitarization of the longitudinal @xmath5 scattering cross section at tev scale energies .
however , although no vestige of the higgs boson may be seen until the lhc , a failure to observe a higgs boson in pre - lhc experiments could significantly challenge the principle motivation for weak - scale supersymmetry , at least in its minimal forms .
if nature is supersymmetric above the weak - scale , the allowable range of higgs boson masses is considerably restricted . in the minimal supersymmetric extension of the standard model ( mssm )
, the lightest higgs boson lies below @xmath6 at tree level , @xmath7 where @xmath8 is the ratio of higgs boson vacuum expectation values .
quantum corrections can lift the light higgs boson mass above @xmath6 @xcite , but the magnitude of these corrections are restricted if supersymmetry provides a successful solution to the naturalness problem .
radiative corrections to the light higgs boson mass in supersymmetry have been calculated by many authors @xcite . from these corrections ,
upper bounds for the lightest higgs boson mass have been computed either by choosing arbitrary heavy masses for superpartners or by demanding the theory remains perturbative up to some high scale @xcite .
while these upper bounds reasonably approximate an important , unexceedable upper - limit on the higgs boson mass , they do not provide a complete picture of our expectations for the mass of the lightest higgs boson in supersymmetric models .
realistically , we expect the higgs boson mass to be significantly lighter . to achieve higgs boson masses as heavy as these upper - bounds requires some or all superpartner masses to be much heavier than the weak - scale .
the appearance of this heavy mass scale in turn requires demonstrably large , unexplained cancellations among heavy masses in order to maintain a light weak - scale .
however , avoiding this fine - tuning is the principle reason that supersymmetry was introduced at the weak - scale . in this article
, we observe that it would be quite unnatural for the lightest higgs boson mass to saturate the maximal upper bounds which have been previously computed .
we compute the natural upper bound on the higgs boson masses in minimal , low - energy supergravity ( mles ) , and we show the extent to which naturalness is lost as the experimental lower bound on the lightest higgs boson mass increases .
section two provides a brief review of naturalness and how it is reliably quantified .
an analysis of the natural upper bound on the higgs boson mass follows in section three .
we find that for @xmath9 gev , if @xmath10 gev , minimal low energy supergravity does not accommodate the weak - scale naturally .
moreover , in the _ most _ natural cases , @xmath11 gev when @xmath9 gev . for modest @xmath2 ,
the natural upper - bound is even more restrictive . in particular , for @xmath12 and @xmath9 gev , if @xmath13 gev large fine - tuning is required , while the _ most _ natural values of the higgs boson mass lie below @xmath6 .
this has important implications for challenging weak - scale supersymmetry at collider experiments .
in particular , if the lightest supersymmetric higgs boson is not observed at cern s @xmath14 collider lep - ii , requiring natural electroweak symmetry breaking in mles will progressively increase the lower bound on @xmath2 as lep - ii increases in energy . in the _ most _ natural cases ,
if the energy of lep - ii is extended to @xmath15 gev , a light higgs boson would be observed provided it decays appreciably to @xmath16 , but it would not be possible to argue that natural electroweak symmetry breaking is untenable in the minimal supersymmetric standard model if the higgs boson lies above the kinematic reach of lep - ii . by contrast , the proposed run - iii of fermilab s tevatron with @xmath17 ( tev33 ) can pose a very serious challenge to the minimal supersymmetric standard model . the projected mass - reach for a standard model higgs boson at tev33 is 100 ( 120 ) gev with integrated luminosities of 10 ( 25 ) @xmath18 @xcite .
if the possibility that the light higgs boson decays primarily to neutralinos can be excluded on the basis of combined searches for superpartners at lep - ii and the tevatron , natural electroweak symmetry breaking in the minimal supersymmetric standard model will no longer be possible if tev33 fails to observe a light higgs boson .
the original and principle motivation for weak - scale supersymmetry is naturalness .
supersymmetry provides the only explicitly known mechanism which allows fundamental scalars to be light without an unnatural fine - tuning of parameters .
naturalness also implies that superpartner masses can not lie much above the weak - scale if we are to avoid the fine - tuning which would be needed to keep the weak - scale light . in this section
, we recall the principle of naturalness and briefly review how it can be reliably quantified .
a more complete discussion of naturalness criteria can be found in ref . @xcite .
although fine - tuning is an aesthetic criterion , once we adopt the prejudice that large unexplained - cancellations are unnatural , a quantitative fine - tuning measure can be constructed and placed on solid footing . for any effective field theory , it is straightforward to identify whether large cancellations occur , and when these fine - tunings are present their severity can be reliably quantified . in non - supersymmetric theories , light fundamental scalars
are unnatural because scalar particles receive quadratically divergent contributions to their masses .
generically , at one - loop , a scalar mass is of the form @xmath19 where @xmath20 is the ultraviolet cutoff of the effective theory , and @xmath21 is a bare term .
the divergence in eq .
( 2.1 ) must be almost completely cancelled against the counter term or the fundamental scalar will have a renormalized mass on the order of the cutoff . in supersymmetry ,
additional loops involving super - partners conspire to cancel these quadratic divergences , but when supersymmetry is broken , the cancellation is no longer complete , and the dimensionful terms in eq .
( 2.1 ) are replaced by the mass splitting between standard particles and their super - partners . in this toy example , the cancellation is self - evident , and no abstract quantitative prescription is needed to determine when the parameters of the theory must conspire to give a light scalar mass .
we are interested in a more complicated example , and this requires a quantitative prescription for identifying instances of fine - tuning . in the toy example , if we examine the sensitivity of the scalar mass to variations in the coupling @xmath22 : @xmath23 where @xmath24 the scalar mass will be unusually sensitive to minute changes in @xmath22 when we arrange for large unexplained - cancellations @xcite : @xmath25 however , the bare sensitivity parameter @xmath26 , by itself is not a measure of naturalness .
although physical quantities depend sensitively on minute variations of the fundamental parameters when there is fine - tuning , fine - tuning is not necessarily implied by @xmath27 .
large sensitivities can occur in a theory even when there are no large cancellations .
in particular , this is true for supersymmetric extensions of the standard model , where it is known that bare sensitivity provides a poor measure of fine - tuning @xcite .
a reliable measure of fine - tuning must compare the sensitivity of a particular choice of parameters @xmath26 to a measure of the average , global sensitivity in parameter space , @xmath28 .
the naturalness measure @xmath29 will greatly exceed unity if and only if fine - tuning is encountered @xcite . to the amount of parameter space giving a unusually light value of @xmath30 .
this criterion is in fact equivalent to the ratio of sensitivity over typical sensitivity @xcite . ]
this definition is a quantitative implementation of a refined version of wilson s naturalness criterion : observable properties of a system should not be unusually unstable against minute variations of the fundamental parameters . in supersymmetric extensions of the standard model ,
as the masses of superpartners become heavy , increasingly large fine - tuning is required to keep the weak - scale light .
naturalness places an upper bound on supersymmetry - breaking parameters and superpartner masses . because the radiative corrections to the higgs boson mass increase with heavier superpartner masses , naturalness translates into an upper limit on the mass of the lightest higgs boson .
this limit is computed in the following section .
following the methods of ref . 6 , we have computed the severity of fine - tuning in the minimal supersymmetric standard model . for definiteness , we consider soft supersymmetry breaking parameters with ( universal ) minimal , low - energy supergravity ( mles ) boundary conditions .
we quantify the severity of large cancellations , and present our results as upper limits on the higgs boson mass as a function of the degree of fine - tuning .
although our quantitative results were obtained in a framework with universal soft terms at a scale near @xmath31 gev , as motivated by mles , we do not expect our bounds on the higgs boson mass to significantly increase in models with more general soft supersymmetry breaking masses provided they have minimal particle content at the weak - scale .
because there are enough free parameters in mles to independently adjust the parameters in the minimal supersymmetric standard model ( mssm ) which most significantly increase the higgs boson mass , more general soft terms could allow one to increase the masses of the squarks from the first two generations above their naturalness limits in mles , for example , but these new degrees of freedom will not significantly increase the upper limit on the higgs boson mass . qualitatively , our results are even more general , if we enlarge the particle content beyond the mssm , the upper - limit on the lightest higgs boson mass can be increased @xcite , but natural values of the lightest higgs boson mass will lie significantly below any maximal upper - bounds .
our calculation evolves the dimensionless couplings of the theory at two - loops and includes one - loop threshold contributions and one - loop correction to the higgs potential . from the resulting weak - scale parameters , we calculate the pole masses for the higgs bosons at one - loop following standard diagrammatic techniques @xcite .
the remaining next - to - leading order corrections to the higgs boson mass arising from the two - loop evolution of dimensionful couplings are small in the natural region of parameter space @xcite .
figures 1 - 3 show the naturalness of the higgs boson mass as a function of @xmath2 , @xmath32 , and @xmath33 , respectively . in all three figures ideally natural solutions
correspond to @xmath34 and fine - tuning is implied by @xmath35 .
figure 1 shows contours where the severity of fine - tuning - @xmath36 exceeds @xmath37 , @xmath38 , @xmath39 and @xmath40 in the @xmath41-@xmath42 plane for @xmath43 gev . from fig .
1 we see that the mass of the lightest higgs boson can not exceed @xmath0 gev without very significant fine - tuning , while in the most natural cases it lies below @xmath1 gev
. when @xmath2 is small these limits are even more restrictive .
figure 2 shows naturalness contours for the lightest higgs boson mass in mles as a function of the cp - odd higgs mass , @xmath32 for @xmath43 gev and arbitrary @xmath2 .
if we restrict ourselves to modest or small values of @xmath2 these curves will become more restrictive in the @xmath42 direction .
figure 3 shows naturalness contours for the lightest higgs boson mass in mles as a function of the top quark mass .
the inset in fig .
3 displays the current uncertainty in the top quark mass , and the projected uncertainties after run - ii of fermilab s tevatron and after tev33 @xcite .
fine - tuning increases both with increasing superpartner masses and with an increasing top quark yukawa coupling .
therefore , in contrast to the case of fixed superpartner masses where the corrections to the mass squared of the higgs boson increases as @xmath44 , for fixed naturalness these corrections increases roughly as @xmath45 .
we can assess the challenge to weak - scale supersymmetry from higgs boson searches at colliders from the natural regions of parameter space identified in figs.1 - 3 .
the dominant production mechanism for light cp - even higgs boson at lep - ii is higgs - strahlung @xmath46 if higgs boson decays into light neutralino pairs , @xmath47 , are kinematically forbidden , @xmath48 will decay primarily to @xmath16 .
an upper bound on the light higgs mass reach in this mode is set by kinematics and scales as @xmath49 ( a few ) gev .
the combined 95% cl exclusion reaches for a standard model ( sm ) higgs boson at lep - ii are 83 ( 98 ) ( ( 112 ) ) gev at @xmath50 175 ( 192 ) ( ( 205 ) ) gev , with integrated luminosities of 75 ( 150 ) ( ( 200 ) ) @xmath51 , per experiment @xcite .
however , it is well known that the observability of the lightest supersymmetric scalar @xmath48 can be degraded with respect to the standard model in two respects . first , the @xmath52 vertex carries a suppression of @xmath53 relative to the standard model vertex , where @xmath54 is the mixing angle of the cp - even higgs scalars .
the departure of this factor from unity can be appreciable for relatively light values of the @xmath55-odd mass @xmath32 , but it approaches one as the mass of the @xmath55-odd higgs increases .
for @xmath56 gev , @xmath57 . if the @xmath55-odd higgs mass is light it may be produced and seen through associated production @xmath58 , but this mode provides a less significant challenge to weak - scale supersymmetry because the cp - odd scalar mass @xmath32 is much less constrained by naturalness arguments ( see fig .
second , the mass reach for the lightest higgs @xmath48 can also be reduced if @xmath48 decays invisibly into a pair of lightest superpartners , @xmath59 .
this branching ratio can approach 100% when allowed @xcite , and this mode becomes more probable as the mass of the lightest higgs boson increases . in the relatively clean environment of an @xmath14 collider
, a higgs with such invisible decays could be seen from the acoplanar jet or lepton pair topologies resulting from the decay of the associated @xmath60 , but the higgs mass reach in this case is reduced to roughly half of the reach when @xmath48 decays visibly @xcite .
when @xmath61 is maximal , in the _ most _ natural cases , lep - ii operating up to @xmath62 gev would observe a light higgs , but this energy is not large enough to argue that natural electroweak symmetry breaking is untenable in minimal supersymmetry if the higgs boson lies above the kinematic reach of lep - ii .
kinematically , the proposed run - iii of fermilab s tevatron with @xmath17 ( tev33 ) @xcite can pose a very serious challenge to weak - scale supersymmetry .
the best single mode for discovery of a light higgs boson at the tevatron is @xmath63 , with @xmath64 @xcite .
tev33 can probe a sm higgs up to 100 ( 120 ) gev with integrated luminosities of 10 ( 25 ) @xmath18 .
a higgs boson mass in excess of @xmath0 gev would be extremely unnatural in the mssm .
however , the @xmath65 cross section from @xmath66 production is also reduced by the factor @xmath67 .
so the significance of the challenge to weak - scale supersymmetry from light higgs searches at tev33 will depend strongly on the ability of searches for neutralinos and charginos at the tevatron and lep - ii to eliminate the possibility of @xmath68 , by raising the limits on the lsp mass .
if this is the case , natural electroweak symmetry breaking in the minimal supersymmetric standard model will no longer be tenable if tev33 achieves @xmath69 and fails to observe any signal of a higgs boson .
natural choices of parameters in supersymmetric models lead to higgs boson masses which lie significantly below the maximal upper - bounds determined previously in the literature .
we have computed the natural upper bound on the higgs mass in mles , and we have quantified the extent to which naturalness is lost as the lower bound on @xmath42 increases .
a higgs mass above @xmath0 gev will require very large fine - tuning , while the most natural values of the higgs mass lie below @xmath1 gev .
the natural values of the lightest higgs boson mass have important implications for the challenge to weak - scale supersymmetry at colliders .
in particular , if the possibility that the higgs decays predominantly to neutralino pairs can be excluded from neutralino mass limits inferred from other superpartner searches , natural electroweak symmetry breaking will no longer be tenable in the mssm if tev33 achieves the projected reach of @xmath70 gev and fails to observe signals of a higgs boson .
ga acknowledges the support of the u.s .
department of energy under contract de - ac02 - 76ch03000 .
dc is supported by the u.s .
department of energy under grant number de - fg-05 - 87er40319 .
ar is supported by the doe and nasa under grant nag52788 .
fermilab is operated by the universities research association , inc .
, under contract de - ac02 - 76ch03000 with the u.s .
department of energy .
99 e. witten , nucl .
b188 * ( 1981 ) 513 ; s. dimopoulos and h. georgi , nucl . phys .
* b193 * ( 1981 ) 150 ; n. sakai , z. phys .
* c11 * ( 1981 ) 153 ; r. kaul , phys . lett . * b109 * ( 1982 ) 19 . y. okada , m. yamaguchi , and t. yanagide , prog . theor .
* 85 * ( 1991 ) 1 ; j. ellis , g. ridolfi , and f. zwirner , phys . lett .
* b257 * ( 1991 ) 83 ; _ ibid._*b262 * ( 1991 ) 477 ; h.e .
haber and r. hempfling , phys .
* 66 * ( 1991 ) 1815 .
r. barbieri , m. frigeni and f. caravaglios , phys .
lett . * b258 * ( 1991 ) 167 ; y. okada , m. yamaguchi , and t. yanagide , phys . lett .
* b262 * ( 1991 ) 54 ; a. yamada , phys . lett . * b263 * ( 1991 ) 233 ; j.r . espinosa and m. quirs ,
b266 * ( 1991 ) 389 ; a. brignole , j. ellis , g. ridolfi and f. zwirner , phys . lett .
* b271 * ( 1991 ) 123 ; p.h .
chankowski , s. pokorski and j rosiek , phys . lett .
* b274 * ( 1992 ) 191 ; a. brignole , phys . lett . * b281 * ( 1992 ) 284 ; h.e . haber and r.hempfling , phys . rev .
* d48 * ( 1993 ) 4280 .
r. hempfling and a.h .
hoang , phys .
* b331 * ( 1994 ) 99 ; j. kodaira , y. yasui and k. sasaki , phys . rev . * d50 * ( 1994 ) 7035 ; j.a .
casas , j.r.espinosa , m.quirs , and a.riotto , nucl .
b436 * ( 1995 ) 3 , erratum-_ibid .
_ * b439 * ( 1995 ) 466 ; m. carena , j.r .
espinosa , m. quirs and c.e.m .
wagner , phys .
lett . * b355 * ( 1995 ) 221 ; m. carena , m. quirs and c.e.m .
wagner , nucl .
b461 * ( 1996 ) 407 ; m.a .
diaz , t.a .
ter veldhuis and t.j .
weiler , preprint vand - th-94 - 14-upd , hep - ph-9512229 .
report of the tev 2000 study group , d. amidei and r. brock , eds .
fermilab - pub-96 - 082 , http://fnalpubs.fnal.gov/archive/1996/pub/pub-96-082chaps.html ; d. amidei et .
al . , tev33 committee report .
anderson and d.j .
castao , phys . lett . *
b347 * , ( 1995 ) 300 ; phys . rev .
* d52 * ( 1995 ) 1693 ; phys . rev .
* d53 * ( 1996 ) 2403 .
k. wilson , as quoted by l. susskind , phys . rev .
* d20 * ( 1979 ) 2619 ; g.t hooft , in _ recent developments in gauge theories _ , ed by g. t hooft et al .
( plenum press , new york , 1980)p . 135 . j.r . espinosa and m. quirs ,
* b279 * ( 1992 ) 92 ; g.l .
kane , c. kolda , and j.d .
wells , phys .
lett * 70 * ( 1993 ) 2686 .
p. tipton , proceeding of the xxviii international conference on high energy physics 25 - 31 july 1996 , warsaw , poland , to be published world scientific .
j. rosiek , a. sopczak , phys .
b341 * ( 1995 ) 419 ; v. driesen , w. hollik , j. rosiek , z. phys .
* c71 * ( 1996 ) 259 .
report of the lep - ii higgs working group , m. carena et.al . , to be published in cern yellow report , cern-96 - 01 , hep - ph/9602250 .
j.f . gunion and h.e .
haber , nucl .
b272 * ( 1986 ) , 1 ; _ ibid .
_ * 278 * ( 1986 ) 449 ; a. djouadi , j. kalinowski , and p.m. zerwas , z. phys .
* c57 * ( 1993 ) 569 .
a. stange , w. marciano , and s. willenbrock , phys .
* d49 * ( 1994 ) 1354 ; * d50*(1994 ) 4491 .
naturalness contours for @xmath71 and @xmath40 in mles displayed in the @xmath2 - @xmath42 plane , for @xmath72 gev .
ideally natural solutions correspond to @xmath73 , while fine - tuning is exhibited by @xmath35 .
naturalness contours for @xmath71 and @xmath40 in mles displayed in the @xmath33 - @xmath42 plane .
more restrictive contours will result if @xmath2 is constrained to be small .
the horizontal error bars indicate the current uncertainty in the mass of the top quark . | we quantify the extent to which naturalness is lost as experimental lower bounds on the higgs boson mass increase , and we compute the natural upper bound on the lightest supersymmetric higgs boson mass .
we find that it would be unnatural for the mass of the lightest supersymmetric higgs boson to saturate it s maximal upper bound . in the absence of significant fine - tuning
, the lightest higgs boson mass should lie below @xmath0 gev , and in the most natural cases it should be lighter than @xmath1 gev . for modest @xmath2 ,
these bounds are significantly lower .
our results imply that a failure to observe a light higgs boson in pre - lhc experiments could provide a serious challenge to the principal motivation for weak - scale supersymmetry . fermilab - pub-96/147-t + hep - ph/9609463 + .5 in * naturalness lowers the upper bound + on the lightest higgs boson mass + in supersymmetry + * 1.cm greg w. anderson@xmath3 , diego j. castao@xmath4 , and antonio riotto@xmath3 + .25 cm _ @xmath3 fermi national accelerator laboratory + p.o .
box 500 , batavia ,
illinois 60510 + .25 cm _ @xmath4dept . of physics , florida state university + tallahassee , fl 32306 usa .
+ _ _ .5 cm |
both , the bohr - mottelson ( bm ) collective model @xcite and the interacting boson model ( ibm ) @xcite have thoroughly been used to study the same kind of nuclear structure problems .
although very different in their formulation , both models present clear relationships . in an approximate way
, the ibm can be interpreted as the second quantization of the bm shape variables @xcite .
more detailed connections between both models were studied during the eighties by several authors @xcite and , more recently , by rowe and collaborators @xcite .
both models have three particular cases that can be easily solved and for which a clear correspondence can be done .
these three cases are : i ) the bm anharmonic vibrator and the dynamical symmetry @xmath7 ibm limit , ii ) the bm @xmath5-unstable deformed rotor and the dynamical @xmath8 ibm limit , and iii ) the bm axial rotor and the dynamical symmetry @xmath8 ibm limit including @xmath9 interactions @xcite ) .
note that although it is traditionally accepted the correspondence of the dynamical symmetry @xmath10 ibm limit to a submodel of the bm , this fact has never been explicitly probed @xcite .
each of these cases are assigned to a particular shape using the hill - wheeler variables @xmath11 @xcite : spherical , deformed with @xmath5-instability , and axially deformed , respectively . for transitional situations
the correspondence between the two models is difficult , as rowe said `` what is simple in one model will be complicated when expressed in terms of the observables of the other '' .
this situation suggests , for the case of transitional hamiltonians , to look for the connection between bm and ibm through numerical studies . among the transitional hamiltonians , a specially interesting case occurs when it describes a critical point in the transition from a given shape to another . in general , for such a situation , where the structure of the system can change abruptly by applying a small perturbation , both
, the bm and the ibm , have to be solved numerically . however , recently iachello has proposed schematic bohr hamiltonians that intend to describe different critical points and that can be solved exactly in terms of the zeros of bessel functions .
the first of these models is known as @xmath1 @xcite .
@xmath1 is designed to describe the critical point at the transition from spherical to deformed @xmath5-unstable shapes .
the potential to be used in the differential bohr equation is assumed to be @xmath12independent and , for the @xmath13 degree of freedom an infinite square well is taken .
similar models were proposed later on by iachello , called @xmath14 and @xmath15 @xcite , to describe the critical points between spherical and axially deformed shapes and between axial and triaxial deformed shapes , respectively .
all these models give rise to spectra and electromagnetic transition rates that are parameter free , up to a scale . in spite of their simplicity ,
some experimental examples were found @xcite , just after the appearance of these models . in this work ,
we concentrate on @xmath1 and related models .
it will be published elsewhere the corresponding study for @xmath16models @xcite .
the formulation of @xmath1 attracted immediately attention both experimentally and theoretically . soon after the introduction of the @xmath1 model , the nucleus @xmath17ba was proposed by casten and zamfir @xcite as a realization of it .
other experimental examples proposed are : @xmath18ru @xcite , @xmath19pd @xcite and , @xmath20pd @xcite . concerning theoretical extensions of @xmath1 , first , arias @xcite proposed a generalization of the e2 operator to be used with the @xmath1 model , then caprio @xcite checked that a substitution of the original infinite well in the @xmath13 variable by a finite one , which makes the model not exactly solvable anymore , provides similar results .
it showed that the @xmath1 description is `` robust in nature '' , _
i.e. _ the main features of the model remain almost unchanged under strong modification of the depth of the potential .
arias and collaborators @xcite were the first authors who tried to analyze in a quantitative way the connection between the @xmath6 ibm critical point and the @xmath1 model
. in particular , they established , looking to few observables that the ibm , at the critical point , gives results close to @xmath1 for a small ( @xmath21 ) number of bosons . however , the ibm results for large @xmath22 nicely reproduce the spectra and electromagnetic transition rates of a bohr hamiltonian with a @xmath23 potential ( in the following @xmath2 ) . once more , the model is not analytically solvable anymore .
lvai and arias @xcite solved the bohr equation with a sextic potential with a centrifugal barrier @xcite , arriving to almost closed analytical formulae for the energies and wavefunctions . immediately after , bonatsos and collaborators explored the possibility of getting numerical solutions for the @xmath5-independent bohr hamiltonian with potentials of the type @xmath24 , with @xmath25 @xcite .
these sequences of potentials allow to go from the vibrational limit , @xmath26 , to @xmath1 , @xmath27 .
in particular , in ref .
@xcite spectra and transition rates for the potentials @xmath28 and , @xmath29 , are given explicitly and compared with the original @xmath1 ( infinite square well potential ) case .
as mentioned above , all these models are produced in the bm scheme and a natural question is to ask for the corresponding equivalence in the ibm .
is the ibm able for producing the same spectra and transition rates ?
if yes , does the ibm hamiltonian correspond to a critical point ?
this work is intended to answer these questions for the @xmath1 and related models ( @xmath30 and , @xmath31 potentials ) and analyze the convergence as a function of the boson number . for that purpose , a large set of @xmath1 and related models results for excitation energies and transition rates
are taken as reference for numerical fits of the general @xmath6 ibm transitional hamiltonian .
this procedure will allow to establish the ibm hamiltonian which best fit the different @xmath0models and their relation with the critical points .
the paper is organized as follows : in section [ sec - fit ] the fitting procedure is described and the obtained results are commented . section [ sec - crit ] is devoted to study the energy surfaces of the fitted ibm hamiltonians and to analyze these in relation to the critical point . in section [ sec - quasi ] the connection between the present results and the concept of quasidynamical symmetry
is discussed .
finally , in section [ sec - conclu ] the summary and conclusions of this work are presented .
the most general , including up to two - body terms , ibm hamiltonian can be written in multipolar form as , @xmath32 where @xmath33 is the @xmath34 boson number operator , and @xmath35 the symbol @xmath36 stands for the scalar product , defined as @xmath37 where @xmath38 corresponds to the @xmath39 component of the operator @xmath40 .
the operator @xmath41 ( where @xmath5 refers to @xmath42 and @xmath34 bosons ) is introduced to ensure the correct tensorial character under spatial rotations .
the electromagnetic transitions can also be analyzed in the framework of the ibm .
in particular , in this work we will focus on the @xmath43 transitions .
the most general @xmath43 transition operator including up to one body terms can be written as , @xmath44,\ ] ] where @xmath45 is the boson effective charge and @xmath46 is a structure parameter .
the @xmath0models are intended to be of use for @xmath5-unstable nuclei having @xmath47 as symmetry algebra . for the construction of an ibm @xmath5-unstable transitional hamiltonian it is sufficient to impose in eq .
( [ ham1 ] ) @xmath48 ( this implies that no casimir operator from the @xmath10 algebra is included ) as can be observed if the hamiltonian ( [ ham1 ] ) is rewritten in terms of casimir operators ( the definition for the casimir operators have been taken from @xcite ) : @xmath49 + \frac{18}{35 } \kappa_4 \,\hat c_2[u(5)]\\ & + & \big(\kappa_1-\frac{\kappa_3}{10}-\frac{\kappa_4}{14}\big)\,\hat c_2[o(3)]+ \big(\frac{\kappa_3}{2}-\frac{3}{14 } \kappa_4\big)\,\hat c_2[o(5 ) ] -\frac{\kappa_0}{4}\,\hat c_2[o(6 ) ] .
\label{ham - cas}\end{aligned}\ ] ] if additionally , we want to construct an ibm transitional hamiltonian that preserves the @xmath47 symmetry , casimir operators for @xmath7 , @xmath8 and @xmath47 can be included but not the quadratic @xmath50 casimir operator .
this condition , translated to the multipolar form language used in eq .
( [ ham1 ] ) , leads to the constraint @xmath51 ( see eq .
( [ ham - cas ] ) ) .
in addition , the structure parameter , @xmath46 , in the @xmath52 operator is usually taken as zero in the standard ibm calculations for @xmath5-flat hamiltonians . in our calculations
we will impose @xmath48 , _
i.e. _ the @xmath5 flatness . to make more simple the later analysis
, we will restrict ourself to the case @xmath53 , leaving as free parameters @xmath54 , @xmath55 , @xmath56 ( plus @xmath57 that fixes the energy scale ) . in practice
, we do not impose the constraint @xmath58 but , as it will be shown , the condition will be fulfilled in every fit . in this section
we describe the procedure for getting the ibm hamiltonian which best fit the different @xmath0models .
the @xmath59 test is used to perform the fitting .
the @xmath59 function is defined in the standard way , @xmath60 where @xmath61 is the number of data , from a specific @xmath1-model , to be fitted , @xmath62 is the number of parameters used in the ibm fit , @xmath63 is an energy level ( or a @xmath64 value ) taken from a particular @xmath0model , @xmath65 is the corresponding calculated ibm value , and @xmath66 is an arbitrary error assigned to each @xmath63 . in order to perform the fit , we minimize the @xmath59 function for the energies , using @xmath57 , @xmath54 , @xmath55 and @xmath56 as free parameters and @xmath67 and @xmath68 fixed to zero . for doing this task we use minuit @xcite , which allows to minimize any multi - variable function .
the labels for the energy levels follow the usual notation introduced for the @xmath1 model : @xmath69 enumerates the zeros of the @xmath13 part of the wave function , and @xmath70 is the label for the @xmath47 algebra , _
i.e. _ the @xmath47 seniority quantum number , which is a good quantum number along all the transition from @xmath7 to @xmath8 . the selected set of levels included in the fit for the different @xmath0models are : * for the @xmath71 band , all the states with angular momentum lower than @xmath72 and @xmath73 .
an arbitrary @xmath74 is used for these states except for the @xmath75 state for which @xmath76 is used .
this latter value allows to normalize all the ibm energies to @xmath77 .
note that the energy of the state @xmath75 is fixed arbitrarily to @xmath78 ( remind that the spectrum is calculated up to a global scale factor ) .
* for the @xmath79 band , all the states with angular momentum lower than @xmath80 and @xmath81 .
an arbitrary @xmath82 is used for these states . * for the @xmath83 band , just the states with ( @xmath84 ) and ( @xmath85 ) are included .
an arbitrary @xmath86 is used for these states . with this selection ,
the number of energy levels included in the fit , @xmath61 , is equal to @xmath87 .
note that the state @xmath88 is not a real data to be reproduced because we are interested just in excitation energies and therefore the ground state is naturally fixed to zero in both , @xmath0models and ibm . in table [ tab - energ - fit ] the states included in the fit are explicitly given .
.states included in the energy fit . [ cols="^,^,^,^",options="header " , ]
one of the most attractive features of the @xmath0models treated in this work is that they are supposed to describe , at different approximation levels , the critical point in the transition from spherical to deformed @xmath5-unstable shapes .
since they are connected to a given ibm hamiltonian , as shown in the preceding section , this should correspond to the critical point in the transition from @xmath7 to @xmath8 ibm limits , _
i.e. _ this hamiltonian should produce an energy surface with @xmath89 .
is this the case for the fitted ibm hamiltonians obtained in the preceding section ? before starting with the discussion it is necessary to establish a measure on how close is a given ibm hamiltonian to the critical point .
an energy surface can be associated to a given ibm hamiltonian by using the intrinsic state formalism @xcite which introduces the shape variables @xmath11 in the ibm . to define the intrinsic state one has to consider that the dynamical behavior of the system can be approximately described in terms of independent bosons moving in an average field @xcite .
the ground state of the system is written as a condensate , @xmath90 , of bosons that occupy the lowest - energy phonon state , @xmath91 : @xmath92 where @xmath93 @xmath13 and @xmath5 are variational parameters related with the shape variables in the geometrical collective model @xcite .
the expectation value of the hamiltonian ( [ ham1 ] ) in the intrinsic state ( [ gs ] ) provides the energy surface of the system , @xmath94 . this energy surface in terms of the parameters of the hamiltonian
( [ ham1 ] ) and the shape variables can be readily obtained @xcite , @xmath95.\end{aligned}\ ] ] the shape of the nucleus is defined through the equilibrium value of the deformation parameters , @xmath13 and @xmath5 , which are obtained minimizing the ground state energy , @xmath96 .
a spherical nucleus has a minimum in the energy surface at @xmath97 , while a deformed one presents the minimum at a finite value of @xmath13 .
the parameter @xmath5 represents the departure from axial symmetry , _ i.e. _ @xmath98 and @xmath99 stand for an axially deformed nucleus , prolate and oblate respectively , while any other value corresponds to a triaxial shape .
an additional situation appears when the energy surface is independent on @xmath5 but presents a minimum in @xmath13 , being the nucleus @xmath5-unstable .
it should be noted that for a general ibm hamiltonian including up to two body terms the shape is either axially symmetric or @xmath5-unstable .
moreover , the hamiltonians considered in this work correspond always to the @xmath5-unstable situation . with the tools described above one can study phase transitions in the ibm @xcite .
first , the parameters that define the hamiltonian are the control parameters and normally are chosen in such a way that only one of them is a variable , while the rest remain constant .
the deformation parameters @xmath13 and @xmath5 become the order parameters , although in our case the only order parameter is @xmath13 . roughly speaking
, a phase transition appears when there exists an abrupt change in the shape of the system when changing smoothly the control parameter .
the phase transitions can be classified according to the ehrenfest classification @xcite .
first order phase transitions appear when there exists a discontinuity in the first derivative of the energy with respect to the control parameter .
this discontinuity appears when two degenerate minima exist in the energy surface for two values of the order parameter @xmath13 .
second order phase transitions appear when the second derivative of the energy with respect to the control parameter displays a discontinuity .
this happens when the energy surface presents a single minimum for @xmath97 and the surface satisfies the condition @xmath89 . in a more modern classification ,
second order phase transitions belongs to the high order or continuous phase transitions @xcite . to determine whether a given hamiltonian corresponds to a critical point or
not , the flatness or the existence of two degenerate minima in the energy surface should be investigated . for the case of one parameter ibm hamiltonian , _
e.g. _ consistent q ( cqf ) hamiltonians @xcite , it is simple to find an analytical expression for the critical control parameter in the hamiltonian . however , for a general ibm hamiltonian it is necessary to rewrite the energy surface in a special way , as the one presented in ref .
@xcite . there
, the authors manage to write the energy surface of a general ibm hamiltonian in terms of two parameters .
the authors make use of some concepts from the catastrophe theory @xcite to define the two essential parameters , ( @xmath100 ) . in terms of these
they find expressions for the locus , in the essential parameter space , that gives a critical point at the origin in @xmath13 , called bifurcation set , and for the locus that gives rise to two degenerate minima , called maxwell set . for the hamiltonians considered in section [ sec - fit ] @xmath48 and @xmath53 , in these cases @xmath101 and @xmath102 can be written as , @xmath103 where @xmath104 note that in the large @xmath22 limit , @xmath57 is proportional to @xmath22 ( see figure [ fig - par - not4t4 ] ) and therefore ( [ r1 ] ) can be approached by , @xmath105 this expression agrees with the use of an energy surface derived through a holstein - primakoff expansion @xcite . in this language
, a critical hamiltonian corresponds to @xmath106 . in figure [ fig - r1-not4t4 ]
the values of @xmath102 as a function of @xmath22 for the ibm hamiltonians obtained from the fit are presented for the different @xmath0models studied . for the @xmath1 model
, the fitted ibm hamiltonian produces @xmath106 for @xmath107 . in the case of the @xmath4 the value @xmath106
is obtained for @xmath108 , while for @xmath3 it is obtained for @xmath109 . for the @xmath2 model
it is known that @xmath106 is reached for very large number of bosons @xcite .
( see text for definition ) as a function of @xmath22 for the fitted ibm hamiltonians.,width=377 ] it is worth to show ( see figure [ fig - ener - r1 ] ) that for all the fitted ibm hamiltonians , the resulting energy surfaces are quite flat in a large interval of @xmath22 values and , therefore , it is justified to say that the fitted ibm hamiltonians are very close to the critical area . as a consequence
, the @xmath0models will be appropriated to describe phase transition regions close to the critical point . , for selected values of @xmath102 ( see text for definition).,width=377 ]
the concept of quasidynamical symmetry ( qds ) was introduced in refs .
@xcite and has been used in the study of phase transitions .
this concept is very useful for working with hamiltonians that present as limits two dynamical symmetries ( depending on the value of a control parameter ) . in this situation ,
the system shows the tendency to hold onto a given symmetry until the control parameter reach a critical value , passing the system , at this moment , onto the other symmetry .
the remarkable feature is that the system can present a set of states that behave as belonging to irreducible representations ( irreps ) of the corresponding symmetry group , although in fact , they do not belong to a given irrep but to a mixture of them . in mathematical terms
, a qds can be defined through the _ embedded representations _
@xcite : `` if a subset of states of a system are in one - to - one correspondence with the states of an irrep of a group g and if all the properties of the subset of states associated with observables in the lie algebra of g ( including their relationships to one another but not necessarily their relationships to states outside of the subset ) are as they would be if the states actually belonged to an irrep of the group g , then the subset of states is said to span an embedded representation of g '' . therefore ,
in the case of a qds , there exist a set of states that behave as belonging to a unique irrep of g , although that is only apparent , because they correspond to a superposition of irreps , but all their observables ( up to certain degree of accuracy ) are identical to the ones of states within a given irrep . in summary
, the states can be expressed as a coherent superposition of irreps that behave as a single one .
note that to show that a qds exists , one has to fix a subset of states and the degree of accuracy for the comparison with the observables of the dynamical symmetry . in our comparison between the ibm and the @xmath0models we observe a phenomenon which resembles the qds , _
i.e. _ part of the ibm spectrum behaves as having @xmath0symmetry , although , indeed they do not have such a symmetry .
we should emphasize that this is not a real qds for two reasons : i ) @xmath0cases are not dynamical symmetry limits of the ibm and ii ) the bm and the ibm have different hilbert spaces .
indeed , it is not possible to define irreps in @xmath0models and therefore embedded representations .
we will call this situation quasi - critical point symmetry ( qcps ) @xcite . in order to study in detail the qcps one has to fix the degree of accuracy to be demanded to the observables . in our study , for the energies an accuracy of @xmath110 for all the states belonging to a given @xmath69 is set while for the @xmath64 values an accuracy of @xmath111 for all the studied intra - band transitions in a given @xmath69 is selected .
tables [ table - states - not4t4 ] and [ tab - be2-comp ] , which correspond to @xmath112 are analyzed below , * @xmath1 : only the states in the @xmath71 band present @xmath1 qcps .
* @xmath4 : only the @xmath71 states present @xmath4 qcps . *
@xmath3 : only the @xmath71 states present @xmath3 qcps . * @xmath2 : all the studied states , @xmath71 , @xmath79 and @xmath83 , present @xmath2 qcps .
these results , regarding the energies , can be extended to larger values of @xmath22 too ( see figure [ fig - chi2-e5-not4t4 ] ) , _ i.e. _ the values of the energies remain stable when @xmath22 increases , while for the @xmath64 values the observed differences become larger , specially in the @xmath1 case .
in this paper , we have studied the connection between the @xmath0models and the ibm on the basis of a numerical mapping between models . to establish the mapping we have performed a best fit of the general @xmath6 transitional ibm hamiltonian to a selected set of energy levels produced by several @xmath0models .
later on , a check to the wavefunctions , obtained with the best fit parameters , has been done by calculating relevant @xmath64 transition rates .
all calculations have been done as a function of the number of bosons .
once the best fit ibm hamiltonians to the different @xmath0models are obtained , their energy surfaces are constructed and analyzed with the help of the catastrophe theory so as to know how close they are to a critical point .
finally , the concept of quasi - critical point symmetry is introduced , as similar to the idea of quasidynamical symmetry .
we have shown that it is possible , in all cases , to establish a one - to - one mapping between the @xmath0models and the ibm with a remarkable agreement for both the energies and the @xmath64 transition rates .
in general , the goodness of the fit to the energies is independent on the number of bosons , but the corresponding @xmath64 transition rates are indeed sensitive to @xmath22 .
this is so specially in the @xmath1 , for which the @xmath59 value reaches a minimum for @xmath22 small ( @xmath107 ) and from there on increases notably as a function of @xmath22 .
globally , the best agreement is obtained for the @xmath2 hamiltonian and the worst for the @xmath1 case . for the case of very large number of bosons and hamiltonians with @xmath47 symmetry
we have confirmed the results of @xcite , _
i.e. _ the only @xmath0model that can be reproduced exactly by the ibm is @xmath2 , corresponding such a hamiltonian with the critical point of the model ( @xmath106 ) .
a consequence of this excellent agreement is that it is impossible , from a experimental point of view , to discriminate between a @xmath1-model and its corresponding ibm hamiltonian when only few low - lying states are considered ( usually the four lowest states in the ground state band , plus @xmath113 and @xmath114 in the @xmath79 band ) .
we have also proved that all the @xmath0models correspond to ibm hamiltonians very close to the critical area , @xmath115 .
therefore , one can say that the @xmath0 models are appropriate to describe transitional @xmath12unstable regions close to the critical point .
we have found that the results presented in this paper are consistent with the existence of something similar to a quasidynamical symmetry , we call this phenomenon quasi - critical point symmetry . finally , it should be noted that the use of a more general @xmath6 hamiltonian , _
e.g. _ using @xmath68 as free parameter , do not change the main conclusions of this work .
we are grateful to d.j .
rowe for a careful reading of the manuscript and for his valuable comments . this work has been partially supported by the spanish ministerio de educacin y ciencia and by the european regional development fund ( feder ) under projects number fis2005 - 01105 , fpa2006 - 13807-c02 - 02 and fpa2007 - 63074 , and by the junta de analuca under projects fqm160 , fqm318 , p05-fqm437 and p07-fqm-02962 .
25 a. bohr , mat .
vidensk .
( 14 ) ( 1952 ) .
a. bohr , b.r .
mottelson , mat .
vidensk .
selsk . 27 ( 16 ) ( 1953 ) .
a. bohr , b.r .
mottelson , _ nuclear structure _ , vol .
ii , ( benjamin , elmsford , ny , 1969 ) .
a. arima , f. iachello , ann . phys .
* 99 * , 253 ( 1976 ) .
a. arima , f. iachello , ann .
phys . * 111 * , 201 ( 1978 ) ; o. scholten , f. iachello , a. arima , ann .
phys . , 325 ( 1978 )
; f. iachello and a. arima , _ the interacting boson model _ ( cambridge university press , cambridge , 1987 ) .
d. janssen , r.v .
jolos , f. dnau , nucl .
a * 224 * , 93 ( 1974 ) .
dieperink , o. scholten , f. iachello , phys . rev .
lett . * 44 * , 1747 ( 1980 ) .
a.e.l . dieperink and o. scholten , nucl .
a * 346 * , 125 ( 1980 ) .
ginocchio , m.w .
kirson , phys .
lett . , 1744 ( 1980 ) .
ginocchio , m.w .
kirson , nucl .
a * 350 * , 31 ( 1980 ) .
kirson , ann .
* 143 * , 448 ( 1982 ) .
ginocchio , nucl .
a * 376 * , 438 ( 1982 ) .
rowe and g. thiamanova , nucl .
a * 760 * , 59 ( 2005 ) .
p. van isacker , phys .
lett . * 83 * , 4269 ( 1999 ) .
d.l . hill and j.a .
wheeler , phys .
89 * , 1102 , ( 1953 ) .
f. iachello , phys .
lett . * 85 * , 3580 , ( 2000 ) .
f. iachello , phys .
87 * , 052502 , ( 2001 ) .
f. iachello , phys .
lett . * 91 * , 132502 , ( 2003 ) .
casten and n.v .
zamfir , phys .
lett . , 3584 , ( 2000 ) .
casten and n.v .
zamfir , phys .
lett . , 052503 ( 2001 ) .
j.e . garca - ramos and j.m .
arias , in preparation .
a. frank , c.e .
alonso , and j.m .
arias , phys .
c * 65 * , 014301 ( 2001 ) .
zamfir , _ et al _ , phys .
c * 65 * , 044325 ( 2002 ) .
da - li zhang and yu - xin liu , phys . rev .
c * 65 * , 057301 ( 2002 ) .
arias , phys .
c * 63 * , 034308 ( 2001 ) .
caprio , phys .
c * 65 * , 031304 ( 2002 ) .
arias , c.e .
alonso , a. vitturi , j.e .
garca - ramos , j. dukelsky , and a. frank , phys .
c * 68 * , 041302(r ) ( 2003 ) .
garca - ramos , j. dukelsky , and j.m .
arias , phys .
c * 72 * , 037301 ( 2005 ) .
g. lvai and j.m .
arias , phys .
c * 69 * , 014304 ( 2004 ) .
ushveridze , _ quasi - exactly solvable models in quantum mechanics _ ( iop publishing , bristol , 1994 ) .
d. bonatsos , d. lenis , n. minkov , p.p .
raychev , and p.a .
c * 69 * , 044316 ( 2004 ) .
a. frank and p. van isacker , _ algebraic methods in molecular and nuclear structure physics _
( john wiley & sons , ny , 1994 ) .
f. james , minuit : function minimization and error analysis reference manual , version 94.1 , cern , 1994 . j. dukelsky , g.g .
dussel , r.p.j .
perazzo , s.l .
reich , and h.m .
sofia , nucl .
phys . a * 425*,93 ( 1984 ) .
p. van isacker and j.q .
chen , phys .
c * 24 * , 684 ( 1981 ) . h.e .
standley , `` introduction to phase transitions and critical phenomena '' , oxford university press , oxford ( 1971 ) .
warner and r.f .
casten , phys .
* 48 * , 1385 ( 1982 ) .
e. lpez - moreno and o. castaos , , ( 1996 ) .
r. gilmore .
_ catastrophe theory for scientists and engineers _ ( wiley , new york , 1981 ) .
arias , j. dukelsky , j.e .
garca - ramos , and j. vidal , phys .
c * 75 * , 014301 ( 2007 ) .
rowe , phys .
lett . * 93 * , 122502 ( 2004 ) .
rowe , p.s .
turner , and g. rosensteel , phys .
* 93 * , 232502 ( 2004 ) .
rowe , nucl .
a * 745 * , 47 ( 2004 ) .
turner and d.j .
rowe , nucl .
phys . a * 756 * , 333 ( 2004 ) .
rowe , private comm . | the connections between the @xmath0models ( the original @xmath1 using an infinite square well , @xmath2 , @xmath3 and @xmath4 ) , based on particular solutions of the geometrical bohr hamiltonian with @xmath5-unstable potentials , and the interacting boson model ( ibm ) are explored . for that purpose ,
the general ibm hamiltonian for the @xmath6 transition line is used and a numerical fit to the different @xmath0models energies is performed , later on the obtained wavefunctions are used to calculate b(e2 ) transition rates .
it is shown that within the ibm one can reproduce very well all these @xmath0models .
the agreement is the best for @xmath2 and reduces when passing through @xmath3 , @xmath4 and @xmath1 , where the worst agreement is obtained ( although still very good for a restricted set of lowest lying states ) .
the fitted ibm hamiltonians correspond to energy surfaces close to those expected for the critical point . a phenomenon similar to the quasidynamical symmetry
is observed . |
after the advent of the brane world scenario @xcite and the ads / cft correspondence @xcite , there has been an increase in interest in the study of higher - dimensional black holes .
the kerr metric was first generalized to higher dimensions in the seminal paper by myers and perry @xcite .
one of the unexpected results is that for some rotating black holes event horizons exist for arbitrarily large values of the rotation parameters .
the stability of such black holes are certainly in question @xcite . these are asymptotically flat black holes .
the first asymptotically non - flat higher - dimensional kerr metric was given by hawking _
et al_. @xcite for five dimensions .
subsequent generalizations to arbitrary dimensions were done by gibbons _
et al_. @xcite , finally resulting in the most general kerr-(a)ds - nut metric by chen , l and pope @xcite .
the study of the wave equations in these black hole spacetimes was initiated by frolov and stojkovic @xcite by analyzing the klein - gordon equation in five dimensions .
the analysis relied crucially on the method of the separation of variables .
the problem of separability of these wave equations in higher dimensions is a difficult one . even for the klein - gordon case ,
early attempts were only aimed at special cases @xcite . finally , using the chen - l - pope metric , frolov , krtous and kubiznak @xcite were able to separate the geodesic equation and the klein - gordon equation in the most general setting . however , the research on the scalar wave equation so far is focused mostly on the case with one rotation , the so - called simply - rotating case .
notably , the stability of the scalar perturbation in six and higher dimensions was considered in the ultra - rotating cases @xcite . apparently no instability was found .
then , due to the interests in ads spacetimes , these considerations were extended to kerr - ads black holes @xcite . here
the expected superradiant instability did indeed show up .
in addition the hawking radiation in these spacetimes @xcite are calculated with possible application to the production and decay of lhc black holes @xcite . in this paper
we would like to explore the cases with two rotations . in the next section we present the general metric of the kerr-(a)ds black hole with two rotations .
the corresponding klein - gordon equation will be separated into one radial equation and two angular ones . in section 3 , we shall develop perturbative expansions of the angular eigenvalues in powers of the rotation parameters .
conclusions and discussions are given in section 4 .
in this section we shall first present the kerr-(a)ds metric with two rotations .
we start with the metric for @xmath1 , that is , for even dimensions
. however , the result we obtain in the end should also be valid for odd dimensional cases . for @xmath1 , there are at most @xmath2 rotation directions so we have @xmath3 , @xmath4 .
this metric , satisfying @xmath5 , can be expressed as follows @xcite .
@xmath6^{2}\nonumber\\ & & \ \
\sum_{\alpha=1}^{n-1}\frac{x_{\alpha}}{u_{\alpha}}\left[\frac{(1+g^{2}r^{2})w}{1-g^{2}y_{\alpha}^{2}}d\tilde{t } -\sum_{i=1}^{n-1}\frac{(r^{2}+a_{i}^{2})\gamma_{i}}{a_{i}^{2}-y_{\alpha}^{2}}d\tilde{\phi}_{i}\right]^{2},\label{metric}\end{aligned}\ ] ] where @xmath7 with @xmath8 .
@xmath9 is the azimuthal angle for each @xmath3 . with the direction cosines
@xmath11 , the metric for a unit @xmath12 sphere is just @xmath13 subject to the constraint @xmath14 .
this constraint can be solved in terms of the unconstrained latitude variables @xmath15 s , @xmath16 the metric for the unit sphere can then be written as @xmath17 with @xmath18 now , to obtain a general metric with two rotations , we take the limit @xmath19 , @xmath20 , @xmath21 , @xmath22 while assuming that @xmath23 . from the definition in eq .
( [ defy ] ) we see that @xmath24 is of the same order of magnitude as @xmath3 . with this in mind ,
we take the limit to arrive at the metric with two rotations for general dimensions @xmath0 , @xmath25^{2}\nonumber\\ & & + \frac{\delta_{y_{1}}}{(r^{2}+y_{1}^{2})(y_{2}^{2}-y_{1}^{2 } ) } \left[\frac{(1+g^{2}r^{2})(1-g^{2}y_{2}^{2})}{(1-g^{2}a_{1}^{2})(1-g^{2}a_{2}^{2})}dt-\frac{(r^{2}+a_{1}^{2})(a_{1}^{2}-y_{2}^{2 } ) } { ( 1-g^{2}a_{1}^{2})(a_{1}^{2}-a_{2}^{2})}\frac{d\phi_{1}}{a_{1}}\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \left.-\frac{(r^{2}+a_{2}^{2})(a_{2}^{2}-y_{2}^{2 } ) } { ( 1-g^{2}a_{2}^{2})(a_{2}^{2}-a_{1}^{2})}\frac{d\phi_{2}}{a_{2}}\right]^{2}\nonumber\\ & & + \frac{\delta_{y_{2}}}{(r^{2}+y_{2}^{2})(y_{1}^{2}-y_{2}^{2 } ) } \left[\frac{(1+g^{2}r^{2})(1-g^{2}y_{1}^{2})}{(1-g^{2}a_{1}^{2})(1-g^{2}a_{2}^{2})}dt-\frac{(r^{2}+a_{1}^{2})(a_{1}^{2}-y_{1}^{2 } ) } { ( 1-g^{2}a_{1}^{2})(a_{1}^{2}-a_{2}^{2})}\frac{d\phi_{1}}{a_{1}}\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \left.-\frac{(r^{2}+a_{2}^{2})(a_{2}^{2}-y_{1}^{2 } ) } { ( 1-g^{2}a_{2}^{2})(a_{2}^{2}-a_{1}^{2})}\frac{d\phi_{2}}{a_{2}}\right]^{2}\nonumber\\ & & + \frac{(r^{2}+y_{1}^{2})(r^{2}+y_{2}^{2})}{\delta_{r}}dr^{2}+\frac{(r^{2}+y_{1}^{2})(y_{2}^{2}-y_{1}^{2})}{\delta_{y_{1}}}dy_{1}^{2 } + \frac{(r^{2}+y_{2}^{2})(y_{1}^{2}-y_{2}^{2})}{\delta_{y_{2}}}dy_{2}^{2}\nonumber\\ & & + r^{2}\left(\frac{y_{1}^{2}y_{2}^{2}}{a_{1}^{2}a_{2}^{2}}\right)d\omega^{2}_{d-6 } , \label{kerradsmetric}\end{aligned}\ ] ] and @xmath26 here we have kept the variables @xmath27 and @xmath28 instead of writing them in terms of angle variables .
this is because the relationship between them , as shown in eq .
( [ defy ] ) , is a bit complicated to write out explicitly . if we solve @xmath27 and @xmath28 in terms of @xmath29 and @xmath30 , we have @xmath31.\end{aligned}\ ] ] however , it is important to note that @xmath27 and @xmath28 must be constrained to be @xmath32 in order to have eq .
( [ defy ] ) well - defined .
the separation of the klein - gordon equation in the general kerr-(a)ds metric has been achieved by frolov _ et al_. @xcite . here
we shall present explicitly the separated equation for the case with two rotations .
writing the klein - gordon field as @xmath33 where @xmath34 the klein - gordon equation @xmath35 can be simplified to the radial equation @xmath36 r_{r}=0,\end{aligned}\ ] ] and the angular equations , for @xmath37 , @xmath38\nonumber\\ & & \ \
+ \bigg\{-\frac{(a_{1}^{2}-y_{i}^{2})(a_{2}^{2}-y_{i}^{2})\omega^{2}}{1-g^{2}y_{i}^{2 } } + 2\omega[m_{1}a_{1}(a_{2}^{2}-y_{i}^{2})+m_{2}a_{2}(a_{1}^{2}-y_{i}^{2})]\nonumber\\ & & \ \ \ \ \ \ \ -2m_{1}a_{1}m_{2}a_{2}(1-g^{2}y_{i}^{2})-\frac{m_{1}^{2}a_{1}^{2}(1-g^{2}y_{i}^{2})(a_{2}^{2}-y_{i}^{2 } ) } { a_{1}^{2}-y_{i}^{2}}\nonumber\\ & & \ \ \ \ \ \ \ -\frac{m_{2}^{2}a_{2}^{2}(1-g^{2}y_{i}^{2})(a_{1}^{2}-y_{i}^{2 } ) } { a_{2}^{2}-y_{i}^{2}}-\frac{a_{1}^{2}a_{2}^{2}j(j+d-7)}{y_{i}^{2 } } -b_{1}y_{i}^{2}+b_{2}\bigg\}r_{y_{i}}=0.\nonumber\\ \label{angulareqns}\end{aligned}\ ] ] in the following we shall solve the coupled angular equations above for the case with two rotations to obtain the angular eigenvalues @xmath39 and @xmath40 .
assuming that the rotation parameters @xmath41 and @xmath42 are small , we shall express these eigenvalues perturbatively as power series of @xmath41 and @xmath42 .
however , the case with @xmath43 is exceptional in which there is only one angular equation which has been dealt with recently @xcite .
therefore , we shall consider in the next section the case with @xmath0 .
with two rotations there are two angular equations involving the latitude coordinates @xmath27 and @xmath28 as in eq .
( [ angulareqns ] ) for dimensions @xmath0 . in this section
we develop expansions in powers of the rotation parameters @xmath41 and @xmath42 for the angular eigenvalues . since the latitude coordinates are restricted to @xmath44 and @xmath45 , it is convenient to change variables to @xmath46 and @xmath47 with @xmath48 where @xmath49 . to further simplify the consideration
we shall specialize to the case with @xmath50 , @xmath51 , and @xmath52 .
the analysis for other values of these parameters can be carried out in a similar fashion . in this special case the angular equations in eq .
( [ angulareqns ] ) are @xmath53 with @xmath37 . here ,
the operators are @xmath54\left\{2-g^{2}a_{1}^{2}\left[(1-x_{1})+\epsilon^{2}(1+x_{1})\right ] \right\}\frac{d^{2}}{dx_{1}^{2}}\nonumber\\ & & \ \ + \left\{\left[\frac{1}{2}(1 + 5x_{1})-\frac{1}{4}g^{2}a_{1}^{2}(1-x_{1})(3 + 7x_{1})\right]\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \left.-\frac{\epsilon^{2}}{2}\left(\frac{1+x_{1}}{1-x_{1}}\right)\left[(1 - 5x_{1})+7g^{2}a_{1}^{2}x_{1}(1-x_{1})\right]\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \left.+\frac{g^{2}a_{1}^{2}\epsilon^{4}(1+x_{1})^{2}(3 - 7x_{1})}{4(1-x_{1})}\right\}\frac{d}{dx_{1}}\nonumber\\ & & \
\ + \left\{\frac{a_{1}^{2}\omega^{2}(1+x_{1})(1-\epsilon^{2})^{2}}{4[2-g^{2}a_{1}^{2}((1-x_{1})+\epsilon^{2}(1+x_{1}))]}\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \left.+\frac{[(1-x_{1})^{2}+\epsilon^{2}(1+x_{1})^{2}][2-g^{2}a_{1}^{2}((1-x_{1})+\epsilon^{2}(1+x_{1}))]}{4(1-x_{1})^{2}(1+x_{1})}\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \ \left.-\frac{\epsilon^{2}(1+x_{1})b_{1}}{4(1-x_{1})}+\frac{b_{2}}{2(1-x_{1})}\right\},\\ { \cal o}_{2 } & = & -\frac{1}{4}(1-x_{2}^{2})\left[2-\epsilon^{2}(1-x_{2})\right]\left[2-g^{2}a_{1}^{2}\epsilon^{2}(1-x_{2})\right]\frac{d^{2}}{dx_{2}^{2}}\nonumber\\ & & \ \ -\left\{\frac{1}{2}(1 - 3x_{2})+\frac{1}{4}\epsilon^{2}(1+g^{2}a_{1}^{2})(1-x_{2})(1 + 5x_{2})\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \left.-\frac{1}{8}g^{2}a_{1}^{2}\epsilon^{4}(1-x_{2})^{2}(3 + 7x_{2})\right\}\frac{d}{dx_{2}}\nonumber\\ & & \ \
+ \left\{\frac{\omega^{2}a_{1}^{2}\epsilon^{2}(1+x_{2})(2-\epsilon^{2}(1-x_{2}))}{8(2-g^{2}a_{1}^{2}\epsilon^{2}(1-x_{2}))}\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \left.+\frac{[2-g^{2}a_{1}^{2}\epsilon^{2}(1-x_{2})][4-\epsilon^{2}(3 - 6x_{2}-x_{2}^{2})+\epsilon^{4}(1-x_{2})^{2 } ] } { 8(1+x_{2})(2-\epsilon^{2}(1-x_{1}))}\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \left.+\frac{\epsilon^{2}(1-x_{2})b_{1}}{8}\right\},\end{aligned}\ ] ] where we have defined @xmath55 . since we have assumed that @xmath56 , @xmath57 .
moreover , we have defined the constants @xmath58.\end{aligned}\ ] ] with these definitions of @xmath59 and @xmath60 , one has only even powers of @xmath61 and @xmath41 .
we shall develop perturbative expansions for @xmath59 and @xmath60 . since we have assumed that @xmath56 , we first expand the operators with respect to @xmath61 .
schematically we write @xmath62 to the zeroth order of @xmath61 , the eigenvalue equations are @xmath63 where @xmath64\frac{d^{2}}{dx_{1}^{2}}\nonumber\\ & & \ \ + \left[\frac{1}{2}(1 + 5x_{1})-\frac{1}{4}g^{2}a_{1}^{2}(1-x_{1})(3 + 7x_{1})\right ] \frac{d}{dx_{1}}\nonumber\\ & & \ \ + \left\{\frac{a_{1}^{2}\omega^{2}(1+x_{1})}{4[2-g^{2}a_{1}^{2}(1-x_{1 } ) ] } + \frac{[2-g^{2}a_{1}^{2}(1-x_{1})]}{4(1+x_{1})}+\frac{b_{20}}{2(1-x_{1})}\right\},\\ { \cal o}_{20 } & = & -(1-x_{2}^{2})\frac{d^{2}}{dx_{2}^{2}}-\frac{1}{2}(1 - 3x_{2})\frac{d}{dx_{2 } } + \frac{1 } { 2(1+x_{2})}.\end{aligned}\ ] ] since @xmath65 does not involve @xmath41 , the corresponding eigenvalue equation can be solved readily . to fix notation and to anticipate the analysis in the following we define @xmath66 .
the first zero in the subscript indicates that the operator involves zeroth power of @xmath61 while the second zero indicates that it involves zeroth power of @xmath41 .
the eigenfunctions and eigenvalues of the equation @xmath67 are @xmath68 where @xmath69 and @xmath70 is the jacobi polynomial .
one can check that the orthonormal condition is @xmath71 to be concrete in the following discussion we take @xmath72 .
hence , we have @xmath73 next , we work on the @xmath74 equation . following
what we have been doing we expand the eigenvalue equation in powers of @xmath41 with the operators @xmath75}{8(1+x_{1})},\\ { \cal o}_{104}&=&\frac{1}{16}g^{2}\omega^{2}a_{1}^{4}(1-x_{1}^{2}),\\ { \cal o}_{106}&=&\frac{1}{32}g^{4}\omega^{2}a_{1}^{6}(1-x_{1})(1-x_{1}^{2}).\end{aligned}\ ] ] similarly we solve the zeroth order eigenvalue equation , @xmath76 , to give @xmath77 with @xmath78 and the orthonormal condition @xmath79 again for simplicity we choose @xmath80 and we have @xmath81 to obtain @xmath82 to @xmath83 , we follow a perturbative procedure @xcite .
the result is @xmath84 if one iterates to higher orders , one can in principle obtain the eigenvalues in powers of @xmath41 and @xmath85 .
here we just list the result . for @xmath50 , @xmath51 , @xmath52 and @xmath86 , @xmath87\nonumber\\ & & + \epsilon^{2}\left[\frac{2}{9}(2\omega^{2}-17g^{2})a_{1}^{2}+\frac{32}{8019}(\omega^{4}-53g^{2}\omega^{2}+196g^{4})a_{1}^{4}\right.\nonumber\\ & &
\ \ \ \ \ \ \ \ \ \ \ \left .
-\frac{32}{2814669}(\omega^{6}-147g^{2}\omega^{4}+5178g^{4}\omega^{2}-18424g^{6})a_{1}^{6}\right]\nonumber\\ & & + \epsilon^{4}\left[-\frac{20}{8019}(\omega^{4}-53g^{2}\omega^{2}+196g^{4})a_{1}^{4}\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \left.-\frac{32}{2814669}(\omega^{6}-147g^{2}\omega^{4}+5178g^{4}\omega^{2}-18424g^{6})a_{1}^{6}\right]\nonumber\\ & & + \epsilon^{6}\left[\frac{20}{8444007}(2\omega^{6}-645g^{2}\omega^{4}+28959g^{4}\omega^{2}-105644g^{6})a_{1}^{6}\right]+\cdots,\end{aligned}\ ] ] @xmath88\nonumber\\ & & + \epsilon^{4}\left[-\frac{4}{8019}(\omega^{4}-53g^{2}\omega^{2}+196g^{4})a_{1}^{4}\right.\nonumber\\ & & \ \ \ \ \ \ \ \ \ \ \ \ \left.-\frac{32}{8444007}(4\omega^{2}-25g^{2})(\omega^{4}-53g^{2}\omega^{2}+196g^{4})a_{1}^{6}\right]\nonumber\\ & & + \epsilon^{6}\left[\frac{4}{8444007}(2\omega^{6}-645g^{2}\omega^{4}+28959g^{4}\omega^{2}-105644g^{6})a_{1}^{6}\right]+\cdots.\end{aligned}\ ] ]
in this paper we have started to study higher - dimensional kerr-(a)ds black holes with two rotations .
first , we have presented here the general metric for these doubly rotating black holes .
we have also separated the klein - gordon equation , writing out the corresponding radial and angular equations explicitly . with these separated equations
we could start to ask questions about stability , the various spectra of hawking radiation and the phenomenon of superradiance .
to get some quantitative understanding of the angular equations , we have developed perturbative expansions in powers of the rotation parameters @xmath41 and @xmath42 for the angular eigenvalues . these expansions could be used in a variety of considerations .
for example , in the evaluation of the frequencies of the low - lying quasinormal modes using a semi - analytic method like the wkb approximation @xcite , it is necessary to keep the frequency @xmath89 arbitrary .
perturbative expansions of the angular eigenvalues that we have developed would be useful .
in fact , we have already considered the quasinormal modes of these doubly rotating black holes using a numerical approach recently @xcite . on the other hand , to discuss the stability of the ultra - spinning black holes , we need to consider the angular eigenvalues in the large rotation limit @xcite . for our case
there are two rotation parameters .
with one small and one large , the situation will be very much like the simply rotating case @xcite and no instability is expected . therefore it is more interesting to consider the other case with both rotation parameters large .
this work will be pursued subsequently .
htc was supported in part by the national science council of the republic of china under the grant nsc 99 - 2112-m-032 - 003-my3 , and the national science centre for theoretical sciences .
the work of jd was supported by the japan society for the promotion of science ( jsps ) , under fellowship no . | in this paper , following the work of chen , l and pope , we present the general metric for kerr-(a)ds black holes with two rotations .
the corresponding klein - gordon equation is separated explicitly , from which we develop perturbative expansions for the angular eigenvalues in powers of the rotation parameters with @xmath0 . |
let @xmath1 be a locally compact group .
an important notion in representation theory is that of a _
model_. broadly speaking , a model is a representation space of @xmath1 ( usually defined geometrically ) which contains any irreducible constituent with multiplicity one .
consider the case where @xmath1 is the @xmath0-points of a reductive group @xmath2 defined over a local field @xmath0 .
there are many examples of models which are realized on induces spaces @xmath3 where @xmath4 is a character of a closed subgroup @xmath5 of @xmath1 i.e. , on the space of smooth functions on @xmath1 which are left @xmath6-equivariant with the right @xmath1-action .
( for simplicity we assume that @xmath5 is unimodular . ) in this case we say that the pair @xmath6 satisfies multiplicity one .
the representations which occur in this model are called distinguished by @xmath6 .
they are characterized by the property that there exists a non - zero functional on them which is @xmath6-equivariant .
if @xmath7 is such a representation then we denote by @xmath8 its realization in the model .
the most well - known case is the _ whittaker model _ in which @xmath1 is quasi - split , @xmath5 is a maximal unipotent subgroup and @xmath4 is a non - degenreate character of @xmath5 ( @xcite , @xcite ) .
the importance of the whittaker model in representations theory and automorphic forms can not be overestimated . for other instances of models , including recent progress cf .
@xcite . in this paper
we address the following question .
let @xmath1 be a reductive group defined over a @xmath9-adic @xmath0 , @xmath10 , @xmath11 two closed subgroups of @xmath1 defined over @xmath0 and @xmath12 , @xmath13 characters of @xmath14 .
suppose that @xmath15 , @xmath13 satisfy multiplicity one and let @xmath7 be an irreducible representation of @xmath1 which is distinguished by both @xmath15 , @xmath13 .
then up to a scalar there is a unique isomorphism @xmath16 can one explicate such an isomorphism , as well as its inverse ?
more precisely , can one find a @xmath1-equivariant integral transform from @xmath17 to @xmath18 with an explicit inversion formula ?
the question is meaningful for any irreducible representation @xmath7 of @xmath1 such that @xmath19 ( whether or not @xmath20 or @xmath21 satisfy multiplicity one in general ) .
we will call such an explicit map @xmath22 a _ model transition_. assume for simplicity that @xmath23 and @xmath24 coincide on @xmath25 . a first attempt would be to consider the map @xmath26 which _ formally _ defines an intertwining operator @xmath27 .
( often , this integral needs to be regularized . )
it turns out however that it is sometimes advantageous to replace the domain of integration by @xmath28 where @xmath29 is a proper subgroup of @xmath11 containing @xmath25 .
( this reflects the fact that we secretly work within a subgroup @xmath30 of @xmath1 containing @xmath10 and consider maps @xmath31 . )
of course this procedure does not give rise to a map from @xmath32 ( or a suitable @xmath1-invariant subspace @xmath33 where the regularized integral is defined ) to @xmath34 .
however , one can hope that in certain cases for ( suitable ) @xmath20-distinguished representations @xmath7 we have @xmath35 and @xmath36 .
we will give several examples of this phenomenon .
the first example is for the group @xmath37 .
let @xmath7 be a tempered representation @xmath7 of @xmath38 .
in particular , @xmath7 has a whittaker model @xmath39 , where @xmath40 is the subgroup of upper unitriangular matrices in @xmath38 and @xmath41 a non - degenerate character of @xmath40 .
let @xmath42 be the centralizer of @xmath43 in @xmath38 , isomorphic to @xmath44 .
we assume that @xmath7 has a model @xmath45 with respect to @xmath42 .
( it is known that , for any @xmath7 , this model is unique if it exists @xcite . )
note that the character @xmath41 is trivial on @xmath46 .
in one direction we define @xmath47 by @xmath48 where @xmath49 is the mirabolic subgroup of @xmath38 .
the integral converges absolutely ( even if @xmath7 is only assumed to be unitarizable and generic ) .
moreover , it is known that the image of @xmath50 indeed lies in @xmath45 @xcite . in the other direction when @xmath7 is tempered , define @xmath51 by @xmath52 the meaning of the regularized integral ( in a more general context ) will be explained in [ sec : stable ] .
it is a slight extension of the stable integrals studied in @xcite .
we show that the maps @xmath50 and @xmath53are inverse to each other ( proposition [ prop : inversionm ] ) .
this result is a version of local rankin
selberg unfolding .
it is closely related to the material of @xcite .
see also @xcite for a similar result in a different setup .
an immediate consequence is that we can define @xmath50 with respect to the other mirabolic subgroup without changing the result ( see proposition [ prop : sametran ] which is analogous to ( * ? ? ?
* corollary 7.2 ) ) .
next , we will ` inflate ' the @xmath54 result to the group @xmath55 ( a subgroup of @xmath56 of rank @xmath57 ) .
we identify @xmath38 with the levi subgroup of the siegel parabolic subgroup @xmath58 of @xmath1 via @xmath59 .
let @xmath60 be the subgroup of upper unitriangular matrices in @xmath1 .
let @xmath5 be the centralizer of @xmath61 in @xmath1 ; it is isomorphic to @xmath62 . with @xmath7 as before , consider the induced representation @xmath63 realized on the space of smooth functions @xmath64 such that for all @xmath65 the function @xmath66 on @xmath38 belongs to @xmath39 .
the representation @xmath67 admits a unique irreducible quotient ( the langlands quotient ) which we denote by @xmath68 . the representation @xmath68 has a unique model @xmath69 ( theorem [ thm : mwl ] ) .
meanwhile let @xmath70 be the extension @xmath41 to @xmath60 , trivial on @xmath71 .
then @xmath68 has also a unique model @xmath72 ( proposition [ prop : tadic ] ) .
it can be realized as the image of @xmath67 under the standard intertwining operator @xmath73 given by @xmath74 for a suitable weyl element @xmath75 .
we define the map @xmath76 here @xmath77 is the mirabolic subgroup of @xmath56 . in the other direction
we define @xmath78 by @xmath79 once again the regularized integral is defined in [ sec : stable ] .
we show that @xmath80 defines a model transition from @xmath72 to @xmath69 , whose inverse is @xmath81 ( theorem [ thm : mwl ] ) . in particular ,
the integral defining @xmath80 is an @xmath5-invariant functional .
note that in general @xmath82 is finite - dimensional ( but not necessarily zero or one - dimensional ) if @xmath83 is an irreducible representation of @xmath1 .
( we do not know whether @xmath84 for all @xmath83 . )
one can also realize @xmath69 by the map @xmath85 given by the absolutely convergent integral @xmath86 we have the following commutative diagram ( proposition [ p : l ] ) : + i(,12 ) & ^(h,1 ) ( ) + & ^(n,_n ) ( ) + where @xmath87 is the standard epsilon factor @xmath88 ( which does not depend on @xmath89 since @xmath7 has a trivial central character ) .
this leads us to the second theme of this paper which is the compatibility of model transitions .
more precisely , given three models with respect to pairs @xmath15 , @xmath90 with transition maps we will be interested in the commutativity of the diagram + ^(h_1,_1 ) & & ^(h_2,_2 ) + & ^(h_3,_2 ) & we will study this question for the representations @xmath68 above .
following ginzburg
soudry consider a certain character @xmath91 on the unipotent radical @xmath92 of the parabolic subgroup of @xmath1 with levi part @xmath93 ( @xmath94 times ) ( see [ sec : etwon ] ) .
the representation @xmath68 admits a model @xmath95 for @xmath68 ( proposition [ l : uniqueetwon ] which is based on @xcite and proposition [ prop : tadic ] ) .
we also define its variants @xmath96 and @xmath97 where @xmath98 and @xmath99 are obtained from @xmath100 by conjugation by certain elements of the normalizer of @xmath5 ( see [ sec : etwon ] ) .
( typically , @xmath101 is infinite - dimensional for a general irreducible representation @xmath83 of @xmath1 . )
we will show the following commutative diagram of model transitions ( propositions [ prop : 14 ] and [ prop : main ] ) : + & ^(h,1 ) ( ) + ^ ( , _ ) ( ) & & ^ ( , _ ) ( ) + & ^(n,_n ) ( ) + the diagonal maps @xmath102 , where @xmath20 is either @xmath103 or @xmath104 and @xmath21 is either @xmath98 or @xmath99 , are given by a suitable regularization of .
actually , the map @xmath105 on ( a suitable subspace of ) @xmath106 depends on some additional choices but its restriction to @xmath72 turns out to be independent of these choices
. see proposition [ prop : main ] for the precise statement .
the main tool for proving the results is repeated applications of fourier inversion ( see appendix [ a : int ] ) .
this idea is of course not new .
it was used by jacquet
shalika in the analysis of the exterior square @xmath107-function @xcite and plays an important role in the descent construction of ginzburg rallis
soudry @xcite .
in fact , this paper owes a lot to their work .
our point of view however is slightly different as we are interested in explicit computations in the various models .
we also mention recent work of bernstein
reznikov @xcite for a closely related theme .
the integral transforms @xmath108 and @xmath109 are related by conjugation by a certain element of the normalizer of @xmath5 ( see [ sec : etwon ] ) .
it follows from the above diagram that on @xmath72 , there is a functional equation relating the integral transforms @xmath110 and @xmath105 .
a version of this functional equation is stated in corollary [ cor : fornextpaper ] .
the results of this paper will be used in our work on whittaker - fourier coefficients of automorphic forms on the metaplectic double cover of @xmath111 .
more precisely , the results ( and their variants proved in [ sec : mfactor ] ) will be used in proving a conjectural local identity stated in @xcite .
this local identity is the local counterpart ( in the case of the metaplectic group ) of a global conjecture which we proposed in @xcite .
( the conjecture in @xcite is for any quasi - split group , as well as the metaplectic group . )
the paper is organized as follows . in [ sec : stable ] we introduce a rather general procedure for regularization of oscillatory integrals involving generic characters . in [ sec : invglnn ] we consider @xmath54 and the model transition between the whittaker model and the @xmath112-model ( for distinguished tempered representations @xmath7 ) .
the rest of the paper will be devoted to the case of langlands quotient @xmath68 on @xmath113 ( with @xmath7 as before ) . in
[ sec : lq1 ] we prove the uniqueness of @xmath114-model for @xmath68 and factor the intertwining operator explicitly through a model transition @xmath81 between the @xmath114-model and the degenerate whittaker model @xmath104 of @xmath68 .
the inverse of @xmath81 is constructed in [ sec : invtrans ] . in [ sec : etwon ] we introduce the additional models @xmath95 , @xmath96 and @xmath97 of @xmath68 . in
[ sec : analytic ] and [ sec : lemma ] we introduce the ( somewhat delicate ) family of operators @xmath115 ( closely related to @xmath105 mentioned above ) .
the compatibility of the various model transitions for the @xmath114 , @xmath104 and @xmath99 models is proved in [ sec : whitdesc ] and [ sec : prf of prop : main ] .
additional results which are variants of the above ( and which will be used in a subsequent paper ) are proved in [ sec : mfactor ] .
there are also three appendices . in the first one we provide the setup for the ` root exchange ' procedure in the spirit of @xcite .
the second appendix is a certain estimate which is a close variant of a result of waldspurger in @xcite .
the last appendix , due to marko tadi , establishes the uniqueness of the degenerate whittaker models @xmath104 for the representations @xmath68 .
we take this opportunity to thank tadi for his contribution .
we also thank joseph bernstein , patrick delorme , atsushi ichino , andre reznikov , yiannis sakellaridis , david soudry and akshay venkatesh for helpful correspondence . throughout the paper
let @xmath0 be a @xmath9-adic field with ring of integer @xmath116 .
we fix a ( continuous ) non - trivial character @xmath117 .
let @xmath118 be the size of residue field of @xmath0 .
thus @xmath119 for @xmath120 .
we take the self - dual haar measure @xmath121 on @xmath0 with respect to @xmath89 .
we take the multiplicative measure of @xmath122 to be @xmath123 .
we typically denote algebraic varieties ( or groups ) over @xmath0 by boldface letters ( e.g. , @xmath124 ) and denote their set ( or group ) of @xmath0-points by the corresponding plain letter ( e.g. , @xmath33 ) .
( in most cases @xmath124 will be clear from the context . ) for an @xmath125-group @xmath126 let @xmath127 be the set of compact open subgroups of @xmath126 .
we denote the group of continuous characters of @xmath126 ( with the compact - open topology ) by @xmath128 . in particular ,
if @xmath126 is abelian then @xmath128 is the pontryagin dual of @xmath126 .
let @xmath129 be the center of @xmath126 , and @xmath130 the identity element of @xmath126 .
if @xmath126 acts on a vector space @xmath131 and @xmath132 is a subgroup of @xmath126 , we denote by @xmath133 the subspace of @xmath132-fixed points . if @xmath134 and @xmath135 are subgroups of @xmath126 and @xmath134 normalizes @xmath135 , we write @xmath136 for the corresponding modulus function on @xmath134 . in particular , we write @xmath137 for the modulus function of @xmath134 acting on itself by conjugation .
if @xmath132 is a closed subgroup of @xmath126 and @xmath4 is a character of @xmath132 , we denote by @xmath138 ( resp . ,
@xmath139 , @xmath140 , @xmath141 ) the spaces of continuous ( resp .
@xmath126-smooth , ] , compactly supported modulo @xmath132 , smooth and compactly supported modulo @xmath132 ) complex - valued left @xmath142-equivariant functions on @xmath126 . when @xmath143 we omit it from the notation .
we use the following bracket convention for iterated integrals : @xmath144 implies that the inner integrals converge as a double integral and after evaluating them , the outer double integral is absolutely convergent .
let @xmath145 be a continuous function on @xmath126 and @xmath146 a function on a subgroup @xmath134 of @xmath126 .
let @xmath132 be a subgroup of @xmath126 .
we denote by @xmath147 the integral @xmath148 and by @xmath149 the integral @xmath150 .
implicit in the notation is that the integral makes sense and converges .
if @xmath151 then we suppress it from the notation .
we denote by @xmath152 and @xmath153 an integral that is regularized in a certain way .
we caution the reader that the regularization depends on the context .
hopefully this will not cause confusion
. let @xmath154 be the identity matrix in @xmath155 , @xmath156 the @xmath157-matrix with ones on the non - principal diagonal and zeros elsewhere ; @xmath158 is the transpose of a matrix @xmath159 .
we use the notation @xmath160 to mean that @xmath161 with @xmath162 a constant depending on @xmath163 .
we denote by @xmath164 the set of irreducible smooth ( complex ) representations ( up to equivalence ) of a group @xmath126 .
more specific notation will be introduced in [ sec : notationgl2n ] and [ sec : notation ] . for the convenience of the reader
there is an index of symbols at the end of the paper .
we extend the definition of stable integral introduced in @xcite .
suppose that @xmath165 is a unipotent group over @xmath0 and @xmath134 is a compact group which acts continuously on @xmath71 .
denote by @xmath166 the set of compact open @xmath134-invariant subgroups of @xmath71 .
this is a directed set .
moreover , any relatively compact subset of @xmath71 is contained in an element of @xmath166 .
[ sec : stint ] fix a haar measure @xmath167 on @xmath71 .
[ def : stablenormal ] let @xmath145 be a continuous function on @xmath71 .
we say that @xmath145 has a _
@xmath168-stable integral _ over @xmath71 if there exists @xmath169 such that for any @xmath170 containing @xmath171 we have @xmath172 in this case we write @xmath173 for the common value .
in other words , @xmath173 is the limit ( if it exists ) of the net @xmath174 with respect to the discrete topology of @xmath175 .
the case where @xmath134 is the trivial group was considered in @xcite under the name _
stable integral_. [ rem : largerk ] it is clear that if @xmath176 and @xmath145 has a @xmath177-stable integral over @xmath71 , then it also has a @xmath168-stable integral ( with the same value ) .
similarly , any ( right or left ) translate of @xmath145 by an element of @xmath71 has an @xmath134-stable integral , with the same value .
suppose @xmath134 is an @xmath125-group acting on @xmath71 and let @xmath145 be as before .
we say that the @xmath134-stable integral of @xmath145 exists if @xmath178 exists for any compact open subgroup @xmath176 ( or equivalently , for any sufficiently small @xmath179 ) .
we write this value , which by the previous remark does not depend on the choice of @xmath180 , as @xmath181 .
[ rem : aut ] it is clear that if @xmath182 is an automorphism of @xmath71 which commutes with the action of @xmath134 and which multiplies the haar measure on @xmath71 by @xmath183 then @xmath184 whenever one side is defined .
similarly for the @xmath134-stable integral .
we consider the following setting .
suppose that @xmath185 is a unipotent subgroup of a group @xmath126 and @xmath171 is a normal subgroup of @xmath185 such that @xmath186 is abelian .
we identify @xmath187 with the group of characters on @xmath185 that are trivial on @xmath171 .
let @xmath188 be a character of @xmath185 and let @xmath189 be a torus of @xmath126 normalizing @xmath185 and @xmath171 .
assume that the restriction @xmath190 of @xmath188 to @xmath171 is invariant under conjugation by @xmath189 .
( in the cases at hand @xmath190 is usually trivial . )
we say that @xmath188 is @xmath191-generic ( or simply @xmath189-generic if @xmath185 and @xmath171 are clear from the context ) if the map @xmath192 is open .
the prototype is the case where @xmath185 is a maximal unipotent subgroup of a quasi - split group , @xmath188 is a non - degenerate character in the usual sense , @xmath193 the derived subgroup of @xmath185 and @xmath189 is a maximal torus in the borel subgroup containing @xmath185 .
assume that @xmath188 is @xmath191-generic .
for any @xmath194 we write @xmath195 for the function on @xmath187 given by @xmath196 where of course the integral over the empty set is interpreted as @xmath197 . note that @xmath198 , if not empty , is a coset of @xmath199 , the stabilizer of @xmath188 under the conjugation action of @xmath189 .
we take the measure on @xmath198 to be the translate of the haar measure on @xmath200 . since @xmath201 is open ,
@xmath202 .
observe that @xmath203 ( where the superscript denotes the conjugation action ) and therefore , if @xmath204 then @xmath205 .
it follows that for suitable normalization of haar measures we have @xmath206 for any @xmath207 where @xmath208 .
( it suffices to check this for @xmath209 .
then the right - hand side defines a @xmath210-equivariant linear form on @xmath211 . ) in particular , for any @xmath212 the fourier transform of @xmath195 is given by @xmath213 [ lem : genstab ] let @xmath214 be as above and assume that @xmath188 is @xmath191-generic .
let @xmath215 be a subgroup of @xmath171 which is stable under @xmath189 and such that @xmath216 is trivial .
let @xmath217 for some @xmath218 .
suppose that @xmath219 for all @xmath220 .
let @xmath221 then @xmath222 descends to a function on @xmath223 which has a @xmath189-stable integral . moreover if @xmath224 , let @xmath225 , then @xmath226 with @xmath227 .
given @xmath228 , let @xmath229 containing the support of @xmath228 .
we take any @xmath230 .
then @xmath231 which by a change of variables @xmath232 becomes @xmath233 note that @xmath222 descends to a function on @xmath223 .
since by @xmath234 is the fourier transform of @xmath235 , it is in @xmath236 . in particular , the above integral is independent of @xmath237 as long as @xmath237 contains the support of @xmath234 .
thus , @xmath238 has a @xmath239-stable integral .
as for any @xmath240 there is @xmath241 with support in @xmath242 such that @xmath243 , we proved that @xmath222 has @xmath239-stable integral for any @xmath240 .
thus it has @xmath189-stable integral . the identity in the lemma follows from the above calculation . in the setting of the lemma ,
we will write @xmath244 we record some straightforward properties of this regularized integral in the following lemma . [ l : stableprop ] let @xmath145 be as in the previous lemma
. then 1 .
[ item : reginv ] @xmath245 .
[ item : extendint ] if @xmath246 , then @xmath247 .
3 . if @xmath248 stabilizes @xmath188 , then @xmath249 .
we end this section with a couple of remarks let @xmath126 be an @xmath125-group and @xmath250 closed subgroups .
suppose that @xmath251 and that the integral @xmath252 converges absolutely .
then the same is true for @xmath253 .
more precisely , ( assuming for simplicity that @xmath254 ) if @xmath255 is such that @xmath145 is right @xmath256-invariant then @xmath257 where @xmath258 is the image of @xmath256 in @xmath259 .
[ rem : partint ] let @xmath165 be a unipotent group over @xmath0 and @xmath260 , @xmath261 closed subgroups .
suppose that @xmath262 and that the integral @xmath263 converges absolutely .
then the same is true for @xmath264 where @xmath265 .
indeed , by induction on @xmath266 we may assume that @xmath260 is a maximal proper subgroup of @xmath165 and hence ( since @xmath165 is unipotent ) @xmath267 .
if @xmath268 then we apply the previous remark .
otherwise , @xmath269 and the statement is obvious . finally , we have the following elementary result .
[ l : elemc ] let @xmath270 be an @xmath125-group and @xmath271 , @xmath272 closed subgroups with a closed embedding @xmath273 .
( we do not assume that either @xmath271 or @xmath272 is abelian , or that @xmath274 is a homomorphism . )
assume that @xmath275 and @xmath276 are such that @xmath277 then @xmath278 is compactly supported .
moreover , let @xmath279 , @xmath280 a compact subset of @xmath281 and @xmath282 a compact subset of the space of closed embeddings from @xmath271 to @xmath281 ( with the compact - open topology ) .
assume that the images ( in @xmath281 ) of all @xmath283 coincide .
is precompact for any compact @xmath284 . ]
then there exists a compact set @xmath285 depending on @xmath256 , @xmath280 and @xmath282 such that for any @xmath286 satisfying , the support of @xmath278 is contained in @xmath285 .
* throughout the paper we fix an integer @xmath287 ( not to be confused with a running variable @xmath94 ) . * @xmath288 is the group @xmath289 .
* @xmath290 is the standard maximal unipotent subgroup of @xmath288 consisting of upper unitriangular matrices ; @xmath291 is the maximal torus of @xmath288 consisting of diagonal matrices ; @xmath292 is the borel subgroup .
@xmath293 is the standard maximal compact subgroup of @xmath288 .
* @xmath294 is the non - degenerate character of @xmath290 given by @xmath295 * @xmath296 is the derived group of @xmath290 .
* @xmath297 represents the longest weyl element of @xmath288 .
* @xmath298 is the outer automorphism of @xmath288 given by @xmath299 .
* @xmath300 is the centralizer of @xmath43 in @xmath288 , isomorphic to @xmath301 .
* @xmath302 is the set of irreducible representations of @xmath288 of metaplectic type ,
i.e. those which admit a nontrivial @xmath300-invariant form .
such an invariant form is unique up to a scalar @xcite
. any @xmath303 is self - dual .
* we write @xmath304 and @xmath305 for the sets of irreducible generic and tempered representations of @xmath288 respectively .
* @xmath306 is the mirabolic subgroup of @xmath155 consisting of the elements @xmath159 whose last row is @xmath307 .
let @xmath308 .
* the lie algebra @xmath309 of @xmath155 consists of the @xmath157-matrices @xmath33 over @xmath0 .
let @xmath310 be the lattice of integral matrices in @xmath309 .
for any algebraic subgroup @xmath311 of @xmath155 defined over @xmath0 let @xmath312 be the lie algebra of @xmath311 .
the lattice @xmath313 of @xmath314 gives rise to a gauge form of @xmath311 ( determined up to multiplication by an element of @xmath315 ) and we use it to define a haar measure on @xmath126 by the recipe of @xcite .
[ [ section ] ] for @xmath316 we would like to realize the @xmath300-invariant form @xmath317 explicitly on the whittaker model @xmath318 .
conversely , we would like to express the whittaker function in terms of @xmath317 .
this is analogous to the situation of the inner product considered in @xcite .
* lemma 2.1 ) [ p : lm ] assume @xmath316 is unitarizable .
then for any @xmath319 the integral @xmath320 converges and defines a nontrivial @xmath300-invariant linear form on @xmath321 .
thus @xmath322 defines a map from @xmath323 to @xmath324 .
writing ( for a suitable haar measure of @xmath325 ) @xmath326 the convergence follows from the estimates of ( * ? ? ? * lemma 2.1 ) and the fact that on @xmath327 we have @xmath328 clearly , @xmath317 defines a @xmath329-invariant linear form which is non - trivial since we can take @xmath330 to be an arbitrary function in @xmath331 .
lemma 2.1 of @xcite states that the space of @xmath329-invariant linear forms on @xmath7 is one - dimensional .
as @xmath7 possesses a nontrivial @xmath300-invariant linear form , @xmath317 must be @xmath300-invariant .
we observe the following fact : [ l : cuspidal ] suppose that @xmath332
. then @xmath333 .
in particular , if @xmath316 and @xmath319 is such that @xmath334 then the function @xmath335 on @xmath336 is compactly supported modulo @xmath337 .
let @xmath338 be _ any _ @xmath339-relatively cuspidal representation of @xmath288 .
( see @xcite for their existence . ) by the condition on @xmath131 , there exists @xmath340 such that @xmath341 .
it is clear from the definition of @xmath342 that @xmath343 for all @xmath344 . on the other hand , by our assumption on @xmath338 the function @xmath345
is compactly supported on @xmath346 .
the claim follows since @xmath347 is closed in @xmath346 .
( identifying @xmath348 with the non - zero vectors in @xmath349 , the orbit of @xmath339 is given by the vanishing of all odd coordinates , hence closed . )
we would like to apply lemma [ lem : genstab ] to define the inverse transform of @xmath350 , at least in the case that @xmath7 is tempered .
to that end we first provide a bound of the generalized matrix coefficient @xmath351 .
let @xmath352 .
we construct an @xmath300-invariant form @xmath353 on the space of @xmath354 by setting @xmath355 where @xmath356 .
this is well defined . in the case
@xmath357 @xmath358 and the convergence follows since the integrand is @xmath359 . in the general case
let @xmath360 be the standard parabolic subgroup of @xmath288 with levi subgroup @xmath361 .
then @xmath362 , @xmath363 , @xmath364 is a borel subgroup of @xmath300 and @xmath365 thus , @xmath366 where @xmath367 .
therefore , the convergence reduces to the case @xmath357 considered above . for @xmath368 denote by @xmath369 the maximum of the absolute values of the entries of @xmath159 and @xmath370 , and
let @xmath371 .
we have @xmath372 , @xmath373 and @xmath83 is bi-@xmath293-invariant .
we define @xmath374 so that @xmath375
. clearly @xmath376 .
let @xmath377 be the unramified vector in @xmath354 such that @xmath378 .
let @xmath379 , so that @xmath380 .
[ lem : ellsigma ] for any @xmath381 @xmath382 we will use the integration formula @xmath383 here @xmath384 and @xmath385 are the images under transpose of @xmath386 and @xmath290 respectively .
let @xmath387 .
as @xmath388 normalizes @xmath300 , by ( * ? ?
* theorem 3.2 ) we have @xmath389 .
we can unwind the left - hand side of to @xmath390 it follows from lemma [ lem : mainbnd018 ] ( with @xmath391 ) that the integral over @xmath392 and @xmath393 converges and that there exists @xmath394 such that for any subset @xmath395 and integers @xmath396 , @xmath397 , the integral over @xmath392 , @xmath398 , @xmath397 and @xmath399 , @xmath400 is @xmath401 .
the lemma follows .
we also need a lower bound on @xmath402 .
let @xmath403 be the centralizer of @xmath404 , which is conjugate to @xmath300 by @xmath405 .
define @xmath406 .
it follows easily from and the relation @xmath407 that for a suitable inessential constant @xmath162 @xmath408 where the integral converges .
let @xmath409 [ lem : lwrbndxih ] @xmath410 for all @xmath411 and @xmath412 .
clearly , for a suitable choice of haar measure we have @xmath413 here we used the fact that if @xmath10 is any locally compact group , @xmath11 is a closed subgroup of @xmath10 and @xmath414 is a compact open subgroup of @xmath10 then up to normalization of measures we have @xmath415 for any non - negative function @xmath416 .
if @xmath417 then @xmath418 for all @xmath412 by ( * ? ? ?
* lemma ii.3.2 ) .
the lemma follows .
[ lem : bndbytmpr ] let @xmath419 and let @xmath107 be an @xmath300-invariant form on the space of @xmath7 .
then there exists @xmath381 such that for any @xmath420 @xmath421 consequently , @xmath422 the second part of the lemma follows from the first one and lemma [ lem : ellsigma ] . in order to prove the first part
, it is convenient to work with the centralizer @xmath423 of @xmath388 , which is conjugate to @xmath403 by a permutation matrix @xmath75 . define @xmath424 and @xmath425 . note that @xmath426 is a @xmath427-split borel group of @xmath288 where @xmath428 .
thus , by ( * ? ? ?
* lemme 2 ) , if @xmath429 is an @xmath430-invariant form on @xmath7 then there exists a vector @xmath431 ( the dual of @xmath7 ) such that @xmath432 for all @xmath417 .
since @xmath7 is tempered it follows ( cf .
* lemme ii.1.1 ) ) that there exists @xmath381 such that can be chosen independently of @xmath7 by @xcite ] @xmath433 let @xmath434
. then @xmath435 and by lemma [ lem : lwrbndxih ] , @xmath436 since @xmath437 ( see e.g. , ( * ? ? ?
* propotion 3.1 ) ) it follows that @xmath438 the first part of the lemma follows immediately .
let @xmath440 and consider @xmath441 .
the character @xmath294 is @xmath442-generic and is trivial on @xmath443 . by lemma
[ lem : bndbytmpr ] , the conditions of lemma [ lem : genstab ] are satisfied and from lemma [ l : stableprop ] part [ item : reginv ] we get a map from @xmath324 to @xmath323 : @xmath444 [ sec : regintng ] the next proposition is analogous to ( * ? ? ?
* lemma 4.4 ) .
[ prop : inversionm ] assume that @xmath445 .
then @xmath446 namely , @xmath447 is the inverse transform of @xmath350 between models of @xmath7 . to prove it is enough to consider @xmath391 . by lemma [ l : stableprop ] ,
the right - hand side of is a whittaker functional . by uniqueness of whittaker functional ,
both sides are proportional and hence we only need to prove the identity for @xmath131 whose restriction to @xmath336 is compactly supported modulo @xmath448 .
this would follow from the following functional equation and lemma [ l : stableprop ] part [ item : extendint ] .
[ l : feglnn ] if @xmath449 then @xmath450 and @xmath451 .
the fact @xmath333 follows from lemma [ l : cuspidal ] .
we identify @xmath329 with the product of @xmath452 and the mirabolic subgroup of @xmath452 .
we write @xmath453 for the character of @xmath329 given by the determinant of the @xmath452 factor and @xmath454 for the determinant of the other factor . for convenience we also set @xmath455 for @xmath456 .
for any @xmath457 let @xmath458 ( resp .
@xmath459 ) be the mirabolic subgroup of @xmath460 ( resp .
the group of upper unitriangular matrices of @xmath460 ) .
we embed @xmath460 ( and its subgroups ) in @xmath288 via @xmath461 .
note that @xmath462 where @xmath463 is the subgroup of unipotent matrices in @xmath464 whose upper left @xmath465 corner is the identity matrix .
then @xmath466 we will prove by descending induction on @xmath467 that @xmath468 on @xmath336 .
the case @xmath469 is the required identity .
the base of the induction , the case @xmath470 , is a tautology . for the induction step
we prove that for any character @xmath4 of @xmath471 @xmath472 for all @xmath344 .
as @xmath473 and @xmath474 we have @xmath475 thus we can write the left - hand side of explicitly as @xmath476 let @xmath477
. then @xmath478 identifies @xmath479 with @xmath480 . for @xmath481 , @xmath482 and @xmath483
, we have @xmath484 where @xmath485 is the standard isomorphism between @xmath486 and @xmath487 and @xmath488 is the standard pairing on @xmath489 , which allows us to view @xmath486 as the dual of @xmath487 .
the set @xmath490 is the non - zero vectors in a subspace @xmath491 of @xmath487 , and @xmath488 restricts to a non - degenerate pairing between @xmath491 and @xmath492 .
also @xmath493 for @xmath494 .
therefore the left - hand side of is : @xmath495 by our support condition on @xmath131 , the function @xmath496 extends to a schwartz function on @xmath491 .
the right - hand side of is @xmath497 .
the left - hand side of becomes @xmath498 thus , the relation follows by fourier inversion .
let @xmath499 be the second mirabolic subgroup of @xmath288 consisting of matrices whose first column is @xmath500 .
then if @xmath501 where @xmath502 is unitarizable , the same argument as in proposition [ p : lm ] shows that @xmath503 converges and defines another nontrivial @xmath300-invariant linear form on @xmath321 . by uniqueness
, @xmath504 is a scalar multiple of @xmath317 .
the following functional equation is in the spirit of ( * ? ? ?
* corollary 7.2 ) .
[ prop : sametran ] for any unitarizable @xmath502 and @xmath501 we have @xmath505 .
we first assume that @xmath7 is tempered .
let @xmath506 . we show that @xmath507 , which will imply that @xmath508 on @xmath321 by proposition [ prop : inversionm ] .
let @xmath509 . then @xmath510 .
for @xmath131 as above , as @xmath511 , @xmath512 and @xmath513 @xmath514 which by proposition [ prop : inversionm ] ( applied to @xmath515 and @xmath516 ) is @xmath517 .
using the classification of irreducible generic representation of metaplectic type ( @xcite ) , a standard argument ( see @xcite ) extends the result from the tempered representations to all unitarizable @xmath518 .
we keep the notation of [ sec : notationgl2n ] and introduce additional notation as follows . * @xmath519 .
* @xmath520 is the siegel parabolic subgroup of @xmath1 , with its standard levi decomposition .
@xmath521 is the opposite parabolic of @xmath522 , with unipotent radical @xmath523 . @xmath524 .
* we use the isomorphism @xmath525 to identify @xmath288 with @xmath526 .
* @xmath60 is the standard maximal unipotent subgroup of @xmath1 consisting of upper unitriangular matrices ; @xmath527 is the maximal torus of @xmath1 consisting of diagonal matrices ; @xmath528 is the standard borel subgroup of @xmath1 . * for a subgroup @xmath33 of @xmath1 , @xmath529 denotes @xmath530 .
in particular @xmath531 and @xmath532 .
* @xmath533 .
we set @xmath534 .
* let @xmath535 be the standard maximal compact subgroup of @xmath1 .
* let @xmath536 be the character @xmath537 of @xmath38 .
extend @xmath536 to a left-@xmath71 right-@xmath535 invariant function on @xmath1 using the iwasawa decomposition .
* @xmath5 is the centralizer of @xmath538 in @xmath1 , isomorphic to @xmath539 . *
write @xmath540 for any subgroup @xmath33 of @xmath1 .
in particular , @xmath541 and @xmath542 .
* let @xmath543 be the twisted transpose map on the space @xmath544 of @xmath157 matrices given by @xmath545 .
let @xmath546 .
* define @xmath547 to be the isomorphism given by @xmath548 .
similarly @xmath549 is the isomorphism from @xmath550 to @xmath551 .
* @xmath552 ; @xmath553 represents the longest @xmath526-reduced weyl element of @xmath1 . *
let @xmath554 be the matrix with one at the @xmath555-entry and zeros elsewhere .
let @xmath556 when @xmath557 and @xmath558 . * @xmath559 ( @xmath560 ) is the one - parameter group @xmath561 .
@xmath562 is the simple root group @xmath563 and @xmath564 .
* @xmath566 is the character on @xmath551 given by @xmath567 .
@xmath70 is the _ degenerate _
character on @xmath60 given by @xmath568 for any @xmath569 and @xmath570 .
@xmath571 is the restriction of @xmath70 to @xmath526 , i.e. , @xmath572 . *
@xmath573=xyx^{-1}y^{-1}$ ] denotes the commutator of @xmath574 and @xmath575 .
* @xmath576 is the group generated by @xmath577 . *
the convention of haar measures will be as in [ sec : notationgl2n ] .
namely , we will use the lattice of @xmath578-matrices integral matrices to define a gauge form ( up to an element of @xmath315 ) for any algebraic subgroup of @xmath579 ( and in particular , any algebraic subgroup of @xmath2 ) defined over @xmath0 .
[ [ section-1 ] ] let @xmath580 .
we consider @xmath7 as a representation of @xmath526 via @xmath581 . for @xmath582 ,
let @xmath583 .
note that @xmath584 is a parabolic subgroup of @xmath5 with levi decomposition @xmath585 .
the opposite parabolic with respect to @xmath586 is @xmath587 .
we have @xmath588 as was observed in the course of the proof of ( * ? ? ?
* , theorem 2 ) , if @xmath589 then @xmath67 admits a non - trivial @xmath5-invariant functional .
more precisely , if @xmath125 is an @xmath339-invariant functional on @xmath7 then @xmath590 this is well defined since by @xmath591 for all @xmath592 .
we will need the following complementary information .
[ prop : distequi ] let @xmath593 .
then @xmath67 is @xmath5-distinguished if and only if @xmath7 is @xmath586-distinguished . in this case
@xmath67 has a unique @xmath5-invariant form up to a constant multiple .
we argue using the geometric lemma of bernstein zelevinsky .
the orbits of @xmath5 on @xmath594 , ordered by dimension , are indexed by integers @xmath595 .
let @xmath596 be representatives of the orbits and let @xmath597 be the stabilizer .
let @xmath598 .
correspondingly , the representation @xmath67 admits a filtration @xmath599 by @xmath5-invariant subspaces such that @xmath600 ( here @xmath601 denotes unnormalized compact induction and the superscript denotes conjugation by @xmath596 . ) let @xmath602 be the standard parabolic subgroup of @xmath1 contained in @xmath522 such that @xmath603 . as described in @xcite we can choose @xmath596 such that @xmath604 where 1 .
the projection of @xmath606 to @xmath526 is @xmath607 , the unipotent radical of @xmath608 ( which is the parabolic subgroup of @xmath526 type @xmath609 ) .
3 . @xmath610 .
it follows by a simple computation that @xmath611 is the restriction of @xmath612 . by a standard argument using frobenius reciprocity ( cf .
@xcite ) we conclude that @xmath613 here @xmath614 is the jacquet module of @xmath7 with respect to @xmath608 .
we show below that @xmath615 when @xmath616 .
this would imply that @xmath617 from which the proposition follows .
in fact , we will show that ( for @xmath616 ) @xmath618 for any irreducible subquotient @xmath619 of @xmath614 .
denote by @xmath620 the central character of an irreducible representation @xmath621 .
as @xmath593 , by @xcite @xmath622 where for @xmath623 , @xmath624 is the unique irreducible quotient @xmath625_{\rho_i})$ ] of @xmath626 .
here we use @xmath627 to denote parabolic induction ; @xmath628 are irreducible supercuspidal representations of @xmath629 with @xmath630 ; @xmath631 .
consider an irreducible subquotient @xmath619 of @xmath614 .
then by the geometric lemma and the description of the jacquet modules of square - integrable representations , there exist @xmath632 , @xmath623 such that @xmath633 is a subquotient of @xmath634 and @xmath635 is a subquotient of @xmath636 where @xmath637_{\rho_i})$ ] and @xmath638_{\rho_i})$ ] . if @xmath639 for some @xmath467 then @xmath640 and @xmath641 with @xmath642 , in which case @xmath633 is not of metaplectic type .
otherwise @xmath635 has the form @xmath643 with @xmath644 .
in particular @xmath635 is generic and hence does not admit a non - trivial functional invariant under @xmath645 by @xcite . in both cases
we have @xmath618 as required .
it is conceivable that in general , any @xmath646 admits at most one @xmath5-invariant functional up to a scalar .
however , this is an open problem except for @xmath647 ( @xcite ) .
for any @xmath648 and @xmath649 define @xmath650 , @xmath65 .
let @xmath651 .
let @xmath652 be the space of @xmath1-smooth left @xmath71-invariant functions @xmath653 such that for all @xmath65 , the function @xmath654 on @xmath526 belongs to @xmath655 .
for any @xmath649 we have a representation @xmath656 on the space @xmath652 given by @xmath657 , @xmath658 .
then the representation @xmath656 is isomorphic to @xmath659 .
it will be convenient to set @xmath660_{s , g}(m)={\delta}_p({\varrho}(m))^{-\frac12}w({\varrho}(m)g){\nu(g)}^s , \ \
g\in g,\ , m\in{{\mathbb m}},\ , s\in{\mathbb{c}}\ ] ] so that @xmath661_{s , g}\in{{\mathbb{w}}^{\psi_{n_{{\mathbb m}}}}}(\pi)$ ] . thus , @xmath662_{s , g_2}=[w]_{s , g_2g_1}$ ] and for @xmath663 , @xmath570 and @xmath65 we have @xmath664_{s,{\varrho}(m)ug}={\delta}_p({\varrho}(m))^{\frac12}{\left|{\det m}\right|}^s\pi(m)[w]_{s , g}.\ ] ] we can explicate as follows .
[ lem : wl ] let @xmath665 , considered also as a representation of @xmath526 through @xmath581 .
assume that @xmath7 is unitarizable .
then a nontrivial @xmath5-invariant linear form on @xmath666 is given by @xmath667_{\frac12,h})\ dh= \int_{h_{\bar u}}{\mathfrak{p}^{h_{{{\mathbb m}}}}}([w]_{\frac12,u})\ du\ ] ] where @xmath317 is defined in proposition [ p : lm ] .
we define the integral transform @xmath668 by @xmath669_{\frac12,hg})\ dh.\ ] ] we will show that if @xmath7 is tempered then @xmath670 factors through the langlands quotient @xmath68 of @xmath671 , and hence gives rise to a model @xmath672 of the langlands quotient .
we will also define an integral transform @xmath81 from @xmath69 to @xmath72 as well as its inverse .
the fact that there is a unique model @xmath72 is proved in appendix [ sec : tadic ] . for any @xmath674
we denote by @xmath675 the space of right-@xmath256-invariant functions @xmath107 on @xmath673 such that for all @xmath65 the integral @xmath676 converges absolutely .
we endow @xmath675 with the topology given by the seminorms @xmath677 we define @xmath678 with the inductive limit topology .
note that @xmath679 .
define @xmath680 as the subspace of @xmath681 consisting of @xmath682 such that @xmath683 .
the character @xmath70 is @xmath684-generic and is trivial on @xmath685 .
thus , by lemma [ lem : genstab ] we can define for @xmath686 @xmath687 note that @xmath688 when the left - hand side is well defined , where @xmath689 on the right - hand side was defined in [ sec : regintng ] .
[ l : stablemodel ] 1 .
[ mod : dense ] the space @xmath690 of compactly supported right @xmath256-invariant functions on @xmath673 is dense in @xmath675 .
hence @xmath691 is dense in @xmath681 .
the expression @xmath692 is a continuous functional on @xmath675 .
the function @xmath693 belongs to @xmath694 . to @xmath695 ,
trivially on @xmath696 . ]
the first part is clear .
the second part follows from ( applied with @xmath697 ) .
the last part follows from lemma [ l : stableprop ] .
we go back to the notation of [ sec : matcoefbnd ] .
let @xmath698 be the function @xmath699 thus @xmath700 where @xmath701 is the unramified section of @xmath702 .
as before let @xmath703 , @xmath65 .
[ lem : conv192 ] for any @xmath394 and @xmath704 @xmath705 note that the integrand is non - negative . by the gindikin
karpelevich formula , the integral @xmath706 converges for @xmath704 and is equal to @xmath707 where @xmath708 and as usual @xmath709 . on the other hand
, it follows immediately from ( * ? ? ?
* lemme ii.3.4 ) that for any @xmath710 and @xmath394 we have @xmath711
. therefore the lemma follows from lemma [ lem : ellsigma ] .
let @xmath651 .
assume that @xmath7 is self - dual .
we recall the intertwining operators @xmath712 given by @xmath713 in the range of convergence .
( the element @xmath538 is introduced in order to preserve the character @xmath571 . )
thus , @xmath714_{-s , g})=\int_u\ell([w]_{s,{\varrho}(e)w_uug})\ du\ ] ] for any @xmath125 in the smooth dual of @xmath7 .
in fact , holds for any functional @xmath125 provided that the right - hand side is absolutely convergent .
indeed , both sides are unchanged if we replace @xmath125 by @xmath715 for sufficiently small @xmath716 .
we also set @xmath717 so that if @xmath7 is tempered then @xmath718 defines a @xmath1-intertwining map from @xmath666 to @xmath72 .
we denote by @xmath719 the sign such that @xmath720 where @xmath125 is a non - trivial @xmath300-invariant linear form on @xmath7 .
( by uniqueness , @xmath719 exists and does not depend on the choice of @xmath125 . )
it is known that for @xmath518 we have @xmath721 ( the ` standard ' @xmath722-factor of @xmath7 , which of course does not depend on @xmath89 since @xmath7 has a trivial central character )
see ( * ? ? ?
* theorem 3.2 ) .
however , we will not use this fact .
[ p : l ] let @xmath419 considered also as a representation of @xmath526 through @xmath581 . then @xmath723 and @xmath724 where @xmath725 let @xmath726 .
we show that @xmath727 .
first , by and for any @xmath728 we have @xmath729_{\frac12,\bar u { { w_0^{m}}}g})\,d\bar u \\=\int_{h_{\bar u}}{\mathfrak{p}^{h_{{{\mathbb m}}}}}(\pi({{w_0^{{\mathbb m}}}})([w]_{\frac12,\bar u g}))\,d\bar u = { \epsilon_\pi}\int_{h_{\bar u}}{\mathfrak{p}^{h_{{{\mathbb m}}}}}([w]_{\frac12,\bar u g})\,d\bar u={\epsilon_\pi}l_w(g).\end{gathered}\ ] ] next we show the convergence of for @xmath730 .
we have @xmath731_{\frac12,w_uvug})\ dv}\right|}\ du \le\int_{u}{\left|{{\mathfrak{p}^{h_{{{\mathbb m}}}}}([w]_{\frac12,w_uug})}\right|}\ du.\ ] ] therefore @xmath732_{\frac12,w_uun})}\right|}\ du\ dn\\= \int_u\int_{n_{m}\cap h{\backslash}n_{m}^{{\operatorname{der}}}}{\left|{{\mathfrak{p}^{h_{{{\mathbb m}}}}}(\pi(n)[w]_{\frac12,w_uu})}\right|}\ dn\ du.\end{gathered}\ ] ] we write @xmath733 where @xmath734 , @xmath735 and @xmath736 . since @xmath661_{\frac12,k}$ ] , @xmath737 range over a finite set of vectors in @xmath321 we conclude from lemma [ lem : bndbytmpr ] that there exists @xmath394 such that for any @xmath738 with @xmath739 @xmath740_{\frac12,w_uu})}\right|}= { \delta}_p(m_1)^{\frac12}{\nu(m_1)}^{\frac12}{\left|{{\mathfrak{p}^{h_{{{\mathbb m}}}}}(\pi(mm_1)[w]_{\frac12,k_1})}\right|}\\\ll_w { \delta}_p(m_1)^{\frac12}{\nu(m_1)}^{\frac12}\xi_{h_{{\mathbb m}}}(mm_1)\sigma_{h_{{\mathbb m}}}(mm_1)^a\\\le ( \xi_h)_\frac12(mw_uu)(\sigma_{h_{{\mathbb m}}}(m)+2\sigma(m_1))^a\ll ( \xi_h)_\frac12(mw_uu)(\sigma_{h_{{\mathbb m}}}(m)+\sigma(u))^a.\end{gathered}\ ] ] the convergence of @xmath741 therefore follows from lemma [ lem : conv192 ] .
we get @xmath727 .
similarly , since @xmath742 we have @xmath743_{\frac12,w_uvug})\ dv\big)\ du \\=\int_u { \mathfrak{p}^{h_{{{\mathbb m}}}}}([w]_{\frac12,w_uug})\ du=\int_u { \mathfrak{p}^{h_{{{\mathbb m}}}}}([w]_{\frac12,{\varrho}(e)w_uug})\ du= { \mathfrak{p}^{h_{{{\mathbb m } } } } } ( [ m(\frac12)w]_{-\frac12,g})\end{gathered}\ ] ] where the last equality follows from .
it follows from that @xmath744 equals @xmath745_{-\frac12,{\varrho}(n)g})\psi_{n_{{\mathbb m}}}^{-1}(n)\,dn\\= { \epsilon_\pi}{\int^{reg}}_{n_{{\mathbb m}}\cap h_{{{\mathbb m}}}{\backslash}n_{{\mathbb m}}}{\mathfrak{p}^{h_{{{\mathbb m } } } } } ( \pi(n)[m(\frac12)w]_{-\frac12,g})\psi_{n_{{\mathbb m}}}^{-1}(n)\,dn={\epsilon_\pi}[m(\frac12)w]_{-\frac12,g}(e)={\epsilon_\pi}m^*w(g)\end{gathered}\ ] ] where the regularized integral @xmath746 is as defined in [ sec : regintng ] and in the last line we used the relation .
next we show that @xmath747 defined above ( and hence @xmath670 ) factors through the langlands quotient @xmath68 .
indeed we will define an inverse transform of @xmath81 .
let @xmath524 .
[ thm : mwl ] let @xmath748
. then 1 .
the integral @xmath749 converges for @xmath750 .
@xmath68 is @xmath5-distinguished with a unique @xmath5-invariant linear form up to scalar multiple .
3 . @xmath81 is a model transition from @xmath69 to @xmath72 with the inverse being @xmath80 .
we first show convergence .
let @xmath751 .
since @xmath752 is unimodular and @xmath753 , we can write @xmath754 as @xmath755 for a suitable haar measure of @xmath756 . we need to show the convergence of the inner integral .
write @xmath757 as @xmath758 with @xmath759 . then @xmath760 . on the other hand , for any @xmath704 we have @xmath761 and @xmath762 is supported on @xmath763 .
thus the convergence follows from is unitarizable since we can take @xmath764 ( ( * ? ? ?
* lemma 2.1 ) ) . ]
@xmath765 we will show that for @xmath766 : @xmath767 this would imply the second and third part of the theorem . indeed , by proposition [ p : l ] and the equation above @xmath768 on @xmath666 . in particular @xmath670 factors through @xmath769 .
this shows @xmath68 is @xmath5-distinguished . by proposition [ prop : distequi ] ,
the @xmath5-invariant linear form on @xmath68 has dimension one .
thus we have a model @xmath69 .
the proposition follows . in order to show
let @xmath770 be the image of the co - character @xmath771 , @xmath772 and let @xmath773 ( resp . ,
@xmath774 ) be the unipotent group in @xmath60 ( resp . , @xmath775 ) , consisting of elements of the form @xmath776 where the middle @xmath777 block of @xmath778 is the identity matrix .
( it is a heisenberg group of dimension @xmath779 . )
then the left - hand side of is @xmath780}}\circ{{\mathcal l}}_{\bar v^k\cap h } \circ\dots\circ{{\mathcal l}}_{\tilde t_1,\nu^{{\mathbf{n}}-1}}\circ{{\mathcal l}}_{\bar v^1\cap h}\circ{{\mathcal l}}_{n,\psi_n^{-1}}^{reg , h } \circ l_w(e).\ ] ] we introduce some auxiliary integrals . for @xmath772 ,
let @xmath781 be the unipotent radical of the standard parabolic subgroup with levi part @xmath782 and let @xmath783 . if @xmath784 then @xmath785 for @xmath786 by remark [ rem : partint ] .
since @xmath787 is @xmath788-generic we can define using lemma [ lem : genstab ] the regularized integral @xmath789 this is a continuous functional on @xmath681 .
similarly for @xmath790 let @xmath791 be the simple root group consisting of @xmath792 .
define a continuous functional on @xmath681 : @xmath793 here we use lemma [ lem : genstab ] with the data @xmath794 , @xmath795 , @xmath796 , @xmath797 . as @xmath798 and @xmath799 ,
follows from the following proposition .
[ prop : iterinv ] let @xmath790
. then 1 . for @xmath784 and @xmath65 ,
@xmath800 is compactly supported on @xmath774 and @xmath801 2 .
let @xmath802 and @xmath766 .
then @xmath803}\ dt = j_{k+1}(l_w;e).\ ] ] let @xmath804 , @xmath805 , @xmath806 and @xmath807 . then @xmath808\subset d\subset n_k$ ] and therefore , since @xmath145 is @xmath809-equivariant , for @xmath810 , @xmath811 @xmath812dc)=\psi_n([c , d])\psi_n(d)f(c).\ ] ] it is easy to check that @xmath813)$ ] defines a homeomorphism from @xmath271 to the pontryagin dual of @xmath814
. thus we can apply lemma [ l : elemc ] to conclude that @xmath815 is supported on a compact set determined by @xmath159 and @xmath256 .
in particular the integration in is a continuous functional on @xmath681 .
thus we only need to prove the first part of the proposition for @xmath816 .
the identity follows by applying lemma [ l : fourier ] for @xmath817 , @xmath818 , @xmath819 , @xmath820 , @xmath821 and @xmath822 for @xmath823 , @xmath824 .
we only need to note that the map @xmath813)$ ] defines a homeomorphism preserving haar measures between @xmath825 and the pontryagin dual of @xmath826 .
we turn to the second part of the proposition .
first we need a form of fourier inversion formula : [ l : genftinv ] let @xmath701 be a smooth function on @xmath0 .
suppose that there exists @xmath827 such that @xmath828 for all @xmath120 , @xmath829 .
assume that @xmath830
. then @xmath831 is well defined and equal to @xmath832 .
we first observe that for any @xmath833 the limit in @xmath162 ( with respect to the discrete topology ) exists . indeed , @xmath834 whenever @xmath835 .
consider @xmath836 for @xmath837 .
the above integrand in @xmath838 is supported on @xmath839 .
thus @xmath840 and hence @xmath841 by assumption .
let @xmath842 clearly @xmath843 is absolutely convergent and equals @xmath832 by fourier inversion .
on the other hand , for any @xmath844 @xmath845 thus , is absolutely convergent and moreover @xmath846 the lemma follows .
let @xmath847 where @xmath848 .
[ l : invinfprep ] there is @xmath849 such that @xmath850 when @xmath851 .
moreover , for some @xmath852 we have @xmath853 as @xmath854 .
let @xmath855 be such that @xmath856 for @xmath857 of the form @xmath858 with @xmath851 . then since @xmath770 stabilizes @xmath859 , @xmath860 by a change of variable @xmath861 .
next we show the bound .
recall @xmath862 here we can choose @xmath863 .
in particular @xmath189 commutes with @xmath864 .
thus by , the integration of @xmath865 can be taken over a compact set which is independent of @xmath574 .
moreover @xmath865 and @xmath864 commute .
thus we may ignore the integration over @xmath865 and it suffices to bound @xmath866 we show that @xmath867 we argue as in remark [ rem : partint ] . consider a sequence of subgroups @xmath868 such that @xmath869 for all @xmath870 and let @xmath871 .
we show that for all @xmath467 @xmath872 altogether we get .
if @xmath873 , this follows from and the fact that @xmath874 is independent of @xmath574 for any @xmath674 . on the other hand , if @xmath875 then @xmath876 from the bound given in the proof of proposition [ p : l ] ( combining and ) the right - hand side of is : @xmath877 for suitable @xmath878 where @xmath879 .
this by the argument of lemma [ lem : conv192 ] is @xmath880 where @xmath881 . as in
, we can unwind the above integral using the definition of @xmath402 to get @xmath882 where @xmath883 .
we integrate the above over @xmath884 and separate the two cases @xmath885 even and @xmath885 odd .
when @xmath885 is even , the same argument as in the proof of lemma [ lem : ellsigma ] shows that the integration of the above expression over @xmath884 is @xmath886 for some @xmath887 .
thus @xmath888 .
when @xmath885 is odd , setting @xmath889 we use the change of variable @xmath890 to bound the integral over @xmath884 by @xmath891 separate the integration over @xmath892 into two parts : @xmath893 and @xmath894 . by a change of variables the first part is @xmath895 which once again as in lemma [ lem : ellsigma ]
is @xmath896 for some @xmath897 . on the other hand ,
the contribution of @xmath894 is @xmath898 which again by the argument in lemma [ lem : ellsigma ] is @xmath899 for some @xmath900 .
the lemma follows .
we can now finish the proof of .
assume that @xmath901 is right invariant under @xmath674 .
let @xmath902}$ ]
. then @xmath903 . thus @xmath904}\ dt=\int_{\tilde t_k}\big({\int^{reg}}_f f_k(\lambda_{2{\mathbf{n}}-k}(x)t)\psi^{-1}(x)\ dx\big)\ dt
\\=\int_{f}\big({\int^{reg}}_f f_k(\lambda_{2{\mathbf{n}}-k}(x))\psi^{-1}(xt)\ dx\big)\ dt.\end{gathered}\ ] ] here the regularized integration over @xmath0 is a stable integral which equals @xmath905 . let @xmath906 .
then by lemma [ l : invinfprep ] , @xmath907 satisfies the conditions in lemma [ l : genftinv ] . by lemma [ l : genftinv ] , the above integral is well defined and equal to @xmath908 as required .
motivated by the results of ginzburg rallis
sodury @xcite we will consider another model for the langlands quotient . let @xmath92 be the unipotent radical of the standard parabolic subgroup of @xmath1 with levi subgroup @xmath909 ( @xmath910 times ) and let @xmath911 be a character of @xmath92 of the form @xmath912 where all @xmath913 .
note that @xmath527 acts transitively on the characters of this form .
the restriction @xmath914 of @xmath911 to @xmath915 is given by @xmath916 let @xmath917 be the stabilizer of @xmath914 under the conjugation action of @xmath527 on the characters of @xmath918 .
explicitly , @xmath919 the character @xmath911 is @xmath920-generic and trivial on @xmath921 ( which is contained in @xmath918 ) . using lemma [ lem : genstab ]
we can therefore define @xmath922 for any @xmath784 .
the following is a consequence of the results of ginzburg rallis soudry and the uniqueness of @xmath104-model of @xmath68 ( proposition [ prop : tadic ] ) .
[ l : uniqueetwon ] if @xmath419 , then @xmath68 has a unique @xmath95 model .
* theorem 5.7 ) and its proof apply to @xmath68 .
therefore , as explained in [ ibid.,5.8 ] , ( * ? ? ?
* theorem 4.3 ) also applies to @xmath68 .
hence , the lemma follows from proposition [ prop : tadic ] .
we get immediately from lemmas [ lem : genstab ] , [ l : stableprop ] and proposition [ p : l ] : @xmath923 defines a continuous functional on @xmath681 .
when @xmath802 , @xmath923 defines an intertwining map from @xmath69 to @xmath924 . later on we will show that @xmath923 is non - zero ( proposition [ prop : 14 ] ) .
we introduce some variants of the above lemma .
let @xmath925 ( resp . ,
@xmath926 ) denote the image under @xmath927 ( resp . , @xmath928 ) of the space of strictly upper triangular matrices in @xmath929 .
let @xmath930 be the unipotent subgroup @xmath931 of @xmath1 .
for instance , for @xmath932 , @xmath930 consists of the matrices in @xmath1 of the form @xmath933 clearly @xmath934 .
let @xmath935 be the character on @xmath930 given by @xmath936 to verify that @xmath935 is indeed a character let @xmath182 be the weyl element in @xmath1 such that @xmath937 , @xmath457 , @xmath938 , @xmath939 , @xmath940 , @xmath941 .
let @xmath942 be the weyl element in @xmath288 such that for @xmath943 , @xmath944 when @xmath945 mod @xmath946 and @xmath947 if @xmath948 mod @xmath946 .
let @xmath949 .
then @xmath950 , @xmath951 and the function @xmath911 on @xmath92 given by @xmath952 , @xmath953 is a character of the form with @xmath954 .
we can define @xmath955 for any @xmath686 .
we get : @xmath108 defines a continuous functional on @xmath681 .
when @xmath802 , @xmath68 has a unique @xmath96 model .
@xmath108 defines an intertwining map from @xmath69 to @xmath956 .
let @xmath957 then @xmath958 normalizes @xmath5 .
define @xmath959 and @xmath960 .
then @xmath961 .
write @xmath962 with @xmath963 .
for instance , when @xmath964 @xmath965 consists of the matrices of the form @xmath966 explicitly , @xmath967 is given by @xmath968 where @xmath567 and @xmath969 we can define @xmath970 for any @xmath686 . [ cor : uniqueetwon ] @xmath109 defines a continuous functional on @xmath681 .
when @xmath802 , @xmath68 has a unique @xmath97 model .
@xmath109 defines an intertwining map from @xmath69 to @xmath971 .
note that @xmath972 if and only if @xmath910 is even .
thus for @xmath973 , @xmath974 , and @xmath975
our next goal is to describe explicit an model transition between @xmath976 and @xmath971 which will coincide with @xmath977 but which is given more directly .
basically we want to integrate over @xmath978 , i.e. over @xmath979 .
this is a little more involved than the previous cases because of convergence issues .
we fix @xmath674 and @xmath980 throughout .
we will define a family of integral transforms from @xmath981 to @xmath982 .
eventually we will show in [ sec : prf of prop : main ] that these transforms all equal to @xmath977 when restricted to @xmath976 .
let @xmath983 be the subgroup consisting of @xmath984 where the last row of @xmath985 ( and hence also the first column ) is @xmath197 .
let @xmath986 be the one - parameter subgroup @xmath987 of @xmath551 .
the character @xmath566 is clearly @xmath988-generic and trivial on @xmath989 . using lemma [ lem : genstab ] we define @xmath990 for any @xmath991 such that the integral @xmath992 converges for all @xmath65 .
if furthermore , the integral @xmath993 converges , we denote it by @xmath994 .
it is clear that if @xmath994 is defined then for any @xmath995 @xmath996 let @xmath997 be the stabilizer of the character @xmath566 of @xmath551 in @xmath288 , i.e. : @xmath998 explicitly , @xmath997 consists of the matrices in @xmath288 whose first ( resp .
last ) column is @xmath999 ( resp .
@xmath1000 ) for some @xmath1001 .
let @xmath1002 .
note that @xmath426 normalizes @xmath989 and @xmath1003 . by remark [ rem : aut ]
we have @xmath1004 for any @xmath1005 , provided that @xmath994 is well defined .
[ l : ureg ] let @xmath1006 and let @xmath1007 be a compact subset of @xmath291 . then for any @xmath1008 @xmath1009 is defined and @xmath1010 for some @xmath1011 depending only on @xmath256 , @xmath1007 and @xmath1012 .
we first show for any @xmath65 the function @xmath1013 on @xmath989 is compactly supported with the support contained in a compact set depending on @xmath1014 and @xmath1007 ( for @xmath1008 ) .
indeed , suppose that @xmath1015 and let @xmath1016 be the iwasawa decomposition , with @xmath1017 .
then @xmath1018 since @xmath1019 . from the support condition of whittaker functions ( on @xmath288 )
, we get that @xmath1020 , @xmath1021 . on the other hand , writing @xmath1022 , the entries of @xmath574 are bounded by @xmath1023 . thus @xmath1024 is well defined .
moreover , we can write @xmath1025 where @xmath1026 depends only on @xmath1007 , @xmath1012 and @xmath1014 . we claim that @xmath228 is compactly supported on @xmath1027 .
let @xmath1028 be a column vector in @xmath1029 and let @xmath1030 .
for @xmath1031 we have @xmath1032\in \bar u_{1,1}\bar u^{\dag}$ ] and @xmath1033)=\psi_{v'}(\bar u):=\psi(\bar u_{4{\mathbf{n}},1}v'_1+\dots+\bar u_{4{\mathbf{n}},2{\mathbf{n}}-1}v'_{2{\mathbf{n}}-1}).\ ] ] note that @xmath1034 stabilizes the pair @xmath1035 under conjugation .
thus @xmath1036 is left @xmath1037-equivariant .
clearly @xmath1036 is also left @xmath1038-equivariant .
we get @xmath1039 let @xmath271 be the image under @xmath928 of the subspace of @xmath550 consisting of matrices @xmath865 such that @xmath1040 unless either @xmath1041 or @xmath1042 . thus @xmath1043 and by lemma [ l : elemc ] we get that @xmath1044 is compactly supported and the support is bounded in terms of @xmath256 and @xmath1007 .
the lemma follows .
[ cor : suppw ] the restriction of any @xmath1045 to @xmath989 is compactly supported .
moreover , for any @xmath674 there exists @xmath1046 such that the support of @xmath1047 is contained in @xmath1048 for any @xmath1049 .
[ [ section-2 ] ] following lemma [ l : ureg ] , for any @xmath1050 , @xmath1051 and @xmath663 define @xmath1052 we observe some properties of @xmath1053 .
let @xmath1054 and @xmath1055 .
[ l : fwproperty ] for any @xmath1056 and any @xmath1057 , @xmath1058 we have : 1 .
@xmath1059 for any @xmath1060 .
2 . @xmath1061 for any @xmath1062 .
3 . @xmath1063 if @xmath1064 , @xmath1031 and @xmath1065 .
note that we also have @xmath1066 for any @xmath1067 but we wo nt use this fact .
let @xmath1060 . from , and
the equivariance of @xmath131 , we get @xmath1068 hence , the first claim . for the second claim ,
we get from : @xmath1069 the last claim follows from . indeed , for any @xmath1057 @xmath1070 motivated by the above properties of @xmath1071 , we define some classes of function on @xmath1072 .
we first introduce some notations and auxiliary unipotent subgroups of @xmath288 .
let @xmath1073 be the closed subvariety @xmath1074 of @xmath288 .
thus , any element of @xmath1075 can be written uniquely as @xmath1076 with @xmath569 and @xmath1077 . for @xmath1058 , define a function @xmath1078 on @xmath1075 by @xmath1079 for @xmath1080 , let @xmath1081 be the unipotent subgroup of @xmath1082 generated by the commuting one - parameter root subgroups @xmath1083 , that is , @xmath1084,2{\mathbf{n}}-[\frac{k+1}{2}])}.\ ] ] let @xmath1085 so that @xmath1086 note that @xmath1087 .
it is clear @xmath1088 when @xmath885 is odd .
define @xmath1089 we illustrate the groups in the case @xmath964 . in the template
below : @xmath1090 the numbers @xmath1091 above and below the principal diagonal mark the non - constant coordinates of @xmath1081 and @xmath1092 respectively .
( the groups @xmath1092 with @xmath1093 will not be used . )
meanwhile for @xmath1094 , each group @xmath1095 corresponds to the numbers @xmath1096 below the principal diagonal and the numbers @xmath1097 above it .
let @xmath1098 .
we observe that @xmath1099 and @xmath1100 where @xmath1101 .
for @xmath939 let @xmath1102 be the one - dimensional torus @xmath1103 note that as algebraic groups ( but not at the level of @xmath0-points ) we have @xmath1104 .
for any @xmath1105 ( resp .
@xmath1106 ) consider the classes @xmath1107 ( resp .
@xmath1108 ) of functions @xmath228 on @xmath1072 satisfying the following properties for any @xmath1058 , @xmath1057 : @xmath1109 for instance , by lemma [ l : fwproperty ] , if @xmath1110 then @xmath1111 .
define @xmath1113 the group @xmath1114 corresponds to the lower unitriangular part in the template .
define @xmath1115 , i.e. @xmath1116 it will be also useful to set @xmath1117 note that @xmath1118 and in particular @xmath1119 .
also @xmath1120 .
denote the simple roots of @xmath288 by @xmath1121 , so that the corresponding one - parameter root groups are @xmath1122 and @xmath564 .
write @xmath1123 where @xmath1124 for a two - variable function such as @xmath1125 there are two possible meanings for the operation @xmath1126 , namely either in the @xmath1127 variable or in the @xmath838 variable .
we use @xmath1126 to denote the integration in the @xmath1127 variable . for any subtorus @xmath1128 of @xmath1129 and @xmath1130
we will write @xmath1131\circ\phi:=\int_{t'}f'(t')\phi(\cdot t';\cdot)\ dt'\ ] ] ( if defined ) .
the following lemma will be proved in the next section .
[ l : mreg2 ] fix a compact subset @xmath1007 of @xmath291 and @xmath1106 . 1 .
let @xmath1132 and @xmath1133 . then 1 .
@xmath1134\circ\phi$ ] is a convergent integral .
moreover @xmath1134\circ\phi\in{\mathcal{c}}_k(t_{{\mathbb m}}\times { { \mathbb m}})^{k_0}$ ] .
@xmath1135\circ\phi$ ] is defined in the sense of lemma [ lem : genstab ] , with @xmath1136 , @xmath1137 , @xmath1138 , @xmath1139 and @xmath1140 .
moreover @xmath1135\circ\phi$ ] belongs to @xmath1141 , and there exists @xmath1142 depending on @xmath1007 , @xmath256 and @xmath1143 such that @xmath1144\circ \phi(t , e)= { { \mathcal l}}_{\omega , \psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_k}}[f_k]\circ \phi(t , e)\ ] ] for all @xmath1008 .
2 . let @xmath1145 .
then for all @xmath1008 , as a function of @xmath1146 , @xmath1125 is supported on some @xmath1147 determined by @xmath1007 and @xmath256 .
moreover @xmath1148 belongs to @xmath1149 . for @xmath1150 ,
set @xmath1151 let @xmath1152 be the torus @xmath1153 .
we can then define the following integral transforms : [ def : tranwav ] identify @xmath1154 with @xmath1155 by @xmath1156 .
for @xmath1157 we define @xmath1158 on @xmath1159 by @xmath1160\circ\cdots \circ{{\mathcal l}}_{{\mathcal{u}}^1}\circ{{\mathcal l}}^{reg}_{{\mathcal{v}}^1,\psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_1}}[f_1]\circ\phi\ ] ] ( and extend @xmath1158 to @xmath1155 by multi - linearity ) . also define for @xmath1050 @xmath1161 where @xmath1162 .
[ cor : abcde ] notations as in lemma [ l : mreg2 ] . 1 .
there exists @xmath1163 depending only on @xmath256 , @xmath1007 and @xmath1164 , such that for all @xmath1165 we have @xmath1166\circ\phi(t;e).\ ] ] 2 .
there exists @xmath1167 depending only on @xmath256 , @xmath1007 and @xmath1164 such that @xmath1168 for all @xmath1008 .
we have @xmath1169 if the right - hand side is absolutely convergent .
4 . if @xmath1170 then @xmath1171 .
all parts except the last one are immediate from lemmas [ l : ureg ] and [ l : mreg2 ] .
it follows from lemma [ l : mreg2 ] and that @xmath1172 , while from lemma [ l : fwproperty ] , @xmath1173 .
recall that by lemma [ l : stablemodel ] , for any @xmath686 , we have @xmath1174 .
we will prove the following functional equation in [ sec : prf of prop : main ] : [ prop : main ] let @xmath1175 .
then for any @xmath686 we have @xmath1176 for all @xmath1177 and @xmath1178 . consequently , by theorem [ thm : mwl ] we have @xmath1179 on @xmath72 . [ rem : compactonly ] note that by corollary [ cor : uniqueetwon ] , @xmath977 and @xmath115 ( for @xmath1170 ) are proportional on @xmath72 with the constant of proportionality depending a priori on @xmath838 , @xmath1164 and @xmath7 . also note that for fixed @xmath838 and @xmath1164 , @xmath1180 is a continuous functional on @xmath681 , being equal to a ( finite ) linear combination ( depending on @xmath256 ) of the values of the function @xmath1181 .
on the other hand @xmath109 is also a continuous functional on @xmath681 .
thus we only need to prove the identity for @xmath816 by lemma [ l : stablemodel ] part .
[ [ sec : deflk ] ] we first introduce a few more families of unipotent groups . for @xmath1182 ,
define @xmath1183 so that @xmath1184 . also set @xmath1185 for instance , in the case @xmath964 and @xmath1094 in the diagram above @xmath1186 corresponds to the numbers @xmath1096 below the principal diagonal and the numbers @xmath1187 above it ; @xmath1188 ( for @xmath1189 ) corresponds to the numbers @xmath1190 below the principal diagonal and the numbers @xmath1097 above it .
observe that @xmath1191 , @xmath1192 and @xmath1193 note that @xmath1188 is a maximal unipotent subgroup of @xmath288 .
in fact for @xmath1194 , @xmath1195 where @xmath1196 are the permutation matrices given by @xmath1197,\\ e_{6{\mathbf{n}}-k-2i+1 } & i > 2{\mathbf{n}}-[\frac{k+1}2],\end{cases}\\\ ] ] where @xmath1198 is the standard basis and @xmath1199 .
[ r : lshape ] for @xmath1200 , consider the isomorphism @xmath1201 from @xmath1188 to @xmath290 .
then we have @xmath1202 and @xmath1203 so that , @xmath1204 } n_{\alpha_{2i+2{\mathbf{n}}-k}},\,\ , \sigma_k(\overline{{\mathcal x}}^k)=\prod_{i=0}^{[\frac{k}{2}]-1}n_{\alpha_{2i+1 + 2{\mathbf{n}}-k}}.\ ] ] if we denote by @xmath1205 , @xmath1206 the entries of @xmath1207 in the simple roots then @xmath1208 for @xmath772 , @xmath1209 is a character on @xmath1186 . for @xmath772 , @xmath1210 is the subgroup of @xmath290 consisting of the matrices whose off - diagonal entries are zero at rows @xmath1211 and @xmath1212 , @xmath1213}{2}$ ] . in particular , for the negative simple root groups @xmath1214 , @xmath1215 $ ] we have @xmath1216 , which are _ almost simple _ root group . hence , @xmath1217 is a character of @xmath1218 that is supported on the almost simple root groups .
similarly , @xmath1219 is the subgroup of @xmath290 consisting of the matrices whose off - diagonal entries are zero at rows @xmath1220 , @xmath1221}{2}$ ] . the simple root groups @xmath562 are stable under @xmath1222 when @xmath1223 and @xmath1224 ( almost simple root group ) when @xmath1225 $ ] .
thus , @xmath1226 is a character supported on the simple root groups @xmath562 ( @xmath1223 ) and the almost simple root groups @xmath1227 ( @xmath1228 ) .
it is now clear that @xmath1229 is a character on @xmath1230 which is supported on the simple root groups @xmath562 ( @xmath1223 ) and the almost simple root groups @xmath1231 ( @xmath1228 ) .
[ lem : lproperty ] 1 .
[ claim : normall ] for @xmath1232 , @xmath1233 and @xmath1234 .
also @xmath1235 .
[ claim : equivl ] for any @xmath1236 and @xmath1077 we have @xmath1237 .
[ claim : charl ] for @xmath1182 , @xmath1238 is a character of @xmath1095 .
[ claim : chitomd ] for @xmath1239 we have @xmath1240 for any @xmath1051 and @xmath1241 such that @xmath1242 for all @xmath1243 .
in particular , when @xmath1177 , @xmath1244 .
by it suffices to prove parts [ claim : normall ] and [ claim : charl ] for @xmath1245 .
part [ claim : normall ] follow from remark [ r : lshape ] . for part [ claim : equivl ] , write @xmath1246 with @xmath569 and @xmath1247 , then @xmath1248 .
let @xmath1249 .
write @xmath1250 with @xmath1251 and @xmath1252 for @xmath1253 . then by part [ claim : equivl ] @xmath1254y_2z_1z_2)\\={{\psi}_{x_{{\mathbb m}},t}}(y_1[z_1,y_2]y_2){{\psi}_{x_{{\mathbb m}},t}}(z_1){{\psi}_{x_{{\mathbb m}},t}}(z_2),\end{gathered}\ ] ] where @xmath1255=z_1y_2z_1^{-1}y_2^{-1}\in\check{{\mathcal x}}^k$ ] by part [ claim : normall ] .
recall that @xmath1078 is a character on @xmath1186 . by remark [ r : lshape ] ,
@xmath1256)$ ] is not supported on any simple root or almost simple root of @xmath290 .
thus @xmath1257)=1 $ ] .
we get from part [ claim : equivl ] : @xmath1258y_2){{\psi}_{x_{{\mathbb m}},t}}(z_1){{\psi}_{x_{{\mathbb m}},t}}(z_2)={{\psi}_{x_{{\mathbb m}},t}}(y_1){{\psi}_{x_{{\mathbb m}},t}}(y_2){{\psi}_{x_{{\mathbb m}},t}}(z_1){{\psi}_{x_{{\mathbb m}},t}}(z_2)\\= { { \psi}_{x_{{\mathbb m}},t}}(x_1){{\psi}_{x_{{\mathbb m}},t}}(x_2).\end{gathered}\ ] ] this shows that @xmath1078 is a character on @xmath1259 .
it is also clear that @xmath1078 is a character on @xmath1260 .
part [ claim : chitomd ] is clear from the definition of @xmath1078
. let @xmath1261 and define @xmath1262 , @xmath1263 and @xmath1264 similarly .
note that for @xmath1245 @xmath1265 let @xmath1266 .
if @xmath1267 and @xmath1268 , then by lemma [ lem : lproperty ] @xmath1269\in [ \overline{{\mathcal m}}^{2k+1},\hat{{\mathcal m}}^{2k+1}]\cap[\hat{{\mathcal m}}^{2k+2},{{\mathcal m}}^{4{\mathbf{n}}-2k-2}]=\hat{{\mathcal m}}^{2k+1}\cap\hat{{\mathcal m}}^{2k+2}=\check{{\mathcal m}}^{2k+1}.\ ] ] a simple calculation shows that @xmath1270)=\psi({\left\langle\bar n , n\right\rangle}_{2k+1,t})\ ] ] where @xmath1271 are non - degenerate pairings @xmath1272 here @xmath1273 and @xmath1274 are immaterial constants ( depending continuously on @xmath838 ) . note that with respect to the standard bases of @xmath1275 and @xmath1276 , the determinant of @xmath1271 is @xmath1277 .
similarly for @xmath1278 and @xmath1279 , by lemma [ lem : lproperty ] @xmath1269\in [ \overline{{\mathcal m}}^{2k},\hat{{\mathcal m}}^{2k}]\cap[\hat{{\mathcal m}}^{2k+1},{{\mathcal m}}^{4{\mathbf{n}}-2k-1}]\subset\hat{{\mathcal m}}^{2k}\cap\hat{{\mathcal m}}^{2k+1}=\check{{\mathcal m}}^{2k},\ ] ] and @xmath1280)=\psi({\left\langle\bar n ,
n\right\rangle}_{2k , t})\ ] ] where @xmath1281 are non - degenerate pairings @xmath1282 where again @xmath1283 and @xmath1284 are some immaterial constants ( depending continuously on @xmath838 ) . first note that the condition is clearly preserved under the operations @xmath1285 , @xmath1286\circ\phi$ ] , @xmath1287 , @xmath1106 , if defined . 1
. suppose that @xmath1132 with @xmath1106 .
it follows from property that @xmath1288 \circ \phi(t , m)=\int_{{t}_k } \phi(t , t'm)\chi_k(t')f_k(t')\ dt'\ ] ] for a certain ( unimportant ) character @xmath1289 of @xmath1290 .
thus by , the integrand on the right hand side is in @xmath1291 .
in particular , @xmath1134\circ \phi(t , m)$ ] is a convergent integral .
+ property for @xmath1134 \circ \phi$ ] follows immediately . on the other hand , property for @xmath1134 \circ \phi$ ]
follows immediately from lemma [ lem : lproperty ] part .
thus @xmath1134 \circ \phi\in{\mathcal{c}}_k(t_{{\mathbb m}}\times { { \mathbb m}})^{k_0}$ ] .
2 . let @xmath1292 \circ \phi(t , m)\in{\mathcal{c}}_k(t_{{\mathbb m}}\times { { \mathbb m}})^{k_0}$ ] . by and for any @xmath1293 and @xmath1294 we have @xmath1295dc)=\psi({\left\langlec , d\right\rangle}_{2k , t}){{\psi}_{x_{{\mathbb m}},t}}(d)\phi_1(t , c).\ ] ] applying lemma [ l : elemc ] we conclude that @xmath1296 is compactly supported on @xmath1297 uniformly in @xmath1008 .
to show that the conditions of lemma [ lem : genstab ] are satisfied for the definition of @xmath1135\circ\phi$ ] , we only need to observe that by , @xmath1134\circ\phi(t;\cdot)$ ] is left invariant under some @xmath1298 .
lemma [ lem : genstab ] also gives the existence of @xmath1142 with the required properties .
+ we show that @xmath1299 .
recall that @xmath1300 .
we check property for @xmath1301 separately for @xmath1302 and @xmath1303 . in the first case ,
follows from the fact that @xmath1304 on @xmath1305 . in the second case note that for any @xmath1306 we have @xmath1307\in \check{{\mathcal m}}^{2k}$ ] and @xmath1308)=1 $ ] ( as @xmath1309 ) by lemma [ lem : lproperty ] .
thus @xmath1310 and @xmath1311 finally , the property follows from the fact that @xmath1312 stabilizes the character @xmath1313 on @xmath1305 .
2 . suppose that @xmath1145 with @xmath1106 .
+ recall that for @xmath1314 and @xmath1315 , we have @xmath1316\in \hat{{\mathcal m}}^{2k+1}$ ] and by , @xmath1317)=\psi({\left\langlec , d\right\rangle}_{2k+1,t})$ ] .
thus for such @xmath162 and @xmath163 we have @xmath1318 .
it follows from lemma [ l : elemc ] that @xmath1319 is compactly supported uniformly for @xmath1008 .
+ we check that @xmath1320 . by lemma [ lem : lproperty ] , @xmath1321 .
since @xmath1322 is trivial on @xmath1276 , we are left to show that @xmath1323 for @xmath1324 . for @xmath1314 and @xmath1324 ,
as both @xmath162 and @xmath94 lie in @xmath1325 , we have @xmath1307\in \check{{\mathcal m}}^{2k+1}$ ] and @xmath1308)=1 $ ] by lemma [ lem : lproperty ] .
thus @xmath1326 and @xmath1327 condition is clear because @xmath291 normalizes @xmath1276 .
this concludes the proof of lemma [ l : mreg2 ] .
before proving proposition [ prop : main ] , we first turn our attention to model transition from @xmath72 to @xmath956 . we will state the analog of proposition [ prop : main ] in this setting ( proposition [ prop : 14 ] ) . the proofs of the two propositions are repeated application of fourier inversion . as the proof of proposition [ prop : 14 ] is simpler , we present it first . for @xmath1045 define @xmath1328 recall that by corollary [ cor : suppw ] the integrand is compactly supported , hence the integral converges .
moreover , by @xmath110 is non - vanishing on @xmath72 ( and in fact on any non - trivial subrepresentation of @xmath1329 for any @xmath1330 ) ( * ? ? ?
* theorem in 1.3 ) cf .
* remark 4.8 ) .
[ prop : 14 ] for any @xmath686 we have @xmath1331 , namely @xmath1332 in particular by theorem [ thm : mwl ] , for @xmath802 we have @xmath1333 and hence , @xmath108 , @xmath109 and @xmath923 are non - zero on @xmath69 . by the remark above and lemma [ l : stablemodel ]
the left - hand side of is a continuous functional on @xmath681 and it is enough to prove the proposition for @xmath816 . define for @xmath1182 @xmath1334,[(4{\mathbf{n}}-k+1)/2]\}}\ ] ] and @xmath1335 thus , @xmath1336 .
for example when @xmath964 , the space @xmath926 consists of the image under @xmath928 of the matrices of the form @xmath1337 where the digit @xmath467 corresponds to @xmath1338 .
for this section , only the groups @xmath1339 are relevant .
the other groups will be relevant for the next section .
for @xmath1182 let @xmath1340 these are unipotent subgroups of @xmath1 .
in fact , they are subgroups of @xmath1341 which is the unipotent radical of a semistandard borel subgroup .
in particular , @xmath1342 .
let @xmath1343 be the closed subvariety @xmath1344 of @xmath1 .
thus @xmath1345 and @xmath1346 for all @xmath885 .
define a function @xmath1347 on @xmath1348 by @xmath1349 note that @xmath1350 .
( for this section it is sufficient to consider @xmath1351 instead of @xmath1348 . ) we write the left - hand side of as @xmath1352 we apply lemmas [ l : fourier ] and [ c : fourier2 ] according to the parity of @xmath885 with @xmath1353 , @xmath1354 and @xmath1355 .
we claim that the data @xmath1356 , @xmath1357 , @xmath1358 , @xmath1359 , @xmath1360 and @xmath1361 satisfy the conditions of lemma [ l : fourier ] ( resp . , lemma [ c : fourier2 ] ) if @xmath885 is odd ( resp . , even ) . for
that we need the following properties . 1 .
[ claim : normalg ] for @xmath1232 , @xmath1362 , @xmath1363 ; both are subgroups of @xmath1364 .
[ claim : charg ] for @xmath1182 , @xmath1365 is a character of @xmath1366 .
[ claim : equivg ] for @xmath1367 , @xmath1368 , @xmath1369 .
[ claim : hinterg ] for @xmath1370 , @xmath1371 while @xmath1372 .
@xmath1347 is trivial on @xmath1373 .
[ claim : prop5 ] for @xmath790 we have @xmath1374\subset \check{{\mathcal n}}^{2k-1}$ ] . moreover , if @xmath1375 and @xmath1376 then @xmath1377)^{-1}=\psi({\left\langle\bar u , v\right\rangle}'_{2k-1})$ ] where @xmath1378 identifying @xmath1379 and @xmath1380 with @xmath1381 where @xmath1382 using the standard coordinates , the determinant of the pairing @xmath1383 is @xmath1384 .
the proof of the first three properties is similar to that of lemma [ lem : lproperty ] .
the fourth property is clear from the definitions of @xmath1385 , @xmath1386 and @xmath1347 .
an explicit calculation gives the last property .
it follows that when @xmath885 is even , @xmath1387 and @xmath1388 , and the conditions of lemma [ c : fourier2 ] are satisfied
. when @xmath885 is odd , the conditions of lemma [ l : fourier ] are satisfied , with @xmath1389 given by @xmath1390 ( see remark [ rem : condfourier ] ) . in both cases
we get the relations @xmath1391 and by induction ( since @xmath1392 and @xmath1393 ) @xmath1394 for @xmath1395 we obtain the statement of the proposition . by
we also conclude : for any @xmath973 we have @xmath1396
for @xmath686 let @xmath1397 be the function on @xmath1398 given by @xmath1399 recall from lemma [ l : ureg ] that the above regularized integral over @xmath551 is defined . define the character @xmath1400 on @xmath290 by @xmath1401 and let @xmath1402 be the function on @xmath1403 given by @xmath1404 recall the subgroup @xmath997 defined in and @xmath1054 .
set @xmath1405 and @xmath1406 .
note that the restriction of @xmath1402 to @xmath1407 is a character which is @xmath1408-generic ( or alternatively , @xmath1409-generic ) and trivial on @xmath1410 .
we can therefore define ( using remark [ rem : partint ] and lemma [ lem : genstab ] ) @xmath1411 for @xmath686 . by lemma [ lem : genstab ] ,
the functionals @xmath1412 on @xmath681 are uniformly continuous for @xmath838 in a compact set .
[ l : nuresult ] for any @xmath686 we have for @xmath1058 : @xmath1413 once again , it is enough to prove it for @xmath816 .
it is convenient to introduce the notation @xmath1414}\circ f(\cdot):=f(x\cdot)$ ] . from the definition of @xmath1415 we get @xmath1416}\circ{{\mathcal l}}^h_{n , \psi_n^{-1}}\circ l)({\varrho}(m)).\ ] ] for @xmath772 ,
let @xmath1417 be the one - parameter subgroup @xmath1418 of @xmath551 .
the integration over @xmath1419 can be written as an iterated integral @xmath1420 .
recall the one - parameter root subgroups @xmath1421 , @xmath1422 , @xmath1423 .
for convenience we also set @xmath1424 . for @xmath1200 ,
let @xmath1425 in particular , @xmath1426 we will prove the following statements ( for @xmath1427 ) : @xmath1428}\circ{{\mathcal l}}^h_{n,\psi_n^{-1}}\circ l={\left|{t_1}\right|}^{-2{\mathbf{n}}}{\left|{\det t}\right|}^{2{\mathbf{n}}}{{\mathcal e}[{\varrho}(t)]}\circ{{\mathcal l}}^h_{\hat{{\mathcal n}}^{4{\mathbf{n}}-1 } , { { \psi}_{x_0}}^{-1}}\circ l,\\ \label{eq : baru_11 } { { \mathcal e}[{\varrho}(t)]}\circ{{\mathcal l}}^h_{\hat{{\mathcal n}}^{4{\mathbf{n}}-1 } , { { \psi}_{x_0}}^{-1}}\circ l= { \left|{t_1}\right|}^{{\mathbf{n}}}{\left|{\det t}\right|}^{-2{\mathbf{n}}}{\delta}_b^{\frac12}({\varrho}(t)){{\mathcal l}}^h_{\hat{{\mathcal n}}^{4{\mathbf{n}}-1 } , { { \psi}_{x_0,t}}^{-1}}\circ l,\\ \label{eq : baru_k1 } \begin{split}&{{\mathcal l}}_{\bar u_{k,1},{{\psi}_{x_0,t}}^{-1}}\circ\dots\circ{{\mathcal l}}_{\bar u_{2,1},{{\psi}_{x_0,t}}^{-1}}\circ { { \mathcal l}}^{reg,\bar u^{\dag}}_{\bar u^{\dag}\bar u_{1,1},{\psi_{\bar u}}}\circ { { \mathcal l}}^h_{\hat{{\mathcal n}}^{4{\mathbf{n}}-1},{{\psi}_{x_0,t}}^{-1}}\circ l=\\ & { { \mathcal l}}_{\bar u_{k,1},{{\psi}_{x_0,t}}^{-1}}\circ\dots\circ { { \mathcal l}}_{\bar u_{1,1},{{\psi}_{x_0,t}}^{-1}}\circ { { \mathcal l}}^h_{\hat{{\mathcal n}}^{4{\mathbf{n}}-1},{{\psi}_{x_0,t}}^{-1}}\circ l= { { \mathcal l}}^h_{\hat{{{\mathcal n_0}}}^k,{{\psi}_{x_0,t}}^{-1}}\circ l,\ \ k=1,\dots,2{\mathbf{n}}. \end{split}\end{gathered}\ ] ] the lemma would follow by combining the relations , , and for @xmath1429 . to prove note that @xmath1430 normalizes @xmath989 , and the modulus function of @xmath1430 acting on @xmath989 is @xmath1431 . thus , @xmath1432}={\left|{t_1}\right|}^{-2{\mathbf{n}}}{\left|{\det t}\right|}^{2{\mathbf{n}}}{{\mathcal e}[{\varrho}(t)]}\circ{{\mathcal l}}_{\bar{u}^{\dag}}.\ ] ] recall the groups @xmath1386 defined in the proof of proposition [ prop : 14 ]
. then @xmath1433 .
thus @xmath1434 . as in the proof of proposition
[ prop : 14 ] we apply , this time for @xmath1435 , to get : @xmath1436 the relation follows . to prove we use the change of variable @xmath1437 and the fact that @xmath1438 .
an explicit computation shows that the jacobian is @xmath1439 .
we obtain @xmath1440}\circ{{\mathcal l}}^h_{\hat{{\mathcal n}}^{4{\mathbf{n}}-1 } , { { \psi}_{x_0}}^{-1}}\circ l(g)= \int_{\hat{{\mathcal n}}^{4{\mathbf{n}}-1}\cap h{\backslash}\hat{{\mathcal n}}^{4{\mathbf{n}}-1}}l(n{\varrho}(t)g){{\psi}_{x_0}}^{-1}(n)\,dn \\={\left|{t_1}\right|}^{{\mathbf{n}}}{\left|{\det t}\right|}^{-2{\mathbf{n}}}{\delta}_b^{\frac12}({\varrho}(t))\int_{\hat{{\mathcal n}}^{4{\mathbf{n}}-1}\cap h{\backslash}\hat{{\mathcal n}}^{4{\mathbf{n}}-1 } } l(ng){{\psi}_{x_0}}^{-1}({\varrho}(t ) n{\varrho}(t)^{-1})\,dn.\end{gathered}\ ] ] note that @xmath1441 . to prove we will apply lemma [ l : fourier ] or lemma [ c : fourier2 ] , according to the parity of @xmath885 with @xmath1353 , @xmath1354 and @xmath1355 .
we claim the data @xmath1442 , @xmath1443 , @xmath1444 , @xmath1445 , @xmath1446 and @xmath1447 satisfy the conditions of lemma [ l : fourier ] ( resp .
[ c : fourier2 ] ) if @xmath885 is odd ( resp .
this follows from the following properties . 1 .
[ claim : normalnu ] for @xmath1370 , @xmath1448 and @xmath1449 , both being subgroups of @xmath1450 .
[ claim : equivnu ] for @xmath1451 , @xmath1031 , @xmath1452 .
[ claim : charnu ] for @xmath1200 , @xmath1453 is a character of @xmath1454 which is trivial on @xmath1455 .
[ claim : hnu ] when @xmath885 is odd , @xmath1456 .
when @xmath885 is even , @xmath1457 and @xmath1458 .
5 . for @xmath1105
we have @xmath1459=\bar u_{1,1}\subset \check{{\mathcal n_0}}^{2k-1}$ ] .
moreover , if @xmath1460 and @xmath1461 then @xmath1462)^{-1}=\psi(\alpha\beta)$ ] .
indeed , the proof of the first three properties is similar to that of lemma [ lem : lproperty ] .
we omit the details .
the fourth property is clear , while the last one follows from a simple computation .
applying lemma [ l : fourier ] ( resp .
[ c : fourier2 ] ) if @xmath885 is odd ( resp .
even ) we obtain the relation @xmath1463 for any @xmath1464 .
in particular , for @xmath1465 , @xmath1466 is absolutely convergent where @xmath1467 . thus by lemma [ l : stableprop ] and the fact @xmath1468 , we get the first equation in .
the second relation in follows from the above identity by induction on @xmath885 .
we now prove proposition [ prop : main ] .
recall that for @xmath1058 : @xmath1469\circ\cdots \circ{{\mathcal l}}_{{\mathcal{u}}^1}\circ{{\mathcal l}}^{reg}_{{\mathcal{v}}^1 , \psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_1}}[f_1]\circ{\tilde{\upsilon}}(l;t , e).\end{gathered}\ ] ] we continue to assume that @xmath816 .
observe @xmath1470 for @xmath569 and @xmath1471 for @xmath1031 .
thus by lemma [ l : nuresult ] , @xmath1472 let @xmath1473 . define @xmath1474 so that @xmath1475 [ lem : fixcompact ] the restriction of @xmath1476 to @xmath997 belongs to @xmath1477 where @xmath1478 .
clearly , @xmath1476 is @xmath1479-equivariant .
since @xmath1480 is compact , the restriction of @xmath107 to @xmath1481 is compactly supported modulo @xmath1482 .
it follows that @xmath1476 is compactly supported modulo @xmath586 .
it remains to observe that @xmath1483 is closed in @xmath1484 , or equivalently , that the orbit of @xmath43 under conjugation by @xmath997 is closed .
however , it is easy to see that this orbit consists of the matrices of the form @xmath1485 where @xmath1486 is an involution with characteristic polynomial @xmath1487 and @xmath1488 are row vectors of size @xmath1489 such that @xmath1490 .
this is clearly a closed subset of @xmath288 .
in particular , @xmath1483 is an affine variety . for @xmath1133 ,
we let @xmath1491 . from and
, @xmath1492 is equal to @xmath1493\circ\cdots \circ{{\mathcal l}}_{{\mathcal{u}}^1}\circ{{\mathcal l}}^{reg}_{{\mathcal{v}}^1 , \psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_1}}[f'_1]\circ { { \mathcal l}}^{h_{{\mathbb m}}^{{\flat}}}_{n_{{{\mathbb m}}}^{{\flat}},{{\psi}_{x_{{\mathbb m}},t}}^{-1}}\circ\lambda(l;e).\ ] ] ( note that @xmath1494 is a function of two variables @xmath1495 through the dependence of @xmath1078 on @xmath838 . )
we will show that for any @xmath816 and @xmath1170 we have @xmath1496\circ\cdots \circ{{\mathcal l}}_{{\mathcal{u}}^1}\circ{{\mathcal l}}_{{\mathcal{v}}^1 , \psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_1}}[f'_1]\circ { { \mathcal l}}^{h_{{\mathbb m}}^{{\flat}}}_{n_{{{\mathbb m}}}^{{\flat}},\psi_{n_{{\mathbb m}},t}^{-1}}\circ\lambda(l;e)\\=c({\mathfrak{f } } ) { { \mathcal l}}^{h_{{\mathbb m}}^{{\flat}}}_{{\mathcal{z } } , { \psi_{{\mathcal{z}}}}^{-1}}\circ \lambda(l;e).\end{gathered}\ ] ] taking into account the definition of @xmath1476 and we get @xmath1497 which is the statement of proposition [ prop : main ] .
it remains to prove .
recall the definition of @xmath1498 in [ sec : deflk ] .
by lemma [ lem : lproperty ] , @xmath1499 is a character of @xmath1498 .
also recall that @xmath1500 and @xmath1244 for @xmath1170 .
thus from lemma [ lem : fixcompact ] , is the case @xmath1501 of the last part of the lemma below .
[ l : iter4 ] assume @xmath1502 .
then for any @xmath1106 and @xmath1058 : 1 .
@xmath1503 is invariant under @xmath1504 where @xmath1505 .
2 . @xmath1506 .
3 . @xmath1507 .
thus , @xmath1508\circ\cdots \circ{{\mathcal l}}_{{\mathcal{u}}^1}\circ{{\mathcal l}}_{{\mathcal{v}}^1 , \psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_1}}[f'_1]\circ { { \mathcal l}}^{h_{{\mathbb m}}^{{\flat}}}_{n_{{{\mathbb m}}}^{{\flat}},{{\psi}_{x_{{\mathbb m}},t}}^{-1}}\circ f\\=\big(\prod_{i=1}^k\int_{{t}_i}f'_i(t')\ dt'\big ) { { \mathcal l}}^{h_{{\mathbb m}}^{{\flat}}}_{\hat{{{\mathcal m}}}^{2k+2},{{\psi}_{x_{{\mathbb m}},t}}^{-1}}\circ f.\end{gathered}\ ] ] the first part follows from lemma [ lem : lproperty ] part [ claim : chitomd ] . for the second and third parts we will apply lemmas [ c : fourier2 ] and [ l : fourier ] respectively with @xmath1509 , @xmath1510 and @xmath1511 as above .
observe that for @xmath1106 @xmath1512 we claim that the data @xmath1513 , @xmath1514 , @xmath1515 , @xmath1516 , @xmath1517 and @xmath1518 satisfy the conditions in lemma [ c : fourier2 ] .
this follows from lemma [ lem : lproperty ] .
thus , the second part follows from lemma [ c : fourier2 ] .
similarly , for the third part , we apply lemma [ l : fourier ] to the data @xmath1519 , @xmath1520 , @xmath1521 , @xmath1522 , @xmath1523 and @xmath1518 . in this case by , @xmath1524 gives an isomorphism between @xmath271 and @xmath1525 preserving haar measure .
finally , the last part follows immediately by induction on @xmath885 .
note that @xmath1526 .
this concludes the proof of proposition [ prop : main ] .
we finish the paper with some auxiliary results ( lemma [ lem : spcltrans ] and corollaries [ cor : mreg ] and [ cor : fornextpaper ] below ) which will be used in a sequel of this paper .
they are variants of the results of the previous sections and are proved in a similar way .
the results do not seem to fit a general pattern and are hard to motivate by themselves .
the only reason to include them here is that they are simple modifications of earlier results .
[ [ section-3 ] ] recall that @xmath499 is the second mirabolic subgroup of @xmath288 consisting of matrices whose first column is @xmath1527 .
let @xmath1528 .
consider the subspace @xmath1529 of @xmath1530 consisting of the @xmath131 s such that is not closed . ]
for @xmath1532 define @xmath1533 note that @xmath1534 since @xmath1535 stabilizes the character @xmath1536 .
recall that by lemma [ l : cuspidal ] ( or rather , its analogue for @xmath499 ) if @xmath1537 then @xmath1538 .
consider @xmath1539 and @xmath1540 then @xmath1541 by ( the analogue of ) lemma [ l : feglnn ] . in this subsection
we prove the following functional identity .
[ prop : factor ] we have @xmath1542 or equivalently @xmath1543 we first write @xmath1544 as an iterated integral .
let @xmath1545 be the image of the co - character @xmath1546 so that @xmath1547 .
[ lem : manipim ] for any @xmath1548 , @xmath1549 where @xmath1550 , @xmath1105 .
using we write the ( absolutely convergent ) integration @xmath1551 over @xmath1114 as an iterated integral over @xmath1552 , and write the integration over @xmath1535 as an iterated integral over @xmath1545 ( @xmath1105 ) .
thus , @xmath1551 equals @xmath1553 where for @xmath1554 @xmath1555 ( we recall that @xmath1313 is trivial on @xmath1556 , @xmath1557 . )
it is easy to see that for @xmath1558 , @xmath1545 stabilizes @xmath1559 under conjugation .
of course @xmath1545 normalizes @xmath1556 for all @xmath467 . for @xmath1560 let @xmath1561 and @xmath1562 . by a change of variable @xmath1563 we have @xmath1564 and ( for @xmath1565 ) @xmath1566 for any function @xmath145 on @xmath1 and @xmath1567 on @xmath1545 , provided that the double integral converges absolutely .
the lemma follows by making the above changes of variables for all pairs @xmath1568 with @xmath1569 ( first type ) and @xmath1565 ( second type ) and noting that @xmath1570 on @xmath1545 .
indeed , @xmath1571 is the modulus function of @xmath1572 acting on the group @xmath1573 , which is easy to compute .
recall from definition ( with @xmath1574 ) that @xmath1575 in particular , @xmath1576 and @xmath1577 .
we apply the lemmas in appendix [ a : int ] for @xmath1578 , @xmath1579 and @xmath1355 .
observe that the modulus function of @xmath1580 acting on @xmath1581 by conjugation is @xmath1582 , i.e. @xmath1583 write @xmath1584 where @xmath1585 [ l : iterim ] for any @xmath1586 and @xmath1105 we have 1 . [ claim : evencase ] @xmath1587 .
[ claim : tavg ] @xmath1588 .
[ claim : rkodd ] @xmath1589 .
therefore for @xmath1590 , @xmath1591 we first recall ( similarly to ) that for all @xmath885 @xmath1592 for part [ claim : evencase ] , use lemma [ c : fourier2 ] , with the data @xmath1593 , @xmath1594 , @xmath1595 , @xmath1596 , @xmath1597 , @xmath1598 and @xmath1599
. from and lemma [ lem : lproperty ] , the conditions of lemma [ c : fourier2 ] are satisfied . for part [ claim : tavg ] , use lemma [ l : fourier2 ] with the data @xmath1600 , @xmath1601 , @xmath1602 , @xmath1603 and @xmath1599 .
the conditions in lemma [ l : fourier2 ] are easy to verify in our case .
in particular , for @xmath1604 , @xmath1605 and @xmath1606 , we have @xmath1607 . thus we apply lemma [ l : fourier2 ] with @xmath1608 and @xmath1609 . taking into account
, we get part [ claim : tavg ] .
for part [ claim : rkodd ] , use lemma [ l : fourier ] , with the data @xmath1610 , @xmath1611 , @xmath1612 , @xmath1613 , @xmath1614 and @xmath1599 .
it is easy to check that @xmath271 normalizes @xmath134 . from and
lemma [ lem : lproperty ] , to check the remaining conditions of lemma [ l : fourier ] we observe that for @xmath810 and @xmath1615 we have @xmath1316\in \check{{\mathcal x}}^{2k-1}$ ] with @xmath1616)=\psi(\sum_{i=1}^{k-1 } c_{2{\mathbf{n}}-k+1+i,2{\mathbf{n}}-k - i+1}(d_{2{\mathbf{n}}-k - i,2{\mathbf{n}}-k+1+i}+d_{2{\mathbf{n}}-k - i+1,2{\mathbf{n}}-k+i}))\ ] ] ( where @xmath1617 ) .
thus @xmath1618)$ ] gives an isomorphism @xmath1619 preserving haar measure .
applying lemma [ l : fourier ] , we get part [ claim : rkodd ] .
finally , the last part follows by induction since @xmath1620 .
to finish the proof of proposition [ prop : factor ] we apply the above lemma to @xmath1621 with @xmath1532 . then @xmath1622 .
since @xmath1623 and @xmath1624 , the proposition follows from the @xmath1625 case of and lemma [ lem : manipim ] .
[ sec : box ] [ [ section-4 ] ] next we note the following simple lemma .
[ lem : switch ] let @xmath1626 and @xmath1627 be two subgroups of @xmath1628 , @xmath4 a character of @xmath1626 , @xmath1567 a function of @xmath1629 .
let @xmath1630 be such that @xmath1631\in g_1 $ ] for all @xmath1632 .
let @xmath1633 be the function of @xmath1629 given by @xmath1634)$ ] .
( it is not necessarily a character . )
then for any @xmath1635 we have @xmath1636}\circ { { \mathcal l}}_{g_2,\xi}\circ f={{\mathcal l}}_{g_2,\xi\chi_c}\circ{{\mathcal e}[c]}\circ f$ ] .
this is just a restatement of the identity @xmath1637
cb\cdot)\xi(b)\ db=\int_{g_2}f(cb\cdot)\xi(b)\chi_c(b)\ db.\qedhere\ ] ] let @xmath1638 . set @xmath1639 and @xmath1640 with some parameter @xmath1641 .
let @xmath1642 so that @xmath1643 .
[ lem : elemv ] for any @xmath1644 , @xmath1051 and @xmath1645 we have @xmath1646\in\check{{\mathcal x}}^k$ ] and @xmath1647)=1 $ ] .
moreover , if @xmath1395 then for any @xmath1170 and @xmath1645 we have @xmath1646\in\check{{\mathcal x}}^k$ ] and @xmath1647)=1 $ ] . if @xmath1644 then @xmath1648 and therefore @xmath1646\in\check{{\mathcal x}}^k$ ] . also , @xmath1647)=1 $ ] since @xmath1649 is a character . in the case
@xmath1395 it is no longer true that @xmath1650 .
however , @xmath1651 and thus @xmath1652\subset [ \overline{{\mathcal x}}^k,\hat{{\mathcal x}}^k]\subset\hat{{\mathcal x}}^k$ ] by lemma [ lem : lproperty ] . similarly , @xmath1653 so that @xmath1652\subset[\hat{{\mathcal x}}^{k+1},{{\mathcal x}}^{4{\mathbf{n}}-k-1}]\subset\hat{{\mathcal x}}^{k+1}$ ]
. thus @xmath1652\subset\hat{{\mathcal x}}^k\cap\hat{{\mathcal x}}^{k+1}=\check{{\mathcal x}}^k$ ] .
moreover , explicit computation ( using the particular form of @xmath1654 and @xmath1078 ) shows that @xmath1647)=1 $ ] .
let @xmath1655 be the kernel of the character @xmath1656 of @xmath997 .
note that @xmath1657 contains any unipotent subgroup of @xmath997 and that @xmath1658 let @xmath1659 be the function on @xmath1535 defined by @xmath1660 explicitly if @xmath1661 then @xmath1662 .
let @xmath1663 for @xmath1532 .
[ lem : spcltrans ] for any @xmath1532 we have @xmath1664}\circ{{\mathcal l}}^{h_{{\mathbb m}}}_{{\mathcal{z}},{\psi_{{\mathcal{z}}}}^{-1}}\circ { { \mathcal l}}^{n_{{\mathbb m}}}_{{\mathcal{p}}_{{{\mathbb m}}}'\cap h_{{\mathbb m}}}\circ w.\ ] ] thus , if @xmath502 is unitarizable , for any @xmath501 such that @xmath1665 we have @xmath1666}\circ{{\mathcal l}}^{h_{{\mathbb m}}}_{{\mathcal{z}},{\psi_{{\mathcal{z}}}}^{-1}}\circ { { \mathcal l}}^{n_{{\mathbb m}}}_{{\mathcal{p}}_{{{\mathbb m}}}\cap h_{{\mathbb m}}}\circ w.\ ] ] as in the proof of proposition [ prop : factor ] , let @xmath1621 ; then @xmath1622 . as in lemma
[ lem : manipim ] we get the following iterated integral expression of @xmath1667 : @xmath1668 here we use the fact that @xmath1669 when @xmath1670 for @xmath1671 .
now use for @xmath1501 and then part [ claim : evencase ] of lemma [ l : iterim ] ( for @xmath1625 ) . recalling that @xmath1577 we get @xmath1672 next note @xmath1673 .
moreover for @xmath1674 , we have @xmath1675\in \hat{{\mathcal x}}^{2{\mathbf{n}}-1}$ ] with @xmath1676)=\xi_{{{\epsilon'}}}(t){\psi}_e({{\epsilon'}})^{-1}$ ] . from lemma [
lem : switch ] we get @xmath1677}\circ{{\mathcal l}}^{h_{{\mathbb m}}}_{\hat{{\mathcal x}}^{2{\mathbf{n}}-1},{\psi}_e^{-1 } } = { { \mathcal e}[{{\epsilon'}}]}\circ{{\mathcal l}}_{t^{\mathbf{n } } , \tilde\xi_{\mathbf{n}}}\circ{{\mathcal l}}^{h_{{\mathbb m}}}_{\hat{{\mathcal x}}^{2{\mathbf{n}}-1},{\psi}_e^{-1}}.\ ] ] from parts [ claim : tavg ] and [ claim : rkodd ] of lemma [ l : iterim ] for @xmath1625 , to prove we are left to show that @xmath1678}\circ{{\mathcal l}}^{h_{{\mathbb m}}}_{{{\mathcal x}}^{2{\mathbf{n}}}_\circ\ltimes\check{{\mathcal x}}^{2{\mathbf{n}}-1},{\psi}_e^{-1}}= { { \mathcal e}[{{\epsilon'}}]}\circ{{\mathcal l}}_{{\mathcal{u}}^{{\mathbf{n}}-1}}\circ{{\mathcal l}}^{h_{{\mathbb m}}}_{{{\mathcal x}}^{2{\mathbf{n}}}_\circ\ltimes\check{{\mathcal x}}^{2{\mathbf{n}}-1},{\psi}_e^{-1}}.\ ] ] from lemma [ lem : elemv ] ( for @xmath1395 ) we get that for any @xmath1679 , @xmath1680\in\check{{\mathcal x}}^{2{\mathbf{n}}-1}$ ] with @xmath1681)=1 $ ] .
the above relation follows from lemma [ lem : switch ] .
when @xmath501 , we also have @xmath1682 by proposition [ prop : sametran ] .
the second assertion follows .
[ [ section-5 ] ] define a character @xmath1683 on @xmath926 by @xmath1684 .
we show the following variant of proposition [ prop : 14 ] .
[ lem : 14var ] for any @xmath686 we have @xmath1685}\circ{{\mathcal l}}_{{\mathcal{a}},{\psi_{{\mathcal{a}}}}^{-1}}^{h , reg}\circ l.\ ] ] again by continuity , we only need to prove the identity for @xmath816 . by corollary [ cor : suppw ] we can write the left - hand side as an iterated integral : @xmath1686 here we note that @xmath1683 is trivial on all @xmath1386 with @xmath1644 . applying for @xmath1687
we get that the above is @xmath1688 note that @xmath1689 and @xmath1690 . from property and
the statements immediately preceding it , ( applied in case @xmath1625 ) we get for any @xmath1691 , @xmath1692\in \hat{{{\mathcal n}}}^{2{\mathbf{n}}-1}$ ] with @xmath1693)={\hat{\psi}_{{\bar u^\triangle}}}(\bar v)$ ] .
thus from lemma [ lem : switch ] the above equals : @xmath1694}\circ{{\mathcal l}}^h_{\hat{{{\mathcal n}}}^{2{\mathbf{n}}-1},{{\psi}_{x_0}}^{-1}}\circ l={{\mathcal e}[{{\epsilon}}]}\circ{{\mathcal l}}_{{\bar{u}}^{2{\mathbf{n}}-1}}\circ{{\mathcal l}}^h_{\hat{{{\mathcal n}}}^{2{\mathbf{n}}-1},{{\psi}_{x_0}}^{-1}}\circ l.\ ] ] the lemma follows now from ( with @xmath1395 ) .
[ [ section-6 ] ] let @xmath1695 be the direct product of the commuting one - parameter root subgroups @xmath1696 .
we supplement lemma [ l : mreg2 ] with the following : [ l : regcomp ] let @xmath674 and @xmath1106 .
then for any @xmath1697 we have @xmath1698}\circ \phi(t,\cdot)= { { \mathcal
e}[{{\epsilon'}}]}\circ { { \mathcal l}}^{reg}_{{\mathcal{v}}^k , \psi_{{{v_{\triangle}}}}}\circ \phi(t,\cdot)\ ] ] if either side is well defined .
similarly , for any @xmath1699 ( and if @xmath1501 when @xmath1170 ) we have @xmath1700}\circ \phi(t,\cdot)={{\mathcal e}[{{\epsilon'}}]}\circ { { \mathcal l}}_{{\mathcal{u}}^k}\circ \phi(t,\cdot).\ ] ] finally , suppose that @xmath1701 .
then for all @xmath1008 a compact subset of @xmath291 , as a function of @xmath1702 , @xmath1703}\circ\phi(t , m)$ ] is supported on some @xmath1704 determined by @xmath1007 and @xmath256 . to prove , we show that for @xmath1705 and @xmath1697 , we have @xmath1706 .
this follows from since by lemma [ lem : elemv ] and , @xmath1646\in \hat{{\mathcal x}}^{2k}\cap { { \mathbb m}}^{{{{\flat}}{{\flat}}}}=\hat{{\mathcal m}}^{2k}$ ] and @xmath1647)=1 $ ] .
similarly , for @xmath1699 we have @xmath1706 for @xmath1707 , hence .
now suppose that @xmath1701 .
note that for @xmath1708 , @xmath1709\in \hat{{\mathcal m}}^{2{\mathbf{n}}}$ ] by lemma [ lem : lproperty ] part [ claim : normall ] and .
thus @xmath1703}\circ\phi(t , nm)=\phi(t,{{\epsilon'}}nm)={{\psi}_{x_{{\mathbb m}},t}}'(n){{\mathcal e}[{{\epsilon'}}]}\circ\phi(t , m)$ ] where @xmath1710 .
for fixed @xmath1711 , let @xmath1712 be the product of @xmath1713 .
let @xmath1714 . then @xmath1715 fix @xmath1711 with @xmath1716 .
let @xmath1717 be the product of @xmath1718
. then @xmath1719 and @xmath1720 when @xmath1721 and @xmath1722 .
moreover , @xmath1723)=\psi({\left\langle\overline c , d\right\rangle}_{y_j , t})$ ] where @xmath1724 is the projection of @xmath162 to @xmath1712 and @xmath1725 is a non - degenerate pairing between @xmath1712 and @xmath1717 depending continuously on @xmath838 .
thus for any @xmath1726 , @xmath1727 and @xmath1722 we have @xmath1728 by condition .
lemma [ l : elemc ] implies that for any @xmath1008 , @xmath1729 is supported on some @xmath1730 which is determined by @xmath1007 , and @xmath256 and independent of @xmath1726 .
it follows from , @xmath1731 is compactly supported on @xmath1732 if and only if it is so on @xmath1733 . since @xmath1734 , we conclude that @xmath1125 is compactly supported on @xmath1735 , uniformly for @xmath1008 .
[ cor : mreg ] let @xmath1050 , @xmath1170 and @xmath1736 .
then @xmath1737\circ\cdots\\ \circ{{\mathcal l}}_{{\mathcal{u}}^1}\circ{{\mathcal l}}^{reg}_{{\mathcal{v}}^1,\psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_1}}[f_1]\circ{{\mathcal e}[{{\epsilon'}}]}\circ{\upsilon}_{w_s}(t , e)\end{gathered}\ ] ] is analytic in @xmath1738 and locally constant in @xmath1170 , uniformly in @xmath1738 .
it equals @xmath1739}\circ { { \mathcal l}}_{{\mathcal{u}}^{{\mathbf{n}}-1}}\circ{{\mathcal l}}^{reg}_{{\mathcal{v}}^{{\mathbf{n}}-1},\psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_{{\mathbf{n}}-1}}}[f_{{\mathbf{n}}-1}]\circ\cdots\\ \circ{{\mathcal l}}_{{\mathcal{u}}^1}\circ{{\mathcal l}}^{reg}_{{\mathcal{v}}^1,\psi_{{{v_{\triangle}}}}}\circ{{\mathcal l}^{{t}_1}}[f_1]\circ{\upsilon}_{w_s}(t , e).\end{gathered}\ ] ] moreover it equals @xmath1740 when the integrals converge . here the character @xmath1313 on @xmath1741 factors through @xmath1114 .
it follows from lemma [ l : ureg ] ,
corollary [ cor : abcde ] and lemma [ l : regcomp ] that , for all @xmath838 in a fixed compact subset @xmath1742 , the iterated integral equals the integration with the domain of integration for @xmath1743 and @xmath1744 replaced by some open compact subgroups @xmath1745 and @xmath1746 , dependent on @xmath131 and @xmath1742 , but independent of @xmath1738 .
thus we get that is analytic in @xmath1738 and locally constant in @xmath1170 , uniformly in @xmath1738 .
the second claim follows from lemmas [ l : regcomp ] , as clearly the operators @xmath1747 $ ] and @xmath1703}$ ] commute .
the last claim follows from lemma [ l : stableprop ] . from proposition
[ prop : main ] , we get for @xmath686 and @xmath1748 , for any @xmath1170 and @xmath1749 , @xmath1750}\circ{{{\mathcal t}}_{(h,1)}^{({\mathfrak{e } } , { \psi_{{\mathfrak{e}}}})}}(l).\ ] ] [ [ section-7 ] ] let @xmath1751
. then @xmath1752 and @xmath1753 .
let @xmath1754 .
it consists of @xmath1743 with @xmath1755 whenever @xmath1756 and @xmath1757 .
we extend the character @xmath1683 trivially to @xmath1758 .
it is proved in ( * ? ? ?
* lemma 4.5 ) that for any @xmath1045 , @xmath1759 is compactly supported .
in particular , the integral @xmath1760 is well defined .
we have the following corollary of the previous discussion .
[ cor : fornextpaper ] let @xmath1761 , let @xmath1762 , then for any @xmath1170 and @xmath1749 , @xmath1763
. moreover @xmath1764}\circ { { \mathcal l}}_{{\varrho}({n_{{{\mathbb m}},\triangle } } ) } \circ{{\mathcal e}[{\varrho}({{\epsilon'}})]}\circ{{{\mathcal t}}_{(h,1)}^{({\mathfrak{e } } , { \psi_{{\mathfrak{e}}}})}}(l)$ ] . by proposition [ p :
l ] , we see @xmath1765 where @xmath1766 . by , @xmath1767}\circ{{{\mathcal t}}_{(h,1)}^{({\mathfrak{e } } , { \psi_{{\mathfrak{e}}}})}}(l).\ ] ] on the other hand , @xmath1768}\circ{{{\mathcal t}}_{(h,1)}^{({\mathcal{a}},{\psi_{{\mathcal{a } } } } ) } } ( l)\ ] ] by lemma [ lem : 14var ] . using
, we get @xmath1769}\circ{{\mathcal e}[{\hat{w}}]}\circ{{{\mathcal t}}_{(h,1)}^{({\mathfrak{e } } , { \psi_{{\mathfrak{e}}}})}}(l).\ ] ] the corollary follows from the relation @xmath1770}\circ{{\mathcal e}[{\hat{w}}]}={{\mathcal l}}_{{\bar{u}_{\triangle}}}\circ{{\mathcal e}[{\hat{w}}]}\circ{{\mathcal e}[{\varrho}({{\epsilon'}})]}= { { \mathcal e}[{\hat{w}}]}\circ{{\mathcal l}}_{{\varrho}({n_{{{\mathbb m}},\triangle}})}\circ{{\mathcal e}[{\varrho}({{\epsilon'}})]}. \qedhere\ ] ]
in this appendix all groups are algebraic over a local field @xmath0 ( or their groups of @xmath0-points ) and all isomorphisms are defined over @xmath0 .
we fix a group @xmath1771 , a closed subgroup @xmath1772 such that @xmath1773 is affine , and a continuous character @xmath1774 .
the following statements are in the spirit of @xcite .
( see also ( * ? ? ?
* lemma 2.2 ) . )
we fix a non - trivial character @xmath89 of @xmath0 .
it determines a self - dual haar measure on @xmath0 , as well as a haar measure on @xmath122 . throughout this section
we assume that haar measures on groups and their semidirect products are chosen compatibly , and all isomorphisms are compatible with haar measures .
[ l : fourier ] let @xmath1775 be closed unipotent subgroups of @xmath270 satisfying the following properties .
1 . @xmath1776 , @xmath272 normalizes @xmath135 and @xmath1777 .
2 . @xmath1778 , @xmath1779 .
3 . @xmath1780 .
@xmath1781 is normal in @xmath272 and @xmath1782 is abelian .
5 . @xmath1783 and there is a homeomorphism @xmath274 between @xmath271 and the pontryagin dual of @xmath1782 , preserving haar measures .
let @xmath1784 be such that for any @xmath1785 , @xmath810 : @xmath1786 where @xmath1787 is the image of @xmath1788 under the projections @xmath1789 .
then we have @xmath1790 for any @xmath1791 , i.e. @xmath1792 first note that the restriction of @xmath145 to @xmath134 is compactly supported modulo @xmath1793 since @xmath1794 is closed in @xmath1795 , as a unipotent orbit on an affine variety . similarly for @xmath180 and @xmath135 .
make a change of variable @xmath1796 on the left - hand side of . by left @xmath1797-equivariance of @xmath145 ,
it is equal to @xmath1798 this integral equals @xmath1799 let @xmath1800 be the function @xmath1801 .
then the above is just @xmath1802 the second equality of is clear . in most cases
we will apply the lemma when @xmath1803 , i.e. @xmath1804 .
[ c : fourier2 ] let @xmath1775 be closed unipotent subgroups of @xmath270 satisfying the following properties .
1 . @xmath1805 .
2 . @xmath1804 , @xmath1779 , @xmath1778 .
3 . @xmath1806 .
let @xmath1784 be such that @xmath1807 for all @xmath1785 , @xmath810 .
then we have @xmath1808 for any @xmath1791 , i.e. @xmath1809 since @xmath1804 and @xmath1805 we have @xmath1810 .
since @xmath1806 we have @xmath1811 .
we will also need the following variant of lemma [ l : fourier ] .
this will only be used in [ sec : mfactor ] .
[ l : fourier2 ] let @xmath1812 be closed subgroups of @xmath270 satisfying the following properties . 1 .
@xmath134 is a unipotent group .
2 . @xmath1804 and @xmath1813 .
3 . there is an isomorphism @xmath1814 .
4 . there is an isomorphism @xmath1815 .
@xmath271 normalizes @xmath134 .
@xmath1783 . 7 .
@xmath1816 .
let @xmath1817 be such that for any @xmath1785 , @xmath1818 @xmath1819 where @xmath1787 is the image of @xmath1788 under the projections @xmath1820 .
let @xmath1821 , @xmath810 be the modulus function of @xmath1822 acting on @xmath1794 and set @xmath1823 , @xmath1818 .
then we have @xmath1824 for any @xmath1791 .
in other words , @xmath1825 we make a change of variable @xmath1796 on the left - hand side of . by left @xmath1797-equivariance of @xmath145
we get @xmath1826 this integral equals @xmath1827 let @xmath1828
. then @xmath1829 and the above integral is just @xmath1830 as required .
[ rem : condfourier ] we will only use the lemmas above in the case where @xmath1831 is the fixed point subgroup of an involution @xmath1832 of @xmath270 which stabilizes @xmath134 , @xmath135 , @xmath271 , @xmath272 .
thus , the conditions @xmath1780 ( resp , @xmath1806 , @xmath1813 ) are automatically satisfied .
also , in the cases at hand the restriction of @xmath146 to @xmath134 and to @xmath180 will be a character ( and @xmath1833 ) .
( note however that @xmath146 is not a character on @xmath1834 . )
thus , the condition on @xmath146 in lemma [ l : fourier ] is that it is right @xmath271-invariant and @xmath1835)={\left\langle\iota(c),\overline d\right\rangle}$ ] for all @xmath810 , @xmath811 .
for this section let @xmath1836 and let @xmath1837 be the standard borel subgroup of @xmath1 where @xmath60 is the subgroup of upper unitriangular matrices and @xmath527 is the diagonal torus .
let also @xmath535 be the standard maximal compact subgroup of @xmath1 .
let @xmath1838 be the permutation matrix corresponding to the longest weyl element .
we denote by @xmath1839 be the maximum of the norm of the entries of @xmath1840 and @xmath1841 , and let @xmath1842 .
we have @xmath372 , @xmath373 and @xmath83 is bi-@xmath535-invariant .
let @xmath354 be the representation parabolically induced from the trivial representation of the @xmath135 .
let @xmath377 be the unramified vector in @xmath354 such that @xmath378 .
for any @xmath1843 let @xmath1844 .
similarly for @xmath1845 .
we show the following variant of ( * ? ? ?
* proposition 3.2 ) : [ lem : mainbnd018 ] for any @xmath1846 there exists @xmath1847 such that @xmath1848 for all @xmath837 and @xmath1849 .
first , we reduce to the case @xmath1850 , i.e. , to the inequality @xmath1851 indeed , write @xmath1852 where @xmath823 , @xmath1853 and @xmath737
. then by a change of variable we get @xmath1854 by matrix multiplication @xmath1855 .
therefore , @xmath1856 and the above is @xmath1857 note that @xmath1858 .
thus , the estimate will give @xmath1859 ( possibly for a larger @xmath1860 ) .
let @xmath1861 be the parabolic subgroup stabilizing the line @xmath1862 .
let @xmath1863 .
we will prove by induction on @xmath1127 .
for the induction step it suffices to prove that for any @xmath1864 there exists @xmath1847 such that @xmath1865 for all @xmath1866 and @xmath837 .
( note that @xmath1867 and @xmath1868 are unchanged if we view @xmath1869 as an element of @xmath1870 . )
clearly there exists @xmath1871 such that @xmath1872 for any @xmath1873 .
let @xmath1874 denote the subgroup of @xmath551 consisting of elements with @xmath1875 .
on @xcite it is proved that for any @xmath1876 there exists @xmath887 such that @xmath1877 for all @xmath1878 .
we will use the following fact ( ( * ? ? ?
* lemme ii.4.2 ) ) : there exists @xmath897 such that @xmath1879 this implies for all @xmath271 and @xmath738 @xmath1880 indeed , writing @xmath1881 where @xmath1882 and @xmath1883 then by a change of variable @xmath1884 the left - hand side of is @xmath1885 by a similar reasoning @xmath1886 we break the integral in according to the regions @xmath1887 and @xmath1888 where @xmath1889 is a constant which will be determined below .
we write @xmath1890 and @xmath1891 for the corresponding integrals so that @xmath1892 write @xmath1893 . in the domain of @xmath1890
we have @xmath1894 .
thus @xmath1895 by .
consider now @xmath1891 . on the domain of @xmath1891 we have @xmath1896 .
also @xmath1897 so that @xmath1898 .
we write @xmath1899 where @xmath1900 and @xmath1901 . once again
, we break the domain of integration of @xmath1891 to @xmath1902 and @xmath1903 where @xmath1904 is a small constant which will be fixed later .
we write @xmath1905 and @xmath1906 for the corresponding integrals so that @xmath1907 in domain of @xmath1905 we have @xmath1908 and @xmath1909 .
thus , by @xmath1910 together with it follows that ( here @xmath1911 ) @xmath1912 taking @xmath1913 and @xmath1914 we get @xmath1915 by . for @xmath1906 , once again write @xmath1893 . by the same reasoning as above , @xmath1916 . on the domain of integration of @xmath1906 we have @xmath1917 and hence @xmath1918 .
therefore for suitable @xmath1919 we have @xmath1920 by .
combining , , , and we get .
the lemma follows .
as in the body of the paper for @xmath1921 we denote by @xmath68 the langlands quotient of the induced representation @xmath1922 where @xmath1923 and @xmath1924 is the ( upper triangular ) siegel parabolic subgroup of @xmath1 ( with @xmath1925 ) . also recall that @xmath60 is the group of upper unitriangular matrices in @xmath1 and @xmath70 is a character on @xmath60 which is trivial on @xmath71 and non - degenerate on @xmath1926 .
the goal of this section is to prove the following uniqueness result .
[ prop : tadic ] let @xmath1927 .
then @xmath1928 in other words , ( since @xmath1929 ) the jacquet module @xmath1930 of @xmath68 with respect to @xmath522 ( a finite length representation of @xmath526 ) has a unique irreducible generic subquotient .
* for any @xmath125-group @xmath126 we denote by @xmath1931 the grothendieck group of the category @xmath1932 of smooth ( complex ) representations of finite length of @xmath126 .
it is an ordered abelian group .
* we denote by @xmath1933 the one - dimensional representation of the trivial group ( @xmath1934 or @xmath1935 depending on the context ) .
* we write @xmath1936 and @xmath1937 . * for @xmath1938 , @xmath13
we denote by @xmath1939 the representation parabolically induced from @xmath1940 ( with respect to the upper triangular parabolic subgroup ) . * for @xmath1941 and @xmath1942
we denote by @xmath1943 the representation parabolically induced from @xmath1944 ( once again , with respect to the upper triangular parabolic subgroup ) . *
if @xmath1945 we denote by @xmath1946 , @xmath649 the twist of @xmath1947 by the character @xmath1948 of @xmath155 . * if @xmath1949 with @xmath1950 we write @xmath1951=\{a,\dots , b\}$ ] . by a segment @xmath1951_\rho$ ]
we mean a set of the form @xmath1952 where @xmath621 is an irreducible supercuspidal representation of some @xmath155 .
to such a segment one associates an essentially square - integrable representation @xmath1953_\rho)$ ] of @xmath1954 .
all essentially square - integrable representations of @xmath1955 are obtained this way , for a unique segment . by convention
the empty segment corresponds to @xmath1933 , i.e. @xmath1956_{\rho})=1 $ ] .
* we write @xmath1957_{\rho_1}\mapsto[a_2,b_2]_{\rho_2}$ ] if 1 .
@xmath1957_{\rho_1}\cup[a_2,b_2]_{\rho_2}$ ] forms a segment which properly contains @xmath1957_{\rho_1}$ ] and @xmath1958_{\rho_2}$ ] , and , 2 .
@xmath1959 with @xmath1960 . *
more generally , for a multisegment @xmath1957_{\rho_1},\dots,[a_k , b_k]_{\rho_k}$ ] ordered such that @xmath1961_{\rho_i}\not\mapsto[a_j , b_j]_{\rho_j}$ ] for all @xmath1962 , we denote by @xmath1963_{\rho_1},\dots,[a_k , b_k]_{\rho_k})\ ] ] the langlands quotient ( i.e. , the unique irreducible quotient ) of @xmath1964_{\rho_1})\times\dots\times\delta([a_k , b_k]_{\rho_k})$ ] .
this gives a bijection between irreducible representations of @xmath1955 and multisegments .
the supercuspidal support of the representation above is @xmath1965_{\rho_i}$ ] ( multiset union ) . * for @xmath1966 let @xmath1967 be the langlands quotient of @xmath1968 .
* we denote by @xmath1969 semisimplification of the jacquet module @xmath1970 of @xmath1942 with respect to the siegel parabolic and by @xmath1971 the generic part of @xmath1972 ( that is , we retain only the generic representations in the composition series of @xmath1970 , with the same multiplicities ) .
* we have @xmath1973 where @xmath1974 is defined in @xcite .
it satisfies @xmath1975 .
* we denote by @xmath1976 the generic part of @xmath1977 . * for our purposes
we only need to know @xmath1977 and @xmath1978 for essentially square - integrable @xmath1947 .
we have @xmath1979_\rho))=\sum_{c\in [ a-1,b]}\delta([c+1,b]_\rho)\times\delta([-c ,- a]_{{\rho^{\vee}}})\ ] ] while it follows from @xcite that if @xmath1980_\rho)$ ] then @xmath1981 equals @xmath1982_\rho,[d+\frac12,a-\frac12]_\rho)\times \delta([\frac12-d , a+\frac12]_{{\rho^{\vee}}},[\frac12-c , a-\frac12]_{{\rho^{\vee}}}).\ ] ] where the sum is over @xmath1983 $ ] and @xmath1984 $ ] such that @xmath1985 .
in particular , @xmath1986 * if @xmath1966 then we can write @xmath1987 with @xmath1988 square - integrable . then @xmath1989 .
it follows that holds for @xmath1966 as well .
we are now ready to prove proposition [ prop : tadic ] . since @xmath1990 it follows from frobenius reciprocity that @xmath1991 is a quotient of @xmath1930 . in particular , @xmath1992 we will show that @xmath1991 is the only generic irreducible subquotient of @xmath1930 , i.e. that @xmath1993 .
consider first the case where @xmath1994 where @xmath1995 are square - integrable and distinct .
we will prove the statement by induction on @xmath885 .
the case @xmath1996 follows from the results of @xcite .
more precisely , suppose that @xmath1997_\rho)$ ] with @xmath1998 .
let @xmath1999 be such that @xmath2000 is reducible .
since @xmath2001 , @xmath67 is reducible and therefore @xmath2002 $ ] @xcite
. also @xmath2003 by the results of shahidi @xcite . by @xcite there exist two inequivalent subrepresentations @xmath2004 , @xmath2005 of @xmath67 ( with @xmath2006 if @xmath2007 ) such that @xmath2008}\delta([-l , a-\frac12]_\rho)\times\delta([l+1,a+\frac12]_\rho),\\ { s_{{\operatorname{gl}}}}(\delta_2)=\sum_{l\in [ \epsilon , a-\frac12]}\delta([-l , a+\frac12]_\rho)\times\delta([l+1,a-\frac12]_\rho).\end{gathered}\ ] ] since @xmath2009}\delta([-l , a-\frac12]_\rho)\times\delta([l+1,a+\frac12]_\rho)\ ] ] we infer that
. we will not need it . ]
@xmath2010}\delta([l+1,a+\frac12]_\rho,[-l , a-\frac12]_\rho).\ ] ] in particular , @xmath2011 for the induction step , assuming that the statement holds for @xmath885 and we will prove it for @xmath2012 .
write @xmath2013_{\rho_i})$ ] , @xmath2014 with @xmath2015 .
let @xmath2016 . since @xmath68 is the unique irreducible quotient of @xmath2017 , it is also a quotient of @xmath2018 ( which is a quotient of @xmath2019 ) .
taking jacquet module we infer that @xmath2020 in particular , by induction hypothesis and @xmath2021}\delta([c+\frac12,a_{k+1}+\frac12]_{\rho_{k+1}})\times\delta([\frac12-c , a_{k+1}-\frac12]_{\rho_{k+1 } } ) \times\pi'\nu^{-\frac12}.\end{gathered}\ ] ] in a similar vein , @xmath2022 where @xmath2023 .
we may assume without loss of generality that @xmath2024_{\rho_{k+1}}\not\subseteq[-a_i , a_i]_{\rho_i}$ ] ( and hence , @xmath2025_{\rho_i}$ ] ) for all @xmath623 .
comparing supercuspidal supports , we observe that there are no common irreducible subrepresentations of @xmath2026}\delta([c+\frac12,a_{k+1}+\frac12]_{\rho_{k+1}})\times\delta([\frac12-c , a_{k+1}-\frac12]_{\rho_{k+1 } } ) \times\pi'\nu^{-\frac12}\ ] ] and @xmath2027 since @xmath2028 occurs in the support of the supercuspidal support of all summands in the former expression , but not in the latter . therefore , by @xmath2029 which gives the induction step .
consider the general case . by the result of matringe @xcite , we can write @xmath2030 where @xmath2031 is tempered and @xmath1947 is of the form considered above .
( in fact , @xmath1947 is uniquely determined by @xmath7 . ) the intertwining operator @xmath2032 is the composition of the intertwining operators @xmath2033 thus , the image @xmath68 is a subquotient of the image of the middle one ( denoted @xmath2034 ) which is @xmath2035 .
( in fact , it is a direct summand since the latter is unitarziable . )
thus , @xmath2036 and hence by the previous part and @xmath2037 as required .
this concludes the proof of proposition [ prop : tadic ] .
avraham aizenbud and dmitry gourevitch , _ generalized harish - chandra descent , gelfand pairs , and an archimedean analog of jacquet - rallis s theorem _
, duke math .
j. * 149 * ( 2009 ) , no . 3 , 509567 , with an appendix by the authors and eitan sayag .
mr 2553879 ( 2011c:22026 ) i. m. gelfand and d. a. kajdan , _ representations of the group @xmath2040 where @xmath535 is a local field _ , lie groups and their representations ( proc .
summer school , bolyai jnos math .
budapest , 1971 ) , halsted , new york , 1975 , pp .
mr 0404534 ( 53 # 8334 ) david ginzburg , stephen rallis , and david soudry , _ on a correspondence between cuspidal representations of @xmath2041 and @xmath2042 _ , j. amer .
* 12 * ( 1999 ) , no . 3 , 849907 .
mr 1671452 ( 2000b:22018 ) herv jacquet and joseph shalika , _ exterior square @xmath107-functions _ , automorphic forms , shimura varieties , and @xmath107-functions , vol .
ii ( ann arbor , mi , 1988 ) , perspect .
, vol . 11 , academic press , boston , ma , 1990 , pp .
mr 1044830 ( 91g:11050 ) to3em , _ on a new functional equation for local integrals _ , automorphic forms and related geometry : assessing the legacy of i.i .
piatetski - shapiro , contemp .
614 , amer .
soc . , providence , ri , 2014 , pp .
261294 .
to3em , _ square integrable representations of segment type ( generic reducibilities ) _ , functional analysis viii , various publ .
( aarhus ) , vol .
47 , aarhus univ . ,
aarhus , 2004 , pp .
mr 2127173 ( 2005m:22019 ) a. v. zelevinsky , _ induced representations of reductive @xmath2050-adic groups .
ii . on irreducible representations of @xmath2051
cole norm .
( 4 ) * 13 * ( 1980 ) , no . 2 , 165210 .
mr 584084 ( 83g:22012 ) | we study the interplay between different models of the same irreducible representation of the @xmath0-points of a reductive group over a local field . |
oblivious transfer ( ot ) is an important two - party cryptographic protocol , as a building block for many general cryptographic primitives . in the first ot system introduced by rabin @xcite ,
a message is received with probability 1/2 and the sender does not know whether the message reaches the receiver .
this is later called the all - or - nothing ot or simply the rabin ot . even et al .
@xcite defined the 1-out - of-2 ot , where the sender has two secrets and the receiver can choose one and only one of them in an oblivious manner .
that is , the sender can not know the receiver s choice and the receiver can not know anything more than one secret .
the two types of ot are shown to be equivalent in the classical world @xcite , in the sense that one form of ot can be used as a building block to construct the other .
for both types of ot in the classical world , it is rather obvious that unconditional , information theoretic security can not be achieved for both sides at the same time . therefore , computational assumptions such as the existence of trapdoor function is required . in the practical sense ,
the hardness of factorization and discrete logarithm are popular assumptions that are often required in many cryptographic protocols .
such assumptions are deeply threatened by the development of quantum computing , due to the shor s algorithm @xcite . on the other hand
, quantum techniques also provide new potential tools for the construction of cryptographic primitives .
for example , the bb84 protocol @xcite is proposed for key agreement with unconditional security , which is impossible in the classical world .
therefore , there is hope that quantum cryptography will develop faster than quantum cryptanalysis , and a new kind of cryptography will be ready to replace the old one before it turns out to be insecure .
moreover , it is interesting to note that the original idea of ot comes from the quantum realm , in the novel paper by wiesner @xcite .
based on the bb84 key agreement , in 1994 crpeau @xcite proposed a novel protocol for 1-out - of-2 ot which resolves most problems known in previous schemes .
the security of this scheme relies on the use of another important cryptographic primitive , the bit commitment ( bc ) .
since then , in @xcite and @xcite it has been formally proved that @xcite is secure with the use of quantum bc .
the construction of ot from bc using quantum techniques is itself an achievement , since such a construction is not known in the classical world .
unfortunately , in 1997 a few major impossibility results were found concerning ot and bc .
these include the work of lo against the 1-out - of-2 ot @xcite and the work of mayers @xcite against bc .
lo and chau @xcite also independently argued that unconditionally secure quantum bc is impossible .
this is a major setback and breakthrough in the research of quantum based ot , bc , and general cryptographic protocols .
since then , new protocols have been proposed , avoiding the known impossibility results in various ways .
for example , quantum bc can be secure based on computational complexity assumptions @xcite or physical assumptions about the technology used by the adversary @xcite .
on the other hand , currently there is no known impossibility results against the rabin ot .
it is unknown if the equivalence between rabin ot and 1-out - of-2 ot in the classical world applies to the quantum world .
if this is the case , then quantum rabin ot is also impossible . in this paper , to investigate the possibility of rabin ot , we study a weakened form of rabin ot . in @xcite , the idea of the bb84 protocol is used to construct ot .
the attack of delayed measurement is the major reason that bc is required . without bc
, there is no suitable time for sending the information about the basis of the qubits . in this paper ,
we propose a weak rabin ot protocol based on the b92 key agreement protocol @xcite .
our observation is that , unlike the bb84 protocol , in the b92 scheme the sender never needs to send the information about the basis of the qubits . despite the general impossibility
, there are some known differences for ot in the classical and quantum world . in the classical world , ot is impossible even in the honest - but - curious model , where the parties must follow the protocol , but could try to gain more information through private computations . in the quantum world , if the parties follow the protocol strictly , ot would be possible . in this paper
, we show that our weak ot protocol has some properties that are not possible in the classical world .
this is another difference between the classical and quantum ot impossibilities .
it also suggests that an impossibility result for rabin ot may not be easy to obtain , since a weak version of it is actually possible .
there may be several ways to weaken the security definition of rabin ot .
for the sake of our study , we give the following definition for the weak rabin ot ( wrot ) protocol , between sender alice and receiver bob . 1 .
the honest alice inputs a random message bit @xmath0 .
2 . the honest bob receives @xmath0 correctly from the honest alice with probability @xmath1 , a value specified in the protocol .
otherwise bob receives zero information about @xmath0 .
bob knows if he has received @xmath0 .
if he does not receive @xmath0 , he outputs @xmath2 .
3 . for any run of the protocol , a cheating alice has limited advantage @xmath3 to guess or change the probability that bob outputs @xmath2 .
that is , in alice s final view , if the probability that bob outputs @xmath2 is @xmath4 , then @xmath5 .
note that a cheating alice needs not prepare any message bit .
4 . a cheating bob always tries to increases the overall rate that he gets @xmath0 , possibly through guessing .
he needs not be certain about the correctness of the output , so he never outputs @xmath2 . for any cheating bob , his advantage is limited to @xmath6 . if the final rate that he gets @xmath0 correctly is @xmath7 , then @xmath8 .
under this definition , alice can cheat either by trying to find out what happens at bob s side when he decodes the bit , or by really changing the probability that bob gets @xmath2 . in the wrot ,
alice can change the probability that bob gets @xmath2 by @xmath3 .
also , bob can increase the rate he gets @xmath0 to @xmath9 .
this is why the protocol is called weak . at the beginning
the two parties specify two non - orthogonal quantum states @xmath10 and @xmath11 of one qubit .
they use @xmath10 to represent bit 0 and @xmath11 to represent 1 . according to @xmath0
, alice prepares @xmath10 or @xmath11 to send to bob , and bob uses the positive operator - valued measure ( povm ) method to distinguish the two states @xcite unambiguously .
since only the angle @xmath12 between vectors @xmath10 and @xmath11 affects bob s ability to distinguish them , we set @xmath13 and @xmath14 , with @xmath15 . in this case
the povm elements are : @xmath16 where the decode probability for bob is @xmath17 in the honest case . in other words , bob gets @xmath2 with probability @xmath18 .
the measurement matrices @xmath19 represent the output symbols @xmath20 of bob , respectively .
the actual implementation of the povm is not important .
the method in @xcite suggests that bob only needs a unitary operation of two qubits followed by two measurements of one qubit .
first , we argue that alice can not use entanglement for cheating .
notice that in our simple scheme , alice sends a qubit to the honest bob and he measures it with the povm . if the qubit is entangled to some quantum states held by alice , the cheating alice should perform her measurement on such quantum states after bob finishes his .
but due to fundamental physical laws , alice could never receive any information about whether bob has performed a measurement .
so there is no difference if alice measures before bob does . in that case , alice would have created and collapsed the entanglement at the same time , before the qubit is sent out .
this gives her no use of the entanglement .
this argument is valid for any implementation of the povm at bob s side . instead of using entanglement
, alice would better create a pure state qubit @xmath21 she wants to send to bob .
for this case , @xmath21 can be any one qubit state .
in general @xmath21 would be @xmath22 where @xmath23 are real numbers and @xmath24 .
the probability of bob getting @xmath2 is changed to @xmath25 .
we can compute alice s advantage @xmath3 by the difference between this probability and the value @xmath18 , the rate of bob getting @xmath2 in the honest case .
a direct calculation gives @xmath26 if alice is cheating , she could either increase or decrease the chance that bob gets @xmath2 .
she can always make sure that bob does not get @xmath2 at all , when @xmath27 .
the lowest ( negative ) value of @xmath3 is always @xmath28 . from ( [ v ] ) , the highest value of @xmath3 , given @xmath18 , occurs when @xmath29 and @xmath30 . in this case ,
@xmath31 . in the wrot
, alice can freely choose any @xmath3 between the maximum and minimum .
since @xmath32 , alice s advantage @xmath33 can be controlled by the choice of @xmath18 in the protocol .
a smaller @xmath18 provides better security against alice .
next , we consider a cheating bob , who would never output @xmath2 , but would rather try to guess @xmath0 .
he accepts some uncertainty , which is unavoidable since the two given states are non - orthogonal . to provide the lowest error rate guessing @xmath0
, he chooses an optimized projection with orthogonal basis @xmath34 and @xmath35 .
optimization is achieved @xcite when @xmath36 is minimized , where @xmath37 is the angle between @xmath38 and @xmath39 for @xmath40 .
this happens when @xmath41 .
call this angle @xmath42 , the error rate of bob guessing @xmath0 is @xmath43 .
therefore @xmath44 here , we observe a difference between classical and quantum world concerning the wrot . for the classical world , even in the honest - but - curious model of weaker attacks , alice can compute exactly what bob can , based on the communication between them .
that is , alice knows bob s view about the random variable @xmath0 .
therefore , either alice can know for certain that bob outputs @xmath2 , or bob can find a way to compute @xmath0 . in terms of the wrot , it is either @xmath45 or @xmath46 . in the quantum case based on our scheme ,
alice never has complete information of bob s view on @xmath0 , as the result of the povm is unpredictable .
it can be seen that @xmath3 can be much smaller than @xmath17 , while @xmath47 as there is no way to perfectly distinguish non - orthogonal quantum states .
we can study the relation between @xmath18 , the maximized @xmath3 , and the maximized @xmath6 .
figure [ fig : one ] shows the graph of @xmath6 and @xmath3 against @xmath18 .
while a trade - off between @xmath6 and @xmath3 may be expected , in our scheme both @xmath6 and @xmath3 can be suppressed with a lower @xmath18 .
but this is not at all a good news because , in very vague terms , when @xmath18 is too small , the usefulness of the rabin ot is also low , as the uncertainty about whether bob receives @xmath2 is small in alice s point of view .
note that the graph only plots the maximum value of @xmath3 , not the minimum .
the minimum value is @xmath48 and it is significant when @xmath18 is larger .
therefore it is reasonable to consider only small @xmath18 .
if the wrot is used to construct 1-out - of-2 ot using standard techniques @xcite , a weak 1-out - of-2 ot will be resulted , but the weakness parameters of the obtained scheme would be rather high . therefore it does not meet the conditions for enhancement to normal ot @xcite . in this way , our protocol neither violates the impossibility of 1-out - of-2 ot nor suggests inequivalence of rabin ot and 1-out - of-2 ot . on the other hand , it is currently unknown if the normal rabin ot is possible or not . but to seek any impossibility result one must avoid our constructive result of weak rabin ot . c. h. bennett and g. brassard : quantum cryptography : public key distribution and coin tossing , in _ proceedings of the ieee international conference on computers , systems , and signal processing _ , bangalore ( 1984 ) , pp.175179 .
i. damgrd , j. killian , and l. salvail : on the ( im)possibility of basing oblivious transfer and bit commitment on weakened security assumptions , in _ advances in cryptology eurocrypt99 _ , lncs 1592 , pp.5673 , 1999 . | due to the commonly known impossibility results , information theoretic security is considered impossible for oblivious transfer ( ot ) in both the classical and the quantum world . in this paper , we proposed a weak version of the all - or - nothing ot . in our protocol
the honest parties do not need long term quantum memory , entanglements , or sophisticated quantum computations .
we observe some difference between the classical and quantum ot impossibilities .
* keywords : * quantum oblivious transfer , information theoretic security |
study of traffic flow relies in no small part on theoretical / mathematical modeling because of the difficulty in extensive measurements and designed experiments , and predictions from simulation and analytical study have particular importance in this field @xcite .
a car - following model is described in general by an equation of the form : @xmath0 @xmath1 denotes the position of the @xmath2th car at time @xmath3 and a positive constant @xmath4 is a delay in time , which corresponds to the driver s reaction time . after this latency , each car reaches the velocity @xmath5 determined by the headway @xmath6 , i.e. , a distance to the next car in front ; hence @xmath7 is nowadays referred to as _ the optimal - velocity ( ov ) function_. in the earlier studies of traffic flow , g. f. newell proposed the car - following model with @xmath8 where @xmath9 , @xmath10 and @xmath11 are positive constants , and found an exact solution for @xmath12 @xcite . in this case , the equation can be transformed into a linear equation and thus allows for superposition of the shocks each of which present a traffic jam propagating upstream at different speeds . subsequently , g. b. whitham obtained an elliptic solution for newell s model in the generic case @xmath13 @xcite .
this presents nonlinear waves beyond newell s _ linear _ solutions , including a solitary - wave in the case that the modulus of the elliptic function goes to 1 .
he also suggested that the solitary - wave solution be stable in the range of parameters for stable linear waves .
both newell s model and the kac van moerbeke system reduce to the same equation if one assumes a steady - profile solution with the phase velocity @xmath14 ; the kac
van moerbeke system is known as one of the integrable lattice systems @xcite .
this fact motivated whitham to derive a solitary - wave solution of newell s model after a long calculation .
( it is notable that these results consist with the theory of nonlinear waves @xcite . )
bando _ et al .
_ proposed an alternative car - following model described by a second - order differential equation @xmath15 $ ] with the sensitivity @xmath16 ( a positive constant ) and the ov function , @xmath17 where @xmath18 is the inflection point of the _ tanh _ function .
they referred to the model as the optimal - velocity ( ov ) model @xcite .
( note that replacing @xmath19 with @xmath20 for small @xmath21 , one formally recovers the ov model . )
now @xmath22 is the most popular choice for the ov function and is actually suitable for mathematical treatment . in @xcite , they showed exact solutions for the car - following model with the ov function , @xmath23 which they call _ the delayed optimal - velocity ( ov ) model_. the car - following model is totally asymmetric in the lattice variable , and which is one of the crucial properties of traffic flow models .
however , in the above works , a steady - profile solution is chosen so that the delay in time should compensate for the asymmetric lattice shift .
consequently , all the solutions obtained , given by elliptic functions , have the same phase velocity of @xmath14 independently of the form of the ov function . an alternative way to exactly solve a traffic - flow model is to assume a shock - wave solution . inspired by hirota s method @xcite to find the soliton solutions of soliton equations , in @xcite they reduced the delayed ov model expressed in @xmath24 , @xmath25 to two coupled equations , and thus found a shock - wave solution .
it is distinct from the previous results introduced above that their solution includes a free parameter which can control the shock velocity .
they also considered the same scenario for newell s model , finding a shock solution with velocity @xmath26 ; this solution also includes a free parameter but does not allow for the variation of the velocity .
in this work , we solve exactly the discrete delayed ov model firstly introduced in @xcite and thereby integrate the previous results of @xcite and @xcite . in section 2
, we derive the discrete delayed ov model and give exact results .
we also consider the continuum limit as the discrete delayed ov model reduces to the delayed ov model . in section 3 ,
we consider the ultra - discrete limit of the results obtained in the previous section . in this limit
, the model is transformed into a cellular automaton ; thus we obtain a delay cellular automaton with an exact solution .
section 4 is devoted to summary and conclusion .
changing the variable as @xmath27 in , we transform the delayed ov model into an algebraic differential equation : @xmath28 where the transcendental function _ tanh _ disappears . then we discretize and obtain its exact solution assuming the open boundary condition that an infinite number of cars run on the road , i.e. , @xmath29
. first we shall review the relationship between and _ the modified lotka
volterra ( mlv ) equation _ given by @xmath30 where @xmath31 and @xmath32 .
if one assumes a traveling - wave solution to , i.e. , @xmath33 with @xmath34 , yields @xmath35 where @xmath36 stands for the derivative .
meanwhile , changing the variable as @xmath37 , from we have @xmath38 then , we assume the traveling - wave solution , @xmath39 with @xmath40 , to and reach the same equation as : @xmath41 we thus see that both and reduce to the same equation if a traveling - wave solution is assumed . here
we introduce _ the discrete modified korteweg
de vries ( mkdv ) equation _
@xcite , @xmath42 where @xmath43 .
this is known as a discretization of the mlv equation .
( we use superscript @xmath3 as the discrete time variable . ) actually , if we let @xmath44 then reduces to in the continuum limit where @xmath45 regarded as the discrete time unit tends to 0 . in the same manner as above
, we change the variable as @xmath46 in and have @xmath47 where @xmath48 .
assume the traveling - wave solution , @xmath49 with @xmath50 ( @xmath51 ) , and yields @xmath52 here we introduce an alternative variable @xmath53 and then consider @xmath54 with @xmath55 .
if @xmath56 satisfies , we have @xmath57 since @xmath54 ( @xmath58 ) , we may replace @xmath59 with @xmath60 and @xmath61 with @xmath62 , and consequently obtain _ the discrete delayed optimal - velocity ( ov ) model _ : @xmath63 note that possesses the asymmetry in the lattice variable @xmath2 which is required for the car - following model .
let us see that the discrete delayed ov model is the discrete version of the delayed ov model .
we restrict ourselves to @xmath64 , for @xmath10 corresponds to the discrete time unit .
if there exists a smooth function @xmath65 that satisfies @xmath66 , we have @xmath67 where we let @xmath68 .
consequently , taking the limit @xmath69 , we successfully recover from .
accordingly , the discrete variable @xmath70 corresponding to the headway @xmath71 should be defined from as @xmath72 in order to find shock solutions of , we assume @xmath73 where @xmath74 , @xmath75 , @xmath76 and @xmath77 ( @xmath78 ) are constants to be determined . since can transform into @xmath79 , we only consider the case of @xmath80 .
substituting into , we obtain @xmath81 where @xmath82 .
elimination of @xmath76 and @xmath77 from yields the two cases : @xmath83 these are the dispersion relations for the traveling - wave solutions .
remember that @xmath53 corresponds to and that the distance @xmath71 of cars must be positive , and accordingly we shall find the solution in the range of @xmath84 in the case of , from we obtain two shock solutions ; however these lie out of the range . in the case of , together with @xmath76 and @xmath77 determined by we obtain the two shock solutions : @xmath85 figure [ figdiscrete ] shows the graphs of an exact solution for the discrete delayed ov model : the headway @xmath70 given by and .
this illustrates a shock wave propagating upstream ; the shock front corresponds to the tail of a traffic jam .
alternatively , will show an upside - down shock propagating upstream , which illustrates dissolution of a traffic jam ; the shock front corresponds to the head of a traffic jam . of @xmath2th car where @xmath53 is given by with parameters @xmath86 , @xmath87 , @xmath88 , @xmath89 at time @xmath90 , @xmath91 , @xmath92 .
] the parameter values used in the simulation are chosen to satisfy the conditions discussed below .
in general , the range of a shock wave is bounded by the limit values at infinity .
hence we have only to investigate these values in order to find the range of the solution . at first , we consider . since @xmath80 and @xmath51 , the upper limit of is smaller than @xmath93 if and only if @xmath94 is satisfied .
now we assume @xmath94 . since @xmath80 , @xmath51 and @xmath95 , taking the limits as @xmath96 , we find that @xmath53 satisfies @xmath97 for all @xmath3 and @xmath2 .
we introduce the function of @xmath11 , @xmath98 and then have the derivative @xmath99 this leads to @xmath100 for all @xmath80 . consequently , if and only if @xmath101 , we find an @xmath11 that satisfies @xmath102 and a shock solution that satisfies . from @xmath103
, we explicitly show the condition to have an exact shock solution : @xmath104 ( we remark that one can find an _ unphysical _ solution , @xmath105 , unless is considered . ) assuming , we can find the maximum value @xmath106 of @xmath11 such that @xmath102 is satisfied .
consequently , we have @xmath107 next we consider the solution . as well as the above discussion , we find the lower limit of is larger than @xmath108 if and only if @xmath94 is satisfied .
we also assume @xmath94 in this case . taking the limits as @xmath96 , we find that @xmath53 satisfies @xmath109 for all @xmath3 and @xmath2 .
now we let @xmath110 and thus find that ranges from @xmath111 to @xmath112 .
then we see the condition to have an exact solution , @xmath113 and that there exists the maximum value @xmath114 of @xmath11 such that @xmath115 since @xmath116 or @xmath117 is equivalent to @xmath118 , we find @xmath119 for @xmath120 .
the phase velocity of the shock solutions and is given by @xmath121 from this formula , we can see that the phase velocity may attain the maximum at an arbitrary @xmath122 , but however it appears complicated . for simplicity
, we will discuss this point for the continuum version .
as described above , @xmath45 or @xmath10 takes the role of discrete time unit and tends to 0 when we consider the continuum limit .
the dispersion relation implies that @xmath11 converges to @xmath93 or @xmath123 in the continuum limit @xmath69 .
however , it is apparent that @xmath123 is never admissible for .
we hence let @xmath124 , which gives @xmath125 as @xmath126 . also , let @xmath127 , @xmath128 and @xmath3 be replaced with @xmath129 to consider the continuum limit .
( note that the latter @xmath3 denotes the continuous time variable . )
thus we have the exact solutions of , @xmath130 and the dispersion relation , @xmath131 which is common to both solutions .
the solution and the dispersion relation obtained here coincide with those given in @xcite .
also , from we have @xmath132 note that @xmath18 ensures @xmath133 and consequently that the right hand side of is positive . for small @xmath4
, the phase velocity @xmath134 may be approximated by a power series in @xmath4 : @xmath135 this suggests that a quick response of drivers to the traffic situation does not always yield fast propagation of a traffic jam .
we find the maximum velocity @xmath136 at @xmath137 . of @xmath2th car where @xmath138 is given in with minus
sign and parameters @xmath86 , @xmath139 , @xmath140 at time @xmath90 , @xmath91 , @xmath92 . ]
figure [ figcont ] shows an exact shock solution : the headway @xmath141 which corresponds to the solution with minus sign in .
as expected , it illustrates a traffic jam propagating upstream .
in this section , we consider the ultra - discrete limit of the discrete delayed optimal - velocity model . in this limit ,
the dependent variable is discretized and thus we obtain a delay cellular automaton from the delay discrete equation . the ultra - discrete limit was firstly developed to directly relate the soliton equations to cellular automata @xcite .
we define _ the ultra - discrete limit _ or _ ultra - discretization _ by the transformation from a positive real variable @xmath142 to a ultra - discrete variable @xmath143 : @xmath144 where @xmath145 is the ultra - discrete parameter .
( in usual , a ultra - discrete variable is denoted by the capital letter corresponding to that of the variable . ) by linking @xmath142 with @xmath143 by @xmath146 multiplication / division and addition in the original variable are transformed respectively into addition and ` max ' in the ultra - discrete variable : @xmath147 thus , an equation including multiplication and addition but no subtraction transforms into the ultra - discrete equation described with addition and ` max ' in the ultra - discrete variable . since addition and ` max ' become integer functions being restricted to integers , we can regard the ultra - discrete equation as a cellular automaton . as mentioned above
, the ultra - discretization runs only for equations described with multiplication and addition .
we hence change the dependent variable as @xmath148 to eliminate subtraction from : @xmath149 then , we introduce the ultra - discrete variable @xmath150 and parameters @xmath151 and @xmath152 as @xmath153 , @xmath154 , @xmath155 , and thus obtain from the ultra - discrete mkdv equation , @xmath156 where @xmath151 and @xmath152 take integer values as well as @xmath150 @xcite .
we remark that the variable changes , @xmath46 , @xmath148 and @xmath153 , done thus far are collected into @xmath157 taking @xmath158 into account , we introduce the ultra - discrete variable @xmath159 corresponding to the car distance @xmath70 such that @xmath160 where @xmath161 is an ultra - discrete parameter corresponding to @xmath162 , i.e. , @xmath163 and @xmath164 .
then , in order to ultra - discretize , we change the variable as @xmath165 which means @xmath166 and @xmath167 accordingly , from we have @xmath168 the condition for there to be an exact shock solution suggests that we should introduce the ultra - discrete parameter @xmath169 corresponding to @xmath10 as @xmath170 where @xmath171 ; this implies that we focus on around @xmath172 .
consequently , substitution of and into , we obtain _ the ultra - discrete delayed optimal - velocity ( ov ) model _ ,
@xmath173 in the ultra - discrete limit . by taking the ultra - discrete limit ,
we obtain exact solutions of the ultra - discrete delayed ov model from the solutions and of the discrete delayed ov model . according to the above procedure to obtain the ultra - discrete delayed ov model , we transform and into @xmath174 k^nl^{t - m-1}}{(1 - 4\gamma)(l-1)(1+k^nl^{t - m-1})}\\ & = \frac{4\gamma\frac{l - l^{-m}}{l-1 } -1}{1 - 4\gamma } \frac{1 + k^{n+1 } l^{t - m}}{1+k^nl^{t - m-1 } } , } \ ] ] and @xmath175 k^nl^{t - m } } \\ & = \frac1{4\gamma\frac{l - l^{-m}}{l-1 } -1 } \frac{1+k^nl^{t - m}}{1+k^{n+1}l^{t - m+1}}. } \ ] ] ( we exploit the dispersion relation . ) by use of the function @xmath176 defined in , we have @xmath177 from the condition , we see that is positive and hence the prefactors in and are positive .
now we introduce the ultra - discrete parameters as @xmath178 , and @xmath179 . since @xmath80
, we note that @xmath180 .
the dispersion relation is transformed into @xmath181 and yields @xmath182 in the ultra - discrete limit .
note that @xmath171 , and one may reduce this equation to _ the ultra - discrete dispersion relation _ : @xmath183 then let @xmath184 and this , together with , transforms into @xmath185 thus we have @xmath186 in the ultra - discrete limit . since @xmath187 and @xmath51 ,
we compare both sides and consequently obtain @xmath188 finally we obtain exact solutions of the ultra - discrete delayed ov model by ultra - discretizing and : @xmath189 these are surely the solutions of as far as the dispersion relation holds .
the following conditions are added so that the headway @xmath159 should take positive values .
the range of @xmath11 given in or determines the range of @xmath190 for or respectively in the ultra - discrete limit . from we have @xmath191 then , let @xmath192 and we have , together with the above results , the condition , @xmath193 for in the ultra - discrete limit .
as well , from we have the condition @xmath194 for .
figure [ figultra ] shows an exact shock solution .
it provides a quite simple expression of traffic jam propagation . of @xmath2th car with parameters @xmath195 , @xmath196 , @xmath197 , @xmath89 at time @xmath198 , @xmath91 , @xmath199 . ]
in this work , we consider the discrete delayed ov model described by a delay difference equation , where the time variable is fully discretized and the delay is hence given by an integer .
in the continuum limit , the model reduces to the delayed ov model , one of classical traffic - flow models , which is described by a delay differential equation .
we moreover work in the ultra - discrete limit where the dependent variable is also discretized to take integer values , proposing a novel traffic - flow model described by a delay cellular automaton , which we call the ultra - discrete delayed ov model .
we find exact solutions presenting a traffic jam of a shock wave form for the discrete delay model considered .
working in the continuum and ultra - discrete limit , we also obtain the corresponding solutions to the delay differential equation and the delay cellular automaton . to our knowledge , this is the first case of a delay cellular automaton with an exact solution .
the significance of the exact solutions presently obtained lies in that , as well as the shock formation of vehicular traffic , we have the velocity of a traffic jam propagating in an exact form from the dispersion relation .
the shock velocity implies that a quick response of drivers to the change of car distance does not always result in fast dissolution of a traffic jam .
we think that moderate response to the deceleration of cars in front will be effective to suppress disturbance of a homogeneous flow .
we note that in previous works @xcite , in contrast with the present study , they give exact results for the fundamental diagram , i.e. , the density - flow plot to investigate a phase transition from homogeneous to inhomogeneous density flow ; but however these results do not show the detailed form of a traffic jam .
the exact solutions can also prove , as discussed in @xcite , that nonlinear waves such as a shock and a solitary wave may appear in the range of parameters where the linear solution , which presents a uniform flow , exists stably .
this means that we can explore nonlinear waves beyond the linear stability theory .
a detailed discussion on this point will be found in our subsequent publications .
we finally add that the present work is motivated by the fact that the delayed ov model shares the same solitary - wave solution with the kac van moerbeke system , one of integrable lattice systems , and we have expected that the soliton theory including discretization and ultra - discretization of integrable systems applies well to the present model .
this work was supported in part by kakenhi grant number 26610033 .
99 schadschneider a , chowdhury d and nishinari k 2011 _ stochastic transport in complex systems : from molecules to vehicles _
elsevier amsterdam newell g f 1961 _ oper .
res . _ * 9 * 209 whitham g b 1990 _ proc .
a math . phys .
* 428 * 49 kac
m and van moerbeke p 1975 _ adv .
_ _ 16 _ 160 whitham g b 1974 _ linear and nonlinear waves _ john wiley & sons inc .
bando m , hasebe k , nakayama a , shibata a and sugiyama y 1995 _ phys .
e _ * 51 * 1035 bando m , hasebe k , nakanishi k , nakayama a , shibata a and sugiyama y 1995 _ j. phys .
i france _ * 5 * 1389 hasebe k , nakayama a and sugiyama y 1999 _ phys . lett .
a _ * 259 * 135 nakanishi k , itoh a , igarashi y and bando m 1997 _ phys . rev .
e _ * 55 * 6519 hirota r 2004 _ the direct method in soliton theory ( cambridge tracts in mathematics no 155 ) _ ed nagai a , nimmo j and c gilson tutiya y and kanai m 2007 _ j. phys .
_ * 76 * 083002 kanai m , isojima s , nishinari k and tokihiro t 2009 _ phys .
e _ * 79 * 056108 tsujimoto s and hirota r 1995 _ rims kkyroku _ * 933 * 105 igarashi y , itoh k and nakanishi k 1999 _ j. phys . soc .
jpn . _ * 68 * 791 tokihiro t , takahashi d , matsukidaira j and satsuma j 1996 _ phys .
* 29 * 3247 takahashi d and matsukidaira j 1997 _ j. phys . a : math .
gen . _ * 30 * l733 kanai m , nishinari k and tokihiro t 2006 _ j. phys . a : math
. gen . _ * 39 * 9071 kanai m 2007 _ j. phys . a : math .
_ * 40 * 7127 kanai m 2010 _ phys .
e _ * 82 * 066107 | we consider a car - following model described by a delay difference equation and give its exact solutions that present propagation of a traffic jam .
this model is a discrete - time version of the delayed optimal - velocity model ; in the continuum limit , we recover the delay differential equation for this model and the exact solutions as well .
we then work in the ultra - discrete limit , obtaining a delay cellular - automaton model , which successfully inherits the solutions .
also the dispersion relation for the present solutions suggests that a quick response of drivers does not always result in fast dissolution of a traffic jam .
_ keywords _ : delay difference equation , exact solution , ultra - discrete limit , cellular automaton , nonlinear waves , car - following model , traffic flow + |
recently , narrowband fluorescent nanodiamonds ( nds ) harnessing the photoluminescence of silicon vacancy ( siv ) centers have attracted research interest due to their promising applications as single photon sources as well as fluorescence labels for in vivo imaging.@xcite for the application as single photon sources , siv centers stand out owing to high brightness , narrow bandwidth single photon emission : single photon rates up to 6 mcps have been reported .
simultaneously , the fluorescence predominantly concentrates in the narrow purely electronic transition , i.e. , the zero - phonon - line ( zpl ) with down to 0.7 nm width at room temperature.@xcite for the application as fluorescence labels , siv centers are promising as they enable optical excitation using red laser light ( e.g. , at 671 nm ) combined with fluorescence in the near - infrared spectral range at approx . 738 nm.@xcite both properties aid to minimize absorption in biological tissue as well as tissue autofluorescence;@xcite additionally , the narrow bandwidth enables efficient spectral filtering of fluorescence and background signal . in ref .
, we reported the production of an aqueous , colloidal solution of siv containing nds from polycrystalline chemical vapor deposition ( cvd ) diamond films using the bead assisted sonic disintegration ( basd ) method ( see refs . ) .
these siv containing nds in solution are promising for applications as fluorescence labels as well as single photon sources .
however , some essential characteristics of the siv fluorescence in basd nds have not been investigated to date . in the context of single photon sources ,
the feasible positioning of nds in solution by nano - manipulation techniques is advantageous : it enables the coupling to photonic nano - structures , e.g. , dielectric resonator structures to build enhanced single photon sources.@xcite to reach efficient coupling of a defect center to a resonator mode , spectral overlap of the resonator mode and the color center s emission line has to be ensured.@xcite cooling of the color center generally reduces the emission linewidth and thus potentially enhances the spectral overlap of an emission line with a narrow resonator mode of a high q cavity .
however , the cooling also induces a spectral shift of the zpl.@xcite thus , knowledge of the zpl shift as well as the temperature dependent linewidth is crucial to ensure spectral overlap at the targeted temperature when coupling color centers to resonator structures . for siv centers in basd nds , neither temperature dependent linewidths nor lineshifts have been investigated to date .
second , the polarization of single photons emitted from a color center is crucial , e.g. , in quantum key distribution applications ( ref . ) or for the frequency conversion of single photons.@xcite at cryogenic temperature , the zpl of the siv center splits into 4 fine - structure components.@xcite however , the polarization properties of these line components have not been reported in the literature in contrast to the room temperature polarization properties.@xcite third , for single photon sources as well as fluorescence labels , spectrally and temporally stable fluorescence emission is essential .
therefore , stabilizing the desired charge state of the color center , corresponding to defined luminescence properties , is crucial . for nitrogen vacancy color centers in nds
, it has been shown that the charge state critically depends on the surface termination of the nds.@xcite nevertheless , no data on the dependence of siv luminescence in nds on the surface termination is reported in the literature . using heating in air ( oxidation ) ,
amorphous carbon and graphitic shells can be preferentially removed before the @xmath0 bonded diamond core of the nd is reduced.@xcite at 400430 @xmath1c , @xmath2 bonded carbon can be etched without loss of diamond even for diamond nanoparticles.@xcite simultaneously , the surface is covered with oxygen containing functional groups such as carboxyl and carbonyl groups.@xcite at higher temperatures , diamond can be etched away and the nds can be reduced in size : at 500 @xmath1c , hpht nds with about 60 nm size are etched at a rate of 1 nm / hour.@xcite however , the etch rate strongly depends on the crystallographic direction.@xcite furthermore , smaller diamonds get etched or even destroyed at lower temperature ( see ref . for work on detonation nds ) . in the present work ,
we use air oxidation to remove damaged regions from the surface and establish a defined surface termination of the basd nds .
we investigate the effects of the surface treatment via raman spectroscopy and we observe the reaction of the siv center luminescence to the same treatments .
we furthermore address the crucial topics introduced above : temperature dependent linewidth and position as well as the polarization of the zpl components .
the investigated nds have been described in detail in ref . .
they are obtained via the basd method from a polycrystalline cvd diamond film grown on a si substrate .
siv centers are created _ in situ _ as a result of the etching of the si substrate and subsequent incorporation of si into the growing diamond .
the nds have a size of 7080 nm . for the following measurements , we use samples where the colloidal solution has been dropcast and dried on an ir multilayer substrate.@xcite a raman spectrum of an ensemble of untreated nds
is depicted in curve 2 in fig .
[ figureraman](a ) .
raman spectroscopy confirms a high crystalline quality of the nds : we find a raman line at 1333.00 @xmath3 0.07 @xmath4 with a width of 4.9 @xmath3 0.4 @xmath4 .
the error is obtained as the standard deviation of the width and position measured on different spots on the dropcast sample . to obtain the width ,
a lorentzian fit has been used .
the observation of a raman line shifted only by 0.5 @xmath4 to higher frequency compared to the case of single crystal diamond witnesses a low absolute ( compressive ) stress in the diamonds ( see , e.g. , ref . )
whereas the broadening of the line indicates the stress distribution.@xcite however , the raman spectrum also reveals a signature of @xmath2 bonded carbon at around 1500 @xmath4.@xcite we note that a very similar non - diamond raman signature has also been present in the cvd diamond film used as a starting material for nd production.@xcite the @xmath2 bonded carbon is supposed to reside mainly on or below the surfaces of the diamonds partly as a result of surface damage during the milling procedure or as a residual of @xmath2 bonded carbon from the grain boundaries of the cvd film . as a consequence of the milling process
being carried out in water , the nds most probably initially reveal a termination with hydroxyl functional groups.@xcite however , acid cleaning has been subsequently employed to remove contaminants from the milling process possibly partially oxidizing the surfaces .
we now subject the sample to two oxidation steps : in a first oxidation step , the sample is heated to 460 @xmath1c in air and kept at that temperature for 90 mins [ raman spectrum : curve 4 in fig . [ figureraman](a ) ] . in a second oxidation step ,
we keep the sample at 480 @xmath1c for 40 mins [ raman spectrum : curve 5 in fig .
[ figureraman](a ) ] . after the heating steps
, the sample is cooled down inside the oven .
we additionally test the effect of a hydrogen plasma ( hot filament plasma , 33 mins , pressure @xmath512 mbar , distance to filament 5 - 6 mm , filament current 16 a , temperature ( sample holder ) 500 @xmath1c ) on untreated basd nds [ raman spectrum : curve 3 in fig .
[ figureraman](a ) ] .
all surface treatments , especially the heating in air to 480 @xmath1c reduce the signal related to @xmath2 carbon which is in accordance with the assumption that the @xmath2 carbon phase is located at the surfaces of the nds .
furthermore , the oxygen surface treatments narrow the diamond raman line in contrast to the hydrogen treatment : after the first oxidation step , we find a width of 4.0 @xmath3 0.2 @xmath4 ; following the second step it further reduces to 3.7 @xmath3 0.2 @xmath4 . as mentioned before
, the stress distribution in the diamond material can induce a broadening of the raman line.@xcite furthermore , diamond with a high defect density has a broader raman line.@xcite thus , the narrowing of the line suggests that the oxygen treatment is capable of removing highly stressed as well as highly defective regions of the milled nds .
the oxidation procedure thus constitutes an enhancement of the crystalline quality of the nds .
we note that further heating in air ( 500 @xmath1c , 2 hours ) , led to a significant reduction of the diamond quantity and no raman spectra could be obtained after that heating step . as a preliminary remark for the next section
, we note that luminescence of siv centers has been observed on the untreated as well hydrogen treated and oxidized dropcast samples .
c , curve 5 oxygen 480 @xmath1c . for comparison , the raman spectrum of a high pressure high temperature single crystalline diamond is given ( curve 1 ) .
the spectra have been recorded using 488 nm laser light , corrected for a linearly rising fluorescence background and shifted vertically for clarity .
( b ) detailed spectrum of the diamond raman line with lorentzian fit .
note that the calibration of the raman spectrometer has been done on the raman line of an hpht single crystal diamond set to 1332.5 @xmath4 .
[ figureraman ] ]
to enable the investigation of the fluorescence of individual nds , we spin coat the aqueous solution onto an ir multi - layer substrate.@xcite these substrates provide low background fluorescence for the luminescence investigation of single siv centers@xcite and highly efficient fluorescence collection.@xcite moreover , these substrates withstand the oxidation steps which we carry out . due to
randomly created spatial patterns of the spin coated diamonds , we reliably identify and relocate distinct nds on the untreated samples as well as after the oxidation steps in a homebuilt confocal microscope setup .
the employed confocal setup uses a microscope objective with a numerical aperture of 0.8 and the internal detection efficiency of the setup is estimated as 25% . for more details on the experimental setup
see , e.g. , ref . .
] the untreated nds mostly contain ensembles of siv centers : we observe no saturation of the fluorescence and intensity auto - correlation ( @xmath6 ) measurements do not indicate single photon emission ( antibunching ) .
an example of the typical luminescence spectrum of an ensemble of siv centers is displayed in fig .
[ figurespectra](a ) : the zpl of the siv center is clearly visible , here at 739 nm . for most nds , the zpl resembles an asymmetric tail towards longer wavelengths , indicating an inhomogeneous broadening of the siv center ensemble zpl with a preferential red shift.@xcite due to the asymmetry as well as the unknown fraction of homogenous and inhomogeneous broadening , we can only estimate the linewidth @xmath7 as well as the peak position @xmath8 .
we use a lorentzian fit to the data as a rough estimate .
we obtain linewidths ranging from 6.5 to 9.1 nm , with peak wavelengths between 738.8 and 739.2 nm for the investigated nds in accordance with our previous observations in ref .
[ see fig .
[ figureoxtrendg2](a ) ] . to estimate the brightness of the nds
, we measure the fluorescence intensity vs. excitation power for each nd . as a measure of the brightness
, we employ the slope of the linearly rising intensity vs. excitation power curve ( in cps/@xmath9w ) .
the results are displayed in fig .
[ figureoxtrendg2](a ) . using 671 nm excitation
, we obtain on average 3900 cps/@xmath9w without any indication of saturation or fluorescence instability up to roughly 0.5 mw .
thus , these nds containing siv ensembles are potential candidates for stable , narrowband fluorescence markers . , brightness and zpl peak wavelength @xmath8 for nds containing siv ensembles .
( b ) @xmath6 function of a single siv center with underlying ensemble emission after first oxidation step , ( c ) @xmath6 function of a single siv center with underlying ensemble emission after second oxidation step .
gray solid lines represent fits to the data using the @xmath6 function of a three level system convoluted with the instrument response of the measurement setup and corrected for background emission ( for details see e.g. ref . ) .
both @xmath6 measurements have been performed using 20 @xmath9w of excitation power .
[ figureoxtrendg2 ] ] we now subject the sample to the oxidation steps introduced in sec . [ sec_raman ] .
the zpl linewidth @xmath7 , the brightness as well as the zpl peak wavelength @xmath8 for 6 nds is summarized in fig .
[ figureoxtrendg2](a ) .
we observe a general trend , which is also discernible in fig .
[ figurespectra](a ) : the zpl linewidth is reduced , especially due to a narrowing of the line at the red wavelength edge , thus slightly diminishing the asymmetry of the line . due to this reduction of the asymmetry , we also obtain a shift of the peak wavelength when fitting a lorentzian line to the data .
the results are summarized in fig .
[ figureoxtrendg2](a ) .
the maximum shift of the peak position is 0.3 nm , the typical reduction of the linewidth is 0.51 nm . furthermore , changes in the sideband spectrum , including vanishing features in the sideband region , have been observed for several nds .
assuming that the absorption coefficient and the brightness of the siv centers as well as the collection and detection efficiency remain unchanged , the brightness of a nd before and after oxidation can serve as a measure of the number of siv centers contained in the nd .
we observe a trend to a reduced brightness of the nds after each oxidation step ; in particular , after the second oxidation step , the average brightness is reduced to 2400 cps/@xmath9w .
thus , one might conclude that the number of siv centers in the nds is reduced .
the zpl is narrowed during the oxidation steps , as probably predominantly red shifted siv centers in stressed regions close to the surface of the nds are removed .
the cleaning of the surface , however , does not seem to enhance the brightness of the remaining siv centers .
this observation is in contrast to the observations reported in ref .
were non - diamond carbon especially in the grain boundaries of polycrystalline diamond films has been found to quench siv luminescence .
thus , one might expect a significant enhancement of the fluorescence upon surface cleaning .
we emphasize that as we do not know how many centers have been removed , we can not conclusively determine the influence on the brightness of single siv centers , especially close to the surface .
in contrast to the spectra of siv ensembles discussed above , we also find nds where a significantly narrowed , bright line dominates over a broader emission .
an example of such a spectrum is given in fig .
[ figurespectra](b ) .
a lorentzian fit of the narrow line yields a peak position of 737.2 nm with a width of 1.5 nm [ see dashed line in fig .
[ figurespectra](b ) ] . additionally , a broader emission is present that is especially visible in the long wavelength tail of the spectrum [ marked with an arrow in fig .
[ figurespectra](b ) ] .
we interpret this spectrum as follows : the narrow bright line originates from the emission of a single , bright siv center , whereas the underlying broad emission originates from the inhomogeneously broadened emission of an ensemble of weakly fluorescing siv centers .
the observed linewidth of the single siv center is in accordance with previous observations for the linewidth of single siv centers ( 0.7 - 2.2 nm , ref . ) .
we note that in ref .
single siv centers with brightness differing by almost two orders of magnitude have been observed , thus indicating a large spread of single emitter brightness for siv centers , suggesting why a single emitter can dominate over a whole ensemble of centers . to prove this interpretation of the spectrum , we perform @xmath6 measurements after the first oxidation step .
the @xmath6 measurement shows an antibunching dip and is displayed in fig 3(b ) .
however , @xmath10 , thus only imperfect single photon emission is observed and one can not straightforwardly conclude that a single emitter is present . in the following ,
we discuss in detail that the measured @xmath6 function is in accordance with our interpretation of the spectrum consisting of a narrow single emitter line and an underlying ensemble emission .
the @xmath6 function indicates a three level internal population dynamics in accordance with previous observations on single siv centers.@xcite we fit the correlation measurements using the @xmath6 function of a three level system . to take into account the timing jitter of the detection setup , which leads to a deviation from @xmath11 , we convolute the @xmath6 function with the instrument response of our hanbury brown twiss setup ( timing jitter 354 ps , gaussian jitter @xmath12 width , for details see ref . ) .
furthermore , we include the underlying ensemble emission as an uncorrelated background following ref . .
the assumption of an uncorrelated emission from the ensemble is justified as for nds containing ensembles of siv centers , we only found featureless @xmath6 functions indicating uncorrelated emission .
we use the spectrum in fig .
[ figurespectra](b ) to estimate the ratio between the emission intensities in the single emitter line as well as in the ensemble .
we note that for the @xmath6 measurements a bandpass filter with a transmission window between 730 and 750 nm has been used , thus photons from the ensemble emission as well as the single emitter contribute to the measurement .
we find a probability of 63% that a detected photon has been emitted into the narrow line and thus stems from the single emitter .
we include this probability into the fit of the @xmath6 function to account for the background emission and obtain a very good agreement with the measured data [ see solid lines fig . [ figureoxtrendg2](b ) ] thus proving our interpretation that the main contribution to the emission , i.e. the narrow , bright emission line , stems from a single siv center .
after the second oxidation , a more pronounced antibunching dip with @xmath13 is observed [ see fig .
[ figureoxtrendg2](c ) ] . furthermore ,
following the second oxidation step , we clearly observe a saturation behavior as expected for a single emitter instead of a linearly rising fluorescence .
we use @xmath14 to fit the fluorescence intensity @xmath15 in dependence of the excitation power @xmath16 here , @xmath17 is the saturation power , where @xmath18 is the maximum emission rate of the single emitter and @xmath19 gives the brightness of uncorrelated background emission , here predominantly due to the underlying siv ensemble .
the fit yields @xmath20 @xmath9w , @xmath21 kcps , @xmath22 cps/@xmath9w . from this saturation measurement
, we estimate the probability that a photon has been emitted by the single , bright siv center in the @xmath6 measurement as 69% .
again , the fitted @xmath6 function well describes the measured data , proving that the bright , narrow zpl belongs to a single emitter .
thus , the underlying ensemble emission is reduced by the second oxidation step as also visible from the spectrum displayed in fig .
[ figurespectra](b ) , where a significant narrowing of the line at the red edge of the zpl is discernible .
thus , the observation coincides with the findings on the nds hosting siv center ensembles .
the fact that the narrow single emitter is not destroyed during the oxidation steps hints at an incorporation of this color center in a region of the nd which is not particularly close to the surface . summarizing , we have successfully employed oxidation in air as a method to reduce the influence of a weakly fluorescing ensemble of siv centers resulting in an increased single photon emission probability from a bright single siv center in a basd nd , thus significantly enhancing its performance as a single photon source .
] at room temperature , the zpl of the siv center is broadened , possibly masking an underlying fine - structure of the line . to gain further information on the line structure as well as on line shifting and broadening
, we thus investigate the luminescence at temperatures down to liquid helium temperature .
for the low temperature investigations , we select the previously discussed nd with dominant emission from a single siv center and weak underlying ensemble luminescence . at temperatures below 100 k ,
a fine - structure of the zpl of the single emitter evolves as displayed in fig .
[ figtempdep](a ) .
note that the low temperature characterization has been performed following the first oxidation step .
for the weakly fluorescing ensemble accompanying the single siv center , a complicated fine - structure evolves , red shifted to the single emitter line [ see e.g. the spectrum at 5 k in fig .
[ figtempdep](a ) , wavelength range 736 - 738 nm , or figs .
[ figpolexc](a ) and [ figpolem](a ) ] .
we interpret this complicated fine structure in terms of overlapping zpl fine - structure components of single emitters or small sub - ensembles under different stress conditions .
the work of clark et al .
@xcite and sternschulte et al .
@xcite as well as our previous work@xcite found a four line fine - structure for the zpl of an ensemble of siv centers in high quality single crystalline diamond .
the four lines arising from a split excited and ground state , arrange in two doublets as discernible from the spectrum in the inset of fig .
[ figtempdep](a ) measured on a thin high quality homoepitaxial cvd diamond film .
the smaller spacing has been experimentally found to represent the ground state splitting of approx .
50 ghz and the larger spacing corresponds to the excited state splitting of 240 ghz .
the emitter observed here shows a significantly altered and poorly resolved fine - structure fitted best using a dominant main line with two weaker peaks . to explain the observed fine structure one has to remark that the zpl fine - structure pattern has been observed to change significantly under stress:@xcite the variations include changes in the relative intensity possibly leading to the disappearance of line components as well as changes in the level splittings and positions .
changes in the fine - structure spectra have been observed for an siv ensemble under defined stress in [ 100 ] direction in ref . as well as for single centers in
cvd grown nds ( unknown stress state ) in ref . .
thus , also the strongly altered fine - structure of the single emitter in fig .
[ figtempdep](a ) might originate from an siv center experiencing an unknown stress state .
it should be noted that it is also possible that weak fine - structure components remained unidentified due to overlap with the underlying ensemble .
to evaluate the temperature dependent linewidth as well as the peak position of the zpl fine - structure components and to separate gaussian ( inhomogeneous ) and lorentzian ( homogeneous ) contributions to the linewidth , we fit the measured data with three voigt profiles ( for details of the procedure see ref . ) .
[ figtempdep](b ) summarizes the results for the main peak as only this peak is clearly visible over a broad temperature range [ see fig . [
figtempdep](a ) ] .
the fits reveal that at low temperature gaussian line broadening dominates ; the lines are broadened by spectral diffusion .
the term spectral diffusion denotes line broadening induced by the stark shift due to temporally fluctuating charges in the environment of the color center .
we find a spectral diffusion linewidth of 76(1 ) ghz ( 2.5 @xmath4 ) .
the spectral diffusion contribution is constant below approx .
50 k. in the following , we will discuss a possible source of the fluctuating charges leading to spectral diffusion .
finds that the main source of spectral diffusion are charges created as a result of the laser excitation of the diamond . under this assumption
, spectral diffusion is expected to be temperature independent which agrees well with our observation .
a major source for laser excited carriers is substitutional nitrogen which is a prevalent impurity in diamond and can be ionized by laser light with a photon energy exceeding 1.7 ev ( wavelength 729 nm).@xcite as a result , the low temperature linewidth for nitrogen vacancy centers has been found to depend strongly on the concentration of substitutional nitrogen in the host diamond : spectral diffusion linewidths for nitrogen vacancy centers in nitrogen rich type ib diamond can amount to 90 ghz .
@xcite the spectral diffusion linewidth observed here is thus comparable . in ref . , spectral diffusion linewidths for single siv centers in the range of 25 - 160 ghz have been reported .
we note that we can not estimate the nitrogen content of our diamond film , however , no measures have been taken to avoid nitrogen incorporation .
the homogeneous linewidth due to phonon dephasing@xcite increases with temperature @xmath23 and is best described using a @xmath24 law , in accordance with observations in cvd grown nds.@xcite the constant @xmath25 is fitted to be @xmath26 k@xmath27@xmath4 .
a @xmath28 law for the homogeneous line broadening has been attributed to phonon dephasing in a defect rich crystal.@xcite simultaneously , the main fine - structure component is red shifted with increasing temperature as shown in fig .
[ figtempdep](c ) . from liquid helium temperature to room temperature , we find an overall red shift of 25.7 @xmath4 in accordance with previous observations of 26 @xmath4.@xcite in ref .
, a @xmath29 dependence of the line shift is deduced as a result of interaction with lattice and local phonon modes and experimentally verified for nitrogen vacancy centers .
we adapt this model and yield a very good agreement with the measured data as discernible from fig .
[ figtempdep](c ) .
the fitted constants are @xmath30 k@xmath31@xmath4 , @xmath32 k@xmath33@xmath4 .
( @xmath34 k , excitation 0.18 mw ) . shifted to longer wavelength from the main line of the single siv center , the underlying ensemble is clearly visible ( wavelength range 736 - 740 nm ) .
( inset : enlarged detail of single emitter line ) , ( b ) summary of visibility and linear polarization direction corresponding to maximum absorption @xmath35for the single emitter lines [ figpolexc ] ] ( @xmath34 k , excitation 0.18 mw ) . shifted to longer wavelength from the main line of the single siv center , the underlying ensemble is clearly visible ( wavelength range 736 - 740 nm ) .
the excitation polarization is fixed to the optimal value .
( inset : enlarged detail of single emitter line ) , ( b ) summary of visibility and linear polarization direction @xmath36 for the single emitter lines [ figpolem ] ] we now analyze the polarization properties of the investigated single siv center as well as the underlying ensemble emission .
first , we determine the polarized absorption for the previously investigated nd with dominant emission from a single siv center via rotating the linear polarization of the excitation laser @xmath37 and recording spectra for the different polarizations .
the resulting spectra are displayed in fig .
[ figpolexc](a ) . via fitting the spectrum of the single siv center
, we obtain the areas @xmath38 of the fine - structure peaks in dependence of @xmath37 . fitting a sinusoidal function @xmath39 yields the visibility @xmath40
@xmath41 for each line as well as the direction of maximum absorption @xmath35 .
the results are shown in fig .
[ figpolexc](b ) .
the high visibility close to 100% indicates that the color center preferentially absorbs linearly polarized light .
furthermore , the direction of maximum absorption @xmath35 is very similar for all line components .
we thus tentatively conclude that the fine - structure components are excited via the same off - resonant pumping transition .
we note that , due to the different orientations of the centers , the polarization reaction of the ensemble ( see fig .
[ figpolexc](a ) , wavelength range 736 - 740 nm ) mostly averages out , we thus do not analyze the polarization behavior of the ensemble in detail .
second , we analyze the polarization properties of the emitted fluorescence via rotating a linear polarization analyzer in the detection path of the confocal setup .
@xmath37 is fixed to the optimum direction .
the resulting spectra are given in fig .
[ figpolem](a ) .
we evaluate the spectra analogous to the polarized absorption spectra .
we obtain again high visibility for all single emitter lines , thus the emission of the color center is linearly polarized on all fine - structure lines whereas the ensemble emission is only weakly polarized .
the direction of the linear polarization of the single emitter lines is equal within 4@xmath1 .
summarizing , the observed single siv center shows fully polarized absorption and emission which is advantageous for the application as single photon source as well as fluorescence marker : via choosing the matching excitation polarization , an efficient excitation is possible suppressing background fluorescence .
the emission of fully polarized single photons is highly desirable for applications in quantum key distribution and frequency conversion of single photons .
the observations are in accordance with previous findings at room temperature that also identified preferential absorption of linearly polarized light as well as polarized emission.@xcite to understand the origin of the polarized absorption and emission further investigations using more emitters as well as identification of the transitions are necessary .
summarizing , we have successfully employed oxidation in air at up to 480@xmath1c as a post - fabrication treatment to enhance the crystalline quality of basd nds : we find a narrowing of the diamond raman line from 4.9 to 3.7 @xmath4 whereas the raman signal due to non - diamond carbon is significantly reduced .
simultaneously , the linewidth of the siv ensemble fluorescence in individual nds is narrowed by up to 1 nm . for a single bright siv center with underlying ensemble fluorescence , the oxidation leads to an enhancement of the single photon emission probability . using temperature dependent luminescence measurements ( 5295 k ) , we investigate the fine - structure of the single center s zpl . at low temperature
, the line is broadened by spectral diffusion ( 76 ghz ) , whereas homogeneous broadening proportional to @xmath28 dominates at higher temperature .
simultaneously , the line exhibits an overall temperature dependent shift of 25.7 @xmath4 . the center s fine - structure components show a common linear polarization in emission and are excited via the absorption of linearly polarized light . as an outlook
, the production of nds from highly si doped diamond films could be harnessed to produce brighter nds as fluorescence labels , whereas the use of high purity material will be advantageous for single photon applications .
we thank m. fischer , s. gsell and m. schreck for supplying the ir substrates , p. elens and c. hepp for help with the fluorescence measurements and a. fuchs and h. schmitt for performing the hydrogen treatment of the diamonds .
we acknowledge funding by the bmbf ( ephquam 01bl0903 ) and the deutsche forschungsgemeinschaft ( for 1493 ) .
31ifxundefined [ 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 10.1063/1.3599608 [ * * , ( ) ] link:\doibase 10.1088/1367 - 2630/13/2/025012 [ * * , ( ) ] link:\doibase 10.1364/oe.20.019956 [ * * , ( ) ] @noop * * , ( ) link:\doibase { 10.1002/adma.200601452 } [ * * , ( ) ] link:\doibase { 10.1021/nn900339s } [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * ( ) @noop * * , ( ) link:\doibase 10.1103/physrevb.50.14554 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.205211 [ * * , ( ) ] link:\doibase { 10.1103/physrevb.82.115449 } [ * * , ( ) ] @noop * * , ( ) http://www.scopus.com/inward/record.url?eid=2-s2.0-80755184843&partnerid=40&md5=e340b1a416e778f243bd30aef7c31ce0 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) link:\doibase 10.1103/physrevb.51.16681 [ * * , ( ) ] @noop * * , ( ) \doibase doi : 10.1103/physreva.64.061802 [ * * , ( ) ] @noop ( ) link:\doibase 10.1103/physrevlett.110.027401 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | we report fluorescence investigations and raman spectroscopy on colloidal nanodiamonds ( nds ) obtained via bead assisted sonic disintegration ( basd ) of a polycrystalline chemical vapor deposition film .
the basd nds contain _ in situ _ created silicon vacancy ( siv ) centers .
whereas many nds exhibit emission from siv ensembles , we also identify nds featuring predominant emission from a single bright siv center .
we demonstrate oxidation of the nds in air as a tool to optimize the crystalline quality of the nds via removing damaged regions resulting in a reduced ensemble linewidth as well as single photon emission with increased purity .
we furthermore investigate the temperature dependent zero - phonon - line fine - structure of a bright single siv center as well as the polarization properties of its emission and absorption . |
markus bttiker was certainly one of the most influential scientists in the field of mesoscopic physics . among all his important contributions ,
time in quantum mechanics has a peculiar flavor since it occupied his mind at the right beginning and at the end of his carrier .
intrigued at first by the traversal time of an electron through a tunnel barrier @xcite , he came back to this topic after the emergence of `` on - demand single electron sources '' @xcite , which he greatly contributed to develop @xcite , via the concept of waiting time distribution ( wtd ) @xcite .
charge transport at the nanoscale is known to be stochastic due to the quantum nature of particles @xcite .
therefore , going beyond the knowledge of average quantities , such as the average electronic current , appears to be unavoidable and extremely fruitful at the same time .
a deep physical insight can indeed be inferred from the fluctuations of the signal and extracted from various observables .
noise @xcite and full counting statistics ( fcs ) @xcite , namely the second moment of current fluctuations and the statistics of charges transferred during a long time interval , are among the most popular quantities and have been proved to be powerful tools . with the development of electron quantum optics @xcite and the progress in single electron detection at high frequencies @xcite
, it is now relevant and possible to consider electron dynamics and time resolved quantities at quantum mechanical time scales ( typically nano - seconds and below ) .
therefore , new theoretical tools have been developed to describe the current fluctuations at such time scales , such as finite frequency noise @xcite and fcs @xcite , wigner functions @xcite , or the wtd @xcite .
the latter , describes the statistical distribution of time intervals between the detection of two electrons and therefore gives accurate information about correlations between subsequent electrons .
is set to an energy @xmath0 above the fermi energy of the normal part and the gap @xmath1 is much larger than the potential difference @xmath0 . ]
the wtd has been studied for particularly simple systems like single and multiple electronic quantum channels connected to two normal leads via a quantum point contact ( qpc ) @xcite , a quantum capacitor @xcite , a double quantum dot @xcite , a train of lorentzian pulses @xcite or a quantum dot connected to a normal and a superconducting lead @xcite , among others . in this paper
we revisit the physics of normal - superconducting ( ns ) junction through the point of view of waiting times in order to illustrate the effect of superconducting correlations and entanglement @xcite on their distribution . indeed , as we will discuss later , such a system may emit entangled electrons in the normal part , and leads to interesting features in the wtd . the paper is organized as follows . in sec .
[ sec : model ] , we describe the model used for the ns junction and the formalism needed for computing the wtd . in sec . [
sec : onespin ] , we discuss the effect of the transparency of the barrier ( the energy dependence of the andreev reflection ) when the detection process is sensitive to only one electronic spin species and a certain range of energy .
section [ sec : twospins ] is devoted to the effect of entanglement between spin up and spin down electrons emitted from the superconducting part , on the wtds .
we finally conclude and discuss some perspectives in sec .
[ sec : conclusion ] .
moreover , for the sake of clarity , technical details are moved to the appendices .
[ app : a ] demonstrates the formal analogy between our setup and a single quantum channel conductor for a specific detection process whereas important steps for the numerical and analytical calculations of the wtd in the entangled case are explained in [ app : b ] .
one very important consequence of superconductivity is the existence of andreev reflection .
such a phenomenon arises because the superconducting device can not accommodate any single particle excitation with energy below the gap @xmath1 .
therefore , if a single particle like an electron or a hole flows from the normal part to the superconducting part with an energy below this threshold it can only be scattered back at the interface . however , there are now two possibilities .
an electron ( a hole ) can be either normally reflected ( specular reflection ) , that is to say , reflected as an electron ( a hole ) or converted to a hole ( an electron ) .
this is the so called andreev reflection which originates from the fact that the incoming electron finds a partner to create a cooper pair which can enter in the superconductor and leave a hole behind . to be more specific , the system of interest is a polarized ns junction ( with an s - wave superconductor ) , at zero temperature , as presented on fig .
[ fig : fig1 ] .
the superconductor chemical potential @xmath2 is set to be at a potential @xmath0 above the fermi level @xmath3 of the normal metal .
in such a situation , there is an incident hole , coming from the metal , that can be either normally reflected or andreev reflected as an electron .
another way of picturing the andreev effect is to think about the inverse configuration where a cooper pair in the superconductor ( at energy @xmath2 and zero momentum for an s - wave superconductor ) splits at the interface and gives birth to an entangled pair of electrons . from now on
, we will take @xmath0 much smaller than the superconducting gap @xmath1 in order to focus on this sub - gap phenomenon .
this also has the benefit to make the andreev time @xmath4 ( the typical time needed for an andreev event ) much smaller than @xmath5 ( the typical time separation of two single particle wave packets emitted in the normal metal @xcite ) .
this allows us to assume that andreev events are instantaneous and make use of scattering theory .
in addition , this assumption allows one to linearize the dispersion relation around @xmath2 as @xmath6 , with @xmath7 and @xmath8 measured from @xmath2 and its corresponding momentum ( or around the fermi level since @xmath9 and @xmath2 ) . at the interface ,
the scattering is in general not perfect and both normal and andreev reflection will play a role . in order to describe this effect , we use the standard blonder - tinkham - klapwijk ( btk ) model @xcite which has been widely used in the literature .
the junction is modeled by a point - like barrier potential @xmath10 , where @xmath11 is the fermi wavelength and @xmath12 is a parameter measuring the strength of the barrier .
it is then possible to compute the scattering matrix of this setup exactly and obtain the normal and andreev transmission / reflection coefficients @xcite .
we do not reproduce these results in the present paper but give the corresponding numerical values of the coefficients when necessary .
figure [ fig : fig1 ] illustrates the scattering processes that we are now going to describe mathematically .
the incident holes of energies @xmath13 lying between @xmath3 and @xmath14 , arriving from the left and propagating to the right will be either normally reflected as holes of the same energies with amplitude @xmath15 or andreev reflected as electrons of energies @xmath16 with amplitude @xmath17 .
the incoming scattering state is therefore a slater determinant of holes of the form @xcite @xmath18 where @xmath19 stands for the filled fermi sea up to @xmath2 in the normal part .
however , in the electron language , this state is just the fermi sea @xmath20 filled up to @xmath3 instead of @xmath2 . in the following
, we will rather use the electronic picture to simplify the notation but both pictures are equivalent @xcite . due to scattering at the interface ,
the outgoing state is therefore a superposition of reflected holes , entangled electrons and non - entangled electrons @xcite @xmath21 indeed , it is pretty straightforward to see that the previous equation , for a given energy , gives birth to three kinds of term with different levels of complexity .
the terms @xmath22 and @xmath23 correspond to non - entangled contributions whereas @xmath24 describes fully entangled electrons originating from the splitting of a cooper pair at the interface .
when andreev reflection is absent ( @xmath25 ) , the fermi sea is unperturbed by the interface and nothing interesting happens .
counter - intuitively , perfect andreev reflection does not lead to perfect entanglement . on the contrary
, the state is a slater determinant of non - entangled electrons and the ns junction acts as a conventionnal electron source @xcite .
it appears that the maximally entangled situation arises when andreev and normal reflection probabilities are both equal to one half .
nevertheless , the wtd of a fully entangled state has never been studied to our knowledge and we will take the opportunity to study it in this paper before considering the general and more realistic state emitted at the interface . in order to conclude this section , we recall a few definitions about wtds . as mentioned in the introduction , the waiting time @xmath26 is defined as the time delay between the detection of two single particles . due to scattering and the quantum nature of particles ,
this time is a random variable , which distribution ( the wtd ) brings an elegant and instructive picture of the physics . for stationary systems , namely when there is no explicit time dependence , the wtd @xmath27 depends on @xmath26 only ( and not on absolute time ) and is closely related to the idle time probability ( itp ) @xmath28 , the probability to detect no electron during a time interval @xmath26 @xmath29 where @xmath30^{-1}_{\tau=0}$ ] is the mean waiting time @xcite . to go further
, we must now specify the detection procedure to compute the wtd .
in what follows we will assume perfect single electron projective measurement but will consider two different situations as described below .
such perfect detection process is still theoretical and extremely challenging at very short time scales ( nano - second and below ) but recent experiments @xcite are very promising about this issue . in order to compute the itp ,
we regularize the scattering problem as usual @xcite .
we discretize energy , ranging from @xmath31 to @xmath0 , in @xmath32 slices and wave vectors as @xmath33 and consider the limit @xmath34 to mimic a stationary process .
the first application of this setup will be to measure the wtd of holes via the detection of andreev reflected electrons or in other words the wtd of andreev events @xcite .
a single electron detector is located at a position @xmath35 far away from the interface . in this section
we assume that the detector measures only one spin orientation that we will choose upward for concreteness .
moreover , we make the additional assumption that it is only sensitive to energies above @xmath2 , using a quantum dot for instance @xcite . as a consequence ,
this allows us to avoid complications due to entanglement @xcite , which will be the subject of the next section .
following @xcite , the itp is defined as @xmath36 with the explicit condition that only electrons with energies above @xmath2 contribute . here
@xmath37 and @xmath38 stands for the normal ordering . in principle , the itp can not be reduced to a single determinant since the many body state is not a slater determinant . however , according to the assumptions introduced above , only one term survives and the final result can be cast as a single determinant @xcite . under these conditions ,
entanglement no longer plays a role and the problem thus boils down to a single quantum channel with energy dependent transmission @xcite , where the role of the energy - dependent transparency @xmath39 of the qpc is played here by @xmath40 , the andreev reflection amplitude ( see [ app : a ] for details ) .
is increased , @xmath41 decreases from one to zero as shown in the inset . for @xmath42 , @xmath43 and
the wtd is described by the wigner distribution ( blue dots , eq .
( [ eq : wigner ] ) ) .
close to perfect specular reflection ( @xmath44 ) the distribution approaches an exponential law except for very small times ( black squares ) . for intermediate @xmath12 ,
small oscillations with period @xmath45 are superimposed to an exponential decay .
together , with the dip at @xmath46 , they are manifestations of pauli s exclusion principle @xcite . ] in fig .
[ fig : fig2 ] , we have plotted the spin up electronic wtd for different barriers strengths @xmath42 ( perfect andreev reflection ) , @xmath47 and @xmath48 ( strong barrier ) as a function of @xmath49 . for information we show as an inset the corresponding andreev reflection coefficient @xmath50 as a function of energy
. however , the energy dependence is very weak since the energy window @xmath0 is supposed to be much smaller than the superconducting gap @xmath1 .
as expected , the wtd for @xmath42 is approximately given by the wigner surmise @xcite @xmath51\,.\ ] ] indeed , in that case the train of free holes is perfectly converted at the interface into free electrons which are described by random matrix theory @xcite .
as @xmath12 is increased , the distribution is broadened since @xmath41 is no longer equal to one and therefore not all holes are converted into electrons .
the situation is exactly equivalent to free electrons injected into a single quantum channel and partitioned by a quantum point contact with transmission probability @xmath41 due to the electron - hole symmetry in the system .
the difference is only conceptual since here the detector is measuring indirectly the statistics of holes converted into electrons by the ns junction that acts as an andreev mirror .
finally , for large @xmath12 , namely , small andreev reflection , most of the holes are normally reflected and the detector collects rare events which are almost uncorrelated and the wtd is exponential ( except for very short times ) .
indeed , following @xcite we derive the asymptotic behavior of the wtd in the long time limit .
for @xmath43 the decay is gaussian with algebraic corrections well described by the wigner surmise and for partial andreev reflection ( @xmath52 ) it is exponential with a rate given by the geometrical mean of the logarithm of @xmath53 over energy in @xmath54 $ ] , namely @xmath55\ , g(\tau/\overline\tau),\ ] ] where @xmath56 is an oscillatory function that decays as @xmath57 and depends on @xmath58 . to conclude this section
, we note that the role of electrons and holes may be interchanged by inverting the polarization ( @xmath59 ) due to electron - hole symmetry .
we now move a step forward and discuss the effect of entanglement on waiting times .
as mentioned before , andreev reflection might be thought in terms of splitting of a cooper pairs in the vicinity of the interface , leading to the injection of two entangled electrons with opposite spins and energy ( with respect to @xmath2 ) in the normal part as shown in fig .
[ fig : fig3 ] ( left part ) .
however , we have seen that the scattering state ( eq .
( [ eq : psiout ] ) ) is a mixture of entangled and non - entangled components which makes the effect of entanglement hard to separate from other physical ingredients .
therefore , we focus here on the wtd of the fully entangled state of the form @xmath60 \vert f \rangle\,,\ ] ] even if this is not the real quantum state emitted at the interface .
indeed , this work on ns junction rises a fundamental question of the effect of entanglement on waiting times that has never been considered to our knowledge and deserves to be investigated with proper care . at the end of this section
we will give a few hints on how the additional terms included in the full state ( eq .
( [ eq : psiout ] ) ) modify the picture but the common thread of this section will be the study of the fully entangled quantum state ( eq .
( [ eq : psife ] ) ) . to point out the specific features of entanglement
, we will compare this situation to the one of two independent electrons which would correspond to the emission of two electrons with opposite spins from a normal metal ( see fig .
[ fig : fig3 ] right ) @xcite .
this time , the single electron detector is sensitive to both spins and energies between @xmath61 and @xmath62 .
using the energy discretization mentioned before , the entangled state reads @xmath63 \ , \vert f \rangle\ , , \ ] ] whereas non - entangled electrons emitted from a normal lead in an energy window @xmath0 ( see fig . [
fig : fig3 ] right ) would be described by @xmath64 . according to the detection process mentioned above , the itp is formally given by @xcite @xmath65 where the quantum average is taken over the state @xmath66 or @xmath67 leading to @xmath68 and @xmath69 respectively . without any further assumption it is clear that the final result can not be expressed as a single determinant but rather as a sum of @xmath70 terms except in the non - entangled case where the itp factorizes to @xmath71 where the averages are taken over the two spin sectors separately @xcite . indeed ,
each term in the product of eq .
( [ eq : psie ] ) can be split into two parts : @xmath72 and @xmath73 .
@xmath74 is then a sum of @xmath75 terms of the form @xmath76 , where @xmath77 can be either @xmath78 or @xmath79 .
we shall denote a particular term by a string made of a succession of @xmath80 s and @xmath81 s defined as follows : @xmath82 .
this is a slater determinant , while @xmath83 can not generally be cast as a simple slater determinant .
@xmath84 will then be the sum of @xmath70 terms of the form , typically , @xmath85 , with @xmath86 and @xmath87 two generally different slater determinants . as previously shown in refs .
@xcite , each @xmath88 is also a determinant .
details about the procedure to calculate the itp are given in [ app : b ] .
we now discuss our results obtained from a direct and exact enumeration of the @xmath70 terms in the itp .
owing to the exponentially growing number of terms , this approach is limited to relatively small values of @xmath32 .
figure [ fig : fig4]a ) presents the wtd for increasing values of @xmath32 up to @xmath89 . as can be seen
, the curves reasonably converge to a limiting distribution that would be obtained for @xmath90 . to insure this ,
we have computed several finite size corrections that we will discuss later on .
figure [ fig : fig4]b ) compares the wtds of entangled and non - entangled electrons @xcite which are qualitatively similar .
however , as we will discuss in more detail in the next subsection , the maximum of the curve in the entangled case is more pronounced and closer to zero than in the independent case .
the presence of such a peak in the wtd is the hallmark of pair rigidity due to entanglement . to be more quantitative
we can evaluate the probability that the waiting time is smaller than the average waiting time @xmath91 ( which is the same in both situations ) .
we find that this probability is about ten percent larger in the entangled case , demonstrating that the entangled electrons are more correlated than the non - entangled ones .
the short time behavior is one of the most interesting characteristics of the wtd as it reflects the short time correlations encoded in the many body state and not the ones due to scattering .
for free electrons it is universal since it is the expression of the pauli s principle .
indeed , two spinless electrons can not be emitted in the same state which enforces the wtd to start from zero ( with a quadratic behavior ) . however ,
if the electrons have other degrees of freedom like spin or if the mesoscopic conductor supports several channels , the wtd may start from a non - zero value but correlations are still visible and universal @xcite .
however , these correlations only originate from electrons of the same channel and not between different channels .
they eventually disappear in the limit of large number of independent channels @xcite . .
the inset shows how @xmath92 converges to the asymptotic value @xmath93 as predicted by eq .
[ eq : stimewtd ] ( full black line ) .
b ) comparison between the wtd of entangled electrons and non - entangled electrons ( see text ) .
inset : short time behavior in the entangled case for @xmath89 ( full red line and green dashed line ( see eq .
[ eq : stimewtd ] ) ) and @xmath94 ( black long dashed lines ) . ] in order to get the short time expansion of the wtd up to second order in time , we need to expand the itp ( eq . ( [ eq : idle2 ] ) ) to fourth order in terms of moments of @xmath95 , @xmath96 and their products .
this is a straightforward but somehow cumbersome calculation that we do not reproduce in detail here .
after some algebra we obtain the short time expansion of the wtd , including finite size corrections .
the final expression for @xmath97 reads , to second order in @xmath26 @xmath98\ ] ] where @xmath99 in the entangled case and one in the absence of ns junction .
the insets of fig.[fig : fig4]a ) and fig.[fig : fig4]b ) show how this prediction is confirmed by our numerical evaluation of the wtd in the entangled case . in both situations , the wtd starts from the same initial value ( in the @xmath34 limit ) with a quadratic behavior but the coefficient is four times larger in the entangled case .
again , this reflects the pair rigidity due to entanglement .
differences between entangled and non - entangled electrons can also be inferred from the tail of the wtd . following the approach developed in @xcite we obtain the asymptotic behaviors of the wtd in the limit @xmath100 ( for details see [ app : b ] ) . in both cases
we find a gaussian decay @xmath101\,,\ ] ] with @xmath102 a constant which is equal to one in the entangled case and two in the non - entangled case .
therefore , the itp or equivalently the wtd decays more slowly in the entangled case , meaning that the two electrons are not at all independent . as mentionned before ,
entangled pairs are only one component of the real out - going scattering state which appears to be much more complex . for abritrary value of the reflection coefficients @xmath103 and @xmath41 , numerical evaluation of the wtd is pretty challenging since the number of terms grows as @xmath104 . in that case , direct numerical evaluations are restricted to very small values of @xmath32 ( typically 4 or 5 ) which is not sufficient to mimic a stationary situation .
it is then necessary to resort to more sophisticated approaches like monte - carlo sampling of the itp for instance .
however , there is a simple but rather interesting limit which is amenable to analytical calculations , namely the tunneling limit for @xmath44 @xmath105 . in that case
, the full many - body state simplifies to @xcite @xmath106|f\rangle\,,\ ] ] namely , to first non - trivial order in @xmath41 , it is made of scarce singlet pairs .
following the same approach as before we obtain the small time behavior @xmath107\ ] ] and the long time asymptotics @xmath108\,.\ ] ] at short waiting times , the distribution starts from a constant value and the physics is dominated by correlations within a single cooper pair .
indeed , simple considerations show that pairs are roughly separated in time by @xmath109 whereas electrons from the same pair are rather in a span of time of the order of @xmath110 .
the situation is then akin to a single pair which has been studied by hassler et al .
@xcite and confirms our predictions . at large time , the decay is exponential with a rate twice smaller than in the non - entangled case , which again is an hallmark of pair rigidity .
we have presented a theory of waiting times in polarized ns junction at zero temperature . in this setup
, andreev reflection brings new characteristic features in the wtd such as entanglement .
if a detector is sensitive to only one type of spin and to energies above the superconductor chemical potential , the situation is reminiscent of a single quantum channel connected to two normal leads via a qpc with energy - dependent transmission .
the interface acts as an andreev mirror and allows to measure the wtd of holes converted to electrons .
however , if the detector measures electrons with both spins in the whole energy window above the fermi sea of the normal conductor , entanglement between electrons leaves fingerprints in the wtd .
although still academical , we have focused on a fully entangled state which is only one component of the full many - body state flowing out of the ns interface . in that case ,
such signatures are visible for both small and large waiting times but the most important feature is the existence of a peak in the wtd centered before the average waiting time . when taking into account all components of the scattering state , we have shown that some charateristics are still visible in the wtd in the tunneling limit . in the near future , it would then be useful to extend this work beyond the ideal situation and the tunneling limit and evaluate the effect of entanglement for arbitrary values of reflection coefficients .
in addition , a natural extensions would be to study correlations between waiting times of different spin species in the spirit of @xcite .
this would probably yield an even clearer signature of entanglement than the wtd itself . among other future investigations
, it would be possible to study the effect of cross - andreev reflection in a superconducting - normal - superconducting junction or the physics brought by exotic states like majorana modes created by majorana guns @xcite .
the paper is dedicated to the memory of markus bttiker who was very enthusiastic about the quantum theory of waiting times .
m. a. , will always be grateful to him for his kindness , trust and freedom he gave to him when m. a. was a post - doc in his group .
markus was a mentor to him and a source of endless inspiration .
we thank d. dasenbrook , c. flindt , g. haack , p. p. hofer and m. moskalets for useful discussions and remarks .
the research of d. c. was supported by the foundation for fundamental research on matter ( fom ) , the netherlands organization for scientific research ( nwo / ocw ) , and an erc synergy grant .
this appendix is dedicated to prove the formal correspondence between the wtd of a single channel normal conductor and the andreev mirror defined in section [ sec : onespin ] .
this happens because the detection process is only sensitive to spin up electrons and energies above @xmath2 , the superconducting chemical potential . to begin the proof we recall the definition of the itp in terms of the scattering state @xmath111 with @xmath112 where @xmath32 corresponds to the number of slices due to the energy discretization between @xmath2 and @xmath62 ( which has to be taken equal to infinity at the end of the calculation ) and @xmath8 is measured with respect to @xmath113 the fermi momentum at energy @xmath2
there are two important steps in the derivation .
the first one consists in expanding the products and rearranging the terms in order make the relevant terms appear ( the ones with spin up and positive energy electrons ) .
if we restrict ourselves to these terms we build terms like @xmath114 .
then , we include the other terms of the product and show that they are irrelevant . indeed ,
in the following example @xmath115 the two off - diagonal terms vanish since the operator @xmath116 conserves the number of particles .
concerning the two diagonal ones , they leave us with @xmath117 then , we can repeat the same argumentation for all the other irrelevant terms from @xmath118 to @xmath32 .
finally , we are left with @xmath119 with @xmath120 . as a second step ,
we rearrange the fock space by moving all the spin @xmath121 to the left .
spin @xmath122 are invariant under @xmath123 , thus we only have to compute @xmath124 we then recognize the expression for the itp of a spinless quantum channel with bias @xmath0 and transmission amplitude @xmath17 @xcite .
in this appendix we give some details on the construction on @xmath28 .
starting from the n - body states reads @xmath125 \
, \vert f \rangle,\ ] ] with @xmath126 is the fermi sea defined in the main text and @xmath127 the probability to detect nothing in a range of time @xmath26 with a detector sensitive to all the energies and all spins can be written as @xmath128 where the average is taken over @xmath129 . from here , it is convenient to separate eq .
( [ wave_function ] ) in two parts @xmath130 and @xmath131 in order to get a sum of @xmath75 terms where each term is a slater determinant . to simplify ,
it is useful to change the notation : now each configuration is associated to a specific ket ( bra ) and each element of this configuration can be mapped to an ising classical spin .
namely , @xmath130 ( @xmath131 ) is associated to @xmath132 .
( [ pitau ] ) becomes @xcite .
@xmath133 where the summation runs over the @xmath70 configurations for the bra / ket ( one example of possible configuration is @xmath134 for a system consisting of three electron pairs )
. each element of the matrix m can be written as @xcite @xmath135 with @xmath136 and @xmath137 .
we now explain how to compute the long time asymptotics of the itp of the fully entangled state . in that case
, the matrix reduces to usual toeplitz matrix which intervenes in the statistics of levels in the gaussian unitary ensemble ensemble of random matrices @xcite .
namely , @xmath138 with @xmath139 and @xmath140 ( except for an unimportant factor @xmath141 in the angle @xmath142 ) .
the asymptotic behavior of the wtd for large @xmath26 is equivalent to the statistics of having no level in a large energy range . using this correspondence
, the main terms read , for large @xmath143 , with @xmath144 a rather complicated constant which can be find in reference @xcite .
the decay is gaussian with algebraic corrections .
00 m. bttiker and r. landauer , `` traversal time for tunneling '' , phys .
rev . lett .
* 49 * ( 1982 ) 1739 .
m. bttiker , `` larmor precession and the traversal time for tunneling '' , phys .
b * 27 * ( 1983 ) 6178 .
g. fve , a. mah , j .-
m . berroir , t. kontos , b. plaais , d. c. glattli , a. cavanna , b. etienne , and y. jin , `` an on - demand coherent single - electron source '' , science * 316 * , ( 2007 ) 1169 .
a. mah , f. d. parmentier , e. bocquillon , j .-
berroir , d. c. glattli , t. kontos , b. plaais , g. fve , a. cavanna and y. jin , `` current correlations of an on - demand single - electron emitter '' , phys .
b * 82 * ( 2010 ) 201309 .
j. dubois , t. jullien , f. portier , p. roche , a. cavanna , y. jin , w. wegscheider , p. roulleau , and d. c. glattli , `` minimal - excitation states for electron quantum optics using levitons '' , nature * 502 * ( 2013 ) 659 .
m. d. blumenthal , b. kaestner , l. li , s. giblin , t. j. b. m. janssen , m. pepper , d. anderson , g. jones , and d. a. ritchie , `` gigahertz quantized charge pumping '' , nat .
phys . * 3 * ( 2007 ) 343 .
j. p. pekola , j. j. vartiainen , m. mttnen , o .- p .
saira , m. meschke , and d. v. averin , `` hybrid single - electron transistor as a source of quantized electric current '' , nat . phys .
* 4 * ( 2008 ) 120 .
f. giazotto , p. spathis , s. roddaro , s. biswas , f. taddei , m. governale , and l. sorba , `` a josephson quantum electron pump '' , nat . phys .
* 7 * ( 2011 ) 857 . c. leicht , p. mirovsky , b. kaestner , f. hohls , v. kashcheyevs , e. v. kurganova , u. zeitler , t. weimann , k. pierz , and h. w. schumacher , `` generation of energy selective excitations in quantum hall edge states '' , semicond .
* 26 * ( 2011 ) 055010 .
s. hermelin , s. takada , m. yamamoto , s. tarucha , a. d. wieck , l. saminadayar , c. buerle , and t. meunier , `` electrons surfing on a sound wave as a platform for quantum optics with flying electrons '' , nature ( london ) * 477 * ( 2011 ) 435 .
a. fujiwara , k. nishiguchi , y. ono , `` nanoampere charge pump by single - electron ratchet using silicon nanowire metal - oxide - semiconductor field - effect transistor '' , applied physics letters .
* 92 * ( 2008 ) 042102 .
lansbergen , y. ono , a. fujiwara , `` donor - based single electron pumps with tunable donor binding energy '' , nano lett .
* 12 * ( 2012 ) 763768 .
tettamanzi , r. wacquez , s. rogge , `` charge pumping through a single donor atom '' , new journal of physics .
* 16 * ( 2014 ) 063036 .
m. bttiker , `` dynamic conductance and the scattering matrix of small conductors '' , phys .
* 70 * ( 1993 ) 4114 .
a. prtre , h. thomas and m. bttiker , `` dynamic admittance of mesoscopic conductors : discrete - potential model '' , phys .
b * 54 * ( 1996 ) 8130 . ya .
m. blanter and m. bttiker , `` shot noise in mesoscopic conductors '' , phys . rep .
* 336 * ( 2000 ) 1 .
m. moskalets and m. bttiker , `` floquet scattering theory of quantum pumps '' , phys .
b * 66 * ( 2002 ) 205320 .
m. moskalets and m. bttiker , `` time - resolved noise of adiabatic quantum pumps '' , phys .
b. * 75 * ( 2007 ) 035315 .
m. moskalets , p. samuelsson and m. bttiker , `` quantized dynamics of a coherent capacitor '' , phys .
* 100 * ( 2008 ) 086601 .
m. albert , c. flindt , and m. bttiker , `` acuracy of the quantum capacitor as a single - electron source '' , phys .
b * 82 * ( 2010 ) 041407
. f. d. parmentier , e. bocquillon , j .-
berroir , d. c. glattli , b. plaais , g. fve , m. albert , c. flindt , m. bttiker , current noise spectrum of a single - particle emitter : theory and experiment , phys . rev .
b * 85 * ( 2012 ) 165438 .
g. haack , m. moskalets and m. bttiker , `` glauber coherence of single electron sources '' , phys .
b * 87 * ( 2013 ) 201302(r ) .
m. moskalets , g. haack and m. bttiker , `` single - electron source : adiabatic versus non - adiabatic emission '' , phys .
b * 87 * ( 2013 ) 125429 .
m. albert , c. flindt , and m. bttiker , `` distributions of waiting times of dynamic single - electron emitters '' , phys .
lett . * 107 * ( 2011 ) 086805 .
m. albert , g. haack , c. flindt and m. bttiker , `` electron waiting times in mesoscopic conductors '' , phys .
* 108 * ( 2012 ) 186806 .
d. dasenbrook , c. flindt , and m. bttiker , `` floquet theory of electron waiting times in quantum - coherent conductors '' , phys .
* 112 * ( 2014 ) 146801 .
l. s. levitov and g. b. lesovik , `` charge distribution in quantum shot noise '' , jetp lett .
* 58 * ( 1993 ) 230 .
l. s. levitov , h. lee , and g. b. lesovik , `` electron counting statistics and coherent states of electric current '' , j. math .
* 37 * ( 1996 ) 4845 .
`` quantum noise in mesoscopic physics '' , nato science series , edited by yu .
v. nazarov ( kluwer , dordrecht , 2003 ) .
e. bocquillon , v. freulon , f. d. parmentier , j .-
berroir , b. plaais , c. wahl , j. rech , t. jonckheere , t. martin , c. grenier , d. ferraro , p. degiovanni and g. fve , `` electron quantum optics in ballistic chiral conductors '' , annalen der physics * 526 * ( 2013 ) 1 . j. basset , h. bouchiat , and r. deblock , `` high frequency quantum admittance and noise measurement with an on - chip resonant circuit '' , phys .
b * 85 * ( 2012 ) 085435 .
r. thalineau , a. d. wieck , c. buerle , and t. meunier , `` using a two - electron spin qubit to detect electrons fying above the fermi sea '' , arxiv cond / mat 1403.7770 ( 2014 ) . k. thibault , j. gabelli , c. lupien , and b. reulet , `` pauli - heisenberg blockade of electron quantum transport '' , phys .
* 114 * ( 2015 ) 236604 .
j. salo , f. w. j. hekking and j. p. pekola , `` frequency - dependent current correlation functions from scattering theory '' , phys .
b * 74 * ( 2006 ) 125427 . a. bednorz and w. belzig , `` models of mesoscopic time - resolved current detection '' , phys .
b * 81 * ( 2010 ) 125112 .
r. zamoum , a. crpieux , and i. safi , `` one channel conductor coupled to a quantum of resistance : exact ac conductance and finite frequency noise '' , phys .
b * 85 * ( 2012 ) 125421 . c. emary , d. marcos , r. aguado , and t. brandes , `` frequency - dependent counting statistics in interacting nanoscale conductors '' , phys .
b * 76 * ( 2007 ) 161404 . c. flindt , t. novotn , a. braggio , m. sassetti , and a. p. jauho , `` counting statistic of non - markovian quantum stochastic processes '' , phys . rev .
* 100 * ( 2008 ) 150601 .
f. hassler , m. v. suslov , g. m. graf , m. v. lebedev , g. b. lesovik , and g. blatter , `` wave - packet formalism of full counting statistics '' , phys .
b * 78 * ( 2008 ) 165330 , g. b. lesovik and i. a. sadovskyy , phys .- usp . * 54 * ( 2011 ) 1007 .
d. marcos , c. emary , t. brandes , and r. aguado , `` finite - frequency counting statistics of electron transport : markovian theory '' , new j. phys .
* 12 * ( 2010 ) 123009 .
d. marcos , c. emary , t. brandes , and r. aguado , `` non - markovian effects in the quantum noise of interacting nanostructures '' , phys .
b * 83 * ( 2011 ) 125426
. n. ubbelohde _
et al . _ ,
`` measurement of finite - frequency current statistics in a single - electron transistor '' , nat . commun .
* 3 * ( 2012 ) 612 .
d. ferraro , a. feller , a. ghibaudo , e. thibierge , e. bocquillon , g. fve , c. grenier , and p. degiovanni , `` wigner function approach to single electron coherence in quantum hall edge channels '' , phys .
b * 88 * ( 2013 ) 205303 .
s. welack , m. esposito , u. harbola , and s. mukamel , phys .
b * 77 * , `` interference effects in the counting statistics of electron transfers through a double quantum dot '' ( 2008 ) 195315 , s. welack , s. mukamel , and y. yan , `` waiting time distributions of electron transfers through quantum dot aharonov - bohm interferometers '' , europhys
* 85 * ( 2009 ) 57008 .
t. brandes , `` waiting times and noise in single particle transport '' , ann .
( berlin ) * 17 * ( 2008 ) 477 .
k. h. thomas and c. flindt , `` electron waiting times in non - markovian quantum transport '' , phys .
b * 87 * ( 2013 ) 121405(r ) .
m. albert and p. devillard , `` waiting time distribution for trains of quantized electron pulses '' , phys .
b * 90 * ( 2014 ) 035431 .
l. rajabi , c. pltl and m. governale , `` waiting time distributions for the transport through a quantum - dot tunnel coupled to one normal and one superconducting lead '' , phys .
* 111 * ( 2013 ) 067002 .
g. haack , m. albert , c. flindt , `` distributions of electron waiting times in quantum - coherent conductors '' , phys . rev .
b * 90 * ( 2014 ) 205429 .
tang , f. xu , j. wang , `` waiting time distribution of quantum electronic transport in the transient regime '' , phys .
b * 89 * ( 2014 ) 205310 .
b. sothmann , `` electronic waiting time distribution of a quantum - dot spin valve '' , phys .
b * 90 * ( 2014 ) 155315 .
d. dasenbrook , p. p. hofer and c. flindt , `` electron waiting times in coherent conductors are correlated '' , phys .
b * 91 * ( 2015 ) 195420 .
p. p. hofer , d. dasenbrook and c. flindt , `` electron waiting times for the mesoscopic capacitor '' , ( 2015 ) arxiv:1506.02801 .
a. braggio , m. governale , m. g. palac , and j knig , `` superconducting proximity effect in interacting quantum dots revealed by shot noise '' , solid state comm .
* 151 * ( 2011 ) 155 .
g. b. lesovik , t. martin , and g. blatter , `` electronic entanglement in the vicinity of a superconductor '' , eur .
j. b * 24 * ( 2001 ) 287 .
e. prada and f. sols , `` entangled electron current through finite size normal - superconductor tunneling structures '' , eur
j. b * 40 * ( 2004 ) 379 .
p. samuelsson , e. v. shukhorukov , and m. bttiker , `` orbital entanglement and violation of bell inequalities in mesoscopic conductors '' , phys .
* 91 * ( 2003 ) 157002 .
p. samuelsson , e. v. shukhorukov , and m. bttiker , `` quasi - particle entanglement : redefinition of the vacuum and reduced density matrix approach '' , new journal of physics * 7 * ( 2005 ) 176 . z. y. zheng , l. zhou , j. hong , and b. li , `` local and nonlocal entanglement for quasiparticle pairs induced by andreev reflection '' , phys .
b * 74 * ( 2006 ) 085312 .
b. braunecker , p. burset , and a. levy yeyati , `` entanglement detection from conductance measurements in carbon nanotube cooper pair splitters '' , phys .
* 111 * ( 2013 ) 136806 .
t. martin and r. landauer , `` wave - packet approach to noise in multichannel mesoscopic systems '' , phys .
b. * 45 * ( 1992 ) 1742 .
g. e. blonder , m. tinkham , and t. m. klapwijk , `` transition from metallic to tunneling regimes in superconducting microconstrictions : excess current , charge imbalance , and supercurrent conversion '' , phys .
b * 25 * ( 1982 ) 4515 .
s. duhot , `` transport lectronique non local dans des structures hybrides supraconductrices '' , phd thesis ( 2008 ) https://tel.archives-ouvertes.fr/tel-00343716 . v. f. maisi , d. kambly , c. flindt , and j. p. pekola , `` full counting statistics of andreev tunneling '' , phys .
lett . * 112 * ( 2014 ) 036801 .
p. recher and d. loss , `` superconductor coupled to two luttinger liquids as an entangler for electron spins '' , phys .
b * 65 * ( 2002 ) 165327 .
b. sutherland , `` quantum many body problem in one - dimension : ground state '' , j. math .
* 12 * ( 1971 ) 246 .
b. tarasinski , d. chevallier , j. a. hutasoit and c. w. j. beenakker , `` quench dynamics of fermion - parity switches in a josephson junction '' , arxiv:1503.04207 ( 2015 ) . n. v. gnezdilov , b. van heck , m. diez , j. a. hutasoit and c. w. j. beenakker , `` topologically protected charge transfer along the edge of a chiral @xmath145-wave superconductor '' , arxiv:1505.06744 ( 2015 ) .
m. l. metha , _ random matrices _ ( elsevier , amsterdam , 2004 )
. i. v. krasovsky , `` gap probability in the spectrum of random matrices and asymptotics of polynomials orthogonal on an arc of the unit circle '' , int .
res . not .
( 2004 ) 1249 . | we consider a normal - superconducting junction in order to investigate the effect of new physical ingredients on waiting times .
first , we study the interplay between andreev and specular scattering at the interface on the distribution of waiting times of electrons or holes separately . in that case the distribution is not altered dramatically compared to the case of a single quantum channel with a quantum point contact since the interface acts as an andreev mirror for holes .
we then consider a fully entangled state originating from spliting of cooper pairs at the interface and demonstrate a significant enhancement of the probability to detect two consecutive electrons in a short time interval .
finally , we discuss the electronic waiting time distribution in the more realistic situation of partial entanglement .
normal - superconducting junction , andreev reflection , waiting time distribution , entanglement 02.50.ey , 72.70.+m , 73.23.hk |
in @xcite , witten studied @xmath2 heterotic string theory on @xmath3 with an @xmath4 , @xmath5 gauge instanton supported on @xmath6 with instanton number @xmath7 .
the effective @xmath8 theory manifests @xmath9 supersymmetry and has gauge symmetry @xmath10 .
the boundary of instanton moduli space contains singular points , at each of which one unit of instanton number is concentrated .
these are called `` small '' instantons and the local moduli space @xmath11 was shown to be @xmath12 dimensional . it was demonstrated in @xcite that @xmath11 can be described as the flat directions of a potential energy function .
this is constructed from @xmath13 hypermultiplets , each transforming as a fundamental @xmath13 representation of @xmath4 and as a doublet under an additional @xmath14 gauge symmetry . at the origin of field space ,
all of these fields become massless and the @xmath14 symmetry becomes exact .
away from this point , however , the @xmath14 symmetry is spontaneously broken and only the @xmath12 moduli of @xmath15 remain massless .
one concludes that as one approaches a small instanton point in moduli space , the theory develops extra ( massless ) species and an enhanced symmetry .
such points are often referred to as being @xmath16
. small instantons also occur in @xmath1-theory on @xmath17 , where @xmath18 is a calabi - yau threefold @xcite .
as first discussed in @xcite , to be consistent these vacua must have , supported on the calabi - yau manifold at each end of the @xmath0 interval , a gauge instanton indexed by a subgroup of @xmath19 .
one can , for simplicity , assume a trivial instanton at one end of the interval .
the non - trivial instanton on the other end is a connection on a holomorphic vector bundle @xmath20 over @xmath18 with structure group @xmath21 .
quantum consistency requires that the instanton number , here generalized to the 2nd chern class @xmath22 , must satisfy the constraint @xmath23 , where @xmath24 is the tangent bundle of the calabi - yau threefold and @xmath25 is an effective class corresponding to m - five - branes in the @xmath26 `` bulk '' space .
the effective @xmath27 theory manifests @xmath28 supersymmetry .
its gauge symmetry depends on the choice of structure group @xmath29 . in a series of publications @xcite ,
holomorphic vector bundles with @xmath30 were constructed .
we will refer to these when specificity is required .
for example , when @xmath31 , @xmath32 and @xmath33 , the effective @xmath27 gauge symmetry is @xmath34 , @xmath35 and @xmath33 respectively .
the boundary of instanton moduli space contains singular points where a portion of the vector bundle becomes concentrated as a torsion free sheaf over an isolated holomorphic curve with effective class @xmath36 .
these are the heterotic @xmath1-theory analogs of small instantons . here , however , the local moduli spaces @xmath11 are more complicated , changing with the structure of the calabi - yau threefold , the holomorphic vector bundle and the choice of @xmath37 . the moduli spaces of small instantons on elliptic calabi - yau threefolds over hirzebruch bases @xmath38 with vector bundles constructed from spectral covers were analyzed in several publications @xcite .
for example , consider @xmath39 , @xmath31 and @xmath40 , with @xmath41 a section of @xmath42 .
then , it was shown that @xmath43 , where @xmath44 are integers satisfying the constraints @xmath45 and @xmath46 .
it is expected that any @xmath11 can again be described as the flat directions of a potential energy function .
this would be constructed from chiral supermultiplets transforming in some representation @xmath47 of @xmath29 and , possibly , as a representation @xmath48 of an additional gauge symmetry @xmath49 . at the origin of field space ,
all of these fields become massless and the @xmath49 symmetry becomes exact .
away from this point , the @xmath49 symmetry is spontaneously broken and only the moduli of @xmath15 remain massless . therefore , as one approaches a small instanton in moduli space , the theory develops extra ( massless ) species and , possibly , an enhanced symmetry .
that is , small instantons in heterotic @xmath1-theory are @xmath16 .
unfortunately , such potential functions have not been explicitly constructed for most moduli spaces .
limited simple examples indicate that such potentials exist , but a general proof is still lacking .
be that as it may , in this paper we will assume that small instanton moduli spaces @xmath15 in heterotic @xmath1-theory can be so described . in heterotic @xmath1-theory , strings wrapping a holomorphic curve in the calabi - yau threefold will generate a non - perturbative superpotential for both geometric and vector bundle moduli @xcite . as first discussed in @xcite
, the superpotential created by a such a `` worldsheet instanton '' has a specific form ; namely , an exponential of geometric and curve moduli multiplied by a `` pfaffian '' .
the pfaffian is a function of the determinant of the dirac operator twisted by the vector bundle restricted to the curve .
it is generically true @xcite that the pfaffian evaluated at any holomorphic curve is a homogeneous polynomial of the moduli of the restricted bundle .
this has important implications for small instantons .
it was shown in @xcite that the moduli of the vector bundle restricted to the small instanton curve @xmath36 are precisely those which parameterize @xmath11 .
hence , strings wrapped on @xmath36 generate a superpotential on @xmath11 which `` lifts''some , or all , of the flat directions .
furthermore , new local minima may now occur substantially away from the @xmath16 at the zero of field space . within the context of @xmath38 , @xmath31 and @xmath50 ,
pfaffians were explicitly computed in @xcite .
for example , for @xmath51 , @xmath52 and @xmath53 , @xmath54 and the pfaffian was shown to be an explicit homogeneous polynomial of degree @xmath55 in @xmath56 of the @xmath57 moduli of @xmath11 . at a small instanton in heterotic @xmath1-theory
, the vector bundle @xmath20 degenerates into the `` union '' of a new holomorphic vector bundle @xmath58 over @xmath18 and a sheaf over the curve @xmath36 . because of the existence of the @xmath26 interval
, the sheaf can `` detach '' from @xmath18 , leaving only @xmath58 behind , and become a component of a new m - five - brane curve in the bulk space .
that is , @xmath59 .
the physical interpretation of this is that the small instanton `` emits '' a five - brane into the bulk space . of course
, this can occur in reverse , a component @xmath59 of a bulk space five - brane being `` absorbed '' into @xmath58 via a small instanton to become @xmath20 .
such a process , either absorbing or emitting a bulk five - brane , is called a small instanton phase transition and was defined and studied in @xcite .
it was shown that there are two types of small instanton transitions , the first in which the structure group of the vector bundle is unchanged while the number of chiral zero modes is altered and a second where the reverse is true , the structure group changes while the number of chiral zero modes remains fixed . for an elliptically fibered calabi - yau threefold @xmath18 ,
the first type of transition occurs when @xmath36 is the pull - back of a curve in the base @xmath60 .
this is analogous to the heterotic @xmath2 case .
the second type of transition takes place when @xmath36 is proportional to the elliptic fiber @xmath61 . in this paper
, we will consider only those small instanton transition for which the structure group is unchanged .
an important aspect of any small instanton transition is that the @xmath1-five - brane wrapped on @xmath36 in the bulk space itself has moduli .
the most conspicuous of these is the translation modulus , the real part of which describes the separation of the five - brane from the end of the @xmath62 interval and whose imaginary part is an axion .
for a calabi - yau threefold @xmath18 elliptically fibered over any base @xmath60 , the complete moduli space @xmath63 of an @xmath1-five - brane wrapped on @xmath36 , both its dimension and geometry , was computed in @xcite .
this space depends strongly on the curve @xmath36 and typically contains many moduli in addition to the translation modulus . combining this with the discussion above
, we see that in heterotic @xmath1-theory the full moduli space of a small instanton phase transition is @xmath64 , where @xmath11 is the instanton phase and @xmath63 the five - brane phase . generalizing the above discussion
, we assume that the complete phase transition can be described by a potential energy function constructed from chiral supermultiplets in some representation @xmath47 of @xmath29 and , possibly , in a representation @xmath48 of an additional gauge symmetry @xmath49 . the origin of field space is an @xmath16 . away from this point , however , the @xmath49 symmetry is spontaneously broken and only the moduli parameterizing @xmath65 remain massless .
the existence of small instanton phase transitions , from the point of view of holomorphic vector bundles and topology , was demonstrated in @xcite . in the present paper , by studying the motion of heterotic @xmath1-theory vacua in moduli space coupled to an expanding friedmann - robinson - walker ( frw ) spacetime , we will show that these phase transition can also take place dynamically .
specifically , we will show that an @xmath1-five - brane in the bulk can dynamically evolve through the esp , the point in moduli space where the five - brane collides with the end of the @xmath66 interval , continuing until the vacuum comes to rest at a local minimum of the non - perturbative superpotential . at this point , which is substantially distant from the esp ,
the sheaf over @xmath36 has combined with @xmath58 to form a smooth vector bundle @xmath20 , completing the small instanton phase transition .
the dynamical system is trapped at this local minimum by the hubble friction of the frw space .
we ignore the effect of moduli trapping due to radiation into massless fields at the esp @xcite , but will argue that this is insufficient to alter our results .
our results are consistent with those found , in different contexts , in @xcite .
these papers ignore any non - perturbative potentials that might arise and discuss flat moduli space .
they find that the moduli are trapped , both by hubble friction and quantum radiation , exactly at the esp , which , they go on to argue , gives it a special physical meaning .
we note that were we to turn off the non - perturbative superpotential in heterotic @xmath1-theory , one would arrive at precisely the same conclusion .
however , turning on this potential completely alters this result and allows the completion of the small instanton phase transition . throughout this paper , for notational and computational simplicity
, we will take all fields to be dimensionless and choose units in which @xmath67 .
in this section , we consider the dynamics of small instanton transitions ignoring the non - perturbative superpotential generated by worldsheet instantons . in this case
, the moduli space @xmath65 is the flat locus of the potential .
as discussed above , the dimension and geometry of this moduli space is completely understood , at least for a wide range of base spaces @xmath60 and small instanton curves @xmath36 .
however , @xmath11 and @xmath63 are both , generically , multi - dimensional . for example , the small instanton discussed in the introduction had @xmath11 parameterized by @xmath57 complex fields .
unfortunately , a detailed calculation of dynamics in a multi - dimensional moduli space is not only unenlightening in its complexity , but is far beyond our computational power . for this reason
, we will restrict our discussion to a simplified moduli space , one in which both @xmath11 and @xmath63 are taken to be one - dimensional .
we denote the complex modulus of @xmath11 , the instanton phase , by @xmath68 .
physically , this modulus describes the `` smoothing out '' of the small instanton sheaf into the smooth vector bundle @xmath20 .
the complex modulus of @xmath63 , the five - brane phase , is written as @xmath69 and represents the translation modulus of the five - brane .
we want to emphasize that the literature describing transitions in different contexts @xcite all employ very similar simplifications .
we begin by considering the dynamics of the moduli @xmath69 and @xmath68 in static flat spacetime ; that is , we turn off their interaction with gravity .
the simplest @xmath28 supersymmetric self - interactions that have precisely a one - dimension instanton phase and a one - dimension five - brane phase are described by the superpotential @xmath70 recall that both @xmath69 and @xmath68 are taken to be dimensionless and , hence , @xmath71 has mass dimensions @xmath72 in planck units .
furthermore , we choose the planck mass @xmath73 to be unity .
the potential function describing the small instanton phase transition is then given by @xmath74 the instanton phase is the one - dimensional complex submanifold defined by @xmath75 , whereas the five - brane phase is the one - dimensional space given by @xmath76 .
the dynamical calculation is carried out in the four - dimensional real space parameterized by @xmath77 , @xmath78 , @xmath79 and @xmath80 . without loss of generality , one can take @xmath81 .
other values of @xmath71 lead to the same conclusions . in this paper , we will assume that the calabi - yau radius and the interval length are chosen , in units where @xmath67 , to be in the phenomenologically acceptable range @xmath82 this implies that in order to achieve the correct phenomenological values for the newton and gauge coupling parameters , the dimensionless volume and interval moduli should be stabilized at ( or be slowly rolling near ) the values of order one . in this paper
, we will assume that this is the case .
this limits @xmath69 to be less than or of the order of unity . for reasons to become clear in the next subsection , we will choose both @xmath69 and @xmath68 moduli to have values much smaller than one .
let us compute the dynamics with the following initial conditions .
take the initial moduli to be @xmath83 and their initial velocities as @xmath84 these values represent the following physical situation .
first , there is a five - brane in the bulk space substantially separated from the end of the @xmath0 interval with a much smaller displacement in the axionic direction .
a similar very small displacement occurs in the vector bundle modulus .
the five - brane has a small negative velocity in the @xmath85 direction which will cause it to collide with the end of the interval and a similar velocity in the axion direction .
substantially larger velocities are given to both @xmath86 and @xmath87 which enable a possible transition from the five - brane branch to the instanton branch .
the equations of motion are solved numerically on mathematica , with the results for @xmath85 and @xmath87 presented in figures 1 and 2 . curve ( a ) in figure 1 shows that the five - brane collides with the end of the interval but then continues to oscillate around that point .
similarly , curve ( b ) shows that the collision initiates a transition in the vector bundle modulus direction which , however , quickly terminates .
the resulting motion is an oscillation around the small instanton in the vector bundle direction as well .
figure 2 shows the overall motion in moduli space near the small instanton at the origin of field space .
note that although the amplitude of any individual real modulus increases and decreases quasi - periodically , the overall amplitude remains constant in time , as is clearly indicated in figure 2 .
the dynamics of the remaining two real moduli , @xmath88 and @xmath86 , is similar , so we will not display them .
these results are easy to explain .
first of all , in both the five - brane and instanton branches , the potential energy function has walls which become progressively steeper as one moves away from the origin of field space .
hence , the moving moduli see a `` concave '' potential that always reflects them back toward the origin .
this accounts for the oscillation around the small instanton at the origin of field space .
secondly , the lack of any damping , be it hubble friction or any other mechanism , means that the overall oscillation amplitude persists .
the final conclusion is that the phase transition is never completed .
instead the system oscillates endlessly around the small instanton .
note that these results are consistent with those found in @xcite when analyzing undamped motion in different contexts .
let us now consider the moduli @xmath69 and @xmath68 interacting with gravity . in this case
, the dynamics is described by @xmath28 chiral multiplets coupled to supergravity and one must specify both a superpotential @xmath89 and a kahler potential @xmath90 .
the superpotential remains the same , namely @xmath91 in general , the kahler potential of moduli in heterotic @xmath1-theory is unknown .
however , near the origin of field space , that is , when the magnitudes of the moduli are much smaller than unity , this can be taken to be `` flat '' .
therefore , as long as we restrict @xmath69 and @xmath68 describing our small instanton phase transition to be small , the kahler potential is given by @xmath92 this is one reason why we chose the initial conditions for @xmath69 and @xmath68 in the previous subsection to be much less than one .
the potential function describing the small instanton phase transition is now @xmath93 where @xmath89 and @xmath90 are given in ( [ 5 ] ) and ( [ 6 ] ) respectively and @xmath94 is the kahler covariant derivative .
this is a rather complicated function of the moduli .
however , for @xmath69 and @xmath68 much smaller than unity , one need only consider the lowest order interactions .
all terms that are higher order in @xmath95 can safely be ignored .
the result is that one can analyze the dynamics with the same potential as in the non - gravitational case , that is , @xmath96 as previously , we will choose @xmath81 and only consider the region of moduli space where @xmath69 and @xmath68 are much smaller than one .
additionally , one must give an ansatz for the spacetime geometry .
we will assume a metric of the form @xmath97 corresponding to a spatially flat manifold of the frw type . furthermore , we will only consider metrics that are monotonically expanding .
there are , in total , five equations of motion that must be solved simultaneously . to begin with
, there is a single einstein equation for the scale parameter @xmath98 .
additionally , there are the four equations of motion for the real and imaginary parts of the moduli @xmath69 and @xmath68 . despite the fact that the potential in ( [ 8 ] ) is independent of gravitationally induced interactions , non - trivial gravity effects do enter each equation as a hubble damping term .
let us compute the dynamics with the following initial conditions .
take the initial value of the scale parameter to be @xmath99 it is unnecessary to fix an initial radial velocity .
we will continue to choose the initial conditions for the @xmath69 and @xmath68 moduli and their velocities to be those given in ( [ 3 ] ) and ( [ 4 ] ) respectively .
these values represent exactly the same physical situation as in the previous subsection . here , however , the moduli evolve with respect to an expanding frw background spacetime
. the equations of motion are solved numerically on mathematica .
the result for the scale parameter @xmath98 is shown in figure 3 , while those for @xmath85 and @xmath87 are presented in figures 4 and 5 .
figure 3 shows that @xmath98 increases monotonically from its initial value , indicating an expanding frw universe . from curve ( a ) in figure 4
, one sees that the five - brane collides with the end of the interval and then continues to oscillate around that point .
however , the average amplitude and frequency of the oscillation are now strongly damped in time , with @xmath85 converging toward zero .
similarly , curve ( b ) in figure 4 shows that the collision initiates a transition into the vector bundle modulus direction which then terminates , reverses and begins to oscillates around @xmath100 .
once again , the average amplitude and frequency of the oscillation in the vector bundle modulus direction are damped in time and @xmath87 converges to zero .
figure 5 shows the overall motion in moduli space near the small instanton at the origin of field space .
note that the overall amplitude decreases in time .
the dynamics of the remaining real moduli , @xmath88 and @xmath86 , is similar , so we do not display them .
the physical explanation of these results is identical to that given in the previous subsection , with one important exception .
the expansion of the universe produces `` friction '' in the motion of the moduli .
this causes their oscillation around the origin of field space to have ever diminishing amplitude and frequency .
the physical conclusion is that after collision and a period of damped oscillation , the five - brane becomes fixed at the end of the interval .
the associated bundle data is the `` union '' of @xmath58 and a torsion free sheaf over the curve @xmath36 .
that is , the system approaches the small instanton @xmath16 and the phase transition to a smooth bundle @xmath20 is never completed .
note that these results are completely consistent with those found in @xcite when analyzing damped motion in different contexts .
having established our preliminary results , we now turn to the main discussion in this paper ; namely , the dynamics of small instanton transitions in the presence of a non - perturbative superpotential generated by worldsheet instantons . in this case
, the moduli space @xmath63 of the curve remains flat .
however , the potential for the vector bundle moduli is now substantially altered , with the result that , generically , the flat moduli space @xmath11 is replaced by one or more isolated local minima . to proceed , we must first be more explicit about the form of the non - perturbative interactions .
it was shown in @xcite that the superpotential generated by a string wrapped on an isolated curve @xmath101 has the form @xmath102 where @xmath103 are a subset of vector bundle moduli and we have assumed , for simplicity , that the moduli space of the five - brane wrapped on @xmath36 is one - dimensional and parameterized by the dimensionless translation modulus @xmath69 .
the coefficient @xmath104 is then dimensionless and given by @xmath105 where @xmath106 is the size of the @xmath0 interval , @xmath107 is the area of @xmath36 and @xmath108 is the eleven - dimensional planck scale . when the calabi - yau radius and the interval length are chosen , in units where @xmath67 , to be in the phenomenologically acceptable range eq .
, @xmath104 is found @xcite to be of order @xmath109 .
let us recall that in our notation the position of the five - brane is dimensionless and bounded by the length of the interval which is of order one . as mentioned in the introduction , @xmath110 is the pfaffian of the dirac operator twisted by the vector bundle @xmath20 restricted to the curve @xmath36 .
it is found to be a homogeneous polynomial of the moduli @xmath103 of the restricted bundle . in general , the pfaffian is difficult to compute .
however , it has been explicitly calculated in @xcite for vector bundles with structure group @xmath111 restricted to an isolated curve @xmath36 in an elliptic calabi - yau threefold with base @xmath38 .
for example , choosing @xmath51 , a specific @xmath112 vector bundle and @xmath50 , where @xmath41 is an isolated curve in @xmath42 , it was shown that the pfaffian is a homogeneous polynomial of degree @xmath55 in @xmath56 out of the @xmath57 moduli of @xmath11 . an important pattern that emerges from the results in @xcite
is that the smaller the dimension of @xmath11 , the smaller the degree of the homogeneous polynomial .
a final comment is that the coefficient of proportionality has not been explicitly computed .
now consider the simplified moduli space in which both @xmath11 and @xmath63 are one - dimensional and parameterized by the moduli @xmath68 and @xmath69 respectively . in this case , the non - perturbative superpotential @xmath113 in ( [ 11 ] ) takes the form @xmath114 where @xmath115 is a small integer and @xmath116 is a coefficient of mass dimension @xmath117 in planck units .
we now proceed to discuss the moduli dynamics in the presence of superpotential ( [ 14 ] ) .
again consider the dynamics of the moduli @xmath69 and @xmath68 in static flat spacetime .
however , the complete superpotential is now given by @xmath118 the potential function describing the small instanton phase transition is @xmath119 where @xmath120 the minimum of this potential has two distinct types of separated loci .
the first , specified by @xmath121 is the moduli space @xmath122 of @xmath36 .
however , the moduli space @xmath11 is now replaced by @xmath123 isolated points located at @xmath124 where @xmath125 . to carry out an explicit calculation of the moduli dynamics , one must specify the values of the parameters in superpotential ( [ 15 ] ) . without loss of generality , choose @xmath81 . for specificity
, we will take @xmath126 the choice for @xmath104 is consistent with the phenomenologically acceptable values of @xmath127 and @xmath128 given in ( [ 13 ] ) .
similarly , in view of the discussion above , @xmath129 is a reasonable value for the degree of the pfaffian polynomial .
the choice of @xmath116 in ( [ 20 ] ) is less well - motivated .
it is taken because it enhances the distance between the local minimum and the esp in moduli space , as can be seen from ( [ 19 ] ) .
for the parameters in ( [ 20 ] ) , the local minima in ( [ 19 ] ) become @xmath130 similar physical conclusions will be reached for a wide range of all three parameters . with these choices ,
we compute the moduli dynamics using initial conditions similar to those given in ( [ 3 ] ) and ( [ 4 ] ) .
the physical situation is identical to that described in the previous section .
the moduli equations of motion are solved numerically on mathematica , with the results for @xmath77 and @xmath80 shown in figure @xmath131 .
figure @xmath132 indicates that the five - brane collides with the end of the interval but then , eventually , continues to oscillate without damping around that point .
now , however , there is a new phenomenon .
note that immediately after the collision the @xmath77 modulus gets `` hung up '' at the negative value corresponding to the local minima in ( [ 21 ] ) .
it lingers there for a considerable time , before rolling back through the origin and continuing its oscillation around the small instanton point .
similar , but more dramatic , behaviour for @xmath80 is shown in figure @xmath133 .
after a few small oscillations around the origin of field space , the @xmath80 modulus rolls all the way out to the value @xmath134 corresponding to one of the local minima in ( [ 21 ] ) .
it stays in the vicinity of this minimum for some time , before rolling back through the origin and beginning an undamped oscillation around the small instanton .
the choice of @xmath134 rather than the other minimum at @xmath135 is due to the initial conditions .
these can be changed to select the other local minimum .
the behaviour of both moduli is graphically presented in figure @xmath136 , where the excursion of the moduli to the local minimum and back is clearly visible .
the other two real moduli , @xmath78 and @xmath79 , simply oscillate with undamped motion around the origin of field space , so we will not display them . comparing these results with the static flat case in section @xmath137 ,
we conclude that the non - perturbative superpotential has a dramatic effect on the dynamics of the moduli .
specifically , the moduli now evolve far from the origin of field space , seeking , and reaching , an isolated local minimum of the potential .
however , since there is no mechanism for damping , the moduli roll through the isolated minimum , returning to oscillate around the small instanton .
were one to wait for a sufficiently long time , the moduli might once again briefly return to one of the two local minima , only to roll out again and oscillate around the origin of field space .
the lack of any damping means that the overall oscillation amplitude persists .
we conclude that , despite the existence of isolated local minima , the phase transition is never completed .
now let the moduli @xmath69 and @xmath68 interact with gravity .
the superpotential remains the same , namely @xmath138 since the values of all moduli we consider , including the isolated local minima , are much smaller than unity , one can continue to choose the kahler potential to be @xmath139 furthermore , since @xmath69 and @xmath68 are much smaller than one , the complicated potential function in ( [ 7 ] ) simplifies to @xmath140 that is , all terms that are higher order in @xmath95 can be ignored .
the potential function is then identical to the non - gravity case with identical minima ; namely , the five - brane branch ( [ 18 ] ) and the @xmath123 isolated minima given in ( [ 19 ] ) . as in the previous subsection , we choose @xmath81 and take @xmath141 for these parameters , the two isolated minima in ( [ 19 ] ) are again @xmath142 additionally , one must give an ansatz for the spacetime geometry .
we will continue to assume an expanding frw metric of the form given in ( [ 9 ] ) .
let us now compute the dynamics of the moduli @xmath69 and @xmath68 , as well as the spacetime expansion parameter @xmath98 , using the same initial conditions as in section @xmath137 , namely , ( [ 3 ] ) , ( [ 4 ] ) and ( [ 10 ] ) .
the physical situation remains the same as previously .
we use mathematica to numerically solve the equations of motion .
the result for the scale parameter @xmath98 is similar to that shown in figure @xmath72 and , hence , we wo nt present it here .
the dynamics of @xmath85 and @xmath87 is given in figure @xmath56 .
figure @xmath143 indicates that the five - brane collides with the end of the interval .
immediately after the collision , the @xmath85 modulus gets `` hung up '' at the negative value corresponding to the local minima in ( [ 26 ] ) .
now , however , the modulus is `` trapped '' near this value and does not return to the origin of field space .
rather , it oscillates with ever decreasing amplitude and frequency around @xmath144 , eventually coming to rest at this point .
the behaviour of @xmath87 is similar .
figure @xmath145 shows that after a few small oscillations around the origin of field space , the @xmath87 modulus rolls all the way out to the value @xmath134 corresponding to one of the local minima of ( [ 26 ] ) .
again , this local minimum , rather than @xmath135 , is selected by the initial conditions . however , unlike in the previous subsection , the modulus is now `` trapped '' near this minimum , oscillating around it with ever decreasing amplitude and frequency .
it eventually comes to rest at @xmath146 .
the behaviour of both moduli is graphically presented in figure @xmath147 . here , the excursion of the moduli to the local minimum , and the phenomenon of their being trapped at that point , is clearly visible .
the other two real moduli , @xmath88 and @xmath86 , simply oscillate with decreasing amplitude and frequency around the origin of field space .
hence , we will not display them . comparing these results with those in the previous subsection
, we see that the effect of the expanding frw geometry on the dynamics of the @xmath69 and @xmath68 moduli is dramatic .
previously , without gravitation , these moduli evolved out to the local minimum of the non - perturbative superpotential .
however , because of the absence of any mechanism for damping , the moduli eventually rolled out of the local minimum and back to the origin of field space .
they oscillated forever with no loss of amplitude or frequency around the small instanton , occasionally , perhaps , temporarily revisiting one of the two local minima .
when coupled to the expanding frw spacetime , however , this behaviour is completely modified .
the results in figure@xmath56 graphically confirm that the moduli now evolve out to the local minimum and are trapped there by gravitational damping .
the physical reason is clear . immediately after collision ,
the moduli have sufficient `` energy '' to overcome a potential barrier and roll into a valley surrounding the local minimum .
however , the energy is rapidly `` red - shifted '' by the expanding geometry .
the result is that the moduli can no longer cross the potential barrier to return to the origin of field space .
the continued loss of energy to gravitational expansion causes the moduli to oscillate with smaller and smaller amplitude and frequency around the minimum , eventually coming to rest there .
that is , they are gravitationally trapped at the isolated minimum . before drawing our final conclusions
, we would like to comment on the effect of radiative damping on the moduli dynamics . as discussed in @xcite ,
radiative damping occurs at an esp , where the number of massless states and , hence , the quantum mechanical amplitude to produce light particles is greatly enhanced .
this leads to radiative `` friction '' on the moduli dynamics , damping their motion . thus far in this paper
we have ignored this effect , considering only gravitational damping .
could the inclusion of radiative damping change our conclusions , perhaps trapping the moduli at the esp ?
the answer is no .
the results reported in @xcite show that although radiative damping is a strong effect , it requires the moduli to oscillate several times around the esp before their amplitude is substantially reduced .
however , as is clear from figures @xmath143 and @xmath145 , the transition of @xmath69 and @xmath68 from the vicinity of the esp to that of the isolated minimum occurs immediately upon collision , within less than one oscillation ( note that the small amplitude oscillations of @xmath87 in figure @xmath145 all occur prior to the collision and not near the esp ) .
furthermore , they never return to the esp .
it is safe to conclude , therefore , that radiative damping at the esp simply has no opportunity to effect the moduli dynamics and can be safely ignored for the process under discussion in this paper .
in fact , radiative damping helps enable our transition to the isolated minimum .
the reason is that there is also an enhanced number of light states at the local minimum .
once trapped in its vicinity , radiative damping will work with gravitational damping to quickly bring the moduli to rest at the isolated minimum .
to conclude : in an expanding frw spacetime , the collision of a five - brane with an end of the @xmath0 interval passes through the small instanton point , described by
the union of a vector bundle @xmath58 and a torsion free sheaf over the five - brane curve , and then continues to an isolated minimum of the non - perturbative superpotential where the vector bundle is smooth .
gravitational ( and radiative ) damping traps the moduli at the isolated minimum , thus completing the small instanton phase transition to the smooth vector bundle @xmath20 .
the moduli are _ not _ attracted to the esp in this process .
it has been demonstrated in this paper that not only is a small instanton phase transition mathematically consistent , as shown in @xcite , but it can also be realized dynamically .
the results were presented for a specific choice of parameters and initial conditions .
however , we have explored the dynamics of phase transitions for a range of parameters and initial conditions and found that the transition is completed for many different values . however , this is not always the case .
some choices do lead to trapping at the esp , presumably because the energy is either too small to roll into the valley of an isolated minimum or so large that it rolls out again .
be this as it may , initial conditions leading to a completed phase transition are ubiquitous and demonstrate that a nonperturbative addition to the potential can dramatically alter the results of @xcite .
we are grateful to volker braun for discussions of the intricacies of mathematica .
the work of evgeny buchbinder is supported by nsf grant phy-0070928 .
the research of alexander borisov and burt ovrut is supported in part by the department of physics and the math / physics research group at the university of pennsylvania under cooperative research agreement de - fg02 - 95er40893 with the u. s. department of energy and an nsf focused research grant dms0139799 for `` the geometry of superstrings . ''
p. horava and e. witten , `` heterotic and type i string dynamics from eleven dimensions , '' nucl.phys . b460 ( 1996 ) 525 - 540 [ hep - th/9510209 ] ; p. horava and e. witten , `` eleven - dimensional supergravity on a manifold with boundary , '' nucl.phys . b475 ( 1996 ) 94 - 114 [ hep - th/9603142 ] .
a. lukas , b. a. ovrut , and d. waldram , `` on the four - dimensional effective action of strongly coupled heterotic stringtheory '' , nucl.phys .
b532 ( 1998 ) 43 - 82 [ hep - th/9710208 ] ; a. lukas , b. a. ovrut , k. s. stelle , and d. waldram , `` the universe as a domain wall '' , phys.rev . d59 ( 1999 ) 086001 [ hep - th/9803235 ] ; a. lukas , b. a. ovrut , k. s. stelle , and d. waldram , `` heterotic m - theory in five dimensions '' , nucl.phys .
b552 ( 1999 ) 246 - 290 [ hep - th/9806051 ] .
r. friedman , j. morgan and e. witten , `` vector bundles and f theory , '' commun.math.phys . 187 ( 1997 ) 679 - 743 [ hep - th/9701162 ] ; r. donagi , `` principal bundles on elliptic fibrations , '' asian j. math * 1 * ( 1997 ) 214 - 223 [ alg - geom/9702002 ] .
r. donagi , a. lukas , b. a. ovrut and d. waldram , `` non - perturbative vacua and particle physics in m - theory , '' jhep 9905 ( 1999 ) 018 [ hep - th/9811168 ] ; r. donagi , a. lukas , b. a. ovrut and d. waldram , `` holomorphic vector bundles and non - perturbative vacua in m - theory , '' jhep 9906 ( 1999 ) 034 [ hep - th/9901009 ] ; r. donagi , j. khoury , b. a. ovrut , p. j. steinhardt and n. turok , `` visible branes with negative tension in heterotic m - theory , '' jhep 0111 ( 2001 ) 041 [ hep - th/0105199 ] ; r. donagi , b. a.ovrut , t. pantev and r. reinbacher , `` su(4 ) instantons on calabi - yau threefolds with @xmath148 fundamental group , '' jhep 0401 ( 2004 ) 02 [ hep - th/0307273 ] ; y .- h . he , b. a. ovrut , r. reinbacher , `` the moduli of reducible vector bundles , '' jhep 0403 ( 2004 ) 043 [ hep - th/0306121 ] .
e. i. buchbinder , r. donagi and b. a. ovrut , `` vector bundle moduli and small instanton transitions , '' jhep 0206 ( 2002 ) 054 [ hep - th/0202084 ] ; e. i. buchbinder , b. a. ovrut and r. reinbacher , `` instanton moduli in string theory , '' jhep 0504 ( 2005 ) 008 [ hep - th/0410200 ] .
k. becker , m. becker and a. strominger , `` fivebranes , membranes and non - perturbative string theory , '' nucl.phys .
b456 ( 1995 ) 130 - 152 [ hep - th/9507158 ] ; e. witten , `` non - perturbative superpotentials in string theory , '' nucl.phys .
b474 ( 1996 ) 343 - 360 [ hep - th/9604030 ] ; j. harvey and g. moore , `` superpotentials and membrane instantons , '' hep - th/9907026 .
g. moore , g. peradze and n. saulina , `` instabilities in heterotic m - theory induced by open membrane instantons , '' nucl.phys .
b607 ( 2001 ) 117 - 154 [ hep - th/0012104 ] ; e. lima , b. a. ovrut , j. park and r. reinbacher , `` non - perturbative superpotentials from membrane instantons in heterotic m - theory , '' nucl.phys .
b614 ( 2001 ) 117 - 170 [ hep - th/0101049 ] ; e. lima , b. a. ovrut and j. park , `` five - brane superpotentials in heterotic m - theory , '' nucl.phys .
b626 ( 2002 ) 113 - 164 [ hep - th/0102046 ] . | the small instanton transition of a five - brane colliding with one end of the @xmath0 interval in heterotic @xmath1-theory is discussed , with emphasis on the transition moduli , their potential function and the associated non - perturbative superpotential . using numerical methods ,
the equations of motion of these moduli coupled to an expanding friedmann - robertson - walker spacetime are solved including non - perturbative interactions .
it is shown that the five - brane collides with the end of the interval at a small instanton .
however , the moduli then continue to evolve to an isolated minimum of the potential , where they are trapped by gravitational damping .
the torsion free sheaf at the small instanton is `` smoothed out '' into a vector bundle at the isolated minimum , thus dynamically completing the small instanton phase transition .
radiative damping at the origin of moduli space is discussed and shown to be insufficient to trap the moduli at the small instanton point . |
nonlinear conduction in correlated electron systems such as one - dimensional mott insulators@xcite and two - dimensional charge - ordered materials@xcite has been of great interest in the past few decades .
they offer intriguing subjects of nonequilibrium physics in condensed matter and possibility for novel functions of electronic devices .
for example , in a typical quasi - one - dimensional mott insulator , sr@xmath9cuo@xmath10,@xcite a dielectric breakdown has been observed experimentally by applying a strong electric field .
a dielectric breakdown has been reported also in an organic spin - peierls insulator , k - tcnq@xcite [ tcnq = tetracyanoquinodimethane ] .
for another organic compound , ( bedt - ttf)(f@xmath9tcnq ) [ bedt - ttf = bis(ethylenedithio)tetrathiafulvalene ] , which is a quasi - one - dimensional mott insulator , metal - insulator - semiconductor field - effect transistor device structures have been reported,@xcite where its field - effect characteristics are different from those of band insulators.@xcite so far , theoretical investigations on nonlinear conduction for interacting electron systems that are initially insulating in their equilibrium states have been done by several authors.@xcite in general , these studies are classified into two approaches depending on whether an external force is written as an electric field@xcite or a bias voltage.@xcite the former approach is to consider electron systems without electrodes . the electric field is usually applied with open boundary condition,@xcite or equivalently with periodic boundary condition by using a time - dependent magnetic flux.@xcite in one dimension , oka and aoki studied the hubbard model under a strong electric field by the time - dependent density - matrix - renormalization - group method.@xcite one of their important results is that the dielectric breakdown of mott insulators is interpreted as a many - body counterpart of the landau - zener ( lz ) breakdown.@xcite this is known to describe the breakdown of band insulators where the one - particle picture holds .
the threshold is given as @xmath11 with @xmath5 being an energy gap .
a mean - field approach to electric - field - induced insulator - to - metal transitions by using keldysh green s functions has been reported in ref . .
the latter approach is to consider an insulator attached to electrodes .
interface structures must be explicitly taken into account in real electric devices .
along this line , okamoto investigated the current - voltage ( @xmath12-@xmath13 ) characteristics of heterostructures that consist of mott - insulator layers sandwiched by metallic leads by combining the dynamical mean - field theory with the keldysh green s function technique.@xcite ajisaka _ et al .
_ studied the @xmath12-@xmath13 characteristics of an electron - phonon system coupled with two reservoirs by a field - theoretical method.@xcite in the studies listed above , okamoto discussed the @xmath12-@xmath13 characteristics in the framework of the lz breakdown , while ajisaka _
proposed a different mechanism with the threshold bias voltage @xmath2 .
these results seem to be inconsistent with each other .
when we consider a nanostructure in which some material is attached to left and right metallic electrodes , the bias voltage @xmath13 applied to the material is described as @xmath14 with @xmath15 and @xmath16 being the chemical potentials of the left and right electrodes , respectively .
here we assume that the work - function difference at the interfaces is absent for simplicity .
in this case , one might expect that the current flows when some energy levels of the material appear in the region between @xmath15 and @xmath16 , indicating the threshold governed by the applied bias voltage @xmath2 .
in fact , for the transport in a field - effect transistor with a small channel , such an explanation has been used frequently.@xcite however , for a bulk insulator with an energy gap @xmath5 , this picture does not hold and should be replaced by the lz mechanism where the threshold is determined by the electric field , @xmath11 . this consideration poses us a question about which parameter determines the mechanism of the breakdown . in particular
, we address the condition for the realization of the lz breakdown in a structure with electrodes .
it is also important to know the relation between the approaches using the structure with electrodes and those which do not include them explicitly .
we point out that the size of an insulator as well as the potential distribution inside it determine the nature of the breakdown . in this paper , we study one - dimensional band and mott insulators attached to two electrodes ( see fig .
[ fig : model ] ) , using the nonequilibrium green s function approach that has previously been used to discuss the suppression of rectification at metal - mott - insulator interfaces.@xcite this approach more naturally describes nonequilibrium steady states than the approach based on the time - dependent schr@xmath17dinger equation because a current oscillation is inevitable in the latter owing to finite - size effects .
the present method can be easily applied to higher - dimensional systems .
preliminary results for the @xmath12-@xmath13 characteristics of two - dimensional charge - ordered systems are reported in ref . .
we show that the applied bias voltage @xmath13 induces a breakdown of band and mott insulators at zero temperature . for both insulators ,
the threshold shows a crossover as a function of the size of the insulating region @xmath1 . for systems with @xmath1 smaller than the correlation length @xmath18 ,
i.e. , the characteristic decay length of the wave function in the insulator , the breakdown takes place when the bias @xmath13 exceeds the energy gap @xmath5 . for @xmath6 ,
it is governed by the electric field , @xmath19 , which is consistent with the lz tunneling mechanism . whether the charge gap is produced by the band structure or by the electron - electron interaction
is irrelevant to the crossover phenomenon .
we will focus on the spatial modulation of the wave functions . for the crossover behavior and the deformation of the wave functions , the spatial dependence of the scalar potential inside the band or mott insulator is important .
this is demonstrated in the appendix by showing the case without the scalar potential , where the bias - induced transition occurs at @xmath20 regardless of the length @xmath1 .
we consider a one - dimensional insulator that is attached to semi - infinite metallic electrodes on the left and right sides as shown in fig .
[ fig : model ] .
the insulator is referred to as the central part , and described by the hubbard model for a mott insulator and a tight - binding model with alternating transfer integrals for a band insulator , both at half filling : @xmath21 ( c^{\dagger}_{i\sigma}c_{i+1\sigma}+h.c.)\nonumber \\ & + & u\sum_{i}(n_{i\uparrow}-\frac{1}{2 } ) ( n_{i\downarrow}-\frac{1}{2})\nonumber \\ & + & \sum_{\langle\langle ij\rangle\rangle}v_{ij}(n_i-1)(n_j-1 ) , \label{eq : ham}\end{aligned}\ ] ] where @xmath22 denotes the creation ( annihilation ) operator for an electron with spin @xmath23 at the @xmath24th site , @xmath25 , and @xmath26 .
@xmath1 is the total number of sites in the central part .
the parameter @xmath27 denotes the transfer integral , @xmath28 its modulation , and @xmath29 the on - site interaction .
we use @xmath27 as a unit of energy . for @xmath30 and @xmath31 , the first and second terms in eq .
( [ eq : ham ] ) become the one - dimensional repulsive hubbard model , whereas they describe a band insulator for @xmath32 and @xmath33 . the long - range coulomb interaction term with @xmath34
is introduced because it is responsible for the potential modulation near the metal - insulator interfaces ( @xmath35 and @xmath36 ) .
the @xmath37 term is treated by the hartree approximation , which is equivalent to the introduction of a scalar potential that satisfies the poisson equation . here
@xmath38 means the summation over pairs of the @xmath24th and @xmath39th sites with @xmath40 in the central part ( @xmath41 ) .
we briefly review our formulation@xcite below to treat steady states with a finite voltage . for the metallic electrodes , we consider noninteracting electrons .
the effects of the left and right @xmath42 electrodes on the central part are then described by the retarded self - energies.@xcite for simplicity , we take the wide - band limit so that the self - energies are independent of energy . in the present case , they become diagonal matrices@xcite @xmath43 where @xmath44 and @xmath45 are the kronecker delta and @xmath46 ( @xmath47 ) denotes the site connected with the left ( right ) electrode .
we consider the case of @xmath48 . within the hartree - fock approximation for the on - site term in eq .
( [ eq : ham ] ) , the retarded green s function for spin @xmath23 is written as @xmath49_{ij}=\epsilon \delta_{ij}-(h^{r}_{\rm hf\sigma})_{ij},\ ] ] with @xmath50 the off - diagonal elements of @xmath51 come from the first term of eq .
( [ eq : ham ] ) and the diagonal elements are written as @xmath52 with @xmath53 . here , the scalar potential @xmath54 is defined by the hartree approximation to the long - range coulomb interaction as @xmath55 with @xmath56 and @xmath57 being constants .
these constants are so determined that @xmath58 satisfies the boundary conditions : @xmath59 for the bias voltage @xmath60 .
when @xmath13 is positive , the left electrode has a higher potential for the electrons and the current flows from left to right.@xcite here we assume that the work - function difference is absent at the interfaces . by diagonalizing the complex symmetric matrix @xmath61 ,
the retarded green s function is obtained as @xmath62_{ij}=\sum_{m}\frac{u^{\sigma}_{m}(i)u^{\sigma}_{m}(j)}{\epsilon -e^{\sigma}_m } , \label{eq : gr}\ ] ] where @xmath63 ( @xmath64 and @xmath65 are real ) is the eigenvalue of @xmath61 and @xmath66 is the corresponding right eigenvector .
the electron density is calculated by decomposing it into the `` equilibrium '' and `` nonequilibrium '' parts@xcite as @xmath67 the `` equilibrium '' part is defined by integrating the local density of states as @xmath68_{ii } f_c(\epsilon ) , \label{eq : neq}\ ] ] where @xmath69 is the fermi distribution function with the chemical potential @xmath70 at the midpoint of the right and left chemical potentials , @xmath71 . since @xmath72 and @xmath73
, we have @xmath74 . for @xmath75 , where the left chemical potential is higher than the right , @xmath76 ( @xmath77 )
is interpreted as the inflow ( outflow ) .
the `` nonequilibrium '' part of the density @xmath78 is obtained from the `` nonequilibrium '' part of the lesser green s function : @xmath79_{ii}. \label{eq : nneq}\ ] ] in order to obtain @xmath80 , we first decompose the lesser self - energy in the wide - band limit@xcite as in eq .
( [ eq : n_parts ] ) : @xmath81 with @xmath82 and @xmath83,\ ] ] where @xmath84 .
then , we employ the keldysh equation @xmath85 where @xmath86 is the hermitian conjugate of @xmath87 .
the expressions for @xmath88 and @xmath78 , with which the numerical calculations are carried out , are obtained by substituting eq .
( [ eq : gr ] ) into eqs .
( [ eq : neq ] ) and ( [ eq : nneq ] ) .
the results are @xmath89 ^ 2\bigl[\frac{1}{\pi}\tan^{-1}\frac{2(\mu_c-\epsilon^{\sigma}_m ) } { \gamma^{\sigma}_{m}}+\frac{1}{2}\bigr ] , \label{eq : neq2}\ ] ] and @xmath90\nonumber \\ &
= & \frac{\gamma_{\alpha}}{2\pi}\sum_{n , m}\bigl\ { { \rm i m } \bigl [ \frac{u^{\sigma}_m(i)u^{\sigma}_m(i_{\alpha})u^{\sigma * } _ n(i)u^{\sigma * } _ n(i_{\alpha } ) } { \epsilon^{\sigma}_m-\epsilon^{\sigma}_n - i\gamma^{\sigma}_m/2-i\gamma^{\sigma}_n/2 } \bigr ] \bigr\}\nonumber \\ & \times & \bigl [ \tan^{-1}\frac{2(\mu_{\alpha}-\epsilon^{\sigma}_m)}{\gamma^{\sigma}_m } -\tan^{-1}\frac{2(\mu_{c}-\epsilon^{\sigma}_m)}{\gamma^{\sigma}_m}\nonumber \\ & + & \tan^{-1}\frac{2(\mu_{\alpha}-\epsilon^{\sigma}_n)}{\gamma^{\sigma}_n } -\tan^{-1}\frac{2(\mu_{c}-\epsilon^{\sigma}_n)}{\gamma^{\sigma}_n } \bigr]\nonumber \\ & + & { \rm re } \bigl [ \frac{u^{\sigma}_m(i)u^{\sigma}_m(i_{\alpha})u^{\sigma * } _ n(i)u^{\sigma * } _ n(i_{\alpha } ) } { \epsilon^{\sigma}_m-\epsilon^{\sigma}_n - i\gamma^{\sigma}_m/2-i\gamma^{\sigma}_n/2 } \bigr]\nonumber \\ & \times & \bigl [ \frac{1}{2 } \ln \frac{(\mu_{\alpha}-\epsilon^{\sigma}_m)^2+(\gamma^{\sigma}_m/2)^2 } { ( \mu_c-\epsilon^{\sigma}_m)^2+(\gamma^{\sigma}_m/2)^2}\nonumber \\ & - & \frac{1}{2 } \ln \frac{(\mu_{\alpha}-\epsilon^{\sigma}_n)^2+(\gamma^{\sigma}_n/2)^2 } { ( \mu_c-\epsilon^{\sigma}_n)^2+(\gamma^{\sigma}_n/2)^2 } \bigr ] .
\label{eq : nneq2}\end{aligned}\ ] ] in the above equations , we recover the electron density in the equilibrium state without the electrodes if @xmath91 , since the bracket in eq .
( [ eq : neq2 ] ) is reduced to the step function and @xmath92 .
the current from the left electrode is expressed by the `` nonequilibrium '' part of the density as@xcite @xmath93 [ f_l(\epsilon)-f_r(\epsilon)]\nonumber \\ & = & \frac{\gamma_l\gamma_r}{2\pi}\int^{\infty}_{-\infty}d\epsilon \sum_{\sigma}|[g^{r}_{\sigma}(\epsilon)]_{i_li_r}|^2[f_l(\epsilon)-f_r(\epsilon)]\nonumber \\ & = & \gamma_{r}\sum_{\sigma}\delta n^{l}_{i_r\sigma}-\gamma_{l}\sum_{\sigma}\delta n^{r}_{i_l\sigma } , \label{eq : cur}\end{aligned}\ ] ] where we set @xmath94 .
in this section , we show the results of @xmath12-@xmath13 characteristics , charge densities , and the spatial dependence of wave functions for band and mott insulators . for both models ,
a breakdown of the insulating state takes place when the bias @xmath13 becomes sufficiently large .
the threshold shows a crossover behavior as a function of the size of the central part @xmath1 , which indicates the mechanism of the breakdown changes according to @xmath1 .
the profile of the scalar potential @xmath54 has crucial importance on the way of the breakdown .
this is demonstrated in the appendix by showing that , if @xmath54 is artificially set to zero for all @xmath24 , the crossover phenomenon disappears .
figure [ fig : band_iv1 ] shows the @xmath12-@xmath13 characteristics for band insulators with @xmath95 and @xmath96 .
the other parameters are @xmath97 , @xmath33 , @xmath98 , and @xmath99 . for comparison , we show the results for the regular transfer integrals ( @xmath30 ) with @xmath97 , @xmath33 , @xmath100 , and @xmath99 . for @xmath30 ,
the current @xmath101 becomes nonzero for @xmath102 since the central part is metallic .
the @xmath12-@xmath13 curve has stepwise structures owing to the finite - size effect . for @xmath103 and @xmath104 ,
the central part is a band insulator with the energy gap @xmath105 . because of the gap
, @xmath101 is suppressed near @xmath103 .
the @xmath12-@xmath13 curves for finite @xmath28 show more complex structures than that for @xmath30 . apart from the fine structures
, @xmath101 increases almost linearly for large @xmath13 , which indicates a breakdown of the band insulator .
-@xmath13 characteristics of one - dimensional band insulators for @xmath95 and @xmath106 .
the other parameters are @xmath97 , @xmath33 , @xmath98 , and @xmath107 .
results for @xmath30 are also shown where we set @xmath97 , @xmath33 , @xmath100 , and @xmath108.,height=215 ] , `` equilibrium '' part , @xmath109 , and ( b ) `` nonequilibrium '' parts @xmath110 and @xmath111 for @xmath95 , @xmath97 , @xmath33 , @xmath98 , @xmath108 , and @xmath112 .
the scalar potential @xmath54 is shown in the inset of ( a).,title="fig:",height=200 ] , `` equilibrium '' part , @xmath109 , and ( b ) `` nonequilibrium '' parts @xmath110 and @xmath111 for @xmath95 , @xmath97 , @xmath33 , @xmath98 , @xmath108 , and @xmath112 .
the scalar potential @xmath54 is shown in the inset of ( a).,title="fig:",height=200 ] , `` equilibrium '' part , @xmath109 , and ( b ) `` nonequilibrium '' parts @xmath110 and @xmath111 for @xmath95 , @xmath97 , @xmath33 , @xmath98 , @xmath113 , and @xmath114 .
the scalar potential @xmath54 is shown in the inset of ( a).,title="fig:",height=200 ] , `` equilibrium '' part , @xmath109 , and ( b ) `` nonequilibrium '' parts @xmath110 and @xmath111 for @xmath95 , @xmath97 , @xmath33 , @xmath98 , @xmath113 , and @xmath114 .
the scalar potential @xmath54 is shown in the inset of ( a).,title="fig:",height=200 ] for finite voltages applied , the charge distributions in resistive and conductive states for @xmath95 are shown in figs .
[ fig : band_res ] and [ fig : band_con ] , respectively . in fig .
[ fig : band_res](a ) , the electron density @xmath115 and its `` equilibrium '' part @xmath116 for @xmath112 are shown .
a @xmath117 oscillation in the charge distribution is induced by the boundaries.@xcite for all @xmath24 , @xmath118 is almost unity as in the equilibrium case ( @xmath119 for @xmath103 ) .
the electron density @xmath120 is basically unchanged by the bias voltage @xmath13 when @xmath101 is small .
the scalar potential @xmath54 has a linear dependence on @xmath24 as shown in the inset of fig .
[ fig : band_res](a ) .
this is because the long - range interaction term in eq .
( [ eq : psi ] ) is small for @xmath121 , so that @xmath54 is determined only by the boundary conditions .
the `` equilibrium '' part @xmath109 , on the other hand , deviates from unity near the left and right electrodes , where the deviation is canceled by the `` nonequilibrium '' parts @xmath122 , as shown in fig .
[ fig : band_res](b ) . the quantity @xmath123 have nonnegative values for all @xmath24 because electrons come in from the left electrode .
although @xmath123 is large near the left electrode , it decays as @xmath24 increases .
on the other hand , @xmath124 have nonpositive values for all @xmath24 because electrons go out to the right electrode .
note that @xmath125 for @xmath48 , no work - function differences , and at half filling.@xcite the behaviors of @xmath123 and @xmath124 indicate that electrons and holes hardly penetrate into the central part and the resistive state is maintained .
the current hardly flows through the central part , because @xmath101 in eq .
( [ eq : cur ] ) is determined by the difference between the density modulation by the left electrode at the right boundary , @xmath126 , and that by the right electrode at the left boundary , @xmath127
. both terms @xmath126 and @xmath127 are vanishingly small , as shown in fig .
[ fig : band_res](b ) . the charge distribution for @xmath114 , where the system is conductive , is qualitatively different from that for @xmath112 as shown in fig .
[ fig : band_con ] .
the spatial dependences of @xmath109 and @xmath118 are nearly the same .
they increase almost linearly from left to right except in the vicinities of the electrodes , where some oscillatory structure appears .
the distributions of @xmath109 and @xmath118 are understood by that of @xmath54 shown in the inset of fig .
[ fig : band_con](a ) .
the electron density is higher ( lower ) on the right ( left ) half where @xmath54 is low ( high ) .
this behavior is caused by the electrons that move through the system in the conductive phase .
the profile of @xmath54 shows almost a linear dependence on @xmath24 although a small deviation from the linearity near the electrodes is visible in contrast to the resistive phase , which is because the charge redistribution ( @xmath128 ) is easier in the conductive phase . as for the `` nonequilibrium '' parts of the density shown in fig .
4(b ) , @xmath123 have positive values for all @xmath24 , while @xmath124 are negative for all @xmath24 because the electrons come in from the left electrode and go out to the right electrode . a finite
current flows through the central part : @xmath126 and @xmath127 are finite and have the opposite signs .
-@xmath13 characteristics of one - dimensional band insulators for several values of @xmath1 with @xmath95 , @xmath33 , @xmath98 , and @xmath108 .
the solid lines show the function @xmath129 which fits to the results.,height=215 ] and the threshold electric field @xmath130 on the size of the central part @xmath1 .
the other parameters are the same as in fig .
[ fig : band_iv2 ] .
the error bars in the fitting are also shown.,height=215 ] next , we discuss the breakdown mechanism of band insulators . in fig .
[ fig : band_iv2 ] , we show the @xmath12-@xmath13 curves for different sizes of the central parts @xmath1 with @xmath95 . to the numerical results , the function , @xmath131 is well fitted , where @xmath56 and @xmath132 are parameters .
this expression originates from the lz tunneling mechanism through which the insulator breaks down with the threshold voltage @xmath132.@xcite for @xmath133 , the current @xmath101 is exponentially suppressed due to the energy gap , while it increases linearly for @xmath134 .
when the central part is large , the fitting works well as shown in fig .
[ fig : band_iv2 ] , so that the breakdown is consistent with the lz tunneling picture , although there exist fine structures in the @xmath12-@xmath13 characteristics which come from the discreteness of the energy spectrum of the central part .
as @xmath1 decreases , the structure becomes more prominent . for @xmath135 , for example , a deviation from the fitting curve due to the stepwise structure becomes large , which indicates the lz mechanism is no longer applicable to small-@xmath1 systems .
figure [ fig : band_cross ] shows @xmath132 determined by fitting eq .
( [ eq : lz_iv ] ) to the data for each @xmath1 , together with the corresponding electric field @xmath136 . for large @xmath1 ,
@xmath132 is proportional to @xmath1 , so that @xmath130 becomes a constant . in band insulators ,
the lz breakdown is known to be induced by the applied electric field.@xcite since the one - particle picture holds in band insulators , this breakdown can be analyzed as a usual interband tunneling problem and the threshold electric field becomes @xmath137.@xcite the breakdown occurs when the energy gain by displacing an electron with charge @xmath138 in an electric field @xmath139 by the distance @xmath3 , @xmath140 , overcomes the energy gap @xmath5 .
here , @xmath141 is the bandwidth and @xmath142 .
we have obtained the threshold @xmath143 , which is comparable with the value obtained by the lz formula,@xcite @xmath144 with @xmath145 . in short ,
the threshold is governed by the electric field .
when the central part is small , the fitting to the @xmath12-@xmath13 curve becomes worse because the finite - size effect becomes severe .
the lz mechanism is not suitable for understanding this breakdown . in this case , another mechanism , in which the threshold is determined by the bias voltage , is more appropriate for the following reason . as @xmath1 decreases , it eventually becomes smaller than the correlation length @xmath18 .
the tunneling occurs when the energy gain by displacing an electron by the distance @xmath1 , @xmath146 , overcomes the energy gap @xmath5 .
this indicates that the mechanism of the breakdown continuously changes around @xmath147 as a function of @xmath1 .
when @xmath13 exceeds @xmath5 , some energy levels of the central part come in between @xmath15 and @xmath16 . for @xmath148 ,
the wave functions do not fully decay in the system : the electron injected from the left electrode with energy higher than @xmath149 can reach the right electrode through these levels so that the current flows .
this can be clearly seen in fig .
[ fig : band_iv2 ] for @xmath135 where the gap is @xmath150 due to the finite - size effect .
the @xmath12-@xmath13 curve shows an abrupt increase at @xmath151 because @xmath15 exceeds the lowest unoccupied energy level of the central part .
each stepwise increase in the @xmath12-@xmath13 characteristics corresponds to the increase in the number of energy levels located between @xmath15 and @xmath16 .
we have numerically confirmed that the results are qualitatively unchanged even if the long - range coulomb interaction strength @xmath152 and the system - electrode coupling strength @xmath153 ( @xmath154 ) are varied .
thus , the threshold shows a crossover as a function of @xmath1 .
when @xmath155 , the lz - type breakdown occurs and the threshold is governed by the electric field . for @xmath148 ,
on the other hand , the current flows when @xmath13 exceeds the energy gap @xmath5 .
it is noted that the spatial dependence of @xmath54 is important for the realization of the field - induced breakdown as well as the spatial modulation of the wave functions as discussed below . for small @xmath13 , @xmath54 has a linear dependence on @xmath24 throughout the central part because the electrons are localized , @xmath156 , so that the effect of the long - range interaction on @xmath54 is small .
when the system is conductive , the charge redistribution occurs near the interfaces where a small deviation from the linearity is seen in @xmath54 .
this charge redistribution weakens the electric field on the sites away from the interfaces in the central part . for comparison ,
we show in the appendix the @xmath12-@xmath13 characteristics that are obtained by artificially setting @xmath157 for all @xmath24 .
this corresponds to a hypothetical case where a sufficiently large charge redistribution occurs near the interfaces .
there is no electric field in the central part : a voltage drop occurs only at the interfaces . in this extreme case
, we obtain the threshold bias voltage @xmath2 regardless of the length @xmath1 , which is in contrast to the lz - type behavior for large @xmath1 in fig .
[ fig : band_cross ] .
such a situation never occurs in our calculations with @xmath54 and for realistic parameters .
as we will discuss in the next section and also in the appendix , the effect of the spatial profile of @xmath54 on the breakdown mechanism of mott insulators is basically the same as in the case of band insulators .
thus , the model without @xmath54 is inappropriate for realistic insulators . with @xmath158
plotted against @xmath159 for @xmath103 , @xmath160 , @xmath161 , and @xmath162 in the case of @xmath95 , @xmath163 , @xmath33 , @xmath98 , and @xmath164 . for @xmath165 , and @xmath162 ,
the lines are shifted upward by @xmath166 , @xmath162 , and @xmath167 , respectively.,height=249 ] in discussing the breakdown for @xmath6 , the spatial dependences of the wave functions @xmath168 are crucial as explained above . in fig .
[ fig : band_uiruil ] , we show @xmath169 as a function of the real part of the one - particle energy @xmath159 for several values of @xmath13 in the case of @xmath163 .
the behavior of this quantity for @xmath170 is the same .
it shows whether a given one - particle state contributes to the current @xmath101 .
note that @xmath101 is obtained by integrating @xmath171_{i_li_r}|^2 $ ] over @xmath172 .
since @xmath173 appears in the denominator for @xmath174_{ij}$ ] [ eq .
( [ eq : gr ] ) ] , the one - particle state @xmath175 with finite @xmath176 in the interval @xmath177 gives a large contribution to @xmath101 .
this quantity directly shows whether the one - particle state @xmath175 is localized or delocalized because it comes from the product of the amplitudes of the wave function at the two interfaces @xmath178 and @xmath179 . if @xmath176 is large , the state has finite amplitudes at both sides of the central part , so that it is delocalized .
if @xmath176 is small , on the other hand , the state has a small amplitude at either of the interfaces . for @xmath103
, @xmath169 shows two bands that correspond to the conduction and valence bands in the band insulator .
because of the energy gap , no state exists in the region @xmath180 for @xmath103 so that the current does not flow at least for @xmath181 .
when @xmath182 , several states appear in the region @xmath183 , corresponding to the leakage of one - particle states from the electrodes to the central part . however , these states do not contribute to the current because their @xmath169 are vanishingly small as shown in fig .
[ fig : band_uiruil ] .
note that the line for each @xmath102 is shifted upward by @xmath184 . for @xmath185 , the number of states around @xmath186 with small @xmath169 increases .
the energy range where these localized states appear becomes wider as @xmath13 increases .
consequently , the delocalized states that contribute to the current depart from the region @xmath172 .
therefore , the current does not flow even if @xmath13 barely exceeds the gap . as we show in the appendix
, the localized states do not appear if we set @xmath157 for all @xmath24 .
it is crucial to take the spatial dependence of @xmath54 into account to obtain the modulation of the wave functions . for ( a ) @xmath187 , ( b ) @xmath188 , and ( c ) @xmath189 in the case of @xmath185 .
the other parameters are the same as in fig .
[ fig : band_uiruil ] .
the corresponding @xmath159 are indicated by the arrows in fig .
[ fig : band_uiruil].,height=340 ] figure [ fig : band_wf ] shows the spatial dependences of the squares of the absolute values of the wave functions @xmath190 for @xmath185 and several @xmath175 whose @xmath159 are located at the positions indicated by the arrows in fig .
[ fig : band_uiruil ] . here
@xmath175 is so labeled that @xmath191 for @xmath192 .
figure [ fig : band_wf](a ) shows the one - particle state in the lower band with @xmath193 ( @xmath187 ) which is inside the gap for @xmath103 .
this state is localized on the left half of the central part .
the reason is as follows .
the scalar potential is high ( low ) near the left ( right ) electrode . within each of the conduction and valence bands ,
the state whose weight is large near the left ( right ) electrode has a higher ( lower ) energy than others . in the present case ,
the indexes @xmath175 for the valence band are @xmath194 , @xmath195 , @xmath196 , @xmath197 , @xmath198 , @xmath199 , @xmath200 , while those for the conduction band are @xmath201 , @xmath202 , @xmath203 , @xmath204 , @xmath205 , @xmath206 , @xmath207 .
as @xmath175 is lowered , the wave function of the one - particle state is generally extended to a wider region and its largest amplitude is shifted to the right , as shown in fig .
[ fig : band_wf](b ) for the case of @xmath208 ( @xmath188 ) . as @xmath175 is lowered further , e.g. , for @xmath209 ( @xmath189 ) in fig .
[ fig : band_wf](c ) , the one - particle state is delocalized to reach the right electrode . then
, its wave function has large amplitudes near both electrodes . in order to overview the behaviors of the one - particle states , we show the contour map of one - particle states @xmath190 on the @xmath210 plane for @xmath185 in fig .
[ fig : band_map ] . for @xmath185 , the states @xmath211 are localized near the left electrode , and the states below are delocalized . at the bottom of the lower band ,
the states are localized near the right electrode because of the low scalar potential near the right electrode . in @xmath210 plane for @xmath185 .
the other parameters are the same as in fig .
[ fig : band_uiruil].,height=302 ] in this section , we consider the case where the central part is described by the hubbard model .
figure [ fig : mott_iv ] shows the @xmath12-@xmath13 characteristics for @xmath30 , @xmath97 , @xmath212 , @xmath98 , and @xmath99 . when @xmath103 , the system is an antiferromagnetic insulator owing to the hartree - fock approximation .
the energy gap @xmath5 is then 0.25 . in the previous studies on the @xmath12-@xmath13 characteristics of metal - mott - insulator
interfaces,@xcite the current was calculated by solving the time - dependent sch@xmath17dinger equation .
it is argued that the results obtained by the time - dependent hartree - fock approximation for the electron - electron interaction are consistent with those obtained by exact many - electron wave functions on small systems .
for example , the suppression of rectification at metal - mott - insulator interfaces is described by both methods.@xcite although the present time - independent hartree - fock approximation is worse , we expect the present approach captures the essential features of nonequilibrium steady states under the bias voltage .
as we increase @xmath13 , the current begins to flow at @xmath213 .
the breakdown becomes a first - order transition due to the hartree - fock approximation , which is in contrast to the case of band insulators in the previous section .
-@xmath13 characteristics of one - dimensional hubbard model for @xmath30 , @xmath97 , @xmath212 , @xmath98 , and @xmath108.,height=215 ] , `` equilibrium '' part , @xmath109 , and ( b ) `` nonequilibrium '' parts @xmath110 and @xmath111 for @xmath30 , @xmath97 , @xmath212 , @xmath98 , @xmath108 , and @xmath214 .
the scalar potential @xmath54 is shown in the inset of ( a).,title="fig:",height=200 ] , `` equilibrium '' part , @xmath109 , and ( b ) `` nonequilibrium '' parts @xmath110 and @xmath111 for @xmath30 , @xmath97 , @xmath212 , @xmath98 , @xmath108 , and @xmath214 .
the scalar potential @xmath54 is shown in the inset of ( a).,title="fig:",height=200 ] , `` equilibrium '' part , @xmath109 , and ( b ) `` nonequilibrium '' parts @xmath110 and @xmath111 for @xmath30 , @xmath97 , @xmath212 , @xmath98 , @xmath113 , and @xmath114 .
the scalar potential @xmath54 is shown in the inset of ( a).,title="fig:",height=200 ] , `` equilibrium '' part , @xmath109 , and ( b ) `` nonequilibrium '' parts @xmath110 and @xmath111 for @xmath30 , @xmath97 , @xmath212 , @xmath98 , @xmath113 , and @xmath114 .
the scalar potential @xmath54 is shown in the inset of ( a).,title="fig:",height=200 ] the charge distributions in resistive and conductive phases at finite @xmath13 are shown in figs .
[ fig : mott_res ] and [ fig : mott_con ] , respectively .
their overall features are similar to those in band insulators . for all @xmath24 and @xmath133 ,
the electron density @xmath118 is almost unity , which is basically the same as in the equilibrium case ( @xmath119 for @xmath103 ) .
the scalar potential @xmath54 has a linear dependence on @xmath24 as shown in the inset of fig .
[ fig : mott_res](a ) . as for the `` nonequilibrium '' parts ,
@xmath123 ( @xmath124 ) is large near the left ( right ) electrode and decays as @xmath24 increases ( decreases ) [ fig .
11(b ) ] . electrons and holes do not penetrate into the central part so that the current does not flow . for @xmath215 ,
the spatial dependences of @xmath109 and @xmath118 are shown in fig .
[ fig : mott_con](a ) .
they increase almost linearly from left to right .
this reflects the profile of the scalar potential @xmath54 that is higher ( lower ) on the left ( right ) half .
the `` nonequilibrium '' parts of the densities , @xmath123 and @xmath124 are extended over the whole system with small spatial dependence [ fig .
12(b)],@xcite which is in contrast to the resistive phase . since @xmath126 and @xmath127 are finite with opposite signs , a finite current flows through the central part . and the threshold electric field @xmath130 on the size of the central part @xmath1 .
the other parameters are the same as in fig .
[ fig : mott_iv].,height=207 ] in fig .
[ fig : mott_cross ] , we show the threshold bias voltage @xmath132 and the corresponding electric field @xmath216 for the first - order transition as a function of @xmath1 .
when the central part is small , i.e. , @xmath217 , @xmath132 is almost a constant near the energy gap @xmath218 . for small @xmath1 ,
the electron injected from the left electrode with energy higher than @xmath149 can reach the right electrode since the correlation length @xmath219 is comparable to @xmath1 .
thus , the threshold for @xmath148 is determined by the bias voltage . for @xmath6 , on the other hand , @xmath132 is proportional to @xmath1 , so that the threshold is governed by the electric field . in recent theoretical studies,@xcite oka and aoki
have proposed that the lz breakdown occurs also in mott insulators by applying the time - dependent density - matrix - renormalization - group method to the one - dimensional hubbard model under an electric field with open boundary condition . in our calculations ,
the scalar potential @xmath54 is linearly increasing with @xmath24 in the resistive phase as shown in fig .
[ fig : mott_res ] ( a ) , which means that the electrons feel a uniform electric field in the central part .
therefore , our model describes the lz breakdown as in the open hubbard chain as long as the electrodes do not affect the nature of the breakdown for @xmath6 .
in fact , the threshold @xmath130 is about 0.0033 , which is comparable to the lz value,@xcite @xmath220 , with @xmath145 .
thus , the threshold shows a crossover as a function of @xmath1 as in the case of band insulators .
the lz breakdown is explained as before by comparing the charge gap @xmath5 and the work which is done by the electric field on an electron moving over the correlation length @xmath18 .
if the work @xmath221 exceeds @xmath5 , the electron in the lower band may go over to the upper band so that the current flows .
according to the results by oka and aoki,@xcite this consideration is applicable to mott insulators where the correlation effects are important .
thus , we expect that the results obtained by the hartree - fock approximation are qualitatively unchanged even if we take account of the electron correlation .
it is well known that the hartree - fock theory overestimates the charge gap @xmath5 .
it predicts the antiferromagnetic spin ordering which is actually destroyed if quantum fluctuations are appropriately taken into account .
however , the overestimated @xmath5 will alter the threshold only quantitatively .
we also expect that the spin ordering does not essentially affect the breakdown itself since only the charge degrees of freedom are relevant to the mechanism .
we note that the time - dependent density - matrix - renormalization - group method has been applied to a mott insulator with electrodes very recently,@xcite and that the @xmath12-@xmath13 characteristics have been consistently explained by the lz tunneling mechanism .
however , it is also shown that physical quantities such as the spin structure factor and the double occupancy do not reach a stationary state in the accessible time window .
therefore , a direct description of nonequilibrium steady states as in the present study is important to deepen our understanding of the breakdown .
since the crossover behavior is obtained for both band and mott insulators , the electron - electron interaction is not responsible for the phenomenon .
we emphasize that the spatial profile of @xmath54 is important for the realization of the lz breakdown . in the appendix , this is demonstrated for the mott insulator by showing the @xmath12-@xmath13 characteristics that are obtained by artificially setting @xmath157 for all @xmath24 . in this case
, no electric field exists in the central part so that the modification of the wave functions does not occur .
the @xmath12-@xmath13 curves do not show any @xmath1 dependence apart from the fine structures coming from the discreteness of the energy spectrum .
we obtain the threshold bias voltage @xmath2 regardless of the length @xmath1 as in band insulators . in fig .
[ fig : mott_uiruil ] , we show @xmath169 as a function of the real part of the one - particle energy @xmath159 for several values of @xmath13 , where each line for @xmath102 is shifted upward by @xmath13 .
the behavior of this quantity for @xmath170 is the same
. with @xmath158 plotted against @xmath159 for @xmath103 , @xmath166 , @xmath162 , and @xmath167 in the case of @xmath222 , @xmath97 , @xmath212 , @xmath98 , and @xmath164 . for @xmath112 , @xmath162 , and @xmath167 ,
the lines are shifted upward by @xmath166 , @xmath162 , and @xmath167 , respectively.,height=249 ] for ( a ) @xmath223 , ( b ) @xmath224 , and ( c ) @xmath225 in the case of @xmath226 .
the other parameters are the same as in fig .
[ fig : mott_uiruil ] .
the corresponding @xmath159 are indicated by the arrows in fig .
[ fig : mott_uiruil].,height=340 ] for @xmath103 , no state exists in the region @xmath180 since the energy gap opens .
@xmath169 shows two bands that correspond to the upper and lower hubbard bands .
when @xmath112 , one - particle states leak from the electrodes to the central part so that several states appear in the region @xmath183 . as in band insulators , these states do not have any contributions to the current because their @xmath169 are vanishingly small . as we increase @xmath13 further , e.g. , @xmath226 , the states with vanishingly small @xmath169 appear in a wider range around @xmath186 .
the number of these localized states also increases .
the appearance of the localized states keeps the central part resistive until @xmath19 reaches the threshold electric field @xmath130 even if @xmath227 holds .
when the system is conductive ( @xmath134 ) , the upper and lower hubbard bands are merged into a single metallic band . in this case , all the states around @xmath228 are delocalized and contribute to the current .
figure [ fig : mott_wf ] shows @xmath190 as a function of @xmath24 for @xmath226 . here
@xmath175 is chosen at @xmath229 , @xmath230 , and @xmath231 , whose @xmath159 are located at the positions indicated by the arrows in fig . [
fig : mott_uiruil ] . in fig .
[ fig : mott_wf](a ) , we show the wave function with @xmath232 ( @xmath223 ) , which is localized on the left half of the central part .
this state belongs to the lower hubbard band .
since the scalar potential is high ( low ) near the left ( right ) electrode , the state is located near the top of the lower hubbard band .
the one - particle states in the band gradually lose their localized nature as @xmath175 is lowered .
this can be seen in fig .
[ fig : mott_wf](b ) for the case of @xmath233 ( @xmath224 ) , where its largest amplitude is shifted to the right compared to that of @xmath223 .
figure [ fig : mott_wf](c ) shows the one - particle state for @xmath234 ( @xmath225 ) , which is completely delocalized .
its wave function has large amplitudes near both electrodes .
the spatial dependences of the one - particle states in the resistive phase are similar to those in band insulators .
we have investigated the @xmath12-@xmath13 characteristics of the one - dimensional band and mott insulators attached to electrodes . a tight binding model with alternating transfer integrals for the band insulator and the hubbard model for the mott insulator
are studied by using the nonequilibrium green s function method .
the applied bias voltage induces a breakdown of the insulating state to convert into a conductive state for both models .
the threshold shows a crossover as a function of the size @xmath1 of the insulators . for @xmath235 ,
the breakdown occurs at @xmath2 so that the threshold is governed by the bias voltage . for @xmath6 ,
the electric field determines the threshold , @xmath8 , which is consistent with the lz breakdown reported previously.@xcite since the crossover is obtained for both band and mott insulators , the electron - electron interaction is not responsible for the phenomenon . the profile of the scalar potential @xmath54 , which is linearly increasing with @xmath24 in the resistive phase so that the electrons in the central part feel an almost uniform electric field , is important for the realization of the lz breakdown and the crossover behavior .
this work was supported by grants - in - aid for scientific research ( c ) ( grant no .
19540381 ) and scientific research ( b ) ( grant no .
20340101 ) , and by `` grand challenges in next - generation integrated nanoscience '' from the ministry of education , culture , sports , science and technology of japan . in this appendix , we show the results when we artificially set @xmath157 for all @xmath24 .
the @xmath12-@xmath13 curves of the band insulator with @xmath106 , @xmath33 , @xmath164 , and @xmath157 for @xmath236 , @xmath203 , and @xmath207 are shown in fig .
[ fig : band_psi0 ] .
it is apparent that the breakdown occurs at @xmath2 regardless of the length @xmath1 , which is consistent with ajisaka _
et al._@xcite this is in contrast to the results in fig .
[ fig : band_cross ] where @xmath132 is proportional to @xmath1 for @xmath237 .
when we fix @xmath157 , one - particle states do not leak from the electrodes to the central part since the electric field is absent from the central part . in fig .
[ fig : band_app_uiruil ] , we plot @xmath169 as a function of @xmath159 for several values of @xmath13 .
the results indicate that the one - particle energies and the wave functions are not affected by @xmath13 .
this comes from the fact that the hamiltonian in eq .
( 1 ) does not depend on @xmath118 for band insulators if we set @xmath157 for all @xmath24 .
therefore , no localized state appears inside the gap for @xmath103 . in this case
, the current begins to flow when @xmath13 merely exceeds @xmath2 since one - particle states with finite @xmath169 appear in the region @xmath238 . in fig .
[ fig : mott_psi0 ] , we show the @xmath12-@xmath13 curves of the mott insulator with @xmath30 , @xmath212 , @xmath164 , and @xmath157 for @xmath236 , @xmath203 , and @xmath207 . as in the case of band - insulators ,
the breakdown occurs at @xmath2 for all @xmath1 .
although the breakdown seems to be continuous for @xmath236 and @xmath203 , a small discontinuity is evident for @xmath163 , which indicates a first - order transition .
the discontinuity is more obvious for large @xmath29 as shown in the inset of fig .
[ fig : mott_psi0 ] for @xmath239 with the gap @xmath240 .
-@xmath13 characteristics in the case of @xmath242 , @xmath212 , @xmath164 , and @xmath157 for several values of @xmath1 .
the results for @xmath239 and @xmath157 are also shown in the inset .
the arrow indicates the location of the gap @xmath5.,height=215 ] 99 y. taguchi , t. matsumoto , and y. tokura , phys .
b * 62 * , 7015 ( 2000 ) .
r. kumai , y. okimoto , and y. tokura , science * 284 * , 1645 ( 1999 ) .
s. yamanouchi , y. taguchi , and y. tokura , phys .
lett . * 83 * , 5555 ( 1999 ) .
f. sawano , i. terasaki , h. mori , t. mori , m. watanabe , n. ikeda , y. nogami , and y. noda , nature * 437 * , 522 ( 2005 ) .
r. kondo , m. higa , and s. kagoshima , j. phys .
* 76 * , 033703 ( 2007 ) .
s. niizeki , f. yoshikane , k. kohno , k. takahashi , h. mori , y. bando , t. kawamoto , and t. mori , j. phys .
jpn . * 77 * , 073710 ( 2008 ) .
f. sawano , t. suko , t. s. inada , s. tasaki , i. terasaki , h. mori , t. mori , y. nogami , n. ikeda , m. watanabe , and y. noda , j. phys .
. jpn . * 78 * , 024714 ( 2009 ) .
t. s. inada , i. terasaki , h. mori , and t. mori , phys .
b * 79 * , 165102 ( 2009 ) . t. hasegawa , k. mattenberger , j. takeya , and b. batlogg , phys .
b * 69 * , 245115 ( 2004 ) .
k. yonemitsu , j. phys .
. jpn . * 74 * , 2544 ( 2005 ) .
k. yonemitsu , n. maeshima , and t. hasegawa , phys .
b * 76 * , 235118 ( 2007 ) .
k. yonemitsu , j. phys .
78 * , 054705 ( 2009 ) .
t. oka , r. arita , and h. aoki , phys .
lett * 91 * , 066406 ( 2003 ) .
t. oka and h. aoki , phys .
lett * 95 * , 137601 ( 2005 ) .
t. oka and n. nagaosa , phys .
lett * 95 * , 266403 ( 2005 ) .
t. oka and h. aoki , phys .
b * 81 * , 033103 ( 2010 ) .
s. onoda , n. sugimoto , and n. nagaosa , prog .
phys . * 116 * , 61 ( 2006 ) .
n. sugimoto , s. onoda , and n. nagaosa , prog .
. phys . * 117 * , 415 ( 2007 ) .
n. sugimoto , s. onoda , and n. nagaosa , phys .
b * 78 * , 155104 ( 2008 ) .
s. okamoto , phys .
b * 76 * , 035105 ( 2007 ) .
s. okamoto , phys .
* 101 * , 116807 ( 2008 ) .
s. ajisaka , h. nishimura , s. tasaki , and i. terasaki , prog .
. phys . * 121 * , 1289 ( 2009 ) .
l. d. landau , phys .
z. sowjetunion * 2 * , 46 ( 1932 ) . c. zener , proc .
london , ser . a * 137 * , 696 ( 1932 ) .
s. datta , _ quantum transport : atom to transistor _
( cambridge university press , cambridge , 2005 ) y. tanaka and k. yonemitsu , physica b * 405 * , s211 ( 2010 ) .
y. meir and n. s. wingreen , phys .
* 68 * , 2512 ( 1992 ) .
n. s. wingreen , a .- p .
jauho , and y. meir , phys .
b * 48 * , 8487 ( 1993 ) . a .-
jauho , n. s. wingreen , and y. meir , phys .
b * 50 * , 5528 ( 1994 ) .
h. haug and a .-
yauho , _ quantum kinetics in transport and optics of semiconductors _
( springer , berlin , 2008 ) 2nd ed . j. m. ziman , _ principles of the theory of solids _ ( cambridge university press , cambridge , 1979 ) f. heidrich - meisner , i. gonz@xmath243lez , k. a. al - hassanieh , a. e. feiguin , m. j. rozenberg , and e. dagotto , phys .
b * 82 * , 205110 ( 2010 ) . | nonequilibrium states induced by an applied bias voltage @xmath0 and the corresponding current - voltage characteristics of one - dimensional models describing band and mott insulators are investigated theoretically by using nonequilibrium green s functions .
we attach the models to metallic electrodes whose effects are incorporated into the self - energy .
modulation of the electron density and the scalar potential coming from the additional long - range interaction are calculated self - consistently within the hartree approximation . for both models of band and mott insulators with length @xmath1 ,
the bias voltage induces a breakdown of the insulating state , whose threshold shows a crossover depending on @xmath1 .
it is determined basically by the bias @xmath2 for @xmath1 smaller than the correlation length @xmath3 where @xmath4 denotes the bandwidth and @xmath5 the energy gap . for systems with @xmath6 ,
the threshold is governed by the electric field , @xmath7 , which is consistent with a landau - zener - type breakdown , @xmath8 .
we demonstrate that the spatial dependence of the scalar potential is crucially important for this crossover by showing the case without the scalar potential , where the breakdown occurs at @xmath2 regardless of the length @xmath1 . |
many natural and technological complex systems can be described as networks connecting a large number of elements whose states evolve in time @xcite .
the collective behavior of a system resulting from interactions between its components can exhibit non - trivial features , including critical phenomena @xcite .
for instance , a system of binary - state elements ( e.g. , representing individuals having opposing opinions on an issue ) connected through a network having modular organization can exhibit ordering dynamics at very distinct time - scales @xcite and under certain circumstances , self - organize into locally aligned clusters that correspond to the communities of the network ( i.e. , subnetworks characterized by a significantly higher connection density compared to the overall density of the network ) @xcite . in many situations , the links of the network ( representing the interactions ) can also evolve over time as a result of the changes in the states of the components that they connect .
such connections may not only be characterized by weights ( indicating the strength of interaction ) but also sign ( representing the nature of the interaction ) .
for instance , in the context of a network of synaptically - connected model neurons , positive links may correspond to excitatory interactions ( whereby activation of one element can result in activation of other connected elements ) while negative links can give rise to inhibition ( i.e. , activation of an element tends to suppress subsequent activation of neighboring elements ) @xcite .
the occurrence of negative links can lead to the emergence of _ frustration _ because of the existence of inconsistent relations within cycles in the network @xcite .
a network is said to be structurally balanced if its positive and negative links are arranged such that frustration is absent .
such a network can always be represented as two communities , with interactions within each community being exclusively positive while those between communities can only be negative @xcite .
this is of interest not only for social systems in the context of which the concept of balance was first introduced @xcite , but also for biological systems .
for instance , it has been recently observed that the resting human brain is organized into a pair of dynamically anti - correlated subnetworks @xcite .
this suggests that the network responsible for this behavior may be structurally balanced .
this possibility is intriguing in view of the observation of non - trivial collective behavior in balanced networks of binary - state dynamical elements , such as , the coexistence of ordered and disordered regions referred to as `` chimera '' order , in the presence of an external field @xcite .
the process by which networks can evolve to a balanced configuration has been explored by considering a link adaptation process inspired by hebb s principle , the basis behind neural plasticity @xcite . according to this mechanism , weights associated with each link change in proportion to the correlation in the activity of the connected elements .
therefore , any initial frustration in the network can be removed by modifying the connection weights in accordance with the dynamical states of the elements .
systems undergoing such link adaptation in the presence of environmental noise show high variability in the convergence time required to reach the balanced state @xcite . in this paper
we consider how the rate of dynamical evolution of interactions in accordance with the hebbian adaptation principle affects the collective behavior of the network .
starting from a fully connected network that is structurally balanced and introducing noise and external field so that the system exhibits chimera order , we show that different rates of adaptation can result in distinct outcomes .
fast learning rates result in persistence of the chimera regime , although the balanced network now comprises communities with asymmetric sizes , viz . , the ordered sub - network containing a much larger fraction of elements of the network .
slow learning , on the other hand , results in a completely ordered state with the interactions becoming only of positive nature . on removing the external field , fast learning results in retrieving a network that is similar to the original one , i.e. , structurally balanced with communities of almost equal size
. however , slow learning gives rise to a fully frustrated network exhibiting disorder .
we observe that there exists a critical interval for values of the learning rate , around which small changes result in the system converging to distinct final states . in the following sections we first discuss the model and adaptation dynamics , followed by description of the results of numerical simulations .
we conclude with a brief summary of our results and a discussion of their implications .
for the purpose of investigating how collective behavior of a complex system is affected by adaptive dynamics of interactions , we use one of the well - known spin models of statistical physics which are generic systems for analyzing cooperative phenomena .
in particular , we use the binary - state ising spin , the spin orientations ( `` up '' or `` down '' ) representing a pair of mutually exclusive choices .
the simplicity of the model makes it applicable to not just magnetic materials ( in the context of which it was originally proposed ) but any system where the elements choose between the two competing states based on interactions with neighboring elements , noise ( represented by thermal fluctuations characterized by a temperature @xmath0 ) and external stimulus ( often represented by a magnetic field @xmath1 ) .
the interactions @xmath2 between any pair of spins ( @xmath3 , @xmath4 ) can be either positive ( promoting connected spins to have the same state ) or negative ( promoting connected spins to have opposite states ) in nature .
for instance , in neuronal networks , one can view the neurons as binary - state devices that are either firing ( active ) or quiescent ( inactive ) .
correspondingly , as pointed out earlier , the excitatory and inhibitory connections between neurons can be represented by positive and negative interactions , respectively . at a different scale
, one can view genetic regulatory networks in a similar vein , with genes being in either of two states , viz .
, getting expressed ( active ) or not(inactive ) .
in this setting , genes promoting the expression of other genes correspond to positive interaction , while inhibiting the expression of other genes correspond to negative interactions . in a different context , the interactions between ising spins can also be used to represent social intercourse @xcite . here , the spin states are considered to be analogous to two competing opinions , while the interactions may represent the nature of the relation between a pair of individuals - positive corresponding to affiliative and negative corresponding to antagonistic relations .
we consider a system of @xmath5 globally coupled ising spins arranged into two sub - populations ( called modules or communities ) each having @xmath6 spins , at a constant temperature @xmath0 .
for the simulations reported below we have chosen @xmath7 .
the interaction between a pair of spins belonging to the same module is positive having strength @xmath8 ( @xmath9 ) , while that between spins belonging to different modules is negative with strength @xmath10 ( where @xmath11 ) . in the absence of an external field and thermal fluctuations ( i.e. , @xmath12 , @xmath13 ) , the two modules will be completely ordered in opposite orientations . when subjected to an external field of strength @xmath1 ,
the energy of the system is described by @xmath14 where @xmath15 is the ising spin on the @xmath3-th node ( @xmath16 ) in the @xmath17-th module ( @xmath18 ) .
as every spin interacts with every other spin , a mean - field treatment should describe the system accurately . for convenience
we define @xmath19 and @xmath20 as system parameters .
the state of all spins in the system are updated stochastically at discrete time - steps using the metropolis monte carlo ( mc ) algorithm with temperature @xmath0 expressed in units of @xmath21 where @xmath22 is the boltzmann constant .
the initial state of the system we have chosen is one exhibiting chimera order @xcite , to obtain which we use @xmath23 , @xmath24 and @xmath25 .
we also allow the possibility that the interaction strengths can change over time through an adaptation process inspired by the principle of hebbian learning , a classical concept in the area of neural networks @xcite .
colloquially often described as
neurons that fire together , wire together " , in this mechanism the strength of synaptic connection between a pair of neurons is incremented when the two exhibit correlated activation . in ref .
@xcite , this idea was used to propose a link adaptation rule that applies to a system of spins coupled via positive , as well as , negative interactions as : @xmath26 which is applied after every mc step .
the adaptation rate , @xmath27 , decides the time - scale over which the interaction strength changes relative to the spin dynamics . starting with an initially frustrated spin system , implementing the above @xmath2 dynamics in absence of thermal fluctuations ( i.e. , @xmath28 ) results in the system converging to a structurally balanced state @xcite .
this can be intuitively understood in terms of changes in the energy landscape over which the state of the spin system evolves .
frustrated systems have rugged energy landscapes comprising a large number of local minima ( a fact which is exploited in neural network models of associative memory where these minima are used to embed desired patterns one wishes the network to memorize " @xcite . thus , beginning at any randomly chosen initial state of spin configurations , the spin dynamics drives the system into the nearest local energy minimum .
the subsequent evolution of the @xmath2s converts this into the global minimum of the system .
however , in the presence of noise ( @xmath29 ) , fluctuations in the spin dynamics can prevent the system from being trapped in any state for sufficiently long duration .
thus , the adaptation dynamics fails to alter the energy landscape sufficiently so as to turn the configuration into the global minimum .
therefore , with increasing temperature , an extremely long time may be required to reach structural balance .
in this paper we use a modified form of the hebb - inspired adaptation rule for interaction strengths that was introduced in ref .
this is to take into account the asymmetry in the strengths of positive and negative interactions necessary for observing chimera order ( typically @xmath30 ) @xcite .
we change the adaptation rule such that the upper and lower bounds for the evolving interactions have the same values as that of the positive and negative interactions ( respectively ) which give rise to the chimera ordered state .
thus after each mc step , interaction strengths are changed according to : @xmath31 .
\label{hebbrulenew}\ ] ] the state of the system at any time is characterized by two quantities .
one of these is the frustration associated with the interactions , measured by calculating the fraction of frustrated triads ( i.e. , connected sets of 3 spins with an odd number of negative interactions ) in the system : @xmath32 which varies between @xmath33 ( no frustration ) and @xmath34 ( completely frustrated ) . and
@xmath35 , in ( a ) chimera ordered state and ( b ) disordered state , shown for mc simulations with @xmath36 at ( a ) @xmath37 and ( b ) @xmath38 , respectively . ]
shown when the spin system exhibits ( a ) chimera order and ( b ) disorder .
the kernel smoothened distribution function is represented by a thick curve .
results shown for mc simulations with @xmath36 at ( a ) @xmath37 and ( b ) @xmath38 , respectively . ]
the other quantity is an order parameter @xmath39 that is used to identify a chimera state in a spin system with two modules that have magnetizations @xmath40 and @xmath35 , respectively . to numerically estimate the order parameter @xmath41 ,
we are confronted with a few potential complications .
first , as the magnetizations of the two modules are stochastically fluctuating ( especially the one which is disordered , i.e. , having lower magnetization ) , we can use their average values .
however , this gives rise to an additional problem as the modules can frequently exchange their order / disorder status , especially when the modules are of the same size [ fig .
[ mag_time ] ( a ) ] .
thus , a method is required to measure @xmath41 that is unaffected by the modules switching their magnetizations , while being able to determine if the time - averaged magnetizations of the two modules are different ( the signature of chimera order ) .
for instance , a simple time average of the absolute value of differences between @xmath40 and @xmath35 will not give correct results as one obtains a finite value ( because of stochastic fluctuations ) even when the time averaged magnetizations are same for both modules [ fig .
[ mag_time ] ( b ) ] . therefore , to estimate @xmath41 we have used the following algorithm .
first , the frequency distribution of @xmath42 is computed . in the absence of chimera ,
the distribution is unimodal [ fig .
[ o_distrib](a ) ] which is approximated as a gaussian function , while for chimera ordering the distribution is bimodal [ fig .
[ o_distrib](b ) ] which is approximated as a superposition of two gaussian functions . for a unimodal distribution ,
the value of @xmath42 corresponding to the peak is used to calculate the order parameter @xmath41 . for bimodal distributions , a weighted average of the values corresponding to the peaks
for example , if one of the peaks occurs at @xmath43 with value @xmath44 while the other is at @xmath45 with value @xmath46 , the order parameter is calculated as @xmath47 . in order to make the automated detection of the peak locations accurate , the frequency distributions need to be smoothed of all fluctuations in the frequencies so that the local maxima at the peaks can be uniquely determined . for this purpose
we have used a kernel smoothing technique @xcite that gives a single peak location for unimodal distributions and locations for two peaks in the case of bimodal distributions .
starting from a structurally balanced network comprising two modules , we first introduce field and thermal fluctuations so as to drive the system into chimera order - i.e. , one of the modules becomes ordered while the other is disordered .
once this is achieved we allow adaptation dynamics of the interactions to take place . as the evolution of the interactions are related to the degree of fluctuations in the spin states , these would occur mostly in the disordered module where the spins are subject to the competing forces of negative interactions with the spins belonging to the ordered module and the influence exerted by the uniform external magnetic field which tries to align the spins in parallel with those of the ordered module .
thus , at any instant , a few spins in the disordered module will get aligned with the spins in the other module and consequently may develop positive interactions with the latter as a result of link adaptation .
this means that they will now no longer be part of the disordered module but become part of the ordered module . as a result , the size of the modules would change over time - the ordered module increasing at the expense of the disordered one .
this has to be taken into account when measuring system properties , such as , magnetization per spin of the two modules . in order to have a consistent definition of module size that would be valid even when the system is
no longer in structural balance , we heuristically measure it as follows . at each update of the @xmath2s , we randomly choose a spin from the ordered module and consider all spins that have positive interactions with it to belong to the ordered module , with the remaining spins comprising the disordered module . following this process , we can follow the time - evolution of modular membership of individual spins , as well as , that of the module sizes
. however , this measure becomes less useful as the system becomes increasingly frustrated .
mc steps ( indicated by broken line ) .
the ordered module expands in size ( @xmath48 ) while the disordered one shrinks .
learning rate @xmath27 = 0.1 ] and @xmath35 of the two modules in the network ( having sizes @xmath48 and @xmath49 , respectively ) before and after link adaptation dynamics is introduced at time @xmath50 mc steps ( indicated by broken line ) . in absence of adaptation
, magnetizations show the characteristic signature for chimera order .
once adaptation dynamics is operational , the ordered module with magnetization @xmath40 is seen to increase in size .
( bottom ) in absence of learning , the network exhibits no frustration as it is structurally balanced .
following the introduction of adaptation dynamics , there is a transient increase in frustration , before the system once again achieves balance , but with modules of asymmetric sizes . ]
[ hebb_module_size ] illustrates the evolution of module sizes with time following the introduction of link adaptation dynamics in the system .
we can now measure the magnetizations per spin , @xmath40 and @xmath35 , of the two modules by taking into account the modular membership of each spin and the module sizes . fig .
[ hebb_m1m2 ] shows these order parameters , as well as , the frustration in the network , as a function of time before and after the link adaptation dynamics is introduced .
as expected , we observe that following the start of link adaptation dynamics , the ordered module begins to grow in size , which is also observed in the time - evolution of the interaction matrix ( shown as snapshots at regular intervals in fig .
[ jsnaps ] ) .
note that the time interval shown here corresponds to the same period over which the frustration in the system initially rises and then decreases again as the system once more reaches a balanced state , as indicated in fig .
[ hebb_m1m2 ] . ]
is introduced ( the time at which the link adaptation is initiated is indicated by the broken line ) . following a brief transient rise in the frustration
, the system again attains a balanced state although with asymmetric module sizes . ]
is introduced ( the time at which the link adaptation is initiated is indicated by the broken line ) .
the system eventually becomes completely ordered with all interactions becoming positive ( resulting in merging of the two modules into a single one ) . ]
is introduced ( the time at which the link adaptation is initiated is indicated by the broken line ) .
initially , the two modules both become ordered .
however , eventually the two modules merge into a single one as all interactions become positive .
the time of merging coincides with the transient rise in frustration . ] is introduced ( the time at which the link adaptation is initiated is indicated by the broken line ) .
we observe that while one of the modules remains ordered , the initially disordered modules is gradually also tending towards complete order .
eventually one expects the two modules to merge into a single one . ]
the time - evolution of the initial chimera - ordered network subjected to link adaptation dynamics with various learning rates ranging from @xmath51 to @xmath52 are indicated in figs .
[ eps1e-1]-[eps1e-4 ] . for fast learning rates ( e.g. , @xmath53 )
, the system immediately converges to a balanced state characterized by a large ordered module ( comprising about @xmath54 of all the elements ) and a small disordered one . by observing the magnetization time - series of the two modules in fig .
[ eps1e-1 ] it is easy to infer that the system still exhibits chimera order .
we note that a rise in frustration is seen for a very brief duration during the initial period immediately following initiation of the link adaptation dynamics . for a slower learning rate , viz .
@xmath55 , the process of expansion in size of the ordered module is different ( fig .
[ eps1e-2 ] ) .
we notice that soon after the beginning of link adaptation , the chimera ordering is lost as the disordered module becomes ordered .
however , the modules still retain their different identities as they are defined based on the sign of interactions of the spins within them ( which should be positive ) and that with spins in the other module ( which should be negative ) . for even slower learning rates such as @xmath56 or @xmath52 ( figs .
[ eps1e-3]-[eps1e-4 ] ) the duration over which the two modules exist while both being ordered is seen to increase . during this period the modular membership of the spins
do not change significantly ( apart from small fluctuations ) .
chimera ordering is lost as the interaction strengths for spins in the disordered module becomes weaker as a result of frequent switchings of their orientations . as a result the effect of the external field becomes dominant , which causes the spins in the disordered module to align with it . as spins in both modules are now aligned parallel to each other , adaptation of their interactions with time would eventually make all interactions positive , thereby merging the two modules into one .
the time at which this happens is indicated by a transient rise in the frustration ( figs .
[ eps1e-2 ] and [ eps1e-3 ] ) , occurring much later for slower rates of adaptation .
the different scenarios we observe at fast and slow learning rates suggest that there is a competition between two processes : if the modular membership of the spins can change rapidly following initiation of link adaptation so that the system adopts a new structurally balanced configuration , the interaction strengths of the spins in the disordered module do not decrease significantly - thereby allowing the two modules to coexist with the chimera ordering intact ( although the modules now have very different sizes ) .
note that the modular membership of the spins will change rapidly as they continually change their interaction strengths because of the fast adaptation rate .
however , the average number of spins in each module will remain fairly steady . if the learning rate is slow , the interaction strengths will not have time to change sufficiently rapidly to allow the system to reach a new structurally balanced state following the initiation of link adaptation dynamics . as a result the field dominates over the weakened interactions of the spins in the disordered module , thereby eliminating chimera order and eventually causing the two modules to merge .
the existence of chimera order even in situations where the two modules have very different sizes is a novel observation as earlier it had only been observed in a system with symmetric module sizes @xcite . in order to see how the region in the field - temperature parameter space where we observe chimera ordering varies with asymmetry in the module sizes ,
we show in fig .
[ ratio ] the dependence of the order parameter @xmath57 on the strength of the external field ( @xmath1 ) , the scaled temperature ( @xmath58 ) and the ratio of sizes of the two modules ( @xmath59 ) in the absence of link adaptation dynamics .
we observe that the region in parameter space where we find chimera order ( higher values of @xmath57 ) increases significantly as we go towards higher module size ratios ( i.e. , @xmath59 ) especially for lower values of temperature . )
parameter space shown for different ratios of the sizes of the two modules .
chimera is indicated in terms of high values of the order parameter @xmath57 ( represented by different colors ) . ]
we now consider the situation when the external field is withdrawn while allowing the link adaptation dynamics to continue .
a difference is expected with the results shown in the earlier work @xcite where the effect of link adaptation on the evolution of a system subject to thermal fluctuations was investigated , as the learning rule is different on account of the asymmetry in the strengths allowed for negative and positive interactions . as the adaptation dynamics used here allows for much stronger negative interactions ( as compared to positive ) , we would expect the system to become more frustrated as negative interactions are much more likely to contribute frustrated triads .
this is indeed what is observed for slow learning rates , e.g. , @xmath60 and @xmath55 ( fig .
[ off_eps1e-1 - 2 ] ) .
note that in a frustrated system , one can not define modules in a meaningful way , and thus the distinction between modules indicated in the time - series for magnetization is an artifact of the measurement method . for both of the slow learning rates we observe that the system converges to a fully frustrated state corresponding to the maximum value of the measure of frustration in the system ( @xmath61 ) .
, before and after an external field is withdrawn ( the time at which the field is switched off is indicated by the broken line ) .
we observe that the system becomes completely frustrated on withdrawal of the field .
( b ) the corresponding time - evolution shown for link adaptation rate @xmath55.,title="fig : " ] , before and after an external field is withdrawn ( the time at which the field is switched off is indicated by the broken line ) .
we observe that the system becomes completely frustrated on withdrawal of the field .
( b ) the corresponding time - evolution shown for link adaptation rate @xmath55.,title="fig : " ] for faster link adaptation rates , however , we observe unexpected behavior in the system . for a learning rate @xmath62 [ fig .
[ off_eps1.3 ] ] , which is only slightly higher than @xmath53 [ shown in fig .
[ off_eps1e-1 - 2 ] ( a ) ] , the system initially becomes fully frustrated after the external field is withdrawn .
surprisingly , after a period of @xmath63 mc steps during which the system remains frustrated , it suddenly converges to a balanced state .
this intervening period between the field being switched off and the system spontaneously splitting into two modules having positive interactions among all elements within them ( and correspondingly , negative interactions between elements belonging to different modules ) becomes even shorter with increasing learning rates ( e.g. , see fig . [ off_eps1.5 ] for @xmath64 ) .
thus , we can identify a critical value of around @xmath65 for the link adaptation rate @xmath27 , below which the system remains frustrated and above which the system converges to a balanced state , on withdrawal of the external field .
indeed our simulations show that at @xmath62 there is a wide variation in the time required for the system to converge to balance after the field is switched off . , before and after an external field
is withdrawn ( the time at which the field is switched off is indicated by the broken line ) .
we observe that the system initially becomes completely frustrated on withdrawal of the field .
however after about @xmath66 mc steps the system suddenly becomes balanced . ] , before and after an external field is withdrawn ( the time at which the field is switched off is indicated by the broken line ) . we observe that the system quickly becomes balanced after a brief transient rise in frustration following the withdrawal of the field . ] in the previous study of structural balance @xcite , it had been observed that the time required to converge to a structurally balanced state exhibits a bimodal distribution for a range of temperatures .
we observe similar features in our model also - for instance , for a link adaptation rate of @xmath62 . for slightly slower learning rates ( e.g. , @xmath67 ) , the convergence time increases significantly . in our simulations
, many realizations did not converge to the balanced state within the time of observation ( @xmath68 mc steps ) . for higher learning rates ,
the system rapidly converges to balance .
however , unlike the convergence to balance seen in the previous study @xcite , the modules in the structurally balanced state that our model system reaches are of almost equal size ( see figs .
[ off_eps1.3 ] and [ off_eps1.5 ] ) independent of the initial state of the network ( including the initial distribution of interactions which was seen to affect the nature of the balanced state that the system converges to in the earlier work ) .
this may appear surprising as starting from a balanced state , one expects that the network will remain in that state as the interactions adapt so as to make the corresponding energy minimum even deeper .
however , what we observe is that , on withdrawing the field the balanced state characterized by asymmetric module sizes is destabilized and the energy minimum moves in the configuration space eventually reaching the region corresponding to balanced state with modules having equal size .
this difference between our results and that of the previous study suggests that the adaptation rule used here ( which is biased in favor of negative interactions ) is responsible for an intriguing meta - dynamics of the energy landscape itself .
we have also observed how the structure of the energy landscape underlying our system evolves as the interactions change through the adaptation rule . in the model studied earlier @xcite , the frustration of the initially disordered system is around 0.5 .
on initiating link adaptation , this value decreases as the system tends towards balance . when frustration reaches a value around 0.49 , we observe that the energy landscape becomes such that starting from any spin configuration one can reach the same minimum energy spin arrangement .
this suggests that a small decrease in the frustration can accompany a a very large change in the basin of attraction of an energy minimum ( which now encompasses a significant fraction of the configuration space ) .
we observe similar behavior in our model system where the adaptation occurs in the presence and subsequent absence of an external field .
for example , for an adaptation rate of @xmath69 , when the field is withdrawn the system initially becomes completely frustrated with the value of frustration measured as around 0.97 .
the corresponding energy landscape has a very large number of minima , each having very small basins of attraction .
as the system evolves towards a balanced state , we note that even when the frustration decreases by a very small amount , e.g. , to 0.93 , the energy landscape transforms to one having an extremely deep energy minimum with a very large basin of attraction .
we believe that this radical transformation of the energy landscape of the system at the initial stages of approach to balance points to features of landscape evolution that deserve further study .
in this paper we have explored the behavior of a spin system that is subjected to an external field , while the interactions adapt in response to the system dynamics .
the results obtained through our simulations that is described above is summarized in fig .
[ schematic ] . from an earlier study
, we had known that introducing link adaptation inspired by the hebbian principle in a frustrated system can result in it converging to a structurally balanced state - corresponding to evolution of the underlying energy landscape - that depends on the initial state of the system . applying this adaptation dynamics to a system
exhibiting chimera order ( as done in here ) allows us to explore how the dynamics of the energy landscape is affected by an external field , as well as , asymmetry in the adaptation rule regarding negative and positive interactions .
we observe that when system is adapting through faster learning rates , the chimera order is maintained throughout its evolution , with the ordered module increasing in size at the expense of the disordered module . for slower learning rates ,
first the interactions of the elements belonging to the disordered module become weak which results in loss of chimera order .
this is followed by the system converging to a fully ordered state where almost all the interactions are positive .
the sizes of the two modules defining the system where one observes chimera order can be varied to see how the ratio of module sizes can affect the subsequent evolution of the system .
indeed , we see that the region in the field - temperature parameter space over which chimera order is seen increases as the ratio increases from 1 .
when the external field is withdrawn , the system returns to a structurally balanced state with modules of similar sizes if the adaptation rate is high .
however , for slower learning rates , the system becomes completely frustrated .
we have also tried to explore how the structure of the energy landscape of the system evolves during learning .
starting from a chimera ordered state that is subjected to adaptation rate and external field , we observe changes in the landscape after the field is switched off . initially the system becomes completely frustrated .
however , after an interval ( which decreases with increasing learning rate ) the system shows a sudden large increase in the basin of attraction of an energy minimum , although the frustration of the system has decreased only negligibly .
this surprisingly radical change in the landscape resulting from relatively small degree of change in the interaction structure of the network is a question that needs to be explored further .
it is of interest to consider the implications of the results reported here for adaptation in biological systems . as connections in the brain
evolve according to long - term potentiation that embodies the hebbian principle that has inspired the link adaptation dynamics used in our study , it is natural to expect that the brain may exhibit at least some of the features observed here .
indeed , as mentioned earlier , the experimental observation of two dynamically anti - correlated subnetworks in the resting human brain @xcite strongly suggests that in the absence of strong external stimulation the underlying network is structurally balanced . on the other hand , when exposed to stimuli , correlated brain activity is indeed observed - although not encompassing the entire network .
this is , to some extent , reminiscent of the chimera ordered state that is seen in our model system .
thus , the brain may be seen as corresponding to a system subject to relatively fast adaptation rate .
however , the hebbian principle can apply to a much broader class of biological systems , e.g. , gene regulation networks where the co - expression of genes has been suggested to result in their co - regulation over evolutionary time - scales @xcite . as the adaptation in this case
is provided by natural selection , which is orders of magnitude slower than the learning process in the brain mentioned earlier , it is probably not unreasonable to conclude that this can be seen as one corresponding to our model system subject to slow rate of adaptation .
it is easy to see that the frustrated system with a large number of energy minima can be considered to be analogous to the cellular differentiation process that allows convergence to any one of a large number of possible cell fates dependent upon initial conditions .
indeed this analogy has been used earlier by kauffman to motivate boolean network models for explaining differentiation during biological development @xcite . extending this analogy
, one may wonder whether the system has a state corresponding to the ordered state seen on exposing it to an external stimulus . as this state is unique and
will be attained by the system independent of all initial conditions , it is tempting to suggest that induced pluripotency in cells exposed to chemical stimuli @xcite may be the biological analogue .
therefore , it may be of interest to study the phenomena reported here in biologically realistic models of networks adapting at different time scales .
we thank rajeev singh , shakti n. menon and r. janaki for helpful discussions .
this work was supported in part by imsc complex systems project ( xii plan ) funded by the department of atomic energy , government of india .
the numerical work was carried out in machines of the imsc high - performance computing facility , including `` satpura '' which is partially funded by dst ( grant no . sr / nm / ns-44/2009 ) . | many complex systems can be represented as networks of dynamical elements whose states evolve in response to interactions with neighboring elements , noise and external stimuli .
the collective behavior of such systems can exhibit remarkable ordering phenomena such as _ chimera order _ corresponding to coexistence of ordered and disordered regions .
often , the interactions in such systems can also evolve over time responding to changes in the dynamical states of the elements . link adaptation inspired by hebbian learning
, the dominant paradigm for neuronal plasticity , has been earlier shown to result in structural balance by removing any initial frustration in a system that arises through conflicting interactions . here
we show that the rate of the adaptive dynamics for the interactions is crucial in deciding the emergence of different ordering behavior ( including chimera ) and frustration in networks of ising spins .
in particular , we observe that small changes in the link adaptation rate about a critical value result in the system exhibiting radically different energy landscapes , viz . , smooth landscape corresponding to balanced systems seen for fast learning , and rugged landscapes corresponding to frustrated systems seen for slow learning . |
there is currently interest in exploiting electron spin for nano - spintronic@xcite and quantum information processing@xcite .
this is partly motivated by electron spin long coherence times @xcite and availability of scalable semiconductor technology . in the simplest approach ,
a physical qubit is identified with the two states of an electron spin , which can be manipulated by applying local magnetic fields .
much progress has been achieved using micro - magnet@xcite technology with electron spin qubits .
an alternative approach is to encode a logical qubit in a two level system constructed with spin complexes@xcite .
this includes singlet - triplet two electron qubit@xcite and a logical qubit encoded in a degenerate ground state of a three spin complex@xcite .
as suggested by divincenzo et al.@xcite , in the framework of a heisenberg model for a logical qubit encoded in a linear chain of three spins , such a logical qubit can be manipulated by the control of the exchange interaction between pairs of spins . the ability to manipulate spin state with a voltage
is related to the relation between the orbital and spin part of the many - electron wavefunction . a detailed microscopic model of a three electron
complex in a single semiconductor quantum dot ( qd ) , exploring the orbital and spin relation , has been investigated and compared with experiment by one of us@xcite .
a similar microscopic model of three electrons localized in a lateral gated triangular triple quantum dot molecule in the plane of a gaas / gaalas heterojunction was proposed and explored by one of us in ref . .
the coded qubit was identified with chirality of three electron spin complex , or equivalently , two possible directions of a minority spin motion . other proposals to implement coded qubits with both triple quantum dots@xcite ( tqds ) and atom traps@xcitealso exist .
the advantages of working with a tqd - based coded qubit are two - fold .
first , every quantum gate can be implemented electrically .
in such a scheme , magnetic field will only be used for initialization and ( or ) measurement of a coded qubit .
second , the coded qubit involves decoherence free subspace@xcite ( dfs ) and is immune to channels of collective decoherence .
this reduces decoherence of a tqd - based coded qubit due to charge fluctuations as discussed in ref . .
other interesting phenomena involving tqd molecules include non - fermi liquid behavior when coupled to leads@xcite , potential for generating maximally entangled three - partite ghz and w states@xciteand manipulation of total spin@xcite . in this work
we develop a microscopic theory of quantum circuits based on coded qubits encoded in chirality of electron spin complexes in tqd .
we use a combination of linear combination of harmonic orbitals - configuration interactions ( lcho - ci)@xcite , hubbard and heisenberg models to determine a set of optimal conditions for single qubit operations and describe the two qubit gate .
we show that there exists an in - plane magnetic field direction and magnitude optimal for single qubit operations . however , the magnetic field will rotate qubit states in an undesirable way ; and we show how this rotation can be controlled by tuning voltages on the gates .
once the optimal sets of magnetic fields and voltages have been obtained , the exact diagonalization techniques are used to establish and verify an effective two - qubit hamiltonian .
it is shown that the effective two qubit interaction is ising - like , leading to a two - qubit phase gate .
the present work establishes both single qubit and two - qubit operations necessary for performing quantum computation using tqd - based coded qubits .
the paper is organized as follows .
[ sec : intro ] contains introduction
[ sec : sys ] describes the quantum circuit based on chirality of 3 electron spin complexes and computational methodology . in sec .
[ sec : cqbt ] , we present a definition of a tqd - based coded qubit , discuss initialization , single qubit operations using voltages , and measurement of the state of the coded qubit . in sec . [
sec : dqop ] , we present an effective hamiltonian of 2 coupled coded qubits and show that it can be canonically transformed to an ising interaction .
[ sec : con ] contains summary .
fig.([fig : layout]a ) shows schematically layout of quantum circuits based on coded qubits encoded in chirality of electron spin complexes in triangular tqds .
the circles denote individual lateral quantum dots formed in the 2d electron gas ( 2deg ) at the heterojunctions of algaas / gaas by metallic gates on the algaas surface .
the gates are set to confine a single electron in each dot denoted by an arrow .
additional gates ( not shown here ) are used to control tunneling between dots in the same tqd molecule , shown schematically as solid lines .
dashed lines indicate tunneling between neighboring tqds .
the tunneling between any pair of qds is responsible for exchange interaction of electron spins localized on each qd . the brackets indicate two tqds isolated from the rest of the circuit . it is assumed that any number of tqds can be isolated from the rest of the circuit by turning off the boundary exchange interactions with gate voltages .
for comparison , fig.([fig : layout]b ) shows the circuit composed of linear , instead of triangular , tqds , similar to the linear chain of spins proposed by divincenzo et al . , in which the coded qubits in the chain are implemented with only two exchange interactions .
fig.([fig : layout]c ) is another possible architecture studied by weinstein@xcite et .
al . for spin system
however , in this design , bringing tqds close together induces interactions between quantum dots beyond the ones indicated by dashed line . in this work
we focus mainly on triangular tqds - based quantum circuit shown in fig.([fig : layout]a ) .
since electrons are well localized in each qd , a system of two tqds in a chain can be very well described by an extended hubbard model@xcite . with @xmath0 electron creation ( annihilation ) operator for the electron with spin @xmath1 on the @xmath2 qd
, the hubbard hamiltonian reads : @xmath3 where @xmath4 labels the two tqd molecules from left to right , indices @xmath5 range from @xmath6 to @xmath7 label each qd from left to right in the @xmath8 tqd molecule .
the intra - tqd hubbard parameters @xmath9 , @xmath10 , @xmath11 and @xmath12 are the tunneling matrix element between the @xmath2 and the @xmath13 qds , the off - site coulomb interaction parameter between any two qds in the same tqd molecule , the on - site coulomb interaction strength for any qd , and the on - site energy for the @xmath2 qd respectively .
the on - site energy depends on spin and magnetic field which is applied in the plane of a tqd molecule .
the hubbard parameters @xmath14 and @xmath15 represent the inter - molecular tunneling matrix element and inter - molecular coulomb interactions .
@xmath16 is the number operator for the @xmath17 electron on the @xmath2 qd .
@xmath18 is the electron charge density operator on the @xmath2 qd .
when we discuss triangular resonant tqds , we drop the subscripts on all the parameters . in our model , we consider the parameters corresponding to regime of strong correlations : @xmath19 . the intra - tqd hubbard parameters , @xmath20 , @xmath21 , @xmath10 , and @xmath11 , in eq.([eq : nhubbard ] ) are obtained from a microscopic calculation for single tqd based on lcho - ci method as explained in ref . .
the inter - tqd hubbard parameter @xmath15 is taken to be the direct coulomb interaction between 2 charges localized on adjacent edge dots of two neighboring tqd molecules . the energy spectrum and eigenstates of the hubbard hamiltonian for one and two tqd molecules
are obtained using configuration interaction technique . for a given number of electons @xmath22
we construct all possible configurations @xmath23 , build hamiltonian matrix in the space of configurations , and diagonalize it numerically . at half - filling
, the low - energy spectrum of the hubbard model can be approximated by a spectrum of a heisenberg model@xcite describing electron spins localized in each dot : @xmath24 the exchange interactions , @xmath25 , for the tqds can be expressed in terms of tunneling matrix elements and quantum dot energies : @xmath26 the exchange interaction can be controlled by either tuning the tunneling matrix element @xmath9 by , for example , additional gates controlling the height of the tunneling barrier , or by biasing the dots and changing their on - site energy @xmath27 .
in this section , we discuss a single coded qubit shown in fig.([fig : bfield]a ) : its preparation , initialization , operation and measurement in the presence of a lateral magnetic field in the @xmath28 direction .
the qubit is encoded in quantum states of a three electron spin complex in a fully symmetric and half - filled tqd .
an example of one of the three possible configurations @xmath29 with @xmath30 is shown schematically in fig.([fig : bfield]a ) .
the three configurations with the minority spin on qd 1 , 2 or 3 form a doubly degenerate ground state with total spin @xmath31 and an excited state with total spin @xmath32 separated from the ground state by @xmath33 , as shown in fig.([fig : bfield]b ) . the doubly degenerate ground state with fixed @xmath30 forms an effective two level system .
we identify the coded qubit levels @xmath34 and @xmath35 with @xmath36 where @xmath37 and @xmath38 .
the three spin complex is characterized by chirality @xmath39 which measures the degree of collinearity of the three spins .
the two coded qubit levels , eq.([eq : chiralstates ] ) , are the eigenstates of chirality operator with eigenvalues @xmath40 .
these two states also can be more intuitively characterized by minority spin moving either to the left or to the right as in resonant valence bond ( rvb ) plaquette @xcite .
we note that in the absence of the magnetic field the coded qubit states in @xmath41 subspace are degenerate with coded qubit states in @xmath30 subspace .
if we are to work in the computational space corresponding to @xmath30 , any process which flips the electron spin will remove the coded qubit from its computational space . in order to separate the computational hilbert space @xmath30 from @xmath42 subspace
, a magnetic field @xmath43 is applied along the @xmath28 direction as shown in fig([fig : bfield]a ) .
we avoid applying magnetic field applied in the @xmath44 direction , because it activates undesirable , higher order spin - spin interactions@xcite , whereas the magnetic field applied in the plane only modifies the on - site energies of qds : for @xmath43 field , qds 1 and 3 energy level as @xmath45 , where @xmath46 is the spatial separation of qds 1 and 3 and @xmath47 is the cyclotron frequency , while the qd 2 energy level only acquires the zeeman term . hence magnetic field in the @xmath28 direction
effectively lowers the energy of the qd 2 by @xmath48 with respect to qds 1 and 3 .
fig.([fig : bfield]b ) shows the energy spectrum of the coded qubit as a function of @xmath49 obtained in lcho - ci method . in our lcho - ci calculation , we use the following parameters for a specific resonant tqd . the inter - dot distance is 10 @xmath50 , where @xmath51 is the effective bohr radius for @xmath52 .
the confining gaussian potential on the @xmath2 qd is of the form , @xmath53 , where @xmath54 and @xmath55 .
@xmath56 is the effective rydberg for @xmath52 . in fig.([fig : bfield]b ) , the optimal cyclotron frequency @xmath57 corresponds to @xmath58 . at @xmath59 , the four - fold degenerate ground state
is separated from the four - fold degenerate spin polarized excited state by @xmath33 .
the two computational hilbert spaces corresponding to @xmath60 separate energetically with increasing magnetic field while the spin polarized states decrease in energy . at magnetic field @xmath61 ,
such that @xmath62 , the energy gap is maximized and should correspond to the working point that can maintain the longest coherence time of the coded qubit .
the coded qubit should operate at this value of magnetic field .
however , as discussed above , the magnetic field effectively biases the qd 2 .
this removes the degeneracy of the two qubit levels and rotates them from their zero magnetic field states .
the energy splitting of the two coded qubit levels as a function of applied magnetic field is shown in fig.([fig : gate]a ) .
the splitting is a fraction of the large energy scale @xmath33 . for the largest gap ,
i.e. @xmath63 , the splitting is @xmath64 . in order to restore the degeneracy of the two coded qubit levels
, one can apply voltage to the qd 2 .
in fig.([fig : gate]b ) , microscopic calculations done in lcho - ci show that a positive voltage bias on the qd 2 can indeed bring the tqd back on resonance in the presence of @xmath65 field .
we have now established the coded qubit and the best conditions for its operation . in order to operate the coded qubit
, we need to be able to initialize it .
we propose to initialize the coded qubit by turning off both interaction between qds 1 and 3 and interaction between qds 1 and 2 .
the only remaining interaction is between qds 2 and 3 .
the ground state of a tqd , @xmath66 , becomes a product of a spin down state of an electron on the qd 1 and a singlet state of electrons across qds 2 and 3 .
this intuitively is a ground state in magnetic field .
the singlet state of two electrons in a pair of qds 2 and 3 can be generated in real time starting from two electrons in a biased qd 3 .
this procedure does not generate directly the coded qubit levels @xmath67 and @xmath35 , but the state @xmath68 is a linear superposition of the two qubit levels : @xmath69 . once interactions are turned on , any state can be obtained from the initial state .
this can be seen by writing the tqd heisenberg hamiltonian in the basis of the two coded qubit levels @xmath70 , @xmath71 if we take @xmath72 , and let @xmath73 , then the heisenberg hamiltonian corresponds to @xmath74 operation .
if we take @xmath75 and @xmath76 , then the coded qubit hamiltonian eq.([eq : singlecodedqubit ] ) corresponds to @xmath77 operation .
the capability to rotate a qubit with respect to two different axes on a bloch sphere allowes us to generate arbitrary single qubit operations . in practice , we tune the exchange interaction @xmath25 through biasing qds and changing their energies . as long as the biasing @xmath78 is satisfied , we expect the quantum state to remain in the qubit subspace during the process of tuning the exchange interactions as discussed in ref . .
next , we discuss the measurement of coded qubits .
several proposals@xcite discusses methods of detecting chirality of a triangular ( three - body ) antiferromagnetic cluster . for our specific tqd - based coded qubit
, we propose to apply a @xmath79 field to split the two coded qubit levels with different chirality .
as discussed earlier , the application of @xmath79 field will give rise to additional terms proportional to chirality operator on top of the heisenberg model , eq.([eq : heis1 ] ) , used for qubit modeling .
since the additional term to the heisenberg model is strictly proportional to the chirality , it splits the two coded qubit levels with the gap given by @xmath80 where @xmath81 is the magnetic flux through the device and @xmath82 is the flux quanta .
this energy gap can be interpreted as the energy difference between two magnetic dipole moments@xcite oriented in opposite directions under @xmath79 . in ref . , scarola et .
al . also showed that the unintended qubit rotation due to the additional chirality term can be significantly reduced for a range of specific weak @xmath79 fields that satisfy @xmath83 and @xmath84 , where @xmath85 is the modified magnetic length , and @xmath86 is the confining frequency used to approximate the potential of a qd .
thus , initializing the system under @xmath79 might not be a good idea as the field has to stay on for the entire period of quantum computation and eventually modify the quantum state of the coded qubit due to magnetic moment coupling with @xmath79 .
however , a measurement done with advanced spectroscopy only requires @xmath79 to be turned on for a relatively short time and significantly limit the extent to which the coded qubit will be modified .
we now turn to discuss double coded qubit operations using both hubbard and heisenberg models . starting with the hubbard hamiltonian , eq.([eq : nhubbard ] ) , we derive perturbatively a heisenberg hamiltonian for a complex of six electron spins with a set of exchange interactions @xmath25 @xmath87 with all other exchange interactions set to zero . for all the hubbard parameters considered in this study ( i.e. weak inter - tqd tunneling and strong coulomb interaction ) , @xmath88 is usually about two order of magnitudes smaller than @xmath89 and @xmath90 , mainly due to the fact that @xmath91 in our study .
we treat the inter - tqd exchange interaction , scaled by @xmath88 , perturbatively to derive an effective interaction in the coupled coded qubits subspace , @xmath92 , where the superscript @xmath93 and @xmath46 stand for left and right tqd respectively .
the effective qubit - qubit interacting hamiltonian reads @xmath94 where @xmath95 and @xmath96 and the direction vectors in eq.([eq : codedqubit4 ] ) are defined as follows : @xmath97 .
thus , the effective interaction for 2 tqd - based code qubits is equivalent to an @xmath98 hamiltonian under a uniform in - plane magnetic field for 2 spins .
if we rotate the effective qubit - qubit interacting hamiltonian , eq.([eq : codedqubit4 ] ) , from the coded qubit basis to the jacobian basis , @xmath99 , the rotated hamiltonian reads @xmath100 where @xmath101 , and @xmath102 .
the jacobian state @xmath103 is defined in sec.[sec : cqbt ] when we discuss the initialization of the coded qubit .
the jacobian state @xmath104 , where @xmath105 is a @xmath106 triplet state of electrons on qds 2 and 3 and @xmath107 is a @xmath108 triplet state of electrons on qds 2 and 3 .
similarly , @xmath109 can also be written as a linear combination of the coded qubit levels .
therefore this rotation of basis only requires single qubit operations applied to each coded qubit ; the two hamiltonians , eq.([eq : codedqubit4 ] ) and eq.([eq:2qubitl ] ) , are locally equivalent@xcite . as ising interaction
can be used to generate cnot gate@xcite , we show how to generate non - local two qubit interactions with 2 tqd - based coded qubits .
fig.([fig : ising ] ) shows results of exact diagonalization of the hubbard hamiltonian for two coupled tqds as a function of increasing coupling @xmath14 between the two molecules .
the inset shows the entire energy spectrum at @xmath110 .
the low energy spectrum consists of four levels , characterized mostly by the two lowest levels of each tqd .
these 4 low - lying energy levels are well separated from the rest of the spectrum and constitute the two coupled coded qubit subspace .
the low - lying spectrum contains a doubly degenerate level , which is a signature of the ising model in an external field . in practice , the tuning of @xmath88 , the inter - tqd exchange interaction ,
should be done via tuning the tunneling parameter @xmath14 .
( [ fig : ising ] ) shows that the ising model features of the energy spectrum are maintained over a wide range of values of @xmath14 , and hence the ising model hamiltonian , derived above , describes the coupling of two coded qubits very well .
in summary , we present a theory of quantum circuits based on coded qubits encoded in chirality of electron spin complexes in lateral gated semiconductor triple quantum dot molecules with one electron spin in each dot . using microscopic hamiltonian and exact diagonalization techniques
we show how to initialize , coherently control and measure the quantum state of a chirality based coded qubit using static in - plane magnetic field and voltage tuning of individual qds .
the microscopic model of two interacting coded qubits is established and mapped to an ising hamiltonian . hence both conditional two - qubit phase gate and voltage controlled single qubit operations
are demonstrated .
the authors thank nserc , quantumworks , cifar , nrc - cnrc crp and ogs for support .
hsieh would like to thank y .-
shim and a. sharma for useful discussions .
subspace under a @xmath43 field .
the @xmath43 field breaks the discrete rotational symmetry , and it affects qd 2 differently from qds 1 and 3 .
the energy spectrum of a triangular tqd as a function of @xmath111 , where @xmath112 is the cyclotron frequency used in lcho - ci calculation .
@xmath57 corresponds to a magnetic field @xmath61 such that the gap between computational subspace ( the lowest energy level in the plot ) and the rest of spectrum is maximized as indicated by the black arrow . , scaledwidth=90.0% ] .
the @xmath43 field splits the doubly degenerate ground state , and @xmath64 corresponds to the gap between the two levels at @xmath57 or equivalently at @xmath113 .
the energy spectrum of the qubit levels under @xmath113 as a function of voltage @xmath114 on qd2 , and @xmath115 is the unbiased voltage at resonant condition when there is no external @xmath65 field .
the splitting of levels under @xmath65 field can be restored via gate voltage tuning.,scaledwidth=70.0% ] .
the energy spectrum resembles that of the ising model with an external field : a doubly degenerate levels corresponding to state @xmath116 and @xmath117 and two unique levels @xmath118 and @xmath119 .
`` @xmath120 '' denotes a doubly degenerate level .
inset : the entire energy spectrum at @xmath121 calculated with lcho - ci method.,scaledwidth=90.0% ] | we present a theory of quantum circuits based on logical qubits encoded in chirality of electron spin complexes in lateral gated semiconductor triple quantum dot molecules with one electron spin in each dot . using microscopic hamiltonian
we show how to initialize , coherently control and measure the quantum state of a chirality based coded qubit using static in - plane magnetic field and voltage tuning of individual dots .
the microscopic model of two interacting coded qubits is established and mapped to an ising hamiltonian , resulting in conditional two - qubit phase gate . |
the electron properties of two - dimensional ( 2d ) graphene , a single - layer carbon sheet , has attracted much attention by both theoreticians and experimentalists ( see @xcite and references therein ) . along with this related structures , namely graphene nanoribbons ,
are also under intensive investigation @xcite .
one of the reason for this is that the long electron mean free path in graphene up to 1@xmath0 m opens a field of carbon - based nanoelectronics , where gnrs are used as interconnects in nanodevices .
the unique electron mobility in graphene structures is caused by the strong bonding between the carbon atoms in the honeycomb lattice of graphene .
this in turn prevents the replacing of the carbon atoms by alien ones .
nevertheless , graphene is not immune to extrinsic disorder and its transport properties @xcite are very sensitive to impurities and defects @xcite .
the theoretical problem of an impurity in 2d graphene was considered originally in @xcite . in the vicinity of the dirac points in @xmath1 space , which are peaks of the double cones of the fermi surface , the low - energy electronic excitations in gapless graphene are described by the equation of the effective mass approximation , which is formally identical to the 2d dirac equation for a massless neutrino , having the fermi speed @xmath2 . in the presence of an attractive impurity centre of charge @xmath3 screened by a medium of the effective dielectric constant @xmath4 the electron states are drastically different for the subcritical @xmath5 and supercritical @xmath6 regions of the strength of the coulomb interaction , where @xmath7 and @xmath8 are the 2d momentum of the impurity electron and that , having the speed @xmath9 , respectively .
clearly , the super- and subcritical regimes can be reached for the dimensionless coulomb potential strength @xmath10 for @xmath11 , respectively .
the difference of the subcritical and supercritical electron states is caused by their different behaviour in the vicinity of the impurity centre @xmath12 .
the subcritical regime admits regular solutions to the dirac equations , while the wave functions corresponding to the supercritical case oscillate and do not have any definite limit .
the physical reason for this is that at the subcritical strengths @xmath13 the centrifugal potential barrier prevents the electron `` fall to the centre '' @xcite , while the supercritical strengths @xmath14 provide the collapse .
clearly , the continuum approach based on the dirac formalism becomes inapplicable .
the lattice - scale physics dominates that in turn requires a regularization procedure , namely the cutoff of the coulomb potential at short distances @xmath15 , where @xmath16 is the c - c distance in graphene .
the physics of the supercritical impurity electron in graphene @xcite closely resembles that of the relativistic electron in an atom having the nuclear charge @xmath17 @xcite .
since , as it follows from below , only the supercritical regime is relevant to the impurity state in gnr , we focus on this case .
numerical and analytical approaches developed on the tight - binding model of the graphene lattice and of the dirac equation , subject to the regularization procedure , respectively , undertaken originally by pereira @xcite et.al .
have revealed the infinite number of the quasi - bound states , having the finite width , arising , as it was shown quasi - classically @xcite , from the collapsed states . if the requirement of the regularity of the wave functions in the vicinity of the source of the electron attraction is to be replaced by the less rigorous condition of its square integrability an infinite number of the strictly discrete energy levels were found to occur .
the coulomb potential cutoff in the gapped @xcite and no cutoff in the gapless graphene @xcite induce the energy series bounded and unbounded from below , respectively . in the gnr , which in principle can be treated as a quasi-1d structure
, we can expect completely different results .
the strictly discrete bound states regular at the impurity centre @xmath18 are realized without the regularization procedure , in particular without the cutoff of the coulomb potential , preventing the collapse .
consequential concerns lie in the well known fact that the reduction of the dimension of the structure increases the stability of the impurity electron . in units of the impurity
rydberg constant @xmath19 the binding energy @xmath20 of the impurity electron in 3d bulk material is @xmath21 , in the narrow 2d quantum well @xmath22 @xcite , and in the thin quantum wire of radius @xmath23 much less than the impurity bohr radius @xmath24 , @xmath25 @xcite . besides , an extremely weak 3d atomic potential not providing bound electron states , transforms in the presence of a magnetic field into a quasi-1d system binding the electron @xcite .
it is relevant to note that these atomic states arise under as weak as one likes magnetic fields i.e. the as large as one likes magnetic lengths playing the same role as the width of the gnr .
note that the confinement attributed to the semiconductor thin films @xcite replaces the 3d coulomb potential @xmath26 by the effective 2d potential of the weaker singularity of the logarithmic character .
it seems that in the quasi-1d gnr the effect of the attenuation of the potential singularity preventing the fall to the centre exceeds that of the vanishing of the 2d centrifugal potential barrier promoting the collapse . clearly
a study of the impurity electron state in graphene structures is important on account of two reasons .
first , these structures provide a realization in solid state physics of remarkable effects of quantum electrodynamics caused by a large `` fine structure constant '' @xmath27 @xcite .
second , we expect a strong impact of impurities on the electronic systems not only for 2d graphene layers possessing an outstanding high electron mobility @xcite but in particular for impurity gnrs whose properties are not widely addressed in the literature yet .
brey and fertig @xcite have shown that the energy spectrum of the electron in an armchair gnr bounded in @xmath28-direction is the sequence of the subbands formed by the branches of the continuous energies of the longitudinal unbounded @xmath29-motion emanating from the size - quantized energy levels @xmath30 , ( @xmath31 is the discrete lable ) , reflecting the ribbons @xmath28-confinement .
the equation for the components @xmath32 of the dirac spinor relevant to the a and b sublattices of the graphene for the electron positioned far away from the impurity centre has the form @xmath33 showing that the armchair gnr manifests itself as gapped structure entailing the bound and unbound impurity states for the energies @xmath34 , respectively . of special interest
is the narrow gnr of width @xmath35 for which @xmath36 , where @xmath37 is the radius of the bound electron state , being induced by the ribbon confinement @xmath35 .
such a gnr provides the expected electron binding energy @xmath38 , where @xmath39 is a some function vanishing at @xmath40 , which is of interest and attractive because of two aspects . on the one hand the ribbon provides a considerable impurity binding energy which could be measured experimentally and on the other hand the impurity potential can be treated perturbatively and an analytical approach to the problem becomes feasible .
a schematic form of the potentials @xmath41 provided in eqs .
( [ e : coulomb ] ) , ( [ e : pot ] ) , ( [ e : limit ] ) at @xmath42 and quasi - discrete @xmath43 ( [ e : rydberg ] ) and continuous @xmath44 ( [ e : outercont ] ) spectra adjacent to the ground @xmath45 ( discrete states ) and first @xmath46 and second @xmath47 size - quantized levels @xmath48 ( [ e : phi ] ) in the gnr of width @xmath35 . ]
a comment concerning the from of the energy spectrum is in order . in the zeroth approximation of isolated size - quantized @xmath31-subbands
i.e. in the single - subband approximation the slow longitudinal motion parallel to the boundaries is governed by the 2d coulomb potential averaged with respect to the `` fast '' transverse @xmath31-states .
the energy spectrum consists of sequence of series of quasi - coulomb discrete @xmath49-levels and continuous sub - bands positioned below and above , respectively relatively to the size - quantized energy levels @xmath48 ( see fig .
1 ) . only the series of the impurity energy levels @xmath50 adjacent to the ground size - quantized energy level @xmath51 is strictly discrete .
the @xmath49-series adjacent to the excited levels @xmath52 come into resonance with the states of the continuous spectra of lower subbands and in fact in the next multi - subband approximation turn into quasi - discrete resonant states ( fano resonances)@xcite .
the corresponding resonant widths @xmath53 determine the auto - ionization rate and life - time @xmath54 of the resonant impurity states being of relevance to an experimental study . also to our knowledge
an analytical approach to the problem of impurities in gnr providing the explicit dependencies of both discrete and especially quasi - discrete electron states on the width of the gnr @xmath55 and the position of the impurity centre within the gnr are not comprehensively available in the literature . in order to fill the above
mentioned gap we perform an analytical study of the strictly discrete and resonant impurity states in a narrow armchair gnr . the impurity centre is positioned anywhere within the ribbon bound by the impenetrable boundaries .
the width of the gnr is assumed to be much less than the radius of the impurity state .
the complete 2d envelope wave function satisfying the massless dirac equation is expanded with respect to the basis formed by the 1d size - quantized subband wave functions describing the fast transverse motion bound by the boundaries of the gnr .
the generated set of equations for the 1d quasi - coulomb wave functions relevant to the longitudinal slow motion is solved in the single- , two- and three - subband approximations , in which the ground , first and second excited subbands are involved .
the mathematical method is based on the matching of the coulomb wave functions with those obtained by an iteration procedure at any point within the intermediate region bound by the ribbon width and the radius of the created coulomb state .
both the real and imaginary parts of the complex energy levels are calculated in a single procedure .
the dependencies of the binding energy and resonant energy shift and width on the width of the gnr and the position of the impurity centre are obtained in an explicit form .
numerical estimates show that for a narrow gnr the binding energy and the resonant width are quite reasonable , to render the impurity electron states in gnr experimentally observable .
our analytical results are in line with those calculated numerically and revealed in an experiment .
we remark that our aim is to elucidate the physics of the impurity states in gnr by deriving closed form analytical expressions for their properties .
we do not intend to compete with the results of computational studies .
the paper is organized as follows . in section 2 the analytical approach based on the multi - subband approximation
is described .
the real quasi - coulomb functions of the discrete and continuous spectrum and the real energy levels determining the binding energies are calculated in the single - subband approximation in section 3 .
the complex energies including the resonant shift and width associated to the first and second excited subbands are found in the double and three - subband approximation , respectively , in section 4 . in section 5
we discuss the obtained results .
section 6 contains the conclusions .
according to the above an analytical description of the stable and metastable impurity electron states in the narrow armchair gnr is of significant interest .
it elucidates the underlying basic physics of the carbon - based nanodevices , in which the highly mobile electrons remain unbound in the 2d graphene monolayers while in their interconnects , namely in quasi-1d armchair gnrs these electrons are trapped by impurity centres .
the latter could modify the overall transport properties .
we consider a ribbon of width @xmath35 located in the @xmath56 plane and bounded by the lines @xmath57 the impurity centre of charge @xmath3 is displaced from the mid - point of the ribbon @xmath58 by the distance @xmath59 the equation describing the impurity electron at a position @xmath60 possesses the form of a dirac equation @xmath61 where the hamiltonian @xmath62 is given by @xmath63 + v(\vec{\rho})\hat{{\rm i}};\,p=\hbar v;\,v=10 ^ 6~\mbox{m / c}\ ] ] with @xmath64 composed by the hamiltonians relevant to the inequivalent dirac points @xmath65 and @xmath66 ( @xmath67 are the pauli matrixes ) presented originally in ref .
the matrix @xmath68 in ( [ e : basic ] ) is the unit matrix and @xmath69 is the 2d coulomb impurity potential , @xmath4 is the effective dielectric constant related to the static dielectric constant @xmath70 of the substrate by @xcite @xmath71 the envelope wave four - vector @xmath72 consists of two vectors @xmath73 describing the motion of the electron in sublattices @xmath74 and @xmath75 of graphene @xmath76 each determined by the wave functions @xmath77
@xmath78 the total @xmath79 state implies the multiplication of the @xmath77 functions with the factors @xmath80 , respectively .
the boundary conditions for the armchair ribbon require the total wave function to vanish at both edges @xmath81 for both @xmath79 superlattices @xcite @xmath82 where @xmath83 is the graphene superlattice constant .
the basis wave vectors @xmath84 and the energies @xmath48 describing the transverse size - quantized @xmath28-states are derived from equation @xmath85 to obtain @xmath86;~ \vec{\phi}_{na } ( x)= \begin{array}{c } \begin{bmatrix } \varphi_{na}^{(+ ) } \\ 0 \\ \varphi_{na}^{(-)}\\ 0 \end{bmatrix } \end{array } ; ~ \vec{\phi}_{nb } = \begin{array}{c } \begin{bmatrix } 0\\ \varphi_{nb}^{(+ ) } \\ 0 \\ \varphi_{nb}^{(- ) } \end{bmatrix } \end{array}\ ] ] , where @xmath87 with @xmath88 \right ) \right ] \right \ } ; \nonumber\\ \varepsilon_n=|n-\tilde{\sigma}|\frac{\pi p}{d};~n=0,\pm1,\pm2,\ldots~ ; \quad\tilde{\sigma}=\frac{kd}{\pi}-\left [ \frac{kd}{\pi}\right]\end{aligned}\ ] ] we consider transverse states with positive energies @xmath89 in the armchair ribbon of width @xmath35 providing the gaped ( insulator ) structure @xmath90 @xcite .
it follows from eq.([e : phi ] ) that the energy levels @xmath48 as a function of width @xmath35 are the oscillations describing by parameter @xmath91 imposed on the decreasing curve @xmath92 .
below we ignore these oscillations keeping @xmath93 the boundary conditions ( [ e : bound ] ) after substitution of @xmath77 by @xmath94 , respectively , are satisfied . the wave vectors @xmath95 form orthonormal subsets , for which @xmath96 the boundary conditions ( [ e : bound ] ) imposed on the wave vector @xmath97 ( [ e : wavevect ] ) force us to expand the wave vectors @xmath73 in series @xmath98 with respect to the basis functions @xmath99 taking for the coefficients @xmath100 substituting the wave vector @xmath72 ( [ e : wavevect ] ) with the wave vectors @xmath73 and the wave functions @xmath101 into eq .
( [ e : basic ] ) and subsequently using the properties ( [ e : ortho ] ) we obtain by the standard method the set of equations for the wave functions @xmath102 @xmath103 @xmath104,\ ] ] where the potential @xmath105 is given by eq .
( [ e : coulomb ] ) . at @xmath106 @xmath107~;\,s=0,1,2,\ldots ;
\ ] ] as expected in the limiting case @xmath108 eqs .
( [ e : set ] ) decompose into the sets describing the 1d coulomb states , while in the absence of the impurity centre @xmath109 we arrive at the wave functions @xmath110 and the energies @xmath111 of free electrons in the armchair nanoribbon @xcite .
below we solve the set ( [ e : set ] ) in the adiabatic approximation implying the longitudinal @xmath29-motion governed by the quasi - coulomb potentials @xmath112 to be much slower than the transverse @xmath28-motion affected by the boundaries of the narrow ribbon .
the coulomb potential ( [ e : coulomb ] ) is assumed to be small compared to the ribbon confinement . in the case of @xmath113
the lowest three sub - bands are specified by indices @xmath114 .
the set ( [ e : set ] ) corresponding to these subbands becomes @xmath115=0~;\\ v_0^{(2 ) ' } + \frac{1}{p}\left(e - \varepsilon_0 -v_{00}\right)v_0^{(1 ) } -\frac{1}{p } \left [ v_{10}v_{1}^{(1)}+ v_{-10}v_{-1}^{(1)}\right]=0~;\\ v_1^{(1 ) ' } - \frac{1}{p}\left(e + \varepsilon_1 -v_{11}\right)v_1^{(2 ) } + \frac{1}{p}\left [ v_{01}v_{0}^{(2)}+ v_{-11}v_{-1}^{(2)}\right]=0~;\\ v_1^{(2 ) ' } + \frac{1}{p}\left(e - \varepsilon_1 -v_{11}\right)v_1^{(1 ) } -\frac{1}{p } \left [ v_{01}v_{0}^{(1)}+ v_{-11}v_{-1}^{(1)}\right]=0~;\\ v_{-1}^{(1 ) ' } - \frac{1}{p}\left(e + \varepsilon_{-1 } -v_{-1 - 1}\right)v_{-1}^{(2 ) } + \frac{1}{p}\left [ v_{0 - 1}v_{0}^{(2)}+ v_{1 - 1}v_{1}^{(2)}\right]=0~;\\ v_{-1}^{(2 ) ' } + \frac{1}{p}\left(e - \varepsilon_{-1 } -v_{-1 - 1}\right)v_{-1}^{(1 ) } -\frac{1}{p } \left [ v_{0 - 1}v_{0}^{(1)}+ v_{1 - 1}v_{1}^{(1)}\right]=0~ ; \end{array } \right \}\end{aligned}\ ] ]
at the first stage we neglect the coupling between the states corresponding to the subbands with different @xmath116 the reason for this is that in the narrow ribbon of small width @xmath35 the diagonal potentials @xmath117 dominate the off - diagonal terms @xmath118 almost everywhere but for a small region @xmath119 ( see eq .
( [ e : onoff ] ) ) . in this case @xmath120 and the set ( [ e : set ] ) decomposes into independent subsets each specified by an index @xmath31 .
the 1d impurity states are then governed by the potential @xmath121 @xmath122 the set ( [ e : set ] ) for @xmath123 with @xmath124 is solved by matching in the intermediate region the corresponding solutions @xmath125 one of which is valid in the inner region close to the impurity centre and the other represents a solution of the outer region distant from the centre .
this method was originally developed by hasegava and howard @xcite in studies of excitons subject to strong magnetic fields and then successfully employed for the investigation of the impurity and exciton states in quantum wells @xcite , super - lattices @xcite and quantum wires @xcite . in the inner region
+ @xmath126 an iteration procedure is performed .
the first integration of the set ( [ e : set ] ) , in which we neglect the terms consisting of the energies @xmath127 , with the trial functions @xmath128 gives @xmath129~;\quad f_{1,2}=\frac{2y}{d_{1,2}}\ ] ] where @xmath130 and where @xmath131 is the dimensionless strength of the coulomb potential . the function @xmath132 can be obtained from eq .
( [ e : step1 ] ) by replacing @xmath133 by @xmath134 and @xmath134 by @xmath135 .
subsequent integration leads to the two independent particular solutions @xmath136 and @xmath137 corresponding to the relationships @xmath138 the linear combination of these solutions provides the general iteration functions , which read in the region @xmath139
@xmath140 where @xmath141 and where @xmath142 and @xmath143 are the arbitrary magnitude and phase , respectively .
since the potentials ( [ e : pot ] ) satisfy @xmath144 = @xmath145 the wave two - vectors + @xmath146 are classified with respect to parity .
further we focus on the even wave vectors for which @xmath147 where @xmath148 with @xmath149 ( @xmath150 is the pauli matrix ) .
the condition of even parity imposed on the wave vector @xmath151 formed by the components ( [ e : iter ] ) , implies that the phases @xmath143 are equal to the half integer of @xmath152 . obviously , an alternative way to derive eq .
( [ e : iter ] ) is to solve eqs .
( [ e : set ] ) for @xmath153 and then to compare the resulting solutions @xmath154 , expanded in series up to the terms of the first order of @xmath155 , with those given by eqs .
( [ e : step1 ] ) .
the calculated constant @xmath74 leads to the functions ( [ e : iter ] ) .
\a ) _ discrete states _ the exact solutions to eqs . (
[ e : set ] ) at @xmath124 for @xmath156 in the region @xmath157 with @xmath158 are calculated by the same method employed in studies of a relativistic electron in hydrogen @xcite and super - heavy atoms with the nuclear charge number number @xmath17 @xcite @xmath159\\ \sinh\frac{\psi_n}{2}\tau^{-\frac{1}{2}}\left [ w_{\kappa,\mu}(\tau ) -\frac{\tanh
\psi}{q}w_{\kappa + 1,\mu}(\tau ) \right ] \end{array } \right.\end{aligned}\ ] ] where @xmath160 and where @xmath161 is the whittaker function associated with the kummer function @xmath162 @xcite @xmath163 with @xcite @xmath164 the functions ( [ e : outer ] ) are normalized to @xmath165 .
the asymptotic behavior of the outer functions ( [ e : outer ] ) at large distances @xmath166 follows from eqs .
( [ e : outer]),([e : whitt ] ) and ( [ e : kum ] ) @xmath167 in the region @xmath168 eqs .
( [ e : outer]),([e : whitt ] ) and ( [ e : kum ] ) lead to the expressions @xmath169 where @xmath170 and @xmath171 are arbitrary constants . matching the eqs .
( [ e : outer ] ) and ( [ e : iter ] ) in the overlapping intermediate region @xmath172 we impose the condition @xmath173 which yields @xmath174 using the properties of the arguments of the @xmath175-functions in eq .
( [ e : theta ] ) for a small parameter @xmath176 and for the quantum numbers @xmath177 @xcite ( see also ref .
@xcite for details ) and choosing @xmath178 we arrive at the equation for the corrections @xmath179
@xmath180 -\ln ( n+\delta_{nn } ) \qquad \qquad \nonumber \\
+ ~~ \psi(1+n)+\ln \frac{|n-\tilde{\sigma}|\pi d}{2d}+2c-1=0,\end{aligned}\ ] ] where @xmath181 is the psi - function ( logarithmic derivative of the @xmath175-function ) , @xmath182 is the euler constant .
the corrections @xmath179 calculated from eq .
( [ e : discr ] ) determine the rydberg series of the discrete energy levels @xmath183 adjacent to the size - quantized energy level @xmath48 @xmath184;\ , & n=1,2,\ldots \\ \frac{\varepsilon_n}{\sqrt{1+\frac{q^2}{\delta_{n0}^2}}};~&n=0 \end{array } \right.\end{aligned}\ ] ] which allow to estimate the size of the coulomb state in eq .
( [ e : asimpt ] ) @xmath185 for @xmath186 and @xmath187 clearly from ( [ e : rydberg ] ) , the existence of the intermediate matching region @xmath188 is provided for excited state @xmath186 by the employed above small parameter @xmath189 and for the ground state @xmath190 by the condition @xmath191 the correction @xmath192 satisfies the transcendental equation @xmath193 + \ln \frac{|n-\tilde{\sigma}|\pi d}{2d}+c-1=0,\ ] ] while the corrections @xmath179 for the excited states @xmath186 can be calculated in an explicit form @xmath194\right\},\ ] ] \b ) _ continuous states _
since our approach to determine the wave function of the continuous states closely resembles that applied above for the wave functions of the discrete states only the basic points will be given below . setting in eqs .
( [ e : outer ] ) @xmath195 we obtain @xmath196\\ -{\rm i}&\sin\frac{\varphi_n}{2}t^{-\frac{1}{2}}\left [ w_{\tilde{\kappa},\mu}(t ) + { \rm i}\frac{\tan \varphi}{q}w_{\tilde{\kappa}+1,\mu}(t ) \right ] .
\end{array } \right.\end{aligned}\ ] ] where @xmath197 the wave vectors ( [ e : outercont ] ) are normalized to @xmath198 . at large distances @xmath199 the wave functions ( [ e : outercont ] ) , ( [ e : whitt ] ) , ( [ e : kum ] ) have the asymptotic form of the outgoing waves @xmath200 further we introduce the real wave functions associated with the standing waves @xmath201,\ ] ] where the functions @xmath202 have the asymptotic form of the ingoing waves @xmath203 and @xmath204 and @xmath205 are the arbitrary magnitude and phase , respectively . in the region @xmath206 eqs .
( [ e : outercont ] ) , ( [ e : whitt ] ) and ( [ e : kum ] ) lead to @xmath207\ ] ] where @xmath208 ; \xi_{+,-}=\arg \gamma ( -2{\rm i } q)\mp \arg \gamma_{+,- } ; m(\varphi)=\frac{\sin\frac{\varphi}{2}+\cos\frac{\varphi}{2}}{1+\sin\varphi}.\ ] ] the wave functions @xmath209 can be obtained from eq .
( [ e : outercont1 ] ) by replacing @xmath204 by @xmath210 and @xmath211 by @xmath212 for @xmath213 the wave functions ( [ e : outercont1 ] ) read @xmath214 ; v_{n \mbox{out}}^{(2)}(t)=d_n\left [ \cos\chi_n + c_n\cot\lambda_n\sin\chi_n \right]\ ] ] with @xmath215 similar to the case of the discrete states we obtain the equation for the phase @xmath216 on equating the ratios @xmath217 taken for the iteration ( [ e : iter ] ) and outer functions ( [ e : outercont2 ] ) @xmath218 since @xcite @xmath219 eq . ( [ e : lambda ] ) acquires for @xmath220 an explicit form @xmath221 - 2c+1}.\ ] ] since at @xmath189 the phase @xmath222 ( @xmath223 is the riemann zeta function with @xmath224 ) is the value of the higher order of smallness @xmath225 we set in eq .
( [ e : omega ] ) @xmath226 . as expected ,
setting in the functions @xmath227 ( [ e : outercont1 ] ) @xmath228 and then matching these functions with the iteration functions @xmath229 ( [ e : iter ] ) we obtain the equation for the phase @xmath205 . substituting this result into equation @xmath230 determining the poles of the @xmath231 matrix with @xmath232 @xcite we arrive at eqs .
( [ e : eqn ] ) , ( [ e : discr ] ) for the discrete energy levels .
below we consider the coupling between the ground @xmath45 and first excited @xmath46 states described by the system of the four upper eqs .
( [ e : set1 ] ) at @xmath233 applying the iteration method with the trial functions @xmath234 we arrive at two particular linear independent four - vectors , having the components @xmath235 calculated for @xmath236 . the linear combination of these vectors taken for @xmath237 $ ] provides the general expression for the iteration four - vector with the components @xmath238 where @xmath239 are an arbitrary constant and phase , respectively .
the parameter @xmath240 \nonumber\\ & + & \sin\alpha_0 \left [ { \rm { si}}\left ( \frac{\pi}{2 } + \alpha_0\right)+ { \rm { si}}\left ( \frac{\pi}{2 } -\alpha_0\right ) \right],~\alpha_0=\frac{\pi x_0}{d}.\end{aligned}\ ] ] consisting of the integral sine @xmath241 @xcite , describes the coupling induced by the potentials @xmath242 ( [ e : pot ] ) . the functions @xmath243 can be obtained from the functions @xmath244 ( [ e : iter2 ] ) by mutual replacing @xmath245 equating the ratios of the functions of the continuous spectrum + ( [ e : outercont2 ] ) @xmath246 and the iteration functions ( [ e : iter2 ] ) @xmath247 and then matching the ratios of the functions of the discrete states ( [ e : outer1 ] ) @xmath248 and the iteration functions @xmath249 we obtain the set of equations where the functions @xmath251 are defined by eqs .
( [ e : qfirst ] ) , ( [ e : iter ] ) , ( [ e : outer1 ] ) , respectively . in the limiting case of negligible coupling
the set ( [ e : setcoupl ] ) decomposes into two independent equations relevant to the discrete ( [ e : eqn ] ) , ( [ e : discr ] ) and continuous ( [ e : lambda ] ) states .
solving the set ( [ e : setcoupl ] ) by the determinantal method we obtain the equation for @xmath252 , which is then expanded in series with respect to the parameter @xmath253 keeping at @xmath178 the terms of the first order @xmath254 we arrive at the equation for the phase @xmath255 in an explicit form @xmath256 substituting eq .
( [ e : cot ] ) into equation @xcite @xmath257 the complex quantum numbers @xmath258 introduced in eq .
( [ e : outer ] ) can be calculated , which in turn determine the complex energy levels @xmath259 adjacent to the size - quantized first excited energy level @xmath260 @xmath261 where the second term in the right - hand part is the rydberg series of the energy levels associated with the quasi - coulomb diagonal potential @xmath262 ( [ e : limit ] ) ( no coupling ) . the following notation in eq .
( [ e : compl ] ) for the resonant shift @xmath263 and resonant width @xmath264 both induced by the inter - subband @xmath265 interaction is used @xmath266 and @xmath267 in eqs .
( [ e : width ] ) and ( [ e : shift ] ) @xmath268 @xmath269 where the corrections @xmath270 can be calculated from eqs .
( [ e : discr ] ) ( [ e : ground ] ) ( [ e : exc ] ) . in the logarithmic approximation @xmath271 , @xmath272 the quantum number @xmath273 can be found from equation @xmath274,\ ] ] with ( [ e : phi ] ) for @xmath275 . in conclusion of this section note that the equation absolutely identical to eqs .
( [ e : compl ] ) - ( [ e : shift ] ) can be derived by matching the real iteration functions @xmath276 ( [ e : iter ] ) and complex functions of the continuous states @xmath277 ( [ e : outercont ] ) having the asymptotic form of the outgoing wave .
in this section we consider the coupling between the discrete states adjacent to the highest size - quantized level @xmath278 and and the continuous states attributed to the low - lying levels @xmath279 and @xmath280 .
below we neglect in the set ( [ e : set1 ] ) the off - diagonal potentials @xmath281 and @xmath282 describing the interactions of the @xmath265 subbands . extending the iteration procedure employed above for the single- and double- subband approximation to the present stage with the trial functions @xmath283 we arrive at two particular linear independent sixfold vectors @xmath284 calculated for @xmath285 . taking @xmath286 $ ] , we obtain the components @xmath287 of the total iteration sixfold vector @xmath288 @xmath289 where @xmath290 is determined in eq .
( [ e : iter ] ) and @xmath142 and @xmath143 are arbitrary constants and phases , respectively .
the parameter @xmath291 ( [ e : couple ] ) , while @xmath292 \nonumber\\ + \sin 2\alpha_0 \left[\mbox{ci}(\pi + 2\alpha_0)- \mbox{ci}(\pi - 2\alpha_0 ) \right],~\alpha_0=\frac{\pi x_0}{d}\end{aligned}\ ] ] describes the coupling induced by the potentials @xmath293 ( [ e : pot ] ) .
the functions @xmath294 can be obtained from the functions @xmath295 ( [ e : iter3 ] ) , respectively , by replacing @xmath296 . as mentioned above
further we match the wave functions of the continuous spectrum @xmath297 having the asymptotic form of the outgoing wave @xmath298 ( [ e : outercont ] ) to give in turn the sixfold vector @xmath299 with the components @xmath300 \nonumber\\ -(1-c_n)\exp\left[-\imath \left(q\ln2k_n y + \xi_{- } \right ) \right],~n=0,1,\end{aligned}\ ] ] where @xmath301 and @xmath211 are given in eqs .
( [ e : outercont2 ] ) ( [ e : outercont1 ] ) and ( [ e : outercont ] )
. the wave functions @xmath302 can be obtained from the functions @xmath303 ( [ e : outercont3 ] ) , respectively , by replacing @xmath304 the wave functions @xmath305 have the form ( [ e : outer1 ] ) , in which @xmath306 ( [ e : outer ] ) and @xmath307 ( [ e : theta ] ) are calculated for @xmath47 . matching the sixfold wave vectors @xmath308 within the intermediate region by imposing the conditions @xmath309 where @xmath310 are given by eqs .
( [ e : outercont3 ] ) and ( [ e : outer ] ) and @xmath311 by eqs .
( [ e : iter3 ] ) we obtain @xmath312 - r_{-1 } q \gamma_{0,-1 } [ a_0^{(-)}\sin\zeta_{-1}+{\rm i } a_0^{(+)}\cos\zeta_{-1}]=0~;\\ r_1 [ a_1^{(- ) } \cos(q + \zeta_1 ) - { \rm i } a_1^{(+)}\sin(q + \zeta_1 ) ] - r_{-1 } q \gamma_{1,-1 } [ a_1^{(-)}\sin\zeta_{-1}+{\rm i } a_1^{(+)}\cos\zeta_{-1}]=0~;\\ r_{-1}\sin(\omega_{-1}-q - \zeta_{-1})-r_{1}q\gamma_{1,-1}\cos(\omega_{-1}-q - \zeta_{1})- r_{0}q\gamma_{0,-1}\cos(\omega_{-1}-q - \zeta_{0})=0~ , \end{array } \right \}\end{aligned}\ ] ] in eqs .
( [ e : setcouple1 ] ) @xmath313 \pm(1-c_{n})\exp\left[-{\rm i } \left(q\ln2k_n y + \xi_{- } \right)\right]\ ] ] and @xmath314 are introduced by eqs .
( [ e : iter ] ) and ( [ e : outer1 ] ) for @xmath47 , respectively . solving the set ( [ e : setcouple1 ] ) by the determinantal method we obtain the equations for the complex quantum numbers @xmath315 ( [ e : outer ] ) , in which we take for the phases @xmath316 and keep the terms of the first order of @xmath189 @xmath317 with @xmath318 and the quantum numbers @xmath319 are obtained from @xmath320\ ] ] with ( [ e : phi ] ) for @xmath278 . the complex quantum numbers @xmath321 calculated from eq
( [ e : image1 ] ) lead to the complex impurity energy levels adjacent to the size - quantized second excited sub - band @xmath278 @xmath322 where the resonant width @xmath323 and shift @xmath324 have the form @xmath325 and @xmath326 the coefficients @xmath327 are defined in eqs .
( [ e : width ] ) and ( [ e : shift ] ) .
we define the binding energy of the electron @xmath328 in the @xmath43-th quasi - coulomb state associated with the @xmath31 size - quantized subband as the real part of the difference between the size - quantized energy @xmath48 ( [ e : phi ] ) of the free electron and the energy of the impurity electron @xmath183 given by eqs .
( [ e : rydberg ] ) , ( [ e : compl ] ) and ( [ e : compl ] ) for the ground @xmath45 , first @xmath46 and second @xmath47 excited subbands , respectively .
since the resonant shifts @xmath329 ( [ e : shift ] ) , ( [ e : shiftn ] ) are of the order of @xmath330 with respect to the rydberg energies determined by the second terms in the right - hand parts of eqs .
( [ e : rydberg ] ) , ( [ e : compl ] ) and ( [ e : compl ] ) the binding energies read @xmath331;~&n=0 \end{array } \right.\end{aligned}\ ] ] where the corrections @xmath179 can be calculated in the single - subband approximation from eqs .
( [ e : ground ] ) and ( [ e : exc ] ) for the ground @xmath190 and excited @xmath186 impurity states , respectively .
it follows from eqs .
( [ e : bind ] ) and ( [ e : phi ] ) that the binding energy @xmath332 and the oscillatory part of @xmath48 ( see @xmath91 in eq .
( [ e : phi ] ) ) decrease with increasing the ribbon width @xmath35 . in an effort to render our calculations close to an experimental setup
, we take below for the estimates of the expected values the parameters @xmath333 @xcite and @xmath334 corresponding to the @xmath335 and sapphire , respectively , employed as substrates for gnr @xcite .
the latter parameter @xmath155 is close to the limit caused by the condition @xmath336 ( see below eq.([e : rydberg ] ) ) .
further we focus on the monotonic dependence @xmath92 and keep the the parameter @xmath337 in eq .
( [ e : phi ] ) for the levels @xmath48 to be @xmath338 . the binding energy @xmath339 ( [ e : bind ] ) of the ground state @xmath190 calculated for(@xmath340 ) as a function of the reciprocal width @xmath341 of the gnr .
impurity is placed symmetrically to the boundaries @xmath342 .
the parameter @xmath343 . ] the dependencies of the binding energies on the width of the gnr @xmath35 for the ground state for different strengths of the impurity potential are given in fig .
2 . these graphs , while ignoring the oscillations , are qualitatively completely in line with the data of the numerical calculations and experimental observations recently performed with the related coulomb systems . the exciton effects in the armchair gnrs were studied in frame of the tight - binding model @xcite and density functional theory @xcite , while han et.al .
@xcite investigated experimentally the influence of the localized states in gnrs on the electron transport .
the relation @xmath344 including oscillations @xcite have been found to occur .
the differences between the impurity states considered here and the exciton and localized states prevent us from a detailed quantitative comparison .
the dimensionless binding energy @xmath345 calculated from ( [ e : bind ] ) , ( [ e : ground ] ) , ( [ e : phi ] ) with @xmath346 for the ground state @xmath347 plotted as a function of the effective index of the corresponding subband @xmath348 and the relative impurity position @xmath349 in the gnr of width @xmath35 . ] the coulomb pattern of the energy levels ( [ e : rydberg ] ) enables to introduce the effective rydberg constant @xmath350 , the bohr radius @xmath351 and the mass @xmath352 for the impurity electron in gnr @xmath353 which additionally illustrate the physical reason of the bonding of the impurity electron , namely the quasi-1d geometry of the gnr .
note that the bound states arise at any finite width @xmath354 .
this result is qualitatively analogous to the effect of anti - diamagnetism caused by the influence of the magnetic field on the weakly bound atomic state .
demkov and drukarev @xcite considered the 3d potential well of small radius and depth to provide the capturing of the electron .
it was shown that the arbitrarily weak magnetic field @xmath75 induces the bound electron state with the binding energy @xmath355 .
the common reason for this is that the finite width @xmath354 and magnetic length @xmath356 transform the graphene monolayer and atomic structure , respectively , into the quasi-1d systems , which are more favorable to generate bound states .
the dependencies @xmath357 and @xmath344 correspond to the different dispersion laws namely @xmath358 and @xmath359 @xmath360 for the atomic ( @xmath361 ) and gnr ( @xmath362 ) electron , respectively . the dependence of the binding energy @xmath328 ( [ e : bind ] ) on the displacement of the impurity centre @xmath363 from the mid - point of the ribbon @xmath58 is contained in the corrections @xmath364 namely in the term @xmath365 in eqs .
( [ e : ground ] ) and ( [ e : exc ] ) , while @xmath328 as a function of the effective number of the subband @xmath366 is given by the sub - band threshold @xmath48 ( [ e : phi ] ) mainly and the term @xmath367 in the correction @xmath179 .
the dimensionless binding energy @xmath368 as a function of the effective quantum number @xmath366 and relative displacement @xmath369 for the ground @xmath190 state is depicted in fig .
3 . clearly , the higher the subband i.e. the greater the value @xmath366 is the less the binding energy @xmath370 .
also the binding energy decreases when the impurity shifts from the mid - point of the ribbon towards the boundaries .
the latter conclusion coincides with those obtained for the quantum well in refs .
@xcite .
the inter - band coupling shifts the strictly discrete excited rydberg series @xmath371 ( [ e : rydberg ] ) calculated in single - subband approximation by an amount @xmath329 ( [ e : shift ] ) @xmath46 , and ( [ e : shiftn ] ) @xmath47 and transforms them to the quasi - discrete levels of width @xmath53 ( [ e : width ] ) @xmath46 , and ( [ e : widthn ] ) @xmath47 .
note that the conclusions made on the base of the first and second excited subbands can be qualitatively extended to others .
since the resonant shifts @xmath372 first are much less than the resonant widths @xmath373 at @xmath374 and second the resonant shifts do not change the discrete character of the energy spectrum ( [ e : rydberg ] ) we focus on the widths @xmath53 .
it is clear from eqs .
( [ e : width ] ) , and ( [ e : widthn ] ) that the widths @xmath375 increase with decreasing the ribbons width @xmath35 .
note that this dependence is opposite to that in a semiconductor narrow quantum well : the narrower the well is the less are the resonant widths @xcite .
the reason for this is that in the quantum well the resonant width @xmath376 where the impurity rydberg constant @xmath377 and the binding energy @xmath328 do not depend on the well width @xmath35 , while the inte - rband energy distance @xmath378 increases and consequently the resonant width decreases with the narrowing of the quantum well . for the ribbon @xmath379 ( [ e : bind ] ) and the inter - subband coupling do not depend on the ribbon width @xmath35 and @xmath380 .
the resonant width @xmath381 ( [ e : width ] ) of the ground impurity state @xmath347 relatively to the corresponding threshold @xmath382 ( [ e : phi ] ) @xmath383 versus the relative impurity position @xmath349 in the gnr of width @xmath35 providing the parameter @xmath343 .
the parameter @xmath155 is taken to be @xmath340 . ]
the dependence of the resonant widths @xmath384 ( [ e : width ] ) , calculated in the double - subband approximation , on the position of the impurity centre @xmath363 is described by the coupling parameter @xmath385 ( [ e : couple ] ) and the corrections @xmath270 ( [ e : ground ] ) and ( [ e : exc ] ) .
the dependencies of the relative resonant widths @xmath386 ( [ e : width ] ) , on the dimensionless shift @xmath387 for the ground @xmath190 state are presented in fig.4 , in which the limitation on the parameter @xmath388 are caused by the condition imposed on @xmath389 placed below eq .
( [ e : rydberg ] ) . for the impurity positioned at the mid - point of the ribbon @xmath390 the resonant width and shift both vanish @xmath391 because of the even @xmath28-parity of the coulomb potential @xmath105 ( [ e : coulomb ] ) in eq .
( [ e : pot ] ) and opposite parities of the neighboring @xmath265 transverse @xmath28-states to give @xmath392 .
both in the quantum well and in the ribbon the shift of the impurities from their mid - points eliminates the even @xmath28-parity of the potential @xmath105 ( [ e : coulomb ] ) in eq .
( [ e : pot ] ) , that leads to the coupling between the @xmath393 subbands .
if the impurity displaces from the mid - point towards the boundaries @xmath394 the resonant widths @xmath264 ( [ e : width ] ) monotonically increases .
this correlates completely with the analogous dependence found for the impurity states in the semiconductor quantum well @xcite . for small shifts @xmath395 in eq .
( [ e : couple ] ) we obtain for the parameter @xmath396 in eqs .
( [ e : width ] ) @xmath397 while for the impurity positioned close to the ribbon edge @xmath398 we obtain @xmath399 . note that the zeroth width of the first excited @xmath400 @xmath401 series in case of the symmetrical @xmath42 impurity position is a consequence of the double - subband approximation . in the multi- subband approximation
the levels of the above mentioned series would acquire finite widths .
the resonant width @xmath402 ( [ e : widthn ] ) of the ground impurity state @xmath347 relatively to the corresponding threshold @xmath278 ( [ e : phi ] ) @xmath403 versus the relative impurity position @xmath404 in the gnr of width @xmath35 using the parameter value @xmath343 .
the parameter @xmath155 is taken to be @xmath405 . ]
the dependence of the resonant widths ( [ e : widthn ] ) of the impurity states corresponding to the second excited subband @xmath47 on the position of the impurity centre is completely different from that related to the first excited subband @xmath46 . equation ( [ e : widthn ] ) shows that contributions to the resonant widths @xmath323 are caused by the coupling with the subbands @xmath406 ( [ e : couple ] ) and @xmath407 ( [ e : couple1 ] ) .
note that the estimated contribution to the resonant width @xmath323 caused by the neglected coupling between the @xmath393 subbands is of the order of @xmath408 in the vicinity of the mid - point @xmath409 the subband @xmath46 contributes mostly @xmath410 , while for @xmath411 both subbands contribute @xmath412 .
the position of the impurity @xmath413 , at which the effects of the subbands @xmath45 and @xmath46 on the resonant width @xmath323 are in balance is determined by the root @xmath414 of the equation @xmath415 to give the result @xmath416 .
the coupling between the subbands @xmath417 provides the nonzero widths @xmath323 and shifts @xmath324 for any positions @xmath363 of the impurity .
the width @xmath402 as a function of the impurity shift @xmath363 is given in fig.5 demonstrating the monotonic drop within the same regions as those corresponding to fig.4 . as mentioned above the presented method is valid under the conditions @xmath189 for the excited impurity states @xmath186 and @xmath418 ( [ e : ground ] ) for the ground state @xmath190 . under these conditions the radius of the impurity state considerably exceeds the width of the gnr so that the ribbon is narrow compared to the impurity size .
it follows from eq .
( [ e : discr ] ) that a `` big logarithm '' can only hardly achieved @xcite i.e. the logarithmic approximation @xmath419 ensures the real smallness of @xmath389 .
however the previous calculations related to the ground state of the quasi-1d diamagnetic exciton @xcite and present estimates show that a reasonably small parameter @xmath155 leads to values @xmath420 , which provide a quite accurate and adequate description of the ground impurity state in gnr . taking into account possible experiments we estimate the expected electron binding energy for the impurity centre placed at the middle point of the gnr of width 1 nm on the sapphire substrate as @xmath421 and on the @xmath335 substrate as @xmath422 . this is less than the data attributed to the @xmath423 substrate @xmath424 because of the relatively small screening of the impurity potential .
also an estimate of the lifetimes @xmath54 yields for the impurity positioned at the mid - point of the gnr @xmath425 for the @xmath426 and sapphire substrate , respectively . for the @xmath427 substrate
the screening of the impurity attraction is less , the lifetime is reduced and therefore less favourable for a corresponding experimental observation .
a shift of the impurity centre @xmath428 generates lifetimes @xmath429 of the same order as @xmath430 at @xmath42 .
the electrons captured onto such short - lived trap states will most likely contribute to the dc transport .
however , the high - frequency response of such electrons may reveal the signatures of localization .
clearly , the above considered quasi - rydberg series ( [ e : rydberg ] ) do not cover the total set of discrete states .
the oscillations of the wave functions ( [ e : iter ] ) and ( [ e : outer1 ] ) caused by the logarithmic term are an indicator of additional energy levels positioned below the series ( [ e : rydberg ] ) . since the possible strong shift of these levels away from the threshold @xmath48 is against the spirit of the employed adiabatic perturbation theory @xmath431 describing the shallow energy levels we are limited to qualitative estimates based on the quasi - classical relativistic approach @xcite and @xcite .
the dimensionless binding energy @xmath432 of the quasi - classical ground state @xmath433 found from ( [ e : class ] ) for @xmath434 and from ( [ e : phi ] ) with @xmath343 for @xmath435 versus the parameter @xmath155 . ] in an effort to elucidate the origin of this additional series let us consider the so called `` logarithmic '' energy levels governed by the logarithmic potential ( [ e : limit ] ) taken for @xmath436 .
these levels can be calculated from the bohr - sommerfeld quantization rule @xmath437 where @xmath438 and @xmath439 are the classical relativistic momentum and turning point , respectively , with @xmath440 ; \nonumber\\ \mathcal{p}(y_0)&=&0.\end{aligned}\ ] ] equation ( [ e : bohrsomm ] ) admits an exact solution which provides for the energies @xmath441,~s_n=\frac{\varepsilon_n d}{2pq}=\frac{|n-\tilde{\sigma}|\pi}{2q},\ ] ] where @xmath442 is the modified bessel function @xcite . the binding energy @xmath443 with @xmath183 calculated from ( [ e : class ] )
reads @xmath444 both for @xmath445 and for @xmath446 .
it follows that the weakness of the logarithmic singularity and smallness of the strength of the impurity potential @xmath431 seems not to provide the bonding of the quasi - classical relativistic electron @xmath447 , while a sufficiently strong attraction @xmath448 could produce a localized impurity state @xmath449 .
the dependence of the binding energy of the quasi - classical ground state @xmath433 found from ( [ e : class ] ) on the parameter @xmath155 is depicted in fig.6 .
the ground `` logarithmic '' level arises at the critical value @xmath450 and shifts towards lower energies to provide for the binding energy @xmath451 .
the above can be considered as no more than only a qualitative evidence of existence of such additional states in gnr that have transformed from the collapsed states in the graphene monolayer governed by the 2d impurity potential @xmath452 @xcite . though the `` logarithmic '' and quasi - rydberg levels in principle correspond to the same region of the parameter @xmath453 any numerical comparison between the results for the quasi - rydberg series based on the dirac equation and those for the `` logarithmic '' levels derived from the quasi - classical method applied moreover to the ground state seems to be incorrect .
the total set of the impurity states in gnr requires a further study of the equations ( [ e : set ] ) with the potential ( [ e : pot ] ) , having the logarithmic singularity in the vicinity of the impurity centre .
we have developed an analytical adiabatic approach to the problem of bound and meta - stable ( fano resonances ) quasi - coulomb impurity states in a narrow gaped armchair graphene nanoribbon ( gnr ) .
the width of the gnr is taken to be much less than the radius of the impurity .
this adiabatic criterion implies a variable width of the gnr and simultaneously the smallness of the coulomb interaction relative to the size - quantized energy induced by the gnr .
the energy spectrum of the impurity electron is a sequence of the series of the quasi - rydberg discrete and resonant states adjacent to the ground and excited size - quantized subbands , respectively .
the binding energies and the resonant widths and shifts attributed to the inter - subband coupling are calculated in an explicit form in the single- and multi - subband approximation , respectively .
the binding energies and the resonant widths both increase with decreasing the gnr width . as the impurity centre displaces from the mid - point of the gnr the binding energies decrease , while the resonant widths of the quasi - rydberg series associated with the first / second excited sub - bands increase / decrease , respectively .
our analytical results are in complete agreement with those found by other theoretical approaches and in particular numerical studies .
estimates of the expected values show that the bound and meta - stable impurity states in gnr can be observed experimentally .
the authors are grateful to c. morfonios for technical assistance . financial support by the deutsche forschungsgemeinschaft is gratefully acknowledge . | an analytical study of discrete and resonant impurity quasi - coulomb states in a narrow gaped armchair graphene nanoribbon ( gnr ) is performed .
we employ the adiabatic approximation assuming that the motions parallel ( `` slow '' ) and perpendicular ( `` fast '' ) to the boundaries of the ribbon are separated adiabatically .
the energy spectrum comprises a sequence of series of quasi - rydberg levels relevant to the `` slow '' motion adjacent from the low energies to the size - quantized levels associated with the `` fast '' motion .
only the series attributed to the ground size - quantized sub - band is really discrete , while others corresponding to the excited sub - bands consist of quasi - discrete ( fano resonant ) levels of non - zero energetic widths , caused by the coupling with the states of the continuous spectrum branching from the low lying sub - bands . in the two- and three - subband approximation the spectrum of the complex energies of the impurity electron is derived in an explicit form .
narrowing the gnr leads to an increase of the binding energy and the resonant width both induced by the finite width of the ribbon . displacing the impurity centre from the mid - point of the gnr causes the binding energy to decrease while the resonant width of the first excited rydberg series increases . as for
the second excited series their widths become narrower with the shift of the impurity .
a successful comparison of our analytical results with those obtained by other theoretical and experimental methods is presented .
estimates of the binding energies and the resonant widths taken for the parameters of typical gnrs show that not only the strictly discrete but also the some resonant states are quite stable and could be studied experimentally in doped gnrs . |
thin film growth processes involve complicated kinetics giving rise to a rich variety of surface morphologies . within this vast domain ,
the study of the growth in the submonolayer regime is of particular interest due to the large impact of the initial kinetics on the resulting film structure .
experiments on thin film growth on well characterized substrates using molecular - beam epitaxy ( mbe ) have provided a large body of information about growth kinetics and morphology , and revealed that for a variety of systems and a broad temperature range , island nucleation is the dominant mechanism for crystal growth @xcite .
diffraction methods such as helium beam scattering @xcite , low energy electron diffraction @xcite and other techniques @xcite , provide information on the collective behavior and the statistical properties of the surface .
these techniques have been used to measure the island size distribution , the island density , and their scaling properties with respect to the coverage and the flux @xcite .
the variation of the island density with respect to the temperature was also studied @xcite .
more detailed information at the atomic scale is provided by scanning tunneling microscopy ( stm ) @xcite .
most notably , stm provides means to study the variety of morphologies encountered in the different systems , or in the same system under different growth conditions @xcite . in some experiments stm was used to acquire information on larger scales , e.g. , island size distributions @xcite . despite the wealth of experimental results at the atomic scale , for decay rates of small islands , mobility of small islands and edge diffusion @xcite ,
the underlying energetics is mostly inaccessible to direct experimental measurements .
thus , one must rely on theory to extract activation energies from the experimental results , and these are usually limited in number , and sometimes are subject to alternative interpretations . the only technique which provides direct access to diffusion processes and activation energies at the atomic scale is field ion microscopy ( fim ) @xcite .
this technique was used to identify the diffusion modes of adatoms @xcite as well as small islands @xcite on fcc(001 ) metal surfaces , to measure their diffusion coefficients , and to determine the sticking process of adatoms to an island @xcite .
recently there were several attempts to use stm to derive such local information directly @xcite .
theoretical studies aimed at providing better understanding of the relation between key processes at the atomic scale and the resulting morphologies have been done using monte carlo ( mc ) simulations @xcite . in simulations of island growth during deposition ,
atoms are deposited randomly on the substrate at rate @xmath0 [ given in monolayers ( ml ) per second ] and then hop , attach to , and detach from existing islands according to some model .
a common approach is to assume some key processes and their rates , and then simulate the growth process @xcite . in some cases information such as diffusion length , typical distance and time between nucleation events
is assumed to be known a - priori , and is put by hand into the simulation in order to accelerate the computation @xcite .
the advantage of this approach is that the models are well defined and use only few parameters .
these models are useful for studies of scaling and morphology but can not provide a quantitative description of diffusion on a particular substrate .
furthermore , they account only for a limited number of processes , that are assumed to be the only significant ones
. a complementary scheme employs the underlying activation energies . in this scheme the hopping rate @xmath1 ( in units of hops per second ) of a given atom to each unoccupied nearest neighbor ( nn ) site is given by @xmath2 where @xmath3 @xmath4 is the commonly used attempt rate , @xmath5 is the activation energy barrier , @xmath6 is the boltzmann constant and @xmath7 is the temperature .
the activation energy barrier @xmath5 depends on the local environment of the hopping atom , namely the configuration of occupied and unoccupied adjacent sites .
two approaches have been taken in the construction of the energy barriers for hopping in the simulations .
one approach was to construct simple models that include the desired features , such as stability and mobility of small islands , and that take into account properties such as bond energies @xcite . in general , this approach encompasses both the virtues and the drawbacks of the simpler approach presented before .
a second approach is based on the use of an approximate many - body energy functional to calculate the hopping energy barriers for a complete set of relevant configurations @xcite .
this approach provides a good description of diffusion processes on the given substrate but only limited understanding due to the large number of parameters . in this paper
we extend and further explore a framework for a _
systematic _ derivation of simple models for self - diffusion on fcc(001 ) surfaces out of a detailed and complicated set of energy barriers .
simple in this context means that only a small number of parameters are involved and all have a definite and intuitive interpretation . using sensible assumptions about the bond energies and diffusion paths we obtain simple formulae for the activation energy barriers .
we then _ optimize _ the parameters of these formulae for each metal separately by using energy barriers obtained from the embedded - atom method .
this procedure gives rise to simple models that have at most four parameters and provide good quantitative description of the landscape of hopping energy barriers . in a previous publication we have introduced the framework and applied it to cu / cu(001 ) growth @xcite .
here we make a three - fold step forward .
( a ) we include four other fcc metals in the model and derive the appropriate parameters for them .
this shows the utility of the models and provide a unifying framework that applies to a large class of metals ; ( b ) non - linear interactions are introduced in addition to the linear interactions considered before .
this allows for a more accurate optimization without increasing the number of parameters ; ( c ) we explore the basis of the assumptions underlying this scheme and the extent of their applicability .
the paper is organized as follows . in sec .
ii we introduce the physical framework of the model and discuss its underlying assumptions .
the results of the eam calculations are given in sec .
the models are introduced in sec .
iv with the fitting to the eam results .
this is followed by a discussion of the results and their implications in sec .
the framework developed in this paper assumes several characteristics of the diffusion processes considered .
it applies to systems for which these assumptions are valid , which includes most of the fcc metals in the moderate temperature regime @xmath8 .
the following assumptions are employed throughout the discussion .
_ bridge - site hopping of adatoms is in general dominant over exchange hopping _
[ fig : terrace_diffusion ] ) .
there has been a controversy concerning this assumption . using semi - empirical methods the barrier for exchange hopping for cu(001 )
was estimated to be @xmath9 ev @xcite , in agreement with the experimental data that was available at that time @xcite .
this is much lower than the barrier for bridge - site hopping .
however , theoretical work using several other methods @xcite including eam indicate that the exchange barrier for cu is much higher ( more than @xmath10 ev ) and therefore bridge - site hopping is dominant .
this conclusion is also supported by recent experimental work and reinterpretation of the previous findings @xcite . in general ,
bridge - site hopping is found to be dominant in all the metals studied here .
( for au , some exchange processes appear to be significant , yet in most cases bridge hopping is favorable ) . due to the exponential dependence of the rate of each process on the corresponding energy barrier ,
it is generally reasonable to take into account only the mechanism which is energetically favorable ( bridge - site hopping , in this case ) and neglect the mechanism which exhibits higher activation energy barrier ( exchange mechanism ) .
_ only nearest and next - nearest neighbor interactions are significant_. within this assumption one can obtain the activation energies for most diffusion processes to a good accuracy .
however there are some processes such as vacancy diffusion , where a larger environment affects the diffusing atom .
_ there is one common attempt frequency for all processes . _ since there is no systematic knowledge about the dependence of the attempt frequency on the local environment , the assumption of one common frequency is the usual practice .
estimations for the attempt frequency can be obtained using md simulations @xcite .
another way is to fit molecular static ( ms ) data to harmonic potential to find an effective force constant , and then deduce the frequency of oscillations .
so far , little work has been done on this subject .
previous works , including interpretation of experimental results , usually pre - suppose some common attempt frequency in the range @xmath11@xmath12@xcite .
the models described in this work are tested by fitting their parameters to energy barriers for self - diffusion of cu(001 ) , ag(001 ) , au(001 ) , ni(001 ) and pd(001 ) surfaces , obtained using eam @xcite .
this method uses semi - empirical potentials and provides a good description of self - diffusion on such surfaces @xcite .
specifically , for all the metals considered here the eam functions developed by adams , foiles , and wolfer ( afw ) @xcite are employed .
these functions are fitted to a similar but more accurate data base as the one employed by foiles , baskes , and daw @xcite .
the calculations are done on a slab of 20 square layers with 100 atoms in each layer .
when an atom on the surface hops into a vacant nearest neighbor site it has to cross the energy barrier between the initial and final sites .
we have used molecular statics in conjunction with the eam functions to find that energy barrier .
this is simply the difference between the energy at the bridge site ( or more precisely , at the point along the path with highest energy ) and in the initial site .
the hopping energy barriers are calculated for all local environments as shown in fig .
[ fig : env ] , where seven adjacent sites , @xmath13 are taken into account , according to the assumptions presented in sec .
each one of these sites can be either occupied ( @xmath14 ) or vacant ( @xmath15 ) , giving rise to @xmath16 barriers .
a binary representation is used to assign indices to these barriers . for each configuration @xmath17 the barrier is given by @xmath18 , where @xmath19 takes the values @xmath20 .
the full set of hopping energy barriers ( given in ev ) is presented in table [ eam_table ] , for cu(001 ) , ag(001 ) , au(001 ) , ni(001 ) and pd(001 ) . to show these values in a compact form ,
each barrier in table i corresponds to a configuration in which the occupied sites are the union of the occupied sites in the picture on top of the given column and on the left hand side of the given row .
the column in table i in which a given configuration appears is determined by the occupancy of sites @xmath21 while the row is determined by sites @xmath22 .
one can define @xmath23 such that for each configuration @xmath24 . to demonstrate the use of table [ eam_table ] , we will check for cu(001 ) the barrier of the configuration in which sites @xmath25 and @xmath26 are occupied and all other sites adjacent to the hopping atom are vacant .
for this configuration , according to eq .
( [ n1n2 ] ) , @xmath27 and @xmath28 ( @xmath29 ) .
the barrier , that is found in the column with the index @xmath30 and the row with index @xmath31 , in the line of cu , is @xmath32 ev . in table
[ eam_table ] we use the symmetries of the configurations in the @xmath33 cell ( fig .
[ fig : env ] ) to reduce the number of entries .
there is a mirror symmetry plane perpendicular to the surface and containing the arrow of the hopping atom .
consequently , the columns of @xmath34 and @xmath35 , in which site @xmath36 is occupied , stand also for the symmetric configurations in which @xmath37 is occupied . in the other four columns , there are some configurations that , due to symmetry , appear twice . in such cases ,
the barrier for the configuration with larger @xmath38 appears in _
italics_. for the purpose of the calculations and parameterization of the model we consider only hopping moves in which a single atom hops each time .
it turns out , however , that in some cases the molecular statics calculations , used to obtain the barriers , give rise to concerted moves . in such moves the atom at site @xmath39 follows the hopping atom and takes the place vacated by the hopping atom
this fact significantly reduces the barrier .
it turns out that for configurations in which concerted moves appear , they can be suppressed by adding a column of three atoms on the left hand side of sites @xmath40 and @xmath26 . in table i , the energy values for those configuration in which a concerted move was found , are shown in parenthesis .
the barrier obtained when the concerted move was suppressed is shown to the left of the parenthesis . to gain a better understanding of the barrier energy landscape we present the barrier height distribution ( without concerted moves ) in fig .
[ histogram ] for the five metals considered .
we observe that this distribution exhibits four groups .
this feature is in agreement with ref .
@xcite where a different method @xcite was used to calculate the barriers .
each group , corresponds to a single or a double column in table i. in general , group i includes very fast moves towards island edges , group ii includes moves along the edge , group iii includes , most notably , the single atom move , while group iv includes detachment moves .
the starting point in the construction of a simple model that describes the hopping energy barriers for all the configurations of fig .
[ fig : env ] , is the assumption that the contributions of all adjacent atoms to the energy barrier add up linearly . to examine this assumption for cu(001 ) , we evaluated directly the binding energies within the eam approach for a series of configurations from which we extracted the relevant bond energies .
the binding energy of a given configuration of adatoms on the surface is evaluated as follows : first we calculate the total energy of the system in that configuration .
then we find the total energy of another configuration in which there is the same number of adatoms on the surface but they are far apart from each other .
( by that we mean that moving them one lattice site in any direction would not change the total energy ) .
the binding energy between the adatoms is given by the difference in the total energies between the two configurations .
an example of the procedure is shown in fig .
[ linearity1 ] .
it appears , that the evaluation of the nnn bond energy is easier since one can construct a sufficiently large set of configurations which include only nnn bonds with no nn bonds . in the case of nn bonds ,
most of the relevant configurations also include nnn bonds [ fig .
[ linearity2 ] ( a ) ] .
similarly , considering an atom on top of a bridge site , typical configurations which include atoms adjacent to the bridge site exhibit nn bonds between them as shown in fig .
[ linearity2](b ) .
therefore , we will first examine the additivity of the nnn bonds employing a series of four configurations in which an adatom has 1 , 2 , 3 and 4 nnn on the surface .
for each one of these four configurations the total energy is compared to that of a configuration with the same number of adatoms , in which they are far apart from each other . in fig .
[ linear : graph](a ) , the total binding energy between adatoms is plotted as a function of the number of nnn bonds .
the best linear fit is drawn , and its slope yields a value of @xmath41 ev .
the next step is to examine the linearity of the nn binding energy using a series of four configurations in which an adatom has 1 , 2 , 3 and 4 nn s on the surface , within a similar procedure .
the nnn bonds in each configuration are deducted using the value obtained before .
the results are shown in fig .
[ linear : graph](b ) and the nn bond energy is obtained : @xmath42 ev . a similar analysis for an adatom on a bridge site is shown in fig .
[ linear : graph](c ) and the binding energy between an atom on the bridge site and an adjacent atom is given by @xmath43 ev .
the energy barrier @xmath5 for a certain process is the difference between the binding energies of the hopping adatom ( to the substrate and to adjacent adatoms ) at the initial position , @xmath44 , and at the bridge site , @xmath45 , namely @xmath46 . on the basis of the additivity feature just demonstrated
, we will now express these binding energies as the sum of the occupation states of the relevant sites .
the first approximation for the energies gives a model ( model i ) with only two parameters , that reproduces the main features of the eam barriers . in order to establish the model
, there are two things to note about the parameters obtained in sec .
[ linearity ] .
first , the values of nn binding energies at the lattice site and bridge site are very close .
this reflects the fact that the nn distance corresponds approximately to the minimum potential of the two - body interaction .
second , both these energies are much larger than the nnn binding energy .
these two features are quite general and common to all the metals we discuss here . for the simplest model we will neglect the effect of the nnn atoms , and assume a single nn binding energy , @xmath47 , for both lattice and bridge sites .
the resulting expression for the binding energy at the initial ( fourfold hollow ) site is @xmath48 the energy of an isolated atom is @xmath49 .
the energy of the hopping atom when it is on the bridge site is given by : @xmath50 where @xmath51 is the energy of an isolated atom on top of a bridge site .
thus , for a given configuration the barrier , @xmath52 , for an atom to hop into an adjacent vacant site is given in model i by : @xmath53 where @xmath54 and @xmath38 is given by eq .
( [ barindex ] ) . in this model
only three sites affect the energy barrier , which can take only four different values , as the expression in the parenthesis can be either 1 , 0 , -1 or -2 .
each of these four barrier values corresponds to one of the four groups in fig .
[ histogram ] .
the parameters of this model , as well as those of the models discussed below , are adjusted to best fit the eam data .
more specificly , we found the parameters that best describe the 128 eam barriers by minimizing the sum of squares : @xmath55 ^ 2 .
\label{optimize}\ ] ] the values obtained for these parameters for the five metals are shown in table [ add_table ] . despite its simplicity , model
i can be used to describe and analyze the main diffusion processes : single adatom hopping , attachment , detachment and edge diffusion .
the barriers obtained from this model can be incorporated in simulations to reproduce ( at least qualitatively ) experimental features such us cluster mobility , island morphology and island density @xcite .
the model presented above describes only the gross features of the diffusion process . in order to get more quantitative results , and to better understand the importance of the different processes , it is necessary to further refine the model
. we will now introduce model ii , in which the effect of nnn atoms in the initial configuration is included .
the expression for the energy at the initial site is now @xmath56 where @xmath57 is the reduction of the energy due to a nnn bond .
the energy barriers are now given by @xmath58 model ii accounts better for processes such as detachment , edge diffusion and vacancy diffusion , which generally involve nnn interactions . in the distribution of the barriers obtained from the model ,
the main groups exhibit certain widths .
yet , they are still significantly narrower than the groups of the eam barriers .
this is due to the fact that during the hopping process adjacent atoms may relax within their potential well .
model ii accounts for these effects only on average and therefore gives rise to narrower groups .
the values obtained as best fits of the model parameters to the eam data for the different metals are shown in table [ add_table ] .
further refinement can be obtained by introducing a distinction between the nn bond energies at the initial four fold hollow site and that at the bridge site , as suggested in ref.@xcite .
this modification , which introduces a fourth parameter into the model , gives only slightly better agreement with the eam results . in the following section we present a more effective refinement based on nonlinear interactions . to obtain models which provide a better fit to the eam barriers , it is necessary to consider effects that are caused by the simultaneous interactions of the hopping atom with several of its neighbors .
such effects may be described by expressions such us @xmath59 or @xmath60 , which are equal to 1 only if all the relevant sites are occupied , and 0 otherwise .
such expressions are clearly beyond the linear bond counting scheme of the previous section .
there is a large number of possible nonlinear interaction terms .
our analysis of the eam calculations , however , indicates that two of them are most significant .
the first term is related to the shape of the diffusion path .
it corresponds to configurations in which sites adjacent to the bridge site , are occupied on both sides of the diffusion path ( namely , at least one of the sites 1 and 2 , as well as at least one of the sites 5 and 6 are occupied ) .
it appears that in these cases the energy barrier is considerably higher than for configurations where sites on only one side of the path are occupied .
this effect is due to the `` stiffness '' of the diffusion path induced by the attraction from two opposite directions . even though there are nine different such configurations , they all contribute about the same energy difference , and hence can be bound to a single parameter @xmath61 ( for opposite ) .
the additional term that is now added to the expression for the barrier is @xmath62 where @xmath63 if at least one of the sites i and j is occupied , and 0 if both are empty . in all the metals we checked , except cu ,
this term is much larger than the nnn bond , sometimes by an order of magnitude .
the second nonlinear interaction term is smaller , and is comparable to the effect of nnn sites .
it is related to the energy of the hopping atom in the initial site .
the eam calculations indicate that if the two nearest neighbor sites 1 and 5 , that are symmetric with respect to the hopping direction are both occupied , then the initial configuration is more tightly bound .
this means that if sites 1 and 5 are both occupied , the energy barrier is expected to be higher .
consequently , the corresponding term would be : @xmath64 .
the fitted value obtained for @xmath65 is very close to that obtained for the nnn binding energy @xmath66 , for all five metals .
we thus included both contributions in the same term , although of different physical origin , to avoid the need for a fifth independent parameter .
the resulting model ( model iii ) for the hopping energy barriers is @xmath67 the values obtained from the best fit for the four parameters @xmath68 , @xmath47 , @xmath61 and @xmath69 for the different metals are given in table [ non_add_table ] .
there are two remarks to be made about eq .
( [ longform ] ) .
first , terms such as @xmath70 are _ not _ in contradiction to the assumption that only nearest and next - nearest neighbor interactions are significant .
these terms are just a manifestation of the simultaneous interactions of , say the atoms in sites @xmath71 and @xmath72 , with the hopping atom , of which they are both nearest - neighbors .
second , @xmath73 and @xmath74 are not included in the last term of eq .
( [ longform ] ) , since we found that their dominant contribution is in the nonlinear term .
the quality of the fit can be viewed in fig .
[ models ] .
the numbering of the configurations is the decimal representation of the binary number @xmath75 , where @xmath76 if site @xmath77 is occupied ( unoccupied ) , and @xmath78 is the opposite of @xmath79
. there are essentially 6 groups of barriers which are marked in the figures .
these groups correspond ( not necessarily in order ) to the six columns of table [ eam_table ] .
as can be seen , groups ii(a ) and ii(b ) are in the same energy range , and together form group ii in fig .
[ histogram ] .
similarly groups iii(a ) and iii(b ) coincide with group iii in fig .
[ histogram ] .
thus , there are actually only four groups as mentioned in sec .
[ eam_section ] . beyond this basic division ,
there are some significant differences among the metals which model iii seems to handle well .
the most important one is the effect of nnn atoms on the energy barrier .
it can be seen from table [ non_add_table ] that for ag and pd this effect is almost negligible , while for cu and ni it has much greater importance .
the effective nnn binding energy for au is even negative .
although it may be possible to construct models with the same number of parameters that would give better agreement with eam results for each specific metal alone , our approach is to find the general characteristics of diffusion mechanisms , common to different substrates .
the agreement between the eam barriers and model iii is slightly worse for au than for the other metals .
this may be due to substrate relaxation effects which are found to be more important in this metal , and are not accounted for in the model .
the models presented in the previous section help to identify the main physical mechanisms that determine the activation energies of self - diffusion processes on fcc(001 ) metal surfaces . although such processes may involve interactions with many substrate and in - plane atoms , they can be well described as the sum of few relatively simple terms .
the first term is the activation energy for hopping of an isolated atom .
the main corrections are due to nearest - neighbor in - plane atoms at the initial site , as well as at the bridge site . the former increase the energy barrier while the latter
decrease it by nearly the same amount .
the next contribution is due to simultaneous presence of atoms on both sides of the hopping atom relative to the hopping path .
this term is important in relatively dense environments .
its typical value is about half that of the nn binding energy .
a third and generally much smaller contribution consists of nnn bonds as well as a term associated with the simultaneous presence of atoms in both sites 1 and 5 ( fig .
[ fig : env ] ) . beyond the physical understanding gained by this analysis , the models can be used to evaluate the activation energy of any diffusion process on the ( 001 ) surface of the metals discussed above .
the models suggest that given a set of few activation energies ( which can be obtained from eam , ab - initio calculations or experiments ) , it is possible to extract the complete set of activation energy barriers . to realize model i , for example
, only two parameters are needed
. they can be obtained e.g. from the activation energy for single adatom hopping , and the dissociation energy of a dimer . to estimate a barrier using model ii ,
a third parameter is needed , which is the nnn binding energy .
this may be obtained if the mobility of a trimer is known .
the fourth parameter which is needed for model iii can be estimated from the activation energy for detachment from an atomic step . since model iii provides an expression for the energy barriers , linear in the parameters , any four barriers that give rise to four linearly independent equations , are sufficient to determine all four parameters , and consequently all the other barriers .
the possibility to construct a full set of activation energy barriers from a relatively small set of parameters is especially useful for simulations . without this knowledge ,
some processes have to be discarded from the simulations as unimportant , or assigned activation energies which are not fully substantiated .
these approaches take much of the power of computer simulations , and deny the possibility of direct quantitative confrontation with experimental data .
even if a list of all relevant activation energies is available , the model can be used to check the self - consistency of the data .
it can also help to interpret simulation results , which depend otherwise on a huge number of parameters .
the models presented here apply for diffusion on flat surfaces and do not describe the motion up / down steps .
such inter - terrace moves involve a large number of possible local enviroments , including flat steps as well as kink sites .
we believe that the approach proposed here can be extended to describe these processes as well . in summary
, we have constructed a family of models which describe self - diffusion on fcc(001 ) metal surfaces and tested them for cu , ag , au , ni and pd . for each one of these metals , the parameters of the models were optimized by comparing the energy barriers to a full set of barriers obtained from semi - empirical potentials via the embedded atom method .
it is found that these models , with at most four parameters , provide a good description of the hopping energy barriers on the fcc(001 ) surfaces .
, in _ scale invariance , interfaces and non - equilibrium dynamics _ , vol .
344 of _ nato advanced study institute , series b : physics _ , edited by m. droz , a. j. mckane , j. vannimenus and d. e. wolf ( plenum , new york , 1994 ) , . + & & & & & & & + & & & & & & & + & cu & 0.01 & 0.25 & 0.18 & 0.48 & 0.48 & 0.81 + & ag & 0.16 & 0.37 & 0.23 & 0.47 & 0.48 & 0.72 + 0 & au & 0.37 & 0.64 & 0.45 & 0.72 & 0.70 & 1.02 + & ni & 0.06 & 0.38 & 0.25 & 0.62 & 0.63 & 1.02 + & pd & 0.15 & 0.48 & 0.34 & 0.70 & 0.71 & 1.08 + & & & & & & & + & & & & & & & + & cu & 0.02 & 0.28 & 0.25 & 0.53 & 0.46 & 0.85 + & ag & 0.16 & 0.37 & 0.24 & 0.46 & 0.48 & 0.72 + 1 & au & 0.31 & 0.54 & 0.39 & 0.61 & 0.64 & 0.85 + & ni & 0.09 & 0.40 & 0.30 & 0.66 & 0.69 & 1.05 + & pd & 0.15 & 0.47 & 0.33 & 0.69 & 0.70 & 1.07 + & & & & & & & + & & & & & & & + & cu & 0.02 & 0.21 & 0.18 & 0.44 ( 0.34 ) & 0.46 & 0.74 ( 0.60 ) + & ag & 0.15 & 0.35 & 0.22 & 0.45 & 0.48 & 0.72 + 2 & au & 0.30 & 0.56 & 0.38 & 0.62 & 0.79 & 1.00 + & ni & 0.09 & 0.35 ( 0.10 ) & 0.26 & 0.57 ( 0.48 ) & 0.61 & 0.95 ( 0.79 ) + & pd & 0.16 & 0.43 ( 0.38 ) & 0.33 & 0.68 & 0.74 & 1.10 ( 0.91 ) + & & & & & & & + & & & & & & & + & cu & 0.05 & 0.25 & 0.24 & 0.48 & 0.54 & 0.78 ( 0.65 ) + & ag & 0.16 & 0.35 & 0.23 & 0.45 & 0.50 & 0.72 + 3 & au & 0.27 & 0.52 & 0.36 & 0.61 & 0.77 & 1.00 + & ni & 0.13 & 0.35 ( 0.33 ) & 0.31 & 0.63 & 0.68 & 0.99 + & pd & 0.17 & 0.49 & 0.35 & 0.70 & 0.77 & 1.13 + & & & & & & & + & & & & & & & + & cu & _ 0.02 _ & _ 0.28 _ & 0.21 & 0.50 & _ 0.46 _ & _ 0.85 _ + & ag & _ 0.16 _ & _ 0.37 _ & 0.24 & 0.47 & _ 0.48 _ & _ 0.72 _ + 16 & au & _ 0.31 _ & _ 0.54 _ & 0.38 & 0.67 & _ 0.64 _ & _ 0.85 _ + & ni & _ 0.09 _ & _ 0.40 _ & 0.30 & 0.65 & _ 0.69 _ & _ 1.05 _ + & pd & _ 0.15 _ & _ 0.47 _ & 0.33 & 0.69 & _ 0.70 _ & _ 1.07 _ + & & & & & & & + & cu & 0.04 & 0.30 & 0.28 & 0.54 & 0.66 & 0.89 + & ag & 0.17 & 0.36 & 0.24 & 0.47 & 0.49 & 0.72 + 17 & au & 0.26 & 0.57 & 0.34 & 0.60 & 0.59 & 0.83 + & ni & 0.13 & 0.43 & 0.35 & 0.68 & 0.75 & 1.09 + & pd & 0.15 & 0.48 & 0.34 & 0.70 & 0.71 & 1.07 + & & & & & & & + & & & & & & & + & cu & 0.05 & 0.27 & 0.23 & 0.49 & 0.52 & 0.80 + & ag & 0.16 & 0.35 & 0.22 & 0.44 & 0.49 & 0.72 + 18 & au & 0.26 & 0.57 & 0.34 & 0.55 & 0.75 & 0.98 + & ni & 0.12 & 0.40 & 0.30 & 0.63 & 0.66 & 1.01 + & pd & 0.17 & 0.48 & 0.34 & 0.68 & 0.75 & 1.11 + & & & & & & & + & & & & & & & + & cu & 0.08 & 0.29 & 0.29 & 0.52 & 0.61 & 0.83 + & ag & 0.17 & 0.35 & 0.24 & 0.45 & 0.51 & 0.73 + 19 & au & 0.26 & 0.54 & 0.32 & 0.64 & 0.82 & 0.99 + & ni & 0.16 & 0.42 & 0.36 & 0.66 & 0.74 & 1.04 + & pd & 0.19 & 0.50 & 0.37 & 0.72 & 0.79 & 1.15 + & & & & & & & + & & & & & & & + & cu & _ 0.02 _ & _ 0.21 _ & 0.18 & 0.48 & _ 0.46 _ & _ 0.74 ( 0.60 ) _ + & ag & _ 0.15 _ & _ 0.35 _ & 0.31 & 0.54 & _ 0.48 _ & _ 0.72 _ + 32 & au & _ 0.30 _ & _ 0.56 _ & 0.61 & 0.86 & _ 0.79 _ & _ 1.00 _ + & ni & _ 0.09 _ & _ 0.35 ( 0.10 ) _ & 0.32 & 0.65 ( 0.52 ) & _ 0.61 _ & _ 0.95 ( 0.79 ) _ + & pd & _ 0.16 _ & _ 0.43 ( 0.38 ) _ & 0.46 & 0.77 ( 0.66 ) & _ 0.74 _ & _ 1.10 _ + & & & & & & & + & & & & & & & + & cu & _ 0.05 _ & _ 0.27 _ & 0.28 & 0.54 & _ 0.52 _ & _ 0.80 _ + & ag & _ 0.16 _ & _ 0.35 _ & 0.32 & 0.54 & _ 0.49 _ & _ 0.72 _ + 33 & au & _ 0.26 _ & _ 0.57 _ & 0.58 & 0.81 & _ 0.74 _ & _ 0.98 _ + & ni & _ 0.12 _ & _ 0.40 _ & 0.38 & 0.69 ( 0.64 ) & _ 0.66 _ & _ 1.01 _ + & pd & _ 0.17 _ & _ 0.48 _ & 0.47 & 0.83 & _ 0.75 _ & _ 1.11 _ + & & & & & & & + & & & & & & & + & cu & 0.05 & 0.24 & 0.28 & 0.50 ( 0.16 ) & 0.55 & 0.78 ( 0.37 ) + & ag & 0.17 & 0.36 ( 0.26 ) & 0.38 & 0.58 ( 0.32 ) & 0.64 & 0.84 ( 0.45 ) + 34 & au & 0.32 & 0.59 & 0.60 & 0.84 & 1.02 & 1.20 + & ni & 0.12 & 0.39 ( 0.16 ) & 0.42 & 0.72 ( 0.29 ) & 0.75 & 1.07 ( 0.54 ) + & pd & 0.20 & 0.48 ( 0.24 ) & 0.53 & 0.83 ( 0.40 ) & 0.94 & 1.24 ( 0.66 ) + & & & & & & & + & & & & & & & + & cu & 0.08 & 0.26 ( 0.08 ) & 0.33 & 0.49 & 0.62 & 0.81 ( 0.47 ) + & ag & 0.18 & 0.36 ( 0.26 ) & 0.39 & 0.59 ( 0.35 ) & 0.65 & 0.84 ( 0.56 ) + 35 & au & 0.29 & 0.54 ( 0.28 ) & 0.57 & 0.81 ( 0.54 ) & 0.95 & 1.13 ( 0.85 ) + & ni & 0.15 & 0.41 ( 0.21 ) & 0.47 & 0.74 ( 0.37 ) & 0.82 & 1.09 ( 0.67 ) + & pd & 0.23 & 0.50 ( 0.27 ) & 0.56 & 0.86 ( 0.48 ) & 0.97 & 1.28 ( 0.82 ) + & & & & & & & + & & & & & & & + & cu & _ 0.05 _ & _ 0.25 _ & 0.28 & 0.51 & _ 0.54 _ & _ 0.78 ( 0.65 ) _ + & ag & _ 0.16 _ & _ 0.35 _ & 0.33 & 0.54 & _ 0.50 _ & _ 0.72 _ + 48 & au & _ 0.27 _ & _ 0.52 _ & 0.53 & 0.81 & _ 0.77 _ & _ 1.00 _ + & ni & _ 0.13 _ & _ 0.35 ( 0.33 ) _ & 0.38 & 0.68 ( 0.57 ) & _ 0.68 _ & _ 0.99 _ + & pd & _ 0.17 _ & _ 0.49 _ & 0.48 & 0.83 & _ 0.77 _ & _ 1.13 _ + & & & & & & & + & & & & & & & + & cu & _ 0.08 _ & _ 0.29 _ & 0.33 & 0.56 & _ 0.61 _ & _ 0.83 _ + & ag & _ 0.17 _ & _ 0.35 _ & 0.34 & 0.54 & _ 0.51 _ & _ 0.73 _ + 49 & au & _ 0.26 _ & _ 0.54 _ & 0.53 & 0.88 & _ 0.82 _ & _ 0.99 _ + & ni & _ 0.16 _ & _ 0.42 _ & 0.44 & 0.73 & _ 0.74 _ & _ 1.04 _ + & pd & _ 0.19 _ & _ 0.50 _ & 0.51 & 0.85 & _ 0.79 _ & _ 1.15 _ + & & & & & & & + & cu & _ 0.08 _ & _ 0.26 ( 0.08 ) _ & 0.34 & 0.51 ( 0.22 ) & _ 0.62 _ & _ 0.81 ( 0.47 ) _ + & ag & _ 0.18 _ & _ 0.36 ( 0.26 ) _ & 0.38 & 0.58 ( 0.34 ) & _ 0.65 _ & _ 0.84 ( 0.56 ) _ + 50 & au & _ 0.29 _ & _ 0.54 ( 0.28 ) _ & 0.52 & 0.83 ( 0.57 ) & _ 0.95 _ & _ 1.13 ( 0.85 ) _
+ & ni & _ 0.15 _ & _ 0.41 ( 0.21 ) _ & 0.47 & 0.73 ( 0.36 ) & _ 0.82 _ & _ 1.09 ( 0.67 ) _ + & pd & _ 0.23 _ & _ 0.50 ( 0.27 ) _ & 0.55 & 0.86 ( 0.47 ) & _ 0.97 _ & _ 1.28 ( 0.82 ) _ + & & & & & & & + & & & & & & & + & & & & & & & + & cu & 0.13 & 0.28 ( 0.12 ) & 0.40 & 0.53 ( 0.36 ) & 0.70 & 0.90 ( 0.69 ) + & ag & 0.19 & 0.36 ( 0.28 ) & 0.40 & 0.58 ( 0.48 ) & 0.67 & 0.85 ( 0.72 ) + 51 & au & 0.30 & 0.55 ( 0.44 ) & 0.53 & 0.83 ( 0.69 ) & 0.91 & 1.20 ( 0.99 ) + & ni & 0.21 & 0.43 ( 0.26 ) & 0.53 & 0.76 ( 0.55 ) & 0.89 & 1.12 ( 0.86 ) + & pd & 0.26 & 0.53 ( 0.34 ) & 0.60 & 0.89 ( 0.66 ) & 1.02 & 1.32 ( 1.04 ) + ( 25,25)(-15,122 ) ( 0,160)(1,0)30 ( 0,170)(1,0)30 ( 0,180)(1,0)30 ( 0,190)(1,0)30 ( 0,190)(0,-1)30 ( 10,190)(0,-1)30 ( 20,190)(0,-1)30 ( 30,190)(0,-1)30 ( 15,175 ) ( 19,175)(1,0)6 ( 5,185 ) ( 15,185 ) ( 5,165 ) ( 15,165 ) | we present a class of models that describe self - diffusion on fcc(001 ) metal substrates within a common framework .
the models are tested for cu(001 ) , ag(001 ) , au(001 ) , ni(001 ) and pd(001 ) , and found to apply well for all of them . for each of these metals
the models can be used to estimate the activation energy of any diffusion process using a few basic parameters which may be obtained from experiments , ab - initio or semi - empirical calculations . to demonstrate the approach , the parameters of the models are optimized to describe self - diffusion on the ( 001 ) surface , by comparing the energy barriers to a full set of barriers obtained from semi - empirical potentials via the embedded atom method ( eam ) .
it is found that these models with at most four parameters , provide a good description of the full landscape of hopping energy barriers on fcc(001 ) surfaces .
the main features of the diffusion processes revealed by eam calculations are quantitatively reproducible by the models . |
it is well - known ( see , e.g. , ( * ? ? ?
* th . 11.9 ) ) that , for @xmath1 , the set of values @xmath2 , @xmath3 , is dense in the complex plane .
in fact , much more is true : it was proved by bohr and jessen that there exists a borel probability measure @xmath4 on @xmath5 , such that the support of @xmath4 is the whole complex plane , and such that the convergence in law @xmath6 holds for @xmath7 continuous and bounded .
the corresponding density question for @xmath8 is , however , still open ( it was apparently first raised by ramachandra during the 1979 durham conference , but seems to appear in print only in heath - brown s note in @xcite ) : the difficulty is that the values @xmath9 , @xmath10 , do not have a limiting distribution , as evidenced already by the hardy - littlewood asymptotic @xmath11 or by selberg s result that @xmath12 , @xmath10 , is asymptotically normal with variance _ growing to infinity _
( see also the work of ghosh @xcite for the imaginary part , and @xcite for a recent proof ) .
in other words , `` most '' values of @xmath9 are rather large , though the zeta function is zero increasingly often as the imaginary part grows . in this paper ,
we show ( corollary [ cor - zeta ] ) how the density of values of zeta on the critical line would follow rather directly from a suitable version of the keating - snaith moment conjectures , which we viewed in our previous work with j. jacod @xcite as a refined version of the gaussian model .
in fact , under suitable assumptions , we could prove a quantitative result , bounding from above the smallest @xmath13 for which @xmath9 lies in a given open disc in @xmath5 .
this argument is based on a very general probabilistic estimate proved in section [ sec - proba ] , which throws some light on the nature of the mod - gaussian convergence that we defined in @xcite .
we hope that this result will be of further use . in another paper ( jointly with f. delbaen ,
see @xcite ) , it will be seen that one can weaken considerably the assumption needed in order to prove the density of values of @xmath9 ( but without quantitative information ) . as applications of the general result
, we will also prove the following theorems in section [ sec - examples ] ( the precise versions are given there ) .
[ th - rmt ] let @xmath14 be arbitrary , @xmath15 such that @xmath16 .
there exists @xmath17 , which can be bounded explicitly , such that @xmath18 provided @xmath19 , where @xmath20 denotes probability haar measure on the unitary group @xmath21 , and the implied constant is absolute .
[ th - eulerproduct ] define @xmath22 for @xmath23 and @xmath3 .
let @xmath14 be arbitrary , @xmath15 such that @xmath16 .
there exists @xmath17 , explicitly bounded , such that @xmath24 for all @xmath19 , where @xmath25 is the lebesgue measure and the implied constant is absolute . in a different direction , we obtain some evidence for the density of @xmath9 by looking at special values of families of @xmath0-functions over finite fields . in doing so , we also consider the analogue of theorem [ th - rmt ] for symplectic and orthogonal matrices .
we refer to section [ sec - ff ] for precise statements and definitions , and only state here one appealing ( qualitative ) corollary : [ th - ff ] the set of central values of the @xmath0-functions attached to non - trivial primitive dirichlet characters of @xmath26 $ ] , where @xmath27 ranges over primes , is dense in @xmath5
. * notation . * as usual
, @xmath28 denotes the cardinality of a set . by @xmath29 for @xmath30 , or @xmath31 for @xmath30 , where @xmath32 is an arbitrary set on which @xmath33 is defined , we mean synonymously that there exists a constant @xmath34 such that @xmath35 for all @xmath36 .
the `` implied constant '' refers to any value of @xmath37 for which this holds .
it may depend on the set @xmath32 , which is usually specified explicitly , or clearly determined by the context .
similarly , @xmath38 for @xmath30 means @xmath29 and @xmath39 , both for @xmath30 .
we write @xmath40 the pochhammer symbol .
* acknowledgments . * the first version of this paper was written while the first author was on sabbatical leave at the institute for advanced study ( princeton , nj ) ; many thanks are due to this institution for its support .
this material is based upon work supported by the national science foundation under agreement no .
dms-0635607 .
the second author was partially supported by snf schweizerischer nationalfonds projekte nr .
200021 119970/1 .
thanks to k. soundararajan for pointing out that the density conjecture for @xmath9 might be a good problem to study using complex moments of the zeta function , and to r. heath - brown for explaining us the history of the question .
many thanks also to n. katz for discussions and explanations surrounding issues of monodromy computations .
many thanks also to f. delbaen for useful discussions concerning the underlying probabilistic framework .
the graphs were produced using sage 4.2 @xcite , relying on the barnes function routines in the mpmath package .
in this section , which is purely probabilistic , we present two versions of `` local '' bounds for probabilities in the case of sufficiently uniform mod - gaussian convergence of sequences of random vectors .
this may be compared with the local central limit theorem ( see , e.g. , @xcite for the one - dimensional case ) .
in fact , as f. delbaen pointed out , one can obtain qualitative statements that generalize both the standard local central limit theorem and recover our results below under weaker assumptions . however , our emphasis is on explicit quantitative lower bounds for local probabilities .
we first introduce the definition , generalizing @xcite to random vectors .
fix some integer @xmath41 , and let @xmath42 be a sequence of @xmath43-valued random variables defined on a probability space @xmath44 ( as is the case for convergence in law , we could work without change with random variables defined on different probability spaces ) .
let @xmath45 be a sequence of non - negative quadratic forms on @xmath43 .
the sequence @xmath42 is then said to be convergent in the mod - gaussian sense with covariance @xmath46 and limiting function @xmath47 if @xmath48 locally uniformly for @xmath49 ; @xmath47 is then a function continuous at @xmath50 and @xmath51 . here
, @xmath52 denotes the standard inner product on @xmath43 .
the intuitive meaning is that , in some sense , @xmath53 is `` close '' to a ( centered ) gaussian vector @xmath54 with covariance matrix @xmath46 . as in @xcite , this notion is of most interest if the covariance `` goes to infinity '' .
however , in contrast with the case of @xmath55 , this can mean different things because there is more than a single variance parameter involved . to discuss this
, we diagonalize @xmath46 in an orthonormal basis . ] in the form @xmath56 where @xmath57 is the necessary ( orthogonal ) change of variable .
then `` @xmath46 goes to infinity , '' in the weakest sense , means that the largest eigenvalue @xmath58 goes to @xmath59 as @xmath60 .
we are interested in the distribution of values of @xmath53 as @xmath61 grows ; clearly , if ( say ) the @xmath46 are already diagonalized in the canonical basis and @xmath62 is constant , these values will have first coordinate much less spread out than the last ones . to simplify our discussion , and because this is the situation in our applications , we will assume this behavior does not occur and that in fact the smallest eigenvalue goes to infinity . for simplicity
, we will assume in fact that for some fixed @xmath63 , we have @xmath64 so that also @xmath65 ( we say that the convergence is _ balanced _ ) . in our main applications , this will be the case with @xmath66 .
note in particular that it follows that the discriminant @xmath67 goes to infinity as @xmath60 , and moreover @xmath68 our question is now the following : given @xmath42 , @xmath69 , as above , with this type of mod - gaussian convergence , can we bound from below the probability @xmath70 where @xmath71 is a fixed open set in @xmath43 ? denoting by @xmath72 the dual quadratic form , the gaussian model suggests that , if @xmath71 is relatively compact ( e.g. , some non - empty open ball ) , we could expect @xmath73 as @xmath60 , since @xmath65 implies that @xmath74 for all @xmath75 .
we will confirm that this holds in certain conditions at least .
we strive especially for lower bounds on @xmath76 , which we wish to be quantitative , so that we can determine some @xmath77 ( depending explicitly on @xmath71 ) for which @xmath78 note that such a quantitative result must depend on the location of the open set @xmath71 , whereas the limit itself only depends on the volume , as seen above . the specific hypothesis we use may seem somewhat arbitrary , but they turn out to be satisfied ( with room to spare ) in the later applications . [ th-2 ] let @xmath41 be fixed and let @xmath42 be a sequence of @xmath43-valued random variables defined on @xmath44 , such that @xmath42 converges in the mod - gaussian sense with covariance @xmath69 , and that the convergence is @xmath79-balanced with @xmath63 , with @xmath80 for all @xmath61 .
let @xmath81 be gaussian random variables with covariance matrices given by @xmath69 , so that @xmath82 assume moreover the following three conditions : _ ( 1 ) _ there exist constants @xmath83 , @xmath84 and @xmath85 such that , for any @xmath23 and @xmath49 such that @xmath86 , we have @xmath87 _ ( 2 ) _ the function @xmath47 is of class @xmath88 on @xmath89 . _ ( 3 ) _ for some @xmath90 and @xmath91 , we have @xmath92 .
let @xmath93 be any number such that @xmath94 then , for any fixed non - empty open box @xmath95 with width @xmath96 such that @xmath97 , we have @xmath98 for @xmath23 , where the implied constant depends only on @xmath99 and the implied constant in _ ( [ eq - strong - convergence])_. in particular , for any fixed non - empty open set @xmath100 , we have @xmath101 provided @xmath19 , where @xmath77 and the implied constant depend on @xmath71 and the same data as above .
note the following elementary lower bound , valid if @xmath102 : @xmath103 where the implied constant depends only on @xmath104 ; this is where the location of @xmath71 enters , since the error term will only be smaller than this , roughly , when @xmath105 .
the growth condition ( [ eq - order-2 ] ) is in fact a consequence of the uniform mod - gaussian convergence , at least provided the sequence @xmath106 does not grow too fast .
for instance , if @xmath107 for all @xmath23 , for some constants @xmath108 , @xmath109 , we can obtain ( [ eq - order-2 ] ) with @xmath110 .
indeed , we can write @xmath111 if @xmath61 is large enough and @xmath86 .
note then that @xmath112 we now fix @xmath23 minimal such that @xmath113 and if this value of @xmath61 is large enough , we get @xmath114 , and hence @xmath115 as desired . on the other hand , if this chosen @xmath61 is too small , @xmath116 is bounded , and the desired estimate is trivial .
let @xmath117 be the smallest eigenvalue of @xmath46 , so that @xmath118 for all @xmath49 and @xmath23 . for simplicity
, we denote also @xmath119 as in ( [ eq - strong - convergence ] ) .
we now first fix @xmath120 such that @xmath121 , and then fix a smooth , compactly supported function @xmath122 on @xmath123 such that @xmath124 where @xmath125 and the implied constant depends only on @xmath126 ( we will define @xmath120 to be a function of @xmath61 at the end , and hence must be careful to have estimates uniform in terms of @xmath120 ; this is provided by using only the above properties of @xmath122 ; the maximal value of @xmath126 used will also be bounded only in terms of @xmath104 , @xmath47 and the data in ( [ eq - strong - convergence ] ) ) .
it is classical that such a function exists ( examples are constructed in @xcite ) .
then define @xmath127 for @xmath75 .
it follows that @xmath128 next , we define @xmath129 and we start our argument with the obvious inequality @xmath130 where @xmath131 is the law of @xmath53 . applying the plancherel formula , we get @xmath132 where @xmath133 is the characteristic function of @xmath53 and @xmath134 denotes the fourier transform of @xmath33 ( the smoothness of @xmath33 guarantees that @xmath135 is in @xmath136 , so the plancherel formula is valid by a simple fubini argument ) .
we have @xmath137 since @xmath138 for @xmath139 and @xmath140 ( by repeated integration by parts ) , we find that @xmath141 the implied constant depending on @xmath126 . using the formula for @xmath142 , selecting for given @xmath143 an index @xmath126 so that @xmath144 and applying the second upper bound above for this index only , if @xmath145 , we derive @xmath146 for any fixed @xmath147 , where the implied constant depends only on @xmath104 and @xmath148 . in particular , for @xmath149
, we will use simply the upper bound @xmath150 .
we now proceed to approximate .
first of all , for any radius @xmath151 , we can estimate the contribution of those @xmath143 with @xmath152 using the estimate above with a value of @xmath147 which will be determined later .
after integrating over @xmath153 , we obtain @xmath154 for any @xmath151 . after selecting @xmath155
, we obtain @xmath156 for @xmath23 , the implied constant depending only on @xmath157 . on the other hand ,
provided @xmath158 , we use ( [ eq - strong - convergence ] ) and ( [ eq - order-2 ] ) to get @xmath159 for @xmath23 , using again the definition of @xmath33 , the implied constant depending on @xmath47 . by ( [ eq - uniformity ] ) ,
the last term can be bounded by @xmath160 for @xmath23 if @xmath161 , the implied constant depending on @xmath162 .
we then split the first term further in two parts , namely where @xmath163 , and where @xmath164 .
the parameter @xmath165 will be chosen later , in such a way that the region @xmath166 ( which is inside @xmath167 ) contains the region @xmath163 ( which is a neighborhood of @xmath50 that contracts to @xmath50 as @xmath60 , if @xmath165 does not grow too fast , since it is an ellipsoid with longest axis @xmath168 ) .
the second part of the integral is bounded by @xmath169 and in the first part , we use the approximation @xmath170 for @xmath166 , coming from the @xmath88 assumption on @xmath47 , to get @xmath171 the implied constant depending on @xmath47 , where the last integral was estimated using latexmath:[\[\|t\|^2{\leqslant}\frac{1}{\delta_n}q_n(t){\leqslant}\frac{\kappa_n^2}{\delta_n},\quad\quad we can rewind the computation for the first term , with the gaussian @xmath54 instead of @xmath53 : for @xmath23 , we have @xmath173 where the implied constant is absolute , and then we write @xmath174 for @xmath23 , where the implied constant depends only on @xmath104 ( the last step is obtained using the density of the gaussian @xmath54 ) . summarizing , we have found @xmath175 ( where we recall that @xmath176 ) .
now if @xmath177 , and @xmath178 we have first ( see ( [ eq - mmu ] ) ) the condition @xmath179 and moreover the third error term is absorbed in the second one , while the last becomes @xmath180 thus if we select any @xmath181 , the previous conditions on @xmath148 hold , and we find that @xmath182 now , we attempt to select @xmath120 to equalize the error terms involving it , i.e. , so that @xmath183 which translates to @xmath184 and two cases arise : \(i ) if @xmath185 , we have @xmath186 ( as is necessary to define @xmath120 in this way ) , and we obtain from the above that @xmath187 where @xmath188 ( note that , since @xmath189 is assumed to be @xmath190 , we have @xmath191 . ) \(ii ) if we have @xmath192 , we take simply @xmath193 and obtain @xmath194 where the implied constant depends on @xmath104 and @xmath148 .
the combination of the two cases leads to the lower - bound in ( [ eq - quantitative ] ) .
the upper bound is proved similarly , using instead of @xmath122 a function which is @xmath195 for @xmath196 and @xmath197 for @xmath198 for some suitable @xmath199 ; we leave the details to the reader .
[ rm - lower - bound ] if we are interested in obtaining a lower bound only ( which is the most interesting aspect in a number of applications ) , it is simpler and more efficient to fix , e.g. , @xmath193 , from the beginning of the proof . for @xmath200
this leads for instance to @xmath201 for any @xmath202 ( by taking @xmath148 large enough depending on @xmath203 ) , where @xmath204 and the implied constant depends on @xmath47 , @xmath205 and @xmath203 .
[ rm - tv ] from the probabilistic point of view , this proposition gives one answer , quantitatively , to the following type of question : given a sequence @xmath42 of ( real - valued ) random variables and parameters @xmath206 such that @xmath207 converges in law to a standard centered gaussian variable ( with variance @xmath208 ) , to what extent is @xmath53 itself distributed like a gaussian with variance @xmath209 ?
here , we perform the comparison by looking at @xmath76 , @xmath71 a fixed open set . and
this shows clear limits to the gaussian model : for example , any integer - valued random variable @xmath53 will have @xmath210 for any open set not intersecting @xmath211 , and yet may satisfy a central limit theorem ( e.g. , the @xmath61-th step of a symmetric random walk on @xmath211 ) .
even more precisely , denoting @xmath212 the kolmogorov distance , there exist integer - valued random variables @xmath53 with @xmath213 where @xmath54 is a centered gaussian with variance @xmath61 , which indicates a close proximity and yet , again , @xmath210 if @xmath214 .
for instance , take @xmath53 with distribution function @xmath215 given by @xmath216 in particular , it is also impossible to prove results like theorem [ th-2 ] using only this type of assumption on the kolmogorov distance . of course ,
if we assume that @xmath217 with @xmath218 ( arbitrarily slowly ) , we get @xmath219 for @xmath220 $ ] , @xmath221 .
but such an assumption is unrealistic in practice . for instance , if one assumes that @xmath42 converges in the mod - gaussian sense with covariance @xmath222 , and if the limiting function @xmath47 is @xmath88 and the convergence holds in @xmath88 topology , one can straightforwardly estimate the kolmogorov distance by @xmath223 which is comparable to ( [ eq - dist - k ] ) , but one can also check that this can not be improved in general to something like ( [ eq - impossible ] ) .
in example [ ex - eigenvalue - count ] in section [ sec - examples ] , we will also describe a much deeper and more illuminating situation concerning the limits of what can be hoped , even with something like mod - gaussian convergence .
[ rm - variant ] other variants could easily be obtained . in particular , it is clear from the proof that if @xmath47 has sub - gaussian growth , i.e. , we can take @xmath224 in ( [ eq - order-2 ] ) , the results can be substantially improved .
however , in our main applications , this condition fails .
also , one could use test functions @xmath33 which decay at infinity faster than polynomials to weaken the uniformity requirement in the convergence condition ( [ eq - strong - convergence ] ) ( for instance , for @xmath55 , it is possible to find @xmath33 which is smooth , non - negative and compactly supported in any fixed open interval and satisfies @xmath225 for any @xmath15 , as constructed , e.g. , in ( * ? ? ?
1.3.5 ) or @xcite ) . again
, for our main unconditional applications , our conditions hold with room to spare , so we avoided this additional complexity .
we can also introduce a linear term ( corresponding roughly to the expectation of @xmath53 ) in addition to the covariance terms in the definition of mod - gaussian convergence ( as in @xcite for @xmath55 ) , but this amounts to saying that @xmath42 converges in the mod - gaussian sense with covariance @xmath69 and mean @xmath226 , @xmath227 , if the sequence @xmath228 converges in our original sense above . but note that the interpretation of a lower bound for @xmath229 , as given by theorem [ th-2 ] for a fixed @xmath100 , is quite different , if @xmath230 is itself `` large '' .
maybe one should see the statements in that case as giving natural examples of sets @xmath231 for which one knows that @xmath232 has the specific decay behavior @xmath233 as @xmath60 .
in particular , lower bounds for @xmath229 do not give control of @xmath76 , and indeed this may be zero for all @xmath61 large enough ( see , for instance the example in section [ ssec - symplectic ] below of values at @xmath208 of characteristic polynomials of unitary symplectic matrices , which is always @xmath234 ) .
we present now some applications of theorem [ th-2 ] ( in particular , proving theorems [ th - rmt ] and [ th - eulerproduct ] ) .
we also give an example that illustrates the limitations of such results , suggesting strongly that one can not replace mod - gaussian convergence with the existence of the limits ( [ eq - converge ] ) only for @xmath143 in a neighborhood of the origin .
[ ex - keating - snaith ] one of the canonical motivating examples of mod - gaussian convergence is due to keating and snaith @xcite
. let @xmath235 where @xmath236 is a haar - distributed random unitary matrix in the compact group @xmath237 ; we view these random variables as @xmath238-valued ( via the real and imaginary parts ) , so if @xmath239 , we have @xmath240 we first clarify the choice of the branch of logarithm : @xmath53 is defined almost everywhere ( when @xmath208 is not an eigenvalue of @xmath236 ) , and for @xmath241 with @xmath242 , such that @xmath243 where the last logarithms are given by the taylor expansion around @xmath50 .
this is the same convention as in ( * ? ? ?
after ( 7 ) ) .
keating and snaith show that @xmath42 satisfies @xmath244 for @xmath245 ( note the asymmetry between @xmath246 and @xmath247 ) , see ( * ? ? ?
* eq . ( 71 ) ) , taking into account a slightly different normalization : their @xmath143 is our @xmath246 and their @xmath248 is our @xmath247 .
it is also useful to observe that , in the connection with tplitz determinants ( see @xcite ) , this characteristic function is the @xmath61-th tplitz determinant corresponding to the symbol which is a pure fisher - hartwig singularity of type @xmath249 ( this is denoted @xmath250 in @xcite , where the formula ( [ eq - ks ] ) is stated as eq . ( 41 ) ; see @xcite for two elementary computations of the corresponding tplitz determinants . )
we rewrite this in terms of the barnes function @xmath251 , as is customary .
we recall that @xmath252 is an entire function of order @xmath253 , such that @xmath254 and @xmath255 for all @xmath256 , and that its zeros are located at the negative integers .
in particular , it satisfies @xmath257 for all @xmath23 and @xmath258 .
thus , we have @xmath259 ( see , e.g. , ( * ? ? ?
* eq . ( 41 ) ) ) .
we now get from ( * ? ? ?
3.2 ) that @xmath260 where @xmath261 hence we have complex mod - gaussian convergence with limiting function @xmath262 here we can take @xmath66 in ( [ eq - grow - not - too - much ] ) . note that this is _ not _ the product of the two individual limiting functions for mod - gaussian convergence of the real and imaginary parts of @xmath53 separately ( which are @xmath263 and @xmath264 ) , although after normalizing , one obtains convergence in law of @xmath265 to _ independent _ standard gaussian variables , as noted by keating and snaith .
we now check that theorem [ th-2 ] can be applied to the sequence of random variables @xmath42 .
the fact that @xmath266 is of class @xmath88 on @xmath238 is clear in view of the analytic properties of the barnes function .
condition ( 3 ) is also obvious .
a uniformity estimate like ( [ eq - strong - convergence ] ) is not found in @xcite or @xcite , though it is proved for @xmath143 in a fixed compact region of @xmath238 in ( * ? ? ?
3.2 ) . in proposition
[ pr - app ] in the appendix , we prove @xmath267 for @xmath268 . in view of @xmath269 ,
this is compatible with ( [ eq - strong - convergence ] ) and ( [ eq - uniformity ] ) , with @xmath270 arbitrarily large , @xmath271 and @xmath189 ( defined by ( [ eq - order-2 ] ) ) can be taken to be any @xmath272 .
thus the constant @xmath273 in ( [ eq - quantitative ] ) can be any number @xmath274 moreover , since @xmath275 we see by using remark [ rm - lower - bound ] ( namely , ( [ eq - quant - lower ] ) ) and ( [ eq - lower - easy ] ) that we have the following corollary : for @xmath276 we have @xmath277 for all @xmath61 with @xmath278 where both implied constants are absolute .
( the first condition on @xmath61 ensures the main term is @xmath279 , while the second ensures that the error term is smaller ; we have taken @xmath280 for definiteness ; any number @xmath281 can replace @xmath282 ) .
now we appeal to the following elementary lemma : [ lm - log ] let @xmath283 and @xmath15 , and denote @xmath284-\pi,\pi].\ ] ] then , provided @xmath16 , we have @xmath285 for all @xmath286 such that @xmath287 for given @xmath288 , non - zero , we get from this and ( [ eq - dense - quant - log ] ) , applied to @xmath289 and to @xmath290 instead of @xmath96 , the following explicit form of theorem [ th - rmt ] : [ th - rmt - quant ] let @xmath14 be arbitrary , @xmath15 such that @xmath16 .
we have @xmath291 for @xmath292 , where @xmath293 where @xmath37 and the implied constants are absolute .
it follows from asymptotic formulas for the barnes function , e.g. @xcite , that we have @xmath294 which is illustrated in figure 1 .
this super - gaussian behavior is the main cause of difficulty in the proof of theorem [ th-2 ] .
, @xmath295,width=384 ] note that if one only wants to say that @xmath296 , for @xmath61 growing , has dense image in @xmath5 , much simpler topological arguments suffice .
another conspicuous example of mod - gaussian convergence is the arithmetic euler factor in the moment conjecture for @xmath9 . in @xcite
, we considered this factor for the real part @xmath297 only , and we first generalize this as in the previous section .
consider a sequence @xmath298 of independent random variables uniformly distributed on the unit circle and indexed by prime numbers , and let @xmath299 where the logarithm is given here by the taylor expansion around @xmath50 .
for each individual term @xmath300 , we have @xmath301 with @xmath302 .
expanding by the binomial theorem and picking up the constant term in the expansion in fourier series , we obtain @xmath303 in terms of the gauss hypergeometric function . arguing as in @xcite
, we see now that @xmath304 and hence , denoting @xmath305 we get @xmath306 as @xmath60 , with limiting function @xmath307 we have here also @xmath66 in ( [ eq - grow - not - too - much ] ) .
now , to check the uniformity required in ( [ eq - strong - convergence ] ) , we write @xmath308 with @xmath309 if we expand the @xmath27-factor using the binomial theorem , we obtain @xmath310 and if we assume that @xmath311 with @xmath90 , crude bounds show that this @xmath27-factor is @xmath312 where the implied constant is absolute , so that if @xmath313 , for instance , we get @xmath314 although this is crude , it already gives much more than ( [ eq - strong - convergence ] ) , both in terms of range of uniformity and sharpness of approximation .
since condition ( 3 ) is also obviously valid here , theorem [ th-2 ] ( or rather ( [ eq - quant - lower ] ) ) applies with @xmath189 any real number @xmath315 , @xmath270 and @xmath37 arbitrarily large , and shows that @xmath316 for any @xmath317 and @xmath318 , provided @xmath319 for some large constant @xmath320 . from this
we deduce easily the more arithmetic - looking statement of theorem [ th - eulerproduct ] .
indeed , let @xmath321 be given by ( [ eq - pnt ] ) . for fixed @xmath61 ,
it is well - known that the random variables @xmath322 on the probability spaces @xmath323,t^{-1}\lambda)$ ] converge in law , as @xmath324 , to @xmath325 where @xmath326 are as above ( independent and uniformly distributed on the unit circle ; the independence is due to the fundamental theorem of arithmetic ) . for any open set @xmath327 ,
is in the boundary of @xmath327 is zero or not , we do not claim or need to claim an equality ; see , e.g. , ( * ? ? ?
2.1 , ( iv ) ) .
] it follows that @xmath328 applying lemma [ lm - log ] as in the proof of theorem [ th - rmt - quant ] , we obtain theorem [ th - eulerproduct ] with @xmath329 for some absolute constant @xmath37 . again , the density of values of @xmath321 for @xmath23 and @xmath3 ( or of @xmath330 , which amounts to the same thing ) is an easier matter that can be dealt with using topological tools .
[ ex - zeta ] the two previous examples are of course motivated by their conjectural relation with the behavior of the riemann zeta function on the critical line ( this is the arithmetic essence of @xcite ) . indeed , keating and snaith conjecture that : [ cj - ks ] define @xmath331 , when @xmath332 is not the ordinate of a non - trivial zero of @xmath333 , by continuation along the horizontal line @xmath334 , with limit @xmath50 when @xmath335 . for any @xmath245
, we have @xmath336 as @xmath324 , where @xmath52 is the inner product on @xmath238 as in _ ( [ eq - inner - product])_. hence , we see in particular that the following holds : [ cor - zeta ] assume there exist @xmath84 , @xmath337 and @xmath338 such that conjecture _ [ cj - ks ] _ holds with the error term @xmath339 replaced by @xmath340 uniformly for @xmath341 .
then the set of values @xmath9 is dense in the complex plane . in fact , there exists @xmath85 , @xmath342 , such that , for any @xmath14 and @xmath16 , there exists @xmath143 with @xmath343 such that @xmath344 of course , such a strong conjecture concerning the imaginary moments of @xmath9 looks quite hopeless at the current time : there is no known non - trivial result available , even assuming the riemann hypothesis .
but example [ ex - eigenvalue - count ] below suggests that ( with this approach ) it is indeed necessary to require that the characteristic function converge uniformly for @xmath143 in a region growing with @xmath345 . in @xcite , jointly with f. delbaen , we will explain how the weaker qualitative statement @xmath346\,\mid\ , \zeta({{\textstyle{\frac{1}{2}}}}+iu)\in v\})\gg \frac{1}{\log\log t}\ ] ] for a fixed open set @xmath327 and for @xmath345 large enough ( which of course suffices to give a positive answer to ramachandra s question ) can be proved under much weaker assumptions than a uniform version of conjecture [ cj - ks ] .
another remark concerning conjecture [ cj - ks ] has to do with the factored form of the limiting function @xmath347 , which seems to imply some asymptotic independence property .
recall that the real and imaginary parts of @xmath348 and @xmath266 are themselves asymptotically independent _
after renormalization _ , but are not products of the limiting functions for the two parts separately .
so conjecture [ cj - ks ] , if correct , is evidence of quite particular probabilistic behavior . and
a few other results do provide evidence of a factored limiting function , with `` random matrix '' term split from the euler factor .
the mod - poisson analogy is also consistent with this , in the case of the number of prime factors of an integer , as discussed in detail in @xcite . ]
[ ex - eigenvalue - count ] our assumptions in theorem [ th-2 ] are probably not optimal .
we now describe an illuminating ( counter)-example in the direction of understanding when a result like this could be true .
we again look at random matrices @xmath236 in the compact group @xmath237 ( as in example [ ex - keating - snaith ] ) , but this time we consider the random variables counting the number of eigenvalues in certain fixed arcs of the unit circle : fix @xmath3490,1/2[$ ] , and let @xmath350 then let @xmath53 be the number of eigenvalues @xmath351 of @xmath236 such that @xmath352 . note that @xmath53 is an integer - valued random variable .
it was proved by costin and lebowitz that @xmath353 converges in law to a standard normal random variable .
wieand @xcite gave a proof based on asymptotics of tplitz determinants with discontinuous symbols ; as noted by basor , this gives the asymptotic @xmath354 as @xmath60 , for all @xmath143 _ with _
@xmath355 ( see , e.g. , ( * ? ? ?
* th . 5.47 ) , applied with @xmath356 , @xmath357 , @xmath358 , @xmath359 , and the condition on @xmath143 is equation ( 5.79 ) in loc .
cit . , or @xcite ) .
this asymptotic is of course of the form ( [ eq - converge ] ) for these values of @xmath143 , but the restriction @xmath355 is necessary , since the characteristic function of @xmath53 is @xmath360-periodic for all @xmath61 .
the convergence is sufficiently uniform for @xmath143 close to @xmath50 to allow the deduction of the renormalized normal behavior ( as wieand did , using the laplace transform instead of the characteristic function ) , but when @xmath203 is rational , the set of possible values of @xmath361 for @xmath23 is a discrete set in @xmath123 . is true for irrational @xmath203 , by dirichlet s approximation theorem , and by the existence of matrices in @xmath237 where the number of eigenvalues in @xmath362 takes any value between @xmath50 and @xmath61
we now consider examples related to @xmath0-functions over finite fields .
our main input will be deep results of deligne and katz , and we are of course motivated by the philosophy of katz and sarnak @xcite . the goal is to make statements about the distribution of values at the central point @xmath363 of @xmath0-functions over finite fields .
the appealing aspect is that these form discrete sets , hence proving that they are dense in @xmath5 ( as in theorem [ th - ff ] ) , for instance , is obviously interesting and meaningful .
we consider examples of our results for the three basic symmetry types in turn : unitary , symplectic , and orthogonal .
for the last two , this means first obtaining a suitable analogue of example [ ex - keating - snaith ] .
the corresponding limiting functions have already been studied in some respect by keating - snaith @xcite and conrey - farmer @xcite , though our expressions seem somewhat more natural .
let @xmath364 be a finite field with @xmath365 elements .
unitary symmetry arises ( among other cases ) for certain types of one - variable exponential sums over finite fields , which are associated to dirichlet characters of @xmath366 $ ] , which we now describe ; these will lead to a proof of theorem [ th - ff ] .
let @xmath364 be a finite field with @xmath365 elements of characteristic @xmath367 .
a _ dirichlet character _
modulo @xmath368 $ ] is a map @xmath369{\rightarrow}{{{\mathbf c}}},\ ] ] defined by @xmath370 where @xmath371 is a group homomorphism @xmath372/g{\mathbf{f}}_q[x])^{\times}{\longrightarrow}{{{\mathbf c}}}^{\times}.\ ] ] this character is non - trivial if @xmath373 , and primitive if it can not be defined ( in the obvious way ) modulo a proper divisor of @xmath374 .
the associated @xmath0-function is defined by the euler product @xmath375 for @xmath376 , where the product ranges over monic irreducible polynomials in @xmath366 $ ] and @xmath377 .
one shows quite easily that this is in fact a polynomial ( which we denote @xmath378 ) in the variable @xmath379 of degree @xmath380 , if @xmath381 is primitive modulo @xmath374 and non - trivial .
the examples used in proving theorem [ th - ff ] arise from the following well - known construction . for any integer @xmath382 , with @xmath383 , any non - trivial multiplicative character @xmath384 such that @xmath385 , and any squarefree polynomial @xmath386 $ ] of degree @xmath387 , we let @xmath388 where @xmath389 is defined to be @xmath50 .
these are multiplicative exponential sums , and have been studied intensively , due in part to their many applications to analytic number theory ( for their generalizations to multiple variables , see the paper @xcite of katz ) .
it is also well - known that one can construct a non - trivial dirichlet character @xmath390 , primitive modulo @xmath374 , such that @xmath391 where @xmath392 denotes the `` companion '' sums over extensions of @xmath364 , namely @xmath393 where @xmath394 is the norm map .
we will denote @xmath395 the corresponding @xmath0-function .
moreover , we have the riemann hypothesis for these @xmath0-functions ( due to a. weil ) , which gives the link with random unitary matrices : there exists a unique conjugacy class @xmath396 in the unitary group @xmath397 such that @xmath398 ( so that , in particular , we recover the weil bound @xmath399 by looking at the trace of @xmath396 ) . for all this
, one can see , for instance , @xcite , which contains a self - contained account .
we will first prove the following theorem , which is clearly a stronger form of theorem [ th - ff ] in view of the preceding remarks : [ th - ff - strong ] for @xmath382 and @xmath400 , let @xmath401 $ ] .
for @xmath27 prime , let @xmath402 denote the set of pairs @xmath403 where @xmath404 is non - trivial character of @xmath405 and @xmath406 .
let @xmath283 and @xmath15 with @xmath16 be given .
for all integers @xmath407 , we have @xmath408 where @xmath409 where @xmath34 and the implied constants are absolute .
this result depends on the mod - gaussian convergence for characteristic polynomials on @xmath237 ( i.e. , on example [ ex - keating - snaith ] ) . indeed , denoting by @xmath410 the space of conjugacy classes in @xmath237 ,
we have the following : [ th - katz ] for any integer @xmath411 , any odd prime @xmath27 with @xmath412 , the conjugacy classes @xmath413 become equidistributed in @xmath414 as @xmath415 , with respect to haar measure .
this is an easy consequence of results of katz ( see ( * ? ?
5.13 ) ) , the only `` twist '' being the extra averaging over all non - trivial dirichlet characters to obtain unitary instead of special unitary ( or similar ) equidistribution .
first of all , it is easy to check that if @xmath412 and @xmath406 is such that @xmath416 , the polynomial @xmath417 $ ] is a strong deligne polynomial in one variable ( in the language of @xcite , these are called `` weakly - supermorse '' polynomials ) .
hence the conjugacy classes in the statement are well - defined
. for simplicity , denote @xmath418 the open subset of the affine @xmath143-line where @xmath416 .
now , according to the weyl equidistribution criterion , we must show that @xmath419 for any ( fixed ) non - trivial irreducible unitary representation @xmath420 of the compact group @xmath397 .
we isolate in the sum those characters @xmath404 where @xmath421 : there are at most @xmath422 of them . for any other character @xmath423 , the inner sum over @xmath424 is of the type handled by the deligne equidistribution theorem .
let @xmath425 be the order of the dirichlet character @xmath426 , where @xmath427 is the real character modulo @xmath27 . by ( * ?
? * th . 5.13 , ( 2 ) ) ( the restriction @xmath428 ensures the assumptions hold ) ,
provided @xmath412 , e.g. , @xmath429 , the relevant geometric monodromy group is equal to @xmath430 with maximal compact subgroup @xmath431 for simplicity , we write @xmath432 , @xmath433 .
the restriction of @xmath420 to @xmath434 is a finite sum of irreducible representations of this group ( possibly including trivial components ) . applying ( * ? ? ?
* th . 9.2.6 , ( 5 ) ) to each of the non - trivial ones ( and the obvious identity for the trivial components ) , we find that @xmath435 where the implied constant is _ absolute _
( this is because we have a one - parameter family , so we can apply @xcite and the fact proved in @xcite that the relevant sheaf is everywhere tame , so the swan - conductor contribution is zero ; the parameter curve @xmath418 has @xmath387 points at infinity , which give the factor @xmath387 above ) . using frobenius reciprocity or direct integration ( using , e.g , ( *
* lemma ad.7.1 ) ) , we find that the multiplicity of the trivial representation in @xmath420 ( restricted to @xmath434 ) satisfies @xmath436 for a given @xmath420 , this is equal to @xmath208 only if @xmath437 , so @xmath438 for some @xmath439 , and if @xmath404 is such that @xmath440 .
the number of such characters @xmath404 is therefore @xmath441 , the implied constant depending on @xmath420 .
hence we find , after adding back the characters with @xmath421 , that @xmath442 where the implied constant depends on @xmath420 .
this confirms the claimed equidistribution .
this an easy consequence of theorem [ th - katz ] : for any open set @xmath443 , we have first that @xmath444 as @xmath27 goes to infinity , and then we can write @xmath445 and then we apply theorem [ th - rmt - quant ] .
in theorem [ th - katz ] , we performed the average over @xmath404 , because for `` standard '' families of exponential sums ( those parametrized by points of algebraic varieties ) , the ( connected component of the ) geometric monodromy group is always semisimple , so its center is finite and its maximal compact subgroup can never be @xmath237 .
however , one may expect that @xmath143 could be fixed in the example above , and that ( for instance ) the conjugacy classes @xmath446 corresponding to the exponential sums @xmath447 ( where @xmath448 , @xmath412 ) would be already equidistributed in @xmath414 as @xmath415 .
very recently , n. katz @xcite has indeed shown that for such families there is always an a - priori `` sato - tate law '' , i.e. , that the conjugacy classes become equidistributed in @xmath449 for some compact group @xmath450 .
in fact , in this work , katz proves the required equidistribution for ( most ) fixed @xmath143 _ in the vertical direction _ where one looks at characters of @xmath451 for fixed @xmath365 and @xmath452 ( see ( * ? ? ?
17.6 , remark 17.7 ) ) .
it seems quite possible that the `` horizontal '' direction we are interested in will also follow from these new techniques .
similarly , it is likely that families of hyper - kloosterman sums of certain types exhibit full unitary monodromy , e.g. , for any integer @xmath453 , any additive character @xmath454 , any multiplicative character @xmath455 , one can consider the sums @xmath456 for which the basic theory ( due to deligne @xcite ) shows that the associated @xmath0-function ( unitarily normalized so the central point is @xmath457 ) is @xmath458 for some unique conjugacy class @xmath459 .
one would then expect ( this was suggested by katz , and is again potentially in the realm of his recent work @xcite ) that if we fix a character @xmath460 for @xmath405 ( e.g. , @xmath461 ) and define @xmath462 for all @xmath41 , the sets of conjugacy classes @xmath463 become equidistributed in @xmath464 as @xmath452 .
a typical example of symplectic symmetry involves families of @xmath0-functions of algebraic curves over finite fields . for simplicity
, we will consider one of the simplest ones , but we first start by proving distribution results for characteristic polynomials of symplectic matrices , which are of independent interest . we first remark that for @xmath465 , the characteristic polynomial can be expressed in the form @xmath466 for some eigenangles
@xmath467 , @xmath468 , and it follows that @xmath469 this positivity is reflected in a shift in expectation in the mod - gaussian convergence ( it also means that the argument is not an interesting quantity here ) . we obtain : [ pr - symplectic ] for @xmath470 , let @xmath471 where @xmath472 is a haar - distributed random matrix in the unitary symplectic group @xmath473 .
then @xmath474 converges in mod - gaussian sense with @xmath475 and limiting function @xmath476 indeed , we have @xmath477 for @xmath478 , where the implied constant is absolute .
figure 2 is a graph illustrating the logarithmic growth of @xmath479 .
, @xmath480,width=384 ] ( compare ( * ? ? ?
4.9 ) ) keating - snaith ( * ? ? ?
* ( 10 ) ) compute that @xmath481 which , together with the formula ( [ eq - ratio - gamma ] ) , gives @xmath482 by applying proposition [ pr - app ] , ( 3 ) in the appendix , we get @xmath483 as claimed
. in particular , the central limit theorem for @xmath484 takes the form of the convergence in law @xmath485 so , for any @xmath486 , we have @xmath487 on the other hand , by applying theorem [ th-2 ] , as we can according to the previous proposition , we can control the probability of the values of @xmath484 in much smaller ( dyadic or similar ) intervals : let @xmath488a , b[$ ] with @xmath486 real numbers .
we have @xmath489 for @xmath490 , where the implied constant is absolute .
in particular @xmath491 if @xmath492 , we apply theorem [ th-2 ] , with the constants @xmath66 , @xmath189 arbitrarily close to @xmath253 , @xmath270 arbitrarily large and @xmath493 , so that @xmath273 can be any number with @xmath494 and in particular @xmath495 is valid . if @xmath496 , we split the interval @xmath497a , b[$ ] into @xmath498 intervals of length @xmath499 and apply the previous case to the interior of those intervals . since the joint distribution of eigenvalues of @xmath472 is absolutely continuous with respect to lebesgue measure , the probability of falling on one of the missing endpoints is zero , and summing over these intervals gives the result .
we now deduce an arithmetic corollary , using families of hyperelliptic curves over finite fields . for any odd @xmath365 , any integer @xmath470 and any squarefree monic polynomial @xmath500 $ ] of degree @xmath501 ,
let @xmath502 be the smooth projective model of the affine hyperelliptic curve @xmath503 the number of @xmath451-rational points on @xmath502 satisfies @xmath504 where @xmath427 is the quadratic character of @xmath505 and the notation is as in section [ ssec - unitary ] .
the associated @xmath0-function ( the numerator of the zeta function ) is defined by @xmath506 or , in other words , we have @xmath507 weil proved that @xmath508 is a polynomial in @xmath509 $ ] , of degree @xmath510 , all roots of which have modulus @xmath511 , and which is _ symplectic _ : there is a unique conjugacy class @xmath512 in @xmath473 such that @xmath513 [ th - hyperelliptic ] let @xmath514 be the set of squarefree , monic , polynomials of degree @xmath501 in @xmath366 $ ] .
fix a non - empty open interval @xmath497\alpha,\beta[\subset ] 0,+\infty[$ ] . for all @xmath374 large enough
, we have @xmath515\alpha,\beta [ \bigr\}\bigr| } \gg \frac{1}{\sqrt{\log g}}.\ ] ] ( note that this is in fact a very weak version of what we can prove ) .
let first @xmath516 be the set of @xmath517 for which @xmath518 . in ( * ?
4.9 ) , we showed , using the relevant equidistribution computation in @xcite that the ( real - valued ) random variables @xmath519 on @xmath516 ( with counting measure ) converges in law to @xmath471 where @xmath472 is a random matrix in the unitary symplectic group @xmath473 , distributed according to haar measure .
the previous proposition shows that theorem [ th-2 ] is applicable to @xmath474 with covariance @xmath475 and limiting function @xmath520 . letting @xmath521 as in the previous section
, we get @xmath522\alpha,\beta[\bigr\}\bigr| \gg \frac{1}{\sqrt{\log g}}\ ] ] for @xmath374 large enough .
since @xmath523 for fixed @xmath374 and @xmath521 ( by an easy application of the equidistribution , see ( * ? ? ?
4.9 ) ) , we get the result stated by exponentiating .
the lower bound ( [ eq - symplectic ] ) is good enough to combine with various other statements proving arithmetic properties of @xmath0-functions which hold for `` most '' hyperelliptic curves .
for instance , from ( * ? ? ? * prop .
1.1 ) ( adapted straightforwardly to all hyperelliptic curves instead of special one - parameter families ) , it follows that if we denote by @xmath524 the set of @xmath517 such that the eigenvalues of @xmath525 satisfy no non - trivial multiplicative relation ,- linear independence of the ordinates of zeros of the riemann zeta function ; non - trivial refers to a relation that can not be deduced from the fact that , if @xmath526 is an eigenvalue , so is its inverse @xmath527 . ]
then we have @xmath528 for some @xmath529 , and hence we get @xmath530\alpha,\beta [ \bigr\}\bigr| } \gg \frac{1}{\sqrt{\log g}}\ ] ] for @xmath374 large enough .
orthogonal symmetry , in number theory , features prominently in families of elliptic curves .
in contrast with symplectic groups , there are a number of `` flavors '' involved , due to the `` functional equation '' @xmath531 for an orthogonal matrix @xmath532 ( the standard maximal compact subgroup of the orthogonal group @xmath533 ) , which implies that @xmath534 is zero for `` trivial '' reasons if @xmath61 is even and @xmath535 or @xmath61 is odd and @xmath536 .
when this happens , it is of great interest to investigate the distribution of the first derivative at @xmath537 of the reversed characteristic polynomial . for simplicity , however , we restrict our attention here to @xmath61 even and matrices with determinant @xmath208 , i.e. , to the subgroup @xmath538 of @xmath539 , where @xmath23 . in that case , it is also true that eigenangles come in pairs of inverses , and therefore we have @xmath540 . as in the previous section ,
we start with random matrix computations .
[ pr - orthogonal ] for @xmath23 , let @xmath541 where @xmath236 is a haar - distributed random matrix in the special orthogonal group @xmath538 .
then @xmath53 converges in mod - gaussian sense with @xmath542 and limiting function @xmath543 indeed , we have @xmath544 for @xmath545 , where the implied constant is absolute .
using ( * ? ? ?
* ( 56 ) ) and ( [ eq - ratio - gamma ] ) , we get @xmath546 and by applying proposition [ pr - app ] , ( 4 ) in the appendix , we get the desired formula @xmath547 if we compare with the symplectic case , we observe the ( already well - established ) phenomenon that the value @xmath534 , for @xmath548 tend to be small , whereas they tend to be large for symplectic matrices in @xmath473 .
our arithmetic corollary is based on families of quadratic twists of elliptic curves over function fields , and we select a specific example for concreteness ( see @xcite ) ; the basic theory , which we illustrate here , is again due to katz @xcite . for any odd prime power @xmath549 , any integer @xmath23 , we consider the elliptic curves over the functional field @xmath550 given by the weierstrass equations @xmath551 where @xmath552 is a parameter such that @xmath256 is not a critical value of @xmath553 .
katz proved that the associated @xmath0-function ( which is now defined by the `` standard '' euler product over primes in @xmath554 $ ] , with suitable ramified factors ) is of the form @xmath555 where @xmath556 is a unique conjugacy class in @xmath539 .
[ th - orthogonal ] fix a non - empty open interval @xmath497\alpha,\beta[\subset ] 0,+\infty[$ ] . for all @xmath61 large enough
, we have @xmath557\alpha,\beta [ \bigr\}\bigr| } \gg \frac{1}{\sqrt{\log n}}.\ ] ] as recalled in ( * ? ? ?
4.4 and before ) , for all @xmath558 and primes @xmath27 with @xmath559 and @xmath560 , the conjugacy classes @xmath561 , for @xmath562 not a critical value , become equidistributed in @xmath563 for the image of haar measure ( precisely , this is stated for the `` vertical direction '' where @xmath27 is fixed and finite fields of characteristic @xmath27 and increasing degree are used ; however , because the parameter variety is a curve with @xmath564 points at infinity and the relevant sheaf is tame , we can recover the horizontal statement as in the proof of theorem [ th - katz ] ) .
in particular , there is a subset @xmath565 with @xmath566 where @xmath567 and those restricted conjugacy classes become equidistributed in @xmath568 .
hence , for @xmath61 large enough , we get @xmath569\alpha,\beta [ \bigr\}\bigr| \\{\geqslant}\mu_{so(2n,{{{\mathbf r}}})}(\{a\,\mid\ , \log \det(1-a)\in]\log \alpha,\log\beta[\ } ) \gg \frac{1}{\sqrt{\log n}},\end{gathered}\ ] ] as desired . obviously , this result ( or its generalizations to other families of quadratic twists over function fields ) has interesting consequences concerning the problem of the distribution of the order of tate - shafarevich groups of the associated elliptic curves , through the birch and swinnerton - dyer conjecture ( which is known to be valid in its strong form for many elliptic curves over function fields over a finite field with analytic rank @xmath50 or @xmath208 ) .
we hope to come back to this question , and its conjectural analogue over number fields , in another work .
we present in this appendix some uniform analytic estimate for the barnes function , which are needed to verify the strong convergence assumption ( [ eq - strong - convergence ] ) for sequences of random matrices in compact classical groups .
note that we did not try to optimize the results .
if we expand the logarithm , defined using the taylor expansion of @xmath578 at the origin , we have @xmath579 for @xmath580 , with an absolute implied constant .
hence we obtain @xmath581 with an absolute implied constant , for @xmath453 and @xmath582 , hence for @xmath571 we get @xmath583 since @xmath584 for @xmath453 , with absolute implied constants . hence , we have @xmath585 for @xmath571 , for some absolute implied constant , and we get the stated formula @xmath586
\(2 ) note first that the conditions @xmath23 and @xmath268 ensure that the values of the barnes function in the denominator are non - zero .
next , let @xmath587 , @xmath588 ; we can express the ratio of barnes function as @xmath589 and @xmath268 gives @xmath590 , @xmath591 , allowing us to apply ( [ eq - power - ratio ] ) three times .
the exponential terms cancel out , leading to @xmath592 which gives the first part of the proposition .
\(3 ) we use a similar computation , applying ( [ eq - power - ratio ] ) six times with the parameters @xmath593 @xmath594 leading , after an easy calculation , to a main term @xmath595 for the ratio of barnes functions , and some further computation leads to the stated result ( each parameter @xmath256 has @xmath596 , so the error term is also as given ) .
e. kowalski : _ the large sieve , monodromy , and zeta functions of algebraic curves , 2 : independence of the zeros _ , international mathematics research notices ( 2008 ) , article rnn091 , 57 pages , doi:10.1093/imrn / rnn091 | in the context of mod - gaussian convergence , as defined previously in our work with j. jacod , we obtain asymptotic formulas and lower bounds for local probabilities for a sequence of random vectors which are approximately gaussian in this sense , with increasing covariance matrix .
this is motivated by the conjecture concerning the density of the set of values of the riemann zeta function on the critical line .
we obtain evidence for this fact , and derive unconditional results for random matrices in compact classical groups , as well as for certain families of @xmath0-functions over finite fields . |
one of the most basic operations of linear optics is the interference between two optical modes in a two - path interferometer . in quantum optics
, this scenario has been widely studied in the context of quantum metrology , where it serves to illustrate the role of non - classical correlations in the phase sensitivity of multi - photon interference @xcite .
unfortunately , these arguments about phase sensitivity often leave out the details of the physics , focusing merely on the perceived figures of merit and not on the mechanisms by which multi - photon interference patterns emerge .
recent experiments on multi - photon interference are providing a much more detailed picture of non - classical effects in two - mode interferences @xcite , and it might be time to ask what these experimental observation of multi - photon fringes can tell us about the nature and origin of quantum interference . to properly address this question ,
it is necessary to distinguish carefully between classical wave interference and quantum interference . in typical experiments ,
multi - photon fringes are detected by measuring the phase dependence of count rates for a fixed combination of input and output conditions .
quantum interference thus appears as a modulation in the phase dependence of a single multi - photon probability .
this is quite different from classical interference , where the output intensity is a deterministic function of the phase shift and probabilities appear only as a technical noise background . in quantum optics , classical interference
is usually associated with the poissonian photon counting statistics of coherent states , which describes a highly localized increase of detection probability for the classically expected intensity distribution at a specific phase shift in the interferometer . in the limit of high photon number
, we can distinguish classical interference and multi - photon quantum interference qualitatively , since only the latter effect is associated with multiple interference fringes in the phase dependence of the output probabilities , and these interference fringes have distinct periodicities that are much shorter than @xmath0 .
it is therefore a non - trivial challenge to explain the relation between the modulations of probability seen in multi - photon interference and the statistics of classical interference effects observed in the phase dependence of output intensities . in the present paper ,
we address this problem by examining the role of the operators representing the coherence of the two field modes .
these operators have a clear classical meaning , describing the intensity differences between any two modes as a component of a three dimensional vector in close analogy to the algebra of quantized spins .
we point out that the phase dependence of the quantum state components representing a specific measurement outcome can be described in terms of weak values of the intensity difference between the two arms of the interferometer .
these weak values can be written as a function of the eigenvalues of the initial and the final state using relations that correspond to the classical causality of two - path interference .
for the case of photon number states in both the input and the output , we can then derive a differential equation describing the phase dependence of a single interference fringe , where the main contribution to the phase dependence originates from the rapid oscillation of the fringes , while the slowly varying envelope describes the statistics expected for a classical interference of two fields when the optical phases of the input fields are completely random .
it is possible to identify the periodicity of the multi - photon fringes that describe the quantum mechanical modulation of output probability with the intensity difference between the paths obtained from the weak value of the squared operator representing this intensity difference .
the periodicity of multi - photon interference fringes can thus be explained in terms of a classical estimate of the intensity difference between the paths obtained from the experimentally controlled input and output conditions .
our results show that multi - photon interference can be understood in terms of the classical relations between the intensities and coherences of the two modes .
specifically , the quantum mechanical phase of the multi - photon interference fringes is given by a classical action that relates the optical phase shift to its generator , the intensity difference between the paths .
the non - classical effects of multi - photon interference can thus be traced to the fundamental role of the action in quantum physics @xcite .
it may be worth noting that the direct identification of the action of phase shifts also provides a more efficient approach to the quantization of fields that starts from the macroscopically observable physics and hence avoids many of the ambiguities associated with the mathematical concepts of superpositions and state vectors .
multi - photon interference may thus help us bridge the gap between quantum mechanical concepts and classical intuition in a new and unexpected manner .
the paper is organized as follows : in sec .
[ sec : phase ] , we review the operator algebra of two - path interference and its relation to the interference fringes observed in multi - photon experiments . in sec . [
sec : wv ] , it is shown that weak values can be used to express the phase dependence of a single multi - photon interference fringe .
based on this observation , we derive a differential equation for the interference fringes obtained with well - defined photon numbers in the input and the output . in sec . [
sec : separate ] , the differential equation derived in sec . [ sec : wv ] is solved approximately by separating the interference fringe into an interference term described by a phase dependent action @xmath1 and an envelope function @xmath2 .
it is shown that the action evolves according to a classical hamilton - jacobi equation , while the envelope describes the statistics of classical random phase interference . in sec .
[ sec : examples ]
we apply the theory to a number of characteristic cases , comparing the approximate results to exact results obtained for eight and sixteen photons .
the results show that the approximation correctly describes the main features of multi - photon interference , especially with regard to the separation of quantum interference effects and classical random phase field interference . in sec .
[ sec : action ] we identify the necessary conditions for the identification of multi - photon interference fringes with a classical action function .
it is shown that weak values can be used to identify the action of multi - photon phase interferences for a wide variety of possible input states . in sec .
[ sec : noon ] we discuss the relation between the general analysis of multi - photon interference using the action and the definition of multi - photon interference based on superpositions of photon number eigenstates in the interferometer paths associated with the well known noon states .
it is pointed out that the noon states represent a special case of the general action - based theory .
[ sec : concl ] summarizes the results and concludes the paper .
the physics of two optical modes can be described in terms of the field operators @xmath3 and @xmath4 that describe the complex field amplitudes of the two modes . due to their mathematical effects on photon number states ,
these field operators are commonly known as annihilation operators , although it should be kept in mind that this mathematical effect is not related to the physical properties described by the operators in any obvious or intuitive manner .
the proper connection between two mode coherences and photon number is obtained by considering the second order products of field amplitudes , @xmath5 each component of this three dimensional vector represents an intensity difference between two orthogonal modes in units of @xmath6 per photon . the motivation for this choice of units is the representation of phase shifts between the modes @xmath3 and @xmath4 , which is generated by the component @xmath7 in the same way that a hamiltonian generates the time evolution .
specifically , the unitary transformation of a phase shift is given by @xmath8 and the effect of a phase shift on an arbitrary state @xmath9 can be described by @xmath10 here , the product @xmath11 in eq .
( [ eq : unitary ] ) represents the action of a phase shift , just as the action of a time evolution is given by the energy - time product @xmath12 . in a two path interferometer
, @xmath7 describes the intensity difference between the two paths in the interferometer .
the intensity difference between the input modes is usually given by @xmath13 .
the remaining component @xmath14 describes the phase coherence between the input fields that results in interferences when a phase shift of @xmath15 is applied . for a phase shift of @xmath15 ,
the intensity difference observed in the output can be given by @xmath16 in the following , we will consider input states @xmath17 that are eigenstates of the input intensity difference @xmath13 , and measurements of the output intensity difference @xmath18 .
specific measurement results can be given by an integer or half - integer value of @xmath19 representing one half of the photon number difference between the output ports .
since the total photon number @xmath20 is conserved , @xmath19 represents a photon number state of the two output ports with output photon numbers of @xmath21 and @xmath22 .
the quantum states @xmath23 representing these measurement results are eigenstates of @xmath18 with eigenvalues of @xmath24 .
we are therefore interested in the phase dependent values of the output probabilities given by @xmath25 where the unitary transformation is used to relate the state @xmath26 to the input state @xmath27 .
experimentally , these probabilities are obtained as multi - photon coincidence rates for the detection of @xmath22 photons in one output port and @xmath21 photons in the other .
quantum interference is observed as a phase dependent modulation of the probability , with a periodicity that can be as short as @xmath28 for @xmath20 photons .
we will now analyze the phase dependence of @xmath29 to identify the origin of these quantum interference fringes in multi - photon interference .
we can apply the transformation of quantum states given by eq.([eq : schroed ] ) to describe the phase evolution of the probability amplitudes @xmath29 associated with specific measurement outcomes @xmath19 . in its most conventional form ,
the resulting differential equation is given by @xmath30 in the textbook approach to quantum dynamics , the operator @xmath7 is usually expanded into its matrix representation in the measurement basis @xmath23 . however , this may not be the most meaningful analysis of the role of the generator @xmath7 in the evolution of the input - output relation @xmath31 . a closer correspondence to classical dynamics can be maintained by identifying the contribution of @xmath7 with its weak value , @xmath32 it is interesting to note that the weak value can be used to express the effects of the operator @xmath7 on the phase evolution of the state component @xmath33 .
the advantage of this approach is that it is often possible to find the weak value without having to solve the complete dynamics in the schrdinger picture .
specifically , weak values can be determined by expressing the operator as a function of two operators , where the initial state is an eigenstate of the first and the final state is an eigenstate of the second operator . in the present case
, @xmath17 is an eigenstate of @xmath13 and @xmath34 is an eigenstate of @xmath18 , which can be related to the operators at @xmath35 using eq.([eq : heis ] ) . since the algebra of the @xmath36 operators is the familiar algebra of spin operators for a total spin quantum number of @xmath37 , it is possible to derive a relation between the phase shift generator @xmath7 and the components @xmath13 and @xmath18 in the @xmath13-@xmath14 plane orthogonal to @xmath7 by using the total length of the @xmath38-vector , @xmath39 since the initial and the final state are eigenstates of @xmath38-vector components orthogonal to @xmath7 , it is not possible to distinguish negative values of @xmath7 from positive values of @xmath7 . as a result of this symmetry ,
the real part of the weak value of @xmath7 is zero at all phases @xmath15 , and the ratio between the amplitude @xmath29 and its phase derivative in eq.([eq : action ] ) is always real .
it is therefore possible to express @xmath29 by real numbers for all values of @xmath15 . to make optimal use of the relation in eq.([eq : jquad ] ) , we now consider the second derivative of the phase dependence , @xmath40 this derivative is described by the weak value of @xmath41 , and this weak value can be determined using the eigenvalues @xmath42 and @xmath19 for the input state and the measurement outcome , respectively .
it is important to arrange the order of the operators so that the operator @xmath13 is always to the right of the operator @xmath18 , and the necessary application of the commutation relations results in a contribution from the imaginary weak value of @xmath7 .
the final relation between the weak values and the eigenvalues therefore reads @xmath43 since both the weak value of @xmath41 and the weak value of @xmath7 appear in the phase derivatives of @xmath33 , we can use this relation between the weak values and the eigenvalues to find a differential equation for the phase dependence of the probability amplitude @xmath33 that does not depend on the probability amplitudes of any other measurement outcomes @xmath19 .
according to eqs.([eq : action ] ) and ( [ eq : quadact ] ) , this differential equation can be written as @xmath44 we have thus derived a general mathematical description of the multi - photon interference fringes observed with any photon number input . in the following , we will consider the physics described by this differential equation and identify the characteristic features of its solutions .
the description of the effects of phase shifts on multi - photon states given by eq.([eq : fringes ] ) makes it possible to separate quantum effects from the classical limit by considering how the relation changes with total photon number .
classical effects should all scale with the total intensity , whereas the magnitude of quantum effects will always depend on absolute photon numbers . in the present case
, such a separation of scales can be achieved by expressing the probability amplitudes @xmath33 by a product of a slowly varying envelope function @xmath45 and a quantum interference effect given by an action @xmath46 that describes the rapid modulation of probability associated with multi - photon interference , @xmath47 the factor of two represents the interference between two classical solutions , as will be seen more clearly in the following discussion .
note that there is no approximation involved at this point , and the separation in eq.([eq : separate ] ) can be used to obtain an exact solution of eq.([eq : fringes ] ) for a specific photon number . however , our main concern is the comparison between the quantum effects that are expressed by the rapid oscillation of @xmath48 and the much slower variation of @xmath45 that corresponds more closely to the classical statistics of random phase interference . in the following ,
we will therefore focus on approximate solutions , where the different scales of the phase dependence allow a complete separation between the dynamics of @xmath46 and the dynamics of @xmath45 .
specifically , the use of @xmath49 in eq.([eq : separate ] ) allows us to expand eq.([eq : fringes ] ) in @xmath49 , where the classical limit emerges when action differences of @xmath49 are not resolved . in the limit of sufficiently large photon numbers ,
we can therefore identify quantum effects as effects that depend on the precise ratio between the macroscopic action @xmath46 and the fundamental constant @xmath49 . in eq.([eq :
fringes ] ) , the derivatives result in contributions that depend explicitly on the fundamental constant @xmath49 . due to the smallness of @xmath49
, we can expand the equation and neglect higher order contributions of @xmath49 in favor of the lower order contributions .
the leading terms of the expansion are proportional to @xmath50 , which means that the left hand side of eq .
( [ eq : fringes ] ) can be represented by the square of the derivative of @xmath46 , @xmath51 note that the largest contribution of the first derivative of @xmath33 in eq.([eq : fringes ] ) is proportional to @xmath52 , so that the contribution of the first derivative to the left hand side of eq.([eq : fringes ] ) can be neglected except at phases where the leading contribution of @xmath53 becomes very small . for non - vanishing values of @xmath53 , the approximate solution of the equation
is given by @xmath54 where the choice of sign is arbitrary since the action @xmath1 is defined by the phase of a cosine in eq.([eq : separate ] ) .
we choose the negative sign because it corresponds to the conventional definition of action in classical hamilton - jacobi theory , as will be seen below .
importantly , the right hand side of this equation has a very intuitive physical meaning : it is the classical value of @xmath7 obtained from the length of the @xmath38-vector and the values of @xmath13 and @xmath18 when the heisenberg relation between the operators in eq.([eq : heis ] ) is converted to a classical relation between the eigenvalues by neglecting the non - commutativity of @xmath18 and @xmath13 .
we can define this classical approximation of @xmath55 as @xmath56 which describes the intensity difference between the arms of a classical two - path interferometer when an input intensity difference of @xmath57 and a phase shift of @xmath15 result in an output intensity difference of @xmath24 .
the approximate quantum mechanical solution to eq .
( [ eq : fringes ] ) given by eq.([eq : sapprox ] ) then corresponds to the classical relation between the generator @xmath55 and the action @xmath1 given by a version of the hamilton - jacobi equation , @xmath58 the reason why this action describes a quantum effect is that it appears as the quantum phase of the multi - photon interference fringes described by eq.([eq : separate ] ) , where the classical action @xmath1 is converted into a quantum phase by dividing it by @xmath49 .
thus , the larger the classical action @xmath1 becomes , the faster the output probability oscillates , resulting in microscopic ( and therefore highly phase sensitive ) modulations of probability in the limit of macroscopic action .
the approximations used above apply whenever the modulation dominates the phase dependence , which is the case whenever @xmath55 is sufficiently larger than @xmath49 .
we can now turn to the approximate solution for the slowly varying envelope function @xmath45 .
since the quantum effects are for the most part described by the interference effects associated with the action in eq.([eq : separate ] ) , we expect the square of the envelope function @xmath45 to represent the classical probability density of the multi - photon interference scenario .
specifically , the phase dependence of @xmath45 can be determined by considering the second largest term in the expansion of eq.([eq : fringes ] ) in @xmath49 , which is proportional to @xmath52 .
the result can be written as @xmath59 the general solution of this equation does not depend on the specific value of @xmath55 and can be expressed by a relation between @xmath45 and @xmath46 given in more compact form as @xmath60 since the phase derivative of the action is given by ( [ eq : hje ] ) , it is possible to express the squared envelope function @xmath61 as a function of @xmath55 and a normalization factor of @xmath62 , @xmath63 this relation corresponds to a classical density of output intensities , where the factor of @xmath49 originates from the quantized distance between two eigenvalues of @xmath13 .
once the factor of @xmath49 is removed , the probability density scales with total intensity , indicating that the envelope function @xmath2 represents statistics that are also observable in the classical limit .
it is in fact possible to derive an expression for @xmath62 based on classical considerations .
if we assume that the probability distribution originates from random phase interference between the two input intensities , we can explain the probability densities in terms of the phase difference @xmath64 between the input modes .
since the output value of @xmath18 is a deterministic function of the fixed input value of @xmath13 and the random phase @xmath64 , the ratio of probability densities for intervals of @xmath65 and intervals of @xmath66 is given by @xmath67 thus the factor in the denominator of eq.([eq : ergo ] ) effectively converts probability densities in the random phase @xmath64 into probability densities in the output observable @xmath68 .
for homogeneous phase distributions , the factor of @xmath62 in eq.([eq : ergo ] ) is given by @xmath69 using these relations , it is possible to find a properly normalized solution for the envelope function @xmath45 .
in the light of this result , we can conclude that the slowly varying envelope function describes the classical statistics originating from the random optical phase of the photon number states in the input .
specifically , the phase dependence of @xmath61 given by eq.([eq : ergo ] ) is independent of the absolute photon number if the photon number differences of @xmath70 and @xmath71 are given by constant fractions of total photon number . in the limit of high @xmath20
, this means that the shape of the envelope @xmath45 does not change much if the photon number changes by about @xmath72 , corresponding to the poissonian photon number fluctuations of coherent states .
it is therefore also possible to observe @xmath45 using random phase interferences between two coherent states of light , where the relative error in @xmath73 caused by photon number fluctuations drops to zero with @xmath74 . on the other hand ,
the multi - photon interference patterns described by @xmath46 disappear in the statistics of coherent states because their periodicities are much shorter than the absolute shot noise value of @xmath72 in the limit of high photon number . we can therefore confirm that @xmath45 is classical in the sense that it can also be observed using input states that can be represented by mixtures of coherent states as described by a positive p - function , while the multi - photon fringes described by @xmath46 are non - classical since they disappear when the input can be described as a mixture of coherent states . in combination with the solution for the action @xmath1 , we now have an approximate solution for @xmath33 that is valid whenever the values of @xmath55 are sufficiently large . for multi - photon states , this condition is usually satisfied if there is a classical solution that relates the input intensity to the output intensity for a phase shift of @xmath15 .
the condition only breaks down when the argument in the square root of eq.([eq : j3 ] ) is very small or negative .
note that eq.([eq : fringes ] ) can also be solved in the regime of negative @xmath75 values , where the differential equation describes evanescent solutions characterized by an exponential suppression of the amplitude @xmath33 with increasing distance from the region of positive @xmath75 .
mathematically , this situation is equivalent to the evanescent wavefunctions known from tunneling .
in fact , a comparison of eq.([eq : fringes ] ) with the time independent schrdinger equation shows that the mathematics are quite similar , and that the method of approximation used above corresponds to the wkb - approximation in the region where the energy is larger than the potential energy .
the regions where @xmath55 drops to zero and becomes imaginary can therefore be treated in close analogy to the wkb approximation for the tunneling problem , by allowing evanescent solutions where the cosine function in eq.([eq : separate ] ) is replaced by a corresponding exponential function .
the intermediate regime close to @xmath76 corresponds to the turning points of the wkb solution and can be approximated by airy functions . in this manner , it is possible to obtain an approximate solution of eq.([eq : fringes ] ) at all photon numbers .
however , we are mostly concerned with the multi - photon interference fringes observed for @xmath77 , so a detailed analysis of the evanescent parts of the phase dependence is not necessary in the present context .
the main merit of the present analysis is that it allows us to separate the quantum interference effects described by the action @xmath1 from the effects of classical field statistics originating from the randomness of the optical phases in the input , which are represented by the envelope function @xmath2 . in the following
, we will apply this analysis to several representative examples of multi - photon interference fringes in order to illustrate how the approximations used here reproduce the characteristic features of multi - photon interference for different input - output relations .
in the measurement scenarios we are considering , the input photon number distribution and the output photon number distribution are both fixed , and the phase dependence of the probability for this specific combination is investigated . for a given total photon number @xmath20 , different measurement scenarios are defined by the combination of input photon number difference @xmath71 and output photon number difference @xmath78 . to illustrate our method of analysis ,
it is best to start with a particularly simple class of scenarios , where the photon numbers in both input and output ports are equal , so that @xmath79 . in this case
, we find that the intensity difference between the two arms of the interferometer does not depend on phase and is simply given by the total length of the @xmath38-vector ,
latexmath:[\[\label{eq : hb } in terms of classical wave interference , it is easy to understand why equal intensities in both the input and the output require a maximal intensity difference between the paths . since both @xmath13 and @xmath18 are zero , we can immediately conclude that @xmath14 should also be zero , so that the @xmath38-vector must point in the direction of @xmath7 .
it is also possible to argue that we do not observe any interferences between the two paths , and this is only possible when all of the light travels along only one of the two paths . as mentioned above
, @xmath55 is actually derived from the weak value of @xmath41 , so the choice of sign in eqs.([eq : j3 ] ) and ( [ eq : hje ] ) is merely an arbitrary convention . in terms of the actual physics
, the scenario is complete symmetric in the two paths , so it is impossible to distinguish positive and negative values of @xmath55 .
the quantum interferences described by @xmath48 in eq.([eq : separate ] ) are a consequence of this path symmetry @xcite . according to eq.([eq : hje ] ) , the distance @xmath81 between two minima of the interference fringes ( @xmath82 ) is given by @xmath83 it is worth noting that the value of @xmath84 is larger than the maximal eigenvalue of @xmath7 , resulting in fringes that have a shorter period than fringes observed for superpositions of @xmath85 and @xmath86 , the states known as noon states @xcite .
the observation of shorter fringes is somewhat surprising because firstly , the noon states achieve the maximal possible phase sensitivity of @xmath20-photon states and secondly the fringes of photon number states can be written as a sum of fringes from superpositions of @xmath87 and @xmath88 , where the noon state contribution is the contribution with the shortest fringe periodicity .
however , the weak value of @xmath41 is not limited by the available range of eigenvalues , resulting in a fringe periodicity that corresponds to an additional one photon intensity difference between the paths .
specifically , the classical relation between @xmath55 and the eigenvalues @xmath57 and @xmath24 does not include any uncertainties , and the total length of the @xmath38-vector is given by @xmath89 because the `` @xmath90 '' contribution represents the necessary quantum fluctuations of an eigenstate with maximal @xmath7 eigenvalue .
the shorter fringes therefore represent the absence of quantum noise in the deterministic relation between input photon numbers and output photon numbers .
importantly , the shorter fringes are only obtained for this specific combination of input and output , which means that this effect is compensated by measurement results with less phase sensitivity and does not result in an enhanced overall phase sensitivity @xcite .
in fact , the experimental results reported in @xcite confirm the observation of fringes shorter than the noon state limit of @xmath28 .
it is interesting that the significance of this experimental result seems to have escaped notice , probably because there was no proper theoretical explanation available at the time .
( 500,480 ) ( 0,0)(480,480 ) [ 0.8 ] comparison of exact solution and approximate separation of the action @xmath1 and the envelope function @xmath2 for ( a ) eight photons and ( b ) sixteen photons .
the dotted lines show the results of the approximation , while the full line is the precise result obtained from the complete state vector in the nine and seventeen dimensional hilbert spaces , respectively .
the approximate amplitude @xmath91 only diverges from the exact solution close to @xmath35 and @xmath92 , where the quantum mechanical solution is limited to a probability of one , while the classical amplitude describes a probability density that diverges to infinity .
, title="fig : " ] we can now find an approximate expression for the amplitude of the @xmath93 in the case of equal photon numbers in the input and output .
there is only the problem of the integration constant for the action @xmath1 .
we can solve this problem by noting that @xmath94 corresponds to the action of a 50:50 beam splitter , where it is known that the probabilities of finding an odd number of photons in the output ports is zero for equal photon numbers in the input .
this means that , at @xmath94 , @xmath95 for odd @xmath96 and @xmath97 for even @xmath96 .
the phase dependence of @xmath1 is therefore given by @xmath98 note that this solution is obtained for phases between zero and @xmath99 , since the approximation that @xmath1 changes more rapidly than @xmath2 breaks down near @xmath35 and near @xmath92 .
to include the range between @xmath100 and @xmath35 , it is convenient to use the absolute value of @xmath15 , for an approximate solution of @xmath101 fig .
[ pairs ] shows the comparison between the exact solution and the approximation for eight photons and for sixteen photons .
as the figure shows , the approximation correctly describes the amplitudes @xmath93 except in the immediate vicinity of @xmath35 and @xmath102 , where the approximation diverges to values higher than one .
this discrepancy is easy to understand , since the classical approximation describes a probability density instead of a discrete probability , so it can exceed a value of one if the probability distribution in @xmath68 is narrower than @xmath49 .
significantly , the discreteness of the outcomes has no effect on the results outside of these two very narrow regions . even for eight photons , the effects of higher order terms in @xmath49 is mostly negligible .
( 500,480 ) ( 0,0)(480,480 ) [ 0.8 ] experimentally observable multi - photon interference fringes for equal photon numbers in the input and the output .
( a ) shows the phase dependent count rates for eight photons and ( b ) shows the phase dependent count rate for sixteen photons .
the dotted line shows the envelope function @xmath103 to indicate that the amplitude of the fringes follows the classical statistics of random phase interference .
, title="fig : " ] since the experimental count rates of a multi - photon interference experiment are given by the probabilities @xmath104 , these are shown in fig .
[ rates ] .
note that the envelope values indicated in the figure correspond to twice the classical probability densities expected for random phase interference with the same input and output conditions .
the coincidence of this envelop with the peaks of the interference fringes in fig .
[ rates ] thus illustrates how classical statistics emerge when the phase resolution fails to resolve the quantum interferences associated with the action @xmath1 . as mentioned above experimental data of this type
has already been reported , e.g. in @xcite for a six photon state , where the minima are separated by phases of @xmath105 .
in general , the positions of the minima are found at @xmath106 , which means that the minima for @xmath107 are located at @xmath108 for @xmath109 , there will be a total of @xmath20 minima , equal to the total number of photons .
however , the width of the peaks at @xmath35 and at @xmath102 is 1.5 times the width of the other peaks , resulting in @xmath110 interference fringes with a width of @xmath111 and two fringes with a width of @xmath112 .
( 500,240 ) ( 0,0)(480,240 ) [ 0.8 ] dependence of the intensity difference between the paths in the interferometer @xmath84 on phase @xmath15 for an input photon number difference of @xmath113 and an output photon number difference of @xmath114 .
the value is given relative to the total length of the @xmath38-vector , which is @xmath115 .
, title="fig : " ] ( 500,480 ) ( 0,0)(480,480 ) [ 0.8 ] experimentally observable multi - photon interference fringes for equal photon numbers in the input and a photon number difference of @xmath116 in the output .
( a ) shows the phase dependent count rates for eight photons and ( b ) shows the phase dependent count rate for sixteen photons .
the dotted line shows the envelope function @xmath103 to indicate that the amplitude of the fringes follows the classical statistics of random phase interference .
, title="fig : " ] next we consider cases where the photon number difference is non - zero in either the input or the output .
experimentally , it is relatively easy to obtain this kind of data from an @xmath117 input state by simply arranging the detectors to detect a non - zero photon number difference of @xmath78 between the output ports . from the output photon number difference @xmath78
, we can then estimate the intensity difference between the paths in the interferometer as @xmath118 is necessary to achieve an output photon number difference of @xmath19 by optical interference , this equation only has valid solutions for a limited range of phase shifts centered around @xmath94 . in all cases
, the magnitude of @xmath19 limits the maximal value of @xmath55 , and hence the periodicity of the multi - photon interference fringes .
[ j3cross ] shows the dependence of @xmath84 on phase for @xmath119 . in this case , the maximal value at @xmath94 is equal to @xmath120 times the maximal value of @xmath115 .
although it is possible to find an analytical solution for the action , the merit of the approximation is that it provides an insight into the physics that determines the periodicity of multi - photon interferences , and into the relation between the action and the envelope function that describes the classical statistics of random phase interference .
we will therefore analyze both aspects using the exact solutions of the interference fringes obtained for specific photon numbers .
[ pcross ] ( a ) shows the interference fringes for eight photons and @xmath121 in the output , and fig .
[ pcross ] ( b ) shows the fringes for sixteen photons and @xmath122 in the output .
the approximate envelop function given by eq.([eq : ergo ] ) is also indicated , showing that the approximation works well within the range of classical solutions . outside this range
, there is a non - zero evanescent solution that quickly drops to zero . within the range of classical solutions , we observe multi - photon interference effects in the form of fringes of width @xmath123 .
this width is easy to observe experimentally , since it can be defined as the phase difference between two consecutive minima of the output count rate .
we can use this definition to obtain an experimental value of @xmath84 for a specific fringe , @xmath124 in order to find out how well the approximation given by eq.([eq : j3 ] ) corresponds to the actual periodicities of multi - photon interference .
for the eight photon case shown in fig .
[ pcross ] ( a ) , probabilities of zero occur at @xmath125 and at @xmath126 , resulting in a single interference fringe with @xmath127 .
the corresponding value of @xmath84 is @xmath128 . according to eq .
( [ eq : m0 ] ) , the maximal possible value of @xmath84 for @xmath129 and @xmath130 should be @xmath131 , indicating that the approximation slightly underestimates the value of @xmath84 . for the sixteen photon case shown in fig .
[ pcross ] ( b ) , probabilities of zero occur at @xmath132 , @xmath133 , @xmath134 , and @xmath135 , resulting in three interference fringes with @xmath136 , @xmath137 , and @xmath136 .
the corresponding values of @xmath84 are @xmath138 , @xmath139 , and @xmath138 .
as expected from the theory , @xmath84 is maximal around @xmath94 , although the theory again underestimates the value , with @xmath140 at @xmath94 for @xmath141 and @xmath142 . as shown in fig.[j3cross ] , the value of @xmath84 drops off symmetrically as the phase is shifted from @xmath94 to higher or lower phases .
we can use eq.([eq : m0 ] ) to find the phase at which the expected value of @xmath84 is @xmath138 .
we obtain phases of @xmath143 and @xmath144 .
both of these phases are located near the center of the interference fringes from which the @xmath84 values are derived . in summary ,
the comparison between the approximate separation of quantum interference and ( classical ) envelope function can also be confirmed for different photon numbers in either the input or the output .
specifically , we can explain both the confinement of output probabilities to a finite phase interval around @xmath94 and the width of quantum interference fringes using the input - output relations given by eqs.([eq : j3 ] ) and eq.([eq : ergo ] ) .
remarkably , both of these relations also apply to classical random phase interference , although classical statistics would result in an output probability given by @xmath145 , without the quantum interference effects associated with the value of @xmath84 .
( 500,240 ) ( 0,0)(480,240 ) [ 0.8 ] dependence of the intensity difference between the paths in the interferometer @xmath84 on phase @xmath15 for an input photon number difference of @xmath146 and an output photon number difference of @xmath114 .
the value is given relative to the total length of the @xmath38-vector , which is @xmath115 .
, title="fig : " ] ( 500,480 ) ( 0,0)(480,480 ) [ 0.8 ] experimentally observable multi - photon interference fringes for photon number differences of @xmath147 in both the input and the output .
( a ) shows the phase dependent count rates for eight photons and ( b ) shows the phase dependent count rate for sixteen photons .
the dotted line shows the envelope function @xmath103 to indicate that the amplitude of the fringes follows the classical statistics of random phase interference .
, title="fig : " ] finally , it may also be good to consider the case where both the input and the output have a non - vanishing photon number difference between the ports . to keep the mathematics simple ,
we choose the case of @xmath148 , where the output photon number distribution happens to be identical to the input photon number distribution .
the intensity difference @xmath84 between the paths is then given by latexmath:[\[\label{eq : mm } note that the maximal value of @xmath84 is the same as the maximal value in eq.([eq : m0 ] ) , but it occurs at @xmath35 instead of at @xmath94 .
the value of @xmath84 drops continuously as @xmath150 increases , until it reaches zero at @xmath151 .
[ j3self ] shows the dependence of @xmath84 on phase for @xmath152 , where the value of @xmath84 reaches zero at @xmath153 .
[ pself ] ( a ) shows the interference fringes for eight photons and @xmath121 in both input and output , and [ pself ] ( b ) shows the fringes for sixteen photons and @xmath122 in input and output . the approximate envelope function is also shown , indicating that eq.([eq : ergo ] ) is a good approximation of the precise results up to the limit of validity near @xmath153 .
we can now evaluate the width of the fringes and obtain the corresponding values of @xmath84 according to eq.([eq : evaluate ] ) .
for the eight photon case shown in fig .
[ pself ] ( a ) , probabilities of zero occur at @xmath154 and at @xmath155 , resulting in a single interference fringe with @xmath156 .
the corresponding value of @xmath84 is @xmath157 . according to eq.([eq : mm ] ) , this value of @xmath84 is obtained at a phase of @xmath158 .
although this phase is found within the fringe , it is much closer to the low phase minimum than to the high phase minimum , indicating that the approximation slightly underestimates the @xmath84 value . in the middle of the fringe , at @xmath159 , eq.([eq : mm ] ) assigns a @xmath84 value of only @xmath160 .
this is consistent with the result from fig.([pcross ] ) ( a ) above , suggesting an error margin of about two percent for eight photon fringe widths obtained from eq.([eq : j3 ] ) . for the sixteen photon case shown in fig .
[ pself ] ( b ) , probabilities of zero occur at @xmath161 , @xmath162 , @xmath163 , and @xmath164 , resulting in three interference fringes with @xmath165 , @xmath166 , and @xmath167 .
the corresponding values of @xmath84 are @xmath168 , @xmath169 , and @xmath170 .
as expected , the values of @xmath84 decrease as the phase shift increases .
the @xmath84 value of the first fringe exceeds the maximal value of @xmath140 obtained from eq.([eq : mm ] ) , indicating once more that the theory underestimates the values of @xmath84 .
for the other two fringes , the values of @xmath84 obtained for their respective widths are found at @xmath171 for @xmath172 , and at @xmath173 for @xmath174 .
both results are very close to the centers of the fringes , confirming that eq.([eq : hje ] ) provides a good approximate description of multi - photon interferences .
the results obtained for all examples show that the physics of multi - photon interference can be understood in terms of the separation between the action @xmath1 that describes quantum interferences and the amplitude @xmath2 that describes the classical statistics of random phase interference . in particular
, it is possible to explain the periodicity of multi - photon interference in terms of the intensity difference between the paths expected from the _ combination _ of input and output conditions .
it is therefore important to recognize that multi - photon interference is not just the result of superpositions in the input state , but also involves quantum coherences in the multi - photon statistics of the measurement outcome . in the case of photon number states in the input ,
the relation is completely symmetric , and the same phase dependent interference fringes will be obtained when input and output are exchanged .
in the analysis presented above , we have studied the multi - photon interference fringes of photon number state inputs in the two input ports of a two - path interferometer . however , the approach developed at the start of sec .
[ sec : wv ] is not limited to these specific states . specifically , eqs.([eq : action ] ) and ( [ eq : quadact ] ) are generally valid expressions of the phase evolution of interference fringes in terms of the weak values of @xmath7 and @xmath41 .
the separation of amplitude @xmath2 and action @xmath1 introduced in eq.([eq : separate ] ) is possible whenever the amplitudes have only real values .
this means that the real part of the weak values of @xmath7 must be zero for all phases @xmath15 , @xmath175 since the phases of different eigenstate components of @xmath7 evolve differently , this criterion can not be satisfied unless each pair of components with opposite eigenvalues have the same amplitude and opposite phases .
if the eigenstates of @xmath7 with eigenvalues of @xmath176 are written as @xmath177 , this criterion can be expressed by the condition @xmath178 usually , this condition is satisfied separately by both the input state and the output state , so that both states are symmetric in the path basis .
specifically , the eigenstates of @xmath18 that represent the results of photon number measurements in the output ports all satisfy the condition @xmath179 therefore , the amplitudes @xmath93 will be real at all phases @xmath15 if the @xmath7 components of @xmath9 satisfy @xmath180 note that it has been shown elsewhere that input states satisfying this path symmetry condition achieve their maximal phase sensitivity in photon number measurements @xcite .
the present analysis investigates the more detailed mechanism by which this is achieved . in general , the approximate solution of eq.([eq : quadact ] )
is given by eq.([eq : second ] ) , so that the multi - photon interference fringes can be described by the phase dependent action @xmath1 with @xmath181 it is therefore possible to derive the periodicity of the interference fringes from the weak value of the squared phase shift generator @xmath41 for any path symmetric state .
note that eq.([eq : quadact ] ) indicates that the weak value of @xmath41 is always real if the states and measurements satisfy eq.([eq : cond1 ] ) at all phases . in the limit of high photon number @xmath20 ,
it is usually possible to find a classical relation that defines the value of @xmath55 as a function of initial conditions @xmath182 , measurement outcome @xmath19 , and phase shift @xmath15 .
it is then possible to relate the multi - photon interference effects to the deterministic relations between initial and final conditions known from classical wave interference . in the present case ,
the action @xmath1 is directly observable as an interference fringe because the amplitudes @xmath93 are all real .
this is a result of the symmetry between negative and positive intensity differences between the paths . if there is only a single classical solution for @xmath55 , the action given by eq.([eq : hje ] ) will merely describe an unobservable phase evolution of @xmath93 . in this sense
, one can explain quantum interference as an interference between the two classical solutions @xmath183 and @xmath184 , both of which connect @xmath182 to @xmath19 under phase shifts of @xmath15 .
importantly , this interference can be described without any reference to hilbert space vectors , since the physical properties @xmath185 can all be defined within classical wave theory .
it may therefore be useful to take a closer look at the relation between classical wave interference and the quantum interference of multi - photon states .
as shown above , multi - photon interference can be understood in terms of the weak values of the generator @xmath7 that describes the phase shift induced by a two - path interferometer . to explain the connection with more conventional ideas of quantum interference as an effect of superpositions in the initial state
, it may be useful to consider a particularly simple solution of eq.([eq : act2 ] ) , where the periodicity of the interference fringes is determined by the input state and does not depend on output photon number or phase .
this situation only occurs if the input state is an eigenstate of @xmath41 , so that the weak value is given by the initial eigenvalue and does not depend on phase anymore . except for arbitrary phase shifts , the complete set of states with constant fringe periodicity
is therefore given by @xmath186 where @xmath177 are the @xmath7 eigenstates with eigenvalues of @xmath176 .
similar to the case of equal photon numbers in input and output , the derivative of the action is constant and a multiple of @xmath49 , @xmath187 for an @xmath20-photon state , the maximal number of fringes is obtained for the so - called noon state , where @xmath85 .
it is therefore tempting to think of multi - photon interference as an @xmath20-fold increase in the number of fringes , or an @xmath20-fold enhancement of phase sensitivity .
however , even the simple superposition states given by eq.([eq : super ] ) show that the number and the periodicity of fringes is not a function of photon number , but a function of photon number distribution between the paths .
in general , the weak value of @xmath41 provides an estimate of the intensity distribution between the paths for a specific combination of input state , measurement outcome , and phase shift .
this estimate identifies the proper physics of multi - photon interference .
what is perhaps most remarkable about the results presented above is that the quantum effects of multi - photon interference can be explained by using the deterministic relations between field coherences already known from classical two - path interference .
the classical interference effect is actually described by the operator relation in eq.([eq : heis ] ) , where the output intensities are expressed as a function of input intensities , coherence , and phase shift . in classical theory ,
probabilities enter the picture only if there is some randomness in the initial conditions . in the case of photon number inputs discussed above
, this randomness can be identified as a random phase difference between the two input fields .
as the approximation shows , the quantum mechanical result reproduces this classical probability distribution over random phases in the form of the envelope function that describes the amplitude of the quantum interference fringes .
importantly , quantum interference is not an effect that emerges gradually from classical field interference as photon number increases , but should be considered as a fundamentally different effect associated with the non - classical relation between the phase shift @xmath15 and its generator @xmath7 .
the only reason why this effect is related to photon number is that , as a quantum effect , its scale is determined by @xmath49 .
all classical effects scale with intensity , only the quantum interference fringes decrease in width as photon number increases .
the analysis in this paper shows that it is possible to separate the quantum scale and the classical scale in the quantum theory of multi - photon interference fringes , highlighting the different physics involved in the two distinct processes .
in conclusion , multi - photon interferences can be explained in terms of the weak values of the intensity difference between the paths of the interferometer . for photon number states ,
it is then possible to describe each multi - photon interference fringe by its own differential equation .
the approximate solution of this differential equation identifies the periodicity of the fringes with the intensity difference between the paths inside the interferometer , which can be determined from the classical relation between input intensity difference , output intensity difference , and phase .
we have shown that it is possible to separate the quantum effects of multi - photon interference from the classical statistics of random phase interference .
specifically , we can trace quantum interference back to the action of the phase shift , which is given by the product of the generator @xmath7 and the phase shift @xmath15 .
the photon number appears in this relation only because @xmath7 is quantized in units of @xmath49 .
the classical units of intensity difference are action units , where @xmath49 is the ratio between the action and the quantum phase .
our analysis thus shows that an detailed analysis of multi - photon interference can be achieved by focusing on the classical field properties if both the input state and the measurement outcomes are treated on an equal footing .
our work provides a general characterization of multi - photon interference that may serve as a road map to future experiments involving larger photon numbers .
we have identified the relevant features of different two - path interference scenarios and explained the origin and the physical meaning of the experimentally observable fringe width .
we believe that these observations will be extremely helpful in the characterization of multi - photon quantum devices and in the development of new approaches to multi - particle quantum physics .
this work was supported by jsps kakenhi grant number 26220712 . | multi - photon interference results in modulations of output probabilities with phase shift periods that are much shorter than @xmath0 . here , we investigate the physics behind these statistical patterns in the case of well - defined photon numbers in the input and output modes of a two - path interferometer .
we show that the periodicity of the multi - photon interference is related to the weak value of the unobserved intensity difference between the two arms of the interferometer .
this means that the operator relations between the photon number differences in input , path , and output can be used to determine the periodicity of the experimentally observed quantum interference , establishing an important link between the classical causality of random phase interference and quantum effects that depend on the superposition of classically distinct possibilities . |
the study of nonleptonic two - body decays of @xmath0 mesons is of primary importance for the exploration of cp violation and the determination of the flavour parameters of the standard model . because of the interference of several competing amplitudes , these processes allow for the presence of different weak and strong - interaction phases , which play a crucial role for cp violation . in the standard model ,
all cp - violating observables are related to the complex phase of the quark mixing matrix , which in turn implies nontrivial angles in the `` unitarity triangle '' @xmath10 . with the standard choice of phase conventions
, one defines the weak phases @xmath11 and @xmath12 , as well as @xmath13 . in the standard model
, @xmath14 can be extracted in a theoretically clean way by measuring the time - dependent rates for the decays @xmath15 .
the measurement of @xmath2 ( or @xmath3 ) is more difficult , since it requires controlling the hadronic dynamics in nonleptonic @xmath0 decays .
a promising strategy for the determination of @xmath2 is based on rate measurements for the charged modes @xmath16 and @xmath17 @xcite .
hadronic uncertainties in this method can be reduced to a minimum by exploiting the structure of the effective weak hamiltonian and using isospin and su(3 ) flavour symmetries @xcite .
if only measurements of cp - averaged branching ratios are available , it is still possible to derive bounds on @xmath2 @xcite , or to determine @xmath2 under the assumption of only a moderate strong - interaction phase @xmath18 between penguin and tree amplitudes .
a different strategy for extracting @xmath2 uses su(3)-symmetry relations between the various contributions to the time - dependent @xmath19 and @xmath20 decay amplitudes @xcite .
the main theoretical limitation of these methods is in the accuracy with which the effects of su(3 ) breaking can be estimated .
the angle @xmath3 can be determined from the time - dependent cp asymmetry in the decays @xmath21 , if the subdominant penguin contribution to the decay amplitudes can be subtracted in some way .
this can be achieved by an isospin analysis @xcite ; however , in practice that route is extremely challenging due to the difficulty in measuring the very small @xmath22 branching ratios .
one must therefore rely on dynamical input for the penguin - to - tree ratio @xcite .
alternatively , @xmath3 can be measured in related decays such as @xmath23 , for which the penguin contribution can be eliminated using a time - dependent analysis of the @xmath24 dalitz plot @xcite .
most of the above - mentioned determinations of @xmath2 or @xmath3 have theoretical limitations , which would be reduced if some degree of theoretical control over two - body nonleptonic @xmath0 decays could be attained . however , in the past this has proven to be a very difficult problem .
even advanced methods such as lattice gauge theory , qcd sum rules , or the large-@xmath25 expansion have little to say about the qcd dynamics relevant to hadronic @xmath0 decays . in recent work ,
we have developed a systematic approach to this problem .
it is based on the observation that , in the heavy - quark limit @xmath26 , a rigorous qcd factorization formula holds for the two - body decays @xmath27 , if the `` emission particle '' @xmath28 ( the meson not obtaining the spectator quark from the @xmath0 meson ) is a light meson .
( it has been argued that perhaps the large-@xmath25 limit may be more relevant to factorization than the heavy - quark limit @xcite .
however , the dramatic decrease of `` nonfactorizable '' effects seen when comparing @xmath29 , @xmath30 and @xmath31 decays shows that the heavy - quark limit is of crucial importance . ) we have previously applied the factorization formula to @xmath32 decays and obtained results for the decay amplitudes at next - to - leading order in @xmath33 and to leading power in @xmath34 @xcite .
the conceptual foundations of our approach have been discussed in detail in @xcite , which focused on the simpler case of decays into heavy light final states ( where @xmath35 is a charm meson ) . in the present work , the qcd factorization formula is applied to the general case of @xmath0 decays into a pair of light , flavour - nonsinglet pseudoscalar mesons .
( preliminary results of this analysis have been presented in @xcite . )
the present analysis contains three new theoretical ingredients in addition to a much more detailed phenomenological analysis : [ [ section ] ] 1 .
+ + matrix elements of electroweak penguin operators are included , which play an important role in charmless decays based on @xmath36 transitions .
this is a straightforward extension of our previous analysis .
however , a sensible implementation of qcd corrections to electroweak penguin matrix elements implies that one departs from the usual renormalization - group counting , in which the initial conditions for the electroweak penguin operators at the scale @xmath37 are treated as a next - to - leading order effect .
[ [ section-1 ] ] 2 .
+ + hard - scattering kernels are derived for general , asymmetric meson light - cone distribution amplitudes .
this is important for addressing the question of nonfactorizable su(3)-breaking corrections , since the distribution amplitudes of strange mesons are , in general , not symmetric with respect to the quark and antiquark momenta .
[ [ section-2 ] ] 3 .
+ + the leading power corrections to the heavy - quark limit are estimated by analyzing `` chirally - enhanced '' power corrections related to certain twist-3 distribution amplitudes for pseudoscalar mesons .
we also discuss potentially large power corrections arising from annihilation topologies , noted first in @xcite .
this is essential for controlling the theoretical uncertainties of our approach .
the second and third items have not been considered in previous generalizations of the results of @xcite to the case of @xmath31 decays @xcite .
chirally - enhanced power corrections were discussed in @xcite , but we disagree with some of the results obtained by these authors .
the qcd factorization approach provides us with model - independent predictions for the decays amplitudes including , in particular , their strong - interaction phases .
the same method can also be applied to other charmless decays , such as vector
pseudoscalar @xcite or vector vector modes .
our main focus here is on the development of the new conceptual aspects of the approach that are important for a comprehensive phenomenological analysis .
this includes a detailed discussion of various sources of potentially large power corrections to the heavy - quark limit .
we then perform a comprehensive study of cp - averaged branching fractions and direct cp asymmetries in decays to @xmath38 and @xmath39 final states , including a detailed discussion of the theoretical uncertainties from all inputs to the qcd factorization approach . in many of the phenomenological applications discussed in this work the dynamical information obtained using the qcd factorization formalism
is used in a `` minimal way '' , to reduce the hadronic uncertainties in methods that are theoretically clean up to `` nonfactorizable '' su(3)-breaking effects .
most importantly , these strategies do not suffer from uncertainties related to weak annihilation contributions . in this way , it is possible to reduce the hadronic uncertainties in the strategies for determining @xmath2 from @xmath40 decays proposed in @xcite to the level of `` nonfactorizable '' corrections that simultaneously violate su(3 ) symmetry and are power suppressed in the heavy - quark limit .
these corrections are parametrically suppressed by the product of three small quantities : @xmath41 , @xmath42 , and @xmath34 . as a consequence ,
we argue it will eventually be possible to determine @xmath2 with a theoretical accuracy of about @xmath43 ( unless @xmath2 is much different from its expected value in the standard model ) .
more accurately , the strategies discussed here can constrain the wolfenstein parameters @xmath44 and @xmath45 with accuracies similar those obtained from the standard global fit to @xmath46 , @xmath47 , and @xmath0@xmath48 mixing .
given the theoretical input discussed in this paper even the present , preliminary data on the rare hadronic decays exclude at 95% confidence level half of the parameter space obtained from the standard fit . the qcd factorization approach provides a complete theoretical description of all @xmath49 decay amplitudes .
this allows for a large variety of predictions , which go far beyond those explored in the present work . in the future
, this will offer the possibility of several nontrivial experimental tests of the factorization formula .
a more detailed discussion of these predictions will be presented elsewhere .
the remainder of this paper is organized as follows : in section [ sec:2 ] , we collect some basic formulae and express the @xmath1 decay amplitudes in terms of parameters @xmath50 and @xmath51 appearing in the effective , factorized transition operators for these decays ( including weak annihilation contributions ) . section [ sec:3 ] contains the technical details of the calculations based on the factorization formula , a discussion of annihilation effects , and a compilation of the relevant formulae for the numerical evaluation of our results .
readers not interested in the technical aspects of our work can proceed directly to sections [ sec : numerics ] and [ sec:4 ] , where we present numerical values for the amplitude parameters @xmath50 and @xmath51 ( section [ sec : numerics ] ) and discuss phenomenological applications of our results ( section [ sec:4 ] ) . specifically
, we consider strategies to bound and determine the weak phase @xmath2 and to extract @xmath52 from mixing - induced cp violation in @xmath53 decay .
we also present predictions for cp - averaged branching fractions and cp asymmetries , and perform a global fit in the @xmath4 plane to all measured @xmath1 branching fractions
. a critical comparison of our formalism with other theoretical approaches to hadronic @xmath0 decays is performed in section [ sec : comp ] .
the effective weak hamiltonian for charmless hadronic @xmath0 decays consists of a sum of local operators @xmath54 multiplied by short - distance coefficients @xmath55 and products of elements of the quark mixing matrix , @xmath56 or @xmath57 .
below we will focus on @xmath31 decays to be specific ; however , with obvious substitutions a similar discussion holds for all other @xmath0 decays into two light , flavour - nonsinglet pseudoscalar mesons . using the unitarity relation @xmath58
, we write @xmath59 where @xmath60 are the left - handed current current operators arising from @xmath61-boson exchange , @xmath62 and @xmath63 are qcd and electroweak penguin operators , and @xmath64 and @xmath65 are the electromagnetic and chromomagnetic dipole operators .
they are given by @xmath66 where @xmath67 , @xmath68 are colour indices , @xmath69 are the electric charges of the quarks in units of @xmath70 , and a summation over @xmath71 is implied .
( the definition of the dipole operators @xmath64 and @xmath65 corresponds to the sign convention @xmath72 for the gauge - covariant derivative . ) the wilson coefficients are calculated at a high scale @xmath73 and evolved down to a characteristic scale @xmath74 using next - to - leading order renormalization - group equations .
the essential problem obstructing the calculation of nonleptonic decay amplitudes resides in the evaluation of the hadronic matrix elements of the local operators contained in the effective hamiltonian .
applying the qcd factorization formula and neglecting power - suppressed effects , the matrix elements of the effective weak hamiltonian can be written in the form @xcite @xmath75 where @xmath76 here @xmath77 , and a summation over @xmath78 is implied .
the symbol @xmath79 indicates that the matrix elements of the operators in @xmath80 are to be evaluated in the factorized form @xmath81 or @xmath82 , as appropriate .
`` nonfactorizable '' corrections are , by definition , included in the coefficients @xmath50 .
the matrix elements for @xmath0 mesons ( i.e. , mesons containing a @xmath83-antiquark ) are obtained from ( [ top ] ) by cp conjugation .
a corresponding result , with obvious substitutions , holds for other decays such as @xmath32 .
the term @xmath84 in ( [ top ] ) arises from weak annihilation contributions and introduces a set of coefficients @xmath85 , which we shall define and discuss in detail in section [ subsec : annihilation ] .
annihilation contributions are suppressed by a power of @xmath34 and not calculable within the qcd factorization approach .
nevertheless , we will include the coefficients @xmath51 in the amplitude parameterizations in this section .
the coefficients @xmath50 multiplying products of vector or axial - vector currents are renormalization - scheme invariant , as are the hadronic matrix elements of these currents . for the coefficients @xmath86 and @xmath87 a scheme dependence remains , which exactly compensates the scheme dependence of the hadronic matrix elements of the scalar or pseudoscalar densities associated with these coefficients .
these matrix elements are power suppressed by the ratio @xmath88 which is formally of order @xmath34 but numerically close to unity . in the following
we shall use the same notation for charged ( @xmath89 ) and neutral kaons ( @xmath90 ) , since the difference is tiny .
a corresponding ratio @xmath91 appears in the discussion of the @xmath32 decay amplitudes . for a phenomenological analysis of nonleptonic
@xmath0 decays it is necessary to estimate these `` chirally - enhanced '' corrections despite the fact that they are formally power suppressed .
a detailed discussion of these corrections will be presented in section [ subsec : power ] , and their numerical importance will be investigated in section [ sec : numerics ] . in terms of the parameters
@xmath50 , the @xmath31 decay amplitudes ( without annihilation contributions ) are expressed as @xmath92 a_{\pi k } \
, , \nonumber\\ - \sqrt2\,{\cal a}(b^-\to\pi^0 k^- ) & = & \left [ \lambda_u\,a_1 + \lambda_p\,(a_4^p + a_{10}^p ) + \lambda_p\,r_\chi^k\,(a_6^p + a_8^p ) \right ] a_{\pi k } \nonumber\\ & & \mbox{}+ \left [ \lambda_u\,a_2 + \lambda_p \,\frac32\,(- a_7 + a_9 ) \right ] a_{k\pi } \ , , \nonumber\\ - { \cal a}(\bar b^0\to\pi^+ k^- ) & = & \left [ \lambda_u\,a_1 + \lambda_p\,(a_4^p + a_{10}^p ) + \lambda_p\,r_\chi^k\,(a_6^p + a_8^p ) \right ] a_{\pi k } \ , , \nonumber\\ \sqrt2\,{\cal a}(\bar{b}^0\to\pi^0 \bar{k}^0 ) & = & { \cal a}(b^-\to\pi^-\bar k^0 ) + \sqrt2\,{\cal a}(b^-\to\pi^0 k^- ) \nonumber\\ & & \mbox{}- { \cal a}(\bar b^0\to\pi^+ k^- ) \,.\end{aligned}\ ] ] here @xmath56 , @xmath93 , and a summation over @xmath94 is implicitly understood in expressions like @xmath95 .
the last relation follows from isospin symmetry .
the cp - conjugate decay amplitudes are obtained from the above by replacing @xmath96 .
we have defined the factorized matrix elements @xmath97 where @xmath98 are semileptonic form factors .
weak annihilation effects contribute further terms to the decay amplitudes , which can be parameterized as @xmath99 b_{\pi k } \ , , \nonumber\\ - \sqrt2\,{\cal a}_{\rm ann}(b^-\to\pi^0 k^- ) & = & { \cal a}_{\rm ann}(b^-\to\pi^-\bar k^0 ) \ , ,
\nonumber\\ - { \cal a}_{\rm ann}(\bar b^0\to\pi^+ k^- ) & = & ( \lambda_u+\lambda_c)\left(b_3-\frac{1}{2}\ , b_3^{\rm ew } \right ) b_{\pi k } \
, , \nonumber\\ \sqrt2\,{\cal a}_{\rm ann}(\bar{b}^0\to\pi^0 \bar{k}^0 ) & = & \mbox{}- { \cal a}_{\rm ann}(\bar b^0\to\pi^+ k^- ) \,,\end{aligned}\ ] ] where @xmath100 the coefficients @xmath101 will be defined in section [ subsec : annihilation ] , but we may note here that @xmath102 are related to the current current operators @xmath103 and @xmath104 in the effective hamiltonian ( [ heff ] ) , and @xmath105 ( @xmath106 ) are related to qcd ( electroweak ) penguin operators .
the @xmath32 decay amplitudes are given by @xmath107 a_{\pi\pi } \ , , \nonumber\\ - \sqrt2\,{\cal a}(b^-\to\pi^-\pi^0 ) & = & \left [ \lambda_u ' ( a_1 + a_2 ) + \frac32\,\lambda_p ' ( -a_7 + r_\chi^\pi\,a_8^p + a_9 + a_{10}^p ) \right ] a_{\pi\pi } \ , , \nonumber\\[0.15 cm ] { \cal a}(\bar{b}^0\to\pi^0\pi^0 ) & = & \sqrt2\,{\cal a}(b^-\to\pi^-\pi^0 ) - { \cal a}(\bar b^0\to\pi^+\pi^- ) \,,\end{aligned}\ ] ] where now @xmath57 , @xmath108 , and @xmath109 the additional annihilation contributions are @xmath110 b_{\pi \pi } \
\nonumber\\ - \sqrt2\,{\cal a}_{\rm ann}(b^-\to\pi^-\pi^0 ) & = & 0 \ , , \nonumber\\[0.2 cm ] { \cal a}_{\rm ann}(\bar{b}^0\to\pi^0\pi^0 ) & = & - { \cal a}_{\rm ann}(\bar b^0\to\pi^+\pi^- ) \,,\end{aligned}\ ] ] where @xmath111 , and @xmath112 neglecting tiny mass corrections of order @xmath113 , @xmath114 these ratios will play an important role in the discussion of su(3 ) violations in section [ sec:4 ] .
the expressions collected above provide a complete description of the decay amplitudes in terms of the parameters @xmath50 and @xmath51 for the various processes .
they are the basis for most of the phenomenological applications discussed in this work .
however , in the literature several alternative parameterizations of the @xmath31 decay amplitudes have been introduced , which are sometimes useful when considering cp asymmetries or ratios of branching fractions .
we briefly elaborate on one such parameterization here , adopting the notations of @xcite .
the dominant contributions to the @xmath31 decay amplitudes come from qcd penguin operators . because the corresponding operators in the effective weak hamiltonian preserve isospin , this contribution is the same ( up to trivial clebsch
gordon coefficients ) for all decay modes .
isospin - violating contributions to the decay amplitudes are subdominant and arise from the current current operators @xmath115 and @xmath116 ( so - called `` tree '' contributions ) , and from electroweak penguins .
the latter are suppressed by a power of @xmath117 , whereas the former are suppressed by the ratio @xmath118 where @xmath119 is the cabibbo angle , @xmath120 is one of the sides of the unitarity triangle , and @xmath44 and @xmath45 are the wolfenstein parameters .
a general parameterization of the decay amplitudes is @xmath121 \ , , \nonumber\\ - { \cal a}(\bar b^0\to\pi^+ k^- ) & = & p\,\big [ 1 + \varepsilon_a\,e^{i\phi_a } e^{-i\gamma } - \varepsilon_t\,e^{i\phi_t } ( e^{-i\gamma } - q_c\,e^{i\omega_c } ) \big ] \,.\end{aligned}\ ] ] ( the phase @xmath122 was denoted @xmath123 in @xcite . )
the amplitude @xmath124 is then determined by the isospin relation shown in the last line of ( [ bkpi ] ) . the dominant penguin amplitude @xmath125 is defined as the sum of all contributions to the @xmath126 amplitude that are not proportional to @xmath127 .
this quantity cancels whenever one takes ratios of decay amplitudes , such as cp asymmetries or ratios of branching fractions .
the parameters @xmath128 and @xmath129 measure the relative strength of tree and qcd penguin contributions , @xmath130 and @xmath131 measure the relative strength of electroweak penguin and tree contributions , and @xmath132 parameterizes a rescattering contribution to the @xmath126 amplitude arising from up - quark penguin topologies .
moreover , @xmath18 , @xmath133 , @xmath134 , @xmath135 , and @xmath136 are strong rescattering phases .
( the strong - interaction phase of @xmath125 is not observable and can be set to zero . )
all parameters except @xmath137 receive weak annihilation contributions .
it is straightforward to express the various amplitude parameters in terms of the parameters @xmath93 and @xmath101 defined earlier .
we obtain @xmath138 a_{\pi k } + \left ( b_3 + b_3^{\rm ew } \right ) b_{\pi k } \right\}\ ] ] for the leading penguin amplitude , and for the remaining parameters @xmath139 } { ( a_4^c+r_\chi^k a_6^c ) - \frac{1}{2}(a_{10}^c+r_\chi^k a_8^c ) + r_a ( b_3+b_3^{\rm ew } ) } \ , , \nonumber\\ \varepsilon_t\,e^{i\phi_t } & = & - \epsilon_{\rm km}\ , \frac{a_1 + \frac32(a_{10}^u+r_\chi^k a_8^u ) - r_a ( b_2+\frac32\,b_3^{\rm ew } ) } { ( a_4^c+r_\chi^k a_6^c ) - \frac{1}{2}(a_{10}^c+r_\chi^k a_8^c ) + r_a ( b_3+b_3^{\rm ew } ) } \ , , \nonumber\\ \varepsilon_a\,e^{i\phi_a } & = & \epsilon_{\rm km}\ , \frac{(a_4^u+r_\chi^k a_6^u ) - \frac{1}{2}(a_{10}^u+r_\chi^k a_8^u ) + r_a ( b_2+b_3+b_3^{\rm ew } ) } { ( a_4^c+r_\chi^k a_6^c ) - \frac{1}{2}(a_{10}^c+r_\chi^k a_8^c ) + r_a ( b_3+b_3^{\rm ew } ) } \ , , \nonumber\\ q\,e^{i\omega } & = & - \frac{3}{2\epsilon_{\rm km}}\ , \frac{a_{10}^c+r_\chi^k a_8^c + r_{\pi k}(a_9-a_7 ) } { ( a_1+r_{\pi k } a_2 ) + \frac32[a_{10}^u+r_\chi^k a_8^u + r_{\pi k}(a_9-a_7 ) ] } \ , , \nonumber\\ q_c\,e^{i\omega_c } & = & - \frac{3}{2\epsilon_{\rm km}}\ , \frac{a_{10}^c + r_\chi^k a_8^c -r_a b_3^{\rm ew } } { a_1 + \frac32(a_{10}^u+r_\chi^k a_8^u ) - r_a ( b_2+\frac32\,b_3^{\rm ew } ) } \,,\end{aligned}\ ] ] where @xmath140 is the ratio of the two factorized amplitudes given in ( [ ratios ] ) , and @xmath141 our notation for amplitude ratios is such that the ratio @xmath142 ( denoted by a capital @xmath143 ) deviates from 1 only by su(3)-breaking corrections , whereas the ratios @xmath144 and @xmath145 ( denoted by a lower - case @xmath146 ) are formally of order @xmath34 in the heavy - quark limit .
however , whereas @xmath147 is indeed very small , @xmath148 ( at a scale @xmath149gev ) is numerically large for real @xmath0 mesons .
finally , note that the electroweak penguin coefficients @xmath150 could be safely neglected in all quantities other than @xmath137 and @xmath151 , because they are tiny compared with the other coefficients @xmath152 . in ( [ params ] ) , they are included only for completeness .
( the systematics of including electroweak penguin contributions will be discussed in more detail later . )
an important quantity affecting the determination of @xmath52 from the time - dependent cp asymmetry in the decays @xmath21 is the ratio of penguin to tree amplitudes . in this case
the tree contribution is no longer ckm suppressed , since @xmath153 is of order unity . the @xmath154 decay amplitude in ( [ bpipi ] )
can be written as @xmath155 where @xmath156 } { ( a_1+a_4^u+r_\chi^\pi a_6^u ) + ( a_{10}^u+r_\chi^\pi a_8^u ) + r_a [ b_1+b_3 + 2b_4-\frac12(b_3^{\rm ew}-b_4^{\rm ew } ) ] } \,,\ ] ] and @xmath157 . in this case the electroweak penguin terms are very small , because they are not ckm enhanced with respect to the tree contribution .
this concludes the discussion of parameterizations of the decay amplitudes .
the following section is devoted to a detailed description of the qcd factorization formalism , the calculation of the parameters @xmath50 for the various nonleptonic decay amplitudes , and an estimation of the annihilation parameters @xmath51 .
the reader mainly interested in the phenomenological applications of our results can proceed directly to section [ sec : numerics ] , where we present numerical results for these parameters , which will be used in the subsequent analysis in section [ sec:4 ] .
based on the underlying physical principle of colour transparency ( see @xcite for early discussions in the context of decays to heavy
light final states ) , supported by a detailed diagrammatic analysis of infrared cancellations at leading power in the heavy - quark expansion , we have shown in previous work that the complexity of the hadronic matrix elements governing energetic , two - body hadronic decays of @xmath0 mesons simplifies greatly in the heavy - quark limit @xmath26 @xcite .
consider @xmath31 decays as an example .
to leading power in @xmath34 , but to all orders in perturbation theory , the matrix elements of the local operators @xmath54 in the effective weak hamiltonian in ( [ heff ] ) obey the factorization formula @xmath158 where @xmath159 are leading - twist light - cone distribution amplitudes , and the @xmath160-products imply an integration over the light - cone momentum fractions of the constituent quarks inside the mesons .
a graphical representation of this result is shown in figure [ fig1 ] . because the energetic , collinear light - quark pair that ultimately evolves into the emission particle at the `` upper vertex '' is created by a point - like source , soft gluon exchange between this pair and the other quarks in the decay is power suppressed in the heavy - quark limit ( colour transparency ) . in other words , whereas the hadronic physics governing the semileptonic @xmath161 transition and the formation of the emission particle @xmath28 is genuinely nonperturbative , `` nonfactorizable '' interactions connecting the two systems are dominated by hard gluon exchange .
the hard - scattering kernels @xmath162 in ( [ fact ] ) are calculable in perturbation theory .
@xmath163 starts at tree level and , at higher order in @xmath33 , contains `` nonfactorizable '' corrections from hard gluon exchange or light - quark loops ( penguin topologies ) .
hard , `` nonfactorizable '' interactions involving the spectator quark are part of @xmath164 . the relevant feynman diagrams contributing to these kernels at next - to - leading are shown in figure [ fig : graphs ] . although individually these graphs contain infrared - sensitive regions at leading power , all soft and collinear divergences cancel in their sum , thus yielding a calculable short - distance contribution .
annihilation topologies are not included in ( [ fact ] ) and figure [ fig : graphs ] , because they do not contribute at leading order in @xmath34 .
these power - suppressed contributions will be discussed separately in section [ subsec : annihilation ] .
we stress that the factorization formula does not imply that hadronic @xmath0 decays are perturbative in nature .
dominant soft contributions to the decay amplitudes exist , which can not be controlled in a hard - scattering approach .
however , at leading power all these nonperturbative effects are contained in the semileptonic form factors and light - cone distribution amplitudes .
once these quantities are given , the nonleptonic decay amplitudes can be derived using perturbative approximations to the hard - scattering kernels .
this allows us to compute perturbative corrections to `` naive factorization '' estimates of nonleptonic amplitudes , which is crucial for obtaining results that are independent of the renormalization scheme adopted in the calculation of the effective weak hamiltonian .
the hard - scattering kernels also contain imaginary parts , which determine the strong rescattering phases of the decay amplitudes . at leading power in @xmath34
these imaginary parts are of perturbative origin . in the remainder of this section ,
we discuss in detail the various issues to be addressed in the evaluation of the factorization formula . in section
[ subsec : wilson ] , a modified renormalization - group treatment of electroweak penguin effects is introduced , which is more appropriate than the standard scheme as far as applications to rare hadronic @xmath0 decays are concerned .
each of the diagrams in figure [ fig : graphs ] contains a leading - power contribution relevant to ( [ fact ] ) and power - suppressed terms , which do not factorize in general . an important class of such power - suppressed effects is related to certain higher - twist meson distribution amplitudes .
these amplitudes are defined in section [ subsec : power ] , and their leading , chirally - enhanced contributions to the nonleptonic decay amplitudes are evaluated .
section [ subsec : airesults ] contains a compendium of the relevant formulae for the calculation of the parameters @xmath50 .
section [ subsec : annihilation ] is devoted to annihilation topologies and the definition of the parameters @xmath51 . in the conventional treatment of the effective weak hamiltonian ( [ heff ] ) ,
the initial conditions for the electroweak penguin coefficients at the scale @xmath37 are considered a next - to - leading order effect , because they are proportional to the electroweak gauge coupling @xmath3 ( see @xcite for a detailed discussion ) . for our purposes , however , it is preferable to deviate from this standard power counting and introduce a modified approximation scheme .
the reason is that the electroweak penguin contributions in @xmath31 decays and other rare processes based on @xmath36 transitions are important only because they compete with strongly ckm - suppressed tree topologies .
electroweak penguins and tree topologies together are responsible for the isospin - violating contributions to the decay amplitudes @xcite .
therefore , their effects are important even though they are suppressed with respect to the leading qcd penguin amplitude , which conserves isospin . the ratio of electroweak penguin to tree amplitudes scales like @xmath165 , where @xmath166 is the wolfenstein parameter .
moreover , the dominant electroweak penguin effects are enhanced by a factor of @xmath167 and @xmath168 .
hence it is not appropriate to count @xmath3 as a small parameter in the renormalization - group evolution , if the effect of interest is related to isospin breaking .
we now describe a systematic modification of the usual leading and next - to - leading approximations , in which the dominant part of the electroweak penguin coefficients at the scale @xmath37 is treated as a leading - order effect .
it is then consistent to include the qcd radiative corrections to the enhanced terms in the initial conditions for the electroweak penguin coefficients and , at the same time , the corrections of order @xmath33 to the matrix elements of the electroweak penguin operators , which represent the next - to - leading order corrections to the hard - scattering kernels in the factorization formula .
using a compact matrix notation , the solution to the renormalization - group equation for the wilson coefficients @xmath169 in ( [ heff ] ) can be written as @xmath170 \vec c(m_w ) \,.\ ] ] the matrices @xmath171 , @xmath172 , @xmath173 , and @xmath174 depend on the ratio @xmath175 and on the anomalous dimensions and @xmath176-function . at leading order
, the evolution matrix reduces to @xmath177 .
the remaining terms shown above are the next - to - leading corrections .
we now expand the coefficients @xmath178 at the weak scale as @xmath179 where superscripts indicate the order in the strong coupling constant @xmath180 .
the term proportional to @xmath3 represents the electroweak contribution originating from photon - penguin , @xmath181-penguin and box diagrams .
we split this term into a contribution @xmath182 containing all terms enhanced by the large top - quark mass and/or a factor of @xmath168 , and a remainder @xmath183 .
as explained above , we treat @xmath182 as a leading effect and hence include the first two terms in its expansion in powers of @xmath180 .
the remainder @xmath183 is considered a next - to - leading effect , and so we only keep the first term in its perturbative expansion
. explicitly , the nonvanishing contributions to the initial conditions in the electroweak sector ( @xmath184 ) are @xmath185 \,,\ ] ] and @xmath186 where @xmath187 with @xmath188 . the inami
lim functions @xmath189 , @xmath190 and @xmath191 can be found , e.g. , in @xcite .
numerically , the remainder @xmath192 is indeed much smaller than @xmath193 , justifying our approximation scheme .
note that @xmath193 is gauge and renormalization - scheme independent .
the remainder @xmath192 is gauge - independent , but it carries the usual next - to - leading order scheme dependence of the electroweak coefficients .
explicit expressions for the qcd corrections contributing to @xmath194 have been obtained in @xcite . using these results ,
we obtain the approximate expressions ( valid for a high - energy matching scale @xmath195 ) @xmath196 in the conventional treatment @xmath194 would be absent , while the sum @xmath197 would be the usual initial condition for the electroweak coefficients at the scale @xmath37 , counted as a next - to - leading order effect .
in addition to the modified counting scheme for powers of coupling constants , we make a further , numerically excellent approximation , which greatly simplifies the systematic evaluation of the wilson coefficients .
in essence , it amounts to neglecting qed effects in the calculation of the wilson coefficients @xmath198 .
the values of these coefficients are obtained using the standard next - to - leading order approximation including only strong - interaction effects . at the same time
, we neglect qed corrections to the matrix elements of the operators @xmath199 .
( in fact , the virtual corrections of order @xmath3 are infrared divergent and require the inclusion of photon bremsstrahlung contributions in order to obtain physical results .
our approximation scheme avoids this complication . )
this treatment can be justified by noting that qed and electroweak contributions to the decay amplitudes in ( [ para ] ) are only important if they contribute to the parameters @xmath137 and @xmath151 . from ( [ params ] )
, it follows that the terms of order @xmath3 contained in the coefficients @xmath200 are enhanced by the prefactor @xmath201 .
it is thus sufficient for all practical purposes to only include the order @xmath3 corrections from the coefficients @xmath200 . on the contrary , qed corrections to the other amplitude parameters
can be safely neglected ( i.e. , the coefficients @xmath202 do not contain terms proportional to @xmath3 ) .
precisely this is achieved by our approximation scheme .
[ cols="<,^,^,^,^,^,^",options="header " , ] = 15.6 cm the uncertainty in the overall magnitude of the penguin amplitude due to weak annihilation effects , and the overall scale uncertainty in the predictions for the branching fractions due to hadronic form factors and ckm elements , are largely eliminated by taking ratios of branching fractions .
the dependence of the six relevant ratios on the weak phase @xmath2 is displayed in figure [ fig : rcp1 ] , in which the curves and bands have the same interpretation as in the previous figure .
( the results presented here are consistent with the preliminary results published in @xcite , which did not include annihilation contributions .
however , the treatment of theoretical uncertainties in @xcite differs from the present work . )
table [ tab : rcps ] gives numerical predictions and theoretical uncertainties for some selected values of the angle @xmath2 .
it is evident that annihilation effects are less important for the ratios .
the small error on the ratios involving @xmath38 final states shows the potential of these ratios to constrain @xmath2 . comparing our results with the preliminary experimental data collected in tables [ tab : data ] and [ tab : rcps ] ( and shown by the horizontal bands ) ,
we note a preference for large values of @xmath2 ; however , the experimental errors are still too large to assign much significance to this observation .
an important remark in this context concerns the ratio involving the neutral mode @xmath203 , which seems to prefer a small value of @xmath2 .
our theoretical results indicate that to a good approximation ( compare the two plots in the first row in figure [ fig : rcp1 ] ) @xmath204 this relation is in slight disagreement ( albeit by no more than @xmath205 ) with present data .
some authors have interpreted this fact as an indication of large rescattering phases ( see , e.g. , @xcite ) . however , the validity of ( [ isospin ] ) is a model - independent consequence of isospin symmetry , which is valid to linear order in the small tree - to - penguin ratios @xcite .
therefore , we expect that with more precise data the values of the two ratios in ( [ isospin ] ) will come closer to each other .
finally , we wish to stress that the difference between the qcd factorization results and naive factorization are largest for the ratio of the @xmath206 and @xmath207 final states . whereas we can accommodate the low experimental value of this ratio for @xmath208 ( at the @xmath205 level ) , this would not be possible for any value of @xmath2 if one were to use the naive factorization model .
it is evident from figure [ fig : rcp1 ] that in many cases the error on @xmath209 constitutes one of the largest uncertainties in the predictions for the ratios of branching fractions .
this is true , in particular , for the interesting ratio of the @xmath206 and @xmath210 final states .
for instance , if @xmath209 is 20% lower than its current central value , then our prediction for the @xmath211 branching ratio is reduced by 40% , whereas the result for the @xmath212 branching fraction remains almost unaffected .
this shows that for future analyses of rare @xmath0 decays it will be important to reduce the theoretical uncertainty in the value of @xmath46 .
we can unfold the strong dependence of the ratios on @xmath209 by presenting the predictions for the branching fractions as functions of the wolfenstein parameters @xmath44 and @xmath45 .
this is illustrated in figure [ fig : rercp ] for the two cases of the ratios @xmath213 and @xmath214 .
the other two ratios involving @xmath38 final states exhibit a dependence similar to the first of these two examples . for not too large values of @xmath45 ,
any of the three ratios involving only @xmath38 final states is a direct measure of @xmath44 , since the dominant dependence on the ckm parameters arises from @xmath215 . for positive @xmath44
the theoretical uncertainty is relatively small , as can be deduced by comparing the three panels of figure [ fig : rercp ] that refer to @xmath213 .
( note that the contour lines in this plot do not coincide with those of constant @xmath216 inferred from figure [ fig : ckmfit ] , since there the parameter @xmath128 is assumed to be fixed and extracted from data , whereas here it is determined from theory and hence proportional to @xmath209 . )
the ratio @xmath217 exhibits a rather different behaviour .
the circular contours reflect the increase of the @xmath206 branching fraction with @xmath46 .
the offset of the center to negative values of @xmath44 results from the interference of the two contributions with different weak phases to the decay amplitudes , which for @xmath218 is constructive ( destructive ) for the @xmath219 ( @xmath38 ) final states ( and vice versa for @xmath220 ) .
= 15.6 cm the current experimental values for the cp - averaged branching fractions are still afflicted by large errors , so a fit of @xmath4 to these ratios may appear premature . nonetheless , it is useful to carry out such an analysis at the present stage , not only to exhibit its potential once the errors decrease , but also to gauge the limitations eventually set by the theoretical uncertainties .
figure [ fig : reband ] shows the results separately for each of the five ratios of cp - averaged branching fractions that have been measured so far .
consider first the dark - shaded band in each plot , which covers the region allowed by theory given the current central experimental value .
the light - shaded bands in figure [ fig : reband ] represent the theoretically allowed regions obtained by using experimental values for the ratios shifted up or downwards by two standard deviations .
hence , the width of each band thus reflects the ( largely irreducible ) theory uncertainty given a certain value of the observable , while the current constraint on @xmath44 and @xmath45 provided by each observable separately can roughly be taken to be the entire region bounded by the two light - shaded bands . the information from all cp - averaged @xmath1 branching fractions ( and , in the future , from the corresponding cp asymmetries )
can be combined into a single global fit , giving regions for the wolfenstein parameters @xmath44 and @xmath45 that are allowed by theory .
we will now determine these regions , using two slightly different treatments of theoretical uncertainties . because our focus is to determine fundamental parameters of the standard model , we adopt a more conservative error analysis than in the previous paragraphs of this section and scan all theory input parameters over their respective ranges of uncertainty ( rather than adding theoretical uncertainties in quadrature ) .
the same conservative treatment was used in the analysis of the weak phase @xmath2 in section [ sec : nr ] .
it is evident from our previous discussion that ratios of cp - averaged branching fractions provide the most powerful constraints on the wolfenstein parameters , since they are afflicted with relatively small theoretical uncertainties .
however , there are two reasons for not performing a global fit to the ratios directly ( see also the discussion in appendix c of @xcite ) : first , with the present large experimental errors there are significant differences in the results of the fit depending on whether the ratios or inverse ratios are used , leading to an element of arbitrariness ; secondly , even if two quantities have gaussian errors , their ratio does not .
it is therefore preferable to perform the global fit for the individual branching ratios , despite the fact that the predictions for the branching fractions have larger ( and correlated ) theoretical uncertainties .
we now present two versions of such a fit . in the first method , we find the allowed region from @xmath44 and @xmath45 by determining first the 95% confidence level contour for a given theoretical input , and then scanning over all models in the theory parameter space that survive a 5% confidence cut .
the theory parameter space is given by the set of all input parameters confined to the ranges specified in table [ tab : inputs ] , variation of the renormalization scale @xmath221 between @xmath222 and @xmath223 , and scanning of the parameters @xmath224 and @xmath225 in the ranges described in ( [ xhparam ] ) and ( [ xaparam ] ) .
the resulting @xmath226 confidence region is bounded by the dark solid lines in figure [ fig : chi2fit ] .
the grey solid line inside this region envelopes the set of @xmath227 points of all models that pass the 5% confidence cut .
the dashed dark and grey lines have the same interpretation except that only theory models with a @xmath227 per degree of freedom of less than 1 are selected .
the @xmath4 range so determined is compared with the standard ckm fit ( also at 95% confidence level ) @xcite , which uses information from semileptonic @xmath0 decays ( @xmath46 and @xmath228 ) , @xmath229@xmath230 mixing ( @xmath47 ) , and @xmath0@xmath48 mixing ( @xmath231 , @xmath232 , and @xmath14 ) .
the constraint corresponding to @xmath233 is indicated by the solid annulus .
a fit to cp - averaged branching fractions can not distinguish between @xmath45 and @xmath234 , and so for simplicity only the upper half of the @xmath4 plane is shown in the figure . at present our contours are symmetric about the @xmath44 axis . in the future
, data on direct cp asymmetries in the various @xmath1 modes should be included in the fit , in which case it will become possible to obtain information about the sign of @xmath45 . before interpreting the results of this analysis
, we discuss a second method for performing the global fit , which follows the strategy outlined in @xcite .
it differs from the method explained above in two ways .
first , the large overall scale uncertainty ( of almost 40% ) due to the error on the value of the form factor @xmath235 , which is common to the theoretical predictions for all branching fractions ( see , e.g. , ( [ brs ] ) ) , is eliminated by fitting the form factor to the data . in that way , the sensitivity to the wolfenstein parameters in increased in the same way as it would be by using ratios ; however , the statistical problems of using ratios ( see above ) are avoided .
specifically , we start from the @xmath236 function @xmath237 where @xmath238 are the experimental values for the branching ratios , @xmath239 are their errors , and @xmath240 are the corresponding theoretical predictions obtained with the fixed form factor @xmath241 .
the scale factor @xmath242 parameterizes the deviations of the product @xmath243 from its central value . for a given set of theory parameters , we determine @xmath244 such that @xmath245 is minimized , provided that the value of @xmath242 is allowed by the theory ranges for the form factor ( see table [ tab : inputs ] ) and @xmath228 . for most choices of theory parameters
this condition is satisfied .
if it is not , we choose the value of @xmath242 inside the allowed range that is closest to the optimal value in ( [ sdet ] ) .
next , we scan over all theory parameters and determine the global minimum of @xmath236 in the @xmath4 plane .
we then plot contours of constant @xmath246 corresponding to the 68% , 90% and 95% confidence levels .
the second difference with respect to our first scanning method is that these contours refer to a constant @xmath236 , whereas in the first method the @xmath236 is different for each of the theory models considered .
the results of this analysis are shown in figure [ fig : chi2babar ] .
the best fit has an excellent @xmath247 ( for 3 degrees of freedom ) , corresponding to a confidence level of 54% .
the parameters of the particular theory model corresponding to this fit are all very reasonable .
in particular , the quantity @xmath225 parameterizing the weak annihilation terms is even smaller in magnitude than our default value .
hence , a good global fit to the data can be obtained without using extreme choices of input parameters , or invoking phenomenological recipes such as largely enhanced annihilation or charm - penguin contributions .
the default parameter set in table [ tab : inputs ] also yields a good fit , which has @xmath248 and 21% confidence level .
note that at 90% confidence level the allowed range obtained by combining our results with the measurement of @xmath209 in semileptonic @xmath0 decays excludes @xmath249 , thus establishing the existence of a cp - violating phase in @xmath250 transitions .
the results of the two scanning methods shown in figures [ fig : chi2fit ] and [ fig : chi2babar ] are similar ( though the range obtained in the first method is somewhat more conservative ) .
the allowed region for @xmath44 and @xmath45 obtained from our analysis of rare hadronic @xmath0 decays is compatible with the standard global fit using information from semileptonic @xmath0 decays , @xmath229@xmath230 mixing and @xmath0@xmath48 mixing .
however , the best fits to the rare - decay branching fractions prefer a larger value of @xmath2 ( of about @xmath251 ) .
this has been noted by many other authors in the past , usually based on the naive factorization approximation and without a theoretical error estimate .
here we have put this analysis on a firmer theoretical footing . on the other hand , we see that small values of @xmath2 are also allowed , provided the value of @xmath46 is lower than its standard central value . combining our results with the standard fit would reduce the allowed region in the @xmath4 plane by about a factor 2 , indicating that even with present experimental ( and theoretical ) errors the information obtained from rare hadronic decays provides important constraints on the wolfenstein parameters .
it will be interesting to follow how the comparison of the standard fit and the fit to rare hadronic @xmath0 decays will develop as the data become more precise .
if the two fits remain consistent with each other even as the allowed regions shrink in size , this would constitute a highly nontrivial test of the ckm model , in which for the first time the phase of the @xmath252 matrix element ( which enters through tree penguin interference ) is restricted by both direct and indirect measurements .
if , on the other hand , the two regions would not overlap at a reasonable confidence level , this may suggest that the standard theory of weak interactions does not account completely for either the @xmath0@xmath48 mixing amplitude ( since this is what determines the left - most side of the allowed region in the standard fit ) or the loop - induced flavour - changing amplitudes in @xmath1 decays .
( another option would be to abandon the theoretical framework advocated here , but we understandably leave it to others to pursue this avenue . )
= 15.6 cm the qcd factorization approach predicts the strong - interaction phases and hence the direct cp asymmetries in the heavy - quark limit . in the following ,
we define the asymmetries as @xmath253 where @xmath254 contains a @xmath255 quark ( rather than antiquark ) .
the dependence of the cp asymmetries on the weak phase @xmath2 is displayed in figure [ fig : acp1 ] .
the asymmetries are typically predicted to be small , concurrent with the fact that strong - interaction phases are suppressed in the heavy - quark limit @xcite .
even the largest asymmetries for the @xmath38 final states are predicted to be of order 10% or less .
an exception to this rule is the final state @xmath256 , for which the asymmetry is sensitive to qcd penguins and to the coefficient @xmath257 , both of which have large uncertainties and a potentially large relative phase . while our approach predicts the generic magnitude of the cp asymmetries , their precise values remain uncertain
this is not surprising , since in contrast to the branching ratios the asymmetries are more sensitive to and in fact generated by
corrections to naive factorization .
hence , they are subject to larger relative uncertainties .
in particular , the cp asymmetries are proportional to the sines of strong - interaction phases , which are of order @xmath258 or @xmath34 . only the leading perturbative contributions to these phases
are calculable . on the contrary ,
the cp - averaged branching fractions depend on the cosines of strong - interaction phases , which are equal to 1 in the heavy - quark limit
. first measurements of cp asymmetries have been published by the cleo @xcite and belle @xcite collaborations .
all results are compatible with no asymmetry , with typical @xmath259 errors of about @xmath260
. these errors are not yet small enough to draw meaningful conclusions , except that very large asymmetries appear already improbable . in the future
, accurate measurements of cp asymmetries will test the generic prediction of the qcd factorization approach that strong - interaction phases are small .
if this prediction is confirmed , further analysis of the cp asymmetries should help to determine the sign of @xmath2 , which can not be probed using cp - averaged branching fractions .
in addition , data on cp asymmetries may help to further constrain annihilation contributions and other theoretical input parameters in the qcd factorization approach .
a general overview of the various approaches to the evaluation of the matrix elements in exclusive hadronic @xmath0 decays has been given in @xcite .
in that reference we have commented on the strategies and methods that had been employed previously and explained how they are related to qcd factorization . in this section
we discuss recent papers on the subject and clarify the differences with our work .
we divide these papers into four categories : those using our qcd factorization formalism , those using a perturbative pqcd approach , those in which nonperturbative effects are estimated using light - cone sum rules or model calculations , and finally those containing phenomenological analyses .
we consider each of these in turn . before discussing other approaches to nonleptonic @xmath0-decays it may be helpful to summarize again the conceptual basis of the qcd factorization approach .
the key ingredient is the systematic analysis of feynman graphs in the heavy - quark limit , from which we deduce the factorization of infrared singularities into hadronic light - cone distribution amplitudes and form factors .
this enables us to establish the factorization formula ( [ fact ] ) .
the separation of long- and short - distance contributions to the decay amplitudes , necessary to establish factorization , holds only to leading power in @xmath34 and is based on considerations analogous to those used to demonstrate factorization in other applications of qcd to hard processes ( such as deep inelastic scattering , drell yan production , and electromagnetic form factors of hadrons at large momentum transfer ) .
the factorization formula leads to a model - independent treatment of exclusive hadronic @xmath0 decays in the heavy - quark limit .
a consistent counting scheme for powers of @xmath34 and a systematic identification of all the leading contributions are crucial for establishing this result .
the framework proposed in @xcite is general and provides a starting point for further theoretical developments , such as the improved understanding of the nonperturbative input ( e.g. , the @xmath261 form factor and the light - cone distribution amplitudes ) and estimates of power corrections .
several authors have used the framework of qcd factorization for applications to two - body hadronic @xmath0 decays @xcite .
we now compare the results reported in these papers with ours and comment on apparent differences and discrepancies .
muta et al .
have generalized the results of our previous work @xcite by including electroweak penguin contributions , and have applied the qcd factorization approach to the decays @xmath31 , @xmath219 .
we would like to point out the following differences with respect to the present work : * the hard - scattering kernels are derived only for symmetric light - cone distribution amplitudes . * in evaluating the @xmath33 corrections to the penguin coefficients @xmath262 and @xmath86 , the existence of the two distinct types of penguin contractions shown in figure [ fig : penguin ] is not taken into account . as discussed in section [ subsec : comments ] , this leads to incorrect terms proportional to @xmath263 in these coefficients . * an incomplete projector for the twist-3 , two - particle distribution amplitude of the pion
is employed .
this gives an incorrect contribution proportional to @xmath264 in @xmath86 .
* we also observe a disagreement with the remaining terms in the correction of order @xmath33 to @xmath86 , which concerns the function @xmath265 in @xcite .
all three components , the @xmath266-dependent part , the constant , and the coefficient of the logarithm , differ from our findings .
we note that the @xmath221-dependence of @xmath267 is in conflict with the requirement of renormalization - group invariance of the product @xmath268 . in addition , the contributions from the radiative vertex corrections to @xmath86 and @xmath87 are missing . *
a minor difference comes from the fact that we neglect electroweak penguin contributions to @xmath262 and @xmath86 , while these are included in @xcite . on the other hand , muta et al .
omit the contributions proportional to @xmath269 , @xmath270 and @xmath271 to the electromagnetic penguin coefficients @xmath87 and @xmath272 . * the twist-3 contributions to the spectator hard - scattering amplitudes and annihilation effects
are not discussed .
as we have seen , these corrections are the potentially most important source of theoretical uncertainty .
the formulae given in @xcite have also been applied to discuss @xmath0 decays into vector
pseudoscalar final states @xcite and final states containing @xmath273 and @xmath274 @xcite .
the previous comments apply also to these papers , as they do to @xcite , which relies on the same expressions except for dropping the formally power - suppressed terms proportional to @xmath86 and @xmath87 . in a series of papers , du et al .
have discussed @xmath0 decays into two light pseudoscalar mesons @xcite .
we focus on the two last papers in @xcite , which contain the most complete and updated results .
similarly to the present work , electroweak penguin contributions and chirally - enhanced twist-3 components of the pion distribution amplitude are included in these papers , but no weak annihilation effects are considered .
the hard - scattering kernels are given only for symmetric distribution amplitudes , and explicit results are presented for the case of @xmath32 decays .
we will therefore not distinguish between @xmath219 and @xmath38 final states in the present discussion .
the formulae for the coefficients @xmath50 can be directly compared with our results .
a minor difference concerns the tiny electroweak penguin contributions to the coefficients @xmath275 , which we decided to neglect in our approximation scheme , but which are retained in @xcite .
next we recall from section [ subsec : comments ] that in the approximation of including twist-3 contributions only when they are chirally enhanced , the equations of motion require the use of asymptotic distribution amplitudes @xmath276 , @xmath277 .
du et al.have considered a more general situation , which would affect the cancellation of endpoint singularities and the renormalization - scale dependence of some results .
use of the appropriate asymptotic distribution amplitudes eliminates these spurious ambiguities . finally , we note two minor discrepancies : * our expression for the twist-3 , hard spectator interaction in ( [ hiexpr ] ) contains a factor @xmath278 in the integrand , whereas the corresponding result given in eq .
( 37 ) of of the third paper in @xcite has @xmath279 using the asymptotic forms of the twist-3 distribution amplitudes @xmath276 and @xmath277 , we find that the two results differ by a factor of @xmath280 . * in our result for the twist-3 penguin contribution @xmath281 in ( [ hatpk ] ) ,
the coefficient of @xmath269 , for example , is @xmath282 the equivalent expression given in eq .
( 23 ) of the third paper in @xcite reads @xmath283 using asymptotic twist-3 distribution amplitudes , one may check that @xmath284 and @xmath285 .
the two results then agree up to a constant @xmath286 .
this difference can be traced back to an inconsistent use of twist-3 projectors in four space - time dimensions within a @xmath287-dimensional loop calculation , before the subtraction of ultraviolet poles is performed ( see the last reference in @xcite ) .
analogous differences occur in the other terms in @xmath86 and @xmath87 .
apart from these discrepancies , the expressions agree with our results for symmetric distribution amplitudes .
cheng and yang have applied the qcd factorization approach to a study of the decays @xmath288 @xcite .
annihilation topologies are discussed and argued to be important numerically .
a few minor discrepancies with our results occur in the expressions for the coefficients @xmath289 .
the penguin contraction of the operators @xmath290 is treated incorrectly , as discussed in section [ subsec : comments ] .
photonic penguin contractions of the operators @xmath291 contributing to @xmath292 are omitted .
also , qcd corrections to @xmath293 are not considered , presumably because the corresponding amplitudes are formally power suppressed .
similarly to the results of @xcite , the twist-3 kernel for the hard spectator interaction contains an additional factor of @xmath280
. finally , we briefly comment on the analysis of weak annihilation contributions . in @xcite the final state
consists of a vector meson and a kaon rather than two pseudoscalar mesons .
the decay amplitudes are estimated in the approximation of using leading - twist distribution amplitudes for the @xmath18 meson , but including chirally - enhanced twist-3 contributions for the kaon . taking this systematic difference into account
, we can still compare the results of @xcite with our expressions .
the two agree in the case of @xmath294 operators , but the results for @xmath295 operators are different .
the latter are the ones that depend on the twist-3 contributions , but we do not have sufficient information to trace back the origin of the discrepancy . recently , keum , li and sanda have presented an extensive study of @xmath31 decays within a perturbative hard - scattering ( or `` pqcd '' ) approach @xcite . while the underlying goal of a separation of soft and hard physics in the @xmath0-decay matrix elements is similar in spirit to qcd factorization , there are fundamental differences in the implementation of this idea . in the pqcd approach ,
the @xmath261 and @xmath296 form factors are assumed to be perturbatively calculable .
this assumption is justified by invoking sudakov effects to regulate the infrared - sensitive contribution from the endpoint region in the integral over the light - cone momentum of the outgoing spectator quark . in other words ,
the contribution from the region where the spectator quark is soft is supposed to be strongly suppressed , and therefore the exchange of a hard gluon is always required .
instead of being of leading order ( in the qcd coupling constant ) , as in the generic case of a @xmath261 form factor dominated by soft physics , the form factor is now counted as being of order @xmath33 in perturbation theory .
thus , the hierarchy of the various contributions to the decay amplitudes in pqcd is very different from that in our approach : * in the pqcd approach , the @xmath31 matrix element is a quantity of order @xmath33 .
to this order , therefore , all the wilson coefficients can then be taken in the leading logarithmic approximation . * the `` nonfactorizable '' hard gluon - exchange contributions to the kernels @xmath297 in ( [ fact ] )
enter at order @xmath298 and are therefore omitted .
note , however , that it is essential to consider such effects in order to establish factorization .
this has never been done explicitly in the pqcd framework . *
the form - factor terms containing the kernels @xmath299 and the hard - scattering terms containing the kernels @xmath164 in the factorization formula are treated as being of the same order in the pqcd approach .
the former are called `` factorizable '' and the latter `` nonfactorizable '' . because both terms vanish simultaneously in the limit @xmath300 , naive factorization is not recovered in any limit of qcd .
this is in contrast to our interpretation of the qcd factorization formula , for which naive factorization is recovered in the heavy - quark limit .
after this brief synopsis of the key features of the pqcd scheme , we now examine critically the main assumptions of this approach in the context of @xmath31 decays . in the pqcd approach the transverse momenta of the quarks
are kept explicitly when evaluating contributions that are potentially infrared sensitive .
for instance , the gluon - exchange kernel for the @xmath261 form factor is written as @xmath301 } + \ldots \,,\ ] ] where the ellipses represent less singular terms . here
@xmath302 ( @xmath303 ) is the light - cone momentum fraction of the spectator quark in the @xmath0 meson ( pion ) . in the conventional collinear expansion adopted in our work ,
the transverse momenta @xmath304 would be treated as higher - twist effects and dropped ( more precisely , the amplitude would be expanded in powers of transverse momenta ) .
it then follows that @xmath305 , and so the convolution with the leading - twist pion distribution amplitude @xmath306 would lead to an integral @xmath307 , which is infrared divergent .
this well - known result can be interpreted as a sign of the dominance of soft endpoint contributions in the @xmath261 form factor @xcite . in @xcite ( see also @xcite ) this divergence is regulated by keeping the transverse momenta .
then a fourier transform from transverse momentum space into impact parameter space ( @xmath255 space ) is performed , and a sudakov factor is included for each of the meson distribution amplitudes .
this factor strongly suppresses the region of large @xmath255 , and the @xmath261 form factor is therefore assumed to be perturbatively calculable .
it should be noted that retaining the transverse momentum dependence in @xmath308 , which is only part of the complete higher - twist contribution , is a model - dependent procedure .
it is precisely in the critical endpoint region @xmath309 that the leading two - particle fock - state description that underlies this analysis breaks down .
a further puzzle to us is the assumption made in @xcite that the spectator quark in the @xmath0 meson has a large momentum component in the `` @xmath310 '' direction , @xmath311 , when the pion momentum is in the `` @xmath312 '' direction .
this is used to derive the sudakov factor for the @xmath0 meson in analogy to the case of a fast light meson .
since the spectator quark in the @xmath0 meson is intrinsically soft , we see no justification for such an assumption . in conclusion , we believe that the perturbative evaluation of the @xmath261 form factor , which is one of the central ingredients of the pqcd analysis in @xcite , is not justified .
the relevance of sudakov form factors in hadronic @xmath0 decays should be investigated in a more systematic way .
however , it seems unlikely to us that sudakov logarithms are sufficiently important at the scale @xmath313 to eliminate the soft contributions to heavy - to - light form factors and thus render them calculable in a model - independent way .
a complete description of the @xmath261 transitions at large recoil in the heavy - quark limit and the derivation of a ( hypothetical ) factorization theorem for these processes has , to our knowledge , never been presented . for recent progress in this direction and further critical discussions of the problem
see @xcite .
keum et al .
claim a dynamical enhancement of ( factorizable ) contributions to the matrix elements of penguin operators ( such as @xmath314 and @xmath315 ) in the effective weak hamiltonian .
the first reason given in support of this assertion is a dynamical enhancement of the @xmath261 form factor of the scalar density @xmath316 contained in ( the fierz - transformed form of ) @xmath315 , relative to the form factor of the vector current @xmath317 relevant for other operators .
a dynamically different structure for the form factor of the scalar density in pqcd is claimed , and a strong sensitivity of the effect to the model employed for the @xmath0-meson light - cone distribution amplitude is noted . however , an enhancement of this nature is not possible , because the form factor of the scalar density is related to that of the vector current by the equations of motion , leading to @xmath318 the erroneous conclusion in @xcite is the consequence of an incorrect treatment of the twist-3 contribution to the pion distribution amplitude ( see below ) .
the second reason given in favor of a penguin enhancement is the choice of a low scale @xmath221 in the ( leading - order ) penguin coefficients @xmath319 and @xmath320 .
this is motivated by arguing that @xmath221 should be a typical scale intrinsic to the dynamics of the @xmath261 form factor in the pqcd evaluation .
however , the scale @xmath221 is unphysical and must be canceled by vertex corrections to the operator matrix elements , and in the case of @xmath320 also by the running quark masses .
such corrections have been properly included in the present work , but they were neglected in @xcite .
the scale @xmath221 in @xmath321 may depend on the scale of non - factorizable hard spectator interactions , but it is clearly independent of the internal dynamics of the form factor .
the arguments made in favour of a low scale @xmath221 and correspondingly increased penguin coefficients are therefore not well justified . in their analysis ,
keum et al .
find that penguin annihilation contributions are numerically very important and give a dominant contribution to some of the @xmath31 decay amplitudes .
we have investigated these effects in the present work and found them to give a significant correction ( compatible with being a power correction of canonical size ) , but not a dominant contribution . nevertheless , it must be emphasized that weak - annihilation diagrams contribute at subleading power in the heavy - quark expansion and , in general , are not calculable in a qcd - based factorization approach .
therefore , the question about the relevance of these effects warrants further investigation . it is claimed that large strong - interaction phases are generated by hard gluon exchange with the spectator quark in the @xmath0 meson , as well as by weak annihilation diagrams .
let us consider the hard spectator interactions in detail . in that case
the effect is calculable within the qcd factorization approach and found to be real to leading order .
the source of the imaginary part found in @xcite is as follows .
the quark propagator entering the hard - scattering diagram is written as @xmath322 working to leading power we would drop the transverse components and find a real contribution to the kernel .
keum et al . , on the other hand , keep the transverse momenta and hence generate an imaginary part proportional to @xmath323 , where @xmath324 .
again , this treatment is model dependent .
the fourier transform of the kernel into impact parameter space results in a bessel function with the imaginary part proportional to @xmath325 . in the evaluation of the matrix element
this function is convoluted with the sudakov factor .
since the bessel function is oscillating rapidly , with the amplitude of @xmath326 growing like @xmath327 , the result of this convolution is very sensitive to the details of the @xmath255-space cut - off .
this implies that the estimate of the strong - interaction phase is both model dependent and numerically sensitive to effects that are poorly under control .
similar comments hold for the estimate of the strong - interaction phases from the annihilation contributions , which are generated in an analogous way .
finally , we remark that the simple @xmath328 structure of the twist-3 projection for the pion employed by keum et al .
is incomplete .
the proper treatment is discussed in section [ subsec : comments ] .
the wrong twist-3 projection is , in particular , inconsistent with gauge invariance .
moreover , the correct asymptotic behaviour of the twist-3 pion distribution amplitude @xmath329 is proportional to a constant , whereas the functional form @xmath330 is assumed in @xcite .
this problem affects all decay amplitudes , including the @xmath261 form factor , the corresponding spurious penguin enhancement , and the nonfactorizable spectator interactions and annihilation contributions .
other analyses in the spirit of the pqcd approach were presented in @xcite , to which similar comments apply . in @xcite ,
the presence of a `` recoil - phase '' effect was advocated , which was claimed to originate within the pqcd framework .
this phase should affect , e.g. , the @xmath261 form factor , which was assumed to be dominated by hard gluon exchange .
it was argued that , when the gluon is exchanged between the spectator quark and the @xmath255 quark , the @xmath255-quark propagator could go on the mass shell because @xmath331 .
this in turn would lead to a complex phase in the form factor . in our opinion
this conclusion is unwarranted , since bound - state effects have to be factorized before a perturbative treatment can be justified and thus can not influence the hard - scattering process .
another fundamental objection to the `` recoil phase '' is that it would contradict the fact that the @xmath261 form factor is a real quantity .
recently , khodjamirian has suggested to study hadronic matrix elements for two - body @xmath0 decays in the framework of light - cone qcd sum rules @xcite .
as an example , he discusses the matrix elements of the current current operators @xmath115 and @xmath116 for @xmath53 decays . the idea is to generalize the light - cone qcd sum - rule analysis of the @xmath261 form factor directly to the case of hadronic two - body modes . the starting point is the correlation function @xmath332|\pi^-(q ) \rangle \,,\ ] ] where @xmath333 and @xmath334 are interpolating currents for the emission pion and the @xmath0 meson , respectively .
the explicit pion state @xmath335 represents the recoil pion that absorbs the spectator quark . according to the qcd sum rule philosophy ,
the correlator is evaluated in two ways : by a direct calculation in qcd , and by inserting complete sets of hadronic states between the operator @xmath336 and the interpolating currents , extracting the desired ground - state contribution with the help of quark hadron duality .
the two sides of the equation are expressed in the form of dispersion relations , and the usual borel transformation is applied . to leading order in @xmath33 , the factorized result for the matrix element
is recovered . a particular type of power correction
is then estimated as a further illustration .
this contribution comes from higher twist , three - particle ( quark antiquark gluon ) fock components of the recoil pion , where the ( nonfactorizable ) gluon couples to the quark lines of the emission current . in the heavy - quark limit the resulting expression
is demonstrated to scale as @xmath34 relative to the leading contribution .
it is estimated to be of relative size @xmath337 with @xmath3380.15gev .
note that this correction has no rescattering phase .
its numerical effect is small , but comparable to the perturbative corrections at leading power .
additional nonfactorizable contributions exist but have not yet been investigated .
examples are the gluon - exchange effects that correspond to the vertex corrections and the spectator scattering diagrams in the qcd factorization formula . in the sum - rule approach
these effects are in principle contained in similar diagrams , together with certain hadronic corrections of subleading power .
a potential difficulty will be to disentangle power corrections to the asymptotic result given by the qcd factorization formula from uncertainties intrinsic to the sum rule method , such as the assumption of quark hadron duality and the approximation of the emission pion by an interpolating current .
it is not fully clear how this can be achieved in a controlled and systematic fashion .
some recent papers @xcite have tried to model soft final - state interactions via the rescattering of certain hadronic channels such as @xmath339 .
we consider such an approach to be problematical for a variety of reasons .
there are many more intermediate channels beyond those taken into account .
systematic cancellations among these channels , which are predicted to occur in the heavy - quark limit , are missed when only a few intermediate states are retained ( see the discussion in sections 3.4 and 7.2 of @xcite ) .
moreover , the hadronic dynamics of multi - body decays is very complicated and in general not under theoretical control .
the main problem of such models is the lack of a systematic approximation scheme based on parametric expansions . in our opinion ,
the use of a purely hadronic language , suitable for kaon decays , is not very helpful in the case of @xmath0 decays , where the number of channels and the energy release are large . in @xcite ,
the coefficients @xmath50 obtained from the qcd factorization approach have been used in conjunction with a hadronic description of final - state interactions .
the general caveats concerning hadronic rescattering models apply also here .
moreover , such an approach faces a manifest double - counting problem , since rescattering effects in the heavy - quark limit are already contained in the qcd coefficients @xmath50 .
a nonperturbative treatment of `` charming penguins '' , matrix elements with charm loops of the type shown in figure [ fig : penguin ] , was proposed in @xcite . in this calculation
operators with flavour structure @xmath340 were split into nonlocal products of currents @xmath341 and @xmath342 .
these were connected by intermediate virtual @xmath343-meson states to yield transitions such as @xmath344 under the separate action of the currents .
however , the operators @xmath340 are local operators in the effective theory at the @xmath255-mass scale
. they would become nonlocal only if probed at the far higher scale of @xmath345 , but certainly not at a scale of order @xmath346 , as implied in @xcite .
we therefore see no basis for the method adopted in this paper .
there are several recent analyses in the literature that approach hadronic @xmath0 decays in a more phenomenological way .
some of the most extensive studies of this type have been presented by ali et al .
@xcite , who elaborate on the concept of `` generalized factorization '' @xcite . in these analyses ,
part of the order-@xmath33 vertex and penguin contributions calculated in the present work are included . as a consequence ,
the decay amplitudes receive strong - interaction phases due to the bander silverman
soni mechanism @xcite , and nonzero cp asymmetries are obtained . however , no systematic attempt to include all such effects ( or to prove factorization ) is made . also , these authors rely on ad hoc model assumptions such as an `` effective number of colours '' @xmath347 , which is introduced to parameterize nonfactorizable corrections .
a recent study of @xmath31 decays presented in @xcite emphasizes general , model - independent parameterizations of the decay amplitudes such as our parameterization in ( [ para ] ) , avoiding as much as possible the use of dynamical input .
for instance , strong - interaction phases are a priori allowed to take any value in the construction of bounds on ckm parameters .
information on these phases may then be inferred indirectly .
such an approach is complementary to the more ambitious goal of exploiting theoretical insight into the hadronic matrix elements .
in fact , the smallness of the rescattering effects parameterized by @xmath348 in ( [ para ] ) , which is a prediction of the qcd factorization approach , represents a useful input to such a phenomenological analysis .
many other phenomenological studies of two - body hadronic @xmath0 decays have been performed in the literature .
recent examples are @xcite , where simplifying assumptions are made and no systematic estimates of theoretical uncertainties are undertaken .
in this paper we have presented a detailed study of @xmath31 and @xmath32 decays based on the qcd factorization formula .
this approach allows us to perform a systematic and model - independent calculation of two - body hadronic @xmath0 decays in the heavy - quark limit .
we have evaluated the hard - scattering kernels entering the matrix elements for @xmath31 , @xmath219 decays at next - to - leading order in @xmath33 , thus obtaining predictions for the decay amplitudes including the leading , `` nonfactorizable '' corrections .
we have included the contributions from electroweak penguins , taking into account the corresponding order-@xmath33 corrections to the dominant terms ( enhanced by @xmath349 or @xmath168 ) in a consistent approximation scheme .
all hard - scattering kernels have been derived for general , asymmetric distribution amplitudes , as appropriate for @xmath229 mesons .
in addition to computing the model - independent leading - twist results , we have identified and estimated those power corrections which are expected to be the largest . we have analyzed the complete set of contributions from light - cone wave functions of twist 3 with a chiral enhancement factor @xmath350 or @xmath351 , as well as the contributions from weak annihilation topologies , including both twist-2 and twist-3 components in the light - meson wave functions .
we distinguish three classes of power corrections : the contributions from operators with ( pseudo- ) scalar currents ( @xmath352 ) , twist-3 effects in the hard spectator interactions , and weak annihilation corrections . while the first two effects are proportional to the chiral enhancement factor , the annihilation terms are quadratic polynomials in this factor .
since these are power - suppressed effects , we do not expect the factorization formula to work in these cases .
in fact , naively evaluating these terms in a `` hard - scattering '' framework one encounters logarithmic endpoint singularities .
( incidentally , the terms of order @xmath33 related to @xmath293 are free of such infrared singularities , but divergent terms are present for the other two contributions . ) therefore , while our results at leading twist are model - independent predictions of qcd in the heavy - quark limit , some model dependence is currently unavoidable in the description of power corrections .
we choose to regulate these divergences by introducing a simple cutoff , leading to the complex phenomenological parameters @xmath224 and @xmath225 .
the motivation for this model - dependent approach to power corrections is twofold : we obtain systematic order - of - magnitude estimates , while automatically keeping track of relative suppressions or enhancements from wilson coefficients and ckm factors .
in addition , we can investigate the sensitivity of the observables on the model parameters .
it turns out that annihilation graphs may in principle give sizable corrections . on the other hand ,
twist-3 effects in the hard spectator contributions appear generally to be less prominent . using this framework ,
we have performed a detailed comparison with the available experimental data on @xmath31 and @xmath32 branching ratios .
the data are beginning to be sufficiently precise to allow for detailed phenomenological analyses .
an important result is that in the qcd factorization approach the @xmath31 branching fractions can be larger than the @xmath32 ones without any tuning of the theoretical input parameters , and without invoking large phenomenological power corrections .
an acceptable fit to the branching fractions is obtained even if we impose that @xmath353 as implied by the standard constraints on the unitarity triangle .
this is in contrast to the naive factorization model , in which the observed branching fractions can either not be reproduced at all or require a large ckm angle @xmath2 , in conflict with other indirect determinations . encouraged by this success ,
we have derived constraints in the @xmath4 plane from a global fit to the @xmath31 , @xmath219 branching ratios , which already provide useful additional information on the unitarity triangle . clearly , the experimental situation will continue to improve and allow us to further test our framework and exploit it in phenomenological studies of hadronic @xmath0 decays . an important role in these analyses
will be played by the tree - to - penguin ratio @xmath128 , which can be reliably determined from @xmath17 and @xmath354 decays and can be unambiguously confronted with the theoretical predictions .
this will be a valuable check that the power corrections are not surprisingly large .
the absence of penguin and annihilation contributions in the @xmath17 mode is a crucial feature in this analysis
. a more precise measurement of this decay will thus be particularly useful .
we would like to conclude by emphasizing again the strategy underlying our treatment of weak annihilation contributions and non - leading , but chirally - enhanced spectator interactions .
while the present data do not require these effects to be larger than expected , the data also do not definitively exclude this possibility .
our error analysis is based on allowing the parameters @xmath355 and @xmath356 to be smaller than 1 ( reflecting the range of our expectations ) , but the error estimates sometimes depend sensitively on the precise choice of this upper limit , in particular in the case of @xmath355 . with more precise data one may contemplate fitting @xmath355 to data ( in practice this implies fitting the qcd penguin amplitude ) in order to decide upon the plausibility of the adopted range of values .
the discovery of large annihilation contributions would by itself constitute valuable insight into the strong - interaction dynamics of nonleptonic decays , but would evidently limit the utility of the qcd factorization approach .
the present data make this scenario appear unlikely , but we look forward to further information to validate our error estimation . in this case , important and reliable information on flavour physics will become available from @xmath31 and @xmath32 decays , as well as from a large class of other two - body hadronic decays into light pseudoscalar and vector mesons . _
acknowledgments : _ we are grateful to a. hcker and h. lacker for helpful discussions and for providing us with a data file of their ckm fit results .
we also thank g. martinelli and l. silvestrini for useful discussions .
the research of m.n . is supported in part by the national science foundation .
c.t.s . acknowledges partial support from pparc through grant no .
gr / k55738 .
m. beneke , g. buchalla , m. neubert and c.t .
sachrajda , preprint pitha-00 - 13 [ hep - ph/0007256 ] , to appear in the proceedings of the 30th international conference on high - energy physics ( ichep 2000 ) , osaka , japan , 27 july2 august 2000 .
a. khodjamirian et al . , phys .
d * 62 * ( 2000 ) 114002 [ hep - ph/0001297 ] ; + a. khodjamirian and r. rckl , in : _ heavy flavours ii _ , a.j .
buras and m. lindner eds .
( world scientific , singapore , 1998 ) pp . 345 [ hep - ph/9801443 ] .
a. ali and c. greub , phys .
d * 57 * ( 1998 ) 2996 [ hep - ph/9707251 ] ; + a. ali , g. kramer and c .- d .
lu , phys .
d * 58 * ( 1998 ) 094009 [ hep - ph/9804363 ] ; phys .
d * 59 * ( 1999 ) 014005 [ hep - ph/9805403 ] .
m. bauer , b. stech and m. wirbel , z. phys .
c * 34 * ( 1987 ) 103 ; + m. neubert and b. stech , in : _ heavy flavours ii _
buras and m. lindner ( world scientific , singapore , 1998 ) pp . 294 [ hep - ph/9705292 ] . | in the heavy - quark limit , the hadronic matrix elements entering nonleptonic @xmath0-meson decays into two light mesons can be calculated from first principles including `` nonfactorizable '' strong - interaction corrections .
the @xmath1 decay amplitudes are computed including electroweak penguin contributions , su(3 ) violation in the light - cone distribution amplitudes , and an estimate of power corrections from chirally - enhanced terms and annihilation graphs .
the results are then used to reduce the theoretical uncertainties in determinations of the weak phases @xmath2 and @xmath3 . in that way ,
new constraints in the @xmath4 plane are derived .
predictions for the @xmath5 branching ratios and cp asymmetries are also presented .
a good global fit to the ( in part preliminary ) experimental data on the branching fractions is obtained without taking recourse to phenomenological models .
cern - th/2001 - 107 + clns 01/1728 + pitha 01/01 + shep 01/11 + hep - ph/0104110 + april 11 , 2001 * qcd factorization in @xmath5 decays and + extraction of wolfenstein parameters * m. beneke@xmath6 , g. buchalla@xmath7 , m. neubert@xmath8 and c.t .
sachrajda@xmath9 + |
prompted by the necessity of a precise calibration of photo - detectors both in the analog and photon - counting regime @xcite , we presented recently @xcite a detailed theoretical analysis on the possibility to use the correlations of spontaneous parametric down conversion ( spdc ) light @xcite for calibrating analog detectors , as an extension of the technique developed in photon counting regime @xcite . however , while the method is suitable for analog calibration at relatively low gain , which means a photon flux lower than @xmath0 photon / s , in higher - gain regime it is limited by the difficulty of collecting the same correlated modes in the two branches @xcite . in this paper
we present a detailed theoretical analysis of a scheme based on the stimulated pdc ( on the other hand a discussion of the uncertainty budget is left to a specifically addressed paper @xcite ) .
we show that this scheme allow to overcome the problem mentioned above .
indeed , in this case the photon fluxes can be varied by varying the power of the coherent seed beam , without increasing the parametric gain @xmath1 . on the other hand , the spdc non - classical correlation at single - photon level which enables the absolute calibration of counting detectors survives , in some form , when a coherent seed is injected and the photon flux becomes macroscopic .
the schematic set - up for calibration of photodetectors by using pdc is shown in fig .
1 . we model the photodetection process in the analog regime as a random pulse train @xcite @xmath2 i.e. a very large number of discrete events at random times of occurrence @xmath3 . the pulse shape @xmath4 is determined by the transit time of charge carriers .
we assume that @xmath4 is a fixed function with characteristic width @xmath5 and a unit area .
the pulse amplitude @xmath6 is a random variable in order to account for a possible current gain by avalanche multiplication .
the statistical nature of the multiplication process gives an additional contribution to the current fluctuations @xcite .
in an ideal instantaneous photocell response , without avalanche gain , all values @xmath6 are equal to the charge @xmath7 of a single electron and @xmath8 . in the case of ideal quantum efficiency , since the probability density of observing a photon at time @xmath9 at detector @xmath10 ( @xmath11 ) is related to the quantum mean value @xmath12 of the photon flux , we calculate the average current output of @xmath13 as @xmath14 where the factor @xmath15 is the average charge produced in a detection event .
we have assumed the response function for the two detectors to be the same , @xmath16 . at the same time
, the auto - correlation and the cross - correlation functions for the currents can be expressed as @xmath17 respectively for @xmath18 where @xmath19 is the auto - correlation function of the photon flux at detector @xmath20 , and for @xmath21 where @xmath22 is the cross - correlation between the fluxes incident on the two different detectors . by the substitution @xmath23
, it is convenient to express them as @xmath24 the second contribution , proportional to the photon flux when @xmath18 , represents the the intrinsic and unavoidable component of the fluctuation that does not depend on the specific property of the field since it generates from the commutation relation of the quantum fields in the free space .
the third one is the normally ordered correlation function of the fluctuation .
actually it is a function just of the difference @xmath25 and its typical variation scale provides the coherence time @xmath26 of the pdc radiation .
now we introduce the quantum efficiency @xmath27 of detector @xmath10 , defined as the number of pulses generated per incident photon . in @xcite a real detector is modelled , as usual , with an ideal one ( @xmath28 ) preceded by a beam splitter with the intensity transmission coefficient equal to the quantum efficiency of the real detector @xcite . within this picture
it is possible to take into account the quantum efficiency by the following substitutions : @xmath29 thus , @xmath30 being time independent , according to eq .
( [ curr ] ) we obtain : @xmath31 eq . ( [ i_ji_k ] ) becomes @xmath32 where we introduced the convolution of the response function of detectors @xmath33 .
here we consider a type i pdc process . in the limit of monochromatic and plane - wave pump approximation ,
only pairs of modes with opposite transverse wave vectors , @xmath34 and @xmath35 , and with conjugate frequencies , @xmath36 and @xmath37 , are coupled such that the energy and momentum conservation hold .
the equations describing the down conversion process are the input - output transformations @xmath38 linking the fields @xmath39 and @xmath40 after the non - linear interaction to the fields @xmath41 and @xmath42 before the interaction started .
the coefficients @xmath43 and @xmath44 must satisfy the properties @xmath45 which guarantee the conservation of the free - field commutation relations @xmath46= \delta_{j , k}\delta_{\mathbf{q},\mathbf{q}'}\delta_{\omega,\omega'}\ ] ] @xmath47=0.\ ] ] here , @xmath48 and @xmath49 define the strength of the process and at the same time the range of transverse momentum and frequencies in which it takes place and thus are named gain functions .
in fact , the finite length of the crystal introduces a partial relaxation of phase matching condition concerning the third component of the momenta . by selecting a certain frequency , the transverse momentum uncertainty ,
i.e. the angular dispersion of the emission direction , is not null ( @xmath50 for the non - collinear pdc , where @xmath51 is the crystal length and @xmath52 is the central angle of propagation with respect to the pump direction ) . on the contrary , by fixing the transverse momentum @xmath34 , or equivalently a direction of propagation @xmath53 , the spectral bandwidth @xmath54 is proportional to @xmath55 , and typically @xmath56s for type i. hereinafter we will focus on the properties of the far field , observed at the focal plane of a thin lens of focal length @xmath57 , placed at distance @xmath57 from the crystal .
the spatial distribution of the far field is , in this case , the fourier transform of the field just at the output face of the crystal .
this special imaging configuration is convenient to show the basic of calculation , nonetheless the validity of the results is more general .
thus , any transverse mode @xmath34 is associated with a single point @xmath58 in the detection ( focal ) plane according to the geometric transformation @xmath59 .
the far field operator in the space - temporal domain is therefore @xmath60 the mean value of the operator @xmath61 is @xmath62 representing the intensity profile of the emission in the detection plane .
this mean value has to be calculated over the following initial state , in which the field 1 is in the vacuum state whereas the field 2 is in a multi - mode coherent state : @xmath63 where @xmath64 is the complex parameter associated to this state .
the seed beam is therefore represented as a coherent state of field 2 with fixed frequency @xmath65 and a certain distribution of the transverse momentum @xmath34 . here
we assume that , for the modes that are stimulated by the seed , the spontaneous component of the emission is negligible with respect to the stimulated one .
this corresponds to the assumption @xmath66 and , at the same time , @xmath67 , that is a typical experimental situation in which a few millimeters length crystal is pumped by a continuous pump and the seed has a power just around the microwatt or more .
the intensities ( [ stim1 ] ) of the two stimulated beams after the crystal can be evaluated by using eq.s ( [ input - output ] ) , with the substitution @xmath68 .
this leads to @xmath69 we used the first property of the gain functions in ( [ unitary cond ] ) into the last line of eq .
( [ stim4 ] ) , in order to show explicitly the two contributions to the beam 2 , one given by the original coherent seed and the other from the stimulated emission , proportional to the parametric gain @xmath70 .
we notice that , aside from the spontaneous emission that is neglected , the generated beams conserve the original momentum distribution of the seed but weighted according to the gain function .
the normal - ordered correlation function of the intensities fluctuation is defined as @xmath71 the calculation leads to the following normal - ordered mean values : @xmath72 @xmath73 fluctuations of the stimulated field 1 are negligible , because they have the same order of magnitude of the spontaneous emission . concerning the cross - correlation we have : @xmath74 let us consider two detectors , @xmath75 and @xmath76 , to register the photons crossing the regions @xmath77 and @xmath78 centered on the two correlated bright beams in the detection plane .
let the two areas be much larger than the dimension of the seed beam on the detection plane .
this simply means to collect all the bright components of the emission , corresponding to the modes stimulated by the seed .
we move now from the discrete modes representation of the electromagnetic field , in which it is artificially enclosed in a fictitious cube of side @xmath79 , to a continuous set of modes for @xmath80 .
thus , in the following , the @xmath81-sum and the kronecker function @xmath82 in eq.s ( [ stim6 ] ) and ( [ stim7 ] ) are substituted respectively with the integral over @xmath81 and the dirac function @xmath83 .
the fact that the second - order correlation function has an unphysical spatial behaviour , dominated by a delta function , comes from the starting assumption of a plane - wave pump , having infinite transverse dimension . in a realistic case the delta function is replaced by the fourier transform of the pump transverse profile .
the quantum mean values of the photon fluxes @xmath84 reaching detectors 1 and 2 can be obtained integrating equations ( [ stim3 ] ) and ( [ stim4 ] ) respectively on @xmath77 and @xmath78 . by applying relations ( [ unitary cond ] )
we have @xmath85 ) and ( [ stim6 ] ) over @xmath86 and @xmath87 respectively , we have the following auto - correlation functions for the photon fluxes @xmath88 latexmath:[\[\begin{aligned } \label{stim11 } \langle : \delta f_{2}(t ) \delta f_{2}(t'):\rangle&\approx&\sum_{\omega}e^{-i(\omega_{0}-\omega)\;(t - t')}\nonumber\\ & & \int_{r_{2}}d\mathbf{x}\:|{\widetilde{v}}_{2}(\mathbf{x},\omega)|^{2 } here and in the following , we have assumed that in the limit of small @xmath1 the gain functions @xmath90 , according to the relations ( [ unitary cond ] ) .
then , integrating eq .
( [ stim7 ] ) over @xmath91 gives the following cross - correlation : latexmath:[\[\begin{aligned } \label{stim12 } \langle : \delta
f_{1}(t ) \delta f_{2}(t ' ) : \rangle\approx\sum_{\omega}e^{-i(\omega_{0}+\omega)\;(t - t ' ) } \int_{r_{1}}d\mathbf{x}\nonumber\\ although the seed beam is monochromatic with frequency @xmath65 , the temporal width of both auto - correlation and cross - correlation in a fixed point @xmath93 still depends on the spectral bandwidth @xmath54 of the gain functions .
this explains why they originate from the correlation between the monochromatic seed and the spontaneous emission that is broadband .
therefore , the coherence time of the stimulated emission is still of the order of @xmath94 , analogously to the spontaneous case .
the auto - correlation of the current fluctuations @xmath95 can be calculated by introducing the result in eq .
( [ stim10 ] ) in ( [ i_ji_knew ] ) , obtaining @xmath96 under the condition of small gain ( @xmath97 ) and large intensity of the seed ( @xmath98 ) the fluctuation of the current at detector @xmath75 is dominated by the shot noise component . by substituting eq ( [ stim12 ] ) in ( [ i_ji_knew ] ) , under the same condition , for @xmath99 and @xmath100
, one has @xmath101 where we assumed @xmath102 .
we stress that it is the usual situation , the coherence time of spdc being on the order of picoseconds or less and the typical resolving time of detectors on the order of nanosecond . in this case
any fluctuations in the light power are averaged over @xmath5 .
the cross - correlation function of the fluctuations has the same form as the one obtained in the case of spontaneous down - conversion in our earlier work @xcite .
this is the evidence that quantum correlations do not disappear when the emission is stimulated .
in fact , the down - converted photons are still produced in pairs , although in the beam 2 the photons of the pairs are added to the bright original coherent beam propagating in the same direction .
the factor 2 , appearing in eq .
( [ stimul - cross - corr ] ) can be interpreted as due to the fact that , for any down - converted photon of a pair propagating along direction 2 , there is also the original photon of the seed that stimulated the generation of that pair .
formally it emerges from the time integration of two identical contribution in ( [ stim12 ] ) , the explicit one and its complex conjugated .
since we are interested in a relatively large power of incident light , we can at first consider detectors without internal gain , and assume that the charge produced in any detection event is equal to the single electron charge @xmath103 , i.e. @xmath104 and @xmath105 .
therefore , according to eq.s ( [ stimul - self - corr ] ) and ( [ stimul - cross - corr ] ) , the quantum efficiency can be evaluated as @xmath106 for detectors in which electrons multiplication occurs the statistic fluctuations of this process do not allow to use eq .
( [ eta - stim ] ) for absolute calibration .
however , it can be performed by integrating eq .
( [ stimul - cross - corr ] ) over time @xmath107 @xcite .
it corresponds to evaluating the power spectrum of the fluctuations at frequencies around zero , namely much smaller than @xmath108 .
we would like to stress that in this case , the assumption @xmath16 is not necessary . since @xmath109
we obtain @xmath110 in eq .
( [ eta - stim2 ] ) the statistic of the electron gain does not play any role .
with the purpose to eliminate some problems that appear when working with spontaneous pdc @xcite , we have extended our method to higher flux regimes , taking advantage of the stimulated pdc as a source of bright correlated beams .
we show that this scheme allows one to overcome these former problem : an important result in view of metrology applications .
indeed this result allows to link the photon counting regime with the photon rates of the traditional optical metrology supplying a single radiometric standard for all these regimes .
finally , we would like to stress once more that equations [ eta - stim ] and [ eta - stim2 ] , between quantum efficiency and current fluctuations , even if identical to the ones obtained in our previous work @xcite , are here derived for a more general condition of stimulated emission .
the turin group acknowledges the support of regione piemonte ( ricerca scientifica applicata e14 ) and san paolo foundation .
the russian group acknowledges the support of russian foundation for basic research ( grant # 06 - 02 - 16393 ) .
both acknowledge the joint project of associazione sviluppo del piemonte by grant rfbr - piedmont # 07 - 02 - 91581-aps . | spontaneous parametric down conversion ( spdc ) has been largely exploited as a tool for absolute calibration of photon - counting detectors , i.e detectors registering very small photon fluxes . in @xcite
we derived a method for absolute calibration of analog detectors using spdc emission at higher photon fluxes , where the beam is seen as a continuum by the detector .
nevertheless intrinsic limitations appear when high - gain regime of spdc is required to reach even larger photon fluxes . here
we show that stimulated parametric down conversion allow one to avoid this limitation , since stimulated photon fluxes are increased by the presence of the seed beam
. 99 h.a .
bachor , _ a guide to experimental quantum optics _ , ( wiley - vch , new york , 1998 ) .
m. genovese , `` research on hidden variable theories : a review on recent progresses , '' phys . rep . * 413*/6 , 319 - 398 ( 2005 ) and references therein .
see , for a general review , n. gisin , g. ribordy , w. tittel and h. zbinden , quantum cryptography , rev .
mod .
phys .
* 74 * , 145 - 195 ( 2000 ) , and references therein .
giorgio brida , maria chekhova , marco genovese , alexander penin , ivano ruo - berchera , `` the possibility of absolute calibration of analog detectors by using parametric down - conversion : a systematic study , '' j. opt
. soc . am .
b * 23 * , 2185 - 2193 ( 2006 ) .
b.y . zeldovich and d.n .
klyshko , `` statistics of field in parametric luminescence , '' sov . phys .
jetp lett .
* 9 * 40 - 44 ( 1969 ) .
d.c .
burnham and d.l .
weinberg , `` observation of simultaneity in parametric production of optical photon pairs , '' phys .
rev .
lett . * 25 * , 84 - 87 ( 1970 ) .
d.n .
klyshko , `` use of two - photon light for absolute calibration of photoelectric detectors , '' sov .
j. quant .
elect .
* 10 * , 1112 - 1116 ( 1980 ) .
a. a. malygin , a. n. penin and a. v. sergienko , `` absolute calibration of the sensitivity of photodetectors using a two - photon field , '' sov .
phys .
jetp lett .
* 33 * , 477 - 480 ( 1981 ) .
d.n .
klyshko and a.n .
penin , `` the prospects of quantum photometry , '' phys .
uspekhi , * 30*(8 ) , 716 - 723 ( 1987 ) .
a. migdall , `` correlated - photon metrology without absolute standards , '' physics today january , 41 - 46 ( 1999 ) .
g. brida , m. genovese and c. novero , `` an application of two photons entangled states to quantum metrology , '' j. mod . opt . * 47 * 2099 - 2104 ( 2000 ) .
v. m. ginzburg , n. g. keratishvili , e. l. korzhenevich , `` absolute meter of photodetector quantum efficiency based on the parametric down - conversion effect , '' g. v. lunev , a. n. penin , v. i. sapritsky , opt . eng . * 32*(11 ) , 2911 - 2916 ( 1993 ) . g. brida , m. v. chekhova , m. genovese , m. gramegna , l. a. krivitsky , and s. p. kulik , `` conditioned unitary transformation on biphotons , '' phys . rev .
a * 70 * , 032332 ( 2004 ) .
g. brida , m. chekhova , m. genovese , m.gramegna , l. krivitsky , m.l .
rastello , `` single - photon detectors calibration by means of conditional polarization rotation , '' j. opt .
soc .
am .
b * 22 * , 488 - 492 ( 2005 ) .
g.brida , m. chekhova , m. genovese , m.gramegna , l. krivitsky , m.l .
rastello , `` absolute quantum efficiency measurements by means of conditioned polarization rotation , '' ieee trans instrum . meas . * 54 * no . 2 , 898 - 900 ( 2005 ) .
g. brida et al . , j. mod .
opt . to appear .
k. soda , i. nishio , and a. wada , `` analysis of noises in photocurrent time - correlation spectroscopy of scattered light , '' j. appl .
phys . * 47 * , 729 - 735 ( 1976 ) .
r. g. mcintyre , `` multiplication noise in uniform avalanche diodes '' , ieee trans .
electron .
devices , * ed-13 * , 164 - 168 ( 1966 ) .
the condition is fulfilled if the electronics is the same for the two detector and has a pass band much smaller than the band of the detectors .
d. n. klyshko , a. v. masalov , `` photon noise : observation , squeezing , interpretation , '' phys .
uspekhi , * 38*(11 ) , 1203 - 1230 ( 1995 ) . |
elemental cerium is stark example of the complicating effects of magnetism in metals . in principle
a simple system , cerium at room temperature and ambient pressure presents a monatomic bravais lattice ( the face - centered - cubic @xmath3 phase ) .
however , under the application of modest pressures , cerium undergoes a spectacular first - order transition to a low temperature @xmath4 phase .
both @xmath4 and @xmath3 are isostructural f.c.c
. phases , but differ by @xmath5 in volume , and have markedly different magnetic susceptibilities and resistivities.@xcite there are competing explanations for the transition , which variously invoke a `` kondo volume collapse''@xcite mechanism , a `` mott transition'',@xcite or some entropic mechanism@xcite . in all cases however , it is generally agreed that the large volume @xmath3 phase contains relatively larger localized magnetic moments ( evinced by the higher susceptibility ) and relatively less itinerant electrons ( evinced by the higher resistivity ) . beyond that
, the nature of the transisiton remains contentious .
one of the main complications in the study of cerium is that although the @xmath2 transition occurs on cooling over a wide range of pressures , it is avoided at ambient pressure . in that case ,
an intervening dhcp phase forms from the @xmath3 phase , and transforms into the @xmath4 phase only at lower temperature.@xcite there are commonly two methods used to resolve this problem .
measurements are either performed under pressure on pure cerium , or else the cerium is alloyed with other elements known to suppress the dhcp phase ( typically thorium ) .
results from these two avenues of investigation are not always consistent - for example , it has been claimed that the phonon entropy plays little role in the @xmath2 transition in a ce@xmath6th@xmath1 alloy,@xcite while studies on pure cerium under pressure have returned the opposite result.@xcite nevertheless , the alloying of cerium has generated a family of related compounds , in which the order of the transition , the magnitude of the volume collapse , and the transition temperature can be smoothly varied.@xcite one such alloy , ce@xmath0la@xmath1th@xmath1 , is of particular interest in regards to spin - lattice coupling at the phase transition .
alloying with thorium suppresses the dhcp phase , while alloying with lanthanum causes the transition to become continuous and supresses the critical temperature.@xcite it has been reported that cooling ce@xmath0la@xmath1th@xmath1 in high magnetic fields can suppress the resistivity change associated with the volume collapse to lower temperature , and that the application of pulsed magnetic fields can induce a large moment below the critical temperature.@xcite these measurements map out a phase boundary for the @xmath2 transition as a function of magnetic field , and suggest that the volume collapse can be fully suppressed by @xmath7 tesla . here
we report a direct study of the structure of ce@xmath0la@xmath1th@xmath1 , using synchrotron x - ray powder diffraction in pulsed magnetic fields as high as 28 tesla .
lattice parameter and structural correlation length are extracted throughout the @xmath8 phase diagram .
samples of ce@xmath0la@xmath1th@xmath1 were prepared at los alamos by arc - melting elemental cerium , lanthanum and thorium in a zirconium gettered argon atmosphere . the button was melted and flipped 15 times to ensure mixing .
the final melt was cooled in a round - bottomed trough , which was 5 mm wide .
the resulting @xmath9 25 mm long sample was unidirectionally rolled @xmath9 10 times to achieve a finished thickness of 50 microns .
a section of the as - rolled material was heat treated at 450@xmath10c for 1 hour to remove the rolling texture before a final sample , 2 mm in diameter , was cut from the sheet with a razor blade .
the sample was subsequently mounted inside a blind hole in a sapphire bracket , and sealed with a sapphire cap and stycast epoxy , to ensure good thermal contact and prevent oxidization .
the sapphire bracket was anchored to the cold finger of the double - funnel solenoid pulsed magnet at the advanced photon source ( aps ) , which facilitates diffraction at up to 22@xmath11 of scattering angle , to temperatures as low as @xmath12 k and in magnetic fields as high as @xmath9 30 tesla.@xcite this system features a resistive magnet coil ( of tohoku design ) in a liquid nitrogen bath , connected to a 40 kj capacitor bank .
the bank is capable of discharging a 3 kv charge through the coil in a @xmath13 ms half - sine wave pulse , generating a peak current of @xmath14 ka and a peak magnetic field of @xmath15 tesla .
the heat generated in the resistive coil during a pulse must be allowed to dissipate before the system can be pulsed again . at peak field values
, this requires a wait time @xmath16 8 minutes between pulses .
the sample to be studied is both thermally and vibrationally isolated from the coil .
this new instrument offers a complementary measurement geometry to the aps split - pair pulsed magnet developed previously.@xcite 14.9 tesla , integrated over @xmath17 0.5 tesla for the exposure .
middle : raw data collected on the image plate for a single 1 ms exposure .
bottom : the same data , radially integrated and converted to @xmath18.,width=306 ] experiments were performed at the 6id - b station at the aps . 30 kev photons were selected using a si(111 ) monochromator , with the mirror removed to provide an unfocused beam .
the monochromator was detuned so as to reduce higher harmonic contamination .
the incident beam size was defined by slits , with an illuminated area on sample of 0.2 mm x 1.0 mm .
a @xmath19 ms pulse of x - rays was defined using a pair of fast platinum shutters , and the arrival of this pulse at the sample was synchronized with the peak of the magnetic field pulse .
scattered x - rays were recorded on a two - dimensional image plate detector . in this configuration ,
our system is similar to the pulsed magnet instrument in operation at the european synchrotron radiation facility ( esrf).@xcite to calibrate our instrument , we repeated the measurements@xcite of detlefs _ et al .
_ on polycrystalline tbvo@xmath20 in pulsed magnetic fields , as previously measured by the esrf instrument.@xcite we reproduced the reported high field peak splitting in the tetragonal phase of tbvo@xmath20,@xcite validating our measurement scheme.@xcite we performed our measurements of ce@xmath0la@xmath1th@xmath1 with the axis of the solenoid coincident with the incident beam , and the detector positioned so as to capture the majority of the solenoid exit window . in this way , at 30 kev the first two powder rings ( corresponding to the @xmath21111@xmath22 and @xmath21200@xmath22 peaks ) are completely captured by the detector , allowing a full radial integration and excellent counting statistics even for a single 1 ms exposure . the current through the solenoid and the x - ray flux incident on the sample were monitored as a function of time by tracing the output of a high - resolution current monitor and an air - filled ion chamber on a digital storage oscilloscope .
a typical trace is shown in fig .
1 , along with data from a single exposure on the image plate . in order to extract the correlation length and lattice parameter from these diffraction patterns ,
each of the two lowest lying peaks was fit individually , generating two independent sets of data , which were subsequently compared . in all cases , the properties extracted from the two peaks aggreed within the errorbars of the measurement , and the quantities quoted in the following are the average of the two
powder diffraction patterns at room temperature show a single phase material , with symmetry and lattice constant consistent with the @xmath3 phase of ce@xmath0la@xmath1th@xmath1 . on cooling , we observe the first two bragg peaks to shift to higher @xmath18 , while also broadening by a factor of @xmath23 and exhibiting a reduction in peak intensity ( see fig .
this is consistent with a collapse in volume , accompanied by a striking reduction in the structural correlation length .
it is interesting to note that we observe a single phase at all temperatures , in contrast to recent neutron diffraction measurements on ce@xmath6th@xmath1 which reported a coexistence of both the @xmath4 and @xmath3 phases over a wide range of temperatures.@xcite our sample was initially cooled directly to 8k to characterize the magnitude of the volume collapse , then warmed to 250 k while collecting data .
the volume was observed to approach the @xmath3 phase value quite gradually .
next we cooled slowly and measured the lattice parameters with a fine point density .
the resulting continuous but hysteretic volume collapse curves are plotted in fig .
surprisingly , on our second cooling we found that the absolute size of the volume collpase was diminished - thermal cycling of the sample is detrimental to the phase transition .
111@xmath22 bragg reflection .
the peak shifts to higher @xmath18 and broadens on cooling .
the sample is a single phase at all temperatures .
resolution limited peaks at high temperature are well described by gaussians , while a voigt shape at low temperature is consistent with an ornstein - zernike scattering function with a short correlation length ( solid lines).,width=321 ] la@xmath1th@xmath1 .
sample was warmed from 8 k to 250 k ( @xmath24 ) , then cooled back to 10 k ( @xmath25 ) . this thermal cycling lead to a reduction in the magnitude of the volume collapse , shown by the dashed lines .
the inset shows the ornstein - zernike correlation length extracted from the cooling measurements ( see text ) .
the correlated regions of the collapsed phase have average size of only @xmath26 nm.,width=321 ] the observed peak broadening can be modelled phenomenologically , by assuming that the high temperature patterns show long - range ordered and resolution limited scattering , while the low temperature broadening arises from a reduction in the ornstein - zernike correlation length .
our high temperature peaks are well described by gaussians , and we are assuming a lorentzian form for the scattering function ( see ref . [ ] for details ) .
therefore , the measured scattering at low temperature should be well described by the convolution of a lorentzian and a gaussian - a voigt function .
since the width of this voigt function is related to the width of the gaussian ( the resolution ) and the width of the lorentzian ( the inverse correlation length ) , the numerically extracted fullwidth of the bragg peaks can be converted to a resolution - deconvoluted correlation length .
the solid lines in fig .
2 demontrate the degree of agreement between these lineshapes and the measured data , and the extracted correlation length is plotted in the inset of fig .
our resolution limit is found to be on the order of a few microns , however , the low temperature collapsed phase exhibits correlation lengths which are roughly two orders of magnitude shorter than this .
below @xmath9 30 k , the @xmath4 phase correlations are seen to extend over an average range of @xmath9 20 nm .
since our alloy is @xmath9 80@xmath27 cerium , the mean distance between dopant atoms ( either lanthanum or thorium ) is on the order of 2 nm .
thus , while the inclusion of lanthanum and thorium in the alloy does not completely hinder the formation of the @xmath4 phase , and each correlated region on average contains multiple dopant atoms , nevertheless these atoms act to strongly disorder the sample at the nanoscale within the @xmath4 phase .
conversely , within the @xmath3 phase , individual grains are well correlated over micron - sized regions despite the same level of chemical disorder .
motivated by the results of ref .
[ ] , we next sought to characterize the effect of applied magnetic field on the volume collapse transition in ce@xmath0la@xmath1th@xmath1 . currently , the available capacitance at the aps limits our measurements to applied magnetic fields @xmath28 30 tesla , well below the 56 tesla estimated to fully suppress the @xmath2 transition to zero kelvin.@xcite however , inspection of the resistivity measurements of ref .
[ ] reveals that our highest achievable magnetic fields should be sufficient to suppress the temperature of the volume collapse inflection point by @xmath9 8 k. we can maximize our sensitivity to this effect by cooling to the inflection point of the volume collapse ( @xmath9 35 k ) and pulsing to high magnetic field .
the slope of @xmath29 is large at this temperature , and therefore an 8 k shift in the inflection point should cause a @xmath9 0.7 @xmath27 change in volume at 35 k , well within our resolution .
however , as can be seen from fig .
4 , we failed to observe this effect .
after cooling to 35 k , we collected five diffraction patterns at zero field , which served to define the zero field bragg peak positions to an accuracy better than @xmath30 . next , we collected ten patterns in pulsed field , with the average applied field over the x - ray exposure being @xmath31 tesla .
we observed no measurable change in the bragg peak positions in any individual pattern . nor did we observe any cumulative effect from multiple high - field pulses .
our measurements therefore constrain any volume changes at ( 35 k , 28 t ) to be less than @xmath9 0.07 @xmath27 , a full order of magnitude smaller than what would be expected.@xcite 111@xmath22 bragg peak at t=35 k. there is no discernable change in the position or width of the peak .
the expected curve is calculated based on the position and width change that would arise from an 8 k shift in the transition temperature , consistent with the results of ref .
[ ] .,width=321 ] ) and ornstein - zernike correlation length ( @xmath32 ) , comparing zero field cooled measurements and measurements where the system is repeatedly pulsed to 28 t on cooling .
no field effect is seen.,width=321 ] it is important to remember that the evidence for field induced suppression of the volume collapse in ce@xmath0la@xmath1th@xmath1 for magnetic fields below 30 tesla comes from resistivity curves , measured by cooling in constant applied magnetic fields .
this is procedurally distinct from the zero - field - cooled pulsed measurements reported here .
discrepancy between field - cooled and zero - field - cooled behaviour is a hallmark of spin glass materials , wherein chemical disorder and competing interactions collude to freeze in disordered ground states.@xcite we have already shown that chemical disorder in our alloy has a strong effect on the development of @xmath4-phase correlations ( see fig .
it is therefore tempting to sugest that there may be some glassy behavior at play in ce@xmath0la@xmath1th@xmath1 .
direct comparison between field cooled and zero field cooled measurements are therefore highly desirable , however , a truly field - cooled measurement is not achievable with pulsed magnets . we have attempted to approximate one by slow cooling , and pulsing in regular intervals . in order to minimize the obscuring effects of thermal cycling , we performed two identical cooling curves , one after the other , warming to 280 k once to reset the system .
for each of these measurements , we cooled at a controlled rate of 1 k@xmath33min from 280 k. during the first of these cooldowns , we pulsed to 28 t at 10 min ( 10 k ) intervals . during the second cooldown
, we collected zero - field data in the same manner , albeit more frequently since there was no need to wait for the coil to cool down .
the results are plotted in fig .
5 . clearly
, magnetic fields applied in this way are no more effective at suppressing the volume collapse than the zero - field - cooled methods shown in fig .
4 . we are therefore forced to conclude that the available magnetic fields were insufficient to induce any structural effects in ce@xmath0la@xmath1th@xmath1 . if the @xmath2 transition were continuous for this alloy ( as claimed in ref . [ ] ) , and the magnetic field induced suppression occured as claimed in ref . [ ] , then we would have expected to see a clear shift in @xmath18 as denoted in fig .
4 for the zero - field - cooled measurements , and a clear suppression in temperature should have occurred for the repetitively pulsed measurements reported in fig .
conversely , if the @xmath2 transition were first order for this alloy ( as claimed in ref .
[ ] ) , then our measurements imply that the phase boundary was never crossed in our cooling curve . this would require a steeper increase of the critical field on lowering temperature than was reported in ref . [ .
] it is also possible that the alloy is dynamically inhibited on the timescale of the magnetic field pulse , which would be consistent with a glassy behavior . finally , it is possible that the effects reported in ref .
[ ] were not intrinsic to ce@xmath0la@xmath1th@xmath1 .
clearly , the issue of field induced suppression of the @xmath2 transition requires further investigation in elemental cerium and its alloys .
as has been noted in the preceeding sections , we have observed a detrimental effect due to thermal cycling . our initial measurements on as - grown samples revealed well - correlated grains on the order of microns in size .
on first cooling , the volume collapsed @xmath4 phase was seen to exhibit nanoscale disorder , and on second cooling the absolute size of the volume collapse had diminished ( see fig . 3 and discussion in section iii ) .
over the course of our week - long experiment , we observed the @xmath4-phase volume to gradually increase with each thermal cycling , while the @xmath3-phase volume was seen to gradually decrease .
in addition , while the @xmath4-phase consistently showed nanoscale disorder with correlation lengths of @xmath9 20 nm , the @xmath3 phase correlations were observed to shrink , from micron - sized grains in the as - grown sample to correlation lengths of only @xmath9 150 nm after 11 thermal cycles .
zero - field cooling curves from the second , seventh , and eleventh cycle are plotted in fig .
6 , to illustrate the effect .
it is likely that the strains associated with the large volume change and spatial variations in doping,@xcite which are manifest in the @xmath4-phase disorder , act to introduce new domain boundaries at low temperature , which thereafter persist as defects .
this effect is insidious , since the suppression by thermal cycling we observe here could easily be mistaken for suppression by an external perturbation , if that perturbation was varied monotonically during the course of an experiment .
hypothetically , had we measured cooling curves through the volume collapse in dc magnetic fields which were monotomically increased during the course of our measurements , we may easily have misinterpreted the gradual buildup of thermal cycling damage as a magnetic field induced effect .
therefore , it is important to take extra care when mapping out phase diagrams in cerium alloys .
more generally , these measurements highlight the need for careful consideration of strain driven effects in all materials studies , which can often obfuscate results . transition .
the effects are clear in both the magnitude of the volume collapse ( @xmath34 ) and the ornstein - zernike correlation length ( @xmath32 ) .
all data was collected on cooling in zero magnetic field .
the three data sets represent the second ( @xmath25 ) , seventh ( @xmath24 ) , and eleventh ( @xmath35 ) thermal cycle.,width=321 ]
in this article , we have reported a synchrotron x - ray diffraction study of the volume collapse transition in ce@xmath0la@xmath1th@xmath1 , as a function of temperature and pulsed magnetic field . in this alloy ,
the @xmath2 transition is smooth but hysteretic , characterized by a @xmath36 volume change and a dramatic reduction in structural correlation length .
the collapsed @xmath4 phase is shown to be unaffected by pulsed magnetic fields @xmath28 28 tesla , both in zero - field cooled conditions and under repetitive pulsing on cooling .
it is suggested that a deviation between truly field cooled and zero field cooled behavior in ce@xmath0la@xmath1th@xmath1 may account for the disagreement between our measurements and previous bulk studies.@xcite this implies a glassy character to the low temperature phase .
we have also shown that repeated thermal cycling through the @xmath2 transition acts to disorder the sample on nanometer length scales , while suppressing the magnitude of the volume collapse transition .
these measurements may constitute a microscopic measure of what has been generally refered to as `` sluggish dynamics '' in cerium alloys.@xcite it is our hope that these results will advance understanding of the exotic @xmath2 transition in elemental cerium and its alloys .
specifically , we believe the data presented here highlights the strong effects of disorder in these systems .
we gratefully acknowledge fruitful discussions with j. lashley and b. toby .
use of the advanced photon source is supported by the doe , office of science , under contract no .
de - ac02 - 06ch11357 .
pulsed magnet collaborations between argonne and tohoku university are supported by the icc - imr .
hn acknowledges kakenhi no .
23224009 from mext .
jpcr aknowledges the support of nserc of canada .
f. drymiotis , j. singleton , n. harrison , j.c .
lashley , a. bangura , c.h .
mielke , l. balicas , z. fisk , a. migliori and j.l .
`` suppression of the @xmath37 structural phase transition in ce@xmath0la@xmath1th@xmath1 by large magnetic fields ''
. condens . matter _ * 17 * l77-l83 ( 2005 ) m.e .
manley , r.j .
mcqueeney , b. fultz , t. swan - wood , o. delaire , e.a .
goremychkin , j.c .
cooley , w.l .
hults , j.c .
lashley , r. osborn , and j.l .
`` no role for phonon entropy in the fcc - fcc volume collapse transition in ce@xmath6th@xmath1 at ambient pressure '' _ phys . rev .
_ b * 67 * , 014103 ( 2003 ) i .- k . jeong , t.w . darling , r.h .
heffner , a. llobet and j.c .
. `` neutron diffraction study on the @xmath37 phase transition in ce@xmath6th@xmath1 alloy '' _ philosophical magazine letters _ vol * 88 * , no . * 8 * , 607 - 613 ( 2008 ) i .- k .
jeong , t.w .
darling , m.j .
graf , th .
proffen , r.h .
heffner , y. lee , t. vogt and j.d .
`` role of the lattice in the @xmath37 phase transition of ce : a high - pressure neutron and x - ray diffraction study '' _ phys .
lett . _ * 92 * , 105702 ( 2004 ) m. krisch , d.l .
farber , r. xu , d. antonangeli , c.m .
aracne , a. beraud , t.c .
chiang , j. zarestky , d.y .
kim , e.i .
isaev , r. ahuja , and b. johansson .
`` phonons of the anomalous element cerium . ''
_ pnas _ vol . * 108 * no .
23 , 9342 - 9345 , ( 2011 ) j.c .
lashley , a.c .
lawson , j.c .
cooley , b. mihaila , c.p .
opeil , l. pham , w.l .
hults , j.l .
smith , g.m .
schmiedeshoff , f.r .
drymiotis , g. chapline , s. basu and p.s .
tricritical phenomena at the @xmath37 transition in ce@xmath38la@xmath39th@xmath1 alloys " _ phys .
* 97 * , 235701 ( 2006 ) zahirul islam , dana capatina , jacob p.c .
ruff , ritesh k. das , emil trakhtenberg , hiroyuki nojiri , yasuo narumi and p.c . canfield .
`` a high - field pulsed magnet system for x - ray scattering studies in voigt geometry . '' in preparation for _ rev .
instrum . _
( 2011 ) zahirul islam , jacob p. c. ruff , hiroyuki nojiri , yasuhiro h. matsuda , kathryn a. ross , bruce d. gaulin , zhe qu , and jonathan c. lang .
`` a portable high - field pulsed - magnet system for single - crystal x - ray scattering studies '' _ rev .
instrum . _ * 80 * , 113902 ( 2009 ) j.p.c .
ruff , z. islam , j.p .
clancy , k.a .
ross , h. nojiri , y.h .
matsuda , h.a .
dabkowska , a.d .
dabkowski , and b.d .
`` magnetoelastics of a spin liquid : x - ray diffraction studies of tb@xmath40ti@xmath40o@xmath41 in pulsed magnetic fields '' _ phys .
_ * 105 * , 077203 ( 2010 ) p. frings , j. vanacken , c. detlefs , f. duc , j.e .
lorenzo , m. nardone , j. billette , a. zitouni , w. bras , and g.l.j.a .
`` synchrotron x - ray powder diffraction studies in pulsed magnetic fields . '' _ rev .
instrum . _ * 77 * , 063903 ( 2006 ) c. detlefs , f. duc , z.a .
kazei , j. vanacken , p. frings , w. bras , j.e .
lorenzo , p.c .
canfield , and g.l.j.a .
`` direct observation of the high magnetic field effect on the jahn - teller state in tbvo@xmath20 '' _ phys .
lett . _ * 100 * , 056405 ( 2008 ) j.p.c .
ruff , b.d .
gaulin , j.p .
castellan , k.c .
rule , j.p .
clancy , j. rodriguez , and h.a .
`` structural fluctuations in the spin - liquid state of tb@xmath40ti@xmath40o@xmath41 '' _ phys . rev
_ * 99 * , 237202 ( 2007 ) | we report synchrotron x - ray powder diffraction measurements of ce@xmath0la@xmath1th@xmath1 subject to pulsed magnetic fields as high as 28 tesla .
this alloy is known to exhibit a continuous volume collapse on cooling at ambient pressure , which is a modification of the @xmath2 transition in elemental cerium .
recently , it has been suggested on the basis of field - cooled resistivity and pulsed field magnetization measurements that the volume collapse in this alloy can be suppressed by the application of magnetic fields .
conversely , our direct diffraction measurements show a robust collapsed phase , which persists in magnetic fields as high as 28 tesla .
we also observe nanoscale disorder in the collapsed phase , which increasingly contaminates the high temperature phase on thermal cycling . |
parameters for our calculations have been obtained from a first - principle approach , as we detail in the following . as an initial step ,
first - principle density functional theory ( dft ) calculations were undertaken for a single substitutional boron impurity in two - dimensional ( 2d ) graphene and in gnrs , @xcite by means of the siesta code , @xcite within the local density approximation and using a double-@xmath4 basis set .
following the methodology in refs .
@xcite , the onsite energies of the @xmath5 orbitals for the ( bulk ) 2d case were extracted . their variation with the distance from the boron atom
is shown in fig .
[ fig2 ] ( inset ) .
the boron atom acts as an electron acceptor : the resulting impurity potential is repulsive for electrons and the screening length is found to be of the order of a few angstroms ( the screening length @xcite is defined as the distance from the charged impurity at which the surrounding mobile charges , rearranging under the electrostatic action of the impurity , reduce its potential by a factor equal to euler s number @xmath6 ) .
the transport properties of `` quasi - metallic '' and semiconducting armchair gnrs ( following ref .
@xcite , we refer to an armchair gnr with @xmath7 dimers contained in its unit cell as an `` n - agnr '' ) with a single boron impurity at different sites across the ribbon width are then computed , using the full _ ab initio _
hamiltonian , in the landauer - bttiker framework .
@xcite boron impurities actually induce a strong hole backscattering , with much weaker effect on the electron side .
a similar result was already obtained in boron - doped carbon nanotubes , @xcite and attributed to strong backscattering at the resonance energies of the quasibound states localized around the boron impurity . due to the lack of the rotational symmetry present in carbon nanotubes , in gnrs
the energy of the conductance dip on the hole side depends on the position of the impurity with respect to the ribbon edge ( as illustrated in fig .
[ fig2 ] and fig .
[ fig3 ] ) . in ref .
@xcite it has been shown that these _ ab initio _ results can be fairly well reproduced with a simpler tight - binding model with a single orbital per carbon and boron atom .
for each atom , the onsite energy is taken as the impurity potential extracted from the _ ab initio _ calculation .
the onsite energy of the boron atom is the only parameter that has to be adjusted in order to achieve the best match between tight - binding and _ ab initio
_ transmission profiles .
the success of such a simple tight - binding model in describing boron impurities originates from the preservation of the @xmath8 hybridization of graphene in the presence of substitutional boron . by using this tight - binding model , transport studies in large armchair gnrs with a random distribution of boron atoms , with concentrations between @xmath9% and @xmath10% ,
have been performed in ref .
the addition of the transmission dips induced by all impurities leads to a _ mobility gap _ in the valence band , _
i.e. _ an energy interval where conductance is nearly suppressed by scattering , while transport in the conduction band is preserved ( the opposite behavior would be found in the case of nitrogen doping ) . this occurs for semiconducting as well as for `` quasi - metallic '' gnrs , independently of ribbon width , as long as the potential can be assumed to be constant along the channel .
the mobility gap increases with boron concentration and ribbon length .
@xcite suggested a possible exploitation of this phenomenon to obtain field - effect transistors ( fet ) with wide gnrs ( @xmath11 nm ) , within the reach of conventional lithography .
the absence of a significant bandgap in these ribbons could be compensated for by the mobility gap .
however , the above mentioned study has been performed for ribbons without any gate - induced charge in the channel , since the impurity potential considered was the one extracted directly from an _ ab initio _ calculation without taking into account the effects of screening when a gate is applied . and
@xmath12 directions , to make the drawing clearer ) . ] in order to simulate the behavior of a realistic transistor structure , we need to perform a self - consistent calculation to include the effects of an applied gate voltage and of a non - zero drain - source bias on the charge density .
this can be done , at a reasonable level of computational complexity , with a tight - binding model set up with parameters properly extracted from the _ ab initio _ results .
however , differently from ref .
@xcite , the impurity potential obtained from _ ab initio _ approaches can not be directly used in a self - consistent simulation , since it would be overscreened , as a result of the poisson / schrdinger iterations .
the approach chosen here is to mimic dft self - consistency with the tight - binding model and a proper distribution of fixed charges ; then the final impurity potential arises from the tight - binding poisson / schrdinger self - consistency .
in particular , as we will detail in the methods section , the _ ab initio _ results can be reproduced considering a fixed charge @xmath13 ( where @xmath6 is the elementary charge ) in correspondence of each carbon atom , and using the following tight - binding representation : null onsite energies are considered in correspondence of all atoms , while , following ref .
@xcite , the hopping parameter is equal to @xmath14 ev for all the pairs of nearest neighbor atoms , with the exception of the edge dimers , for which a value of @xmath15 is used instead . to validate this approach , a self - consistent poisson / schrdinger calculation ( described in the methods section )
is performed for an isolated graphene layer with a single impurity , as in the _ ab initio _ calculation .
as shown in fig .
[ fig2 ] ( inset ) , the obtained self - consistent potential around the impurity is very close to the _ ab initio _ result .
also the _ ab initio _ transmission spectra are accurately reproduced for all positions of the impurity , keeping the boron onsite energy equal to zero and with no further parameter adjustment .
transmission spectra comparisons are shown in fig .
[ fig2 ] . and @xmath16 , where @xmath17 is the hexagonal lattice constant .
the bold lines represent the results with our self - consistent tight - binding model .
the thin dotted line represents the _ ab initio _ results .
inset : comparison between the results for the impurity potential in 2d graphene obtained with self - consistent tight - binding and _ ab initio _ approaches . ] before performing full negf calculations , the effect of a gate with a nonzero applied voltage was analyzed on short ribbons with a single impurity .
the ribbon is surrounded by a double gate with a dielectric thickness of 1 nm , as shown in fig . [ fig1 ] .
the source - drain voltage and the fermi level are set to @xmath18 .
the charge and potential on the ribbon adjust self - consistently as a function of the gate potential @xmath19 .
first , the calculation was performed at @xmath20 with vacuum chosen as gate dielectric .
the obtained potential is very close to the result discussed above for the neutral isolated ribbon .
this gate configuration does not induce any significant additional screening .
a more realistic model is obtained by considering a silicon dioxide dielectric between the gate plates and the ribbon , with relative dielectric permittivity @xmath21 .
however , thin vacuum layers are kept just below and above the ribbon , with a thickness of @xmath22 nm , of the order of the calculated graphene / sio@xmath23 distance .
@xcite the impurity potentials and the corresponding transmission spectra are shown in fig .
[ fig3 ] , for @xmath20 and @xmath24 v. screening is enhanced at a negative gate voltage , due to the accumulation of holes in the ribbon .
this has a strong effect on transmission : the conductance dip is less pronounced and is shifted in energy , which could compromise the presence of a clear mobility gap in a transistor configuration .
it was verified that the effect of the gates is much larger than that of the thin air layers , in terms of the global electrostatics , and , in particular , of the screening .
this is the reason why in the runs for the evaluation of the transistor characteristics we will not introduce the air layers , that would make convergence of the poisson - schrdinger iterations much slower . from the edge .
the calculation was performed with silicon dioxide as gate dielectric and a thin vacuum layer surrounding the graphene sheet .
inset : influence of the gate voltage on the impurity potential . ] to assess the performance of complete boron - doped graphene transistors biased in a typical operating point , we have used the open - source code nanotcad vides .
@xcite in this code , the 3d poisson equation is solved self - consistently with the schrdinger equation ( with open - boundary conditions ) , within the non - equilibrium green s function ( negf ) formalism .
@xcite schottky contacts have been modeled at the gnr ends with the phenomenological approach described in ref .
@xcite , including self - energies that mimic metallic contacts .
the considered device structure is reported in the inset of fig .
the device is a double - gate fet and the channel is a 32-agnr ( 3.81 nm wide ) embedded in sio@xmath23 dielectric .
this is consistent with typical nanofabrication techniques , in which the dielectric is deposited after the creation of the drain and source contacts .
top and bottom field oxides are 1 nm thick , and the channel length is equal to 20 nm . in fig .
[ fig4 ] , we show the transfer characteristics of the gnr - fet for a drain - source voltage @xmath25 = 0.1 v , for different boron doping concentrations ( defined as the ratio of the number of boron atoms to the total number of atoms in the channel ) .
dopant atoms are randomly placed along the channel .
results in fig .
[ fig4 ] refer to a single distribution of dopants . ) , for a given random distribution of the dopant atoms .
the transfer characteristic for the undoped device is also reported . ]
as expected , transfer characteristics for the undoped device are symmetric with respect to @xmath26 v. as soon as some boron doping is considered , the p - branch of the @xmath27-@xmath28 curve is suppressed , as previously observed also in carbon nanotubes .
@xcite the p - branch suppression increases as the doping concentration is increased . from these results , it appears that with the considered boron doping concentrations , while still not obtaining a sufficient @xmath29 ratio for logic applications @xcite with the considered 32-agnr ( which , however , having a `` quasi - metallic '' behavior in the absence of doping , represents , in this regard , a worst - case scenario ) , a clear onset of unipolarity occurs .
increasing boron doping , while further reducing the current in the p - branch ( due to holes ) , also leads to a degradation of the device performance above threshold . in fig .
[ fig5](a)-(b ) , we show the transfer characteristics for 23 different distributions of doping atoms along the channel , for the same doping concentrations as in fig .
the thick solid lines correspond to the average transfer characteristics computed over the considered statistics of the doping distribution .
as can be seen , different doping distributions lead to a large dispersion of the transfer characteristics , which could be reduced if several nanoribbon devices were operated in parallel ( as can be necessary in some cases in order to increase the on current ) .
nevertheless , we remark that the electron - hole asymmetry is clearly developing with chemical doping , and , based on the results of ref .
@xcite , it can be strengthened by increasing the gate length up to few hundreds of nanometers ( especially for the case with @xmath30 ) . indeed
, longer channel lengths could result in cumulated contributions of multiple scattering effects and localization , which will even more inhomogeneously impact on transport characteristics ( as shown in equilibrium situations @xcite ) . due to computational limitations
, we are not reaching such channel lengths here .
computational hurdles , in terms of difficulties in achieving self - consistence , also limit the doping level that can currently be simulated with our method to about 0.6% .
thus , at present we can not handle the extremely high doping concentrations , of the order of 10% , that have been reached in a recent experiment @xcite reporting unexpectedly large gap opening effects .
in conclusion , by coupling a dft - based tight - binding parameterization ( validated _ via _ a multi - scale approach based on _ ab initio _ techniques ) with self - consistent negf quantum - mechanical transport simulations , we have found the onset of an electron / hole transport asymmetry and mobility gaps in boron - doped graphene nanoribbon transistors , confirming that chemical doping could serve as a way to shape the current - voltage characteristics of such devices .
this is just a first step , because the achieved @xmath29 ratio , although for a worst - case condition , is still far too small for practical applications .
there is however ample room for improvement and several different approaches could be used to selectively incorporate chemical entities into graphene , thus engineering separate chemically modified areas from a single graphene layer .
a doping strategy could therefore be elaborated to implement pn junctions , logic gates and complex architectures directly patterned on the same underlying graphene layer .
conventional state - of - the - art patterning methods , such as e - beam lithography , could be exploited to design superimposed material architectures ( with either metallic or other semiconducting materials ) onto the selectively doped and functionalized graphene - based substrates .
although our transistor simulations have been performed for graphene nanoribbons with a width below 5 nm ( due to the otherwise prohibitive computational cost ) , we wish to point out that the scaling analysis detailed in refs .
@xcite suggests that our main finding ( electron - hole transport asymmetry ) should be confirmed for ribbon widths above 10 nm , that is within the reach of conventional fabrication lithography techniques .
finally , longer ribbon lengths and low temperature are also factors which are capable of widening the mobility gaps .
the simulation approach is based on the observation that the dft results can be recovered performing a self - consistent poisson / schrdinger calculation , within the negf formalism , on the system , which is described using a tight - binding model and a proper distribution of fixed charges . in detail , we found that , following this procedure , the tuning of a single parameter of the tight - binding model is sufficient to properly reproduce the _ ab initio _ impurity potential and transmission spectra .
the method is inspired by what actually occurs in the dft calculation .
the total charge of a carbon nucleus and of its core electrons is @xmath31 , while it is @xmath32 for boron . in the isolated atoms ,
this charge is compensated by @xmath33 valence electrons for carbon and @xmath34 valence electrons for boron . in graphene
, @xmath34 electrons per atom ( including , in the case of doped graphene , the boron atoms ) tend to hybridize in a `` @xmath8 '' fashion , while the remaining valence electron on each carbon lies on the `` @xmath35 '' bonding subband formed by the @xmath5 orbitals .
also the @xmath5 orbital on the boron atom is not empty , since it tends to attract an electron from the neighboring carbon atoms , and this determines the shape of the impurity potential shown in fig .
[ fig2 ] . in the single @xmath5 orbital tight - binding approach , the `` @xmath8 '' valence electrons
are included in the global ionic charge of the atoms , which becomes @xmath13 for carbon and @xmath18 for boron .
therefore this is the net charge `` localized '' on each atom that we have to sum to the negative charge of the @xmath5 atomic orbitals considered in the simulation . as far as the tight - binding parameters are concerned , the onsite energy for carbon atoms is @xmath18 , while the hopping energy between nearest neighbors is set to @xmath36 ev .
the hopping energy of edge dimers is adjusted to @xmath37 in order to reproduce the small bandgap obtained in dft for `` quasi - metallic '' n - agnrs ( @xmath38 , where @xmath39 is an integer ) .
@xcite only the onsite energy of boron may be further adjusted , performing a self - consistent poisson / schrdinger calculation for an isolated graphene ribbon with a single boron atom and comparing the obtained results with those deriving from the _ ab initio _ simulation of the same structure . for the poisson equation
the same boundary conditions are used in the _ ab initio _ and in the tight - binding calculations : neumann boundary conditions ( _ i.e. _ derivative of the potential energy , and thus electric field , constant and , in particular , zero ) in the longitudinal direction , at the ends of the nanoribbon , and periodic in the transverse direction , with a separation of 2 nm between the adjacent edges of nanoribbon replicas ( indeed , with a large enough separation between the replicas this is equivalent to a neumann boundary condition ) . the gate stack has been treated considering a constant permittivity , the one of silicon oxide , between the nanoribbon and each gate , where a dirichlet boundary condition is assumed . at each iteration ,
the eigenvalues of the hamiltonian are computed and the fermi level is adjusted to obtain charge neutrality . the local charge density is then calculated and the poisson equation is solved .
the obtained electrostatic potential is subtracted from the onsite energies and the process is repeated until self - consistency is achieved . for all of the meshes tested in this work , it was possible to fit the _ ab initio _ data simply by adjusting the boron onsite energy .
a typical choice was to spread the charge of each atom in a spatial region of half - width @xmath40 , the interatomic distance , and to choose a mesh interval of 0.1 nm , small enough to achieve convergence of the poisson solution .
this charge repartition around each atom was defined using a smooth envelope function . in this case ,
the best fitting was obtained for a value of the boron onsite energy of about zero , thus substantially coincident with that for carbon .
b.b . acknowledges financial support from the juan de la cierva program and the fis2008 - 05805 contract of the spanish micinn .
acknowledges the fondation nanosciences _ via _ the rtra dispograph project .
we acknowledge the use of the software vmd @xcite for graphical design .
this work was partly funded by the european union under contract no .
215752 grand ( graphene - based nanoelectronic devices ) , by the french national research agency ( anr ) , in the framework of its 2009 program in nanosciences , nanotechnologies & nanosystems ( p3n2009 ) , through the nanosim - graphene project n@xmath41 anr-09-nano-016 - 01 , and by the italian ministry for the university and research ( miur ) through the granfet project ( n. 2008s2clj9 ) . part of the calculations were performed with the support of the genci ( grand equipment national de calcul intensif ) initiative .
iannaccone , g. ; fiori g. ; macucci m. ; michetti p. ; cheli m. ; betti a. ; marconcini p. perspectives of graphene nanoelectronics : probing technological options with modeling . in _ proceedings of iedm 2009 _ , baltimora , usa , december 7 - 9 , 2009 ; ieee conference proceedings : 2009 ; pp . 245248 .
marconcini , p. ; macucci , m. the @xmath42 method and its application to graphene , carbon nanotubes and graphene nanoribbons : the dirac equation .
nuovo cimento soc .
fis . _ * 2011 * , _ 34 _ , 489584 .
see also http://brahms.iet.unipi.it/supplem/reviewgraph.html ( accessed may 28 , 2012 ) .
cai , j. ; ruffieux , p. ; jaafar , r. , bieri , m. ; braun , t. ; blankenburg , s. ; muoth , m. ; seitsonen , a. p. ; saleh , m. ; feng , x. ; _ et al .
_ atomically precise bottom - up fabrication of graphene nanoribbons .
_ nature _ * 2010 * , _ 466 _ , 470473 .
moser , j. ; tao , h. ; roche , s. ; alzina , f. ; sotomayor torres , c. m. ; bachtold , a. magnetotransport in disordered graphene exposed to ozone : from weak to strong localization .
b _ * 2010 * , _ 81 _ , 205445 . leconte , n. ; moser , j. ; ordejn , p. ; tao , h. ; lherbier , a. ; bachtold , a. ; alsina , f. ; sotomayor torres , c. m. ; charlier , j. c. ; roche , s. damaging graphene with ozone treatment : a chemically tunable metal - insulator transition .
_ acs nano _ * 2010 * , _ 4 _ , 40334038 .
leconte , n. ; lherbier , a. ; varchon , f. ; ordejon , p. ; roche , s. ; charlier , j .- c .
quantum transport in chemically modified two - dimensional graphene : from minimal conductivity to anderson localization . _ phys .
b _ * 2011 * , _ 84 _ , 235420 . bruzzone s. ; fiori g. _ ab - initio _ simulations of deformation potentials and electron mobility in chemically modified graphene and two - dimensional hexagonal boron - nitride . _ appl .
lett . _ * 2011 * , _ 99 _ , 222108 .
lherbier , a. ; biel , b. ; niquet , y .- m . ; roche , s. transport length scales in disordered graphene - based materials : strong localization regimes and dimensionality effects . _
lett . _ * 2008 * , _ 100 _ , 036803 .
zhao , l. ; he , r. ; rim , k. t. ; schiros , t. ; kim , k. s. ; zhou , h. ; gutirrez , c. ; chockalingam , s. p. ; arguello , c. j. ; plov , l. ; _ et al .
_ visualizing individual nitrogen dopants in monolayer graphene
. _ science _ * 2011 * , _ 333 _ , 9991003 .
usachov , d. ; vilkov , o. ; grneis , a. ; haberer , d. ; fedorov , a. ; adamchuk , v. k. ; preobrajenski , a. b. ; dudin , p. ; barinov , a. ; oehzelt , m. ; _ et al .
_ nitrogen - doped graphene : efficient growth , structure , and electronic properties .
_ nano lett . _ * 2011 * , _ 11 _ , 54015407 . soler , j. m. ; artacho , e. ; gale , j. d. ; garca , a. ; junquera , j. ; ordejn , p. ; snchez - portal , d. the siesta method for _ ab initio _ order-@xmath7 materials simulation .
_ j. phys . : condens
. matter _ * 2002 * , _ 14 _ , 27452779 .
stern , f. elementary theory of the optical properties of solids . in
_ solid state physics .
advances in research and applications _ ; seitz f. , turnbull d. , eds . ; academic press : new york , 1963 ; vol . 15 , pp 299408 .
avriller , r. ; roche , s. ; triozon , f. ; blase , x. ; latil , s. low - dimensional quantum transport properties of chemically - disordered carbon nanotubes : from weak to strong localization regimes . _
b _ * 2007 * , _ 21 _ , 19551982 .
fiori , g. ; iannaccone , g. simulation of graphene nanoribbon field - effect transistors .
_ ieee electron device lett . _ * 2007 * , _ 28 _ , 760762 .
code is available at `` nanotcad vides , '' doi : 10254/nanohub - r5116.5 .
( accessed may 28 , 2012 ) .
tang , y .- b . ; yin , l .- c .
; yang , y. ; bo , x .- h . ; cao , y .- l . ; wang , h .- e . ; zhang , w .- j . ; bello , i. ; lee , s .- t . ; cheng , h .-
et al . _ tunable band gaps and p - type transport properties of boron - doped graphenes by controllable ion doping using reactive microwave plasma .
_ acs nano _ * 2012 * , _ 6 _ , 19701978 . | we report fully quantum simulations of realistic models of boron - doped graphene - based field effect transistors , including atomistic details based on dft calculations .
we show that the self - consistent solution of the three - dimensional ( 3d ) poisson and schrdinger equations with a representation in terms of a tight - binding hamiltonian manages to accurately reproduce the dft results for an isolated boron - doped graphene nanoribbon . using a 3d poisson / schrdinger solver within the non - equilibrium green s functions ( negf ) formalism ,
self - consistent calculations of the gate - screened scattering potentials induced by the boron impurities have been performed , allowing the theoretical exploration of the tunability of transistor characteristics .
the boron - doped graphene transistors are found to approach unipolar behavior as the boron concentration is increased , and by tuning the density of chemical dopants the electron - hole transport asymmetry can be finely adjusted . correspondingly , the onset of a mobility gap in the device is observed .
although the computed asymmetries are not sufficient to warrant proper device operation , our results represent an initial step in the direction of improved transfer characteristics and , in particular , the developed simulation strategy is a powerful new tool for modeling doped graphene nanostructures .
the discovery of graphene has opened a promising alternative to silicon - based electronics .
@xcite reported charge mobilities in undoped graphene layers are actually orders of magnitude larger than those measured in silicon , but the unfortunate zero - gap semiconductor nature of this material severely limits the achievable @xmath0/@xmath1 ratio @xcite ( ratio infinite for an ideal switch of the current flowing through the device in the conducting state to that flowing in the nonconducting state ) for graphene transistors , making them not yet able to compete with mainstream silicon technologies .
@xcite another adverse feature of graphene - based devices is their ambipolar electrical behavior , which precludes the development of a complementary logic architecture .
a way to induce an energy gap in graphene is to reduce its lateral dimension .
@xcite by using e - beam lithographic techniques and oxygen plasma etching processes , graphene nanoribbons ( gnrs ) down to a width of 10 nanometers can be successfully fabricated .
@xcite the resulting energy gap , however , is not sufficient for the achievement of a practically useful @xmath0/@xmath1 ratio in a transistor structure .
chemical , in - solution techniques have been developed that allow one to obtain nanoribbons with a smaller width @xcite but are hard to integrate into a large - scale device production technology .
a mixed approach has been followed by wang and dai , @xcite who use lithography to define the nanoribbons , whose width is then shrunk by means of gas - phase chemical etching . although quite interesting , this procedure could have some limitation in terms of the achievable device density .
furthermore , energy bandgaps are very unstable with regards to edge reconstruction and defects , @xcite thus preventing the fabrication of graphene transistors with performances comparable with their silicon counterparts . indeed , almost ideal graphene nanoribbons have been obtained by means of careful chemical synthesis , @xcite however such an approach
can not be applied to the fabrication of complex circuits on insulating substrates in a straightforward way .
strong chemical damage of graphene could in principle be used to tune the current flow , but with the drawback of a significant decay of the conductance that leads to poor performance in terms of current drive .
@xcite some of us have recently proposed a new class of devices based on the chemical doping of graphene by boron or nitrogen .
@xcite using first - principles and mesoscopic quantum transport calculations ( within the landauer - bttiker formalism ) some indications of different behaviors for electrons and holes were predicted in ribbons with lateral sizes as large as 10 nm , with mobility gaps in the order of 1 ev .
@xcite here the mobility gaps are typically defined by the corresponding energy window for which the conductance is smaller than a typical value of @xmath2 ( @xmath3 ( see ref .
@xcite for details ) .
recent experimental findings confirmed the possibility to include boron and nitrogen impurities in graphene , @xcite therefore suggesting doped graphene transistors as a genuine alternative to ultranarrow gnrs with a more controlled tuning of the conductance features .
however , fully self - consistent simulations of doped graphene transistors are mandatory to ascertain the real impact of graphene doping .
we explore the effect , at room temperature , of boron doping in graphene - based field effect transistors using the device geometry illustrated in fig .
[ fig1 ] .
self - consistent negf calculations are performed , allowing for a proper treatment of the screened impurity potential and a more robust assessment of the doping effects on the electron - hole transport asymmetry and on the onset of mobility gaps . |
the lagrange - mesh method @xcite is a simple and often accurate method for solving quantum - mechanical problems .
it can be seen as an approximate variational calculation using a basis of lagrange functions associated with a mesh and the gauss quadrature associated with this mesh to evaluate most of the matrix elements . for reasons
not perfectly understood @xcite , the lagrange - mesh method is in many applications as accurate as the corresponding variational approach .
however , the success of the lagrange - mesh method relies ultimately on the validity of the gauss quadrature .
therefore , when the quality of the gauss - quadrature approximation is poor , the lagrange - mesh method is inaccurate . in particular
, it happens when the potential is singular but surprisingly , only for the lowest partial waves @xcite .
for some problems , the high accuracy of the lagrange - mesh method can be restored by means of a regularization technique @xcite . for other problems ,
such a regularization technique is not available , yet .
it is the case for instance for the study of a coulomb system with more than three particles . in the case of three particles ,
the presence of singularities in the potential restricts the high quality of the lagrange - mesh method to some specific systems of coordinates @xcite . a better understanding of the regularization process , its effects and its necessity , could help to develop new regularization techniques and thus , broaden the scope of applicability of the lagrange - mesh method . in this paper , a regularized lagrange mesh based on orthonormal lagrange - laguerre functions is developed and applied to bound - state calculations .
a comparison with other lagrange meshes provides interesting insight into the regularization process .
the lagrange - mesh method is also extended to the phase - shift calculations by evaluating integral relations derived from the kohn variational principle and adapted from @xcite .
the integral relations are computed with the gauss quadrature associated with the lagrange mesh , which make their evaluation particularly easy .
the accuracy of the approach is analyzed . in sec .
[ s2 ] , the application of the lagrange - mesh method to the study of bound states and to the calculation of phase shifts is outlined .
the lagrange meshes considered in this work are defined in sec .
formulae associated with these meshes and newly derived are given . in sec .
[ s4 ] , the lagrange - mesh method is applied to different test cases and the obtained results are discussed .
concluding remarks are given in sec .
finally , in the appendix , the lagrange - mesh method is applied to a somehow exotic but illuminating problem : the study of bound states in a two dimensional space .
for the partial wave of orbital angular momentum @xmath0 , the radial schrdinger equation for a particle with mass @xmath1 in a central potential @xmath2 reads @xmath3u_{k\ell}(r)=e u_{k\ell}(r)\ ] ] with the boundary condition @xmath4 where @xmath5 is the energy of the particle , @xmath6 is the wave number , @xmath7 is the radius in spherical coordinates , and @xmath8 .
the potential @xmath2 is assumed to have singularities , if any , only at the origin ( @xmath9 ) or at infinity ( @xmath10 ) .
moreover , at the origin , the potential @xmath2 is assumed to be less singular than the centrifugal term , @xmath11 with these hypotheses , the radial wave function behaves at the origin as @xmath12 let us first consider a purely discrete spectrum .
equation ( [ sch ] ) can be solved approximately by expanding the radial wave function @xmath13 in some set of square - integrable functions @xmath14 , @xmath15 then , the energies and the coefficients @xmath16 are determined variationally by solving the generalized eigenvalue problem @xmath17 for @xmath18 .
the eigenvalue problem is not generalized if the norm matrix is the identity , i.e. , if the functions @xmath19 are orthonormal .
now , let us consider that the potential @xmath20 tends asymptotically to the coulomb potential @xmath21 , @xmath22 \mathop{\longrightarrow}\limits_{r\to\infty } 0,\ ] ] where the coulomb potential is given by @xmath23 with parameter @xmath24 .
this potential has a possible discrete spectrum of negative energies and a continuum spectrum of positive energies .
for any strictly positive energy , the schrdinger equation ( [ sch ] ) has a non - square integrable solution with an asymptotic behavior given by @xmath25,\ ] ] where @xmath26 , @xmath27 is a normalization coefficient , @xmath28 is the phase shift corresponding to partial wave @xmath0 and energy @xmath5 , and @xmath29 and @xmath30 are the regular and irregular coulomb functions . for such a potential ,
the solutions of ( [ eigen ] ) with negative energies are again variational approximations of the bound - state wave functions while the solutions of ( [ eigen ] ) with positive energies correspond to pseudostates , which are square - integrable approximations of continuum states .
although these pseudostates do not have the proper asymptotic behavior ( [ asymp ] ) of a positive - energy wave function , they can be used to determine the phase shift via integral relations derived from the kohn variational principle @xcite , @xmath31 u_{k\ell}(r ) f_\ell(\eta , k r ) \rmd r } { \int^\infty_0 [ v(r)-v_c(r ) ] u_{k\ell}(r ) \tilde{g}_\ell(\eta , k r ) \rmd r+i_\gamma},\ ] ] where @xmath32 with @xmath33 is a regularization of the irregular coulomb function @xmath30 and @xmath34 - 2 ( 1-\rme^{-\gamma r})\frac{\rmd}{\rmd r}\right\}g_\ell(\eta , k r ) \rmd r.\end{aligned}\]][ig ] indeed , the integrals in ( [ kvp ] ) and ( [ ig ] ) are short - ranged and therefore , only the internal part of the wave function has to be described accurately to provide accurate phase shifts .
relation ( [ kvp ] ) is a straightforward generalization for an arbitrary angular momentum of relations given in @xcite . in theory , this relation is valid for any positive value of @xmath35 , including @xmath36 with @xmath37 .
however , in practice , the radial wave function is not known exactly and some approximation of @xmath38 , in a finite basis for instance , like ( [ uexp ] ) , is used . in this case ,
relation ( [ kvp ] ) is only accurate for some range of @xmath35 values .
several @xmath35 values need to be tested to find a plateau where the sensitivity of the results to the @xmath35 parameter is weak .
this sensitivity study gives a valuable information about the numerical accuracy of the phase shift which is reached for a particular calculation . in this work
, the radial wave function @xmath13 is expanded in a lagrange basis @xcite .
such a basis is associated with a set of @xmath39 mesh points @xmath40 and the gauss quadrature associated with this mesh .
each lagrange function vanishes at all mesh points except one , @xmath41 where the constants @xmath42 are the weights of the gauss quadrature associated with the mesh , @xmath43 the abscissae @xmath44 and the weights @xmath45 of the gauss quadrature are defined implicitly by the relations @xmath46 which are to be true for any polynomial @xmath47 of degree lower than or equal to @xmath48 .
the weight function @xmath49 defines the type of gauss quadrature which is used .
the choices of @xmath49 considered in this work are specified in the next section .
direct consequences of property ( [ lagcond ] ) are that , at the gauss approximation , the lagrange functions are orthonormal , @xmath50 and the potential matrix is diagonal , @xmath51 where @xmath52 is the kronecker delta and the subscript @xmath53 indicates that the gauss quadrature is used . calculating the overlap and potential matrix elements by applying the gauss quadrature makes their evaluation particularly simple and , in many cases , does not reduce significantly the accuracy of the method . however , in presence of _ singularities _ , the application of the gauss quadrature can reduce the accuracy of the results by several orders of magnitude @xcite . in this context
, the term _ singularity _ means that in the matrix element , the integrand divided by the weight function @xmath49 associated with the gauss quadrature is singular .
for instance , the centrifugal term ( @xmath54 ) , @xmath55 is said _ singular _ if the factor @xmath56 $ ] is singular .
the presence or absence of _ singularities _ depends on the operator and also on the considered gauss quadrature or equivalently on the considered lagrange - mesh .
studying in which cases the presence of a _ singularity _ reduces the accuracy of the lagrange - mesh method is one of the aims of this paper .
three lagrange meshes are considered . all
are based on a gauss - laguerre quadrature , i.e. , a gauss quadrature based on a weight function @xmath57 with @xmath58 .
each mesh is made of @xmath39 lagrange functions defined over @xmath59 : the lagrange - laguerre functions @xcite , @xmath60 the lagrange - laguerre functions regularized by @xmath61 , @xmath62 and the lagrange - laguerre functions regularized by @xmath7 @xcite , @xmath63 with @xmath64 .
the mesh points @xmath65 are the zeros of the generalized laguerre polynomial @xmath66 . the normalization coefficients @xmath67 are given by @xmath68 all lagrange functions are zero at each mesh point but one , @xmath69 where the @xmath42 are the gauss - laguerre weights .
they are defined such that the gauss quadrature @xmath70 is exact if @xmath71 is any polynomial of degree lower than or equal to @xmath48 times @xmath72 .
while not explicitly denoted , the gauss weights @xmath42 depend on @xmath39 and @xmath73 .
note that the set of basis functions @xmath74 , @xmath75 , and @xmath76 define the same vector space .
however , when the gauss quadrature is used , the results obtained with these three meshes differ , sometimes significantly .
while lagrange meshes based on @xmath74 and @xmath77 lagrange functions have widely been used for solving radial schrdinger equations @xcite , the interest of lagrange mesh regularized by @xmath61 , @xmath78 , has been ignored .
therefore , i mainly focus on this mesh in the rest of this section .
the exactness of the gauss quadrature for the overlap matrix elements of the @xmath79 functions combined with relation ( [ fjxi ] ) shows that @xmath78 is a set of orthonormal functions , @xmath80 like the set of non - regularized functions @xmath81 but contrary to the set of regularized functions @xmath82 .
the derivatives of the regularized functions @xmath83 evaluated at the mesh points can be deduced from the derivatives of the non - regularized function @xmath74 at the same mesh points , which are given in @xcite .
the first derivative reads , for @xmath84 , @xmath85 and is zero for @xmath86 , @xmath87 the second derivative reads , for @xmath84 , @xmath88 and , for @xmath86 , @xmath89.\ ] ] the matrix elements of @xmath90 and @xmath91 between the lagrange functions are exact at the gauss quadrature because the integrands are a polynomial of degree lower than or equal to @xmath48 times @xmath72 .
the matrix elements of @xmath90 are thus given by @xmath92 and the matrix elements of @xmath91 by @xmath93 for @xmath84 and @xmath94 for @xmath86 .
the matrix element of @xmath91 is not used in this work but is given for sake of completeness .
the matrix elements of @xmath95 , @xmath7 , and @xmath96 between the lagrange functions are not exact at the gauss quadrature but still have compact expressions , @xmath97
@xmath98 and @xmath99 in the matrix element of @xmath95 , @xmath100 is assumed to avoid a divergence . in equations ( [ 1x2])-([x2 ] ) , the first term of the r.h.s
. corresponds to the matrix element calculated at the gauss approximation .
equations ( [ 1x2 ] ) and ( [ 1x ] ) can be deduced from the matrix elements of @xmath90 and @xmath96 between non - regularized functions @xmath74 , given in @xcite .
relation ( [ x2 ] ) can be proven , by following an approach similar to the one proposed in appendix 1 of @xcite , based on the decomposition @xmath101 the integrals involving the last two terms are calculated from ( [ 1 ] ) and ( [ 1x ] ) .
the integral including the first term is easily calculated when the lagrange functions @xmath102 and @xmath79 are expressed from ( [ defnreg ] ) and ( [ defregsq ] ) as functions of the generalized laguerre polynomials @xmath66 .
the matrix element of @xmath95 involves an integrand which after division by the weight function @xmath72 is singular . in spite of this fact
, it is shown in the next section that using the gauss quadrature for evaluating this matrix element still leads to accurate results .
the matrix elements of @xmath103 between the lagrange functions are not exact at the gauss approximation except for @xmath104 .
they are given , for @xmath84 , by @xmath105\ ] ] and , for @xmath86 , by @xmath106,\ ] ] where @xmath100 is assumed to avoid a divergence .
the last term in the r.h.s of ( [ dx2 ] ) and ( [ last2 ] ) can be seen as a correcting term added to the gauss approximation .
this term is identically zero for @xmath104 .
relations ( [ dx2 ] ) and ( [ last2 ] ) can be proved by following an approach similar to the one proposed in appendix 1 of @xcite .
relations ( [ first])-([x2 ] ) , ( [ dx2 ] ) , and ( [ last2 ] ) have been verified numerically .
the corresponding formulae for the non - regularized and @xmath7-regularized meshes can be found in @xcite . for the sake of clarity , the unscaled versions of the lagrange
meshes have been presented in this section .
however , to reach high accuracy with small numbers of mesh points , it is required to adapt the mesh to the size of the considered physical problem by an appropriate scaling , i.e. , by considering the scaled lagrange functions @xmath107 instead of the lagrange functions @xmath108 and similarly for the other meshes .
parameter @xmath109 is called the scaling factor .
first , the non - regularized lagrange mesh ( [ defnreg ] ) with @xmath110 , the @xmath61-regularized lagrange mesh ( [ defregsq ] ) with @xmath104 , and the @xmath7-regularized lagrange mesh ( [ defreg ] ) with @xmath111 are applied to the study of bound states . since these lagrange meshes correspond to the same vector space of basis functions , they are ideal for testing the impact of the gauss quadrature on the accuracy of the different meshes . at the origin ,
all considered lagrange functions behave like @xmath7 .
they are thus able to reproduce the near - origin behavior ( [ ori ] ) of the radial wave function for any partial wave , provided @xmath112 .
two solvable potentials are considered , namely the harmonic - oscillator potential , @xmath113 , and the attractive coulomb potential , @xmath114 .
for each one , the partial waves @xmath115 , @xmath116 , and @xmath117 are studied . to study the accuracy of the lagrange - mesh calculations in presence of singularities ( centrifugal term and/or coulomb ) ,
several calculations are performed : a variational calculation ( referred as `` var '' in the tables ) based on the @xmath61-regularized lagrange mesh , for which all matrix elements are calculated exactly by the formulae given in the previous section ; two mesh calculations based on the regularized lagrange functions @xmath118 and @xmath119 , referred , respectively , as `` reg @xmath61 '' and `` reg @xmath7 '' , for which all terms are calculated at the gauss approximation ; two mesh calculations based on the non - regularized lagrange functions @xmath19 : one ( referred as `` non reg '' ) for which the gauss approximation is made for the potential term and the centrifugal term and the other ( referred as `` non reg @xmath120 '' ) for which the gauss approximation is only made for the potential term .
the relative errors on the lowest - state energies for each considered partial wave ( @xmath115 , @xmath116 , and @xmath117 ) for the harmonic - oscillator potential and for the coulomb potential are given in tables [ tab1 ] and [ tab2 ] , respectively .
they are defined by @xmath121 where @xmath122 and @xmath123 are the approximate and exact bound - state energies . with this definition ,
the variational principle holds if the relative error @xmath124 is positive .
the number of mesh points @xmath39 and the value of the scale parameter @xmath109 are given in the table captions .
the value of the scale parameter is chosen to provide sensible results but is not really optimized since the aim is only to compare the accuracy of different approaches , not to get the most accurate solutions of the schdinger equations , which are analytically known anyway . to evaluate accurately the relative errors ,
all calculations have been performed with quadruple - precision arithmetic .
.relative error @xmath124 on the lowest - state energy of the harmonic - oscillator potential @xmath113 for the @xmath115 , @xmath116 , and @xmath117 waves obtained from several lagrange meshes ( see text for details ) with @xmath125 and @xmath126 . the notation @xmath127 $ ] stands for @xmath128 .
[ cols= " > , > , > , > , > , > " , ] the number of mesh points @xmath39 and the value of the scale parameter @xmath109 are given in the table caption .
the variational calculation is performed with a @xmath129-regularized lagrange basis of @xmath130 functions , with @xmath110 .
therefore , in both the mesh and the variational calculations , the radial wave function is expanded as a polynomial in @xmath131 of degree up to @xmath130 times @xmath129 times an exponential .
the lagrange - mesh method is as accurate as the variational approach while it involves the substitution of divergent integrals by their gauss - quadrature approximation ! following the reasoning made in the analysis of the accuracy of the lagrange - mesh method for the three - dimensional radial schrdinger equation , this striking property could be anticipated .
indeed , the gauss quadrature induces a huge loss of accuracy in presence of a singularity in the quotient @xmath132 , not in the quotient @xmath133/w(\rho)$ ] .
however , the quotient @xmath134}{w(\rho ) } \mathop{\approx}\limits_{\rho\to 0 } \frac{f_i(\rho ) [ \mathcal{h}_\ell \rho^{\ell+1 } \rme^{-\rho/2h}]}{w(\rho)}\ ] ] is regular for all values of @xmath135 .
11 baye d and heenen p - h 1986 204159 | the lagrange - mesh method is an approximate variational approach having the form of a mesh calculation because of the use of a gauss quadrature .
although this method provides accurate results in many problems with small number of mesh points , its accuracy can be strongly reduced by the presence of singularities in the potential term . in this paper , a new regularized lagrange - laguerre mesh , based on _ exactly _
orthonormal lagrange functions , is devised .
it is applied to two solvable radial potentials : the harmonic - oscillator and coulomb potentials . in spite of the singularities of the coulomb and centrifugal potentials ,
accurate bound - state energies are obtained for all partial waves .
the analysis of these results and a comparison with other lagrange - mesh calculations lead to a simple rule to predict in which cases a singularity does induce or not a significant loss of accuracy in lagrange - mesh calculations .
in addition , the lagrange - laguerre - mesh approach is applied to the evaluation of phase shifts via integral relations .
small numbers of mesh points suffice to provide very accurate results .
_ keywords _ : lagrange - mesh method , gauss quadrature , potential model , bound states and continuum |
Subsets and Splits